John Levon <[EMAIL PROTECTED]> writes:
| On Mon, Aug 19, 2002 at 12:24:31PM +0200, Jean-Marc Lasgouttes wrote:
>
>> John> Mmm, I think so. I need an efficient way to update. I still
>> John> really want to see specific signals for these things like
>> John> lastfiles.
>>
>> Are you kidding? How many clock ticks are you expecting to gain with
>> you 'efficient' way to update all the 4 entries in lastfiles? 20?
>>
>> My take on that is that 'efficient' updating only counts for the TOC.
>> Until you have working, efficient, bugfree and robust code to do that,
>> forget about intelligently updating anything else. In the meantime,
>> since we want to have working Qt menus, why not do what klyx does ad
>> be done with it?
>
| Hmm, I forgot about klyx.
>
>> Let's be serious: have you seen anybody complain about the speed of
>> the Navigate menu in the xforms frontend. I just tried on my p2/366
>> with userguide, and I cannot see a delay. And this thing has never
>> been profiled! I agree the code is tasteless, but in practice, I
>> cannot see the problem.
>
| I've said it before the problem is simple: I am entirely unsure that the
| fill-on-demand thing is even possible. I've already proven that "update
| all the menus" is unworkable, I suppose I shall have to try the same
| with fill-on-demand and see what happens.
I wonder... can we stop making the Qt frontend the best that ever was
and just make a working Qt frontend now? Then we can get 1.3.0 out,
and also get some broader audience for the Qt frontend.
Just make it work!
--
Lgb
