bug#18935: gexp->derivation test failure
Mark H Weaver skribis: > The same failure happened to me in two consecutive build attempts: > once when running "guix system build", and then again when running > "guix build -K guix" so that I could get the failed test logs. Oh, that’s surprising. I’ve not stumbled upon it in my working tree, and it has not failed on Hydra: http://hydra.gnu.org/job/gnu/master/guix-0.7.14e84b2.i686-linux/all http://hydra.gnu.org/job/gnu/master/guix-0.7.14e84b2.x86_64-linux/all http://hydra.gnu.org/job/gnu/master/guix-0.7.14e84b2.mips64el-linux/all Ludo’.
bug#18935: gexp->derivation test failure
l...@gnu.org (Ludovic Courtès) writes: > Mark H Weaver skribis: > >> The same failure happened to me in two consecutive build attempts: >> once when running "guix system build", and then again when running >> "guix build -K guix" so that I could get the failed test logs. > > Oh, that’s surprising. I’ve not stumbled upon it in my working tree, > and it has not failed on Hydra: Update: the same failure has happened to me in *five* consecutive build attempts. I guess I'll have to add another local patch (to update guix in guix) to get my system updated. Thanks, Mark
bug#18831: Error handling in guix import?
Thanks for the report and analysis. This is fixed in 467a3c9. Ian Denhardt skribis: > starting download of `/tmp/guix-file.5J1eOy' from > `https://pypi.python.org/pypi/schema/json'... > ;;; Failed to autoload make-session in (gnutls): > ;;; ERROR: missing interface for module (gnutls) > ERROR: In procedure module-lookup: Unbound variable: make-session As for this part, note that GnuTLS is actually required here, because pypi.python.org is only accessibly over HTTPS here. We’ll make that clear in the manual. Ludo’.
bug#18935: gexp->derivation test failure
Mark H Weaver skribis: > l...@gnu.org (Ludovic Courtès) writes: > >> Mark H Weaver skribis: >> >>> The same failure happened to me in two consecutive build attempts: >>> once when running "guix system build", and then again when running >>> "guix build -K guix" so that I could get the failed test logs. >> >> Oh, that’s surprising. I’ve not stumbled upon it in my working tree, >> and it has not failed on Hydra: > > Update: the same failure has happened to me in *five* consecutive build > attempts. Then either I was mistaken and there’s another bug hiding, or scheduling and I/O are very much deterministic. You’re not using --cache-failures, are you? > I guess I'll have to add another local patch (to update guix in guix) > to get my system updated. Then could you try adding just applying 834ea02? TIA, Ludo’.
