00:02
<
glyph >
hello pypy!
00:02
<
glyph >
my website is down because of this
00:02
<
glyph >
so I think pypy:3 is hosed
00:02
<
glyph >
I do not know what released binaries this corresponds to unfortunately
00:12
<
tos9 >
looks like it should likely be any, since yeah `lib_pypy/_crypt/__init__.py` has `import thread` in it even on the py3.6 branch
00:16
<
glyph >
tos9: well 'thread' is a stdlib module that should be there
00:16
<
glyph >
trying to figure out how to roll back locally
00:16
<
tos9 >
glyph: it's called _thread in py3
00:16
<
tos9 >
(which is what the issue is)
00:17
<
tos9 >
you should be able maybe to just rename it locally
00:17
<
tos9 >
and then the fix is trivial
00:18
<
tos9 >
but I don't know why the buildbot doesn't fail so someone else who knows things will have to chime in
00:18
<
tos9 >
as usual the test suite leaves... what to be desired from cpython, but there does seem to be one
00:20
* glyph
squints at his build server and wonders why it's rebuilding znc
00:26
<
glyph >
tos9: thanks for the fix!
00:26
<
glyph >
I'm just manually rebuilding the previous docker image locally and switching my FROM line
00:26
<
glyph >
I really need a self-test for these images before they go live
00:27
glyph has quit [Quit: End of line.]
00:27
glyph has joined #pypy
00:28
<
tos9 >
I'd merge it but yeah I'm not comfortable knowing why it doesn't fail so will let someone else do it
00:28
<
glyph >
OK, back up.
00:56
jcea has quit [Read error: Connection reset by peer]
00:58
jcea has joined #pypy
02:49
oberstet has quit [Remote host closed the connection]
02:55
jcea has quit [Ping timeout: 260 seconds]
03:11
Mortir has quit [Ping timeout: 256 seconds]
03:38
lritter has quit [Ping timeout: 272 seconds]
04:09
petronny has joined #pypy
06:53
Taggnostr2 has joined #pypy
06:55
Taggnostr has quit [Ping timeout: 260 seconds]
06:56
Taggnostr has joined #pypy
07:00
Taggnostr2 has quit [Ping timeout: 260 seconds]
07:14
ronan has quit [Remote host closed the connection]
07:14
ronan has joined #pypy
07:14
ronan is now known as Guest51626
07:35
Taggnostr has quit [Quit: Switching to single player mode.]
07:35
Taggnostr has joined #pypy
08:04
lritter has joined #pypy
09:18
lritter has quit [Ping timeout: 272 seconds]
09:25
_whitelogger has joined #pypy
09:49
oberstet has joined #pypy
10:07
nopf has quit [Remote host closed the connection]
11:06
<
arigato >
tos9: found out the reason for why test_crypt doesn't fail: that's because importing "crypt" fails, so the test
*skips*
11:09
<
arigato >
someone should try to make the "catch ImportError" in test.support.import_module() more precise
11:10
<
arigato >
something like checking
11:10
<
arigato >
that the error message contains ``'modulename'`` or something
11:12
<
mattip >
thanks for the fix
11:13
<
mattip >
another option would be to report the ImportError failures and filter the ones we expect to be gettng
11:13
<
arigato >
even better, yes
11:14
<
arigato >
I can try to do that
11:16
<
arigato >
the incremental solution would be to add an argument "must_be_present=False", which if True doesn't catch any ImportErrors, and start adding a "must_be_present=True" or "must_be_present=sys.platform in xyz" to tests
11:34
omasanori[m] has quit [Quit: Bridge terminating on SIGTERM]
11:34
the_drow[m] has quit [Quit: Bridge terminating on SIGTERM]
11:34
Orimendix has quit [Quit: Bridge terminating on SIGTERM]
11:34
astrojl_matrix has quit [Quit: Bridge terminating on SIGTERM]
11:35
toad_polo has quit [Quit: Bridge terminating on SIGTERM]
11:43
astrojl_matrix has joined #pypy
11:53
Mortir has joined #pypy
12:01
Orimendix has joined #pypy
12:01
omasanori[m] has joined #pypy
12:01
toad_polo has joined #pypy
12:01
the_drow[m] has joined #pypy
12:01
jacob22_ has quit [Read error: Connection reset by peer]
12:04
jacob22_ has joined #pypy
12:41
<
tos9 >
arigato: ahh :/
12:53
<
arigato >
mattip: fwiw, test_re fully works now (you looked at a pre-merge status before I fixed the last two remaining failures)
12:58
<
arigato >
and now I find out that import_module() has already grown a similar argument in py3.6
13:02
lritter has joined #pypy
13:49
<
mattip >
any thoughts on running py3.7 nightly instead of py3.6?
13:50
<
cfbolz >
works for me
13:50
<
mattip >
arigato: about test_re, it was just an example of something significant that our summary page does not really show
13:54
jcea has joined #pypy
13:55
forgottenone has joined #pypy
14:37
<
arigato >
mattip: right, yes
14:54
<
mattip >
on the other hand, showing all the stdlib failures would be quite sad,
14:54
<
mattip >
on py3.6 there are 20
14:55
<
mattip >
on py3.7 ~200
14:55
<
mattip >
and that is linux64, I don't even want to look at win32
14:56
Guest51626 is now known as ronan
14:58
<
mattip >
hmm, at least 34 are due to dec.decode, so more utf_8_decode mischief?
15:37
gsnedders has quit [Ping timeout: 246 seconds]
15:38
gsnedders has joined #pypy
16:12
jacob22_ has quit [Read error: Connection reset by peer]
16:16
jacob22_ has joined #pypy
16:27
marvin has quit [Remote host closed the connection]
16:28
marvin__ has joined #pypy
18:01
bbot2 has quit [Quit: buildmaster reconfigured: bot disconnecting]
18:02
bbot2 has joined #pypy
18:18
<
arigato >
branch 'import-module-must-be-present' killed, implemented much more simply as a single "extra_tests/test_pypy_modules.py" file
18:36
forgottenone has quit [Quit: Konversation terminated!]
18:39
<
arigato >
merged that in py3.6 and adapted the test, which found that there's an "_opcode" module that we should implement but don't
18:40
<
arigato >
needed because otherwise dis.stack_effect() is silently missing
18:41
<
arigato >
(NO CLUE why test__opcode.py doesn't just say "import _opcode" and tests it)
19:14
jcea has quit [Ping timeout: 260 seconds]
19:43
lritter has quit [Ping timeout: 260 seconds]
19:50
jcea has joined #pypy
20:04
<
mattip >
arigato: _opcode seems to have a single function _opcode.stack_effect(opcode, oparg=None, jump=None)
20:05
<
mattip >
with docstring "Compute the stack effect of the opcode", and internally it calls PyCompile_OpcodeStackEffectWithJump
20:13
<
mattip >
our pycompile.py seems to be built differently, I am not sure I see where we do something similar
20:14
jcea has quit [Remote host closed the connection]
20:26
xcm has joined #pypy
21:25
andi- has quit [Ping timeout: 272 seconds]
21:39
andi- has joined #pypy
23:13
Mortir has quit [Ping timeout: 246 seconds]
23:20
Mortir has joined #pypy