I just touched on this in another thread but thought I'd start a new topic as it seems timely with the docsprint coming up, release of 1.0, and changes in management positions. Please take my opinions with much salt and so on, and of course know that I am very thankful for all the work gone into TG and its componenents....
I strongly feel that the largest obstacle to TG adoption right now is the way it is very scattered and inconsistent, and thus perceived as an unreliable way to build apps where the developer must know they will succeed. I personally am surprised we are giving it the big "1.0" as things stand. My feeling is that TG desperately needs a much stricter separation of dev and stable branches. This would include stable *frozen* documentation and code with the only work on those features being bug fixes and corrections. At the moment, from my two months experience with TG, I would not feel comfortable saying that I am sure I can deploy a simple TG app quickly. I also would not feel comfortable recommending TG to someone who needs to be able to work through the docs and get a solid product at the end that does exactly what they expect in a manner consistent with what they read. I'm at the point where I *expect* surprises! ( not the good kind ... ) I strongly feel that for increased adoption by serious users ( which will eventually produce more developers/documenters/testes) we need: 1.0 Stable: - all features are documented in one central place with docs detailing exactly what they are *known* to do reliably, and no mention of new unstable features - no features are present that are not properly documented and tested ( no surpises! ) - all stable features properly tested with a downloadable set of publicly documented tests - the packages encompassing the above are always available in a reliable manner with mirrors and frozen tarballs ( ie if sqlobject X.X passed our test suites and is the documented version, we stick with X.X in that download ). These are available in multiple formats from multiple servers. 1.x devel: - the new features go here, changes in packages, etc - add on docs that are known to be moving targets I am curious to find out whether I am in a minority in this perception or whether it is something that bears further discussion. I like all the new features that are coming up, and I'm very excited to get to some of them. But it seems we have some slightly out of control feature creep going on. If someone offers me a job on a tight schedule, I would prefer to have a setup where I *know* which tools and features are really at my disposal. I'd rather have to roll my own with no surprises for a few features than wonder about the lot. I'm very glad to see the discussion of TG's future go in a distro'ish direction. I think we should all remember that the job of a distro ( IMHO ) is to provide a known stable set of tools with solid documentation, testing, and availability. For me, the debian model is a very good one, stable, unstable, testing. Thoughts and discussion? Iain --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "TurboGears" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/turbogears?hl=en -~----------~----~----~----~------~----~------~--~---
