Andre Poenitz <[EMAIL PROTECTED]> writes:
>> - We should stop implementing new stuff and start fixing and completing
>> the existing features (after Martin has committed his box inset, of
>> course)
>
| The idea is ok, but I'd not be too strict about it. There are a few
| people which probably can't help much with the real problems, and I
| don't think we need to stop them working on non-intrusive improvements.
The problem is getting those who know to do stuff.
>> - We should stop polishing the kernel. LyX 1.3.3 works despite a lot of
>> design flaws. Thus, I see no reason why we shouldn't be able to release
>> a stable 1.4.0 based on the current code
>
| We are somehow in a transitional state. Going from there to a 'it
| somehow works' state is probably the same amount of work as fixing it
| properly.
It is the probably I do not like. Also the underlying attitude that:
No need to fix bugs, finishing the trasition will fix it.
| In fact, I think there's not too much to be fixed left.
Go look at cursor posistions, drawing of InsetText etc.
there is not need to move into advanced inset usage to see *loads* of
regressions.
| critical bugs right now, coming in two suits: A somehow unitialized
| LyXText and a *ed up table access. The latter is probably easily spotted
| by someone spending some time reading the code. The former might be a
| bit harder. But again, it could be that simply swapping a few lines
| fixes it.
>
| Than there are 153 'normal' bugs. Quite a few of them are
| reLyX/round-trip related. No biggo. Should be verified against tex2lyx
| by someone...
reLyX round-trip bugs are not important, regressions drawing errors
cursor placement bugs are the important ones. My guess is that most of
these have not even been entered into bugzilla.
>
| Ok. Left with say ~4 people and ~12 weeks until Chrismas, this makes
| three bugs to be fixed per week and active developer?
>
| Does this sound unreasonable?
>
| I don't think so.
Get grinding.
--
Lgb
