04:11 MasterDuke left 07:52 sena_kun joined
ab5tract say I've added a new field in a struct (`MVMIOFileData`): how do I actually bubble this new information up into NQP / Rakudo? 15:49
lizmat Bump MoarVM in NQP and then Bump NQP in Rakudo ? 15:50
ab5tract My guess is that a new function is required in MoarVM, which is then exposed to QAST via MAST, and then a new nqp::* on top to access it?
lizmat: if it's that simple I will be very happy indeed :D
lizmat I guess nqp::filemode, looking at nqp::fileislink / filereadable / filewritable implementations ? 15:51
ab5tract excellent 15:55
a bit surprised that none of those take BOOTIO, but they're great starting points
nine New syscall most likely 16:14
That will be available in Rakudo directly with no NQP changes required 16:15
See stat-is-readable and friends 16:17
ab5tract ah, it looks like the issue at hand (R#1373) can be solved in a simpler manner without leaving MoarVM... but the question is, do we want a way for HLLs to access the requested open modes?
linkable6 R#1373 [closed]: github.com/rakudo/rakudo/issues/1373 [LTA] LTA error when trying to exclusive-lock a read-only IO::Handle
ab5tract oops, nevermind. it requires a combination of the mode + the arguments to lock, so the mission continues. thanks nine for the suggestion 16:20
lizmat and yet another Rakudo Weekly News hits the Net: rakudoweekly.blog/2024/05/13/2024-...nnouncing/ 17:23
jnthn nine: It's short for `class Foo { my $foo; method foo() { $foo } }` iirc 18:11
Geth MoarVM: ab5tract++ created pull request #1803:
Provide some mechanisms for introspecting file mode
18:13
ab5tract it's a bit surprising to me that we don't have an `MVMbool` type that could be backed by a short (or C99's bool) 21:01
feels a bit weird to stuff a single bit of information into an MVMint64, but maybe I'm holding something wrong? 21:02
(the sort-of-pun in that sentence was entirely unintentional :) 21:03
lizmat could be very well that jnthn never felt the need for it 21:10
23:40 sena_kun left