Re: Invitation

Egbert Eich wrote:
> Keith Whitwell writes:
>  > 
>  > Actually, I asked in part because I didn't know there was any other team but 
>  > the core team, so I was confused by your words and wanted clarificiation.
>  > 
>  > What is this other XFree86 team?  What are the rights and responsibilities 
>  > associated with membership?
> 
> I've used the word 'team' for those who consider themselves as part
> of the XFree86 developers community. Maybe I used the word team in
> the wrong context - I don't have my dictionary at hand, but I've
> received English lessons before on this list.

That's fine.  It's just that there have been quite a few teams associated with 
XFree86 - for instance, from coreteam.html:

  Egbert Eich, Germany
     * joined team, 1995
     * joined Core, 1999

I'd been led to understand that the 'team' no longer existed as a formal 
concept, only the core team & BOD remain and that the webpage was out of date, 
which makes it confusing when you start issuing invitations to join another 
'team'.

> Also I don't understand the 'membership' issue here. Can't anyone
> just consider herself/himself as part of a group without formal
> membership? 

I wouldn't have thought so either, but in the face of a formal looking 
'invitation to join', you should be able to forgive some confusion.

> 
>  > 
>  > CVS access seems only to be granted to core team members, and that is why I 
>  > consider the idea of opening that group up to new members to be significant. 
>  > If CVS access were granted to non-core-team-members, that would change.
> 
> I do see an issue with timely application of patches.
> Do you feel you need commit access to be able to do
> your work properly?

Why else would I be rabitting on about it like a deranged fool?

(Pauses for a minute to think about all the code I've got in XFree86, and the 
prospect of having had to do all that through fixes@...  I'd probably rather 
spellcheck slashdot...)

I personally can cope because I'm working on a forked version of XFree86 -- 
namely the DRI tree, which doesn't have the fearful/paranoid approach to cvs 
rights we see in XFree86.  You know what?  Our code still goes into XFree86, 
our tree isn't of demonstrably worse quality (in fact they stay pretty closely 
in sync), but our developers sure seem happier and don't do stuff to 
deliberatly provoke dissent...

Like I said - in my experience, people respond very positively to trust, and 
work exceptionally hard to try & live up to that trust.

An alternative is a Linux/Linus model where noone has access, with fanning out 
rings of trust and responsibility.  I might endorse that if I didn't think it 
would basically boil down to the core team saying: "OK, we're a Linux model 
now, just start sending us patches", and after a couple of months the 
excitement wears off & things end up pretty much as they were before.

Anyway, I'm repeating myself now and it's totally boring for everybody, so I 
think it's time to go & do some actual work.

Keith