| ab5tract | ugexe: regarding meta6.json: this is mixing of "list of either strings or objects" makes parsing in a static language kind of annoying gist.github.com/ab5tract/8a491c02f...0e3654105f | 08:40 | |
| Geth | rakudo/main: ccf01a599b | (Elizabeth Mattijsen)++ | t/04-nativecall/06-struct.t Add test for #3851 |
09:04 | |
| roast: 6e71e5575f | (Elizabeth Mattijsen)++ | S06-operator-overloading/infix.t Add test for #3895 |
09:21 | ||
| rakudo/main: 5680202e28 | (Elizabeth Mattijsen)++ | src/core.c/Int.rakumod Make Int ** Int up to 15% faster By splitting off the <0 >=0 cases into implementation-detail methods so that inlining becomes possible for the separate cases. In response to github.com/rakudo/rakudo/issues/3910 |
10:12 | ||
| rakudo/main: fa370cb8e2 | (Elizabeth Mattijsen)++ | src/core.c/Int.rakumod Simplify native coercers Making them up to 50% faster. Sadly .uint and .uint64 still need the workaround in the legacy grammar. Once 6.e is out, the workaround can be removed |
10:51 | ||
| rakudo/main: b7d1089561 | (Elizabeth Mattijsen)++ | t/12-rakuast/xx-fixed-in-rakuast.rakutest Add test for .uint|uint64 workaround As these appear fixed in RakuAST, most likely due to absence of static optimizations |
11:08 | ||
| roast: 36b4414151 | (Elizabeth Mattijsen)++ | S06-multi/type-based.t Add test for #3920 |
11:17 | ||
| rakudo/main: 498b91427a | (Elizabeth Mattijsen)++ | t/12-rakuast/xx-fixed-in-rakuast.rakutest Add test for #3935 |
11:25 | ||
| roast: d8f0ad76cb | (Elizabeth Mattijsen)++ | S02-types/signed-unsigned-native.t Add tests for #3936 |
11:32 | ||
| roast: 34ac62e0eb | (Elizabeth Mattijsen)++ | S32-exceptions/misc2.t Added test for #3949 |
11:39 | ||
| nqp/main: 026d4c0d7e | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | src/HLL/Compiler.nqp Allow --repl-mode even with parameters As a start of providing a fix for github.com/rakudo/rakudo/issues/3952 |
11:47 | ||
| nqp/main: 724ab6a912 | (Elizabeth Mattijsen)++ | src/HLL/Compiler.nqp Make --repl-mode truly leading 026d4c0d7ee5812da only checked --repl-mode if there were command line arguments (other than "-"). This broke all of Rakudo's REPL tests |
12:45 | ||
|
12:54
MasterDuke left
|
|||
| Geth | rakudo/main: 40cbb1944e | (Elizabeth Mattijsen)++ | tools/templates/NQP_REVISION Bump NQP to make --repl-mode leading As part of fixing #3952 |
13:09 | |
| nqp/main: 04eab468db | (Elizabeth Mattijsen)++ | src/HLL/Compiler.nqp Give interactive mode also access to command-line arguments |
13:28 | ||
| ugexe | ab5tract: github.com/Raku/problem-solving/issues/394 | 13:32 | |
| Geth | rakudo/main: 2825646ae3 | (Elizabeth Mattijsen)++ | 4 files Allow CLI arguments in the REPL with --repl-mode=interactive This commit: - Bumps NQP to get "interactive" method to receive CLI args - Makes sure the default REPL input is from STDIN - Adjust help message: only -e overrides --repl-mode - Add test for REPL with additional CLI arguments visible in @*ARGS Fixes #3952 |
14:09 | |
| rakudo/main: 91f2f3c19a | (Elizabeth Mattijsen)++ | t/12-rakuast/xx-fixed-in-rakuast.rakutest Add test for #4062 |
14:30 | ||
| tbrowder | lizmat: i have successfully added basic tests for log2 and, thanks to timo, have those passing spec testing. however, i have NOT added tests for the ENTIRE domain of invalid entries, including entries for numbers x <= 0, and the undefined domain for complex numbers. | 15:00 | |
| those will take more time and most likely smarter math ppl. BUT i do have some good links to excellent Wikipedia articles to put in a new or updated issue. | 15:03 | ||
| so, i can clean up what i have for the incomplete PR, or leave as is for more work. | 15:04 | ||
| my desire is to submit the PR as it is with appropriate caveats. | 15:05 | ||
| lizmat | tbrowder++ | 16:50 | |
| notable6__: weekly | 17:30 | ||
| notable6__ | lizmat, 4 notes: gist.github.com/000dc73712d61c2bef...f251e36df2 | ||
| timo | still working on that ddebian stuff haha | ||
| lizmat | notable6__: weekly reset | 17:33 | |
| notable6__ | lizmat, Moved existing notes to “weekly_2024-10-14T17:33:23Z” | ||
| lizmat | and yet another Rakudo Weekly News hits the Net: rakudoweekly.blog/2024/10/14/2024-...rritonads/ | 18:35 | |
| [Coke] | my brain really wants to split that into two words in an unfortnate place. :) | 18:40 | |
| timo | burritornado | 18:48 | |
| lizmat | .oO( a dirty mind is a joy forever ) |
18:49 | |
| RakuAST **not** being slower: github.com/rakudo/rakudo/issues/41...2411999148 | 18:51 | ||
| timo | cool | 18:52 | |
| Geth | roast: 690c9148fb | (Elizabeth Mattijsen)++ | S12-enums/misc.t Add test for #4134 |
18:56 | |
| bartolin | Oh, I slowly remember where I left off the last time I looked at the jvm backend. precompilation was busted there (and still is). I *think* I pinpointed it to github.com/rakudo/rakudo/commit/efeb9d7742. If I'm not mistaken, the underlying problem was that the null checks in 'method pun' didn't work and one got a NQPMu back for 'blob8.^pun'. (As if TWEAK was not executed.) | 20:31 | |
| sorry for the rambling, but maybe it rings a bell for someone. | |||
| m: say blob8.^make_pun ## same result on jvm | 20:34 | ||
| camelia | (Blob[uint8]) | ||
| bartolin | m: say blob8.pun ## on jvm: Method 'gist' not found for invocant of class 'NQPMu' | 20:35 | |
| camelia | No such method 'pun' for invocant of type 'Blob[uint8]' in block <unit> at <tmp> line 1 |
||
| bartolin | m: say blob8.^pun ## on jvm: Method 'gist' not found for invocant of class 'NQPMu' | ||
| camelia | (Blob[uint8]) | ||
| bartolin | oh, I guess, I was very close back than: If I use just the following: 'method pun($target) { $!pun := self.make_pun($target) }' then loading a module works again. Not a clean solution, but at least no longer a strange unknown failure. | 20:39 | |
| I think, I was not precise before: It's not precompilation, but module loading in general that's busted. | 20:40 | ||
| Geth | roast/log2-tests: 55c808c89a | (Tom Browder)++ | 2 files Add tests for log2 |
22:15 | |
| roast: tbrowder++ created pull request #863: Add tests for log2 |
|||
| ab5tract | bartolin++ | 22:42 | |
| The JVM surely appreciates the love | |||
| *backend | |||
| [Coke] | bartolin++ | 23:08 | |
| tbrowder | librasteve: i have yr log2 tests on deck ready for a new PR for roast issue #862. if you have no objections, i'll submit the PR. | 23:17 | |
| and close the issue if its accepter. | 23:18 | ||
| *accepted | |||
| see new file spec/S32-num/complex-logarithms.t on roast branch 'log2-tests' | 23:30 | ||
| probably ought to make new branch... | 23:33 | ||
| new branch is 'add-log2-test-file' | |||
| .tell librasteve ^^^ | 23:34 | ||
| tellable6 | tbrowder, I'll pass your message to librasteve | ||