[Dbix-class] 5/5 Why "a stability tzar" is a ridiculous proposition

Matt S Trout mst at shadowcat.co.uk
Wed Oct 12 14:11:21 GMT 2016


On Wed, Oct 12, 2016 at 02:02:16PM +0100, Colin Newell wrote:
> I think it might be worth going back to frew's suggestion of the
> Collective Code Construction Contract and think about whether we can
> try to adopt the risk management into the general governance of the
> project.  We all seem to agree on what stability we want.  We are
> mostly arguing over how we achieve it.

I liked your PEP idea. Notably that initial proposal should require

(a) explanation of what the feature achieves
(b) next nearest possible code without said feature for comparison
(c) explanation of why it needs to be in core
(d) risk assessment

Of course, people less familiar with the codebase may not be able to
produce a fully accurate risk assessment at first attempt, but I think
requiring them to think about it anyway doesn't seem unfair.
 
> I would like it to be easy enough that a developer new to the project,
> without needing to get deep in irc communication or mailing lists, can
> contribute a feature/improvement with a reasonable expectation of it
> being accepted.  That's actually relatively rare with most open source
> projects, but I don't see why it needs to be.  If there is clear
> documentation regarding what is acceptable from a stability point of
> view, how to contact people when you want to ask questions or for
> opionions, and explaining what level of tests are required it is quite
> possible that they will be able to produce a reasonable change without
> much assistance.  By having the documented process it should also make
> it easier for the reviewers to complain about stability breaking
> things without needing to worry too much about how to put it, as the
> documentation should have the words.

That would be really nice. and I wonder if we can rig up some sort of
list<->github gatewaying ala p5p's list<->RT gatewaying to make sure
everybody can still be involved. Not sure.

Hopefully once the poo flinging stops and the list is clear to discuss
actual governance we can get into ideas like this in more detail.

-- 
Matt S Trout - Shadowcat Systems - Perl consulting with a commit bit and a clue

http://shadowcat.co.uk/blog/matt-s-trout/   http://twitter.com/shadowcat_mst/

Email me now on mst (at) shadowcat.co.uk and let's chat about how our CPAN
commercial support, training and consultancy packages could help your team.



More information about the DBIx-Class mailing list