PMXBOT Log file Viewer

Help | Karma | Search:

#pypa logs for Wednesday the 20th of November, 2019

(Back to #pypa overview) (Back to channel listing) (Animate logs)
[09:21:24] <pombreda_> techalchemy: re "you can be slightly less patient because you can just wait for me to finish the updates i'm working on" --> ahahah!!
[15:57:15] <astronavt> hi pypa devs. does anyone know why `pip install .` would try to copy the entire current directory to /tmp, even when i have recursive-exclude set in MANFIEST.in? setup: https://bpaste.net/show/UGMLI manifest: https://bpaste.net/show/PIFWY
[16:02:15] <astronavt> this is a problem because my current project contains several very large data files that take a long time to copy and also take up too much space if duplicated
[16:02:24] <astronavt> i tried with --no-build-isolation and that didnt help
[16:02:43] <astronavt> using pip 19.3.1 and setuptools 41.6.0.post20191030
[16:13:39] <ngoldbaum> astronavt: pretty sure this is https://github.com/pypa/pip/issues/2195
[16:15:08] <astronavt> from 2014, :(
[16:15:14] <ngoldbaum> your MANIFEST.in would matter if "pip install ." was equivalent to building an sdist then installing that sdist
[16:15:17] <ngoldbaum> but it's not
[16:15:25] <astronavt> ah thats what i assumed it was tbh
[16:15:27] <ngoldbaum> so it doesn't matter if you exclude those files in the sdist, they'll still get copied
[16:15:52] <ngoldbaum> you should definitely still exclude those files in the sdist though
[16:16:03] <astronavt> indeed, but normally i dont need all this recursive-exclude stuff
[16:16:19] <astronavt> i just graft and set packages= and it works
[16:16:35] <astronavt> ive worked with projects like this before, its weird to me that its only a problem now... is there any workaround?
[16:17:02] <ngoldbaum> make a shell script that builds the sdist and installs that and use the shell script instead of "pip install ."?
[16:17:22] <ngoldbaum> if this is for local development, "pip install -e ." won't copy either
[16:17:30] <astronavt> ah. ok let me try that
[16:17:30] <ngoldbaum> i think
[16:17:36] <astronavt> i probably should be using -e anyway
[16:17:49] <ngoldbaum> yeah, assuming you know what it does
[16:19:17] <astronavt> that seems to work although now im having other problems with disk space on our undersized tmp partition
[16:19:22] <astronavt> ill pop back in here if it still fails