Re: Mediation

[Mark Wielaard]

Hi,

On Tue, 2004-12-28 at 02:34 +0000, Andrew John Hughes wrote:
>       I like this idea of a mediator/librarian role.

Me to!

>   Having a physical human
> contact for new developers would certainly decrease the learning curve,
> and help integrate the newcomer into the project.  Also, as you say,
> there are a lot of things which occur on the mailing lists and in IRC
> that are not formally noted, with developers instead relying on a kind
> of secret wisdom.

And certainly not deliberately. It is often just a time issue (or not
enough time to be more precise).

>       Again, GNU Classpath is fairly non-standard in that:
>       a) code developers have to be untainted
>       b) its a GNU project with FSF-assigned copyright, so there is a formal
> record of all code developers.
> This applies more to code than anything else.

I don't think this is so non-standard. All projects must have some
process in place to make sure they can actually distribute all
contributions. We are just a little bit more upfront about it since we
want to make as sure as possible not to get surprizes later on. It is
standard practise for FSF GNU projects (see Keeping Free Software Free:
http://www.gnu.org/prep/standards/html_node/Legal-Issues.html). And
Apache projects have something similar (See the
Contributor License Agreements http://www.apache.org/licenses/#clas).
Even the Linux kernel these days has a way to track contributors
(Documenting How Patches Reach The Kernel
http://kerneltrap.org/node/3180).

>       Whether these issues are good or bad, I'll leave for your thesis.  I
> think there is definitely an advantage, in having no formal boundary for
> entry into the team.  As you mention, there are informal boundaries in
> 'getting into the groove', adopting the methodology being used and
> finding out how things work.  Could this new role solve this without
> removing the openness of the project?

I think it definitely can. I actually hope it will make the project seem
much more open to contributors then it is seen now.

When I was on vacation in Berlin I met Robert (thanks for showing the
town!) and I suggested to setup a part on developer.classpath.org (a
wiki?) to collect all results of the mediator. That way we can keep the
information around (and hopefully update it) when Robert his thesis is
finished.

>       At a wider scope, I think you have a more general insight into how
> formal methodologies (the idea of the mediator being just one) can be
> applied to a FOSS project.  I'd be interested to see your findings on
> this.

Yes. I am also very interested in your findings. Thanks for doing this.

Cheers,

Mark

signature.asc
Description: This is a digitally signed message part