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.
01:04 benchable6 left, bisectable6 left, coverable6 left 01:05 nativecallable6 left 01:06 mst left 01:07 nativecallable6 joined, coverable6 joined, benchable6 joined, bisectable6 joined, mst joined 01:08 MasterDuke left 06:19 bartolin left
Woodi lizmat: but list and hashes are treadsafe - operations on them ? but in case of "metamodels" that probably can be treated as separate subsystem that is moustly used at boot time or pre-object-creation so can be moustly read-only, which could help with coping. and if metamodels can evolve then some locking protocol on them will be needed so they can stay predictibly r-o. checking version or lock on evry 07:44
usage would be bad for performance and metamodels don't need to be modified often
but I have no clue about details of implementation...
07:46 sena_kun joined
Geth MoarVM/main: a1a18e9842 | (Stefan Seifert)++ | src/6model/reprs/P6num.c
Fix uninitialized is_unsigned flag in P6num

A num without explicit size had an uninitialized is_unsigned flag causing at least flapping tests in NQP.
10:10 sena_kun left
Geth MoarVM/main: 6a9c523dd5 | (Daniel Green)++ | src/core/exceptions.c
Slightly optimize creating backtraces

Instead of using snprintf just to convert a number into a C string and then converting that C string into an MVMString, just use MVM_coerce_u_s to convert the number directly an MVMString.
MoarVM/main: 504a4d7b86 | MasterDuke17++ (committed using GitHub Web editor) | src/core/exceptions.c
Merge pull request #1790 from MasterDuke17/slightly_optimize_creating_backtraces

Slightly optimize creating backtraces
14:40 kjp left, leont left 14:41 leont joined, kjp joined 15:57 epony left 18:42 sena_kun joined
lizmat And yet another Rakudo Weekly News hits the Net: rakudoweekly.blog/2024/02/12/2024-07-dr-raku/ 19:35
19:44 epony joined 22:15 sena_kun left