🦋 Welcome to the IRC channel of the core developers of the Raku Programming Language (raku.org #rakulang). This channel is logged for the purpose of history keeping about its development | evalbot usage: 'm: say 3;' or /msg camelia m: ... | Logs available at irclogs.raku.org/raku-dev/live.html | For MoarVM see #moarvm
Set by lizmat on 8 June 2022.
00:00 rypervenche left, [Tux] left, reportable6 left, rypervenche joined, rypervenche left, rypervenche joined 00:01 reportable6 joined 00:03 Geth left, Geth joined, [Tux] joined 00:09 finanalyst left 00:19 sivoais left, sivoais joined 03:27 quotable6 left, benchable6 left, bloatable6 left, shareable6 left, evalable6 left, bisectable6 left, coverable6 left, statisfiable6 left, linkable6 left, unicodable6 left, committable6 left, notable6 left, reportable6 left, nativecallable6 left, sourceable6 left, tellable6 left, greppable6 left, releasable6 left, squashable6 left 03:28 sourceable6 joined, tellable6 joined, reportable6 joined, quotable6 joined, benchable6 joined, shareable6 joined, unicodable6 joined 03:29 committable6 joined, coverable6 joined, linkable6 joined, greppable6 joined, bisectable6 joined, releasable6 joined, statisfiable6 joined 03:30 squashable6 joined, evalable6 joined, bloatable6 joined, notable6 joined, nativecallable6 joined 04:15 Voldenet left, Voldenet joined 04:32 lucs left, lucs joined 04:36 elcaro left, elcaro joined 05:21 kjp joined 06:00 reportable6 left, reportable6 joined 08:08 ab5tract left 08:10 ab5tract joined, ab5tract left, ab5tract joined 09:23 ab5tract left
Geth rakudo/main: f950ca1384 | (Elizabeth Mattijsen)++ | src/Raku/Actions.nqp
RakuAST: streamline backslash handling in quote langs

Handle vertical whitespace handling in a helper method
rakudo/main: 97fac4920b | (Elizabeth Mattijsen)++ | src/core.c/RakuAST/Raku.rakumod
RakuAST: don't show processors if there aren't any
rakudo/main: 66b6f78876 | (Elizabeth Mattijsen)++ | src/Raku/Actions.nqp
RakuAST: streamline some of slang escape handling
09:43 ab5tract joined
Geth rakudo/main: bc16a7ac2e | (Elizabeth Mattijsen)++ | src/Raku/Actions.nqp
RakuAST: streamline dottyop/methodop some more
10:41 nine left, camelia left 11:04 nine joined 11:16 ab5tract left, ab5tract joined
ab5tract Interesting... I've got things mostly working with flip flop but prefix:sym<^> seems to match `$a ff^ $b` (`$a ff $b` and `$a ^ff $b`) work fine 11:25
it tries to create a ^$b range object in the ($a ff^ $b) case 11:26
lizmat hmmm... maybe the precedence of ff is wrong? 11:27
or of prefix:<^> ? 11:28
ab5tract it should be auto-getting stuff from OperatorProperties, right?
lizmat yes, but maybe I copied it incorrectly into OperatorProperties: there were several sources of information in that respect, and not all of them had the same info :-( 11:29
I mean, if it ASTs into a RakuAST::Infix, yes
ab5tract yeah, it does
11:29 camelia joined
lizmat ff should have precedence j= 11:30
ab5tract what's the best way to check the precedence? 11:31
lizmat looks like prefix:<^> has precedence v= which is higher... which feels... hmmm
ab5tract yeah, not sure how the parser is supposed to interpret that one
lizmat checked with legacy grammar, looks like OperatorProperties is in sync with that 11:34
ab5tract shouldin't it be a case of LTM? 11:35
lizmat aaah... I think I now what I borked 11:36
*know 11:37
testing.... 11:39
Geth rakudo/main: cd32d6a3fd | (Elizabeth Mattijsen)++ | src/Raku/Grammar.nqp
RakuAST: fix thinko with ff^ ff^ ^ff& and ^fff^
lizmat ab5tract ^^
ab5tract lizmat++
lizmat pretty sure that will fix it
Geth rakudo/main: 381cf92173 | (Elizabeth Mattijsen)++ | src/Raku/Actions.nqp
RakuAST: streamline some character class handling
12:00 reportable6 left 12:03 reportable6 joined 12:24 [Tux] left
Geth rakudo/main: ae83388466 | (Elizabeth Mattijsen)++ | src/Raku/Actions.nqp
RakuAST: simplify action method handling for ternaries

And some better grouping of action methods
12:30 [Tux] joined 13:58 nativecallable6 left, coverable6 left, statisfiable6 left, reportable6 left, unicodable6 left, squashable6 left, committable6 left, tellable6 left, bloatable6 left, greppable6 left, releasable6 left, sourceable6 left, evalable6 left, notable6 left, bisectable6 left, shareable6 left, quotable6 left, linkable6 left, benchable6 left 13:59 bloatable6 joined, bisectable6 joined, coverable6 joined, notable6 joined, linkable6 joined, reportable6 joined, shareable6 joined, releasable6 joined 14:00 benchable6 joined, committable6 joined, sourceable6 joined, squashable6 joined, quotable6 joined 14:01 nativecallable6 joined, statisfiable6 joined, unicodable6 joined, greppable6 joined, evalable6 joined, tellable6 joined
ab5tract grrr... so close to a finalized flip flop 14:55
the only problem comes from `/F/ ff *` 14:56
it's definitely a tricky one to solve, all I have to go on is: `No exception handler located for catch` 14:57
[Tux] Rakudo v2023.08-224-gae8338846 (v6.d) on MoarVM 2023.08-3-g2f8234c22
csv-ip5xs0.780 - 0.804
csv-ip5xs-205.109 - 5.213
csv-parser3.503 - 3.704
csv-test-xs-200.336 - 0.372
test6.235 - 6.397
test-t1.351 - 1.397
test-t --race0.819 - 0.856
test-t-2019.756 - 19.867
test-t-20 --race6.128 - 6.789
(after a postponed reboot)
nemokosch how to interpret these values? what do they mean to someone who looks at this chat? 15:11
MasterDuke tux.nl/Talks/CSV6/speed4.html 15:20
there are some other speed<N>.html pages also 15:21
nemokosch I get this part but why does this need to be sent into the chat periodically? 15:25
wouldn't it make more sense to at least only send it on certain noticeable events?
16:00 linkable6 left, evalable6 left, evalable6 joined 16:02 linkable6 joined
lizmat nemokosch: in the past, test-t (as we call it) as been a very useful canary for performance regressions.. so it's good to see it done frequently 16:06
nemokosch It has no meaning out of context. Do you know by heart which test should run under what time? 16:13
I think it's as if geth was set up to post the number of issues in Raku-related github repositories 16:14
MasterDuke but it has ~8 years of context. i couldn't tell you all the names and their usual times out of the blue, but when i look at that list those numbers seem normal 16:21
the context is all the years of posting them on irc and the discussions about them. though they are much less frequent than they used to be, i assume mostly because there aren't very many large changes anymore 16:23
lizmat irclogs.raku.org/perl6-dev/search....iases=True 16:30
nemokosch the same argument could be made for Geth and Raku-related issues, it's kinda self-justificatory 16:35
I'd say the difference is exactly that there are no discussions or anything, there is nothing to be talked about, and then is it really mandatory to dump it into IRC? 16:39
17:02 evalable6 left, linkable6 left 17:03 linkable6 joined 17:05 evalable6 joined 17:14 discord-raku-bot left, discord-raku-bot joined 17:20 timo left 17:21 timo joined
Geth rakudo/main: 7ebd9bb39f | (Elizabeth Mattijsen)++ | 2 files
RakuAST: couple alpha infixes to static action methods

This will allow easy renaming of alpha infix operators in non-english language versions of Raku.
17:29 coleman left 17:31 coleman joined
Geth rakudo/main: 42a92611bd | (Elizabeth Mattijsen)++ | 2 files
RakuAST: couple alpha prefixes to static action methods

This will allow easy renaming of alpha prefix operators in non-english language versions of Raku.
rakudo/main: 5d784a9c98 | (Elizabeth Mattijsen)++ | src/Raku/Actions.nqp
RakuAST: streamline regex building actions
18:00 reportable6 left 18:02 reportable6 joined
ab5tract the flip flop operator is... wild 18:02
18:02 nebuchad` left
ab5tract feels like some ultimate YAGNI to me, but I'm sure it makes someone out there very happy 18:03
18:03 nebuchad` joined
ab5tract what's the use case of this thing? 18:09
nemokosch extracting out a part of a text stream based on some delimiting patterns 18:10
... or so I heard 18:11
I think it's too "clever" to be actually useful
lizmat it's something engrained in sed users apparently 18:20
[Coke] I need it maybe once every 2-3 years, it's nifty but not something I rely on
Geth rakudo/main: 1b6f5791f3 | (Elizabeth Mattijsen)++ | src/Raku/Actions.nqp
RakuAST: streamline quantified_atom action metho
18:53 coleman left 18:55 coleman joined
leont Crap, if I had finished a blogpost of mine, I could have explained it easily, but I haven't. 19:04
The flip flip operator is absolutely essential when using the language as an awk replacement. And almost entirely useless for any other use-case 19:05
That means in Perl it's absolutely essential, but Raku much less so. Mainly because Raku's improvement tend to be on the other side: being a better language at scale. 19:07
Perl is the best awk that was ever written, Raku never was about being a better awk than perl, but about being a better language at scale. IMO. 19:08
ab5tract well, it's working in RakuAST now :) 19:41
20:05 linkable6 left, evalable6 left, linkable6 joined 20:06 evalable6 joined
ab5tract errr.. I spoke to soon. I thought I had it. 20:20
21:07 NemokoschKiwi joined
Geth rakudo: ab5tract++ created pull request #5383:
RakuAST: Add support for flip-flop ops
21:26 NemokoschKiwi left
lizmat ab5tract: but now you haz it ? 21:27
ab5tract yeah, they are working 21:50
but the patch conflicts with something you pushed today 21:51
+2 spectest
Geth rakudo/main: 4c0ffe67dd | ab5tract++ (committed using GitHub Web editor) | 4 files
RakuAST: Add support for flip-flop ops
rakudo/main: f05f774660 | (Elizabeth Mattijsen)++ | src/Raku/Actions.nqp
RakuAST: make ff and friends actually work

  ab5tract++ for the hard work
rakudo/main: 770ac47bf3 | (Elizabeth Mattijsen)++ | src/Raku/Actions.nqp
RakuAST: streamline regex quantifier handling
lizmat 953 +2 indeed! 22:41
22:50 evalable6 left, linkable6 left, evalable6 joined 22:52 linkable6 joined