00:05
yuyichao_ has quit [Ping timeout: 255 seconds]
00:20
yuyichao_ has joined #pypy
00:57
tbodt has joined #pypy
01:39
fryguybob has quit [Ping timeout: 268 seconds]
01:41
marr has quit [Ping timeout: 256 seconds]
02:13
fryguybob has joined #pypy
02:36
pilne has quit [Quit: Quitting!]
02:59
<
fijal >
lots of green, yay?
03:10
ArneBab_ has joined #pypy
03:15
ArneBab has quit [Ping timeout: 256 seconds]
03:15
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
03:25
tbodt has joined #pypy
03:46
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
03:52
dddddd has quit [Remote host closed the connection]
04:55
fryguybob has quit [Ping timeout: 256 seconds]
05:08
lazy1 has joined #pypy
05:15
tav` has joined #pypy
05:17
tav has quit [Ping timeout: 268 seconds]
05:17
tav` is now known as tav
05:22
lazy1 has quit [Quit: Leaving]
05:23
fryguybob has joined #pypy
06:37
<
kenaan >
arigo default c166d499b6d8 /pypy/module/thread/: Issue #2689 thread.Lock objects are weakrefable in CPython
06:40
tayfun26 has joined #pypy
06:49
<
kenaan >
arigo default ecedc8227476 /pypy/module/thread/: thread._local() should be weakrefable
06:49
<
kenaan >
arigo default cd23d604d4bc /pypy/module/struct/: struct.Struct() objects should be weakrefable
08:22
fryguybob has quit [Ping timeout: 265 seconds]
08:22
fryguybob has joined #pypy
08:43
marr has joined #pypy
09:03
commandoline has quit [Read error: Connection reset by peer]
09:08
commandoline has joined #pypy
09:09
exarkun has quit [Read error: Connection reset by peer]
09:11
asmeurer_ has quit [Quit: asmeurer_]
09:12
asmeurer_ has joined #pypy
09:18
exarkun has joined #pypy
09:31
asmeurer_ has quit [Quit: asmeurer_]
09:32
asmeurer_ has joined #pypy
10:41
antocuni has joined #pypy
10:47
yuyichao_ has quit [Ping timeout: 264 seconds]
10:51
exarkun has quit [Ping timeout: 264 seconds]
10:52
exarkun has joined #pypy
11:23
squeaky_pl has joined #pypy
11:29
antocuni has quit [Ping timeout: 264 seconds]
11:40
fryguybob has quit [Ping timeout: 240 seconds]
11:44
fryguybob has joined #pypy
11:47
dddddd has joined #pypy
11:48
<
squeaky_pl >
In the process of releasing PyPy 5.10 I discovered that ncurses can be built with utf-8 support which makes every ncurses SO get an extra letter "w" for wide I believe. This way you can have libncurses and libncursesw side by wide. But most distributions these days rename libncursesw to libncurses and ship in fact libncursesw.
11:51
<
squeaky_pl >
side by side*
12:10
trfl has joined #pypy
12:31
raynold has quit [Quit: Connection closed for inactivity]
12:37
yuyichao_ has joined #pypy
12:46
lud0r has joined #pypy
12:57
<
Rotonen >
fijal (and to whomever else it may concern):
https://pastebin.com/nyV62DA7 <- so this is what we're doing for cpython on macos for embedding into a sandboxed app targeting a range of macos versions building on the latest macos
13:00
Rhy0lite has joined #pypy
13:08
yuyichao_ has quit [Ping timeout: 256 seconds]
13:19
fryguybob has quit [Ping timeout: 240 seconds]
13:24
fryguybob has joined #pypy
13:35
antocuni has joined #pypy
15:07
lud0r has quit [Ping timeout: 268 seconds]
15:15
demonimin has quit [Quit: No Ping reply in 180 seconds.]
15:17
demonimin has joined #pypy
15:17
demonimin has joined #pypy
15:36
tayfun26 has quit [Quit: tayfun26]
15:58
antocuni has quit [Ping timeout: 272 seconds]
17:24
marky1991 has quit [Ping timeout: 256 seconds]
17:30
marky1991 has joined #pypy
17:59
yuyichao_ has joined #pypy
18:09
mat^2 has joined #pypy
18:12
Hotpot33 has quit [Max SendQ exceeded]
18:12
Hotpot33 has joined #pypy
18:12
Hotpot33 has quit [Max SendQ exceeded]
18:12
Hotpot33 has joined #pypy
18:13
Hotpot33 has quit [Max SendQ exceeded]
18:14
Hotpot33 has joined #pypy
18:18
Rhy0lite has quit [Quit: Leaving]
18:46
ceridwen has quit [Quit: Ex-Chat]
19:12
idnar has quit [Read error: Connection reset by peer]
19:13
sbauman has quit [Ping timeout: 252 seconds]
19:13
avakdh has quit [Ping timeout: 252 seconds]
19:13
mitsuhiko has quit [Ping timeout: 264 seconds]
19:13
phlebas has quit [Ping timeout: 264 seconds]
19:13
samth has quit [Write error: Connection reset by peer]
19:13
jeroud has quit [Ping timeout: 255 seconds]
19:14
ionelmc has quit [Ping timeout: 250 seconds]
19:14
DRMacIver has quit [Ping timeout: 264 seconds]
19:14
cfbolz has quit [Ping timeout: 240 seconds]
19:14
altendky has quit [Ping timeout: 272 seconds]
19:14
rguillebert has quit [Ping timeout: 272 seconds]
19:14
wallet42 has quit [Ping timeout: 272 seconds]
19:14
Hexxeh____ has quit [Ping timeout: 272 seconds]
19:14
pchiusano has quit [Ping timeout: 250 seconds]
19:14
_kcr has quit [Ping timeout: 250 seconds]
19:14
untitaker has quit [Ping timeout: 264 seconds]
19:14
krono has quit [Ping timeout: 240 seconds]
19:14
Alex_Gaynor has quit [Ping timeout: 240 seconds]
19:14
graingert has quit [Ping timeout: 240 seconds]
19:19
mattip has joined #pypy
19:23
jamesaxl has joined #pypy
19:31
<
mattip >
squeaky_pl: our cffi wrapper for ncurses builds a cffi-extension module which, at compile time, searches for first ncursesw then ncurses
19:31
<
mattip >
so we ship linked to whichever happened to be found on the buildbot
19:35
phlebas has joined #pypy
19:35
idnar has joined #pypy
19:37
jeroud has joined #pypy
19:39
<
squeaky_pl >
mattip, it is more complex. minimal_curses (the one you run in platform checks) doesnt look at ncursesw at all
19:40
<
squeaky_pl >
then when you search for cursesw you dont link to libpanelw,
19:40
altendky has joined #pypy
19:40
<
squeaky_pl >
the situation is much more complex then the cases you account for
19:41
<
squeaky_pl >
curses can be built in dozen of ways
19:41
_kcr has joined #pypy
19:41
<
squeaky_pl >
but I wont bother you too much since every distro does that differently
19:42
Hexxeh____ has joined #pypy
19:42
rguillebert has joined #pypy
19:43
<
mattip >
uhh, what is _minimal_curses?
19:43
<
squeaky_pl >
it's a check you run before translation
19:43
<
squeaky_pl >
i think it has something to do with libtinfo
19:43
<
squeaky_pl >
which is part of ncurses
19:44
avakdh has joined #pypy
19:44
<
mattip >
ok, and I see we have a pypy module by that name, but why does it exist?
19:44
<
squeaky_pl >
libpypy-c links to libtinfo
19:44
wallet42 has joined #pypy
19:45
<
squeaky_pl >
which is part of ncurses, it can be a separate shared object or not
19:45
fijal has joined #pypy
19:45
DRMacIver has joined #pypy
19:46
sbauman has joined #pypy
19:46
<
mattip >
we only link to it because we have a module called _minimal_curses, my question is why do we have that module by default?
19:47
<
mattip >
is it a left-over from the days before cffi-based _curses ?
19:47
<
squeaky_pl >
it might be, but I assumed it's because of libtinfo
19:48
graingert has joined #pypy
19:48
mitsuhiko has joined #pypy
19:48
<
mattip >
no, the other way around. We have libtinfo because we use _minimal_curses
19:49
<
squeaky_pl >
yes, that library is the database of terminals
19:49
<
mattip >
even the documentation says "don't use this"
19:50
<
squeaky_pl >
yes but you cannot translate without it
19:50
krono has joined #pypy
19:50
<
squeaky_pl >
i mean if the checks fail there translation is aborted
19:50
pchiusano has joined #pypy
19:51
cfbolz has joined #pypy
19:51
<
mattip >
you tried --withoutmod-_minimal_curses ?
19:52
<
squeaky_pl >
no, but since it's default on I though you do it on purpose
19:52
<
mattip >
it seems to be a mistake
19:52
<
squeaky_pl >
anyway I already patched it and i've been always building this like so and nobody complained
19:52
samth has joined #pypy
19:56
<
Rotonen >
less is more, take your diligence as a victory? :P
19:57
<
mattip >
got it, we use it in pyrepl. I wonder if we can use _curses these days?
19:57
<
squeaky_pl >
I vaguely remember arigato tried to fight this one once
20:00
<
mattip >
ok, so ncursesw vs. ncurses built into pypy is only an issue for pyrepl, "normal" curses users get the cffi-based implementation
20:06
Alex_Gaynor has joined #pypy
20:06
<
mattip >
more confusion. If we are linking to ncurses/curses and tinfo, why does only libtinfo.so show up in ldd libpypy-c.so ?
20:06
<
mattip >
just for the platform checks?
20:06
<
mattip >
s/^/is it/
20:13
<
mattip >
on to my next confusion
20:14
<
mattip >
why didn't we merge py3.5-mac-embedding into default, do the problems it solves not appear on pypy2?
20:15
tbodt has joined #pypy
20:16
jamesaxl has quit [Quit: WeeChat 1.9.1]
20:19
pilne has joined #pypy
20:23
<
mattip >
py3.5-mac-embedding adds download-and-build to create_cffi_import_library for _ssl (LibreSSL/libressl-2.6.2.tar.gz) , lzma, _gdbm
20:23
<
mattip >
on darwin, but only for py3.5
20:26
marky1991 has quit [Ping timeout: 256 seconds]
20:47
<
mattip >
I guess that would only help for the cffi-based _ssl in py3.5, the one in py2.7 is still rpython-based
20:56
mattip has left #pypy ["bye"]
21:29
untitaker has joined #pypy
21:32
ionelmc has joined #pypy
21:32
mat^2 has quit [Quit: Leaving]
21:43
ceridwen has joined #pypy
21:43
ceridwen has quit [Changing host]
21:43
ceridwen has joined #pypy
21:55
lesshaste has quit [Ping timeout: 248 seconds]
22:23
marky1991 has joined #pypy
22:37
marky1991 has quit [Ping timeout: 265 seconds]
22:56
<
njs >
do you use libressl instead of openssl? curious about why if so.
23:01
jcea has quit [Remote host closed the connection]
23:02
jcea has joined #pypy
23:02
ssbr` has quit [Ping timeout: 255 seconds]
23:08
jcea has quit [Ping timeout: 265 seconds]
23:16
ssbr` has joined #pypy
23:24
marky1991 has joined #pypy
23:31
yuyichao_ has quit [Ping timeout: 260 seconds]
23:41
marky1991 has quit [Remote host closed the connection]
23:41
marky1991 has joined #pypy
23:49
jcea has joined #pypy