Fire is step THREE! | github.com/perl6/toolchain-bikeshed | Channel logs: irclog.perlgeek.de/perl6-toolchain/today | useful prior art: metacpan.org/pod/CPAN::Meta::Spec
Set by moderator on 6 March 2016.
02:11 FROGGS_ joined 02:49 ilbot3 joined
moderator Fire is step THREE! | github.com/perl6/toolchain-bikeshed | Channel logs: irclog.perlgeek.de/perl6-toolchain/today | useful prior art: metacpan.org/pod/CPAN::Meta::Spec
03:24 sevvie joined 07:17 FROGGS joined 07:25 domidumont joined 07:30 domidumont joined
nine ugexe: the precomp files don't necessarily have to be in the same repo. Makes me wonder if it makes any difference when we delete them. We have to be able to deal with stale precomp files anyway. 08:56
10:22 FROGGS joined 12:27 pnu joined 13:11 sufrostico joined 13:39 FROGGS joined
ugexe right. but if someone wants to delete them off their system there should be a way to do it not just a way for perl6 to ignore them. 17:00
on a different note: precomps that are built in $*CWD/lib/.precomp don't get put into a `/<rakudo commit id>` sub directory. so switching rakudos to test a local module will end up in failure many times 17:03
if it were to follow the examples of site/home and put into `lib/precomp/<rakudo commit id>` i imagine this would not be as rampant a problem 17:05
17:48 sufrostico joined 18:13 Kassandry joined 20:50 patrickz joined
nine Are you sure? There's only one precomp store implementation and it stores in <compiler-id>/<first-two-chars-of-precomp-id>/<precomp-id> 22:19
The compiler-id changes with every rakudo compilation. Switching between compilers should be very safe. 22:20
I actually would very much like to get rid of the rakudo commit id subdirectories in .perl6 as it prevents you from sharing module installations between compiler versions. 22:21
You can already share (or continue to use after upgrading rakudo) dists installed into perl, vendor and site but not those in home.
ugexe i have no <compiler-id> directory under `<project-name>/lib/.precomp` 22:38
it has lib/.precomp/<some sha1>/<first 2 chars> 22:39
i guess it must be hashed into that sha1, but why do that instead of following the other usage which uses the un-hashed compiler-id as the directory? 22:41
nine It looks like E83EB6D42A841EA75E462EFBD204D9B69BB05826.1457281436.61762 which is the sha1 of the compiler id followed by the timestamp of compilation 22:57
The name that's used in .perl6 is not the compiler id but the compiler version which is not unique enough as precomp files really are tied to the perl6 binary that they were built with 22:59
23:01 stmuk joined
ugexe do you happen to know what situations muddy up lib/.precomp? it seems like sometimes i end up running old code until i delete lib/.precomp 23:11
23:54 pnu joined