PostgreSQL and the SQL standards process

Posted Sep 22, 2011 18:59 UTC (Thu) by petereisentraut (guest, #59453)
Parent article: PostgreSQL and the SQL standards process

I have been to SQL standards committee meetings. Actually, the process is quite relaxed and open, and there is friendly cooperation across vendor lines, once you find your way in. So I don't believe the conspiracy theories. And there are no requirements about secrecy of the proceedings. You can read up on everything if you become a member of the working group or even just a temporary guest.

The main problems are (1) ISO politics and (2) the effort required to do the actual work.

First, the participants don't like the ISO politics either, but they're there, so what are you going to do? ISO has existed for 50+ years. Who else are you going to charge with maintaining your standard that is 20+ years old and should perhaps have another 20 years or so to go?

Second, writing actual technical standards requires experts and a lot of time. Who is going to do it? The work is currently done by a dozen people or so who have quasi-guru status within their sponsoring companies. Who is going to put in the time like this in a volunteer organization?

PostgreSQL and the SQL standards process

Posted Sep 23, 2011 9:22 UTC (Fri) by cmccabe (guest, #60281) [Link] (1 responses)

I'm sure that people at SQL standards committee meetings are friendly. But flying halfway across the world at great expense and doing things behind closed doors is a 19th century way of doing things. Modern standards like HTML5 are developed in the open and discussed on mailing lists and IRC. This just makes things easier and cheaper for everyone-- especially users and developers of related software, who actually get a chance to see what developers are discussing.

Also, not all people who work on open source databases are volunteers.

PostgreSQL and the SQL standards process

Posted Sep 23, 2011 13:24 UTC (Fri) by petereisentraut (guest, #59453) [Link]

Well, sure, some people get paid to work on PostgreSQL. But participating usefully in the community, especially when it comes to extra activities like the one discussed here, goes beyond merely being told by some boss to work on some code base. Participating in the community is a volunteer effort, even if it's ultimately part of your job.

PostgreSQL and the SQL standards process

Posted Sep 23, 2011 10:55 UTC (Fri) by nix (subscriber, #2304) [Link] (1 responses)

You could use the same reasoning to argue that nobody in the free software community would sit on the C++ committee, or that nobody in the free software community would get involved with POSIX. But the former has substantial free software representation and the latter is by now pretty much entirely driven by the needs of the Linux community.

So your reasoning appears somewhat flawed. (Or, rather, contains an erroneous implication: that it requires quasi-gurus, sure. That there is something preventing these quasi-gurus from ever participating unless they are backed by a non-'volunteer organization', possibly. Your implication that non-volunteer organizations are necessarily proprietary software companies is plainly false. If this implication is dropped, your last paragraph becomes a non sequitur.)

PostgreSQL and the SQL standards process

Posted Sep 23, 2011 13:26 UTC (Fri) by petereisentraut (guest, #59453) [Link]

I think you are implying a lot of implications. I'm just stating the facts and asking some questions. Now we have to find the answers.