Welcome to the main channel on the development of MoarVM, a virtual machine for NQP and Rakudo (moarvm.org). This channel is being logged for historical purposes.
Set by lizmat on 24 May 2021.
00:07 reportable6 left 00:09 reportable6 joined 00:40 [Coke] left, [Coke] joined 02:10 [Coke] left 02:16 [Coke] joined 03:02 frost joined 05:25 ggoebel left 06:08 reportable6 left 06:09 reportable6 joined
Nicholas good *, * 06:22
It's "unfunny jokes all over the Internet" day 06:23
07:28 nebuchadnezzar left 09:16 evalable6 left, linkable6 left 09:17 linkable6 joined
jnthnwrthngtn Nicholas: No, that's tomorrow! 09:17
09:19 evalable6 joined 10:10 ggoebel joined 10:30 ggoebel left 10:34 Altai-man joined 11:45 MasterDuke left, linkable6 left 11:48 linkable6 joined 11:51 MasterDuke joined
Geth MoarVM: MasterDuke17++ created pull request #1696:
JIT atposref_u + getattrref_u
11:59 frost left 12:00 frost joined
lizmat ooh, sounds interesting! 12:02
12:07 reportable6 left 12:08 reportable6 joined
Geth MoarVM: b2a91e6c3e | (Daniel Green)++ | src/jit/graph.c
JIT atposref_u + getattrref_u
MoarVM: 9e9d303aa4 | MasterDuke17++ (committed using GitHub Web editor) | src/jit/graph.c
Merge pull request #1696 from MasterDuke17/jit_atposref_u_and_getattrref_u
lizmat MasterDuke: bump I guess, eh ? 13:06
[Coke] I like that the JIT of those was just "yes, just use the existing JIT, thank you."
13:06 frost left 13:11 TempIRCLogger joined 14:17 evalable6 left, linkable6 left, linkable6 joined 14:18 evalable6 joined 15:10 discord-raku-bot left 15:11 discord-raku-bot joined 15:51 Altai-man left
lizmat 8 minutes until Transporter 4 launch 16:17
nine jnthnwrthngtn: does RakuAST do module loading yet? 17:02
jnthnwrthngtn nine: Loading, yes, but not precomp (so you need to have had them precompiled using non-RakuAST) 17:03
nine jnthnwrthngtn: wrong! :D 17:07
gist.github.com/niner/0422e5ca8a00...73717f787b gets us through repository setup, precompilation and loading 17:08
Of course Test.rakumod would be a bit of a steep hill to climb next. Especially since we already stumble over the unit module declaration 17:09
Nicholas I see the date on that ... :-)
but I assume that this is a double bluff 17:10
nine My first impression is: RakuAST is really nice to create high level Raku ASTs, but makes it somewhat hard to create low level code like those proxy readers
And why don't we need to call the compunit here in the old QAST implementation?? gist.github.com/niner/0422e5ca8a00...-diff-L316 17:12
jnthnwrthngtn nine: I'm not immediately sure if the proxy reader should be done via RakuAST :)
nine Took me days to figure out that all it needed was that () at the end to get the actual pointy block that we wanted, not the comp unit's mainline
jnthnwrthngtn Oh. Hmm. :) 17:13
nine jnthnwrthngtn: I suspected that, but it's what came up (the missing optimize stage) and I figured it's at least a lesson in how RakuAST works
jnthnwrthngtn nine++ # getting into RakuAST now too :) 17:14
Was the hllize introduction purely for the sake of the proxy reader thing?
(I'm not really sure that concept should be part of the RakuAST API, is all...) 17:15
nine Yes 17:16
I assume I'll write quite a bit of code that will not end up anywhere because it violates the design. But then, I'm a total beginner in this area and most of the work is currently just learning. 17:25
jnthnwrthngtn :) 17:28
Feel free to ask anything about the design 17:29
I'm hopeful of having more time for working on this in the months ahead compared to the last however many
For now, though, dinner won't cook itself...bbl 17:31
18:06 reportable6 left 18:07 reportable6 joined 18:08 sena_kun left 18:14 sena_kun joined 19:09 ggoebel joined
MasterDuke jnthnwrthngtn: did my reply (and change) in github.com/MoarVM/MoarVM/pull/1690 make sense? 19:15
19:46 sena_kun left 19:54 sena_kun joined 20:03 Kaiepi left
Geth MoarVM: c77e1623dd | (Daniel Green)++ | 2 files
Mimalloc should only release memory to the OS at exit if --full-cleanup was used
MoarVM: 35453e9e16 | (Jonathan Worthington)++ (committed using GitHub Web editor) | 2 files
Merge pull request #1690 from MasterDuke17/mimalloc_should_not_release_memory_to_os_on_exit_unless_full_cleanup_is_used

Mimalloc should only release memory to the OS at exit if --full-cleanup was used
nine Just pushed my first commits to rakuast :) 21:04
With that, I think I've earned myself some Doom before going to bed ;)
MasterDuke i was just reading about the new raytracing mod for doom. i finished hl: blue shift a couple days ago, was going to start hl2 in a couple min. getting distracted by this project euler problem though... 21:09
21:33 Kaiepi joined