cfbolz changed the topic of #pypy to: PyPy, the flexible snake (IRC logs: https://quodlibet.duckdns.org/irc/pypy/latest.log.html#irc-end ) | use cffi for calling C | if a pep adds a mere 25-30 [C-API] functions or so, it's a drop in the ocean (cough) - Armin
jacob22_ has joined #pypy
jacob22 has quit [Remote host closed the connection]
oberstet has quit [Remote host closed the connection]
jvesely has quit [Quit: jvesely]
wleslie has joined #pypy
_whitelogger has joined #pypy
dddddd has quit [Ping timeout: 256 seconds]
jcea has quit [Quit: jcea]
tsaka__ has joined #pypy
tsaka__ has quit [Ping timeout: 272 seconds]
lritter has quit [Ping timeout: 260 seconds]
lritter has joined #pypy
lritter has quit [Quit: Leaving]
jvesely has joined #pypy
tsaka__ has joined #pypy
jvesely has quit [Quit: jvesely]
tsaka__ has quit [Ping timeout: 272 seconds]
oberstet has joined #pypy
<mattip> tos9: or any macOS dev: the macOS buildbot owner recently updated Xcode
<mattip> and now the bot cannot find libffi. Are there new instructions we should add to the build guide?
<mattip> do we need to install things via home brew?
lazka has joined #pypy
<lazka> mattip, brew install libffi; export PKG_CONFIG_PATH="$PKG_CONFIG_PATH:$(brew --prefix)/opt/libffi/lib/pkgconfig"
<cfbolz> heh, lazka appears and knows the right incantation
<lazka> :)
<mattip> thanks. Is that the only one or is more needed for translation
<lazka> :/ no idea, this is just how I use libffi for another project
<mattip> also what happens at runtime: do we need to ship libffi?
<mattip> maybe it is already handled in package.py
<kenaan> cfbolz record-known-result 5db9d6a810bb /: merge default
<kenaan> cfbolz record-known-result 452d076e43fa /rpython/: optimize int(str(i)) and same for rbigints. (Note that the other direction is not necessarily correct,...
<kenaan> cfbolz record-known-result e0ec664096c7 /rpython/translator/c/src/support.h: add missing op
<kenaan> arigo default d9468d86914f /pypy/: (cfbolz) graft 4c6c15e3e3bf to get the useful names "listview_ascii", instead of the confusing listview_utf8 on def...
<lazka> mattip, I think you need to bundle it, and adjust rpaths
<lazka> make_portable.py already mentions ffi, so maybe it "just works"?
<kenaan> cfbolz py3.6 3aabca3d88f7 /: merge default
<cfbolz> arigato: around?
tsaka__ has joined #pypy
tsaka_ has joined #pypy
tsaka__ has quit [Read error: No route to host]
dddddd has joined #pypy
tsaka_ has quit [Ping timeout: 272 seconds]
tsaka_ has joined #pypy
tsaka_ has quit [Read error: No route to host]
tsaka_ has joined #pypy
tsaka_ has quit [Read error: No route to host]
<mattip> cfbolz: it seems that translation on default is broken
<cfbolz> Ouch, sorry, on it
<mattip> ahh, it was arigato but thanks
<cfbolz> No
<cfbolz> It was me who grafted his commit
<cfbolz> (it's a mystery to me how py3.6 works with the same code)
dddddd has quit [Ping timeout: 260 seconds]
<kenaan> cfbolz default 57362e78ee31 /pypy/objspace/std/unicodeobject.py: fix translation (why does the same code work on py3.6?)
<cfbolz> mattip: done, sorry about that
<mattip> ¯\_(ツ)_/
lazka has quit [Remote host closed the connection]
oberstet_ has joined #pypy
oberstet has quit [Read error: Connection reset by peer]
tsaka_ has joined #pypy
tsaka_ has quit [Ping timeout: 272 seconds]
tsaka_ has joined #pypy
tos9 has quit [Ping timeout: 256 seconds]
tos9_ has joined #pypy
tos9_ is now known as tos9
jcea has joined #pypy
jcea has quit [Ping timeout: 256 seconds]
jcea has joined #pypy
tsaka_ has quit [Ping timeout: 272 seconds]
dnshane has quit [Quit: http://quassel-irc.org - Chat comfortably. Anywhere.]
dnshane has joined #pypy
<tos9> Hm. Nice. I tried to register on heptapod and it "blocked" me, so I gave up and decided not to bother, but now someone from there emailed to say they fixed it
<tos9> Hooray for decent monitoring.
tsaka_ has joined #pypy
jvesely has joined #pypy
dddddd has joined #pypy
speeder39_ has joined #pypy
<kenaan> cfbolz py3.6 7666a721d2fe /pypy/objspace/std/: slightly overengineered code to improve the performance of str.join (helps both the list in some situations and the ...
CrazyPython has joined #pypy
tsaka_ has quit [Ping timeout: 272 seconds]
<mattip> tos9: so now you can log in via the atlassian OAuth?
CrazyPython has quit [Read error: Connection reset by peer]
<mattip> picking up the sphinx benchmark profile again, I see that PyPy spends 21% of its time in _fast_traverse, CPython does not call it at all
<mattip> this is inside of docutils.nodes.traverse
oberstet_ has quit [Remote host closed the connection]
lritter has joined #pypy
ceridwen has joined #pypy
ceridwen has quit [Client Quit]
ceridwen has joined #pypy
ceridwen has quit [Changing host]
ceridwen has joined #pypy
ceridwen has quit [Client Quit]
ceridwen has joined #pypy
ceridwen has joined #pypy
ceridwen has quit [Changing host]
ceridwen has quit [Client Quit]
ceridwen has joined #pypy
ceridwen has quit [Changing host]
ceridwen has joined #pypy
ceridwen has quit [Remote host closed the connection]
ceridwen has joined #pypy
ceridwen has left #pypy [#pypy]
<mattip> the profiling info was misleading. Both call _fast_traverse the same number of times. On CPython the time is insignificant
<mattip> on cpython the time is 9 secs, on pypy 26 secs. It is a 6 line function, called ~17 million times
<tos9> mattip: GitHub in my case
<mattip> tos9: how did you previously log into bitbucket?
CrazyPython has joined #pypy
speeder39_ has quit [Quit: Connection closed for inactivity]
CrazyPython has quit [Ping timeout: 240 seconds]
<tos9> mattip: uhhh a bit of a mess of ways
<tos9> I had a bitbucket account I think before they offered any SSO
<tos9> and then they added Google SSO, and then GitHub SSO, and at some point I tried to be able to log in with those other things, but it basically didn't go very well
<tos9> So I think the only thing that works for me on bitbucket is straight normal u+pwd login
<mattip> ok. It would be nice to confirm that the atlassian OAuth integration works since that is the mapping between bitbucket user name and heptapod user name
<tos9> mattip: it's not just "up for grabs"?
<tos9> as in you can grab whatever username you want?
<tos9> it used my GitHub username when I logged in, I think it just defaulted to picking the username from the SSO you used to get in
<tos9> I can try logging in via atlassian too though and see what happens
<mattip> when we import the pypy issues and repo, the commit and issue user names will be the bitbucket ones.
<mattip> the only way to know for certain they map correctly is to use the atlassian logon OAuth
<mattip> although the real risk from someone "stealing" a user name is quite low
CrazyPython has joined #pypy
CrazyPython has quit [Read error: Connection reset by peer]
<tos9> ah, interesting, ok
<mattip> numpy may start to use metaclasses to define the dtypes
<mattip> the pr crashes on pypy, I think in cpyext.pyobject.fromref, when it checks isinstance(fromref(ref.ob_type), w_type)
<mattip> too tired, try again tomorrow
marky1991 has joined #pypy
marky1991 has quit [Remote host closed the connection]
marky1991 has joined #pypy
jacob22_ has quit [Quit: Konversation terminated!]
CrazyPython has joined #pypy