Re: RFC: Adding zlib dependency to libgio
- From: nf2 <nf2 email gmail com>
- To: Matthias Clasen <matthias clasen gmail com>
- Cc: gtk-devel-list gnome org, thiago kde org, Alexander Larsson <alexl redhat com>
- Subject: Re: RFC: Adding zlib dependency to libgio
- Date: Wed, 11 Nov 2009 03:54:53 +0100
On Tue, Nov 10, 2009 at 1:41 PM, Matthias Clasen
<matthias clasen gmail com> wrote:
> On Tue, Nov 10, 2009 at 5:49 AM, nf2 <nf2 email gmail com> wrote:
>> On Tue, Nov 10, 2009 at 9:44 AM, Alexander Larsson <alexl redhat com> wrote:
>>> On Mon, 2009-11-09 at 23:03 -0500, Paul Davis wrote:
>>>> On Mon, Nov 9, 2009 at 6:23 AM, Alexander Larsson <alexl redhat com> wrote:
>> As I'm reading the word Gtk+ here more often: I still believe that a
>> VFS API shoudn't be tied to a certain UI toolkit. That would be
>> repeating the mistake KIO did. What about Mozilla, OpenOffice, VLC and
>> many many others. They should all link GIO (as a VFS API). If libgio
>> dupulicates too many things they already have, they might be put off.
> Whether GIO contains DBus support or not has probably minimal
> influence on whether or not Mozilla or OpenOffice use it. They make
> their own decisions about what platform to build on and don't care if
> you think they 'should' use anything in particular.
The modularity of the GLib stack is a nice feature. A lot of things
which are built with it, might be very useful far outside Gtk+ and
Let's take for instance libsoup: It already links GIO. But for what
sake would a HTTP client library always need to carry D-Bus around?
On my system, Gtk+ links 44 libraries. I guess one less or more won't
make any difference. Or, for instance a "gvfs-ls /" , which probably
has to load about 15 libraries, takes 0.03 seconds. Therfore - I
reckon - gathering unrelated APIs into a single *.so won't have any
significant performance benefit. But maybe i'm wrong.
] [Thread Prev