On 23 Sep 1999 15:33:51 +0200, Lars Gullik Bj°nnes wrote:
>I'm a little curious why you think that using inline funcs can help
>us not using heap mem.
Stuff like this is unfortunately endemic e.g. in figinset.C:
[...]
{
[VIP (do Very Important Processing here)]
FOO *p = (FOO *)malloc(sizeof(FOO *));
[VIP, not using p]
while (1) {
[VIP using p]
if (BAR) break;
}
[VIP, not using p]
free(p);
[VIP]
}
So you are better off extracting:
static inline void
VIPfunc(FOO *p)
{
while (1) {
[VIP using p]
if (BAR) break;
}
}
and:
[...]
{
[VIP (do Very Important Processing here)]
[VIP, not using p]
[VIP]
}
>
>| >I will try to spell out my thoughs on figinset.C later, but in short:
>| > Ditch It! De-couple figinset and insetfig. transformation of
>| > eps (or others) into pixmap can be done by external programs.
>|
>| aka toolkits, I suppose.
>
>if you consider gs a toolkit I suppose.
direct ghostview-like action is what LyX does now!
the TK should do this (and invoke gs)
>| No problem to grab and forward summaries I can find.
>|
>| >I'll give you write access to the cvs
>| >repository if you do this. (at least that part of the repository)
>|
>| hmm... How can this save any work for anybody, compared with committing
>| a simple text via list? Do you really want to start to manage a mess of
>| accounts, sub-accounts, limited sub-accounts while re-organizing cvs?
>| Things will inevitably go wrong and users will get confused. Better
>| forget this idea ;-).
>
>If we are to have a document like that (or several) we need someone to
>maintain them. And I am not going to do that.
>
>And I have finished the cvs re-organizing. And we are now using access
>controll list to limit access to the different modules (and
>sub-modules)
I see. BTW cvs manual is frightening. So give me a little time to get
get accustomed to this beast before you let me break everything....
>| >You can even hunt in the mailing list archives to see if there has
>| >been any other nice mails that could go into a document like that.
>|
>| The 'hunter' should preferably be the one who has at least a vague idea
>| that there was something interesting written about 'foo', so he would
>| know he has to search for 'foo': The author himself! We should just
>| 'help' him not to 'bury' things he has invested work in.
>
>So we should ask everybody that has ever posted on the list to see if
>they ever wrote anything interesting about foo? :-)
Well, at least we should ask those well known (notorious) writers of
long, important mails like Allan or this one (help! What am (was) I
thinking :-)
----snip---
On Wed, 22 Sep 1999 09:19:22 -0400, Amir Karger wrote:
>Do either of you have an archive [...] of the thoughts we had?
---snap---
>If we are to have a document like that (or several) we need someone to
>maintain them. And I am not going to do that.
I'm a bit helpless how to maintain the 'thoughts you had' ;-) without
the necessary clues delivered by the 'thinkers'.
Cheers,
Arnd
