installing pybigtools fails due to inability to compile ring v0.17.8

[dmalzl]

Hi,

This is probably a third party issue but I figured maybe you have a simpler solution for this here. I am currently trying to install pybigtools v0.1.2 (as required by another package) with Python version 3.11 but fail to do so as the maturin command fails at compiling ring v0.17.8. I am not familiar with Rust and it's ecosystem so I am kinda out of ideas here. I already tried to build ring from source but this also failed with a similar error. Did you experience something similar before? Could you maybe tell me what I need to do in order to get this to run?

Thanks in advance

[ghuls]

Unless you have a very uncommon architecture, you should be able to install pybigtools from a binary wheel:
https://pypi.org/project/pybigtools/0.1.2/#files

[dmalzl]

Thanks for the reply and the suggestion. After a day of troubleshooting and googling I think that this resulted from an incompatibility of openssl versions between Rust and Python > 3.11.x (not really sure though as I did not dig too deep here). In the end I downgraded the Python 3.11.10 I was using before to 3.11.0 and everything compiled as expected so everything good now. Main message: Looks like incompatibilities between some Python 3.11 requirements and Rust which is solved by simply using 3.11.0 (at least when installing with conda)

[jackh726]

So, I just tested and can definitely install pybigtools with a python 3.12 environment. (From source directly and using conda). It's difficult because actually building pybigtools shouldn't be affected by the python version.

There's going to be a lot that could be effecting this. To really dig into this, I'd at least need to start by getting the exact error(s) you were getting.

[dmalzl]

Hi, thanks for looking into this. I eventually got it to run and I think it was more about my OS than actually Python. I am working on CentOS7 which seems to have an incompatible gcc version. After updating gcc I could install it easily. So I guess it is really my fault here

[dmalzl]

Ah and sorry, about the error:
I don't remember exactly but it was some gcc related compilation error. I'll try to reproduce it and post it here