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
marvin_ has quit [Remote host closed the connection]
marvin has joined #pypy
xcm has quit [Read error: Connection reset by peer]
xcm has joined #pypy
xcm has quit [Remote host closed the connection]
marvin has quit [Remote host closed the connection]
marvin_ has joined #pypy
Rhy0lite has joined #pypy
BPL has joined #pypy
lritter has joined #pypy
<petronny>
Hi, is rpi4 (4GB) powerful enough to translate pypy3?
<cfbolz>
petronny: to translate a 32bit version, probably. it might just take a while
<cfbolz>
64 bit, no
<petronny>
How about the 8GB one?
<arigo>
yes, 8GB is enough
<petronny>
Great, I'll buy one then.
<pmp-p>
rock64 is more powerfull and same price range
<pmp-p>
really sure it can't do it with only 4GiB ?
Dejan has quit [Read error: No route to host]
<petronny>
pmp-p: No, it can't. I just tested that.
<pmp-p>
:(
jcea has joined #pypy
lazka has quit [Quit: bye]
marvin_ has quit [Remote host closed the connection]
marvin_ has joined #pypy
lazka has joined #pypy
astrojl_matrix has joined #pypy
astrojuanlu has joined #pypy
<astrojuanlu>
hi folks, I'm at the SciPy US 2020 sprints trying to get something done: I have this proof of concept https://github.com/astrojuanlu/erfa-cffi that uses CFFI to call a C library. however, as you can see in the README I got it working by cloning the C source somewhere and using C_INCLUDE_PATH, LIBRARY_PATH and LD_LIBRARY_PATH. what would be the best
<astrojuanlu>
way to ship the C code in the way it's one with "usual" extensions?
<astrojl_matrix>
(hope messaging from Matrix also works)
astrojuanlu has quit [Remote host closed the connection]
xcm has joined #pypy
YannickJadoul has joined #pypy
<mattip>
astrojl_matrix: still around?
<astrojl_matrix>
mattip: yes!
<mattip>
hi
<mattip>
you either need to include the sources (easiest), make sure the library is available via some system installer, or include the compiled library
<mattip>
so vendoring the c source code in your repo via a submodule in git is one way
<mattip>
psycopg takes the "you must take care of getting libpq yourself" approach, as does the cffi fork
<astrojl_matrix>
I see - I'm more interested in the vendoring approach, the "bring your own library" is easier for the packager but more difficult for the user
<astrojl_matrix>
mattip: but opencv-python does not use cffi, does it? I don't see `cffi_modules` in their `setup()` call, or `cffi` anywhere in their repository
<astrojl_matrix>
(also, their `setup.py` is +400 lines long... I'll try to avoid that as much as I can)
<mattip>
I was just giving examples of approaches to bundling source code with your repo. Maybe I didn't understand the problem
<astrojl_matrix>
sorry, I didn't explain myself - I see how can I use `git submodule` to vendor the C sources, but I'm missing the link between how to have that vendored source next to the `setup.py`, and make cffi automatically locate the sources without having to use any of the "magic" environment variables. I thought there was some configuration option I was not seeing
<mattip>
you can pass any distutil kwarg to ffi.compile
<astrojl_matrix>
oh I see, so it's done in the `ffi.compile` call - I though it was supposed to be in the `setup` somehow