On 02/08/2013 12:56 AM, Baptiste MATHUS wrote:
2013/8/2 Ron Wheeler <[email protected]
<mailto:[email protected]>>
On 01/08/2013 5:55 PM, Jonathan Sharp wrote:
I think John C raises an interesting case here, where the
voting process
can fall down.
A large code dump like that can hurt the quality of
documentation and
support (in addition to team morale).
It depends on how good the code and documentation is!
If the PMC can not deal with the amount of code being proposed,
they need to increase the PMC.
In the meantime, they can always defer the acceptance.
I would love to have the problem of my team producing too much
tested code and documentation for me to read!!!
That's the point, not really talking of "your team" here. The (somehow
a bit hypothetic, granted) point raised here is where you would find
yourself in front of say one year of coding being proposed. And IIUC,
we're only talking about *PMC members* and values here. Sure, one
company/individual could work on his fork and propose a lot of changes
later when he finally wants/is allowed to, and btw he hasn't any
direct commit access to the project and so on.
It will be harder to say yes to the new code and the coder may find that
he/she has to do a lot of rework to resynchronize with the current base
but you are still getting enhancements that otherwise would not be done.
The coder is always free to make a fork and continue with a "better"
product if the work submitted is not acceptable to the PMC or the PMC
becomes moribund.
I think that this is the normal way to handle this.
Obviously, if the PMC can not keep up with a group of coders that are
pursuing an aggressive vision of Maven enhancements, this may cause the
PMC to become an unacceptable bottleneck and the fork will become the
most active version and may start to attract more of the community.
About growing the PMC, I suppose we're looping here ;-). IIUC, again,
I think the point is precisely to define those values/rules to be able
to induct more serenely new PMC members while asking them to adhere to
those definitions.
I think that, so far, this idea has not found much favour with the
broader community.
I am not sure how you get someone to promise not to undertake
enhancements that result in too much code or add too much functionality.
The actual wording of the proposed litmus test only indirectly addressed
this and seemed to be an attempt to codify a personality conflict with
rules that would hurt the long-term viability of the project by
replacing innovation with the group-think of an inner clique.
It also seemed to go against the principle of open source.
At least, that was the way it appeared to me as a community member
outside the PMC.
Ron
Cheers
--
Ron Wheeler
President
Artifact Software Inc
email: [email protected]
skype: ronaldmwheeler
phone: 866-970-2435, ext 102
