|
04:47
vrurg joined
04:48
vrurg_ left
|
|||
| lizmat | timo: I'd say: yes, but what ? | 08:13 | |
| *what with | |||
|
08:31
sena_kun joined
|
|||
| Geth | rakudo/main: 066afa401c | (Elizabeth Mattijsen)++ | t/12-rakuast/xx-fixed-in-rakuast.rakutest Add test for #2360 |
09:51 | |
| rakudo/main: 8b61f16776 | (Elizabeth Mattijsen)++ | t/12-rakuast/xx-fixed-in-rakuast.rakutest Add test for #2362 |
10:16 | ||
| roast: c3e2e7685b | (Elizabeth Mattijsen)++ | S04-phasers/enter-leave.t Add tests for #2380 |
10:47 | ||
| lizmat | m: sub a(|c) { dd c }; a(1;2) # where is this behaviour documented ? | 11:00 | |
| camelia | \((1,), (2,)) | ||
| lizmat | [Coke] ^^ ?? | 11:01 | |
|
11:18
sjn left
|
|||
| Geth | nqp/main: d280677bd7 | (Elizabeth Mattijsen)++ | src/QRegex/Cursor.nqp Micro-optimize Match!cursor_init a bit If working on a freshly created object, it is not necessary to initialize the $!from / $!pos attributes, as they're native ints that are created with 0 |
11:29 | |
|
11:30
sjn joined
|
|||
| Geth | rakudo/main: 427f8b7a1a | (Elizabeth Mattijsen)++ | tools/templates/NQP_REVISION Bump NQP for Match!cursor_init micro-opt |
11:45 | |
|
12:46
sjn_ joined
|
|||
| Geth | roast: 8498f84f10 | (Elizabeth Mattijsen)++ | S02-names-vars/signature.t Add test for #2416 |
13:53 | |
| [Coke] | docs.raku.org/language/signatures#...parameters | 13:58 | |
| timo | lizmat: literally anything that's got some "binary" bytes in it | 14:30 | |
| lizmat | so maybe it's a matter of just changing the extension of that file? | 14:31 | |
| notable6__: weekly | 14:39 | ||
| notable6__ | lizmat, 1 note: 2024-10-04T09:22:02Z <lizmat>: github.com/zag/raku-knowledge-base | ||
| lizmat | notable6__: weekly reset | ||
| notable6__ | lizmat, Moved existing notes to “weekly_2024-10-07T14:39:48Z” | ||
| timo | no the extension is already .bin, lintian looks at the contents and sees that we're trying to smuggle an elf through | 14:59 | |
| i mean, we could also turn the first couple of bytes into something not elf-compatible | 15:03 | ||
| but clearly someone just took any random library or executable from parrot, chopped it off at the desired size and plopped it in there | |||
| lizmat | wouldn't something like: my $a := buf16.new; $a.push($_) for ^1000; "wop".IO.spurt($a) do the trick? | 15:05 | |
| [Coke] | timo: where is this .bin file? | 15:12 | |
| timo | t/spec/S32-io/socket-test.bin | 15:15 | |
| lizmat: it probably could, actually | |||
|
15:15
sjn_ left
|
|||
| timo | this is only vaguely related, but "magical binary files in the test data" was how the xz-utils backdoor was disguised | 15:16 | |
| well, it was one part of the disguise | 15:27 | ||
| Geth | roast: efe9245fd4 | (Elizabeth Mattijsen)++ | S32-io/socket-test.bin Create non-sensical binary file for testing |
15:28 | |
| lizmat | timo: ^^ would this be better ? | ||
| timo | looks good | 15:30 | |
| lizmat | actually, we don't need the file at all | 15:41 | |
| Geth | roast: 2306979015 | (Elizabeth Mattijsen)++ | 3 files Don't need a binary file to test binary IO Just create a blob during testing in memory. This should make the debian people a lot happier |
15:46 | |
| lizmat | timo: ^^ a better fix | ||
| timo | well, "the debian people" didn't really complain, i just saw an automated checking tool point it out | 15:47 | |
| lizmat | hehe... well, if debian people would look at that output, they would complain no? | ||
| timo | i'm actually not 100% sure if the t/spec/ should be part of the source package at all, and i'm not sure if maybe i broke something to cause them to be there when previously they weren't | 15:48 | |
| lizmat | do they run "make spectest" ? | ||
| jdv | iirc its 6.d that in the src tarball, probably by acvident | 15:56 | |
| 6d errata spec i mean | |||
| timo | only "make test" is run for the "autopkgtest" thing | 16:01 | |
| there are a whole bunch of warnings when building that files have been removed or something that i don't yet really understand what causes them | 16:03 | ||
| salsa.debian.org/timotimo-guest/ra...86574#L146 here's an example output from lintian mentioning the .bin file, timotimo-guest.pages.debian.net/-/...ntian.html has it in pretty html (but the links that are supposed to explain the error codes are all dead links) | 16:16 | ||
| [Coke] | Looks like lots of instance of 'allows to' in rakudo src, but not sure where the man page is defined, I'm sure I've seen it recently on the debian side... | 16:30 | |
| timo | yeah it's generated from stuff in the debian/ folder | 16:31 | |
| pod2man --name=perl6 docs/running.pod debian/perl6.1 | 16:32 | ||
| [Coke] forgot how nice "authorize git over ssh with touch id" was. | 16:44 | ||
| Geth | roast: 9f75bc5d2c | (Elizabeth Mattijsen)++ | S14-roles/composition.t Add test for #2418 |
16:46 | |
| rakudo/main: a60b4e7d2b | (Will Coleda)++ | docs/running.pod Fix grammar issue noticed by debian++ build process. |
|||
| [Coke] | Not to erase history, but is it worth keeping Reini & Moritz listed explicitly on the docs/running.pod page? | 16:49 | |
| lizmat | and yet another Rakudo Weekly News hits the Net: rakudoweekly.blog/2024/10/07/2024-41-knowlite/ | 16:52 | |
| timo | how do you feel about me picking nits in the weeklies? | 17:05 | |
| lizmat | I welcome nits :-) | 17:06 | |
.oO( they're a great band ) |
17:07 | ||
| www.youtube.com/watch?v=Un0VLiPKKyg :-) | 17:12 | ||
| Geth | rakudo/main: c9d03adff9 | (Elizabeth Mattijsen)++ | src/Raku/ast/statements.rakumod RakuAST: PAST is a thing of the past In light of github.com/rakudo/rakudo/issues/2431 |
17:26 | |
| rakudo/main: 0774293522 | (Elizabeth Mattijsen)++ | src/core.c/Num.rakumod Fix LTA error message on num % / ** num The problem is really that *returning* a Failure from a sub that has a native return constraint **dies**, rather than passing on the Failure. If however a CONTROL message is used to return a Failure, all is well. So work around this issue by using `fail` with the Failure rather than just returning it. Fixes #2434 |
19:08 | ||
| roast: e2d90988f9 | (Elizabeth Mattijsen)++ | S02-types/num.t Add tests for #2434 |
19:09 | ||
| rakudo/main: 0fc7d45f66 | (Elizabeth Mattijsen)++ | 5 files Make Iterable:U.flat do the right thing This also exposed a bug in Iterable.flat in that it was **NOT** marked as "nodal". Now that there's only one proto with "is nodal" the Iterable:D.flat also became nodal. This in turn exposed a faulty test in S03-metaops/hyper.t, which will be fixed shortly. Fixes #2442 |
19:51 | ||
| roast: 7afb05d9c3 | (Elizabeth Mattijsen)++ | 2 files Adapt some .flat tests in light of github.com/rakudo/rakudo/commit/0fc7d45f66 |
19:52 | ||
| roast: 789419a1ec | (Elizabeth Mattijsen)++ | S02-types/flattening.t Add tests for #2442 |
19:57 | ||
|
20:31
[Coke] left
22:28
sena_kun left
|
|||