What should we do when some large FUDdy commercial Colossus whose motives we don't necessarily have confidence in wants to do something we pretty much had wished they would do? My recommendation: just lie back and enjoy it honey!
This is what I thought of Microsoft with OOXML and I have the same reaction to IBM's recent publication of its I.T. Standards Policies.
I have had quite a few emails asking me to write a blog defending ISO, irritated that IBM's moves are grandstanding, transparent marketing, last refuge of soundrels, a monopolist complaining about lack of openness, trying to entrench a world-view based on the discredited allegations of the OOXML process, and so on. And, truth be told, of course there is an aspect of rump Plan C about IBM's efforts here.
But I really don't think the Plan C aspects will have legs even in the medium term: people see through that, which leaves us free to consider the good effects that this can have. Within hours of releasing the press release, IBM's Bob Sutor was busily back-pedaling that Notwithstanding the various sexy headlines I've seen today, leaving a standards group would be a last resort.
So what positives can we hope for?
For a start, you cannot nowadays find a supporter for the ISO fast-track procedures. If IBM's machinations helps alter these, that would be good.
If it prompts more governments to get serious about participation in all kinds of standards bodies, that would be good.
If it prompts more technical experts to get involved in any level, that would be good.
If it promotes FOSS people to get more serious about standards, so that inter-project requirements as well as developer energy and corporate agendas drive open source projects, that would be good.
Clearly there is a group of people startled by the discovery that the voluntary standards regime can be unsatisfactory for use for setting mandatory exclusive technologies for all sorts of reasons. If they stop blaming the standards organizations for being oranges (and claiming to be oranges) rather than the desired apples of their dreams, and instead concentrate on the positive step of trying to get bodies or procedures that are more satisfactory, why wouldn't that be a good thing too?
For example, at OASIS there are no limits AFAIK on the membership of committees: I've reported before that the ODF TC had for the most of the last year a rollcall dominated by two particular vendors. Should there be rules which prevent vendors from having more than one vote? Should there be rules on associates, and if so how on earth could they be enforced? Should there be rules based on market sectors, so that in any committee vendors and users cannot exercise more than 50% of the vote? That might force vendors to go out and recruit users more, which would be a good thing...or is that then just "stacking"?
And what about second tier organizations? Organizations such as ISO have, as their voting members, not individuals and not corporations and not countries, but national bodies: each of which has a different setup. The NB of China is a government agency, of course, and you can have certain expectations about the attitude to public participation. Other NBs are NGOs. Some charge fees, some (like Australia) do not.
I have never paid a cent in fees for any standards body I have ever participated in with the exception of when I worked for Academia Sinica as their representative at W3C: they paid the W3C fees. And I think that is the way things should be in more countries; Bob Sutor is factually wrong that participation in ISO necessarily involves more fees than participation in other bodies (any fees vary according to the policies of the National Bodies and the particular projects), but right in raising it as an issue.
Many national and international standards bodies for years funded or subsidized themselves as publishing houses. For the last 15 years they have all been having a debate about the role of this model in the age of WWW access and distribution, and many have moved over to a model where they are happy if non-revenue generating standards get farmed out to the interested consortia and fast-tracked if necessary.
But there are other models. In New Zealand, I believe, the standards body negotiates with government for grants to participate, so there is a direct tie in between standards participation and government policy. (In Australia, Standards Australia also applies for grants on specific requirements: I was able to participated in the OOXML BRM because of a grant for travel expenses, for example.)
If citizens lobby public policy makers to realize the central importance of a viable standards organizations in the modern economy, I don't see any downside. Many governments have ignored the area for far too long, and they need to stop focussing on stimulating industry by privatising rights to use things (e.g. patents) and more opening up the rights so that the public can use them (e.g. standards —in the larger sense of RAND-z, QA-ed, open specifications— and orphan technologies).
And one of the challenges for standards bodies is how to get the FOSS community involved. I have repeatedly mentioned why I think the royalty-bearing MPEG work at ISO is nowadays counter-productive: if this IBM effort promotes people to look at all the important standards that are not currently standardized, then we should welcome that: a standards home for Ogg Vorbis, ZIP, etc. would be good.
Now obviously I have particular views about the form that standards need to take: my blog item yesterday The Cathedral and the Bazaar and Standards gives some. Openness needs to exclude NIH (Not Invented Here) notions, but just as much as openness needs to exclude domination by single players it needs to exclude cartelization (see my Is our idea of Open Standards good enough: verifiable vendor neutrality.)
Now my proclivities are that standards need to stimulate the bazaar and the market. They have a particular role in serving niche requirements where the big boys have no interest, and in fact, as a song says, we are all minorities if you look far enough.
But the more that governments want to mandate unique standards outside the areas of health, safety and the environment, the more that the voluntary standards process as it currently stands is suspect. Now I think that this applies even more to consortia which allow vendors or vendor employees multiple votes, or consortia where all detailed work is done by such dominable committees and which only have up-down acceptance votes post-committee.
If this effort prompts national and international standards bodies to liaise with the appropriate competition commissions, that would be a good thing. (I think the ISO/IEC Secretariats need to take the lead in this consultation, for example, and not solicit information rather than awaiting NB pushes.)
While it is very important to have good developer buy-in and commitment at the committee stage, it is also important to have adequate review. This review is a real strength of the ISO and W3C systems, for example. OASIS and many other consortia need more attention to systematic or scheduled review and defect fixing.
So all up I am not at perturbed by IBM's latest caper: to the contrary I would be delighted if it goes well and bears fruit, especially more participation. This is because I really don't see governments replacing their national standards bodies, nor the national standards bodies starting another organization apart from ISO to do what ISO already does. But extra impetus for improvements to the Directives would be wonderful and quite achievable I think, if there is will. And I can see scope for some International Open Standards Organization, with the aim of making it easier to standardize (at some level) FOSS projects: perhaps a super OASIS.
What we don't want is a Global Vendors' Conspiracy.
More and more nations are including definitions of open source and open standards into legislation, and so it does not seem strange that there would be desire to rate organizations: though I expect that this would rapidly prove wrong-headed (throw out ISO beach flag standards because of OOXML? I don't think so...) and instead a more realistic regime of rating actual standards and individual committee processes for openness would need to be implemented. And I think it would die at that stage, because large companies want processes they can set the agenda for: they want standards that make things better for them, not standards that shift things sideways or backwards. The more that developers set the agenda in committees, rather than responding to user requirements, the less credible the claims for openness of the technology should be.
So, maybe I am overly optimistic and incrementalist, but I recommend lying back and enjoying it!