----- Original Message -----
> [...] I didn't bother create a stripped-down test case at the time.
Nor when asked repeatedly, nor now, nor tomorrow I'd imagine. While
trying to get these patches merged I got a vibe less "collaboration"
and more "steamroller". Let's continue to work on that - it's a two
way street, and the tests are very near and dear to me.
> (IMHO, this much discussion & revision is out of proportion [...]
I'm sorry you feel that way. I was trying to manage risk, and explain
what I was doing as clearly as possible, with a number of changes that
arrived unexpectedly in the lead up to a release.
I think we both have very different ideas about being conservative and
managing risk. The API breaking change, the lack of testing, a change
to a global variable affecting every PCP program, ever - for me, these
things represent high risk.
But we can manage those differences by
- ensuring planned release dates are well known in advance; (tick)
- whenever in doubt, or asked, provide test cases showing problems;
- continue refining HACKING guidelines to keep our thinking in sync;
- reviewing and merging each others code, and merging to dev early on
in each release cycle.
Let's do it better, together, next time. On that note, I'm away for
several days, recharging the batteries - see y'all in about a week!
cheers.
--
Nathan
|