Marko Lindqvist writes:
> On 28 January 2014 02:01, Jacob Nevins
> <0jacobnk....@chiark.greenend.org.uk> wrote:
>> Marko Lindqvist writes:
>>> When you need to know which weekend to reserve for the releasing?
>> I don't currently have any other constraints, so right now I think we
>> can decide pretty much up until the day (Saturday).
> Could you do it on Sunday?
I could, but it would be a bit tight. I'd have to finish by 1530 UT, so
I would have to kick things off on Sunday morning (giving all of
Saturday for testing).
(The release doesn't take much of my time, but the uploads take a while,
and have to complete before I can complete all the announcement
> I don't think I would be confident enough with the testing to release
> Saturday. While Sunday release might mean that Windows binaries are
> built week later than sources are released, this would still give us
> some chance to get 2.4.2 to Ubuntu LTS.
Indeed, this is desirable.
On the other hand, I was just writing an email that I'm starting to get
a little nervous about releasing this weekend, as we're getting a number
of scary-looking changes quite close to the end of the week --
bug #21523 (pathfinding; maybe I shouldn't be scared of this?),
bug #21547 (find_next_free_port(); scary because platform-dependent)
with little opportunity to find further regressions.
On balance, if at least one other person is doing organised testing on
Saturday, then I'm still content to go ahead, but if we find any more
last-minute issues I'm inclined to slip a week.
>> - the possible Gtk3 regression you mentioned
> I have not reproduced it on Win7, so it's probably XP specific. I
> would accept the corner-case that experimental gtk3-client does not
> work on almost-obsolete WinXP.
OK, will not treat this as a blocker.
(For what it's worth, I just tried starting a local game briefly with
cproc's testing/Freeciv-2.4.1-win32-gtk3-setup.exe on a WinXP machine,
and the graphics did not seem completely broken. So that forms some kind
of baseline, maybe.)
Freeciv-dev mailing list