00:19 Altai-man_ joined 00:22 sena_kun left 00:57 djinni` left, greppable6 left, committable6 left, bisectable6 left, reportable6 left 01:02 djinni` joined, greppable6 joined, committable6 joined, bisectable6 joined, reportable6 joined 01:03 AlexDaniel` left, unclechu left 01:07 unclechu joined 01:22 AlexDaniel` joined
Geth roast/roast_657: 17 commits pushed by (Vadim Belman)++
review: github.com/Raku/roast/compare/07c0...6ec9a2d937
01:33 sena_kun joined 01:35 Altai-man_ left 01:49 Kaiepi left 02:42 Altai-man_ joined 02:44 sena_kun left 03:38 sena_kun joined 03:40 Altai-man_ left 04:10 [Coke] left 04:17 [Coke] joined, [Coke] left, [Coke] joined 04:22 Altai-man_ joined 04:24 sena_kun left 04:53 [Coke]_ joined 04:54 [Coke] left 05:02 sena_kun joined 05:04 Altai-man_ left 05:13 Altai-man_ joined 05:15 sena_kun left 05:32 sena_kun joined 05:34 Altai-man_ left 06:02 reportable6 left 06:05 reportable6 joined 06:11 Altai-man_ joined 06:14 sena_kun left 06:31 patrickb joined 06:35 patrickb left 06:47 patrickb joined 07:42 Kaiepi joined 07:51 [TuxCM] left 08:12 sena_kun joined 08:14 Altai-man_ left 08:21 squashable6 left 08:23 leont joined 08:24 squashable6 joined 10:01 Kaiepi left 10:02 Kaiepi joined 10:03 Kaiepi left, Kaiepi joined 10:06 JJMerelo joined
JJMerelo Just a reminder that the call for authors of the 20th anniversary calendar is open github.com/Raku/advent/tree/master/20th 10:08
tellable6 2020-07-28T19:31:01Z #raku <[ptc]> JJMerelo having dug a bit further, I've found that the whateverable Travis build uses htmlify in the test section, which no longer exists
2020-07-28T19:31:41Z #raku <[ptc]> JJMerelo should therefore the whateverable part of the build be removed? I mean, it doesn't work atm anyway... Maybe better to get rid of dead code?
JJMerelo .tell [ptc] it's a reasonable option to remove that, yes. It's not running anyway, and hasn't for some time.
tellable6 JJMerelo, I'll pass your message to [ptc]
JJMerelo There are also a couple of examples now, just in case you need to be inspired github.com/Raku/advent/tree/master...h/articles 10:09
We are looking for short articles that refer to specif RFCs and show how they ended up in the real Raku :-) Everyone is welcome to participate! 10:11
10:11 Altai-man_ joined 10:13 sena_kun left 10:15 MasterDuke left 10:37 TimToady left
JJMerelo We have almost seven now github.com/Raku/advent/tree/master/20th Anyone else? 10:44
10:44 TimToady joined
lizmat hopes TimToady could be persuaded to write one 10:45
moritz lizmat: I recommend writing an email to him; he hasn't been very active on IRC lately 10:49
lizmat moritz: TimToady doesn't answer my emails, so that is not an option :-( 10:50
possibly *any* emails, but that I do not know
moritz :-( 10:51
lizmat JJMerelo: I'm tempted to do raku.org/archive/rfc/28.html but I probably shouldn't 11:05
JJMerelo lizmat double LOL 11:06
lizmat Guess I'm a programmer, so I won't
JJMerelo lizmat we could email Gloria :-) Maybe she can be persuaded to write one, too :-)
11:08 JJMerelo left, JJMerelo joined
JJMerelo has been kicked out again... 11:13
11:13 Kaiepi left 12:12 sena_kun joined 12:14 Altai-man_ left 12:34 squashable6 left
[Tux] Rakudo version 2020.07-46-ga3af5833d - MoarVM version 2020.07-10-g7a2881a85
csv-ip5xs0.824 - 0.866
csv-ip5xs-208.281 - 8.678
csv-parser24.741 - 28.468
csv-test-xs-200.402 - 0.415
test8.309 - 8.756
test-t1.999 - 2.159
test-t --race0.862 - 0.871
test-t-2032.528 - 34.046
test-t-20 --race9.289 - 9.300
13:21 Kaiepi joined 13:30 JJMerelo left 13:46 [Coke]_ is now known as [Coke], MasterDuke joined
raku-bridge <Vendethiel> oh, that's an interesting idea. 13:49
14:10 lucasb joined 14:11 patrickb left, Altai-man_ joined 14:12 patrickb joined 14:14 sena_kun left 14:56 sivoais_ joined 14:58 sivoais left 15:00 squashable6 joined 15:04 lichtkind joined 15:55 raku-bridge left, raku-bridge joined, raku-bridge left, raku-bridge joined 16:01 patrickb left 16:12 sena_kun joined 16:14 Altai-man_ left 16:48 lichtkind left
gfldex is there a Raku-land equivalent to `$/ := nqp::getlexcaller('$/');`? 17:52
lizmat CALLER::<$/> ? 17:54
17:54 JJMerelo joined
gfldex doesnt work in the same way 17:55
line 23-25 in gist.github.com/gfldex/1193b1eaf86...541a0adad3 17:56
descenting into nqp-land makes line 52-54 possible 17:57
But I feel dirty already for using augment. :-> 17:58
Do we have a place to grep in all the source of all the modules in the ecosystem? 18:00
I would like to know how many modules augment and/or `use nqp;`.
timotimo aye, an irc bot
otherwise you can clone perl6-all-modules from mritz' github user
greppable6: use nqp
greppable6 timotimo, 134 lines, 85 modules: gist.github.com/8315e8af66af09cd8f...c2fc24a9d2
gfldex greppable6: augment class 18:01
greppable6 gfldex, 49 lines, 26 modules: gist.github.com/4088b5b8e7b51d9427...500c240a5f
gfldex I feel a little less dirty already.
18:11 Altai-man_ joined 18:13 sena_kun left
gfldex Does the occurrence of ~~ introduce $/ into the local scope? 18:32
timotimo $/ lives in every block, jus tlike $! and $_ 18:33
gfldex is that case i might have found a bug 18:35
gist.github.com/gfldex/8a02dd9f9c6...03e29a5d77 outputs „caller lexicals: ($_)“ unless I uncomment line 50. Then it outputs „caller lexicals: (Exitcode $¢ $/ $=finish ::?PACKAGE GLOBALish $?PACKAGE $! $ex $_ Shell $=pod !UNIT_MARKER EXPORT)“ 18:37
Is the optimiser optimising? 18:38
timotimo it's possible
CALLER:: stuff should only be able to access variables that are "is dynamic" i think?
gfldex I think roast says so. Rakudo is using the nqp-way in Str.match . 18:40
18:48 JJMerelo left
gfldex I got it working by using ‚CALLER::CALLER::<$/> := $/;‘. given/when seams to introduce an additional frame that does not inherit $/ 18:59
Does this qualify as a Rakudobug? 19:04
19:11 lichtkind joined
lizmat gfldex: that's the whole problem with Str.match 19:49
each method needs to either pass on the original $/, or needs to fetch it from their caller, and be careful about that 19:50
the fact that you need two CALLER:: to get at the correct $/ feels correct to me
gfldex Str.pm6 is way longer then it should be …
lizmat it *is* another frame, is it not
gfldex: tell me about it
yet separating it into separate files, like I've done with Supply, generally does not sit well by some people :-( 19:51
gfldex All you do is moving complexety around.
I mean the bug part is that this additional frame does hide $/. 19:52
So it's not really a block.
lizmat fwiw, I would be very glad if we could lose $/ altogether
but then how would $0 and $<foo> work ? 19:53
gfldex Those magic variables are super useful. It's a good thing they are few tho.
lizmat but having to pass them on is a pain :-), as you've found out :-) 19:54
gfldex If it would have been documented, I would likely not spend 3 hours to make it work.
OUTER/CALLER/etc are barely mentioned 19:55
lizmat it's mainly because they're really hacks and feel like they should be handled differently 19:56
gfldex If two MMD candidates have the same "precedence", is it defined behaviour which one will win?
lizmat and they come with performance penalties, as anything doing a CALLER:: basically cannot be optimized, if I remember correctly
gfldex I believe to remember the first found MMD candidate will win. I'm not sure if that is by design or by implementation. 19:59
Actually I will just put it into my next blog post. On the internets roviding an answer that is wrong is the fastest way to get the right answer. 20:01
20:12 sena_kun joined 20:14 Altai-man_ left, MasterDuke left 20:51 sivoais_ left 21:18 sivoais joined 22:11 Altai-man_ joined 22:14 sena_kun left 22:26 leont left 23:06 lichtkind left