Re: Basic questions about the triage process

[CHENG Gao]

*On Mon, 28 Dec 2015 17:21:22 -0800
* Also sprach John Wiegley <address@hidden>:

>>>>>> Lars Ingebrigtsen <address@hidden> writes:
>
>> Andrew Hyatt <address@hidden> writes:
>>> That sounds like a good idea for bugs of reasonable recentness. I'm
>>> going through old bugs that are years old now. To me, it feels a
>>> bit awkward to suddenly ask people to confirm anything after years
>>> have passed - just closing seems like a more reasonable approach to
>>> me. But I'll follow your advice for bugs in the last year. If you
>>> feel strongly that time elapsed shouldn't matter, though, I'm happy
>>> to do it your way all the time.
>
>> Yeah, if they're really old and aren't reproducing, closing them may
>> be the right thing to do.
>
> Andrew, I think your strategy is good, but can we turn that clock back
> to two years? Emacs doesn't move all that rapidly. If you can't
> reproduce something From 2013 or earlier, close it as cannot reproduce
> with a CC to the original reporter. Otherwise, ping the submitter with
> a CC to the bug address saying it can't be reproduced, but leave it
> open.

Maybe the strategy needs a clarification of version supporting policy or
should be based on said policy if it exists.

Say for example 24.x line supported while 23.x support dropped.

The priority could be as below:

Number one: emacs-25 branch related
Should have highest priority since they'll block the release.

Number two: emacs git
They slow down moving train.

Number three: emacs 24.x
Maybe a policy to include accumulated fixes in a new release untill
support dropped, for example yearly bugfix on Dec. 25 or Dec. 31.
Bug fixes only, no new feature backports.

Number four: emacs 23.x/22.x etc (justing kidding. No kidding?)
If bothered by "cannot sleep thinking users are abandoned in darkness"
syndrome, accept users submitted patches and release accumulated bugfix
minor version each year as above.

Maybe a cycling poicy can be built.

Just my RMB2 cents.