Re: GNOME Namespace Management - ARC & GNOME

[Mark McLoughlin <markmc redhat com>]

Hi Brian,

On Wed, 2004-12-15 at 15:14 -0600, Brian Cameron wrote:
> Murray:
> 
> >>Sorry that my emails are so long, I
> >>hope they aren't putting you to sleep.
> > 
> > You could make them shorter just by answering direct questions. Please do
> > try to be concise and self-editing.  Just give us a URL to previous stuff
> > in the archive instead of repeating yourself. Personally, I want this to
> > succeed, and I think that brevity and clarity are how to do it.
> 
> Good point.  

[snip another huge mail]

	It is a good point. As interested as I am in the discussion, I've had
to tune out just because of the sheer volume of these mails.

	I think a point is being missed in this whole discussion, though. We've
totally been concentrating on ABI stability, but that wasn't the subject
of your original mail. Indeed, I think GNOME and ARC are pretty much on
the same page when it comes to ABI stability - I don't think the ARC
would have much problem with gtk+ etc. ABI stability since GNOME 2.0[1].

	The topic of your original was "namespace management with regards to
file installation locations". I'd summarise your concerns as:

  1) There is no guarantees about file format/location stability in 
     GNOME. We either have never thought about file locations as being
     an "interface" just as much as a libraries API is or we just 
     consider these interfaces to be entirely private.

  2) We've made a mess out of /usr/share and scattered stuff all over 
     the place with zero naming convention and no attempt to isolate
     all GNOME data files under the something like /usr/share/gnome

	So, lets get back on that topic. Going over the well worn path of ABI
stability isn't gaining us anything here.

Cheers,
Mark.

[1] - Indeed, I went through the "ABI diffs" when you guys were 
      preparing for your recent ARC case and the only ABI change that
      was of concern was because Sun shipped the gtk+ multihead ABI
      before it was finalized upstream.

      Now, API behaviour stability is a slightly more contentious 
      topic, and GNOME/gtk+ may be slightly more willing to change 
      behaviour in order to fix what are essentially bugs, but the 
      rationale Owen just gave is something I think the ARC can 
      certainly understand/relate to. The gap in thinking on this issue 
      is not large by any means, IMHO.