On Wed, Feb 05, 2003 at 11:03:02AM +0100, Michael Schmitt wrote: > IMHO there should also be a list of open issues/bugs that must be fixed > under all circumstances. > Such a list should be set up by all developers > about a month before the next release. Today there is bugzilla with the > opportunity to set a "target" and to set the level of "severity" but > actually no formal procedure is established for "interpreting" these > entries. (Currently, there are 7 open issues with target "1.3.0" and 10 > bug reports with severity "major" or "critical". So are were really > ready for 1.3.0?)
Undo crashes are critical for me as I currently get them about once a day although I know that I have to avoid multiple undos. And I already lost a few lines although this is "impossible". Best practice right now to cope with that is "cvs commit, save, 20 times Undo, 2 redos, wait for crash, be happy if it doesn't". So this _is_ critical, but currently nobody is able to fix it. And nobody will be able to fix it in a freeze. That's why I am in favour of a fixed date and a "all or nothing" on that date: Either the next release ships, or development is completely open again to regain the possibility to rip out the guts, redo some things from scratch and try again with a new fixed date. > It really hurts me that fundamental things like search&replace (in > combination with insets) are broken. Unfortunately, I am not able to fix > it by myself (no time, no knowledge about the LyX kernel) but even if I > were, I guess you wouldn't accept the patch because it is too late for > the 1.3.0 development cycle... It is too late during any freeze. With your proposal of "stuff to be fixed under all circumstances" we just run into a deadlock: Some problem will be discovered in a freeze, nothing clean can be done about it because of the freeze, so it won't be fixed, so we can't ship, so we stay in freeze, we lose developpers and potential contributors (I think I have seen three or four sensible patches from "new" people which have not been applied due to the freeze. That's "encouraging"...) There must be a means to break such deadlocks. Your proposal just increases the probability that deadlocks occur, whereas Angus' suggestion is at least theoretically a way to break them. Andre' -- Those who desire to give up Freedom in order to gain Security, will not have, nor do they deserve, either one. (T. Jefferson)
