[12:42:38] <Crys> Hi, this is Christian Heimes from Python core and Python Security Response Team. We just had somebody reporting already known issues with passwords in ~/.pypirc and "sudo pip install can destroy systems" on the security mailing list.
[12:44:21] <Crys> We told the reporter that these issues are known for years and that the reporter is free to talk about them. His reply sounded like he might write an angry blog post about pip.
[12:45:10] <Crys> I would like to forward the mails to some pip core devs and give you a chance to make a public statement in advance.
[14:07:25] <ngoldbaum> angry blog posts, the horror
[14:12:27] <Crys> He is now complaining about the fact some other core dev wrote a lengthy explanation and included him in CC.
[14:15:28] <mgedmin> will there be a website with a logo and a cute name?
[14:17:50] <Crys> I wanna have the shirt! And the plushy toy!
[15:34:11] <tos9> Crys: (FWIW probably you know about keyring support for pypirc?)
[15:34:41] <tos9> I think that the ticket asking to remove suggesting putting passwords in pypirc plaintext even got some progress recently
[15:42:01] <Crys> tos9: yes, the initial mail referred to https://github.com/pypa/packaging.python.org/issues/297 and https://github.com/pypa/pip/issues/4575
[15:57:13] <techalchemy> I mean it's not a bad idea
[16:12:22] <cooperlees> Anyone got a link - I'm another who needs to remove password.
[16:13:05] <cooperlees> Is token uploads done? Or am I dreaming that.
[16:13:59] <techalchemy> cooperlees, yeah that's done but I'm not sure if it's still in beta
[16:14:37] <techalchemy> someone may need to flip a switch or something
[16:14:51] <techalchemy> you may need to flip it yourself if you have access :p
[16:15:10] <cooperlees> I have 0 access in Python lands :(
[16:19:52] <techalchemy> yeah i was at a security discussion with some other packaging folks from some other langauges / orgs and we seem to actually be pretty advanced on this front
[16:20:43] <techalchemy> maybe mid-tier on the package signing front because we are playing with distro maintainers
[16:21:53] <cooperlees> Hopefully my donated Facebook money will make that better H1 next year :D
[16:25:13] <techalchemy> cooperlees, interesting, I also just joined the Canonical Security Team in a partly sponsored role to work on packaging related things
[16:27:30] <cooperlees> Love the "Security history"
[16:29:31] <techalchemy> cooperlees, if you do know anyone who is looking and you'd recommend I don't know if we are actually hiring but I think overall the team is growing and favors experienced developers
[16:29:39] <techalchemy> i guess security tends to be that way
[16:30:15] <cooperlees> I'll keep an eye out. I may be next year. Been here 7 years in January ... crazy.
[16:30:33] <techalchemy> i was at my last role for 5
[16:30:40] <techalchemy> i just kept checking that they were paying me market value
[16:34:53] <cooperlees> WE always use latest and greatest things ... Which is nice.
[16:34:58] <techalchemy> that's so different, such few places like that
[16:35:11] <techalchemy> everywhere just gets stuck in an old thing that sucks usually
[16:35:18] <techalchemy> but they're too scared to migrate forward
[16:35:21] <cooperlees> We have legacy, don't get me wrong.
[16:35:34] <cooperlees> But very few places try move out of it as aggressively
[16:36:01] <techalchemy> it's hard to do cost benefit on the tradeoffs there
[16:36:14] <techalchemy> there's a tipping point though and it's probably a lot earlier than most people think
[16:38:35] <cooperlees> We're employed at a "tech company" so if it makes sense and we get benefits from a new technology etc. we should be using it ...
[16:39:00] <cooperlees> You do get to focus deep on your part of the product / infra here
[16:40:23] <techalchemy> my last job i could kind of do whatever i decided was important which was nice
[16:41:21] <techalchemy> but in the same way it seems like i might need to learn a few things pretty deeply here because canonical's security team needs to be able teo make statements about stuff and not be wrong etc
[16:41:46] <cooperlees> Test and POC - Never assume.
[16:42:38] <techalchemy> jus take everyone at their word, all the best code gets written that way