<oriansj>Hagfish: well there is only one way to get digital providence right now and I doubt many people will be interested in doing 3+ years of work to just get to where we are now <Hagfish>but fortunately once provenance is established, it's established forever (although people are welcome to shorten or clarify or document the steps better) <oriansj>xentrac: the ability to prove that the source code actually corresponds to the binary <xentrac>right, but you said "providence", and I thought maybe you meant "provenance" <fosslinux>i actually don't use gentoo as my day-to-day os, i use void linux. i do like gentoo a lot though and just thought that it would be a good fit for something like this project <janneke>fosslinux: i'm lurking on #gentoo-x86, i haven't found a bootstrap- or reproducible-specific channel yet <janneke>fosslinux: they should change their name to void gnu/linux <amz3>At some point, I cross-bootstrap'ed a gentoo. That is, create a gentoo base root fs for another arch. It was not easy at all. <amz3>two weeks later, I could not figure what I have done in the first place. <janneke>amz3: that's a grand bootstrapping effort <stikonas>amz3: is it that hard to bootstrap gentoo once you have python3 (so that you can run portage)? <stikonas>I thought EPREFIX=/new_gentoo emerge --nodeps would just work... <stikonas>janneke: I'm now testing guix pull with your branch, so I think it works for me (which I guess was expected) <amz3>maybe modern gentoo has some niceties to do that work easier. To get started, one need a cross build toolchain, to compile things for another architecture. Then, there were various bugs (that I reported) in the package definitions of the software that are required by the seed. <stikonas>it started building M2-Planet now compiling gcc-4 <stikonas>maybe I should try to test that, with guix it should not be hard to get a container with toolchain, python and nothing else... <janneke>stikonas: i am always very happy when you test sommething that is `expected' to work :-) <janneke>it may easily take us a year to land this in guix; and much may have changed... <stikonas>still very exciting that you can build gcc from basically just source <stikonas>there should be some blog to spread the news :D <janneke>sure, but the first step is having someone else succeed; some review, etc. <oriansj>stikonas_: well we have the guix blog and janneke's but and hacker news is more likely to delete our articles than put it on the first page <stikonas_>well, guix blob posted that previous work on reduced bootstrap seed ***stikonas_ is now known as stikonas
<amz3>why hackernews will delete the articles? <ng0>because it's generally trash? <amz3>ng0: I tend to agree with you. I am trying https://lobste.rs... at the end of the day, it seems all those are a waste of time. <amz3>that said, it is a good infrastructure for free software, that is place to gather and share ideas with broader audience. Maybe that is also the purpose of gnu-system-discuss the misc discuss. <amz3>It is very difficult to get feedback, even on purposefully built communities like HN and lobsters. The best feedback and support, I found it here, in SRFI mailing list as part of my database work and also in GNU related mailling lists. <amz3>I already said that here I think.