The article recently published in Australian media and picked up by others and rerun in other languages is a direct result of my talk on developing OOo. The comments were obviously cherry picked from the whole content for maximum media effect.
I have no inside running on true numbers, I did comment that I was impressed by the ability of Sun to provide most of the features with so few developers. Whether it be 50 or 100 is probably not relevant, think how many programmers the competitor has and then consider how rich OOo really is. Certainly there are very few programmers working on OOo that are not paid by someone this includes RH, Novell, Sun, ... The point I was making was the huge commitment when you cannot work an 8 hour day on the code, how do you learn and follow it. I made the important point that developers should work out what they want to do and focus on a specific feature or tool and do it well, it is literally impossible to 'understand OOo code', I doubt anyone would claim that. There are many experts on pieces and some excellent generalists but none know it all. Why do we not attract and keep these programmers? Those that know will realise I push the ability to build OOo from EVERY download. I personally think that this is not important enough to some, not all, people within Sun. This includes breaks that break because of changes in the developers machine, like the recent xslt problem, I think it is a P1 but to others it only effects a few programmers. This is philosophical these are my opinions, I wish there was a clear indication of consensus from those directly involved in OOo. I recommend that new developers start at the http://ooo.ximian.com site using the build utility that they have. The reason is that on the whole it 'just works' which a new developer starting a build using the Sun supported process it is confusing as hell where to build from and what the real status of anything is. I think that patch acceptance and build handling would benefit greatly from a distributed Source Control Management with the centre being Sun supported, with Ximian supported, and even my own version that all inter-operating and patches flow more easily because of advances in SCM technology. Distributed SCM will reduce the merge hell that we currently go through. I do not suggest doing this immediately because the SCM space is very confused right now due to the Bit Keeper problem. Calling office suites uninteresting is REALLY missing the point. Just about every presentation in the LCA was done in OOo, this is an app that many people use. Every programmer I have talked to considers OOo a killer application for the desktop. This is an interesting project. It is a complex and we must work on that complexity. What I did get from developers was that licensing with LGPL was actually a problem. The ability to remove code and use it in a commercial application was a real worry to some people. Licensing is an issue and I know that it has been discussed at length previously. I hope that this clears things up a little. Thanks Ken Foskey OpenOffice.org developer --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
