00:25 sivoais left 00:26 sivoais joined 00:31 Xliff joined
Xliff j #webkitgtk 00:31
tellable6 2020-11-17T09:49:13Z #raku <patrickb> xliff: ^ This may be relevant for your GTK modules.
2020-11-17T17:17:14Z #raku <holyghost> Xliff my macbook is back online, no need for the keys, thanks
01:28 Kaiepi joined 02:29 melezhik left 02:53 lucasb left 04:30 Xliff left
Geth rakudo: usev6++ created pull request #4044:
[JVM] Use archetypes to figure out coerciveness		 06:58
07:30 MasterDuke left 07:47 domidumont joined 08:00 frost-lab joined 08:12 TreyHarris left
Geth ¦ problem-solving: JJ assigned to codesections Issue There's no effective way to communicate something to all people with commit bits in the different Raku repos and organizations github.com/Raku/problem-solving/issues/246 08:41
08:50 TreyHarris joined 09:05 sena_kun joined
releasable6 Next release in ≈3 days and ≈7 hours. 3 blockers. Please log your changes in the ChangeLog: github.com/rakudo/rakudo/wiki/ChangeLog-Draft 11:00
11:14 squashable6 left, squashable6 joined 12:10 Altai-man joined 12:12 sena_kun left 12:49 frost-lab left
Geth rakudo: 39eedacf07 | (Patrick Böker)++ | docs/running.pod
Update `docs/running.pod`

  `docs/running.pod` is duplicated in
  `Raku/doc:doc/Programs/03-environment-variables.pod6`. But the two versions
ran out of sync. Merge the two files again.		 13:18
rakudo: fb32d0ca32 | (Patrick Böker)++ (committed using GitHub Web editor) | docs/running.pod
Merge pull request #3994 from patrickbkr/update-running-rakudo

Update `docs/running.pod`
nqp: b606a21da4 | (Patrick Böker)++ | tools/templates/windows/nqp-m-build.c
Build: No need to escape " in C code

I'm not entirely sure how the compiler reacts to `'\"'`. Possibly the `\` is simply ignored. Then the code worked fine before already. But better be clear about what happens. Note: The file makes a trip through the templating system, which unescapes all the double backslashes.		 13:19
nqp: 545a988502 | (Patrick Böker)++ (committed using GitHub Web editor) | tools/templates/windows/nqp-m-build.c
Merge pull request #678 from patrickbkr/no-build-runner-double-quoting

Build: No need to escape " in C code
rakudo: 857fbfd5c4 | (Patrick Böker)++ | tools/templates/moar/rakudo-m-build.c.windows
Build: No need to escape " in C code

I'm not entirely sure how the compiler reacts to `'\"'`. Possibly the `\` is simply ignored. Then the code worked fine before already. But better be clear about what happens. Note: The file makes a trip through the templating system, which unescapes all the double backslashes.
rakudo: 9976cab18f | (Patrick Böker)++ (committed using GitHub Web editor) | tools/templates/moar/rakudo-m-build.c.windows
Merge pull request #3985 from patrickbkr/no-build-runner-double-quoting

Build: No need to escape " in C code
rakudo: patrickbkr++ created pull request #4045:
Fix building on Windows with MinGW		 13:29
16:03 patrickb joined
Geth rakudo: f0808da601 | (Patrick Böker)++ | src/vm/moar/runner/main.c
Prevent a compiler warning about an invalid cast

  `_get_osfhandle` returns an `intptr_t` which can be cast to `HANDLE`, but
if `io.h` is not included the compiler seems to assume it returns `int` and thus rightfully warns.		 16:08
rakudo: 200821cd90 | (Patrick Böker)++ | 2 files
Fix building on Windows with MinGW

The flag `--subsystem=windows` is possibly supported by the gnu linker, but it's not by GCC. So we use `-mwindows` which is supported by GCC. It is possible that flag also automatically makes GCC link some additional libraries, but that won't hurt.
... (5 more lines)
rakudo: 8310690d5e | (Patrick Böker)++ (committed using GitHub Web editor) | 2 files
Merge pull request #4045 from patrickbkr/fix-mingw

Fix building on Windows with MinGW
16:11 sena_kun joined 16:12 Altai-man left 16:36 sena_kun left, sena_kun joined
Geth rakudo/no-fast-native-shaped-access: bd9e15c7b7 | (Elizabeth Mattijsen)++ | src/core.c/native_array.pm6
Remove native shaped array fast access

Mainly to be able to test this in Blin for regressions.		 17:08
lizmat sena_kun ^^^^ for Blin testing
Geth roast: 5aee10dfc3 | (Elizabeth Mattijsen)++ | S32-list/seq.t
Add tests for GH#4039
 17:12
linkable6 GH#4039 [open]: github.com/rakudo/rakudo/issues/4039 [tests needed] Regression? Sequence of junctions wrapped in singleton lists
Geth rakudo: f6cdbd8147 | (Elizabeth Mattijsen)++ | 2 files
Fix regression in Font::FreeType

Thanks to Altai-man++ and dwarring++ for keeping nudging me. In the end it turned out to be a stupid copy-pasto :-(		 17:44
lizmat ^^ this could have benefitted from a warning of a parameter not being used inside the sub 17:45
*parameter with a name
18:09 patrickb left 18:17 domidumont left
sena_kun ⏳ 598 out of 1551 modules processed 19:03
Geth rakudo: 6dbce0acee | (Christian Bartolomäus)++ | src/vm/jvm/runtime/org/raku/rakudo/Binder.java
[JVM] Use archetypes to figure out coerciveness

That matches the way .coercive is used in many other places. Fixes github.com/rakudo/rakudo/issues/4041		 19:19
rakudo: e074f3c562 | (Christian Bartolomäus)++ (committed using GitHub Web editor) | src/vm/jvm/runtime/org/raku/rakudo/Binder.java
Merge pull request #4044 from usev6/jvm_gh4041

  [JVM] Use archetypes to figure out coerciveness
sena_kun ⏳ 1097 out of 1551 modules processed 19:39
⏳ 1398 out of 1551 modules processed 20:06
20:06 MasterDuke joined 20:10 Altai-man joined 20:12 sena_kun left
Geth rakudo: b452566565 | (Vadim Belman)++ | 2 files
Fix incorrect assignment passing with coercion

Fix for #4040. `my Rat(Str) $foo = 1` will now fail because `Int` is not passing the type check.
Also made `Metamodel::CoercionHOW::coerce()` more strict about the value it's trying to coerce. Now it will throw if it doesn't match against
  `$!constraint_type`. Normally this shouldn't happen to a user unless it
invokes the method directly, but should be of great help in catching core errors.		 20:33
linkable6 RAKUDO#4040 [closed]: github.com/rakudo/rakudo/issues/4040 [JVM] Failing test in S12-coercion/coercion-types.t after switch to new coercion semantics
rakudo: 824c5adc06 | (Vadim Belman)++ | 2 files
Fix a native string passing instead of Str

It was breaking JVM backend installation.
rakudo: 6f314947ea | (Vadim Belman)++ (committed using GitHub Web editor) | 4 files
Merge pull request #4043 from vrurg/rakudo_4040

Fix incorrect assignment passing with coercion		 20:34
nine vrurg: looks like much of our ecosystem depends on our failure to make module loading really lexical. I.e. use Zef; use Zef::Client; and magically you get a Zef::Distribution available as well. 21:13
My replacement for globals merging accidentally fixes this bug, but trips up all dists I try :/ 21:14
vrurg nine: Too bad. Because this may have side effects if I declare a `my class Zef::MyOwnSomething { }` – and Zef gets it into its namespace to. And if it already has something named same way – ... 21:17
Ok, if it's my code – I can do something about it. But what if two modules clash?? I see no solution to it whatsoever. 21:19
I'd say it's better to get some bravery and take care of the failing modules than preserve such a huge pitfall. 21:20
nwc10 we can't change from tabs to spaces in makefiles because there are too many users 21:21
(ish)
nine Wenn in this non-representative sample we're talking about 80 % of modules 21:27
And IO::Socket::Async::SSL's failure mode caused by a missing use statement was a straight segfault
nwc10 fxing this will be easier than trying to get all of CPAN to run with `use strict` enabled by default 21:28
nine Well, I'll still have a look at how to re-create the bug in my implementation. Once I got that I'm pretty sure we can make it language version dependent. 21:29
The good news is that I've made it far enough to actually try multiple distributions ;) 21:36
21:47 MasterDuke left
vrurg nine++ 21:48
22:01 patrickb joined
lizmat nine: feels to me that this does not need to be version dependent, though 22:38
as fixes to modules wouldn't break under the old, buggy implementation, would they ?
Altai-man nine, can you create a branch with the fix so I could blin it? 22:40
tellable6 2020-11-17T20:30:03Z #raku-dev <[Coke]> altai-man: pong.
23:02 Altai-man left 23:17 patrickb left 23:47 travis-ci joined
travis-ci Rakudo build passed. Elizabeth Mattijsen 'Fix regression in Font::FreeType 23:47
travis-ci.org/rakudo/rakudo/builds/744481739 github.com/rakudo/rakudo/compare/8...cdbd81474c
23:47 travis-ci left 23:56 patrickb joined 23:57 patrickb left

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!