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 21 January 2016.
00:12 leont joined 07:04 domidumont joined 07:11 domidumont joined 08:19 hankache joined 08:38 hankache joined 11:36 camelia joined 12:37 leont joined 16:11 hankache joined
nine_ ugexe: regarding ci.appveyor.com/project/ugexe/zef/...0.13#L1120 I wonder why the precomp file is still opened by any process. Loading a module should just open the file, read it and close it again. As at least currently only one process can load modules from a repo at any given point in time, I don't see why deleting files should conflict. 17:32
tony-o_ nine_: i was looking through the source yesterday, it looked like that would be more easily solved if we could read the module in memory and load from that but it needs apis exposed from the vm. thats what it looked like, anyway. otherwise its reading and holding the file 18:30
i may be misremembering that, planes are exhausting
ugexe due to it not being reproducible locally it almost seems like a race condition between that delete precomp folder stuff and some type of file handle cleanup (maybe specific to windows) 18:31
18:44 leont joined
nine_ probably the file handle is only closed when the GC kicks in 20:16
leont Which is really something to avoid 20:30
I'm wondering if we should have a language construct that automatically closes something at the end of scope, much like a using block in C# 20:31
jdv79 you mean a phaser like CLOSE or LAST or something like that 20:38
i haven't used those yet though
i meant LEAVE
leont Yeah, but doing that well often requires creating a new scope 20:44
Or manually checking if the handle is defined
In fact, without that check it will go belly up if an exception is thrown during opening (because the handle isn't defined by definition) 20:45
21:22 hankache joined