Re: new version of MAD

[Steef Boerrigter <sxmboer sci kun nl>]

Pavel Roskin wrote:
> 
> Hi, Steef!
> I see.  This means that you are not aware of the current state of memory
> allocation and debugging in MC, which has changed dramatically.

My goodness. Either you are really pulling my leg (and succeeding very
well) or you really don't READ your email before you answer them.
I'm comparing my mad.c which evolved over two years versus the one in
mc-2001-07-29-14.tar.gz. Did that version chance dramatically? I doubt
it.

> Strange, I searched both mc gnome org and mc-devel gnome org and could not
> find the word "key_translator".  It must have been long ago and I probably
> didn't participate.
pfff, you must be overworked or something. I was refering to the patches
in the file edit/edit_key_translator.c that YOU put in TWO WEEKS ago
(2001-07-10). Just search on "Steef Boerrigter" in the Changelog.

> The real problem now is that MAD doesn't work at all because it doesn't
> know about g_get_current_dir(), so MC crashes as soon as it tries to free
> the result.
Fair enough, that sounds like a priority issue.

> Maybe this can be completely solved by using glib-based functions?
I shall look into the glib safety measures and get back on this subject
as soon as I know whether glib does in fact support bounds checking and
other type memory errors that are treated by MAD. If there's really no
point in putting it in MC than forget about it. Actually, some people do
use other libraries than glib and other computers than intel boxes.

> I meant your future patch, if you ever make one.
I need little additional disencouragement. And if you really do not like
to answer a simple question, fine!