cfbolz changed the topic of #pypy to: PyPy, the flexible snake (IRC logs: https://botbot.me/freenode/pypy/ ) | use cffi for calling C | the secret reason for us trying to get PyPy users: to test the JIT well enough that we're somewhat confident about it
<wondiws>
simpson, well, my build works :)
jamesaxl has quit [Quit: WeeChat 2.2]
<wondiws>
but needless to say it all goes south when I copy it to /usr/local/bin
<simpson>
Hm. I'm not sure why that's the case. How do things break, exactly?
<wondiws>
simpson, it expects the original tree as in the source, but I just copied the entire tree to /usr/local/lib, and made a symlink, so I'll keep that for now
<tos9>
you can't relocate that binary without rewriting things in it.
<tos9>
(which I am fairly sure is in the build docs somewhere)
<tos9>
a symlink is reasonable, or IIRC there might be a cmdline flag to build the binary slightly differently.
<njs>
I would certainly hope that both gender-neutral people and gender-neutral language are both welcome here
johanfforsberg has quit [Ping timeout: 264 seconds]
lritter has quit [Remote host closed the connection]
<kenaan>
arigo py3.6 b72ee387ca84 /pypy/objspace/std/: Merged in jKorvin/pypy/fix-float-deprecation-warning (pull request #618) Add deprecation warning if type of result o...
<kenaan>
Evgenii Gorinov fix-float-deprecation-warning 817e823be049 /pypy/objspace/std/: Add deprecation warning if type of result of __float__ is float inherit class
<kenaan>
arigo py3.6 77bcaf8bbc5f /pypy/interpreter/: Merged in raduciorba/pypy/py3.6 (pull request #619) fix async generator bug when yielding a StopIteration
<kenaan>
raduciorba py3.6 d890ba3db546 /pypy/interpreter/test/test_coroutine.py: add failing test for yielding a StopIteration from an asyncgen
<kenaan>
raduciorba py3.6 aac246060deb /pypy/interpreter/generator.py: fix async generator bug when yielding a StopIteration
Gonsor has joined #pypy
raynold has joined #pypy
speeder39 has quit [Quit: Connection closed for inactivity]
<cfbolz>
vstinner: cool, fingers crossed that this can go somewhere
<vstinner>
cfbolz: i have a plan!
speeder39 has quit [Quit: Connection closed for inactivity]
<simpson>
vstinner: +1, tough road ahead, good luck.
<simpson>
Unrelated: What happens if I have an @elidable function that is super-small? Can it be inlined? Does the @elidable vanish entirely after inlining?
<simpson>
I'm emitting RPython code using Cozy (https://github.com/CozySynthesizer/cozy) and Cozy supports small foreign snippets of code, with the caveat that the synthesizer will treat them as pure. I figured that, if they have to be pure, I might as well do the courtesy of @elidable on them!
<simpson>
So my hope is that @elidable still allows inlining, and that inlining removes the annotation entirely.
<vstinner>
simpson: my plan rely on a compatibility layer to keep "backward compatibilty" for the *new* API, *and* also collaboration with C extensions maintainers (maybe even propose patches)
<cfbolz>
simpson: no, elidable prevents inlining
<cfbolz>
simpson: anyway, elidable is mostly useful if the function contains operations that don't look pure
<cfbolz>
Which is not the case in your example
<simpson>
cfbolz: Hm. So maybe I should textually inline instead, and the user can pick up the pieces?
<vstinner>
but i will not say that it's an easy plan. i waited one year to propose it "officially" :)
<vstinner>
a good news is that my manager supports this plan ;)
<xorAxAx>
cool!
<vstinner>
FYI i'm now paid to maintain Python for Red Hat
<cfbolz>
simpson: why not leave of the annotation?
<cfbolz>
vstinner: cool!
<simpson>
cfbolz: Ah, yeah, that's a good plan.
<cfbolz>
simpson: is there some info how cozy works? Sounds interesting