00:29
oberstet has quit [Remote host closed the connection]
01:44
_whitelogger has joined #pypy
01:49
epony has quit [Quit: system upgrades again]
02:08
jcea has quit [Ping timeout: 240 seconds]
03:09
jacob22_ has quit [Read error: Connection reset by peer]
03:10
jacob22_ has joined #pypy
03:37
epony has joined #pypy
04:14
lritter has joined #pypy
04:41
todda7 has joined #pypy
04:47
_whitelogger has joined #pypy
04:54
mjacob has joined #pypy
05:42
oberstet has joined #pypy
05:52
mjacob has quit [Quit: leaving]
06:16
glyph has quit [Quit: End of line.]
06:16
glyph has joined #pypy
06:29
_whitelogger has joined #pypy
06:49
todda7 has quit [Ping timeout: 272 seconds]
07:19
jacob22_ has quit [Read error: Connection reset by peer]
07:22
bbot2 has quit [Quit: buildmaster reconfigured: bot disconnecting]
07:22
jacob22_ has joined #pypy
07:22
bbot2 has joined #pypy
07:26
<
mattip >
hmm, not so simple to have 32/64 bit windows buildbots: they want to build the same target
07:26
<
mattip >
more logic needed?
07:28
<
mattip >
for this build, pausing the 32-bit buildbot should be sufficient
07:40
todda7 has joined #pypy
07:58
rubdos has joined #pypy
07:59
rubdos_ has quit [Ping timeout: 240 seconds]
08:39
bbot2 has quit [Quit: buildmaster reconfigured: bot disconnecting]
08:40
bbot2 has joined #pypy
09:07
bbot2 has quit [Quit: buildmaster reconfigured: bot disconnecting]
09:07
bbot2 has joined #pypy
09:09
<
mattip >
heh, there were win64 builds in 2012
09:47
Taggnostr has quit [Ping timeout: 240 seconds]
10:02
_whitelogger has joined #pypy
11:02
<
mattip >
there is some magic missing in the win64 branch to allow virtualenv to work correctly
11:02
<
mattip >
see the buidbot failures
11:02
<
mattip >
somehow it is not getting the correct sys.path set up at initialization
11:54
phlebas has quit [Excess Flood]
11:54
phlebas has joined #pypy
12:11
Taggnostr has joined #pypy
12:50
mattip has quit [Ping timeout: 264 seconds]
12:52
mattip has joined #pypy
13:14
lritter has quit [Ping timeout: 264 seconds]
15:17
jcea has joined #pypy
15:25
<
mattip >
let's try that again, this time after installing cffi
15:47
<
mattip >
nope. Need to build cffi
16:15
ambv has joined #pypy
16:58
oberstet has quit [Quit: Leaving]
17:34
<
mattip >
it seems the win64 branch needs to build cffi from source when running tests
17:35
<
mattip >
it is being installed via a "requirements.txt" to a virtualenv
17:35
<
mattip >
is there a way to make "--no-binary cffi" conditional in requirements.txt?
17:43
jacob22_ has quit [Read error: Connection reset by peer]
17:46
jacob22_ has joined #pypy
18:34
ambv has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
18:41
ambv has joined #pypy
18:44
<
mattip >
the own win64 cpyext tests are failing because the tests are re-creating the bridge API dll over and over again
18:44
<
mattip >
instead of reusing the already-created dll
18:45
<
mattip >
and on windows the linker cannot open the dll because it is has been loaded into the process
18:45
<
mattip >
why doesn't this happen one the default branch with win32?
18:53
<
mattip >
( I guess we should just skip the cpyext and cppyy tests on that branch for now and try again for python3)
19:47
<
mattip >
pytest 6.1.1 added the collect_ignore_glob global variable that allows skipping globs of files, which would be handy to skip cpyext, cppyy
19:53
nulano has joined #pypy
20:25
JStoker has quit [Ping timeout: 244 seconds]
20:27
speeder39_ has joined #pypy
20:27
JStoker has joined #pypy
20:27
JStoker has joined #pypy
20:27
JStoker has quit [Changing host]
20:31
mjacob has joined #pypy
21:05
<
mattip >
nice progress with win64, I think it is almost ready for merging
22:37
speeder39_ has quit [Quit: Connection closed for inactivity]
22:40
<
nulano >
there are a few failures on win32, I think I can take a look at them tomorrow
23:41
_whitelogger has joined #pypy