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.