00:56 squashable6 left 00:58 squashable6 joined 01:48 leont left
Geth roast: vrurg++ created pull request #674:
Fix some coercion-related tests
02:15
03:08 linkable6 left, evalable6 left, MasterDuke left, evalable6 joined 03:10 linkable6 joined
Xliff m: my $a = Pair.new('a', 54); $a.value = 2; 05:29
camelia Cannot modify an immutable Int (54)
in block <unit> at <tmp> line 1
lucs Xliff: Maybe this can help: search for "worth noting" at docs.raku.org/type/Pair 05:42
Xliff lucs: Thanks, but unfortunately, I'm using a hash from another scope and trying to change it while iterating over %myhasn.pair 05:44
lucs Try iterating on %h.kv maybe? 05:52
Not sure how you can change the original %h, might need to copy. 05:54
(where are the experts! :-)
Oh, something like this: for %h.kv -> $k, $v { %h{$k} = $v + 2 } 05:55
m: my %h = <a 1 b 2>; for %h.kv -> $k, $v { %h{$k} = $v + 2 }; say %h 05:57
camelia {a => 3, b => 4}
lucs & # ZZ 06:00
Xliff m: say ('a', 1, 'b', 2, 'c', 3).Hash.gist 07:04
camelia {a => 1, b => 2, c => 3}
Xliff m: say 8 % 2 07:08
camelia 0
07:16 Altai-man joined 07:19 sena_kun joined 07:21 Altai-man left 08:59 sena_kun left, sena_kun joined 10:17 raku-bridge left, raku-bridge joined, raku-bridge left, raku-bridge joined
Geth rakudo/rakuast: 898cfff995 | (Elizabeth Mattijsen)++ | 2 files
Fix typos in comments
10:25
rakudo/rakuast: eea0716f70 | (Elizabeth Mattijsen)++ | 6 files
Introducing Formatter class and associated Grammer

This is basically taking my work on the "newer-sprintf" branch into the RakuAST era, and also making that work a first class citizen.
The Formatter class takes a sprintf format string and returns a Callable that can be called, with the necessary parameters, to have ... (26 more lines)
rakudo/rakuast: 852bdb8773 | (Elizabeth Mattijsen)++ | src/core.e/Formatter.pm6
Tweaks to actually make it work

  ./rakudo-m -e 'use v6.e.PREVIEW; say sprintf(q/%02d: %s/,42,"foo")'
11:07
11:18 Altai-man joined 11:21 sena_kun left
Geth rakudo/rakuast: 5c7c58d669 | (Elizabeth Mattijsen)++ | src/core.e/Formatter.pm6
Refactor create-format a bit

So that it's more about wrapping things that we got so far, to make transition to AST easier.
11:37
12:09 leont joined
Geth rakudo/rakuast: cc2072d7a8 | (Elizabeth Mattijsen)++ | src/core.e/Formatter.pm6
RakuASTify no-arg formats

And return the format for one or more arg formats for now.
12:14
lizmat my first successful RakuASTification! :-)
moritz lizmat++ 12:23
[TuxCM] Rakudo version 2020.08.2-59-ge65466fcd - MoarVM version 2020.08-91-g590bac47e
csv-ip5xs0.971 - 1.042
csv-ip5xs-208.838 - 9.850
csv-parser27.278 - 27.486
csv-test-xs-200.392 - 0.398
test7.760 - 7.857
test-t1.957 - 2.103
test-t --race0.873 - 0.921
test-t-2033.110 - 33.518
test-t-20 --race9.534 - 9.628
12:30
13:28 dogbert17 joined
Geth rakudo/rakuast: ce4f40e0c0 | (Jonathan Worthington)++ | t/12-rakuast/code.t
Correct a couple of test code descriptions
15:17
rakudo/rakuast: 361093c1bd | (Jonathan Worthington)++ | 4 files
Further parameter support in RakuAST

  * Named parameters
  * Support for basic type constraints on scalar parameters
  * Support for optional scalar parameters
  * Use Mu or Any as the parameter default depending on if it's on a
   block or a routine
  * Wire these things up in the RakuAST-based compiler
rakudo/rakuast: d3a8cce33b | (Jonathan Worthington)++ | 4 files
Fix lookup of routines using &name
15:20 sena_kun joined 15:21 Altai-man left 15:52 squashable6 left 15:53 squashable6 joined 17:03 domidumont joined 17:05 ggoebel joined
ggoebel what is the appropriate channel to discuss documentation wrt to what can be found at raku.org, rakudo.org, moarvm.org, etc? 17:18
[Coke] raku is probably best. 17:19
(in general) 17:20
moarvm has #moarvm, might make sense for that one in particular.
ggoebel thx
17:39 MasterDuke joined
sena_kun releasable6, status 17:44
releasable6 sena_kun, Next release in ≈9 days and ≈1 hour. 1 blocker. Changelog for this release was not started yet
sena_kun, Details: gist.github.com/9159dd74ebf1734a90...ecd6ef11fd
17:51 rypervenche left 17:56 rypervenche joined 18:03 ggoebel_ joined 18:06 ggoebel left 18:22 squashable6 left 18:24 squashable6 joined 18:49 domidumont left
vrurg .ask jnthn Could you have another look at github.com/Raku/problem-solving/issues/227? It might have consequences for RakuAST too. 18:55
tellable6 vrurg, I'll pass your message to jnthn
nine vrurg: I came across your github.com/rakudo/rakudo/pull/3284 18:57
vrurg: I've been working on in-process precompilation for quite some time this year already and the biggest obstacle is how we deal with stashes right now. 18:58
My latest attempt is to instead of serializing the stashes after compilation, generate code that will assemble them at deserialize time 18:59
vrurg nine: Look similar to what I was trying to do. I never finished it because I was hitting a SEGV and misdiagnosed it as a moar bug. 19:01
nine Well at the very least having two people come up with the same solution for two different problems makes me more confident, that this is indeed the right direction 19:03
vrurg It certainly is. I investigated it down to the point where the conflict of restoring two different stashes of the same parent package is possible but can't be resolved because it would require the backend to know about Stash object and its content. Hence the fixup at deserialization point. Don't know what's your case, apparently. :) 19:07
BTW, the problem I was resolving was reported by Xliff. I think he can check if your solution fixes that one too. 19:09
nine Well my trouble comes from those stashes being pretty global and get changed by module loading. So if the same process compiles multiple precomp files, the latter ones may pick up dependencies they shouldn't have, just because stashes they refer to have been filled with references by previous compilations 19:16
19:18 Altai-man joined 19:21 sena_kun left
vrurg Hm, that's pretty much opposite problem. In Xliff case symbols were missing because later loaded module overrode earlier one. 19:22
nine Yet the solution could be the same. 19:26
20:00 ggoebel_ left 20:03 ggoebel joined
Xliff vrurg++: You have that soution in a branch? 20:17
vrurg Xliff: there is a PR, but it lags to far behind. Not applicable anymore. 20:31
Xliff: but there is a branch in my fork.
20:39 Altai-man left
Xliff vrurg: Let me know if you want me to test! 21:49
vrurg Xliff: I think this time it's better to wait for nine to finish it's work. I think his solution would cover more aspects of the problem. 21:51
s/it's/his/
Geth rakudo/rakuast: 2eb407d19e | (Elizabeth Mattijsen)++ | src/core.e/Formatter.pm6
RakuASTify subs "parameter" and "handle-integer-numeric"

Also add RAKUDO_AST parameter to actually EVAL the AST, rather than EVALling the string representation that is also generated with but. Add some helper subs as well, for creating string and integer literals.
22:03
lizmat sleep& 22:07
Xliff vrurg: OK! 22:12
Geth rakudo/rakuast: b4139d9135 | (Jonathan Worthington)++ | 4 files
RakuAST support for slurpy parameters
22:26
23:00 pamplemousse joined
Xliff jnthn: How is work on RakuAST progressing? 23:36
tellable6 Xliff, I'll pass your message to jnthn