How Canonical Can Do Ubuntu Right: It Isn't a Technical Problem

By Caitlyn Martin
April 11, 2010 | Comments: 185

I knew in advance that venting my frustrations with Ubuntu in the form of an article yesterday would stir up a hornet's nest. Sometimes when I criticize a Linux distribution there is a tendency by the community around the distro to circle the wagons and to go on the attack. Sometimes, particularly in cases of smaller distributions like VectorLinux and Absolute Linux, the response is very positive: an acknowledgment of the issues involved and a commitment to correct the issues. Developers of distributions in the second category never fail to impress me with their commitment to their product and the community and they deserve kudos for their approach.

So far, with only a very few exceptions, the comments and discussion around my criticism of Ubuntu has been respectful and on topic, even when people strongly disagreed with me. This says something very positive about the Ubuntu community.

Having read all the comments I'd like to clarify my thoughts on the subject. First and foremost, with all my criticism of Ubuntu, I am not questioning the competence or the expertise of the developers at Canonical. Far from it. As I pointed out in the article the folks behind Ubuntu have proven they are capable of delivering a quality product. That isn't at issue.

I feel the problem I am describing is probably rooted in policy or business decisions that have been made which, in turn, tie the hands of the developers. I also don't think there is any one answer or any one solution that is the right one. There are a number of ways that the issues I raise can be tackled. One would be a simple marketing change: tout LTS as the stable "Linux for human beings" and the six month releases as cutting edge. In other words, Canonical can change the expectations of the user community without substantially changing the product and then meet those expectations. This is essentially the Fedora/Red Hat approach and that company is certainly doing well.

The other approach, of course, is to work on better QA for the six month releases. That is a more costly fix because it requires either more developers working on bugs or allowing release dates to slip.

I do appreciate that Canonical has chosen to dedicate a lot of resources to the consumer desktop market when most other Linux companies have not. I do feel that Canonical, through Ubuntu, has a unique capability to speed Linux desktop adoption if they do a good job with the distribution. That, of course, is where my complaint lies. Other distributions which target the desktop and the wider consumer market do a much better job from a technical standpoint. They produce a better product. What they don't have is the mindshare or marketing capability that Ubuntu has. I really believe that the reason many people think the names Ubuntu and Linux are interchangeable is almost solely due to marketing.

"It works for me" is never a valid argument. Imagine for a minute if I, as a professional Linux/UNIX consultant, told a customer who is having a problem that "it works for me" and went no farther. I'd lose all my customers in short order and I'd be out of work. I have no doubt there are people for whom Ubuntu consistently works well. That isn't at issue. There are simply too many people for whom it doesn't work well, particularly when compared to some other distros focused on the desktop.

In fairness I really do need to focus on one area where Canonical and Ubuntu have done an outstanding job in reaching new users outside the Linux community: preloaded systems. Working with vendors like Dell, Hewlett-Packard and a host of smaller companies has allowed system preloaded with Linux that do "just work" to become real options for consumers. Yes, small Linux boutique vendors like zareason and System 76 have been doing this for years and doing it well. Having said that, getting major brand name companies on board undoubtedly is a great help to Linux adoption and Canonical has done a better job of that than any other company out there.

Even considering all of that I still feel that the downloaded Ubuntu offerings more often than not have been substandard when compared to other distributions. I am not going to produce a list of serious bugs over the past five years. What would be the point? The real focus now should be what happens in the future, not what has gone wrong in the past.

I don't want Canonical to fail. Quite the contrary, I want them to succeed and deliver a product the Linux community can feel good about. As about half of the comments to yesterday's article so far demonstrate I am hardly alone in feeling that Ubuntu fails in this respect. Now, at least, I can humbly say I've offered a few potential solutions. Until the day comes when we can point to Ubuntu as a model Linux distribution I still believe it is incumbent on the wider Linux community to decouple the name Linux from Ubuntu.

Yes, as some have argued, having a myriad of choices out there can be confusing. On the other hand, are you intimidated by the breakfast cereal aisle at the supermarket or do you like the variety you have to choose from? We need to make the choices in Linux as appealing as the choices in breakfast cereal, where you can find something suited to any taste. Limiting choice, as Microsoft does with Windows or Apple does with MacOS, is not a good idea at all. Choice is one of the things that distinguishes Linux from its competitors in a positive way.

I am not one of those who subscribe to the idea that Linux is for geeks or computer intelligentsia. At the moment the Linux desktop supports more hardware out of the box than Windows 7 ever will and is more than ready for the general public. It can be very easy to use. No operating system, especially not Windows, is necessarily easy to install if you aren't at least a little bit computer savvy. Ubuntu is far better than Windows in this area. My point, in part, is that other distributions are far better than Ubuntu in precisely the same area.

I also don't subscribe to the notion that writing critical articles about a Linux distribution or an aspect of Linux is harmful. I think some honest examination and airing of issues is a positive thing if it gets people thinking about how to make Linux more appealing and how to solve problems.

At this point I recommend Mandriva 2010 for newcomers to Linux. No, it is not bug free. No distribution is. Mandriva's developers are simply more responsive to bug reports and get issues fixed, usually in a timely manner. In addition, while Mandriva has had a few less than stellar releases they have, more often than not, done a pretty good job of getting things out that work. As always, your mileage may vary.

Others recommend distributions like PCLinuxOS, Simply Mepis or Linux Mint (an Ubuntu derivative). While I have no specific complaints about any of these I simply, perhaps because of my background in the business world, prefer a distribution with a company behind it rather than an individual developer or a small team of volunteers. Pardus, which is backed by the Turkish government, will be a serious contender once they have a 64-bit release. They don't as of yet.

Would I like to recommend Ubuntu in the future? You bet! Without some serious rethinking at Canonical I don't see that happening any time soon, unfortunately.


You might also be interested in:

185 Comments

I completely agree with you and this was the reason why I switched from Ubuntu to Archlinux. People should await some troubles with their distribution is a rolling release distribution like Archlinux, not with "stable" Ubuntu.

After 1,5 year using Arch and Ubuntu, I am surprised by the fact, that the Arch was always more stable and better working than Ubuntu.

You had written all I think for a long time and I am glad, that some professional agrees with me, good work

Canonical is excellent at one thing & one thing only.
That is marketing. That drowns many more people to Linux and FOSS & that is the reason i dont have strong(er) feelings about their way of doing business.
Fact is that the company has a very aggressive policy towards its users. To me it seems that they are not interested in keeping them satisfied at the very least. Instead they just care about adopting new users. They prrove that by connstantly changing everything even at the end of the LTS release cycle. Just look at the forthcoming 10.04 release and the relation it has to 9.10. The only relation i can find is that its still almost entirely written in Python. :)

Ubuntu seems like a transitional filter from the Windows/Mac OS world to the FOSS world. Its bond to Linux is getting looser each day that goes by. Besides the use of the Linux kernel, the interface behaves entirely different than any other Linux distribution out there.
eg. the mounting of optical media/usb sticks.
That proves to me that Ubuntu doesnt want to have Linux users either.
Their target group is Windows users. Users who are accustomed to using products they play no role in. That gives the company control over them. Ubuntu is bait. Bait for ignorant users who heard about it and decided to give it a try. Truth is i wouldnt recommend it to anyone, let alone to people who rely on their machines for their work. I dont believe there isnt a single OS i cant think of that does the job better.
FWIF if i were making money over FOSS like Red Hat and Novell do, i'd be thrilled there is a finally another distribution after Fedora, one backed by a company that is willing to spend money on introducing more & more people to FOSS.
With the way that Canonical works, the others will gain more than them in the long run.
Hopefully us users, will benefit from that hype and all, by better compatibility with closed source applications eg. MS Office, and more support & software availability from companies writing software for MS Windows.

The goal for ubuntu is to release an OS better than Mac OS X

If Canonical-Ubuntu want to make a better OS than Mac OS-X, then it has to be better than a case of GUI mods & Apps. Surely, what makes an OS 'better' is the actual underlying OS?

As much as this, documentation that genuinely informs users how to operate the system without making silly mistakes is not a bad idea either. It's not degree-level stuff: I'm not a programmer, but even I can figure out to tailor a Linux OS to my requirements, & I know enough to not let viruses in (I hope so, at least).

Sure, no OS is UTTERLY secure, but a little bit of useful information goes a long way. A manual (that would take up a few KB of space on an ISO) included in a release would be a good idea. Even electric kettles come with manuals!

One good manual may well be worth more than a million blogs on forums. Besides which, if general end-users/consumers are prepared to pay hundreds or thousands of bucks for an M$ product, why wouldn't they be prepared to donate a little bit for a far superior & inherently secure system such as those offered by the 'Linux community'.

There's no need for a 'single' standard bearer, but a 'bottom line standard' would be a bloody good idea.

Considering that 'Linux' in general, including Ubuntu, is second only to OpenBSD in terms of security, & OS X is no.10 on this list (http://www.secpoint.com/Top-10-Most-Secure-Operating-Systems.html), your statement makes no sense.

Ubuntu 10.04 'Lucid Lynx' does look a little Mac-esque!

Well, I don't know what to answer. You say saying "it works for me" is not an answer, but all you write here is based just on your experience and triggered by the printer problem. How can you say that Ubuntu is not such a good product as others? Your experience, someone's else experience? How can we measure distribution quality? There are 12 million Ubuntu users out there. Ubuntu has been gaining more and more users since it was founded. Do you think it would happen if Canonical constantly provided a product with lesser quality than its competition? I'm a Linux publicist and one of my jobs is trying different distributions. I try tens of distributions every year. I'm maily focused on Mandriva Linux and Ubuntu. I have been using ML for 3 years and Ubuntu for 4 years and although both distributions are good, in my opinion Ubuntu is superior to ML. It provides the best combination of reliability and cutting-edge software. It's clear that Debian Stable provides better reliability, but who wants to use such old software (the same with RHEL)? And Debian Unstable? It's not more reliable than Ubuntu. Actually, I have much worse experience with Debian Unstable than just a printer problem.

Actually, no, my article is not based just on my personal experience. Please take a look at some of the most significant bugs in Launchpad and how long they go without any activity at all. The "printer problem", as you call it, is almost certainly a devicekit or udev problem. You'll note that lsusb indicated that a USB device was detected but not correctly identified. While the Launchpad bug report I linked in the first article referred to just one HP model of printer that is hardly the only thing affected. Here is another bug, another HP model: https://bugs.launchpad.net/ubuntu/+source/cups/+bug/489750 Here it is again with a Minolta printer: https://bugs.launchpad.net/ubuntu/+source/cups/+bug/487546 I suspect that if I dig through bug reports I'll find more similarly affected devices and, I suspect, more than just printers.

If you go through the Intel video driver issues from Ubuntu 9.04 (Jaunty Jackalope) in Launchpad which rendered that release unusable you will see that the developers made a conscious decision not to fix in that release. What percentage of current and recent model PCs, especially laptops and netbooks, use Intel video chipsets? That's a huge number of impacted users and no fix was forthcoming in a release cycle. Mandriva made the same mistake (including Intel drivers which simply weren't ready) in 2009.1. The difference between Mandriva and Ubuntu is that Mandriva fixed the problem. Ubuntu did not.

f you go through the intel_sda audio chipset issues from Ubuntu 8.10 (Intrepid Ibex) in Launchpad you will once again see that developers knew what the solution was and decided not to implement it until the next release. The solution offered in the Ubuntu forums was to download a patch to ALSA and to compile it from source. Sure, that worked for me, but do you really think that's acceptable in a distro that touts itself as being ready for the masses?

Sorry, but everything I have reported is documented by Canonical themselves and is part of the public record.

You ask if I think Ubuntu can build a user base with an inferior product? Yes, with superior marketing, absolutely. I'm old enough to remember VHS vs. Betamax and the struggle for dominance in the VCR market. Beta was, by all accounts, a superior product but VHS captured the market. Windows rise to dominance is another example. Ubuntu is just one of many.

Debian Unstable lives up to it's name. It delivers what it promises. Can you say the same about Ubuntu given what I've just documented for you?

One of our products is based on Ubuntu, so Launchpad is my second home. I know the situation is not ideal, but it's getting better! I remember Ubuntu bugzilla two years ago. I reported a bug and it stayed there without any attention for a year. I have been using Ubuntu 10.04 for a month, reported several bugs and all of them have been confirmed and most of them fixed. Don't forget that Ubuntu is the most popular distribution and it receives far more bug reports than any other distribution.
Mandriva did NOT solve the problem. Mandriva Linux 2009.1 suffered from it as well as Ubuntu. The only difference was that Mandriva used a newer kernel (2.6.29) and a newer version of the driver which had very bad performance, too, but a bit better than the version Ubuntu used. Ubuntu developers decided to use 2.6.28 because surprisingly Ubuntu is more conservative than Mandriva or Fedora. The problem was far more complex than just replacing the older version of driver by a newer one. Why? Because the new version of Intel driver uses functions of DKMS which were finished mainly in 2.6.30. And do you know what a last-minute change of the entire kernel means? It could break many other things. BTW there were repositories with new kernels and drivers provided by Ubuntu developers. I used them and the performance was much better, but for example hooking up a projector didn't work. Ubuntu development has a very clear rule: no newer versions after the repository freeze. This generally leads to better stability even if it doesn't look like at first glance in some situations.
Your comparison to the VHS vs. Betamax case is not accurate. That was single-moment decision. Once VHS had been chosen there was no way to change it. If I choose Ubuntu I can always switch to another distribution. Switching costs are actually very low for a home user. Hence Ubuntu is in very competitive environment and marketing would not be enough in a long-term prospect.
Ubuntu has problems, but it is not worse than other distributions. You pointed out Mandriva. My colleague has 10-year, very deep experience with Mandriva and he can tell you a lot of bummers made by Mandriva developers.

First, that Ubuntu is the "most popular distribution" is debatable if not outright false. Fedora claims 20 million users to Ubuntu's claim of 12 million users. Are those numbers accurate? Possibly not, as I don't think there is any one methodology for counting Linux users that works. Having said that I don't think we know which distro is most popular on the desktop. We do know that there are two clear leaders and that's all.

Where Linux is strongest is in the server room and in that arena it's clear that Red Hat dominates. On the corporate (not consumer) desktop I think you will find that Red Hat also is leading simply because companies that choose Linux on the desktop prefer to have one Linux vendor to deal with rather than two. Novell/SUSE is number two and Ubuntu, at least in the server room, is a very distant third and not really a major player as of yet.

Your claim that Mandriva did not solve the problem is false. They actually offered two fixes: a newer kernel and driver set (you needed both) or an older, working driver. Offering the older driver is how Slackware 13 avoided the problem entirely. Fedora avoided the problem by not pushing out code until it was ready. Ubuntu failed in two key areas: they pushed the code out when it wasn't ready just as Mandriva did, and then they failed right up until today to deliver a fix at all in Jaunty. I'll point out that 9.04 (Jaunty) is still under support. If not fixing a major regression is "support" then it is incredibly poor support indeed.

I am very aware of the interrelationship between the kernel, X.org code and the Intel drivers. I will point out that Red Hat has precisely the same rule as Ubuntu regarding no new versions but they backport newer hardware support into their kernel to get around the issues Ubuntu faced. In addition, unlike Ubuntu, neither Red Hat nor Fedora have locked in, fixed release schedules. If the code is severely broken they simply will not release. Ubuntu will with disastrous results as we have repeatedly seen. Once again, I'll ask you to read the Linux Magazine article on Karmic Koala: http://www.linux-mag.com/id/7600

VHS vs. Beta most certainly does apply, as does Windows vs. OS/2 or any other instance where superior marketing won out and superior technology lost. Just because you don't like a specific example doesn't change the fact that marketing often trumps technology.

Mandiva developers made mistakes? No doubt about it. So do developers for ever major distribution. The question is quantity of mistakes, severity of mistakes, and how the company/distribution fixes those mistakes. Ubuntu has a very poor track record compared to other distributions in all three areas.

Whoa!

Fedora does NOT claim 20 million users. Please, please read over the counting methodology we use to establish our Unique IP address counts. To claim the 20 million+ active userbase is an over reach based on what we are doing.

We publish our methodology specifically to allow our conclusions to be vetted and to provide a starting point for further discussion on how we can all do a better job of getting accurate userbase numbers so we can make a really good stab at getting a strong estimate for linux market penetration.

We aren't just throwing numbers over the wall like Canonical executives are. We are telling you exactly what and how we are counting. Please take the time to try to understand what that means and to ask questions of us if the methodology doesn't makes sense.

And on that note, please make sure you ask Canonical to provide their counting methodology. Counting is hard, if we are talking about the method we are using, then how can we consider the resulting numbers as reliable?

-jef

That Linux Magazine "review" of Ubuntu by Christopher Smart that you cite was a hatchet job. As Ars Techicha put it [1], in reference to that review:

"It's also worth noting that Ubuntu is not unique in issuing releases with known and documented bugs. In fact, most Linux distributions—and most software products in general—will include an overview of errata in their documentation or release notes. It's beyond disingenuous to suggest that this is symptomatic of a failed release."

Full of canards like, "Free software is supposed to improve with each new release. Take OS X..." (Who knew that OS X was free software!), Mr. Smart wrote a hit piece that is, well, beyond disingenuous.

Your points are well-taken, but that particular "review" is not credible.

[1] http://arstechnica.com/open-source/reviews/2009/11/good-karma-ars-reviews-ubuntu-910.ars/11

Fedora claims 24 million MACHINES. Ubuntu claims 12 USERS. This is apples and oranges.

OK, and that has precisely what impact on my criticism of Ubuntu? I said clearly in my original article that I didn't know which was more popular, only that these two distributions are the clear leaders.

I have to agree with Franz' comment above. While I don't deny that there are bugs in the Ubuntu releases, as you say this is the case with any distribution. And while "it works for me" is not a sufficient reply to someone when they have a complaint, "it doesn't work for me" is certainly not license to label the entire distribution as broken or shoddy.

There are certainly a lot of complaints about Ubuntu, sometimes "high profile" (Steve Rosenberg's columns come to mind), but there are also a lot of *users* of Ubuntu. Comparing number of complaints, and even number of bugs isn't necessarily the best approach to comparing the quality of distributions. Some sort of normalized curve of "complaints per user" would be a good starting point (of course, now we have the problem of accurately measuring the number of users!).

Specifically regarding hardware issues, this is also what LiveCDs are *for*. A user can try out Ubuntu on their hardware to see if it is a good fit. If so, I believe you should have no reservations about recommending Ubuntu, and certainly "un-recommend" it if the LiveCD doesn't work out (this is what I do with my friends who are interested in Linux).

Personally, for me Ubuntu has been phenomenal (I came from Debian unstable). While there are certainly some quirks each release, they are very quick to work around, and I have no doubt that different versions of those quirks exist in all distros.

Is there an Ubuntu LoCo in your town/city? The kind of stuff you are lamenting about is solved by having a vibrant local community that can do community based troubleshooting and even QA. The magic of the Ubuntu distribution? Friends and neighbours!

Cheers,
Randall
Ubuntu Vancouver Buzz Generator

A local community, while an excellent thing to be part of, won't solve fundamental breakage in a distribution or an unwillingness by the developers to fix problems within a release cycle. Also, local communities aren't unique to Ubuntu. Fedora claims a larger user base than Ubuntu and has similar user resources. It's not my first choice for newcomers to Linux and that isn't the target audience but if it's resources you want then Fedora and Red Hat are second to none.

That depends how large your local community is, and what they're doing.

Our local community is tackling QA and user administration. This is not a theory. It's live.

The community is what fixes Ubuntu. No one else is going to do it.

Are their Fedora groups that meet in your city with regularity? How big are they? How often?

Cheers,
Randall.

Randall, I live in the Raleigh, North Carolina area. This is the home of Red Hat so I dare say there is more Red Hat/Fedora than anything else here.

You know what? That's irrelevant. It isn't the job of the community to fix the distro. That responsibility belongs to the developer. Canonical is a well funded company with plenty of engineers and developers on staff.

The other reason it's irrelevant is that Ubuntu doesn't market themselves as a distro for Linux hobbyists or Linux professionals. Those are the people who attend LUGs. Ubuntu is for the masses, for humans of all sorts, or so Canonical keeps telling us. These aren't people who want to fix bugs and tinker or are even vaguely capable of doing so. They want their computer to "just work" and have no interest whatsoever in what it takes to make it work. If that is the target audience for Ubuntu then community LUGs are truly meaningless.

I agree with Caitlyn. My wifi on 3 of the Ubuntu machines at home worked perfectly with 8.04.

8.10 NO!
9.04 NO!
9.10 NO!

QED

So my family is stuck at 8.04, and I'm waiting for 10.04 to see if things will be fixed.

If not, I may have to consider buying newer wifi cards to replace the older DLink ones.

Although Ubuntu is getting better, hardware working in one version and not working in a later version is something Mark Shuttleworth has to seriously consider not to break as a development policy.

It's not that difficult is it?

No, it's not hard - it's practically impossible. I doubt there's ever been a distribution release which didn't have *some* hardware support regression. There's just too much hardware out there. Almost every time you change any line of hardware-related code, you break _something_.

(This, of course, is where Apple has a gigantic advantage.)

Adam, I don't disagree with you at all. The issue here is that one distro has more bugs than most (easily shown through documentation they provide) and that they don't fix things promptly when others, including Fedora, Mandriva and openSUSE (the main competitors to Ubuntu) most certainly do.

So, for 3 releases in a row some hardware does not work and an Ubuntu user still does not know if the kernel module is missing, or it is not configured properly, or the hardware support has been dropped - very unlikely, but still possible.

This can never happen to a Slackware, or Gentoo, or Arch, or Debian user. That guys would know what exactly is going on, if not having the problem fixed.

Ubuntu does not educate its users, it pretends it always works (or will work soon) out of the box. That is a lie. Nothing works or will ever work both always and out of the box, including Windows.

It will not help Ubuntu to fix more bugs since some will remain. It will not help to position 6 month releases as bleeding edge since they are not bleeding edge. In general, Linux will never be like Windows.

I have run the testing Debian version for several years with great success, but when Squeeze had trouble with Intel graphics I switched to Ubuntu (started on 9.04 then to 9.1) and I find the "close but no cigar" level of Ubuntu stability quite frustrating. I can't wait tp get back to Debian when Squeeze is released.

Others recommend distributions like PCLinuxOS, Simply Mepis or Linux Mint (an Ubuntu derivative). While I have no specific complaints about any of these I simply, perhaps because of my background in the business world, prefer a distribution with a company behind it rather than an individual developer or a small team of volunteers.

When you claim to judge things based on performance rather than "ideals" you should be consistent and not fall back to a mistaken faith in business over what can be measured and observed. Large companies still don't contribute much to free software, it's all driven by small teams of programmers, but those small teams are able to devote more resources to problems than even the biggest companies can afford. Ubuntu has always been a little buggier for me than Debian itself, but that's to be expected from their goal of pushing fresh software out to users. Mepis, which is more conservative than Ubuntu is somewhere in between. Mepis, is now based on Debian not Ubuntu. Novell has taken a wrecking ball called Mono to Suse and performance has suffered as developers are channeled into that wasteful project, software for iPhones, Microsoft Windows and other nonsense. If you judge software by rhetoric, you should favor distributions that claim to stand up for your rights and community rather than those who claim to be good at flaying customers by making deals with Microsoft. Big business is not always bad but your favorites are.

I don't understand why so many folks are getting so excited about Caitlyn's articles (this and the preceding one). Look folks, she's unhappy with Ubuntu. She couldn't get her HP printer to work with it, but beyond that she provided little to substantiate her assertions. Let her rant. If she's not going to provide any substance, ignore it. You shouldn't argue with someone like that.

As for me, I can't get my HP printer to work with my OS either, in this case Windows XP. I could give you an exceedingly long list of complaints to substantiate my assertion that XP is rubbish, but I couldn't be bothered. I'm about to upgrade to Linux in reaction. Tried Windows 7 and was stunned at how much worse than Windows XP it is - not going to try to substantiate that either. I came to Caitlyn's article because I thought it might be useful with regard to considering Ubuntu as my new OS. Unfortunately it wasn't, substance-less rants being what they are. I have tried the Ubuntu 9.10 live CD and it looks good and worked well for what little I did with it so far. Why Ubuntu? Well, it does indeed have mind share. Those suggesting this is marketing finesse from Canonical are clearly seeing something I'm not. I've never seen any kind of marketing from Canonical - does it really exist? I just keep running into people who actually use and like it as well as reading online materials by end users who like it. Maybe these are all masochistic idiots, but I'm betting they're not. Mandriva? Hmm, interesting - don't know of anyone using it. Oh, well yes, Caitlyn likes it, but her rants don't provide me with anything like useful information.

Sid, my article is unsubstantiated? Please read my reply to Franz where I do, in fact, substantiate precisely what I am saying about Ubuntu and tell you where to find all the documentation you could ever want. Sorry, this isn't just a rant. If what I wrote for Franz isn't enough I can provide links, detail more issues going back for the last five or six years, and provide all the substantiation and useful information you could ever want.

If you don't want to take my word for it and want another source, how about Chris Smart over at Linux Magazine? Chris also used to write for DistroWatch and was a Linux distro developer as well. Take a look at his report on Karmic Koala (9.10) at: http://www.linux-mag.com/cache/7600/1.html He also provides helpful links back to Ubuntu source information. He also was kind of enough to provide a link back to the Ubuntu forum where a poll showed that only 10% of users had a flawless Karmic install. See: http://ubuntuforums.org/showthread.php?t=1305924 I'm certainly not the only one seeing the problems with Ubuntu and these aren't empty rants devoid of information. I can provide more similar reports if you like.

Mandriva has been around for more than a dozen years. I doubt they would have survived as a company if nobody used their product. Just because your circle of friends and acquaintances doesn't use it doesn't mean the quality isn't better. Perhaps more importantly, Mandriva engineers fix bugs within a release cycle rather than leaving their users hanging.

Dismiss me all you like. You have a right to ignore any facts that disturb you or prove inconvenient. That doesn't change the facts.

In regards to "He also was kind of enough to provide a link back to the Ubuntu forum where a poll showed that only 10% of users had a flawless Karmic install".

That is just bad statistical hype. People posting on a forum that something doesn't work, will be verbal and voting. Those that do work will be happily doing computing stuff and not trawling the forums.

4k response is a tiny drop of the Ubuntu user base, but that means nothing due to the context of the poll.

Statistical hype? I don't think so. Try real world experiences of many users instead.

For the sake of argument, let's discount the poll as you suggest. What part of Chris articles aren't accurate? Which of the problems I or other users have listed aren't real?

You and the other Ubuntu faithful can be in denial all you like but that doesn't change the fact that Ubuntu has had many very well documented problems.

Your response does your argument no justice, especially when you judge my overall stance from a single post.

I was merely pointing out that arguments that are based on forum polls are weak at the best of times. Now create a poll during the release of an OS and ask how well did the install go and what do you expect the response to be?

In my opinion you will have 3 big types of posters.

1. People who really do have a problem. They will most likely be trawling the web for answers to their problems and a poll is a good place to vent.

2. People who will stoically defend the OS and will gloss over any problems.

3. Astro turfers who will do their best to appear like they like the OS, but are trying to discredit the release as much as possible.

I believe that a high percentage of real users won't be on the forums, because real users are busy using their computers to do things that they require their tool to do.

More importantly the above 3 groups will ruin any result of a poll because of their objectives. You always have to consider the human element when dealing with humans!

If you wish to label me then please label me as a right tool for the right job faithful. I make my money from best of breed solutions and I try my best not to be blinkered by fanatically beliefs, especially those fuelled by hype.

"my article is unsubstantiated?" I don't even know what that means. What I said was that your articles provided little to no substantiation of your panning of Ubuntu, as in not much in the way of information I could use to learn from.

I did read your response to Franz. I was already willing to accept you couldn't get your printer to work. The rest was of interest but minimally so since it doesn't sound all that much different from every other OS out there. If it wasn't for Franz, we wouldn't have even that from you. More meat, less vitriol.

Not to worry, I am looking elsewhere for articles with real meat, Chris Smart's included.

"a poll showed that only 10% of users had a flawless Karmic install". That's almost too asinine to dignify with a response, but... Are you serious? Almost by definition the people who install something and then go scurrying off to a forum where they discover such a poll are there because they had trouble of some sort. Do you honestly expect a statistically significant number of people who have a flawless install will have even visited such a forum at all, much less answered the poll? That defies logic, Caitlyn.

Don't get me wrong on the Mandriva thing. I fully intend to check it out too. I'm not dismissing it or saying it's flawed just because I haven't run into anyone using it.

Try filling out your articles next time with information people like me can make use of. I'm not ignoring facts. I'm looking for them. Perhaps you'll provide something substantive next time.

Your response reads at: I don't like your facts so I am going to dismiss them an "unsubstansive". So there!

OK... so, please explain to my readers what was unsubstantial about the Intel video driver issues in Jaunty (9.04) or the intel_sda audio driver and ALSA in Intrepid (8.10). How many users did those two bugs alone affect? Were they ever fixed? You know the answer but don't want to admit that not fixing bugs is simply unacceptable.

Andrew Wyatt was right: it's garbage salad indeed.

The article which u gave madam, http://www.linux-mag.com/cache/7600/1.html , is kinda history now.. Look at this > http://www.linux-mag.com/cache/7740/1.html from the same author, yes the very same author, Chistopher Smart, who tried ubuntu 10.04 and praises it now.. He calls it very close to the Perfect Consumer Operating System..
So ppl, have some patience and wait for 10.04, things will mostly get better as the author says.. I agree, 9.04 and 9.10 had some problems, (9.04 had a problem on a frnd's comp and 9.10 on mine) mainly because the developers were rewriting some important parts of the disto, HAL, Upstart and USC..
KDE 4.0, when it was written from scratch with the new Qt toolkit, had thousands of bugs when it arrived that even the core KDE fans din't want to upgrade.. But look how neat, posh, modern and stable the desktop has turned out since KDE 4.3 and 4.4.. It makes Mac and Win7 jealous.. Ubuntu might follow the same lines.. So best thing is to wait for a month, and after the stable version, ubuntu 10.04 is released, install it and enjoy..
Btw, I tried Kubuntu 10.04 netbook remix (beta) and it sure is awesome.. my gprs modem which never worked (in previous versions of MintKDE, OpenSuse, Kubuntu) with the default network manager worked flawlessly with 10.04 beta (had to use wvdial)..

[facepalm]

The WHOLE POINT is that Ubuntu has a long history of severe problems that they don't fix' History is what this whole article is about and the fact that things don't get better from release to release. 10.04 isn't even finalized our out yet and lots of people have reported problems with the beta. Please, spare the the Canonical marketing spiel.

The point, which you utterly ignore, is that other distributions do better. Much better. For that reason and that reason alone equating the names Linux and Ubuntu is problematic. Everything you've written here basically ignores my articles and is beside the point.

I still fail to understand your motive. You don't like linux equating with Ubuntu - but what is the alternative? Equating it with a bunch of other competing distros? Equating it with one of the other distros you mention? How would either of those advance linux? Having used none I could only speculate that they are not as easy to install as Unbuntu.

Clearly, as some other posters have pointed out, Ubuntu has done a great job getting new people - and that is really the point of the distro IMHO - getting people to switch to it, not to achieve a completely stable, pristine OS. It does a great job of removing the "scare" factor from Linux.

My motive is to promote Linux. I do think equating Linux with a "bunch of other competing distros" is, indeed, better. Choice is one of the great strengths of Linux. Windows gives you no choices.

With all due respect, your speculation that other distros are all harder to install than Ubuntu is absolutely incorrect. You admit you've tried no other distribution. Trust me on this, some others are easier, not harder.

There is no such thing as a "pristine OS" but providing a poor experience drives people off for good. My point is that Ubuntu is not doing a "great job" in some very fundamental areas. The "scare factor" is misinformation, plain and simple. Getting an accurate picture of Linux out there is the best way to remove that. Accepting mediocrity is not.

Bad teeeeeeeease Sid!!! :) Ya, both these articles and comments have no useful information for people, have become some fun article to read.. good timepass :)

It is impossible to make everyone happy all the time!

I will say I have run into bugs in many distributions. I have been a Linux user since around 1998 and have tried about every one of them. I am using Ubuntu 10.04 beta now and it works good. I am sure at some point I will run across something that aggravates me, knock on wood this hasn't been the case yet.

I think the thing we need to remember here is the OS is given away freely! Try that in Windows! But, really... This is Linux and it has made huge strides forward in the last couple years, it truly is getting better no matter the version. Innovation is something that takes time and innovation takes someone/people that are willing to break the mold to make it happen. Ubuntu breaks the mold and sometimes does not do it in the most convenient way possible but they do.

Having used Red Hat, Fedora, and SuSE in the past, I was immediately impressed by two things when I moved to Ubuntu: 1) superior fit and finish, and 2) superior community support. Back in the days when I had a Palm, the Palm support on Fedora and SuSE invariably seemed to be broken, whereas it "just worked" when I moved to Ubuntu. Sure, I have had to work around some issues in Ubuntu, but that's where I have found the ample documentation and enthusiastic community support to be so valuable. I don't feel threatened by Canonical; rather I am nurtured by Ubuntu's warm and vibrant community.

I don't feel threatened by Canonical nor have I had complaints about the Ubuntu community. My complaints are two: a large number of bugs and a failure to fix them. I did not hold up either Red Hat/Fedora or openSUSE as examples of superior desktop Linux distributions. Both companies are clearly focused on the server room and, perhaps to a lesser degree, the corporate desktop, not the consumer desktop.

Fit and finish is a matter of personal taste. Have you looked at recent Mandriva releases? PCLinuxOS? Pardus? I don't think Ubuntu is in any way special in this area nor does it particularly fall down in this area either. My complaints with Ubuntu mainly are in the area of hardware support and functionality, not whether it's pretty enough.

Also, you talk about "back in the days". Back whenever lots of things which work now in Linux in general didn't work. There has been progress in every distribution, Ubuntu included. Let's stick with comparing apples and apples, other distros today vs. Ubuntu today. Try that and then come back and let me know how Ubuntu stacks up.

Both articles about *buntu were spot on and mirror my experiences, I think it has been said that Ubuntu should take a step or a half step back from the edge in their releases. That things are broken is normal, but having major item like printing and Xorg or to many items broken, is counter productive for everyone involved.

""It works for me" is never a valid argument"

But it's an argument you must take under consideration. It means that you problemes ain't the same that "Ubuntu sucks".

Three months ago my older graphic card make "kaputt" so i have to buy a new one. Shopkeeper accepted that i can bring the ATI Radeon 4650 back if i'm not lucky. And indeed - kernel 31 didn't give support to that card. So i have to change it to Nvidia GeForce 9500 GT. Kernel 32 support that ATI-graphic card.

I tried Linux first time in 1999 and i must say that hardware support has come much, much better since those days. And if i was a geek i would have managed to install that 32-kernel and use that graphic card. Besides hardware can be even more horrible issue in Windows world.

So Linux and Ubuntu sucks? Not at all. We all know that this graphic card issue is proprietary vs. open source question. And that's just one example why Microsoft wants that nothing will change. But the times have changes since 1999. Have you noticed that huge majority of new Windows 7-computers have ATI Radeon graphic card? Is it coincidence? Fighting against Linux?

I have never, ever said "Ubuntu sucks". I have never said that the distro is without value. Indeed, I pointed out some things that Ubuntu does well in this article. I certainly don't believe that "Linux sucks". Quite the contrary. My entire point is that other Linux distributions do a very good job in areas where Ubuntu does not.

Gee, some people don't read very well, do they?

I agree with Caitlin. I use Ubuntu, but I have been thinking if Debian (stable) wouldn't be better for me. I'm mostly a user, not so much of a tinkerer. I just want it to work. The only things that kept me from doing it so far is that some software versions, like OpenOffice (2.4), are fairly old.

I appreciate the announcement of Canonical that 10.4 will be based on Debian testing iso unstable. That could be the right version for me. Build Ubuntu LTS releases on a more solid foundation and go cutting edge with the in between releases. That would be a good solution. Also goes back to the point of user expectations that was mentioned in the article.

I've used K/Ubuntu since 6.06, it's what got me started using Linux.
I've have never experienced any show stopping problems using it and have always been quite happy.

I am interested that you recommend Mandriva, you feel that KDE gives a more stable distro?

The main reason I've gone back to Ubuntu is the KDE based distros seem to be less stable at the moment.

Not having a pop at KDE, it's always been my favourite.

Dave

Just to note--Mandriva is not a KDE-specific distro. It uses either KDE or Gnome and does both pretty solidly. It also ships with various other desktops. The "Free" version has all those on the download. The "MandrivaOne" live CD version comes in a KDE and a Gnome flavour. I use Mandriva with Gnome, myself, although I'll pop into something else now and again to have a look-see and I use some of the KDE apps.

There is a slight emphasis on KDE over Gnome, but only slight. I believe Mandriva's Control Centre is actually written on GTK+.

So the issue described is not about the specific technology, it's about quality control. As to the RPM thing--eh, it doesn't matter nowadays. Mandriva's urpmi and its graphical installation thingie work about the same as apt and Synaptic. Mandriva's repository is smaller than Debian's, but that's not really a deb vs rpm issue, and it's pretty rare I run into something I want to install that I'd need to go outside the repositories for.

Dave, I agree with you, the KDE based distros have been a little more unstable, but I found the majority of the bugs went with KDE 4.3.
Kubuntu at the end of the month ships KDE 4.4, something I have found to be solid from the last Alpha of Kubuntu 10.04. They've done a magnificant job in the KDE community to get KDE back to the stability we're used to, and I especially like Kubuntu for staying faithful in the most to KDE and QT apps.

I think that Caitlyn started a good discussion. Seen from her perspective I think shes right all the way. I am using linux for a long time on my desktop and server. It is interesting to follow linux development and opensource(both as a hobyist and as a professional).
I am also getting tired of bugs that aren't fixed for a long time, or as they return with the next release.
But, if you are a compagny that wants to make money you have to show the world that you are serious. Canonical is doing many things right. It's the first one who dares to take the glove (desktop for everyone) and has a good compagny behind it. Now that they are in the spotlights, they have to show they are worth it becuase many it-professionals are nowdays reviewing the ubuntu hype.
We all hope that ubuntu wil make it and want to warn about the problems (because there are a few).
Nothing more, nothing less...

I remember at a conference many years ago, being told by this speaker that the one form of advertising that is the most valuable is the one you cannot buy: a word-of-mouth recommendation from a happy customer.

Who cares how much money Mark Shuttleworth can afford to throw into boosting the profile of Ubuntu? At some point, the rubber has to hit the road: real users have to get real work done on the system, and that’s what counts. If Ubuntu is getting popular, it must be because lots of people are succeeding in doing exactly that.

And they’re not just doing it with Ubuntu, they’re doing it with other Linuxes as well. Once you become familiar with the Linux community, you discover that there are loads of other Linuxes out there, with their own pros and cons and their own target markets. And it’s really very easy to change, there is no vendor lock-in.

So ultimately a rising tide floats all boats: as Ubuntu becomes more widely known, Linux as a whole becomes more widely known.

I respect Caitlyn's view of Ubuntu. That was her experience so she can only sincerely write about it, can she not?

That said, Ubuntu has been the only consistently reliable distro EVER on my two desktops and one laptop. The only other distro that can match that reliability is Linux Mint, and guess where it's based on. Laugh all you want, but Ubuntu just works! It has been the most reliable, easily configured, easily administered, distro. It produced the clearest fonts on my desktop LCDs. Debian maybe Ubuntu's superior, but no version of Debian and OpenSuse including the latest have ever detected my monitors' correct resolution. How am I supposed to fix that? With Ubuntu, all versions I've tried get them right the first time.

I have tried most of the other popular distros and have seriously worked to adapt to them, even intending to replace Ubuntu with them in my desktops. But OpenSuse, Mandriva, Fedora all had annoying bugs and behaviors that drive me back to Ubuntu... sort of "better the devil you already know" kind of thing.

You know, in Caitlyn's article, all mention of Ubuntu could be replaced by Fedora (or OpenSuse) and the article would still be correct! This is based on my disappointing experience with OpenSuse and Fedora.

I agree with one thing though. SalixOS is a fine distro. It's what I replaced Ubuntu to run on my laptop which has less abundant resources and power.

Thanks for your article, Caitlyn. If any, it should shake Ubuntu into a more vigilant and productive state.

You can agree with Caitlyn or not. You can say she has facts, or not. Consider her a thermometer. How well is Ubuntu doing? Let's find out by seeing what people are saying about it, formerly happy users who are now unhappy or who have left Ubuntu.

I have only set up 20 or 30 computers with Ubuntu over the last 5 years. I have rolled people forward to get their hardware working, I have rolled people back due to issues. If I decide to stick it out with Ubuntu, it would be lovely to see articles telling me why sitcking it out is so good. If I decide to leave for something else, it would be nice to see articles from people leaving for the same reason and going to where I decided to go. I have to make up my mind for myself.

I can tell you my brother has a Dell Vostro 1520 with an Nvidia card and he likes eyecandy. Does not matter if it is compiz or kwin, Ubuntu for 3 releases has been crap on it. I switched him to Sabayon. Life has been great for him ever since.

My wife has an Aspire One 751h with the GMA500 chipset. At this point it will not run with anything newer than a 2.6.31 kernel. So upgrading is not an option, and as bad as Ubuntu is, it is still known as the best distro to run on that thing for hardware support.

I don't like Ubuntu, it has always been Kubuntu or Xubuntu for me. Both of those distros are treated like bastard children. Ubuntu keeps getting new stuff and they are left in the dust. Mark Shuttleworth seems to have a MAC fetish. Says we need to be as slick as a Mac to go anywhere. In 2 years time he has introduced notifications, and nothing else. So to make progress after 2 years, by fiat they change the buttons placement to match a mac, the color scheme to match a mac, and the wallpaper to match a mac. Let's not forget now using 1000 units instead of 1024 units. On top of that we have dropping Gimp and shoehorning in any Mono app and calling it "best of breed and simpler to use".

I reported a bug, Xubuntu could not autostart DVDs or CDs because Thunar had dvd:/ instead of dvd:// listed. Took over 2 years for them to fix it. They had to add 2 slashes and it took 2 years.

Bottom line for me. I am lazy, I want an easy to use system I don't have to spend time futzing with to keep working. Ubuntu has been that for me for a very long time. Ubuntu Political Issues PLUS Ubuntu not working right problems are making it enough of a hassle that I am testing new Ubuntu versions AND trying out Sabayon and Arch. If Mint moves to debian, they move to that "I gotta try list as well".

#Elder-Geek#

You want an easy to use OS? Have you ever actually tried Linux Mint 8?

After upgrading to Ubuntu 9.10, which I'd understood to be a stable/final/wicked/sorted/top-dog/in-your-face distro, everything went wrong, the forums were a source of confusion, & I went back to M$.

After two months of dissatisfying 'distro-hopping' several 'Linuxes', & a close call with Kubuntu & Xubuntu (which wouldn't recognise audio CDs & which was apparently murdered by Wine's 'black half-screen of death'), I decided to give LM8 a try.

Installation was flawless, everything works, it's a very amenable, user-friendly system, & it comes with a standard whole lot of useful stuff (Java, drivers, installers, etc.) that has to be downloaded in Ubuntu; yet it still fits onto a single 700MB CD, which has something to do with compression of the ISO packages, as well as not wasting valuable ISO space on retro games.

Ubuntu & the other 'Linuxes' could learn a great deal from Linux Mint. It's far more than just '"an Ubuntu derivative"' (as Caitlyn put it) or 'Ubuntu with a makeover' (which is the general message I get from a lot of the reviews), & it's far more than another of Ubuntu's '"bastard children"' (& it's not blue!).

As well as having its own small development team (which can actually be an advantage), it comes with full access to Ubuntu updates & Canonical supported repositories. In other words, it's Ubuntu-based, but much, much better. Regarding upgrades, the Linux Mint (very neat & tidy) website advises against upgrading, if what you have works fine & you're happy with it! Clement Lefebvre & co. have done quite a job.

I have an OS I don't have to fight with or spend half my life dredging forums for often badly explained solutions, an OS on which Firefox doesn't crash all the time (eg Ubuntu & spawn), on which I still have GIMP (unlike Ubuntu), an OS which simply WORKS!

From personal experience, PCLinuxOS (with '357' different desktop options but no guts) & Simply Mepis don't even come close, neither does Pardus (besides which, I'm a little dubious about using an OS backed by a (morally/ethically questionable) government). I didn't go further than "sexier than ever" (BLEH!) Mandriva's messy website. Why pay for something that has far less, when you can get far better for free?! I honestly don't understand why Caitlyn recommends Mandriva as a Linux for beginners.

Personally, I'd recommend linux Mint to M$ refugees, I'd certainly recommend it to complete newcomers, & I definitely recommend it to the friends who first introduced me to Ubuntu. What more can I say? "IT WORKS FOR ME!"

Mandriva is available for free download and always has been. You don't have to spend a dime. If you tried it perhaps you'd know why I recommend it to newcomers :) Don't dismiss something you haven't tried based on a website.

Firstly, many thanks for your excellent hospitality.

Mandriva's website does leave a lot to be desired: & as a website is a first point of contact, it is very important that it makes a good first impression: 'Faster & Sexier Than Ever!' - that is such banal, puerile, & unprofessional marketing! Since when is an OS 'sexy'? This reminds me of the Bad Religion song 'I Love My Computer' - "click me here"... Thanks, but no thanks.

I'm quite happy with LM8 - indeed, my 'distro-hopping' days may even be over, for now, at least, so I don't really need to look at Mandriva. I have an OS on which I can just do stuff, learn a thing or two, & maybe even be able to develop & contribute something to it.

Caitlyn,

Has Mandriva finally fixed their control panel? I'm remembering the network control applet specifically that make me want to beat my computer to death with a hammer, release after release it just wouldn't get fixed.

Finally I just gave up and changed distros, I am writing this on Ubuntu 9.10 and I don't know what to say, it works great.

I guess I should move back to Mandrake since you recommend it and Ubuntu has all these problems you mention.

Has Mandrake improved package management, because that blew also. God it was like rain in the dessert when I used apt-get for the first time. I would chew off my own leg to avoid RPM and dependency hell.

And trying to get software loaded that wasn't in the official pool for a distro, that blows to. Though it is much easier for Debian/RH/Ubuntu as *somone* *somewhere* seems to have packaged it up, specifically for that distro as opposed to "well this is kinda close see if it doesn't bork your system".

There are a lot of aspects that make a good distribution and a good computing experience and frankly Ubuntu has many of them. Do they have their weak spots? Sure, *all* the distros do.

But if Ubuntu was as flawed as you are claiming it certainly would have been toast on my systems by now.

I will probably go back to spinning some other distros just to compare but honestly Ubuntu does mostly "just work".

I will say an interesting point that I notice is that a few distros/users don't really say much bad about other distros and some seem to have vengeance (PCLinuxOS I'm looking at you) toward others, often Ubuntu.

I don't quite understand this.

I don't quite understand the point of your article either. You think all of us ignorant Ubuntu users are just waiting to be enlightened to what a shoddy Distro we are using. You think 12 million people are going to go "duh! thanks Caitlyn I didn't know, let me go hit Mandrakes download servers."

Maybe their is a reason that Ubuntu has become GNU/Linux's "standard bearer", as you call it, on the desk top. Maybe on the whole Ubuntu offers far more positives than negatives for the people that use it.

I know I certainly appreciate a distro/user commuity that spends it's time building itself/themselves up rather than tearing others down.

Maybe you should try it.

Maybe I should try it? I have used Ubuntu extensively for five years. What do you take me for? Don't assume that I or my readers are idiots.

Your "it works for me" argument, as I explained in the article, simply does not fly and is no argument at all. Your "Mandriva blows" argument is contrary to my experience. Your claim of universal joy and happiness by millions of Ubuntu users is arrant nonsense. This whole post comes down to fanboyism at it's worst. Please, if you have facts then back them up with links and supporting documentation.

Caitlyn,

Firstly permit me to say, your style seems a little hard and combative (maybe just Net armour). But I kind of dig that. You don't mind mixin' it up, I respect that as well, as long as it's backed up by facts which you seem keen on. I am keen on them also.

Secondly you have misunderstood a few things in my post.

I wasn't saying Mandriva blows (though one could reasonably infer that at some point I reached that conclusion as I left it), but that installing software outside of the "pool" blows. That is not specific to Mandriva, but back when I was using Mandrake their pool was not the largest so it became a factor more frequently.

I wasn't saying you should try Ubuntu but rather try to accentuate the positive. I read all the/your comments so I get that it's not your style. blunt direct etc.

What do I take you for? Though I don't know you, I take you for someone who is opinionated, sensitive, intelligent and knowledgeable, and I would bet money on honest as well.

Me take you or your readers for idiots? You have it backwards. My statement was speaking to a perceived attitude of you treating Ubuntu as "garbage salad" hence, I would assume anyone using it to be, to you at least, uninformed (you are the only one who has used the word idiot).

To the contrary, I personally, think both your readers and your comments to have plenty of intelligence. If I didn't I wouldn't spend the time interacting.

As far as the "works for me" argument not flying, of course it does. The "me" is simply a single data point, as is every other happy install/use for another an additional data point.

Every bit as valid as each failure, or/additionally bug if you prefer. They are all data points, though folks may ascribe different weights/values to them.

Roll-outs and deployments are evaluated on just these sorts of "hard" and "soft" data points, are they not?

Here comes the hard part.

Caitlyn,

I don't dispute your bug count or bug age assertions.

I have read all the links you posted attesting to Ubuntu's bugginess, and quite a few more.

I agree it would be better if Ubuntu closed more bugs, certainly any hardware bugs/regressions.

I agree Ubuntu should relabel their inter-LTS releases to reflect their more unstable nature.

I'm with you on all that.

I think where I lost you and felt a desire to post the above was the reading of your comments, which on the whole seemed less balanced and reasoned than your actual blog posts.

This response is a good case in point.

"Your claim of universal joy and happiness by millions of Ubuntu users is arrant nonsense. This whole post comes down to fanboyism at it's worst. Please, if you have facts then back them up with links and supporting documentation."

This is... what, I am not sure? You dispute that Ubuntu has a base of millions? Or are you asserting that they all/most are trapped somehow? unhappy? Again "universal joy and happiness" are *your* explicit characterizations not mine.

And by what logic or fact do you assert that the millions (do we agree on this yet?) of users are dissatisfied with Ubuntu?

A couple of developers, a handful of tech writers, and a "bunch" of end users? Is that bunch a hundred? a thousand? ten thousand?

Take all those numbers and double them does that cover it?

You certainly haven't provided even that much "link or supporting evidence" but I am willing to grant it to you.

Caitlyn, they.are.all.data.points.

You want to throw "fanboyism" around? You want to make a call for fact? You want to disparage what info others bring while holding up your own as some iron clad proof?

If so you should rethink that.

I have made few hard claims other than Ubuntu's user base size and a presumption of satisfaction. My own personal experience and opinion. And a plea for a more rational balance in addressing the topic at hand. And an assertion that no system is without faults.

You are the one who has set out such an extensive case and come to the table light on evidence.

And once again:

I don't dispute your bug count or bug age assertions.

I have read all the links you posted attesting to Ubuntu's bugginess, and quite a few more.

I agree it would be better if Ubuntu closed more bugs, certainly any hardware bugs/regressions.

I agree Ubuntu should relabel their inter-LTS releases to reflect their more unstable nature.

I'm with you on all that.

regards Caitlyn,

Clifton

@Clifton: I must say the tone of this post is much different than the previous one. I also must say that based on this second post I misjudged you and where you are coming from. I apologize for that. It seems I'm not the only one who can be combative to use your chosen word. Again, I'm sorry that I misinterpreted what you were saying.

You did say, explicitly, that Mandriva package management "blew". rpm was never the cause of "dependency hell". Poor repository management causes that and it can happen equally with Debian packages as with rpm packages. rpm is the only standard for package management in the Linux Standards Base and is used by the two big corporate Linux distros for good reason: it's mature, stable and reliable. FWIW, I think that is equally true of Debian packaging. With that clarification, no, there hasn't been dependency hell in Mandriva in quite a long time. It just works.

There is software I like and use which is in the Mandriva repository and not in the Ubuntu repository. The reverse is also true, of course. I don't find it any more difficult to use unpackaged software in Mandriva than in other distros. You are correct that commercial vendors and business oriented applications are more likely to be packaged upstream for Red Hat and SUSE (the corporate leaders) than anything else. Ubuntu? Not so much. That is NOT a knock on Ubuntu, BTW. In any case you are looking at requesting a package and waiting or building your own. IMHO, both Ubuntu and Mandriva have large and excellent repositories.

In the first post you disputed what I saw as flaws and in the second post you agreed with me so I'm not quite sure what to say to that. I will say that I do want Ubuntu to succeed, not fail.

Caitlyn,

Well it seems the fires here are finally burning down to embers. I had wanted to respond to your reply, which was a generous and thoughtful one, I tip my hat and thank you for that.

But honestly as I have read more of your work and more of your responses to comments, and then reread a bit, both my desired response and perception of you continued to appreciate and expand.

I am glad you want Ubuntu to succeed. I want the *aims* of Ubuntu to succeed, with or without Ubuntu, and thereby the proliferation of FOSS, and the breaking of the stranglehold of proprietary monopolies and data formats. One of Ubuntu's bugs I think we share equal concern for is bug #1.

So generally to my first post, I should have posted it to your previous blog entry as that is where it rightly belonged. Honestly though, it was a bit of a sarcastic (and correctly noted, combative ;) heated response; that as I read more, and understood better, you and where you are coming from, I deemed inappropriate. This entry was of a different order and had I read it first or at least with clear eyes, I might have generated less heat and more light myself. My apologies to you as well, I misjudged and misinterpreted you also.

As far as Mandrake and RPM go, I used Mandrake for several years not long after it started. It was my first sustained usage of Linux. I am far less technically savvy than you, today, and back then much less so. (disclosure, several years back I pulled a paycheck running a small IT shop for a local non-profit for a couple years (video production, web sites/apps, 50 users on a network I built out (Windows, blech!)); seems like a lifetime ago now). There were none of the higher level tools for RPM as there are now, and with the state of Mandrake's repos back then forcing me out to find individual packages, it was an unappreciated hurdle for my meager skill set. I am going to give Fedora and Mandriva (and thanks for the tip on SL) a spin just to update myself on the current state of affairs in RPM land, and I will update myself on RPM workings as well.

As to your statement that I disagreed with you in my first post, well, that is a mistake. I didn't disagree anywhere in the post.

I had started to write you a much longer post about the miscommunication we have had, and the miscommunications that seem to be happening with many of the commenter's here, but I decided that it might not be appropriate for the setting and our familiarity with each other.

Suffice it to say that I have gone over some of your other writings and comment threads and I think you have great stuff to offer but perhaps aren't getting through to folks as much as you could.

My mom, who is now past, gave me a lesson once. She taught me the difference between being right and being understood, they are two completely different things.

I do want to say also that I find the level of your participation down here in comments to be rare, and refreshing. To me it speaks well of your intent to actually inform. I think many writers don't look beyond the self-reinforcement the article at the top gives them.

You have gained a reader, be well.

Clifton

I have found that each version of Ubuntu gets worse, as far as video is concerned. I have an HP dv4000 laptop with an Intel 915 chipset.

Ubuntu 8.10 - OK.

Ubuntu 9.04 - Doesn't let me choose the correct resolution for my external monitor (Sony 40" Bravia TV, 1366x768), but it's almost correct - still usable.

Ubuntu 9.10 - No picture on external monitor. Turned off video effects, now I get a picture. But when I try to play a video, the video player fails. So I can't use my laptop to play videos on my 40" TV.

Also, playing 720p .mkv files (on the laptop screen) slows down to about 2-3 frames per second - frequently - which is completely unwatchable. Ubuntu 9.04 could play the same videos perfectly. (Linux Mint 8 vs Linus Mint 7 shows the same issue).

On the positive side, the "100 paper cuts" project has made the overall Ubuntu user interface much nicer. But Ubuntu is no good if there's a serious error that I can't fix.

I tried the Beta 1 Ubuntu 10.04 Live CD and the video problems haven't been fixed.

So what now? Mepis?

I totally agree with you and have always been of view that normal ubuntu releases are just beta releases for their LTS. We only keep LTS versions at our computing facility for this very reason.

Upgrading to newer version of ubuntu is time consuming and having to do this twice a year simply isn't a viable solution. Put it this way, their release cycle sucks.

Having said that, I feel ubuntu is facing same thing that Microsoft does... that is once you have a large customer/user base, it gets more and more difficult to get everything going. In order to remove bugs, you keep on introducing new bugs.

I think there is no solution apart from the fact that their is a price you have to pay for simplicity, ease of use and stability.

One of the recurring complaints I see about mandriva/fedora is rpm hell. Why are people still using rpm and not urpmi/yum? Using rpm is like using dpkg on debian/ubuntu and you will have dependency problems also. Furthermore, for any OS that targets desktop users there should be no reason to use the console, whether it be apt,yum,urpmi or whatever. Ubuntu/fedora/mandriva all have graphical package managers that work quite well.

Huh? Dependency hell refers to problems installing a package due to failed dependency issues that are not easily resolved. It hasn't been a problem with Fedora or Mandriva in a number of years.

yum, urpmi, pirut and a host of others are just front-ends to rpm. Most end users do not use rpm from the command line but it remains available.

hmmm... Mandrake only allows a 32bit free download?

is this strike 1#

Nonsense! You can download 64-bit Mandriva for free. For Mandriva Free 64-bit see: http://www2.mandriva.com/downloads/?p=linux-free Only the live CD is limited to 32-bits.

Since you can't defend Ubuntu you're attacking Mandriva now and you can't even get your facts right when you do it.

Yes, they do offer a 64 bit community download, had to dig a bit.

I am glad to see it. I stand corrected. ;)

Well, I've tried and used quite a few distros over the past 10 years or so, including Mandriva (it was my first Linux love, back when it used its maiden name of Mandrake ;-).

I agree with Caitlyn that Ubuntu sometimes regresses in its hardware support. I lost Compiz support a couple of releases ago due to a video driver bug (and switched to Linux Mint for a few months as a result - God intended windows to burn up when closed, darnit!); I currently can't use the scanner on my Canon mp330 due to a driver bug, but haven't yet looked for a distro to run in VirtualBox for my rare scanning needs.

Despite the glitches, though, I stick with Ubuntu for several reasons. One is the huge, friendly, supportive community, well-presented through Ubuntu Forums. I've yet to find another community that fits me so well (Red Hat isn't bad, but it's a distant second; Mandriva is just nowhere close). Another is that, despite the frustrating glitches, Ubuntu "just works" far more often than any other OS product with which I've dealt - including Windows XP and 7, and Sun Solaris. "Not perfect, but better than the others I've tried" still gets my business, though of course I hope they improve.

But what I love most about Linux is that I (and Caitlyn, and you) have a choice. Freedom totally rocks, and THAT is what I'll always defend without question.

George (http://ricegf.com)

@ricegf

haha...yeah that is how long ago I dumped "Mandriva", I thought it was a dumb name when they picked it and I still do, it will always be Mandrake to me. ;)

"But what I love most about Linux is that I (and Caitlyn, and you) have a choice. Freedom totally rocks, and THAT is what I'll always defend without question."

Couldn't have said better.

Let's just pick on Mandriva for a bit, shall we?

"Posted: Fri Jan 22, 2010 12:30 am Post subject: Reply with quote Report
I've never had even close to 16 partitions on my system. As far as installing iso's - they're just disk image files, you don't install them. If you refer to installing operating systems, it doesn't mean squat how many times you do it if you format the partition. I bet you that if I booted 2009 right now I'd be able to clean-install it into the same partition where it used to reside and I might just do that if I find the time in the next few days to prove this point. I'm not considering however dumping all my working OS's for a dubious experiment.

That's just the main problem with Mandriva: they can't help themselves to brake it all every six months when nothing needed fixin'. It's just sad that Linux can't ever mature pass the stage of a hobby for me - I really considered sticking to it full-time instead of always being a curious visitor."

memories...of the way we were.

If you are going to use anecdotal reports from single users you can tear apart any distro you choose. As I said in my other response to you it proves absolutely nothing and is a pointless exercise.

Once again: where Mandriva succeeds and Ubuntu fails is that Mandriva has fewer serious bugs per release. This information is based on release notes published by the two distros. Mandriva also fixes their bugs in a timely manner within a release cycle. Ubuntu is unwilling or unable to do the same. I happen to think they are perfectly able but management chooses not to.

I will point you again to what Chris Smart at Linux Magazine wrote about Karmic Koala, the current release: http://www.linux-mag.com/id/7600 One users comments (including mine) mean very little. A documented and sustained pattern of failure means a whole lot.

"If you are *going to use anecdotal reports from single users you can tear apart any distro you choose. As I said in my other response to you it proves absolutely nothing* and is a pointless exercise."

You realize that the link you provided is to an article that contains

1) the author's personal anecdotal report
2) the author's second hand anecdotal reports from single users
3) the author's opinions
4) a bug count totaling 40 in a RC version

bit of irony don't you think?

I think his follow on article is much better, have you read that one?

http://www.linux-mag.com/cache/7607/1.html

I think this is much more usefull. He is not just pointing out a flaw, which is easy to do especially when others have already done it, as you pointed out every one is doing it.

He is offering something constructive. And at least dealing with the subject with some balance and insight.

As I said above; I am not disputing the bug/code part of your argument, because either way an increased attention to quality code is a Good Thing.


Caitlyn,

I really don't want to fight with you so let me try this another way.

I greatly enjoy seeing GNU/Linux get better, all of it (Mandrake, Redhat, Debian, Ubuntu, all of it).

I have no endearing attachment to Ubuntu, honestly Mint has been tempting me for a while. You got the fanboy part wrong.

My initial post was probably an overheated attempt for the kind of balance and useful insight that he offered in his follow-on post from you.

I fear sometimes, because a certain person doesn't find any, or much, value in a distro/project/license they sometimes overlook the value that *is* there. If not for themselves than others, and can end up doing unnecessary harm to a pointless end with their words.

It's easy to generate heat, that's most all do, but a few generate light. Now that is worth something.

be well,

Clifton

Caitlyn, your Ubuntu experiences sound much like mine. Quality control is definitely lacking.

I'd also like to point out that the Ubuntu/Dell partnership doesn't always work favorably, either. A few months ago, a buddy of mine bought a Dell netbook with Ubuntu installed on a 4-Gigabye SSD. The first thing he did when he received it was to try to update it. Of course, with only a 4-Gig drive, it ran out of diskspace and the update failed. He ended up sending the thing back to Dell.

Also, though I know that your article is focused on Ubuntu Desktop, I'd like to point out that they also have major issues with their server version. Hardy Heron, an LTS release, had to be almost completely rebuilt before it could be considered usable. (Things like shipping it with an outdated version of Logical Volume Manager don't inspire confidence.)

Memories may be beautiful and yet,
What's to painful to remember,
We simply choose to forget,
...The way we were.

"Posted: Wed Nov 04, 2009 5:49 am Post subject: Upgraded to 2010, Now Boots Into the White Screen of Death Reply with quote Report
Hello -- newbie here.

After installing 2010 86/64, Mandriva seems to have detected all hardware correctly. When attempting to boot, however, I get a white screen that slowly bleeds a few vertical translucent stripes.

I have been used to this with my Dell Studio 1535, having tried to install numerous Linux distros that all had the same trouble running the Intel 965 chipset that comes with this machine. Mandriva 2009 Spring was the first one to work -- but now 2010 does not.

I've attempted to go over the install several times using the generic VGA and VESA graphic drivers, but they lead to the white screen, as well.

Any idea why this is happening, and how I can fix it?

Thanks."

@Clifton: All distros have bugs. Mandriva is no exception. Where Ubuntu fails (and Mandriva does not) is in terms of fixing bugs promptly and getting it right within a release cycle. Mandriva also has fewer bugs, particularly fewer serious bugs, per release cycle than Ubuntu does.

I haven't forgotten anything. I can point to anecdotal reports and significant bugs in every major release. Where Ubuntu stands apart is in terms of quantity and inability or unwillingness to fix the bugs.

He is describing a regression, one of the things you are tagging Ubuntu with; on that same Intel chip no less, or is it a different one I wonder.

Interesting.

Nope, it's the same. The distributors who went with a new Intel driver and a new X.org before the code was ready for prime time and while it was horribly broken are all to blame. Not every distro did that. Fedora didn't. Slackware didn't. They avoided the rather serious breakage that occurred as a result.

The case of Ubuntu is especially egregious because they utterly failed to fix the bug or even try for very long under Jaunty Jackalope (9.04), a release that is supposedly still under support. If the distributor fails to fix a major failure with a known cure while the release still is under support and will be for some time then that support is worthless.

What people really need to remember is that if you want stability you need to stick with the LTS versions. As was stated in the comments someone said that they could only use the last LTS release in order to keep their wireless working. I don't really understand the problem they are having. It just sounds strange but 10.04 should work just fine for them. Anyway I've always been use to upgrading every 6 months and now I don't really see any benefit in it. It's easy for a person to keep a stable system and still update your applications when needed just using a LTS release. Common sense would tell a person that the 6 month versions are basically testbeds for the next LTS release. I don't use what is called a one man distro and it's not because they are not good but I've been bitten by them in the past and using Ubuntu is one less thing I have to worry about. I've used several different distros and I just don't really see a lot of difference in any of them. Fedora was the most unstable distro I've ever used but that's fine because I look at Fedora as a testbed for Redhat. Mandriva is very good, OpenSuse is fine but has never impressed me for some reason. So to answer the question, is Ubuntu giving Linux a bad name? Of course not. That is just hogwash. None of the other big boys have stepped up and done for Linux what Ubuntu has in the eyes of the general public. Ubuntu is not for the die-hard linux geeks in the world. It's for everyday people and it seems to me that these everyday people have less problems then the Linux veterans. Just upgrade every two years. Problem solved.

LTS means long term support, not long term stable.

"one of the other big boys have stepped up and done for Linux what Ubuntu has in the eyes of the general public."

If it wasn't for the big boy distributions like RedHat and SuSE, Linux wouldn't be in the datacenter today. That IS in the eyes of the general public.

LTS means long term support, not long term stable.

But that's precisely the point: Caitlin argues that LTS *should* mean STABLE, and I happen to agree. From the article:

tout LTS as the stable "Linux for human beings" and the six month releases as cutting edge. [...] This is essentially the Fedora/Red Hat approach and that company is certainly doing well.

That's my opinion exactly. I don't even care about better Quality Assurance for the intermediate unstable releases; it would be nice to have, for sure, but I would be perfectly happy to accept that "if it breaks, it breaks; if it cannot be fixed, then it will remain broken." The LTS releases, however, should be developed with stability and superior quality assurance as the highest priority.

@Eddie: The point you are entirely missing is that Ubuntu touts the regular releases as ready for the masses, not just LTS. If you reread my article you'd see that I made clear that I would have no valid complaint if they touted LTS as the releases ready for mass consumption and the six month releases as cutting edge. The problem is that Ubuntu doesn't do that. They sell the public a bill of goods about how wonderful and easy all their regular releases are and then consistently disappoint.

Common sense doesn't tell the general public anything. When you start with no real knowledge of Linux whatsoever all you have to go by is what you've read and seen. Joe and Jane User don't know about LTS because that isn't what's sold to them.

Anyway, I think it's your conclusion that Ubuntu isn't harming Linux adoption with shoddy products that's "hogwash", to use your chosen word. I also think the Ubuntu community is loyal to a fault, to the point that they are looking at this with blinders on and are deeply in denial about real problems.

Normal Ubuntu releases are based on Debian unstable. Why should Ubuntu manage to release a stable working version every 6 month when Debian needs ~2 years ?


Instead of integrating all the software Canonical tend to focus on marketing, trends and new directions (ubuntu One, music & software shop, stupid themes).

"In fairness I really do need to focus on one area where Canonical and Ubuntu have done an outstanding job in reaching new users outside the Linux community: preloaded systems."


Well, that does not seem that outstanding: I suppose it is easier to have a consistent set of drivers for a given HW configuration than in a general case, where hundreds of boards, disks, etc have to be handled.
I had the feeling (from using unetbootin'ed liveCDs of many linuxen, my restriction being that the GIMP and gcc should be installed, and my keymap eqsily handled: if these constraints were not present, I stopped/did not download) that many distributions can work almost flawless on the "net""books" I -and my neighbors- could directly buy in a small (thanks heaven) french province town.

My favorite internetcafe landlord found Mandriva2010.0 very nice, when I showed it to him (and he liked Scientific Linux, too) and did not like W7 (too slow: that is astonishing) which is preinstalled on the same "net""book"

What is outstanding about what Canonical have done with OEMs is a tribute to excellent marketing. Mandriva, SUSE, and Fedora all tried to compete in this area and failed. Xandros, gOS and Linpus actually succeeded first, landing ASUS, Acer and Everex, and had their distributions preloaded on the first generation or two of netbooks. They were all locked down, severely limited and, in the case of gOS, severely broken (see: http://broadcast.oreilly.com/2009/02/netbook-nightmare-my-experienc.html ). Despite robust sales for ASUS and Acer their competition, and eventually Acer and ASUS as well, all looked for better Linux solutions.

HP and MSI briefly went with SUSE Linux Enterprise Desktop but it was, unfortunately, poorly implemented and Novell lost a golden opportunity to make inroads on the consumer desktop. Almost everyone else (with Emtec/Gdium as a notable exception) went with Ubuntu with much, much better results. In some cases Ubuntu migrated up the product line from netbooks to laptops and desktops as well.

The success of Ubuntu in preloads was due to two factors. First and foremost is marketing. When it comes to marketing desktop Linux I really believe Canonical has no peers. They know how to use free publicity, to generate media coverage, they understand product placement, and so on. The second reason for their success is that they, unlike Novell/SUSE, worked successfully with the vendors to insure that the product they delivered actually worked. By focusing efforts on a limited set of hardware to be sold they accomplished on preloads what they have failed to accomplish in general.

Is that outstanding? In my view, yes. They succeeded where other Linux distributions failed. It also demonstrates that if Canonical decide to have their Ubuntu developers focus on a project they invariably get really good results.

"What is outstanding about what Canonical have done with OEMs is a tribute to excellent marketing. "
I agree with you, but, from a technical (not PR|marketing) point of view, many (not all : it is ironical those who were chosen to be preinstalled often behaved the worst way) other distributions can do as well or even better;

I quoted Wolvix *Beta* (has a preinstalled GIMP -my nephew really needs the GIMP and I know I am not that kind with UBUlinux!- and gcc,and it is easy to have any keymap -Sabayon seems to become more and more complicated in this respect, on live [i.e demo] DVDs)
This was a Public Relation competition, and, if they do not follow technically, people will get upset (MSI'Suse I could manage by recompiling everything I needed, after building a statically linked gcc binary on RH; sorry, but can a newcomer imagine that? I was before I knew of unetbootin).


"HP and MSI briefly went with SUSE Linux Enterprise Desktop but it was, unfortunately, poorly implemented and Novell lost a golden opportunity to make inroads on the consumer desktop."
I never had sawn such a lousy thing (do beginners read the Linux From Scratch book? in such conditions, they should!)... but now, MSI has Windows Seven in it.... what will they do with the next release??

" Almost everyone else (with Emtec/Gdium as a notable exception) "
Well, this is (should be?) an interesting exception, though off-topic: a friend of mine thought it would be a change from the Intel(amd) family, ant I told him that, unless the Gdium had a huge success and various demands, he would have to add most of the softwares he loves by some cross compiling and that would be long, difficult and unpleasant.
I hope I was wrong, but it does not seem that successful in Europe...

One way to tackle the problem would be to investigate what exactly led to yesterdays printer problem. Was it an upstream degredation? Was a previous Ubuntu bugfix not sent upstream and then forgotten by Ubuntu packagers with the next release? It would be interesting to make a couple of documentation cases about these degradation issues. What happened, how happended it, how to prevent it.

Also I believe while there is healthy competetion, there is still too much elbowing in the Linux distribution landscape. Distribution packagers here and there reimplement (or forget) the same bugfixes independently of each other.
There should be more SHARED MAILING LISTS between distributions. Especially for shared core driver packages, like your HPLIB printer thingy. It's not sufficient that some of the Dist packagers hang around on upstream mailing lists / forums, if there ARE any. More exchange is required to advance Linux as a whole.

@mario: I've already indicated the problem is in either udev or devicekit. Looking at the Launchpad bug report on the similarly affected Minolta printers shows me that the Ubuntu developers have come to precisely the same conclusion. hplip is not part of the problem.

The point, once again, isn't one bug that happened to affect me. The problem is a history of lots of bugs (more than other distros) and a failure by Ubuntu to respond to those bugs and fix them on a timely basis.

haha. Don't tell my about udev. Just reinstalled my system yesterday. - Just imagine it's not your printer that doesn't work anymore, but mouse and keyboard won't.
(I've been using Linux since Slackware '96. But I'm totally clueless how to fix the fragile Ubuntu base system mess at times.)

One thing that his (as do most blogs lately) seem to forget is that Ubuntu is not just Canonical, but also a community. More so than Fedora or Mandriva (how many non-employees attend Fedora or Mandriva developers summits every 6 months?).

Another thing people forget is that in order to fix bugs in releases, the fix has to be tested on a wide variety of systems to make sure it doesn't break something else (x86, x86_64, IA64, PPC, Sparc, Arm, PS3). When a distro only supports 2 hardware models (x86, x86_64) then testing a fix is relatively fast and painless. However, if a patch requires a rebuild of a package across the whole spectrum of supported platforms listed above, then it really can take a lot more time. For example, OpenOffice takes a few hours to build on current Intel (AMD) hardware, but it takes almost 3 days to build on arm. Also, you can walk into any computer store/Fry's/Walmart/etc., and buy x86_64 hardware (which will also run x86 code), but I challenge you to go and do the same for Sparc, PPC, IA64, or Arm. Lack of hardware in these architectures limits what can and cannot be reasonably built, let alone tested.

Even if the quantity of hardware available wasn't an issue, you still have the issue of limited testers. Once a product is released and development focuses on the next release, QA follows the developers, sometimes testing new versions of the same packages multiple times a day, depending on the build speed. The testers' hardware is also now running from the new code base. To go back and retest fixes on the previous release requires (in some cases) either;

A) A drive dedicated to the previous release (some hardware, like arm, don't have the luxury of a visable bootloader like grub for dual/multi-os booting).

-or-

B) Reimaging/reinstalling the previous release, wiping out the currently active test environment. In some cases, just the reimaging process can take over an hour to install plus update just to get to the point where these tests can be run. Yes, the argument that hard drives are really cheap has been made, but on some of the platforms I mentioned, the image can live in NAND memory, or on SSD with limited numbers of writes (and I personally have killed a few of both just with the normal installation testing of a development image).

An easier interim solution to specific issues (aka, your printer or some alsa issue) is that when the bug does get fixed in a newer version of the software, even if it is for the next OS release, you can still install it by pulling it down from launchpad or from upstream. Maybe someone at Canonical can do up a wiki page providing instructions on how to do this.

I have been a tester of various software and hardware systems now for 20 years. Some bugs that I have found back in '93-'94 in certain proprietary systems are just now being fixed. Others never will. Ubuntu (and Linux in general) have at least been very good about fixing bugs relatively quickly, even if it means jumping from release to release every 6 months.

@Tobin: What you are completely overlooking is that other distributions have a much better track record at fixing the major, show-stopping bugs than Ubuntu has. If Fedora or Mandriva or openSUSE can do it then certainly Ubuntu can do it as well. Nobody else leaves things like support for a major category of graphics chipsets broken through an entire release cycle. Note that I am comparing Ubuntu to other distros with corporate backing and large communities. It's an apples-to-apples comparison and one where Ubuntu does not come out looking good at all. I don't see how on earth you can claim they are doing a "good job". Indeed, most of what you write is absolutely true but entirely beside the point.

I beg to differ about your assessment of Fedora's contributor base being less community centric than Ubuntu's. Have you done the analysis of the break down of Fedora's FUDCon participant list? If you haven't you are making some very unsubstantiated claims.


Here's an overview of the analysis I did just on our active package maintainer.

http://jspaleta.livejournal.com/47023.html

As of Aug 21, 2009 the Fedora Package Maintainers looked like this:
Total Maintainers : 770
Red Hat Maintainers: 282

There's more analysis in the blog entry. Please, don't spread assumptions about what you think the Fedora contributor base looks like if you haven't done the work necessary to back up the claim. I spend a lot of my volunteer time trying to get an accurate picture of it. When people make off-the-cuff remarks about it, it upsets me..greatly.

If you have a data-centric view to back up your claim that Fedora's contributor base is somehow more Red Hat dominated than Ubuntu's contributor base is with regard to Canonical dominance...please..by all means..publish your analysis.

I would love to see anyone inside the Ubuntu community step up and do an analysis similar analysis to get an accurate picture of how much of the work really is community driven and how much is really Canonical driven as a starting point for a comparative discussion. Are you the person whose going to step up and do that analysis?

-jef

I like the way you think. Reading your responses to comments is honestly more enjoyable than the article itself (also a good read).

"I also don't subscribe to the notion that writing critical articles about a Linux distribution or an aspect of Linux is harmful. I think some honest examination and airing of issues is a positive thing if it gets people thinking about how to make Linux more appealing and how to solve problems"

We need more of this. Linux religion (blind worship of broken distributions, processes, methods) needs to stop so that we may finally get on the path to platform success.

In order to become ready for critical mass we need more criticism. Bravo.

Thanks. Lots of words flying around here. You are SO patient (and/or committed). Etc.

I gave up on Linux in 1996, and came back in 2008 with Kubuntu 8.04, which is fine, except for various, random problems that come and go as I continue to do updates, with different problems on different computers.

Then 9.10 disabled a brand new machine (from Zareason, no less), so I began looking. Mint is nice but oops! another unexpected, unexplainable problem.

So Mepis 8.5 is installed now, and looks promising. I'm staying away from RPM-based systems due to previous experience, but hoping to find a Linux home that I can just run, and get things done with, without fuss.

Best wishes, and good luck with the continuing flood of responses.

First, thank you for the very nice compliment. I've never thought of myself as a patient person.

When was the last time you used an RPM based distro? I ask because an awful lot has changed since 1996. There is good reason why RPM is the package management system that is part of the Linux Standards Base. Red Hat is the dominant presence in the corporate server room with SUSE second. If rpm was problematic or unreliable that wouldn't be the case.

I've got the impression Caitlyn that your complain about Ubuntu has two different components which are in general no too well differentiated but should be (or at least to what extend and in which way they are connected should be clarified). You obviously dislike Canonical and its marketing machine. And rightly so. You are right when you say they are peerless at it. Probably one of their worst offenses is not to clarify that releases between LTRs are pretty much betas (or work in progress) of the next mayor release. There is probably a not too openly-acknowledged reason: that the base of users who go on to continuously upgrade to the next release are being used as early testers without being told so. On this context it "makes sense" that resources that could go to providing better stability and fixing bugs, go in fact to developments on desktop integration or ever-growing number of "supported" hardware. Canonical obviously has an enormous manpower to test and report, test and report. That Canonical and the man behind are AMBITIOUS isn't arguable.

A different thing from Canonical is the product itself, Ubuntu. I think you're probably over emphasizing Ubuntu's technical failures. Once we agree that releases between LTRs are what they are (in spite of Canonical "lack of clarification" on this) we can see that LTRs haven been so far landmarks (and 10.4 indeed looks like the best one to come). By this I mean landmarks of the Linux desktop. 10.4 indeed makes you think of 12.4. But the honest recommendation here is: if you are happy with how your computer does things, upgrade only in two-three years time: leave the 6-months upgrades to geeks and the like. And on June 2011 install April 1010.

(Having said all that I really, really hope that my sound will some day work out of the box. Once the forums for 10.4 are at work I hope I'll mange to fix it again. So far even removing pulse audio hasn't done the trick)

"A different thing from Canonical is the product itself, Ubuntu. I think you're probably over emphasizing Ubuntu's technical failures."
I do not think Caitlyn Martin is overemphasizing a abnormally (with respect to other linuxen) great mean time to repair (else, every one who buys /gets buyed lots of PCs, cars, bicycles would be a moron, as it is one of their decision criteria). It is a valid criterium if they claim to be ready for the masses (and AFAIK, the masses do not care about the numerology subtle differences : they prefer having something which does not catch bug reintroduction, and, if any, which gets quickly fixed).

Caitlyn, while there are some things about your blogs that I do agree with, your trollish attitude is more "holier than thou", and that turns people off. Your responses to peoples comments are both hostile and argumentative.

A lot of the folks posting here are posting sincerely and just wish to share their opinion, but you openly attack them, and this is not the Open Source way of doing things. People will disagree - just how it is, but rather than be hostile, we should be inclusive and accept that different people have different views, and who knows, maybe some of them are right and you are wrong! Without an open mind, you will never be able to see the difference.

On topic, I do agree that Canonical does have a history of failing to fix some fairly significant bugs. I do not know the reason for this, and neither do you. I do also know that nearly all the major Distro's out there are guilty of doing the same at various times and for different reasons. Mandriva is no "one stop shop solution", and there are many other Distro's out there that do very well indeed on the Desktop, like Debian Stable, CentOS, and even others like those you mention or ArchLinux or heck, even Gentoo does a great job.

@davemc: I haven't attacked anyone or been rude to anyone yet. I've been neither hostile nor argumentative IMNSHO. I've looked through my comments and I have, if anything, been more than patient. I'm certainly not closed minded. Some people seem to interpret honest disagreement or sticking to a position as closed-mindedness. It isn't, nor is it in any way exclusive.

Perhaps you misinterpret direct and to the point and sometimes blunt as attacking or argumentative. A lot of people say that about New Yorkers or Israelis (I'm a bit of both) and I dare say that's a cultural difference more than anything else.

I am direct. I say what I mean and I mean what I say. I understand that you and some other people expect or want a softer approach. All I can say is that, sorry, that's not me.

Offered for your consideration is this 2007 article for O'Reilly by Carla Schroder: http://www.oreillynet.com/linux/blog/2007/12/ubuntu_innovates_excuses.html Here's one more by Akkana Peck: http://shallowsky.com/blog/linux/ubuntu-aumix.html

They essentially said what I am now saying more than two years ago. Nothing has changed in all that time. How many prominent voices in the Linux community saying the same thing must I provide before some people realize there really and truly is a problem here?

I've been a Mac user exclusively since I got my first computer 22 years ago. But I'm also attracted to the idea of FOSS, and would like to see some form of *nix-based desktop (besides Mac OS X) get some real market share, if only because Apple could use some real competition -- but also because a healthy ecosystem is a complex ecosystem, and Microsoft's blanket dominance has not been good for anyone. (Nor would I wish to see Apple take Microsoft's place; Apple is arrogant enough already.)

I have no interest in the apparently interminable infighting in Linux land, because I don't live there; what I'm interested in is what may emerge from the FOSS world to be a player where I do live, the desktop environment of the mass of ordinary home computer users. (I've been a Mac support consultant since the early 90s, so have a pretty good idea of what non-tech users want/need.)

What I'd really like to see is a *nix-based computer world, where anyone can try their hand at building a better OS, but all would be similar enough to allow easy porting of applications among them (as is now possible between the Linux/BSD and Mac environments) -- and then let Microsoft see if they can compete on merit rather than chicanery.

So far as I'm aware, the Ubuntu project is the first (and so far only) serious effort in the Linux/FOSS community to break out of the geek ghetto and make a place in the Real World for a free-based operating system. So I've been following it, if only from the sidelines, with interest. I've installed Ubuntu in a VM on my Mac a couple times and played with it a little: kinda primitive, but certainly usable, with a refreshingly open and fun feeling to it -- sort of like the Mac was in the early years.

I was pleased to see Mark Shuttleworth issue a challenge to the FOSS community, to make besting Apple their goal. Not gonna be easy, as Apple is not standing still, but aiming for excellence rather than mere Microsoftian "good enough" is a good idea.

But in light of recent developments, I'm afraid I I'm beginning to have some serious doubts. As you say, it seems Shuttleworth wants not to compete with Apple, but to be Apple -- which is quite a different thing. Imitation may be the sincerest form of flattery, but in the end that's all it is: the imitator is always a follower, never a real contender.

The new Ubuntu branding looks good: more serious, less like a toy. The new themes I can't really comment about from the little I've seen -- though making the dark one the default (?) makes no sense. Different just to be different? Nope.

As for Ubuntu's bugs, I can't speak from experience, but have been puzzled to see so many problems reported on the Ubuntu forums after each release. I mean, given that the project is not encumbered with the pressures faced by commercial ventures like Apple or Microsoft, it would seem to make sense to take the time available to get it right before releasing it -- since they can (or could, if they cared to).

I wasn't aware until recently (reading here and elsewhere in the last few days) that Ubuntu's new-release bugginess actually is not typical of major Linux distros, which makes it even more puzzling. The carved-in-stone six-month release cycle has always seemed to me entirely too rushed; Apple takes one and a half to two years between major versions -- which still always have major problems.

But it's the window button brouhaha that really brings me to serious doubt and dismay about the future of Ubuntu -- and thus, sadly but realistically, the prospects for FOSS on the desktop.

If Shuttleworth wants to compete with Apple by putting the window buttons on the left, all I can say is it ain't gonna work -- unless he also moves the menus out of the window to a bar at the top of the screen, Macintosh-style. (Oh, and a few more little details.) Which seems highly unlikely. As it stands, from a Fitts' Law UI science perspective, Ubuntu has now accomplished what I would have regarded as highly unlikely: taking a firm third place, after Mac OS and Windows, in usability. Not aiming very high, that.

Second, If Shuttleworth believes he can compete with Apple by imitating Steve Jobs' management style, I don't think that will work either. Jobs gets away with it because he's usually right (though not always) -- and because in 1997 he returned to an Apple which, though faltering, still had some 30 million very loyal customers with nowhere else to go. In fact, when the first release of OS X bore little but cosmetic resemblance to the classic Mac OS, even Jobs felt he had to respond to user outrage by reversing several key decisions.

While there is a small if vocal minority who worship the Very Ground on which Steve Walks, that's not who pays Apple's bills; it's the millions of users who simply find the Macintosh the most user-friendly computing experience. If anything, the few Mac users who might be tempted to switch would likely be motivated more by distaste for Jobs' personality (and its reflection in Apple's) than any other factor -- so finding another Jobs at Ubuntu would not exactly be a draw.

Similarly, most FOSS forums (including this one) seem to be populated mostly by vociferous geek types who haven't a clue as to what it will really take for FOSS to get a real foothold in the Real World. Until recently, it has seemed Shuttleworth did have a good idea, and was implementing it. Unfortunately though, he seems to have lost touch with his good sense.

Now is exactly not the time to be making such drastic UI changes; now, as Ubuntu is poised on the threshold of real success, is the time to emphasize stability and reliability, both in the sense of "it just works" (not rendering peripherals inoperable) and not crashing, and in stabilizing on a UI that won't confuse Grandma by moving such major elements as window buttons from one release to the next (and making Ubuntu significantly different from, I gather, every other major Linux, as well as Windows) -- without even any explanation.

That many Ubuntu forum posters offer a solution that requires the command line to move the buttons back -- apparently assuming that this solves the problem, thus that such major customization will be the norm among users, along with an understanding of the command line -- only highlights how far even Ubuntu is from the Real World that Linux desktop-enthusiasts presumably want to conquer.

So what happens now? If Ubuntu is going to fumble the ball, can another distro take its place? At Distro Watch, the next two in popularity seem to be Fedora and Mint. Fedora, it seems to be generally agreed, is too cutting edge and geeky to take on the role -- nor is it intended to. Of course, there are numerous other distros that, in terms of usability and community spirit, could probably serve as well as Ubuntu. But none have even a tiny fraction of Ubuntu's momentum -- or resources.

So the only other contender for "FOSS standard bearer in the Real World" (I write FOSS instead of Linux so that other *nixes, i.e. BSD, aren't excluded -- though at the moment their chances in this field are minuscule) at this point seems to be Mint. Aside from being based on Ubuntu (though I gather it may connect directly to Debian instead), the problem with Linux Mint can be seen right in its name. Note that Ubuntu is just "Ubuntu"; I believe "Linux" was part of its name once, but has fallen away -- in fact "Linux" no longer appears anywhere on the home page.

Like it or not, the average computer user, if e has heard of Linux at all, knows of it as geek-land, a place where e wouldn't be comfortable. Mac OS X is based on BSD, but it isn't called Apple BSD ("What's BSD?"). Branding matters, especially if you're trying to break into a market with already dominant players. So I wouldn't see any OS with "Linux" in its name going far; it needs to be presented as complete and sufficient in its own right -- and of course, not only equal to the Big Players but better than them in at least some respects.

But "Mint" by itself is too short, and vague: does it refer to "mint" as in money, or "mint" as in tea? And what does either have to do with computing? It's a cute name for a distro in geek-land, but will it fly in the Real World? Most other Linux distro names are also just that -- clever names for distros, not brands that can make a place for themselves by themselves outside the Linux world.

As a somewhat-jaded veteran of the Sixties, I admit I was a little put off at first by the slightly excessive "kumbaya" flavor of the Ubuntu "philosophy", but it does at least have some solid rationale behind it that can be built on for marketing. I note that "Linux for Human Beings" is no longer the slogan; "Free Software for Everyone" might work.

The other problem with Linux Mint, of course, is that Ubuntu is so far ahead in market share -- is, in fact, the only FOSS desktop OS with any market share outside the geek ghetto. The reason Dell is offering computers with Ubuntu installed is because Canonical/Shuttleworth has done the hard, tedious work to establish credibility and enough market share to get Dell interested. Even so, if Dell doesn't find Ubuntu worth the trouble, they'll certainly drop it -- which will make the whole process even harder, much harder, for the next effort.

If Shuttleworth/Canonical really do manage to alienate enough users, perhaps Mint will become "the #1 GNU/Linux Distribution in the world". But that's not my point; it's not that Ubuntu is currently the #1 Linux, it's that Ubuntu is on the threshold of becoming a contender in the arena with Apple and Microsoft. Which is a whole 'nother game. And if Ubuntu blows that chance, when will there be another? Google Chrome might take off, which would not be a bad thing, diversity-wise, but it's really not a desktop OS, nor, though FOSS-based, a FOSS OS in spirit.

So what's the solution? Frankly, I don't know. The best solution would be for Mark Shuttleworth to somehow come to his senses, and (a) learn how to manage his temper better (I know he has to put up with a lot of stupid carping from clueless idiots, but he needs to distinguish between that and intelligent, constructive criticism -- and not alienate valuable allies); (b) put the window buttons back where they belong; (c) find some UI "experts" who know what they're talking about; (d) while making it clear that Ubuntu is "not a democracy" (with which I have no argument), make it also clear just what he means by "community" by showing some respect for users' thoughts and feelings -- and taking the trouble to discuss and give real reasons for such a major change before just implementing it by fiat.

Oh, and of course, (e) revise the release schedule so that the developers can (and must) take the time to get it right before sending it out -- thus leveraging Canonical's advantage by releasing software without major problems.

Will any of that happen? I dunno. If not, can any other FOSS OS become the "standard bearer"? Seems doubtful, unless Ubuntu really does commit total suicide -- which would, of course, make it doubly difficult for anyone else. "Linux for the desktop? Oh, you mean like that Ubuntu mess? Sure...."

The whole situation makes me kind of sad. I've been nudging a Windows-user friend (who probably will never be receptive to moving to Mac) to try out Ubuntu on one of his two laptops. But now I'm wondering: Shall I tell him, "Ubuntu is the best, but first you have to run this undocumented command in Terminal to put the window buttons back where they belong"? Or: "Ubuntu is the best, but be warned that the guy who runs it is prone to making arbitrary, major UI changes for no discernible reason"? That'll go over great.

Or should I recommend something else? "Well, there's A, or B, or C... or maybe you might prefer D or E." And when I put a new hard disk and 10.6 'Snow Leopard' in my MacBook Pro next week, which FOSS OS should I install in a VM to start learning about it? I really don't have the time to learn half a dozen distros on the chance that one of them might break out of the pack. If not Ubuntu, who is the "Standard Bearer"?

Well, just some thoughts from over here in Macintosh land.

About that 1020 printer... mine has never worked properly with any Ubuntu since 8.04.

I wonder what makes your 1020 or your system different from mine.

I don't know. After a couple of years of posting about the issue, I've given up. My 1020 will not load properly at boot. It must be off at boot time and can then be turned on later, after which it works until the next boot.
The odd thing is that this did not happen with earlier versions of Ubuntu (8.04).

The problem has something to do with USB startup by the computer. It is possible it is a Gigabyte or BIOS issue. Why others do not have it, I have no idea, but I can replicate it on my machine anytime for anybody.

Apparently the haters are already at work.
"Caitlyn Martin is a troll!"
http://yalb.net/?p=159

This one clearly doesn't get the point of this post, and accuses you of trolling. I think they're misguided, or a complete idiot. Assuming they're already an Ubuntu fanboi/grl, so anything bad you say must be trolling. Idiot.

This happens anytime any writer criticizes someone's favorite distro. If you want to win popularity and have lots of friends don't ever write critically about Linux. That's a sad fact about our community. For the writer it's simply part of the territory. This guy is mild compared to one I saw on LinuxToday.

I definitely agree on what all you say here Caitlyn. I have been an avid user of Ubuntu since 8.04 and till today I feel that was one distro which felt solid to work with.

I am more of an adventurous kind and like to shuffle things here and there in the system and also a programmer (not the most nifty at), but what I felt working with Ubuntu is that, it's greatest strength is not the quality it delivers (as you mentioned other distros often are better than Ubuntu in this respect) but it's the huge community and the wealth of knowledge that is spread in the Internet about Ubuntu right now. That is something I personally think no other distro can match at this point of time.

But for an user with no Knowledge at all about the world of Linux, I would prefer them to go with Linux Mint.

As for Mandriva, if ever anything breaks down, I am afraid if I would get the solution of it as easily as I get it in Ubuntu.

But yes, Mandriva is far better in delivering a product focused at the end user.

My verdict:
New User:- Linux Mint (it's Ubuntu underneath anyway ;-))
For me:- Ubuntu

Honestly, how hard is it to get the point here? She's only re-stated it numerous times. The fact that numerous bugs are present every single release, and the majority are not even fixed, only left to be done in a later version, is shoddy and unprofessional, especially for a distro that claims to be for the "average user". It can do a lot well, but these issues should be taken seriously, the quality control needs definite tightening up, and the damn bugs need to be fixed before, or at least during the release life-time, not ignored or put-off for a later release. It's shoddy, and lazy. The good it does is nice, but other distros manage to fix their stuff in reasonable time, and during release life, so why the hell don't they? Caitlyn has raised some very reasonable questions, no "trolling" involved. The lightweight losers who accuse her of this are just affronted that their "precious" has had it flaws high-lighted. These are real issues, and deserve to be looked at in a reasonable manner. Caitlyn has remained nothing but polite and professional during this entire exchange, which is surprising considering the ugly tone of some of the posters, who seem intent on focussing narrowly on certain things, and make the issue about minor points that were raised than look at the overall picture being painted. Claiming this is a "beginners" distro, and for the "average" person is flawed and faulty belief if fixing issues requires "hacking" the damn thing to death on the users side. The issues need to be fixed, they've had years to straighten this stuff out, yet seem intent on throwing the latest and freshest stuff at people as unpaid Beta testers. This is what Fedora is accused of. It's either for "human beings", or it's for geeks. Geeks are human, but not all humans are geeks. And if your system doesn't work as it's advertised to, and your not a geek, you're pretty much screwed. Bugs need to be fixed, not ignored. And people should not flame someone for raising issues just because they don't like the content. "It works for me" is nice for you, but there are many it doesn't "work for". Look at things reasonably, and not thru your own narrow focus and viewpoint. And be good to one another. We're all supposed to work together. Flaming someone who raises valid points, and doesn't just heap blind praise on your favourite, is not productive. Neither is people calling "troll" when they hear something they do not like. Learn to deal with the real world.

Interesting perspective, Caitlyn. You mention your background in business; I can relate, because I run a (very) small business. I'm a doctor (www.averyjenkins.com) who owns a small clinic in a small town (www.docaltmed.com). I'm computer savvy, but by no means an IT guy. And about 18 months ago, faced with some aging computers, aging software, and Vista, which blue-screened me 5 times per day, I went in search of an alternative.

After running across Mandriva, I settled on Ubuntu. Why? Precisely because of the level of community support. No, things didn't work perfectly, but community-supported Ubuntu is 6 times better than company-supported Vista or Windows. I actually get answers to my questions. I get work-arounds. I get my business going. Mandriva, as nice as that distro is, just doesn't have as profoundly active a community.

Today, I'm happy to say, my clinic is 100% Ubuntu and 99% FOSS. There is no windows software left except for one vendor-supplied legacy program in VM. And, honestly, networking with Ubuntu was a cakewalk compared to my years of struggle with XP's networking hilarities.

So my experience, coming from an in-the-trenches business background is just the opposite of yours. My experience is that Ubuntu Just Works. It gets the job done 24/7 at a far better ROI and far less maintenance expense than anything else I have used in my 15 years of running a clinic.

One thing which you highlight and I most certainly agree with is that if you compare Ubuntu to Windows n terms of reliability and stability then there is no doubt that Ubuntu wins. I also have not said one negative thing about the Ubuntu community. There is no doubt that the community support is there. That isn't unqiue to Ubuntu, though. I've found that even some smaller distros have outstanding community support. VectorLinux immediately comes to mind as an example. While community is undoubtedly a strength for Ubuntu it is not a unique strength.

I'm glad you don't have problems with Ubuntu. I never said everyone did. I also suspect in your business you don't automatically upgrade every six months but rather do so when you have a need or when a version is going out of support. That most certainly increases your chances of success with Ubuntu.

I suggest you at least take a look at https://bugzilla.redhat.com

There are tons of bugs that hasn't been fixed for many years too.

This is a common problem for all Linux Distros.

By the way, the HP 1020 also has problems with Fedora too, the latest bug was filed on Sept. 13, 2009, and it isn't fixed yet.

Check this out:

https://bugzilla.redhat.com/buglist.cgi?quicksearch=HP+Laserjet+1020+printer

IMHO, all popular Linux distro. that has large user base are potentially in the list of 'Poor Standard Bearer for Linux'

"There are tons of bugs that hasn't been fixed for many years too."
How many? (a number is more pertinent than a weight)
how many ears?
Can one get a hint on the time one has to wait till a bug is fixed (for each -closed- bug, just make the difference between the time it was repaired and the time it was fixed; then make an histogram and an average of theses times) for :

* Red Hat (I did not try)

* Launch Pad /Ubingtu (from their site, stats do not seem that open and visible). Why do they hide them?

* others....

and another hint of the number of bugs waiting to be fixed, and since when (the same very simple calculations apply if the structure )

It is not a matter of loove, "tons of", etc... but relevant figures (in many professional domains).

>> How many? (a number is more pertinent than a weight)
>> how many ears?

well, approximately 13946 numbers since 20000. : bug list of "can't fix" , "wont' fix" , and "upstream bugs" since year 2000

Redhat bug list here

sorry for my post above, the url is too long. here is the tiny url version:

Fedora Bug list/history of "can't fix" , "wont' fix" , and "upstream bugs" : http://tinyurl.com/ydmagsl

Redhat bug list here : http://tinyurl.com/y7bfa4v

I reviewed the list you linked. At lot of these aren't bugs at all but rather feature requests, i.e.: adding Nagios to Fedora. The vast majority are tagged low or medium. Even the ones listed as high or urgent (and there are very few over a very long time period) often impact a relatively small number of users.

By comparison Ubuntu has a lot of high impact issues (i.e.: the Intel video driver issue in Jaunty, the intel_sda sound issue in Intrepid) that they decided not to fix despite huge impacts on large numbers of users. Comparing Launchpad to Bugzilla shows the difference between Ubuntu and Fedora is huge and doesn't weigh in Ubunu's favor at all. If anything you have helped illustrate my point effectively.

Here is the updated search result: http://tinyurl.com/yengkpc

In case I make any mistake again, you could make an advance search by yourself.

I'm not saying that Fedora is as bad, or It's not Ubuntu's fault, my point is that this is actually a common problem for all Linux Distros.

I do NOT for a minute believe that one list of Fedora bug reports proves that this is "a common problem for all Linux Distros." Show me a comparable example for Slackware, please. So much for "all". Second, it's an issue of quantity and severity of bugs which do not get fixed. Taking numbers in a bug report ignores the nature of the problem.

One other factor you are overlooking which is critical to this discussion: Ubuntu is, in their own words, "Linux for human beings." Fedora is a cutting or bleeding edge test bed and basically sells itself that way. What is acceptable for Fedora, which is aimed at geeks looking to test the latest and greatest, is not acceptable for a distro touting itself as Linux for the masses. Now, if you were making the exact same argument about Mandriva or PCLinuxOS, which tout themselves as user friendly and easy for newcomers, you'd be making a cogent point. Nobody I know of claims that Fedora is ready for mass consumption.

When I change priority and severity both (urgent,high, medium), this count halves -7352-... for any kind of HW (I noticed one linked with PPCs : does Ubingtu support it?...)

It remains exactly as long if I make it begin in 2004 (UBUlinux birthdate). There was one(1) bug before 2007,
4000 in 2007,
1500 in 2008, 1700 in 2009-2010 (they have high standards: are issues with atlas -a math library one can recompile- that common and severe?).

Once they have started, I notice they (Fedora) had a huge improvement between 2007 and the next years (twice less bugs : does UBU linux improve in such a dramatic way?); and the number of unfixed bugs seems to go on decreasing, but not that fast.... Is it the same thing with Ubingto
And does UBUlinux count severity the same way? I hope they do.... else, it would be biased.

I support Red Hat professionally so I probably spend more time in Bugzilla than most. Do some low impact or minor bugs go unfixed? Yes, unless a commercial customer pushes on them. It definitely happens. What makes Ubuntu different is that major problems which effect a large number of users go unfixed and are deliberately, as a matter of policy, put off for the next release. No distro gets 100% of everything fixed and you are absolutely right about that. What sets Ubuntu apart is the quantity of unfixed bugs, the severity of same, and a basic decision by Canonical that fixing within a release cycle isn't important. That does NOT describe most other Linux distributions. If there are any others which essentially do the same things Ubuntu does vis a vis bugs then the same criticism applies equally to them. So, no, all distros are not the same in this regard, the problems are not the same and they all wouldn't be equally poor as a standard bearer.

Perhaps a better question is this: do we even need a standard bearer? If we believe choice in operating system is as valuable as choice in the purchase of a car or a box of breakfast cereal then why not tout the choices as a strength?

That's a tough criticism to quantify. How do you measure the difference between high and low impact objectively? I haven't seen a criteria set forth that we could use across different bug trackers to see how one distro is doing compared to another in an objective way.

-jef


One criteria would be the number of users impacted. It's clear that the Intel video driver issue was very high impact. The problem I experienced with udev and/or devicekit clearly has a lower impact.


There will always be high notoriety bug reports. I don't think you can point to two reports with extremely high assumed impact and derive a set of conclusions about policy. I'm surely not going to throw stones...we all have our own head-slapping migraine inducing notoriety bugs.

The question is, how do we quantify "number of users impacted" in a systematic way across many open hardware and software issues? I don't have an answer to that. And until I have an answer to that question about methodology, I've no way to take an objective look at our historic performance inside Fedora, let alone a comparative look to measure performance relative to anther project.

I would ask you to refrain from making comparative statements about performance unless you can articulate analysis methodology to produce your conclusions that other people can re-verify. Please by all means get passionate about individual deficiencies, but don't fall into the trap of speaking about overall comparative bug handling performance without having a way to do a data driven analysis which backs up the conclusions you are drawing.

-jef


"
There will always be high notoriety bug reports. I don't think you can point to two reports with extremely high assumed impact and derive a set of conclusions about policy. I'm surely not going to throw stones...we all have our own head-slapping migraine inducing notoriety bugs."
You cannot assume the Intel driver bug has no notoriety and might have affected a small number of users, as every sort of "net""books" have an Intel chipset!
These "net""books" are considered as being the growing part of the PC market , mostly in emerging countries; 30% of them are linux ones in these countries (not in Europe: people prefer Windows and they would buy a cheaper linux netbook if it was much better than a XP one!): they are meant for beginners, some of them having no (or lousy) IT connections, and not natively OmericonEnglish speakers -which makes access to the nice_helpful_vibrating fora a joke when it comes to fixing)
And this bug was -under Mandriva- and remains -under UBUlinux very annoying : clicking a a menu took seconds : how could the pour souls bug report in such conditions (the trusworthy XP????).


"The question is, how do we quantify "number of users impacted" in a systematic way across many open hardware and software issues? "
If one accepts to give equal weight to every bug (knowing there is at least a terrible exception), another quantification can/could be based on the automatical calculation of the mean difference between :
* the time the bug was fixed
minus
* the time it was detected would be trivial (10 lines of R) if these data were
easily accessible (they seem cryptic, in free format, with IT links to navigate in) and
legally usable!

list of "fix committed", "fix relased", and "fixes in progress" in Karmic: http://tinyurl.com/y6lgfkx

Except for one fix released(" 359392 [i965] X freezes starting on April 3rd" -how did the pour soul bug report, with a frozen screen?) I do not see how I can have an estimation of the time people had to wait to have their bug fixed :
MTTR it is an important indicator evey one can understand : if a garage taxes 5 minutes to repair your brakes, it is much better than taking 5 hours.

Caitlin, I'm a regular user of both Ubuntu 8.04 LTS Desktop & Server and 9.10 Desktop and am very happy with all 3. With them I run a small business and have NO problems. However, I don't modify the out-of-the-box hardware and I stay with the official software repositories because they meet my software needs.

However, once beyond the regular hardware and software, watch out for breakage! And this applies to any Linux distro any time. If I had a new computer that wouldn't run Ubuntu successfully, I'd return it to the manufacturer for a refund or replacement after trying a clean reinstall of Ubuntu.

After reading your complaints about Ubuntu and preferring something else, I'd suggest that you STAY with those something else distros if they work better for you.

Caitlyn,

Hold your ground.

The success of 'Linux' hinges on at least one commercial Distro breaking into the 'minds' of the masses as being as simple and easy to use, but strongly differentiated by 'safe'.

Every day, I see Zero-Day reports for Microsoft Windows. It is a product repleat with security defects which has spawned an industry of 'security experts' who live off of supporting a defective operating system.

Your experience with printing issues exemplifies an issue which will remain--not all defects can be fixed by Canonical as Ubuntu is comprised of 'moving parts' coming from upstream committers who march to a different drum.

Mark Shuttleworth has brought up the cadence issue more than once and continues to seek 'cooperation' with Debian and other Distros to simplify the patch process.

But I feel personally that Ubuntu Linux has the best chance of succeeding on the Desktop as being percieved as being the top contender for alternative Linux Distros.

I hope you get your printer issue resolved.

Dietrich T. Schmitz
Linux Advocate

 

I have had both very good and not so good experiences with Ubuntu and Launchpad. My guess is that the criticism has considerable justification and that Canonical should listen and see what can be done to respond to it.

Nevertheless, as maddening as some of the annoyances have been, broadly I agree completely with Avery Jenkins.

It's a good community. Help is available. And it's getting better all the time.

I agree, Ubuntu has helped pave the road for linux in many aspects.
But in my opinion, their last decent version was 9.04. From that I've seen, they botched hardware support in 9.10 some how, and some don't like what they see in the upcoming 10.04. I think a 1 year release cycle would be better for them, looks like the cut corners to make 6 months.
I myself have been working on getting away from ubuntu and ubuntu based distros, and going to it's parent Debian. As pointed out in the other article, other smaller distros are doing the same. CrunchBang is currently in Alpha 1 of their debian version which looks very promising. WattOS was looking at debian, but haven't seen anymore on it. I look forward to Mint going Debian, it is a nice looking distro, with some of there own config tools and such, felt like they were trying to fill some of the gaps in ubuntu, but ubuntu made it feel slow & bloated when I tried to use it.

I agree that Mandriva is a good distro for someone new to Linux. I've used it on occasion and have had few problems that couldn't be easily solved consulting the forum. The thing that drove me back to Ubuntu was the dreadful software updater. It is a slow, convoluted, aggravating mess that is in serious need of some work.

Caitlyn,

I've worked my way through both Ubuntu blog posts and the zillions of comments to both.

Above all, I have found the commentary around Ubuntu fascinating since I have almost no direct experience with it.

One thing that I have learned from the comments from the Ubuntu faithful is how many have failed to get your primary point. Forget entirely those who came away with the idea that you were saying, "Ubuntu sucks." I lost count of how many times people dismissed your criticism as being unhappy with your printer failing to work. Few mention that this was a regression, almost none address your main criticism that Ubuntu fails to fix bugs until the next release.

Why are so many ignoring the original posts and ignoring your patient, multiple restatements of the original point? I think that goes back to Penguin Pete's blog about not being able to hack people: you can make a useful observation and many will ignore the main point. You can correct them and they'll ignore the correction.

It must be difficult to learn much with that approach. For me, I'll say thanks Caitlyn for the insights and to all, thanks for the commentary.

Caitlyn is not broaching some new topic here. Take a quick search over at the Ubuntu Forums and you will see many dozens page posts about some critical regressions from release to release. The main culprits tend to be Audio (ALSA, PulseAudio), Graphics (Nvidia/ATi), Printers (somewhat less posted about), and Multimedia issues in general. Here is the rub though - these issues are 95% of the time not specific to Ubuntu, but affect systems or subsystems common to all Linux distro's. Ubuntu has the mindshare and the bulk of the "new users" though, so the complaints are far more visible over there than most others. These PulseAudio issues are nothing new to Fedora, as they pioneered it, but you typically wont see posts like, "My Audio wont work no more!!!! FIX IT NOW!!!!", like you regularly see over in *buntu land.

Ubuntu does have some bugs specific to them however few, and they sometimes are slow to fix them, if ever. That is the issue, imo.

Dear Caitlyn,

I don't mean to dig up the past or anything, but I couldn't help myself after reading this article. I looked up your name to find some more of your work and stumbled upon your blog. I read an interesting opinion of yours published here:

http://ever-increasing-entropy.blogspot.com/2009/08/end-of-centos-netbook-experiment.html

To refresh your memory, you argue that CentOS is not suitable for netbooks. A valid argument, certainly nothing controversial there, but then you take the argument to the next level. You suggest that CentOS isn't suitable for servers.

After spending the majority of your blog defending your claim that CentOS should not be used on netbooks (which I think most people understand and would probably agree with) you then briefly explain why it should be ousted from the server market (by briefly, I mean one paragraph). You said something to the effect that the CentOS team is tardy on their updates, citing a critical Firefox update that arrived a week later than the upstream update.

Before you condemn me to the isle of irrelavancy, let me point out the striking parallel between the CentOS blog and this article. You start by making an outrageous claim, cite one problem that you have run into, then spend the rest of the article giving your opinions.

In the CentOS case it goes like this:

1. CentOS doesn't work well on my netbook.

2. CentOS did not release a firefox update in a timely manner, therefore

3. CentOS should not be used on servers.

In the Ubuntu case it goes something like this:

1. Ubuntu is broken

2. I tried to use my printer with Ubuntu and it didn't work, therefore,

3. Ubuntu should not be the standard bearer for Linux

Statements like "Ubuntu should not be the standard bearer for Linux" and "CentOS should not be used on servers" should be backed up by more than one or two problems you've ran into. I realize that you've spent some time citing more problems and backing up your argument in your responses to the readers here, but that sort of thing needs to be in your article proper, not just an afterthought to deal with disgruntled readers.

The point I'm trying to make is that this type of article most certainly belongs in a blog or on facebook (as your CentOS opinion was), a place where opinions are welcome and facts do not need to be placed front and center. Instead, you post this piece in the guise of "news", which is misleading and probably the source of the amazing amount of feedback you have received.

Wow, you didn't read or grasp what I said about CentOS at all. What I said about servers had NOTHING to do with my netbook experience. At the time CentOS was going through some well documented internal turmoil. When I say well documented, I mean well documented on the CentOS website. They had, at the time, been months late with a whole slew of updates (pretty much all the security patches) which the developers admitted and said, in response to my article for O'Reilly, that they had to do better. I notice you failed to link the O.Reilly article. Let me help you out here: http://broadcast.oreilly.com/2009/08/the-future-of-centos-and-crite.html

So...you've completely misrepresented what I wrote about CentOS. The entire reason I recommended against CentOS on servers at that time was that the management of a small, community oriented distro was in turmoil and security patches weren't being released on a timely basis. The relationship between that and my netbook is precisely zero, as in none.

You have equally misrepresented what I said about Ubuntu as Don has tried to explain to you. Let's try this again: Ubuntu has a history of major breakage and regressions since Feisty Fawn. I provided links to articles by Akkana Peck, Chris Smart, Andrew Wyatt and Carla Schroder written over a two year period to back that pup. They, in turn, provided direct quotes from Canonical acknowledging that serious bugs are not fixed during release cycles as a matter of policy.

The regression in udev or devicekit that impacts my printer and, as it turns out, also breaks support for a huge list of printers that use the foo2zjs driver including various models by HP, Minolta and Konica. As you see I haven't stopped working on it and the more I learn the more I realize that this isn't a minor thing affecting a few printers but rather a significant regression. Even so, it isn't the basis of my complaint about Ubuntu. My article said "a case in point". In English that means this is only one example of many. That's the key concept you either deliberately ignore or fail to grasp.

Once again, there have been regressions that impacted nearly every Intel graphics chipset produced over the past fiive or six years, a regression that killed external audio output for anyone with an incredibly common intel_sda audio chipset, a released version of Xubuntu Feisty with an utterly non-functional desktop on the live CD (you'd think the most basic QA would catch that), wifi problems causing networkmanager to lock up and/or drop connections, and a list of additional serious issues that would take many pages to cover. Do you know what all these problems have in common? They weren't fixed during the release cycle despite a minimum of 18 months of supposed support during each release.

Then, of course, you go on to attack my writing based on a completely false set of premises about it. Nice. Really nice.

Oh, and I do mean to dig up the past here. CentOS developers solved their issues and now release much more timely patches. They acknowledged a problem and fixed it. I said if they maintained a decent patch record over the course of a full year I would recommend it again. They have four months to go and I have very little doubt that I will be writing a very positive follow up article this summer. Canonical, on the other hand, has not cleaned up it's act despite the complaints of users and a number of prominent writers in the Linux community. CentOS is worlds ahead of Ubuntu in that regard.

@Deano,

Wow. Amazing. You come right along and prove my point about not seeming to be able to grasp the point Caitlyn has made over and over in the blog posting and her replies to comments like yours.

Deano, Deano, Deano: Regarding Ubuntu, it isn't the issue of her printer not working. Stop and reread that--I'll write it again: it isn't that the printer didn't work. It isn't even that Ubuntu has some bugs with each release. Keep with me here Deano: It is because Ubuntu fails to fix the bugs in a timely fashion. Other distros do, Ubuntu doesn't is her point.

Is her point false? I don't know. I use Fedora and don't use Ubuntu. But you didn't address her point. You set up a straw man and attacked that. That's not useful. Try to do better.

In my experience there has been no distro that is what I would call stable or reliable, certainly none of them have been good enough for the masses, the target audience of Ubuntu. Ubuntu is still probably marginally the best, but the margin is small. If you are a an ordinary user who wants a very reliable, stable operating system, one that simply works, then don't bother with Linux.
I fail to understand the need for Ubuntu to have a six month release schedule. Whats wrong with a year? And why does it have to be done to a particular date, rather than released when its ready? Do we really want to have to reinstall our operating system every six months? Progress is nice, but stability is nice too.

I support Linux in the enterprise (business, government) for a living and have done for 12 years. There are some Linux distributions that are second to none, as in the best OS possible, in terms of stability and reliability, with uptime measured in years. I cannot disagree with your assertion to the contrary strongly enough. Examples include Red Hat Enterprise Linux and free clones (CentOS, Scientific Linux) and Slackware. For the desktop I have found a number of Slackware derivatives that aim at being easy to use such as Vector Linux and SalixOS are rock solid stable as well. An ordinary user most certainly can have a stable operating system that simply works with Linux.

@Caitlyn

I can't help but continue to read the comments here.

The level of idiocy from some of these posters is like crack. hah

Some commenters are either incapable of understanding what you wrote or are intentionally choosing to ignore what you have written, and are attacking you for speaking down about their Ubuntu religion.

There seems to be a pattern to the lack of comprehension though that leads me to believe that the majority of commenters either have zero real-world experience, are still in grade school, or are intentionally deflecting the issues because they fear that it will hurt Ubuntu.

I would say that those of us with a deeper understanding .. that have been in the "trenches" as long as we have get it though.

Unfortunately, the commenters attacking you for talking about problems are only hurting Ubuntu even more. It's a sad state really, having to publicize issues to get them fixed and then having to deal with all of the people that can't or won't see the bigger picture.

I remember when running Linux on my desktop was a fun experience. now? .. not so much.

I'm glad to see that there are others reading that see the big picture though. I have to wonder what the folks that don't comment are thinking. Do they agree, do they disagree?

Well said Fewt, well said. :)

Obvoiusly I agree with your take on this. Most of the comments during the first day after the first article was published were excellent, even if they disagreed with me sharply. Sadly it has deteriorated from there into what I wrote about last November: http://broadcast.oreilly.com/2009/11/the-problem-with-the-linux-com.html

Unfortunately, it always does. :(

This is such a sad comment you make; it leads me to believe you came here just to reinforce your sense of self.

Caitlyn, I think rather than joining this little self congratulation circle the three of you have going here, you should make a serious attempt to understand why you are having such difficulty getting your point heard.

It is, in the main, a useful and valid point.

If you think the reasons are what fewt said above, I think you are selling yourself and the people who have come here short.

The author's statement concerned telling the difference between the handling of high impact versus low impact bugs. The thesis is one organization is clearly doing better than another overall..holding up two examples of high impact bugs that affected one distro but not the other.

I don't think that's a fair analysis. Even though its in my benefit to agree with the author's conclusions, I can't, because I don't have the data to back up that sort of broad finding. The ends don't justify the means.

As a Fedora contributor, I care as much if not more about reusable analysis methodology than I do about knowing how well Fedora is doing in comparison to another project. And so far I'm not seeing an analysis methodology that I can take back and use inside Fedora to trend our own performance to help us optimize our own processes for the benefit of our users and contributors. And without that sort of internal project analysis as a baseline, how do we draw any comparative conclusions of our bug handling versus another distro?

I'm not arguing that the Intel graphics problem was not a high impact bug. I'm saying its not clear to me that its not an outlier in terms of overall responsiveness to high impact bugs by Ubuntu. I want a fair analysis methodology that we can all use that allows us to visualize the response handling of high impact bugs. If you can't articulate how we systematically define high impact, then well, we can't even get started with a critical objective look. Critical opinion not backed by a data-centric view its too easily brushed aside as biased opinion or a hit job and ends up creating more heat than constructive change.

If you weight all bugs equally you are ignoring impact...which is central to the arguments the author is making. Ideally high impact bugs should take priority and get fixed faster than low impact bugs...that's the argument being presented. An equal weighting of all bugs doesn't give us a systematic analysis tool to address the issue of impact response priority.

-jef

"If you weight all bugs equally you are ignoring impact."
I agree, and I know what I (would) do, deliberately, but as impact cannot be objectively measured, maybe it is (and will remain) the least bad way of measuring and the only starting point I could imagine -I am old and silly- ...

As far as now, an automatic measure (to average or to show an histogram) of the time before repairing (or not repairing) can suffer from biases (some bugs might be hidden as features) but would be feasible and easy (10 lines of R; 30 lines of (g)awk )
if the structure of launchpad databases(and others, maybe, but I stopped there) had the time the bug was detected and the time it was fixed (or closed as unfixable/upstream) filled or accessible.

I believe that detecting an issue and being critical, even if technically one cannot -yet- back up this criticism up with objective figures can, on the long term, make users more aware of what they buy (or download/try to install and fix : time and frustration have some economical value which may vary whether one is a buddhist or a stressed atheist, or whether one is in a poor or a rich country : variations and the timely impossibility to measure something are not a proof of its non existence!)
I sometimes work on physical teledetected measures wich are known wit a 1/10 uncertainty (ad one tries to reduce it). But they are thought of as interesting and, if dealed with care, useful...

I have to agree with @Jef. I have to add that certain distros by virtue of their exposure to, and popularity with a large cross section of users will tend to have more bug reports. Looking at the above thread, with contributors from both distributions, I find it unfair to form conclusions based on statistical observations.
I had problems with Intel mode-setting driver from as far back as it was included, I believe Fedora 6, but there was always a way out with the proprietary driver. Perhaps what we need is an across the board basic standard that everything should work at install time, which is pretty much hit and miss as things are right now. i doubt that this will happen.
I have resigned myself to not expect too much, that with limited space on live CD installers there is little likelihood at install time that everything will work, especially exotic hardware.
Perhaps there should be a "paper cuts" approach to significant bugs for all the distros, not just Ubuntu.

"popularity with a large cross section of users will tend to have more bug reports."
This has nothing to do with the time to repair (it is just a time to detect a bug, or a probability of detection, and is more difficult to quantify than the mean time to repair, which could measured once a bug is detected if there are/were convenient databases).
And bug detection can be done, too, without users, by code review (IBM,RH,Suse -alph order- do)
I try to uncouple notions which are uncoupled in the everydays world -cars, toxical gases detection, etc..-

But its a different question than what the author is trying to talk about.

And I'm not convinced that you can completely decouple the rate of defect reporting from rate of defect fixing when there is finite manpower available to actually fix bugs. A larger collection of bugs can in turn mean a longer wait to see them fixed in the average even if everyone is doing an equally god job of dealing with high impact bugs quickly. In fact dealing with high impact bugs as a priority instead of easy fix bugs can actually make the average wait time longer across all reported bugs due to manpower constraints.

As many bugs fixed as fast as possible does not mean the high impact bugs get fixed as fast as possible. It's not as clear an analysis as you would hold up.

I don't think that will tell us anything useful. What matters in understanding performance is getting a handle on response prioritization. People have figured out a criteria for security response prioritization. Surely there's a way to a similar breakdown for impact. It will take real effort however to do that analysis.

-jef

"But its a different question than what the author is trying to talk about."

I agree it is a different question (or at least part of the topic) , but as Caitlyn Martin is writing about quality assurance among UBUlinux and others, it might be ***partly*** an answer (not a good one, as there are no ad hoc databases!) to her issue (giving some numbers instead of 'more often than not', and some statistics instead of feelings, though feelings can trigger some more quantitative approaches) :

"(..) I still feel that the downloaded Ubuntu offerings more often than not have been substandard when compared to other distributions. I am not going to produce a list of serious bugs over the past five years. "

Part of this quality assurance, from an user point of view (and Ubingtu claims to be a linux for the masses : neither Fedora, nor Scientific, nor .... do) is the
time spent/(wasted if it is not fixed) by a
poor user, who crawled (especially with a broken/frozen/slooow terminal) to bug report and
hopes it gets fixed : it this time is too long, whatever the impact is , he might get upset and not recommand linux to his friends (I suppose if his friends are would-be Fedora/Gentoo/Linux-from-Scratch happy users, they wonot listen to his recommendations and perhaps lighten this poor souls sufferings; but are they that many?).

Giving equal weight to any bugs is just assuming the time waiting to have one's leisure PC (not a server, not a production oriented desktop/workstation but a PC to play on) is what understands an ordinary user, unaware of safety issues and of the difficulty of fixing, say, a new KDE (that needs an idea of the complexity of software 80% of the people do not have, IMO!).


"In fact dealing with high impact bugs as a priority instead of easy fix bugs can actually make the average wait time longer across all reported bugs due to manpower constraints. "
You are right, but starting with a quite simple measure and refining seems to me better than not measuring /quantifying (even with a 1 .. 10 uncertainty in some geophysical measures 30 years ago, now 1...2 : if one had not started, if would have remained 1..10 uncertainties!)

BTW : I was very surprised, from bugzilla's histories, that Fedora heroically managed to fix most of KED-4 issues which aroused in 2008 : there were twice less unsolved bugs in 2008 than in 2007, though their courageous tests (which were useful for evey linux using KDE ) might have impacted a too coarse measure of quality!)

An indicator like the time to repair divided by the number of devs/fixers would be fairer to some GNUlinux distributions (and useful if one wants to know if a policy is good), but not satisfiying from a consumer point of view!

Okaaay ... it took a minute for the light bulb to come on. Are you saying Ubuntu sucks just because! My not so well framed argument is better explained with an analogy. The cave man, back in the day, would only have had dinner after running off to the jungle and catching his dinner. The progression of things would have it that he eventually realised that he could have dinner without actually having to catch dinner by weaving a net or fabricating a knife or spear, trading his work for dinner.
Does this relate to Linux, probably not, but perhaps it just my way of entertaining you :-) Seriously its foss in its earliest incarnation and value was perceived even then as it is today as usefulness for purpose. The collective effort was much more worthwhile and valuable than individual toil. Specialization emerged and the rest is history. Today you have airbags, parachutes, toothpaste because someone suffered for it and cared to let someone fix the problem. That's one of life's realities.
What I was trying to explain is process. Users generally have service/product quality expectations. This entire thread speaks to the ideal that users are disappointed by unfixed bugs in a/multiple releases of their choice distro. Users perceive the value of free software they are using, as with commercial software, as the benefit exceeding the cost of obtaining/using it. On an individual level the impact or cost is the inability to use something or reduced quality that they actually incur from using it, in Caitlyn's case not being able to print is a bad thing, for someone else it might be display, sound or networking. So what's the solution: Quietly jumping ship, sending out bug reports, swearing off Ubuntu or whatnot. I think as users of free software we should send out bug reports when possible, and more importantly send out our hardware profile(s) so that developers know the install base thats affected by that once reported blocker bug.
To simplify what I am saying encourage users to keep on using your choice distro. The greater number of users encourages developers to fix problems they know will benefit someone. Developers will flee from unpopular projects. This is almost purely an unavoidable behavioral problem with foss.
The idea that problems can be fixed without users is odd; from back in the day cave man era and for thousands of years since we know that taking users out of the loop is the reason we are even talking about this stuff. User testing might have quickly dispensed of the idea of using square wheels to increase operational the efficiency of whatever square wheels might have been used for.
We live in an imperfect world with limited resources and way too many interests to satisfy. Canonicals business decisions intersect with reality in a way that I can best describe as tough love. I think I have used enough distros to know that problems often go unreported, therefore remain unresolved in many of the smaller distros.
What's clear in this discussion is that however people stand their ground there are multiple parts to this problem beyond getting bugs reported and fixed. It is apparent that we need to measure and report things, a business issue that will make the hidden problems apparent. I see allusions to objective data already being reported, in another part I see protests that interpretation and context of data does not lend credence to the topic being argued, and suggested solutions. Perhaps we should separate each issue and its resolution. After all this is software, and when have we ever been completely satisfied. We always want that extra bit that seems elusive, whatever keeps the bar of achievement high. For instance backporting fixes to years-old software in LTS vs. upgrading software in older releases match newer releases, and measure the fixes effect on user satisfaction measures. We might very well spawn new opensource business models while we are at it. This exercise will require the commitment that is not available by merely conscripting volunteers.

from your post :
300 lines unuseful and, then the looog awaited sentence :
"To simplify what I am saying"!

Just three simple questions, each two lines long (you see I do not want to hide something in a seamless mantra) :

* Should the expectations of the masses be pissed off (and they do not do anything about fora, etc).

* What can one think of a compagny which uses its free products to claim they are meant for the masses?

* What whill people who want to buy a GNU linux think from such attitude?

I am sure that this discussion was focused on the challenges of improving quality for all the users not attitude of commercial companies or motivation of contributors.

I don't mean to be condescending in tone in this post but I am just speaking to the questions posted. Since its an open forum anyone if free to add their opinion.

Q. Should the expectations of the masses be pissed off (and they do not do anything about fora, etc).
A. Did you not read my post! While you are obviously irritated by the length of the post you fail to recognize that F/OSS is organic in the sense that everybody's desires and expectations are incorporated. The limitation in size of the live CD necessitates tough love on the so called expectations of the people. That's why we have so many distributions and derivatives. Volunteers have a life have a life to get back to and often good projects are shelved in order for that to happen: in other words they owe you nothing. If you don't like it try closed sourced software and don't complain about what you already have in Free and open source software.

Q. What can one think of a "compagny" which uses its free products to claim they are meant for the masses?
A. Welcome to planet earth. This is the way of the world. Distributions are just that; a compilation of disparate pieces of software which many times have little work going into the core software functionality other than the compilation effort. The reasons behind the inclusion of software in a distro belong to whoever is distributing the software (help thy neighbor), not necessarily the actual author of the software. In spite of many philosophical arguments that everyone is your customer, the reality is that not everyone is (with few exceptions like churches, or the salvation army); anyone that thinks otherwise needs a serious reality check.

Q. What "whill" people who want to buy a GNU linux think from such attitude?
A. Nobody forces you to buy anything. "Gnu Linux" is free software, please read the license. The gist of free software is that if you don't like it make your own or otherwise do what you can to help yourself out of your situation. If you read about Richard Stallman you will recognise his philosophy and where he is coming from when he advocates free software. For more on that read more about Stallman's 4 freedoms at http://www.gnu.org/philosophy/free-sw.html.

I have also recently read perspectives on free and open source software. I just wonder that the statistical realities work against long term commitment to individual pieces of open source software. There shall always be new software coming up with new problems and the cycle shall continue.

"I am sure that this discussion was focused on the challenges of improving quality for all the users not attitude of commercial companies or motivation of contributors."
But if my boss wants/needs to buy a commercial linux, what will happen if one knows they introduce bugs/do not fix them in free versions (which are often used as tests before decions)


I don't mean to be condescending in tone in this post but I am just speaking to the questions posted. Since its an open forum anyone if free to add their opinion.

Q. Should the expectations of the masses be pissed off (and they do not do anything about fora, etc).
A. Did you not read my post! While you are obviously irritated by the length of the post you fail to recognize that F/OSS is organic in the sense that everybody's desires and expectations are incorporated. The limitation in size of the live CD necessitates tough love on the so called expectations of the people. That's why we have so many distributions and derivatives. Volunteers have a life have a life to get back to and often good projects are shelved in order for that to happen: in other words they owe you nothing. If you don't like it try closed sourced software and don't complain about what you already have in Free and open source software.

....QUEL cata plasme! (3000 lines skipped!of redundant generalities)
"
Will the link to the 4 freedom would have been enough,
though they are already known:
The looongest credos are the more booooring....
"
In the real world (this planet you seem to like), when someone has something to hide, he makes a text full of generalities :
perhaps cutting and pasting the Encyclopedia Brittanica next time will contribute to hide the facts that :
a) Canonical , with poor ordinary users supports, might discredit Linux:

b) there exists simple ways of measuring the time wasted by users to get their software repaired (like for cars : brakes and painting fixes are treated exactly the same way: once a car arrives, security is not more an issue : these has been known for decades).

c) these simple ways are not available through UBUlinux databases (as some of my colleagues produce code, I know these indicators have some value)

D) Complicated pseudomethodologies , looooooooooooong credos, overredundant generalities are the best way of not doing anything, except some typo correcting (I have not an automatic spell checker for Omericoin)


If I understand you correctly I presume that we are talking about the same thing all along. Thank you. Please elaborate on (b) measures of impact to users, and (c) how contributors can determine which bugs to fix, and backport fixes to prior releases.

Before we can get to your point (a) and (d) we have to get people like yourself and colleagues that produce code to contribute to improving process (or methods), a management issue that comes up even in smaller projects. We have already determined, and @Caitlyn has said it over again, that writing of code does not seem to be the problem.

For instance at this link on Fedora, "Why My Bug isn’t Being Fixed:" http://rambleon.usebox.net/post/533515447/why-my-bug-isnt-being-fixed, highlights just the reporting of the problem. We also need to know how to channel fixes in the face of whatever limitations we may collectively have.
Also look at, "Moral obligations of Free Software authors:" http://changelog.complete.org/archives/1463-moral-obligations-of-free-software-authors

All I say is that all these comments are relevant. We should Identify the real problems first. Chronically unfixed bugs may present to be a symptom of the real problems. You fix the symptoms as you should, but also keep an eye out, and commit resources to fixing the real problems.

The take away from reading all the passionate comments, including your own, is that all the good thoughts need to go back to the Ubuntu project in order to be of any use.

@Fellini : thanks for being concise this time.

"(b) measures of impact to users",
I suppose very simple (though unfair) measures of the result of bug fixing are enough :
a car user when he comes to have his brakes fixed/ his painting renewed is no more concerned about safety (and whether it is difficult to fix or not is not his problem!neither whether pieces are to be brought from upstream or not) , but just of

* how much he has to pay (in FLOSS, it is not relevant at all! that makes things simpler)

* how long it will take (and even seemingly "no impact" issues are relevant: if an expert in security is color blind, and needs curves, having easy to read curves, tested with some options of the GIMP or with R's dichromat package, is more serious than it seems at first look, it is easier to fix than a frozen screen issue : both have the same impact in my case).

Therefore, time to fix, as seen from an user point of view, seems essential.

What surprises me is that it is difficult to automatically computed in a distribution which claims to be ignorant masses - oriented.... (I suppose LFS or gentoo do not have the same audience!)


(c) how contributors can determine which bugs to fix, and backport fixes to prior releases.

Well , freely... (and performances can be evaluated a posteriori); but ordinary users should have some accesses to/knowledge of the consequences of their free choices!

I am just trying to find a quantization of the users short term interest (ex. Fedora putting KDE4.0 might be handicapped by this measure: I know it is, and, in the long term, other measures are of interest: but real life objects can be caracterised by their composition in chemical toxics, volume, weight and many other things : if you want to have them carried, weight is the only relevant measure in short term.)


I have to agree with you 100%. I kicked Windows to the curb in 1998 and GNU/Linux has been my main desktop since. Even in most of the place I've worked since then, I've managed to use GNU/Linux for as much of my work as was possible.

I started with RedHat, and later used SuSE, Mandrake/Mandriva, Gentoo, Debian, Arch, Zenwalk, and many others. I started using Ubuntu 5 years ago and, like you, I have really wanted Ubuntu to succeed. But also like you I have had an absolutely terrible time with it. It seems to me that I can a version of Ubuntu working then it works fine. But rarely have I been able to move to the next version without horrible difficulty.

Late in 2007 I bought one of the Dell desktop machines with Ubuntu pre-installed. However, there was a newer version of Ubuntu out and I wanted KDE so I immediately installed Kubuntu-7.10 and it worked well. However, when I tried to upgrade to Kubuntu-8.04 it was a horrible nightmare. I also tried Ubuntu-8.04 and Xubuntu-8.04 and none of them would install either. I wasted hours and hours trying and finally gave up and re-installed 7.10. When 8.10 came out I spent a few hours and couldn't get it to work either. I finally switched to gNewSense on that machine and it has worked pretty well.

I had the same problem with my laptop, only with different versions. I had a really old version of Kubuntu on it (I think it was 6.10) but when the next version came out, it was no dice. I tried almost every version after that on it with no luck. Last year I tried to install 9.04 on it and had all sorts of problems. I literally spent hours fooling around to get the video working, which I finally did. But the Ethernet just wouldn't work. I messed with it for weeks, and never could get it to work. I even had the Admin where I work look at it and he couldn't get it to work either. I finally gave up and installed Ubuntu-8.10 and wham everything worked perfectly out of the box. I just tried to install 9.10 on it a few days ago and it was a non-starter, literally. The graphics sat there and flashed off and on.

As I said I really like Ubuntu and I really would like it to succeed. But my experience with upgrading has been a nightmare. It is really frustrating that you can have it installed on a machine and working fine. But when you try to upgrade it just won't work, even after spending hours and hours working on it. I have filed bug reports with Ubuntu, and have been happy that they do seem to try to look at the bugs. But they have so many bugs, newer versions are usually out by the time they get to them and then it is like: what is the point?

I think I am going to try some of the distros that you have mentioned here, that I haven't tried before and see if there are any of them that float my boat. Like you said that is one of the great things about Linux, you don't like one flavor there are plenty of others to try.

I have to agree with you 100%. I kicked Windows to the curb in 1998 and GNU/Linux has been my main desktop since. Even in most of the place I've worked since then, I've managed to use GNU/Linux for as much of my work as was possible.

I started with RedHat, and later used SuSE, Mandrake/Mandriva, Gentoo, Debian, Arch, Zenwalk, and many others. I started using Ubuntu 5 years ago and, like you, I have really wanted Ubuntu to succeed. But also like you I have had an absolutely terrible time with it. It seems to me that I can a version of Ubuntu working then it works fine. But rarely have I been able to move to the next version without horrible difficulty.

Late in 2007 I bought one of the Dell desktop machines with Ubuntu pre-installed. However, there was a newer version of Ubuntu out and I wanted KDE so I immediately installed Kubuntu-7.10 and it worked well. However, when I tried to upgrade to Kubuntu-8.04 it was a horrible nightmare. I also tried Ubuntu-8.04 and Xubuntu-8.04 and none of them would install either. I wasted hours and hours trying and finally gave up and re-installed 7.10. When 8.10 came out I spent a few hours and couldn't get it to work either. I finally switched to gNewSense on that machine and it has worked pretty well.

I had the same problem with my laptop, only with different versions. I had a really old version of Kubuntu on it (I think it was 6.10) but when the next version came out, it was no dice. I tried almost every version after that on it with no luck. Last year I tried to install 9.04 on it and had all sorts of problems. I literally spent hours fooling around to get the video working, which I finally did. But the Ethernet just wouldn't work. I messed with it for weeks, and never could get it to work. I even had the Admin where I work look at it and he couldn't get it to work either. I finally gave up and installed Ubuntu-8.10 and wham everything worked perfectly out of the box. I just tried to install 9.10 on it a few days ago and it was a non-starter, literally. The graphics sat there and flashed off and on.

As I said I really like Ubuntu and I really would like it to succeed. But my experience with upgrading has been a nightmare. It is really frustrating that you can have it installed on a machine and working fine. But when you try to upgrade it just won't work, even after spending hours and hours working on it. I have filed bug reports with Ubuntu, and have been happy that they do seem to try to look at the bugs. But they have so many bugs, newer versions are usually out by the time they get to them and then it is like: what is the point?

I think I am going to try some of the distros that you have mentioned here, that I haven't tried before and see if there are any of them that float my boat. Like you said that is one of the great things about Linux, you don't like one flavor there are plenty of others to try.

I agree that Ubuntu has problems, but why do you feel safer using a distro provided by a company compared to a team of volunteers? That seems extremely hypocritical to me when Debian itself is just a 'team of volunteers' and they are the ones responsible for 99% of what Ubuntu is (a Debian-derived distro). Personally my best Linux experience has been with sidux, which is also run by a small team of volunteers. Don't knock community efforts!

I don't knock community efforts. However, it takes money in most cases to keep something going. The Debian Foundation handles the fundraising for Debian, the distro, so claiming there is no institutional support isn't quite true. A small team of volunteers can never handle the demands of business or enterprise customers. I never said a company has to back a distro, but to me corporate, institutional or government backing is essential for a distro to be more than a toy for hobbyists and be accepted in the business world. That business world, in turn, employs all the leading Open Source developers who write the upstream code both Debian and sidux depend on.

Ubuntu users are great, the more the better. I would recommend newcomers to Linux, use Fedora. Like Windows 7, Fedora 12 is leading edge Linux. Most Windows users buy the most current version. They don't buy an "easier" Windows Lite version. Those that don't want as many updates or don't need to be current can stay one version behind the latest version of Fedora. Ubuntu is suppose to be easier to use, but for me Fedora seems easier. I think Fedora is easier to install / update software. Some of yesterday's issues with Fedora are just that yesterday's issues. For me Fedora just works. In the end use what ever Linuxdistro you prefer. That's why we like Linux.
.
Get manufacturers to list on there products that they work with Linux. Where is our little sticker that say Linux Compatible.
.
Fedora users working together with Ubuntu-GNU/Linux users.
.
(using: HP Laptop, HP All in one printer)

"While I have no specific complaints about any of these I simply, perhaps because of my background in the business world, prefer a distribution with a company behind it rather than an individual developer or a small team of volunteers."

Ubuntu may be backed by Canonical but virtually all of the packages are lifted straight from Debian testing at the start of a new release. As far as I am aware, Debian is not backed by a company yet Debian servers are found in a huge variety of commercial ventures around the globe where it is renowned for its stability and hugely trusted.

Most of the problems with Ubuntu are fairly trivial, and having used it since 2006 I can say it is vastly improved. I've been running Lucid since one of the early Alphas and it is a superb OS of which I think Ubuntu should and will be proud.

Debian has a foundation backing it and, AFAIK, they aren't terribly particular about who joins or contributes to that foundation. Debian has other problems which are beyond the scope of this discussion.

How can you call regressions that affect almost ever user of one of the most popular graphics chipsets "fairly trivial"? Popular sound cards? The whole pulseaudio debacle? This is clearly a new definition of trivial with which I am not acquainted.

To those who want to switch to Linux, I *NEVER* recommend the most popular distros. One of the reasons is that they usually come with poor programs, very poor in the case of the *buntu family. You can install what you want of course but when you know nothing about Linux...

Other reasons: their concepts are hardly in the average and the way they work, bugs included, makes they're not the easiest systems for beginners.

Oncle Jean

Your article doesn't mention one specific issue that you have with ubuntu - in fact, you do nothing but mention positives, and positives with other distros, but I can't figure out what Canonical could do different, other than vague references to the corporate atmosphere or spending more time on qa. What are some specifics? What can the corporate atmosphere do for you that would make you feel better. I feel like you wrote this article to hear yourself talk. I worked for Novell with opensuse, and I can tell you - that company doesn't give a rats ass about quality. Ubuntu works, whereever, whenever, and on whatever hardware you put it on. OpenSuse works as long as you don't use anything that isn't included in your base red hat distro - as soon as you try to branch out, its thousands of bugs, custom installers, and custom packages just to put you on par with one of the other distros.
I have my own list of things ubuntu could do better (ditching the 1999 gnome ui that red hat has been using in their enterprise distro)- other than that, I feel like I have feature parity with all good distros, and I can even go beyond with all the apps available. Red hat hasn't updated RHEL (except bugs) since 2000, and the age of the distro shows. Even the next release is over 6 years behind the current fedora release.

Oh my... you've got so many things wrong I don't know where to begin.

". OpenSuse works as long as you don't use anything that isn't included in your base red hat distro" Huh? openSUSE is a Novell product. The distribution was originally based on Slackware. SUSE (open or Enterprise) have never had anything to do with Red Hat. They are competitors. I would also dispute your claim that Novell doesn't care about quality since lack of quality always translates to lack of customers. I guarantee you that Novell cares about making money and that means keeping their enterprise customers happy.

"Red hat hasn't updated RHEL (except bugs) since 2000" I don't know what you've been smoking but it must be pretty strong stuff. No changes between Red Hat Linux 6.2 Enterprise Edition (a/k/a RHEL 1) released in 2000 and Red Hat Enterprise Linux 5, released in 2007? There were huge changes in RHEL 2, 3, 4 and 5 at every possible level: kernel, libraries. drivers, applications.... you name it. In addition the dot releases, such as the newly released RHEL 5.5, do include additional hardware support which is backported into the RHEL kernel with the required drivers added. You claim that RHEL hasn't had substantial updates except bugfixes since 2000 is completely at odds with reality.

If you read my articles and my replies to comments your statement that "Ubuntu works, whereever, whenever, and on whatever hardware you put it on. " is absolute nonsense. Tell the folks who had Intel graphics chipsets that Jaunty worked for them. Tell the people who had intel_sda audio chipsets that Intrepid just worked for them. Tell anyone with a printer affected by the regression I've repeatedly described that Karmic just works for them. Once again, your comment defies reality.

If you can't find what I said negative about Canonical in my two articles or didn't see my suggestions for improvement then you either didn't read them at all of have a severe reading comprehension problem. If you worked for Novell I can see why they got rid of you. You are either making up your own facts or history or are utterly incapable of separating truth from fiction.

I endorse Elder-Geek's comment in it's entirety.

I would add that including Mono anything in Ubuntu's distribution is a big mistake. Microsoft is biding its time. When it springs its trap, it will be the effective end of Canonical and Ubuntu. ... Here's a quote for you:

"They’ll get sort of addicted, and then we’ll somehow figure out how to collect sometime in the next decade." -- Bill Gates

I rest my case. ... In them mean time, I've changed distributions from Ubuntu to Fedora for now.

I just reread Elder-Geek's comments and he doesn't say anything at all about Mono. When it comes to Mono, I'm sorry but I can't agree. Let's say Microsoft tries some sort of patent infringement suit. What is to stop Canonical and Ubuntu from simply removing Mono from Ubuntu and and applications that depend on Mono as well?

Wow, Caitlyn. Have much free time besides taming the trolls?

Anyways I pretty much agree overall with what you've said. I gave Ubuntu a shot up through 8.04 and moved on after repeated hit and miss releases. It just wasn't dependable enough for any real world production, a.k.a. using it as an employee of a local web dev company.

In the end I ended up back in Slackware after leaving it around 10.1. I'm back in with Slackware 13 x64 and couldn't be happier.

Ubuntu didn't cut it, but is that a mark against Linux based systems across the board? Nope.

Thanks for the controversy!!!
And let me pose a few questions myself

1. Which is the best Live Distro?

From personal experience I have tried on and off since the early 1990's to successfully install (from 3,5 Disk or CD) Linux just like I installed Windows 3.1 till CD became the standard in the computer industry about the release of Windows 98. I had all sorts of notebooks and none worked right out of the box until after nearly 10 years of windows savvyness accumulated I tried several linux distros out of the box. Suse didn't cut it for me, Red Hat was too dusty and Debian worked on 1/3 of the computers that I had at my disposal.
Then I came to Ubuntu and got the CDs per snail mail. I put it in my pc and it worked! Since then I have become a penguin advocate for the rest of my friends and associates alike.
Do things work all the time? No. But I do not know of any OS that does.But I am glad to say that I can count on fixing things myself because I have one place to go in order to find an answer: help.ubuntu.com and that is what cuts the cheese for me (at least in the beginning). As an example: I got a new asus eepc and my wireless card did not work. I finally got the answer in the ubuntu forums and got my atheros wireless chip going...although it was a learning curve ----I went to just ONE forum and found an answer. Now several years later I kind of like looking back and regretting not have given more of my time to any linux distribution and "getting it to work" instead of countless nights of searching the "right" program for my M$ OS.
Note for the mac advocates: Please forget about Mac since I burned my fingers on a niche product called Atari...

Now I have become addicted to apt-get and aptitude in the cli. And this is the one point where my penguin colleagues in the midst of the M$ Admins do side with me is that the package manager is a kind of a religious-political decision. But even the RPM lovers agree with me that Debian got it right and now Red Hat and Co. have been trying to keep up with this great manager.
So until the RPM politicians come up with something to overcome my love for apt-get and aptitude I will stick to any Debian based distro.

Please do keep up writing and controversing. But be fair and admit that you are kind of stuck on the RPM side.

2. Which is the best Debian Based Desktop and why?

Why don't you write about which Debian Distro can make a good compromise between Debian and Ubuntu in your opinion?

3. Best Debian Based Enterprise Distro?
Also tell us in your opinion which debian packed distro is the one to really challenge the RPM Supremacy of Red Hat and Suse (now Novell) in the business world.

Instead of rambling what one company does wrong, I'd rather see, reasons what does it take to become one of the 10 major Distributions in the distrowatch list. (And to that extent tell us why Ubuntu, has made it so fast to be one of them.)

Please be assured that I will be the first to read avidly your suggestions and gladly replace the OS (M$ Vista) on my mother's (Age 73) laptop (January 2010) for good.

Because no matter what happens one thing is for sure, if we do not unite, we will never overcome the M$s' Desktop share of over 90% of the last 20 years, despite the fact that Linux has made very many of the present gadgetry and the ubiquitous internet possible.

The end user does not care, what is and what's not, he just wants an operating system that works, despite the fact that the majority of pc users has gotten used to pcs not working for any reason, thanks M$s'OS for that. It's time to give them the comfort of a reliable system with the openness to choose the software that he wants to run on the hardware of his choice.

Let's follow the example of Firefox which has broken IE's dominance, at least in Europe, of the browser market and say: Linux users unite.

@AmourduNet:

may be you should (have) read the last two paragraphs of this post, where
- Mandriva (rpm -based),
- Mepis , Mint (both *.deb),
PClinuxOS (I do not know on what it is based),
Pardus (made their own package management, like many do) are recommanded by Caitlyn Martin;
Therefore, where is the rpm bias in this text?
"But be fair and admit that you are kind of stuck on the RPM side. "

I suppose, if I made a link between urpmi (a recursive package manager from Mandriva) and apt-get, I could type (at least basic) debian installation commands clis under a Mandriva; ditto for some other distributions...
"but even the RPM lovers agree with me that Debian got it right and now Red Hat and Co. have been trying to keep up with this great manager."

Have you got factual or based on some analysis arguments (IT links) to convince anyone that *.deb are exceptionally great? and that RH and Co are switching to *.deb??

The main issue with software in general is how long does it take to correct bugs (and are they corrected at all) and not package mangement tools (if tools are never used, or to ship fishy packages, they cannot be that useful...)

Canonical dosen't fix any bugs for the software it includes. They wait for upstream releases.

You're thinking about RedHat. They actually backport changes and fix bugs. Canonical has occasionally been known to release ppa patched software which corrected an issue. But in general it's just the recent beta from the upstream developers.

Wodim should be removed yet Canonical won't because it's in the Debian sources. They stay true to their upstream.

Selinux really isn't necessary as it's disabled by default on the desktop version of Ubuntu. But why they include it is because it's in the Debian sources. It'd be alot of work to remove.

Gnome has suffered majorly. 8.04's gnome was lightening compared to 9.10's molasis.

But again you won't find them fixing the kernel if it is. Got to wait for Linus to release a new patch or version.

Ext4 has been getting slower.

Honestly all the distributions have gotten slower because of the kernel's they use.

Slackware 12.2 runs like shet compared to 11.0/12.0 I noted the kernels used. 12.2 2.6.27 / 12.1 2.6.24 / 12.0 2.6.21 / 11.0 2.6.17?

SL5.4 is using the redhat 2.6.18 which is amazingly fast so what does redhat know that we don't? :D

gnight

Even when high priority, hardware detection bugs are likely to not be fixed for a while, for a simple reason: the (relevant) developer must have that piece of hardware.

A few people have posted in the comments that they have had trouble with Ubuntu and their hardware. I'm sorry that you have had bad luck, but in my experience, Ubuntu does an incredible job of properly detecting hardware, and having working drivers (moreso for internal things like WiFi & graphics cards than external things like printers). Of course I've found some devices that other distros do better with, but I've found that Ubuntu just does better with more devices. These discrepancies are of course due to the simple reason above.

Ubuntu is the only distro I have found that properly handles all the hardware in my laptop automatically (WiFi is generally the biggest problem, as of about 8 months ago, Debian offered non-functioning proprietary drivers). For this reason I have been using Ubuntu for the last few years; as a developer it's not totally weird that I destabilize my system. It's nice to know that even if the thing's not working, if something urgent comes up, I can have it back up with a fresh install in under half an hour.

However, it is for other reasons that I agree: Ubuntu is not going in the right direction. The on-system documentation is sub-par, man and info pages are either missing, incomplete, or out-of-date (the on-line, community docs are pretty high-quality, though). Less of an emphasis is being placed on FOSS: for most of the proprietary stuff they just tag the licenses as `unknown'. As long as I'm on a Debian-like system, I'd like vrms to be able to report my Flash Player. Several free-software packages bring along recommended proprietary packages without any warning, since they are tagged as `unknown'.

Perhaps I'll see if the latest FreeBSD, Debian, or plan9 support my hardware...

In my opinion, it started going downhill after 8.04.1.

With reference to both Luke Shumaker's comment & Caitlyn's original (previous) blog:

I tested Ubuntu 10.04 yesterday. Guess what? It doesn't recognise a Huawei E1550 USB modem. Now, I can understand that 9.04 didn't because of it being newer technology; but since 9.10 did recognise it, it seems a bit odd that it is incompatible with 10.04.

Does this mean that hardware more than 6 months old is effectively obsolete in Ubuntuland, until the 'bug' is fixed? I don't see this as a 'bug' at all; more like a case of neglecting to maintain driver/installer compatibility.

Another example is the z600cups/z600llpddk printer drivers that have worked for years in Ubuntu, & upon which many Lexmark & Dell printers have successfully relied upon to work with Ubuntu. Suddenly, with the 'upgrade' to 9.10, they no longer worked, because there were 'bugs' in the drivers. What, so suddenly? I think not!

Okay, the problem was sorted out, a couple of months down the line, & there's a page on Ubuntu's website with all the links & instructions; but is it not a problem that need not have arisen in the first place? (Since I didn't actually install 'Lucid Lynx', because of modem incompatibility, I can't say whether or not the z600 drivers work in it, although a couple of blogs assert that they can).

Does 'bleeding edge' necessarily mean an 'open wound' upon which 'bugs' will settle to multiply? Can Ubuntu realistically ever be expected to produce a genuinely stable edition? Does not such 'conservatism' run against the Linux grain (brush Tux the wrong way)?

Perhaps 'stable' needs to be redefined to mean 'My Linux system works just fine & I have no intention of upgrading (& would anyone with compatible hardware like an ISO snapshot)?'. Come to think of it, it's probably a very good idea to keep an ISO snapshot album. ;-}

I use Ubuntu for about six months now, not that I don't have prior Linux experience. I installed 9.04 version and at the very beginning, I had problems. Sound was choppy (still is, although not like before - minimizing a program makes sound skip for a fraction of a second... which is mildly annoying). When I upgraded to 9.10, X wouldn't start (had to drop ATI proprietary drivers for my Radeon9800Pro; although this card isn't supported any more, which I found out later, Synaptic still installed it), and compiz crashed the Gnome (which I solved by uninstalling and installing compiz a couple of times, then running compiz --replace... now things seem to work, but I switched to Metacity). But still, I think it's not that bad - you get to know how things work in Linux by trying to fix stuff that's not working. So, in a way, it's like breaking something to see what's inside and how it works. It's just that in Ubuntu, things break on their own. :)

I think Canonical is doing what all the other distributions just tried for many years. Offer a competitive product to Windows and Mac, without 'dirty games' . The ideals from 'Free Software' still remains.

Offcourse the 'quality problem' exists, and there is a long road to be traveled. But try to think before Ubuntu. Which Linux distribution provided what Ubuntu offers today?

I'm not saying Ubuntu don't have problems, every distributions have, but is relative. I'm saying that the only distribution that can offer this 'desktop-approach', with so many resources, so many diferent solutions with minimal user interference is Ubuntu.

Someone can have problems with wi-fi, with a specific hardware. Too bad!!! The beauty of 'free software' is you have options!! The distribution is not working for you, try another one...Complaints are so boring!

You can report this error for Ubuntu. And I'm sure you will have a response. Try to remember when you're using Linux, you are on the 'Free Software' world. It's another concept of relation between people and machines. Think about it.


PS: Fedora sucks!!!! argh!!!

"PS: Fedora sucks!!!! argh!!! "

Thanks for the rationally motivated opinion about the factual bugs of RH 's alpha versions. It perfectly illustrates the reasoning abilities of an UBU linux specialist....

"Offcourse the 'quality problem' exists, and there is a long road to be traveled. But try to think before Ubuntu. Which Linux distribution provided what Ubuntu offers today?"
Gentoo, Mandriva, Red Hat (clones), Slackware and many other, in terms of ease of installation, use or security (one could choose, or , with some luck, have everything at the same time).

"You can report this error for Ubuntu."

There were two already reported errors Caitlyn Martin quoted :
* one with a printer not cold plugging (but it did with every other linux she tried : THE FIX WAS KNOWN) : it was marked as "high priority" in UBUgs databases... and left unattanded (thet makes your sentence "And I'm sure you will have a response" very funny, as the answer is 'UBUlinux pissses you off!")
* the other was Intels driver for netbooks (they have a great audience), leading to very slow interactions. Slackware (derivative) avoided it, Fedora and Mandriva fixed it... and the responsive UBUlinux "offered this 'desktop approach " by not fixing it.... while the solution was known!!!

BTW, what makes me uneasy in this post are the recommandations for beginners, as they seem based on ease of installation :
Mandriva is very easy to install, but I know people who, whatever the ease of installation/ configuration, will remain paralyzed if they have a trivial configuration to do (they cannot decide, it is too mysterious). Preinstalled linuxen or live CD (images, with unetbootin, if they have no CD reader) can workaround this ease of configuration issue (Debian are not easy to install : the ordissimo -a preconfigured PC, with selected Debian application) seems to have success (in France) among old, computer ignorant people...

As a recent convert to Linux, I am often just completely demoralised at the learning curve. I am after something that just works, the content of my web page or document or powerpoint is what matters to me. Windows does a pretty good job of it,( so good, in fact that I keep wanting many of the features of windows to be there still).

Ubuntu does try very hard to accommodate what they call newbies, ie just normal users. And it is still the operating system of my choice having abandoned Vista earlier this year in great frustration after it refused to update itself.

I shall never get into any use of typed commands into terminals, point and click of a mouse is where the progress has got us so far so why need we go backwards? I shall never understand the architecture of Linux. I am a NEWBIE for life and proud of it. In the same way, I don't know how the chip in my SAAB maintains the operation of the car, I just drive it using the human interface.
Show me another Linux distribution that cares about me more than Ubuntu and I will use it, but so far I dare not change, despite any minor glitches you may find.

@Andy: My complaint against Ubuntu is that all too often it doesn't "just work". Most newcomers to Linux want precisely what you want. My point is that Ubuntu doesn't have just "minor glitches," to use your words. It has major, show stopping problems for a significant number of people that absolutely prevent it from working on their systems. Whether it's a whole class of printers which use a specific driver or most every Intel chipset on the planet, versions of Ubuntu not only didn't "just work", but often they didn't work at all out of the box for too many users. Other distributions fix such problems within a release cycle. Canonical, the people behind Ubuntu, have made a conscious choice not to do those fixes. I don't see how that is caring about newcomers.

Other distributions that "care more" include any and all Linux distributions aimed at the desktop that fix bugs. They are more responsive to the problems newcomers and experienced hands alike face when running into problems. I'll repeat my recommendation for newcomers: try Mandriva. It is definitely better than Ubuntu in this respect and is just as aimed at providing an easy to use desktop as Ubuntu is. Yes, the company is in financial trouble but they are in negotiations right now to be sold. New owners should solve the financial issues, at least for now, and in the meanwhile development on 2010.1 continues apace.

Your comment about Windows doing a "decent job" is off base. Yes, their user interface is not terrible, but the malware of all sorts make it a nightmare to use and support and eventually everyone suffers from it. Using Linux (as you do) or MacOS or FreeBSD mostly gets around those issues provided you do basic security and that can be automated. In the case of Ubuntu and Mandriva it is well automated out of the box.

Regarding your analogy to a car: You do need to learn how to drive a car and most people do want to know how to change a tire as well. Yes, I'd rather call AAA (or some other road service) but they aren't always available quickly. Knowing how to do the basics really helps. Learning a little about your computer makes the entire computing experience easier as well. You don't have to be the equivalent of a mechanic but saying you are proud of not knowing is ridiculous.

Oh, and in regard to the command line, professional Windows admins use it too and will tell you that even in Windows there are some things that can only be done from the command line or are worlds easier from the command line. Ditto MacOS. Using the command line isn't "going backwards." It is very much an essential part of every modern operating system. Whether you choose to avail yourself of it or limit yourself by ignoring it is entirely up to you. That's no different in Linux than in MacOS (which is UNIX under the hood) or Windows. I even have an excellent book on Windows shell scripting on my shelf. The idea that this is some archaic, backwards looking Linux thing is just not accurate at all.

Remember what drove you from Windows? The inability to upgrade in place, right? While that seems to work in the current version of Ubuntu there were versions in the not too distant past where that was broken as well. You've made clear that simply isn't acceptable to you. I am essentially agreeing with you on that point. It's not acceptable. More importantly, if it is broken it should be fixed.

Caitlyn: I believe you've finally hit the nail on the head:

'Whether it's a whole class of printers which use a specific driver or most every Intel chipset on the planet, versions of Ubuntu not only didn't "just work", but often they didn't work at all out of the box for too many users. Other distributions fix such problems within a release cycle. Canonical, the people behind Ubuntu, have made a conscious choice not to do those fixes. I don't see how that is caring about newcomers.' SPOT-ON!

If, as you state, that 'Canonical ... have made a conscious choice not to do those fixes[.]', then, that's reprehensible on their behalf; not least because they seem to be offering their product as a bona fide alternative; not least if 'cash-strapped' Mandriva are somehow still doing their best.

I know this may seem facetious, but, if they're trying to compete with MSW, or, be offering their product as an alternative, then can they be expected to do 'a decent job'? (Look out for the 'necessary' anti-virus adverts!)

And with reference to your 'original', and final point: if it works; why break it?!

There are three points which make me uneasy , the link between a brand -and its flaws- and linux adoption, the fate of Mandriva and that you forgot Victor:

* I agree that often, Ubuntu is a synonym for GNU/linux (though I do not know their contribution to Linux kernels and to useful applications _in the case of the GIMP, it is a negative one_ ) : I read it in LeMonde some days ago ("MAacOS, Windows and Ubuntu" sorry LeMonde has no free a
And the fact that 95rchives) and it is very common.
But it is not the 1rst time brand names are taken as synonyms ("Singer" for sewing machines in Western francophone Africa, "Kodiarana" -from goodyear- for tyres in Madagasikara, "frigidaire" for fridge in French) but there is no confusion between the brand and its synonym ; though people know what tyres, sewing machines and fridges are, they would not boycott all the tyres, ... if Goodyear , (...)had an anti-user policy...
And the fact that 95% of the users do not know what an OS is (ask a tailor, an ice-cream sellar, a car fixer) leads mo to believe that UBUlinux anti user policy wonot have any consequence on GNU/linux adoption (if such an adoption happens).

* There are little links between the sad (as usual) financial state of Mandriva and what can be achieved technically by Mandriva; else, if an analogy with physicians were valid, a physician who had a lousy financial state would be allowed to let his patients suffer and die (perhaps it is valid in the Canonical world).
The lousy financial state of the anti-pr (cf http://distrowatch.com/weekly.php?issue=20061002) distribution has been known for at least 5 years, and , if there are nothing new and interesting in the free world, "journalists" without inspiration are glad to link to rumors and unverified facts.... This often happens in spring with Mandrive -some day/year/century, it might come true just like flipping coins-
I tested Mandriva 2010.1 RC1 and 2, and though it is overearly to have a stable opinion, I found nice things :
the Gnome version did not eat too much CPU while trying to hibernate (that annoyed me with a 2010.0 install (not a live image, an install able to use their package manager without efforts) on an USB-stick : I had removed the swap partition, but I did not understand what was wrong.
When they switched from 2010 RC1 to RC2, some gtk libraries were less broken (and therefore I could compile very new R plugins : I could not with RC1); as sometimes alphas are better than betas (it was the case of PCBSD in 2008, as progress is not always monotonous), I was glad there were improvements in this case (for me and my colleagues, the state of Gnome -they do not love KDE-4 at all (who would?) and of the applications they need is likely to be sufficient).

*I was appealed with VectorLinux life CD : it is shipped with KDE3, and it can be unetbootin'ed (this is easier than installing, as there is no babysitting at all -just copies- and needs less mouse clicks (2 or 3 : it is less difficult than installing OpenOffice under Windows -M$ word is likely even worse-) and it has every header I need to compile what I miss from source # R, octave , gnuplot , grass, qt4 , vym and lyx-
The only little flaws are with keymap layout (a 4 level menu in KDE control center, Scientific Linux or some versions of Sabayon are easier) but I suppose it wonot make US inhabitants sad... and the absence (in the live version) of the GIMP and OpenOffice -I can get R's plugins compiles, which emit ODF formats, but cannot put in documents. And I could not see how to install it (normally, the install button should be hidden, as live CDs can be lent to children, but it seems well hidden!!

Ubuntu and its derivatives are probably perfect for 'converts', for the simple reason of similarities to Windows. You will be offered all the 'care' spiel, and you WILL be prompted to install 'important updates', and if you do install them, you may well find that your system becomes gradually slower, much like MSW.

Unlike MSW (XP anyway) 'Ubuntu' updates have to be installed (and therefore activated) by 'root' password permission. If, however, the update contains badly written code, it can crash your system or generally mess it up: in other words, do exactly what certain 'viruses' do.

Guess what happened to me overnight? Yes, my system was infected by a 'Linux virus'! OK, I fixed it, using the live disc, and I didn't lose any data. It had only crashed the kernel and 'freaked out' GRUB.

This brings me to a very significant point: if you want to run 'Linux' (or Ubuntu!), then you have to learn a bit about how the system works; although you don't need to be either a software engineer or a micro-electrician (who has to be somewhat diminutive to get inside the machinery!).

Mandriva is started to look very tempting - I may just take Caitlyn's advice.

Caitlyn

Why is it not possible for a Home User to buy a version of Linux that just works as simply as Microsoft Windows does?

I have tried several versions of Linux and found them all to be too complex and unreliable for the Home Desktop. They seem to work for a while but in the end they just fail in some respect or other.

Yes, Linux is free but the only thing that I have ever learned from it is to beware of people bearing gifts.

Microsoft Windows, on the other hand costs a lot of money, but the Home User can use it and it does work all of the time, presents the non-technical Home User with an opportunity to learn how to use a computer without too much heartache.

In my own expereince: Microsoft Windows Joyful; Linux Miserable


@Glenn: Do you have more money than sense?

Some Linux distros work as simply as Microsoft Windows, if not more so; they do however behave slightly differently, if that's what you mean by complexity. You make no mention of which linux distros you've tried, which version of Windows you prefer, or what your 'desktop' requirements are.

Microsoft's products are squarely aimed at consumers: they are not interested in people learning. If you really want to learn how to use an operating system, then you have to look inside, simple as that.

One of the most common statements I've heard from people who've started using Linux is 'it just works!'. Likewise, one of the most common complaints I've heard from Microsoft users is 'it won't let you do anything!' (Vista especially).

However, seeing that you're quite happy to waste/pay a lot of money, & presumably continue to do so; why not spend it on a Mac, rather than continue to pay for an inherently insecure, inferior, Microsoft product? At least you'll get something half-decent.

If you do find a Linux distro that works for you, you can always make a much appreciated donation; besides which, there are a few Linuces that you do have to pay for, if you want the full version.

I agree that more choice is better.. However I/We seem to be alone in that area. Think about the iPhone, does it matter that its closed? That you get treated poorly more often then not? No, fact is if you develop some thing that looks good and just works people will flock to it. The iPhone/iPad are really good looking devices, however Android phones offer more choices and have way more potential are are more often then not cheaper, yet iPhone is dominating the market. Fact is for the general public they want some thing that looks good, just works and works well and is easy to use. I think if ever there is a distribution that comes out where they have great hardware a slick interface and get lots of goodies to go with it, that is what would speed up Linux on the desktop. In reality most people (non techies) just like what they like and it simply does not matter if its closed or open or limits choice. At least that is what I have seen. Form over function wins the day.

I like how Caitlyn sticks to her guns. :) Also, I like how Mandriva, once Mandrake, has always been at the forefront of supporting the principles of Free Software and produced a top-notch distribution, despite the dramas. I've been a productive user of this distro since the Linux Mandrake 7.1 Deluxe boxed set.

I know this thread is old; however, its still true to this day. And we're in 2011.

I have to agree with Miss Martin. Ubuntu does not fix bugs. They brush them under the carpet and move on; leaving a binary mess in their wake. These guys have manifested into the worst and laziest pushers of software junk that man has ever known.

We have just loaded Ubuntu 10.04 onto several "NEW" machines at the office.

All three machines have various problems that render them fairly useless. The worst issue is hanging and total lock ups. One hangs every few minutes, just sitting idle. I see that the hanging issue was brought to their attention back in 2007. Ubuntu blew it off. It's now Jan 25th, 2011 and we are still seeing posting about this issue. Meanwhile;Shuttleworth is off on another new release caper without addressing anything. His mind really is in the clouds.....literally.

This was the last straw for us. We're looking at Linux Mint for our employees now. I hear that Mint has some people working for them that still use common sense. Or we will go back to windows. At least it's stable now. Ubuntu is about as stable as a fart in a match factory.

I believe Shuttleworth is so power thirsty that he's begun drinking the toilet water.

Yea, Ubuntu is a slug and C@N0N1C@L should concentrate on other things. Linux community is held by the 40-megaton dead slug that stinks of MSFT-like unsupport.

1. I trust Mozilla more than i trust Canonical/Ubuntu
2. Ubuntu is not a great OS (lol), is/was a great idea until Unity.
3. windows Xp has 40% market share that’s a nice reasonable target… don’t go after windows7/MacOS (LOL)
4. SLOW DOWN! ubuntu is moving to fast! in all directions!! it’s unpredictable, buggy, etc
5. Hire some good graphic designers, double the number of developers working on the software center
6. Follow the history of Apple … post 1997 and then forget about it, because Canonical is not Apple and you are not Steve Jobs.
7. fork gnome (kidding)
Tanguy, my new site: http://bit.ly/lXcOjd

News Topics

Recommended for You

Got a Question?