[vlc-devel] Update on the VLC project

[Rafaël Carré]

Le 03/05/2012 15:53, Pierre Ynard a écrit :

>> In order to avoid the same situation we've been into, I suggest a few
>> things:
>>  - the creation of a technical arbitration committee, to avoid the
>>    YES/NO discussions, and whose decision will be final, like Debian or
>>    Rockbox,
> 
> As you and I just agreed upon, the technical decision process is the
> least of our problems. A technical arbitration committee wouldn't
> help with the lack of orientation for PR/legal action, the lack of
> orientation for technical developement, the lack of roadmap, the lack of
> orientation for fostering our third-party developer community, and the
> lack of management of personal (as in, not technical) conflicts.
> 
> I don't think that the current technical arguments mislead the final
> decisions. Let me ask you one question: will Rémi be in the committee?
> If so, then the problem will still be there, but shifted, and possibly
> worse if it gives more authority to Rémi to assert his opinions. If
> not, then you're arbitrary excluding him from the decisions. Not that it
> would solve a big problem: there are decisions that have been made in
> the past, lua stuff for example, despite his disapproval, and it didn't
> prevent development from going on fine anyway. The only thing that this
> would achieve would be mitigating the tensions and harshness of the
> "YES/NO discussions", by hiding them behind a phantom committee, which
> is not the right solution for the right issue.
> 
> To finish on this topic, I'll come back on a "committee decision" that
> happened in the past: when the association decided, without consulting
> the rest of the developers, without getting the blessing of the lead
> developer, that the licensing questions around Apple's ToS were a gray
> area, and that this justified giving a green light to the iOS port,
> leading to the whole App Store fiasco. I can imagine this "technical
> arbitration committee" as being something obscure and irrelevant again.

I agree with you that trusting a committee is weird and I will not judge
the proposal, but instead I give you my own experience with the Rockbox
Steering Board.

It has 5 members chosen by all committers, elected each year (each year
we forget the deadline so they are effectively elected for a bit more).

3 of them are old hardcore hackers which now contribute not often at
all, if not never, but they are always present on IRC channel. (2 of
them founded the project).

The other 2 contribute more often but mostly in non-technical parts
(i.e. documentation and/or plugins)

So they would both fit in non-technical-enough and not-present-enough.

However in 4 years they were contacted only once and took no action this
time.

They made a statement that the denounced action was indeed not correct,
but that the author fixed it in an acceptable way a bit later.


So it seems that the mere existence of this committee is enough for devs
to behave and sort their problems themselves.