RE: Glib main loop update
- From: Martynas Kunigelis <mkunigelis alna lt>
- To: "'jgarzik pobox com'" <jgarzik pobox com>
- Cc: "'gtk-devel-list redhat com'" <gtk-devel-list redhat com>
- Subject: RE: Glib main loop update
- Date: Tue, 24 Nov 1998 11:30:37 +0100
-----Original Message-----
From: Jeff Garzik [SMTP:jgarzik@pobox.com]
Sent: 1998 m. Lapkritis 23 d. 17:29
To: gtk-devel-list@redhat.com
Subject: Re: Glib main loop update
Owen Taylor wrote:
> [ On a different subject, what do people think about splitting
> up glib.h, which is 2200 lines before these main loop
> additions? ]
Yes! /usr/include/glib/*.h would be great.
The GTK+ headers are very easy to read, and my on-line GTK+ reference
consists of "cd /usr/include/gtk ; less <file>.h". It would be great if
GLib were organized along those same lines.
Jeff
[Martynas Kunigelis] I second that. How the header should be split up is
not a trivial question, though. For GLib objects like linked lists or hash
tables, the choice is obvious (<glib/glist.h>, <glib/ghashtable.h> etc.),
but the basic things like the types (guint32 and friends) and endian
conversion macros need some thought IMHO. <glib/gtypes.h> and
<glib/gendian.h> probably are good choices. Where the simple utility
functions like g_basename() or g_dirname() belong, I have no idea. Maybe
<glib/gutil.h> ?
Please excuse me if this message seems malformed or such, I'm using MS
Outlook to send it :( Company policy, you know.
-Martynas
--
To unsubscribe: mail gtk-devel-list-request@redhat.com with
"unsubscribe" as the Subject.
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]