On Tue, Oct 08, 2002 at 10:58:31AM +0200, Lars Gullik Bjønnes wrote: > | I think we need some automatism to solve the dead-lock issue. Like "if pre1 > | is not out after a four week freeze, the freeze is revoked"... > > Yes, but in what way does that help?
We could go back to infrastructure work and move LyX into a position where we are able to fix bugs instead to work around them as we usually do in the "freeze" stages. If the infrastructure is "ready", freeze times should be shorter as bugs can be fixed faster, and by a larger number of people for that matter. > Let's say that we are in review state instead: No patches go into cvs > unless it has been reviewed by at least one "core" developer. Call it > "slush" state if you want to. Well, this does not help very much if it is unclear what kind of changes are allowed. Even if it was clear I have a strong feeling that stuff that's high up on _my_ list (mostly IU, multiparagraph environments and simplifying text*.C) does not fall into that category. [I still have the urge do use a big hammer in some areas. Yesterday's two-hour-hunt for the cursor positioning problem (which, surprisingly is not seen by others using the same WM, compiler and LyX version...) did not really help to suppress that urge...] > And I am not sure about who should be in the "core" group. That does not really matter when looking for a concept. > But the objective is not to cram as many features as possible into the > next release, but to get the next release out of the door. Well, I am really not talking about features, rather about getting things straight. And I am still living under the impression that "tiny patches just to make things work" are in the long run vastly inferior to "get them right once even if it takes a while" -- be it in developper's time or code quality. > The best would be for all of us to look at what needs to be done and help > doing it, even if it means fiddling with code that is not "our own". > I know that I am not good at that, mostly because I am a bit daunted by > the the whole guii, and the task of learning qt. Indeed. That's the core problem which obviously haunts all of us. It's a fun project after all, and if we were to spend time on "no fun" work we could go back to our daytime jobs... Seriously, I think that "people just don't do it" is something that has to be accepted as a precondition, not as something that might be changed. [Although a dayly post with "Help now!" might help a bit ;-}] Andre' PS: For heresy's sake: I think the 1.2 series is in an acceptable shape for most users. So I see no _real_ need for 1.3 (anymore). "Power users" always could use two week old CVS and "Adventurous power users" current CVS... -- Those who desire to give up Freedom in order to gain Security, will not have, nor do they deserve, either one. (T. Jefferson)
