Re: [forum] an observation from the sidelines

[Carl Worth <cworth@east.isi.edu>]

On Mar 22, David Dawes wrote:
 > Why do you equate CVS access with determining the technical direction
 > of a Project?

See below...

 > Getting back to this particular thread, there are no old non-active
 > XFree86 guys telling Carl Worth what he can work on.  From his
 > description of his work, it sounds like it could stand on its own
 > independently of XFree86.  If Carl Worth would like his work to
 > become a part of XFree86, then he should present it to the XFree86
 > developers via the public devel@xfree86.org list.

If I propose new work on devel@xfree86.org, the greatest effect it
could have on the technical direction of the XFree86 Project would be
if it were added to the XFree86 CVS tree. Thus, CVS access does have a
direct role in determining the technical direction of the project.

Now, I've read several messages in this forum from members of the core
team that seem to be attempting to establish the following points:

	1) The technical direction of XFree86 is in fact determined by
	a consensus of the "team" of developers.

	2) This "team" is comprised of more than the core team, but
	also includes people like myself that subscribe to
	devel@xfree86.org and submit code, (eg. to fixes@xfree86.org
	or to core team members).

(This is my understanding, clarifications are always welcome).

My perception of XFree86 has not been consistent with the above
points. I have seen changes that did not seem based on consensus of
any public group, (eg. the sudden merging of many mailing lists
without any advance public discussion). Also, the opaque and gated
submission process has always left me feeling like an outsider rather
than a team member.

I don't think that these points can be easily dismissed as a large
"public perception problem". It may very well be that I have formed
opinions about the workings of XFree86 that are inconsistent with
fact. But, as pointed out by others, the sense of belonging to a team
is nothing more than a issue of perception. Witness the following
exchange in which a "member" is not even aware of the existence of the
"team" to which he is regarding as "belonging" by another:

	On Mon, 24 Mar 2003 14:47:46 +0000, Alan Hourihane wrote:
	> On Mon, Mar 24, 2003 at 02:13:49PM +0000, Keith Whitwell wrote:
	> > What is this other XFree86 team?  What are the rights and
	> > responsibilities associated with membership?
	>
	> That's just it - there isn't. In my eyes Keith, you are part
	> of the XFree86 team already because of the DRI work you
	> already do.

There are things that a project like XFree86 could do to engender a
stronger feeling of membership and community from developers like
myself. Describing and offering rights and responsibilities of
membership is one important aspect.

For example, if XFree86 were to accept some libraries that I have
written, and gave me the right to maintain them in the XFree86 CVS
tree, I would accept the responsibility to not touch anything in CVS
outside of those libraries. I would be glad to submit patches to
maintainers for any changes I wanted outside of the code I would be
maintaining.

In that type of situation, I would feel like a member of the XFree86
team. As it now stands, I'm just a programmer that has some libraries
that happen to use X, (in my own CVS tree). I've also got some code in
the XFree86 X server that I'm not very motivated to maintain given the
current lack of direct maintenance possibilities.

-Carl

-- 
Carl Worth                                        
USC Information Sciences Institute                      cworth@isi.edu