[19:55:54] <palate_> I am trying to embed c++ code in a python package using manylinux, and I'm facing issues there
[19:56:11] <palate_> And with them, I get to wonder if manylinux is a good solution or if it is EOL.
[19:56:29] <palate_> Aren't there security implications in building c++ on such an old system as CentOS5?
[19:57:11] <ngoldbaum> palate_: there are ways to get a new C++ stdlib in
[19:57:35] <ngoldbaum> there are also plans for a new manylinux distribution but as in all things python packaging it's blocked on people having time to work on it
[19:58:12] <sumanah> di_codes1: ^ do you have some thoughts on the manylinux thing?
[19:58:45] <ngoldbaum> palate_: hints in this issue https://github.com/pypa/manylinux/issues/118
[20:18:01] <palate> ngoldbaum: at least it needs the header
[20:18:10] <palate> ngoldbaum: and I don't feel like embedding those headers in my project
[20:19:01] <palate> ngoldbaum: so building depends on TCGETS2, TCSETS2, BOTHER and a struct called "termios2". Then I guess running depends on the kernel, and I don't know what happens if it is not supported
[20:19:19] <palate> ngoldbaum: but manylinux doesn't seem to have those headers
[20:22:40] <ngoldbaum> it's just that the people who have bandwidth to work on this are geographically distributed and are not being paid to work on it