Canary System for trunk

KaiRo - Robert Kaiser kairo at kairo.at
Sat Jun 19 00:02:12 UTC 2010


Ben Bucksch schrieb:
> On 19.06.2010 01:09, KaiRo - Robert Kaiser wrote:
>> Trunk is trunk is supposed to break and be fixed.
>
> Unfortunately, there's no way to know which m-c version was supposed to
> go with that older c-c version, that's why Dan and me pointed out.

The one "supposed" to go with it is the one from the default "branch" of 
the same point of time IMHO. Of course, that's not entirely well-tracked 
in hg, but with our usual practices the commit timestamps should fit in 
a usable way.

And if an m-c checkin breaks us, that's IMHO nothing we should hide or 
be able to ignore reasonably, but something we just need to get fixed 
ASAP. And yes, I hate breakages from other sources as well, but I've 
grown used to them, as almost all SeaMonkey breakages are either 
Thunderbird's or the platform's fault, and I'm happy we are bound to 
deal with them sooner rather than later, because we've tried the "later" 
solution and burned our fingers massively.

Robert Kaiser



More information about the tb-planning mailing list