Re: Discussing issues

[Havoc Pennington <hp@redhat.com>]

On Tue, Apr 15, 2003 at 12:30:12AM -0400, Mark Vojkovich wrote: 
>    What's the matter with defining "large enough" to be "has made
> significant contributions over many years"?  That sounds like as
> good a metric as any, perhaps better than all others.  Though I can
> pretty much guarantee you that by that definition the vote would
> leave pretty much all of the current board and core intact. 
> 
> 
>    This appears to disturb some people.  Why is that?  And what
> justification do you have that this is not the best way for things
> to be - with the people who have contributed the most having the
> most say?  
> 

The way I would justify any decision for how to draw these lines would
be to look at whether it was working in other projects, whether it was
working in XFree86, and for areas where it isn't working, I'd try to
understand specific reasons why not.

When I say "working" I mean in terms of results. One metric is how
efficiently the technical aspects of the project are moving; are all
submodules covered; is there steady progress; do bugs/patches get
handled in a timely fashion. Another metric is whether the project is
able to handle personality conflicts or issues such as company
involvement without creating technical problems or endless
distraction. Does the organization have confidence that it is fair,
and that a company can't override the fairness? Another metric is
whether non-code contributors such as docs, i18n, bugzilla triagers,
web site maintainers, etc. are effectively used. Another is ability to
make and execute shared decisions such as releases. A key
XFree86-specific metric for me is whether it's driving much-needed new
library APIs and extensions in a convincing fashion, as that's my pet
agenda. And so on, you can think of lots of stuff, much of it has been
brought up. XFree86 is good on some metrics and bad on others.

To me the key point about a working project is that it's scalable.
Scalable means you are set up to take advantage of the efforts of
hundreds of people, with many different interests, talents,
personalities, and motivations. While at the same time, keeping the
ability to make shared decisions when necessary.

I'm not thinking of this in a priori moral terms, I'm just looking at
what outputs we get for certain inputs, based on the empirical
experience that's out there.

One random thought: so far XFree86 core team members have only really
mentioned the Linux kernel as an example of how other projects
work. I'd caution you that the kernel is a huge anomaly, it's the
freak in the list. And despite being a poster child, the kernel does
poorly on some of the metrics I listed, compared to other projects.

Havoc
_______________________________________________
Forum mailing list
Forum@XFree86.Org
http://XFree86.Org/mailman/listinfo/forum