»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_log/perl6 | UTF-8 is our friend! 🦋
Set by Zoffix on 25 July 2018.
buggable 🎺🎺🎺 It's time for the monthly Accidental /win Lottery 😍😍😍 We have 0 ballots submitted by 0 users! DRUM ROLL PLEASE!... 00:00
And the winning number is 42! Congratulations to Zoffix! You win a can of WD40!
timotimo rigged!
00:01 Kaiepi left 00:02 Kaiepi joined 00:52 kalkin--- left 01:17 Sgeo_ left
Geth doc: f544565048 | (Zoffix Znet)++ | doc/Language/routines.pod6
Document `undefine`

For 6.c, plenty of spec in roast. 6.d deprecation spec: github.com/perl6/roast/commit/e325718ca2 Rakudo deprecation: github.com/rakudo/rakudo/commit/72bac67080
synopsebot Link: doc.perl6.org/language/routines
01:51 lizmat left 01:53 Sgeo joined, [Coke] joined
[Coke] mst: spam hitting #perl6-dev & #moarvm too 01:54
and #perl6-toolchain
01:55 Zoffix joined
Zoffix So now they get through even with +r? 01:55
oh wc 01:56
01:56 Zoffix left
[Coke] looks like it's ramping up right now, actually. 01:56
01:58 haxmeister left, haxmeister joined 02:07 aindilis left 02:14 aindilis joined
Geth doc: mendel++ created pull request #2234:
mention hyped form of infix assignment ops in traps
doc: mendel++ created pull request #2235:
run the Docker images interactively & on a tty from the Makefile
doc: cdba5a841f | Coke++ | doc/Language/glossary.pod6
synopsebot Link: doc.perl6.org/language/glossary
02:49 vrurg left 02:52 Kaiepi left 02:54 Kaiepi joined 03:30 vrurg joined 03:32 vrurg left 04:11 cgfbee left 04:12 angelds joined 04:18 angelds left 04:20 kerframil left 04:46 vrurg joined 04:47 curan joined 05:04 spycrab0 left 05:10 vrurg left 05:40 sauvin joined, sauvin left 05:42 angelds joined 05:52 sauvin joined 06:29 rindolf joined 06:32 cjkinni left 06:33 cjkinni joined 06:38 domidumont joined 06:39 jmerelo joined
jmerelo squashable6: status 06:41
yoleaux 31 Jul 2018 15:05Z <AlexDaniel> jmerelo: yes??
squashable6 jmerelo, ⚠🍕 Next SQUASHathon in 2 days and ≈3 hours (2018-08-04 UTC-12⌁UTC+14). See github.com/rakudo/rakudo/wiki/Mont...Squash-Day
jmerelo .tell AlexDaniel we'll have to set up something today. The list of modules to work on is there, and it's long-ish...
yoleaux jmerelo: I'll pass your message to AlexDaniel.
AlexDaniel . 06:43
yoleaux 06:41Z <jmerelo> AlexDaniel: we'll have to set up something today. The list of modules to work on is there, and it's long-ish...
06:44 domidumont left 06:45 domidumont joined
AlexDaniel jmerelo: so a repo with tickets populated, one for each failing module, yes? 06:46
Geth doc: e865904d1d | (Norbert Buchmueller)++ | Makefile
run the Docker images interactively & on a tty from the Makefile

This way eg. the test progress is shown when running `make docker-xtest`.
doc: bb19c56d17 | (Juan Julián Merelo Guervós)++ (committed using GitHub Web editor) | Makefile
Merge pull request #2235 from mendel/run-docker-images-interactively-from-makefile

run the Docker images interactively & on a tty from the Makefile Makes sense. Thanks.
07:30 AlexDaniel left
jmerelo AlexDaniel: can we use ecosystem? 07:31
.tell AlexDaniel: Can we use perl6/ecosystem, maybe with a label, and also watch the repos mentioned in those tickets?
yoleaux jmerelo: What kind of a name is "AlexDaniel:"?!
jmerelo .tell AlexDaniel Can we use perl6/ecosystem, maybe with a label, and also watch the repos mentioned in those tickets?
yoleaux jmerelo: I'll pass your message to AlexDaniel.
Geth doc: 1cdd59e9ba | (JJ Merelo)++ | doc/Language/routines.pod6
Slight rephrasing
synopsebot Link: doc.perl6.org/language/routines
07:42 angelds left 07:52 lizmat joined
lizmat good *, #perl6! 08:03
lizmat clickbaits opensource.com/article/18/8/migrat...l-5-perl-6
jmerelo lizmat: hey!
lizmat the first of a series of 12 :-)
jmerelo lizmat: one per month? 08:04
lizmat every 2 weeks is the plan
08:07 vrurg joined, dakkar joined
jmerelo lizmat: great! 08:07
lizmat: migration, or #perl6 in general?
lizmat it's basically a blog post version of all of the research I did for my perl 5 -> perl 6 workshop that I gave at TPCiSLC and will be giving at TPCiG 08:08
jmerelo "the sum of Perl 5 and Perl 6 becomes greater than its parts" So true, lizmat 08:10
08:16 vrurg left 08:19 domidumont left 08:21 telex left 08:26 telex joined 08:28 scimon joined 08:31 epony left 08:33 telex left 08:42 sena_kun joined
El_Che notable6: weekly opensource.com/article/18/8/migrat...l-5-perl-6 08:50
notable6 El_Che, Noted!
El_Che just in case :)
lizmat :-) 09:00
09:19 vrurg joined 09:20 vrurg left 09:24 ilmari[m] joined 09:25 ilmari[m] is now known as ilmari_ 09:26 [particle] joined 09:27 Sgeo_ joined 09:28 kerframil joined, scott joined 09:29 Sgeo left 09:32 telex joined 09:35 dotdotdot joined 09:46 Kaiepi left 09:51 domidumont joined 09:53 HaraldJoerg joined 10:00 benjikun left 10:01 AlexDaniel joined
AlexDaniel . 10:01
yoleaux 07:31Z <jmerelo> AlexDaniel: Can we use perl6/ecosystem, maybe with a label, and also watch the repos mentioned in those tickets?
jmerelo Hi, Alex
AlexDaniel o/
I'd prefer a separate repo
tbrowder_ lizmat: excellent article! great graphics, too! 10:02
lizmat tbrowder_: thank you :-)
jmerelo AlexDaniel: just the repo, or also watching issues on some modules we'll be working on.
AlexDaniel jmerelo: we can't really watch other repos 10:03
jmerelo: need admin access to set up webhooks
tbrowder_ tyil: watching? 10:04
tyil tbrowder_: I am here, though I was not watching this channel in particular rn 10:05
I have time tho
tbrowder_ tyil i am now on your side, defn will have term attribute and definition as @contents of para blocks 10:06
tyil \o/ 10:07
what happened? :p
tbrowder_ rereading S26 shows the subtle but precise nature of Damian’s writing
that convinced me
tyil ah 10:10
it works for me :D
(mostly because now I wont have to update a module I'm working on if it uses .contents like the rest) 10:11
tbrowder_ thinking about how =defn would be best rendered in text, html, and other formats also helped
jmerelo AlexDaniel: I didn't know that. I thought it was polling or somethign like that.
AlexDaniel: anyway, I don't know if I have permission to set up a new repo in perl6?
AlexDaniel jmerelo: you should
tbrowder_ is your module publicly available now?
jmerelo AlexDaniel: should I try? Like bitrot-august-18?
AlexDaniel: or just bitrot and we use it for other bitrot squashathons? 10:12
AlexDaniel jmerelo: just ecosystem-bitrot will do
or unbitrot?
jmerelo github.com/perl6/ecosystem-unbitrot AlexDaniel 10:13
tyil tbrowder_: the sources are, but its not on cpan yet, as its still under construction
tbrowder_ link? 10:14
jmerelo AlexDaniel: all members of the perl6 team have access to it. Any other thing I should set up? Maybe a README with links to places?
AlexDaniel jmerelo: so, we need a squashathon guide
scimon Can you have a different trait to `is` ? 10:15
jmerelo AlexDaniel: OK, I'll do that
AlexDaniel jmerelo: something similar to this: github.com/rakudo/rakudo/wiki/Raku...thon-Guide
jmerelo: but specific to ecosystem stuff
tyil tbrowder_: pm
jmerelo AlexDaniel: first bitrot squashathon, right?
AlexDaniel jmerelo: yeah
jmerelo AlexDaniel: OK, I'll have a go at it a bit later. Meanwhile, feel free... 10:16
AlexDaniel jmerelo: I think I'll write a script to populate issues… 10:17
jmerelo That would be great
AlexDaniel I'm actually out of town, so not sure how much internet I'd have 10:18
jmerelo AlexDaniel: there's a module for that... Give me one second.
tbrowder_ tyil: good, thanks
AlexDaniel jmerelo: I've never seen a proper module for github api
jmerelo: just stuff that creates gists
jmerelo AlexDaniel: there's this one github.com/fayland/perl6-WebService-GitHub 10:19
AlexDaniel for a throwaway thing like this I can just do a regular POST and that'd be it
jmerelo and then my own version: github.com/JJ/perl6-WebService-GitHub/ which I used during the grant.
AlexDaniel github.com/fayland/perl6-WebServic...sues.pm#L6 10:20
jmerelo AlexDaniel: It would be a good addition to the module. Anyway, if you do the throwaway stuff I might incorporate it into my module
scimon For those of us joining half way into the conversation is this a squahathon to handle broken modules post a compiler update?
jmerelo scimon: right
AlexDaniel wrong?
scimon: during this squashathon we try to fix modules that were red for a long time 10:21
jmerelo AlexDaniel: I mean, it's not a big deal to add that to my version of the WebServices module. I'm thinking about publishing it, BTW... Since fayland is not too keen on accepting my pull request.
AlexDaniel scimon: we try really hard not to break modules with rakudo releases 10:22
there are exceptions, however… like if a module depends on a bug or something like that
but mostly I submit PRs before the release myself :) 10:23
scimon: toast.perl6.party/ 10:24
scimon: most of the red modules you can test with some 2017.* release and they'd be red, it's not release related… just some oooold modules :) 10:25
scimon: hm, do you have any idea on how we can prioritize them? 700+ modules to fix is maybe too much… 10:26
jmerelo AlexDaniel: we have the "river" score... 10:32
sena_kun those who have native dependency can be omitted(or need more complex check than toast do), I believe, there are a lot in this category.
jmerelo sena_kun: again, the river score takes care of that. 10:33
You fix something higher up the river, then those which are downriver from it are automatically fixed.
AlexDaniel jmerelo: I think they're talking about ‘missing blabla.so’ kind of failures
sena_kun jmerelo, oh, then never mind.
I do.
jmerelo AlexDaniel: Ah, OK.
sena_kun e.g. all ZeroMQ modules are failing. 10:34
AlexDaniel sena_kun: I don't know… it would be nice to know the list of libs that need to be installed
jmerelo AlexDaniel, sena_kun there was a proposal by samcv, if I remember correctly, to include non-Perl6 dependencies into meta6.json
sena_kun I have a feeling we were here before...
AlexDaniel sena_kun: so this one is `apt install gnuradio-dev` I think? 10:35
sena_kun AlexDaniel, I can probably take a look on such a list if time allows after $daily-comma-job.
AlexDaniel, sorry, what do you mean by gnuradio-dev? 10:36
AlexDaniel sena_kun: well, that package has libgnuradio-zeromq.so file, but I'm not sure if it's the right one
sena_kun AlexDaniel, ah. for zmq support you need to look for `zeromq` package I believe. gnuradio-zmq one is probably some in-between layer. 10:38
AlexDaniel hm interesting 10:39
there's zeromq3 package but I don't see it from my machine
anyway, afk
sena_kun I see ubuntu doesn't have such a package, huh. they offer by hands build instructions.
e.g. www.digitalocean.com/community/que...4-04-4-x64
or gist.github.com/katopz/8b766a5cb0c...9407e83d00 10:40
anyway, zmq is not the main issue, we want a list of dependencies, so I'll look what I can do. Now back to making new fancy features~ 10:41
10:41 HaraldJoerg left 10:46 chsanch joined
xq packages.ubuntu.com/search?keywords=libzmq-dev 10:54
scimon Got it (Sorry $WORK got in the way). 10:57
xq same for gnuradio thing packages.ubuntu.com/search?keyword...dio-zeromq
scimon When is it planned for? 10:58
(He says before leaving to get lunch) 10:59
11:00 robertle joined 11:01 chsanch left 11:04 HaraldJoerg joined 11:12 Xliff joined
Xliff \o 11:12
Are there any web pages that discuss how to deal with grammers that have recursive rules?
tyil Xliff: theres a book about it 11:13
Xliff: www.amazon.com/Parsing-Perl-Regexe...1484232275
Xliff Yes. I know. I have it.
It does not mention anything about recursive rules. Though I have to admit, I haven't read it all the way through. 11:14
moritz it talks about recusion
Xliff moritz++
I was about to .tell you a question. Can you tell me the chapter?
P 112? 11:15
moritz I'm searching for my ebook copy right now :-)
page 113, Left Recursion and Other Traps 11:16
Xliff Thanks. That's exactly what I am looking for. 11:17
11:17 Zoffix joined
Zoffix .tell MasterDuke don't know if you get notifications, but would you know the cause of github.com/rakudo/rakudo/issues/2157 ? Also, would you add comments documenting the magic numbers? Why would nqp::iseq_i(0x1ffffffff,8589934591) be false on 32bit for example, or what's the significance of 16 characters (~53bit number)? 11:18
yoleaux Zoffix: I'll pass your message to MasterDuke.
Zoffix .tell MasterDuke nm the `iseq_i`, I see on HEAD it's not an `add_i`, but what's special about 53-bit numbers? 11:21
yoleaux Zoffix: I'll pass your message to MasterDuke.
Zoffix *now
11:22 Zoffix left
jmerelo Problems again with "trying to assign to an immutable value" when using roles... stackoverflow.com/questions/516323...ed-in-role 11:43
11:44 Kaiepi joined
lizmat jmerelo: wouldn't it make more sense to just create an Enum and mix the $.index into it ? 11:47
jmerelo lizmat: I want to use Enumeration by itself, to create the doc. 11:48
lizmat Ah, ok, I was immediately glued to the code :-)
jmerelo lizmat: Enumerations have $!index, and theoretically it's up to the MOP to change index. But it should be possible to create your own Enumerations
lizmat but is the index a public attribute? 11:49
jmerelo lizmat: no, it's private.
lizmat because if it is not, you can't set it with bless
jmerelo lizmat: you can't set it from BUILD or TWEAK since it yields an immutable error 11:50
lizmat and if it is an native int, there's currently an issue about not being able to set native ints in roles
from the class consuming the role
jmerelo lizmat: Ah. That's it, then.
lizmat you might want to verify, I might be wrong :-) 11:51
jmerelo lizmat: probably not. OK, without it the example does not make a lot of sense anyway... Can it maybe be changed to Int while that is fixed? Or would that be a big performance issue? 11:52
lizmat I'm not sure actually
wrt performance
jmerelo lizmat: OK, thanks anyway 11:53
lizmat: I'm AFK now. See you!
lizmat sorry I couldn't be of help more, am deep somewhere else in the code :-)
11:53 jmerelo left 12:02 Zoffix joined, araujo left
Zoffix .tell MasterDuke nm, the bisect is bogus it seems as the commit is in Rakudo, but the problem exists in nqp as well 12:02
yoleaux Zoffix: I'll pass your message to MasterDuke.
12:02 Zoffix left 12:03 pmurias joined 12:15 curan left 12:16 curan joined 12:21 AlexDaniel left 12:22 AlexDaniel joined 12:23 AlexDaniel left 12:24 AlexDaniel joined
AlexDaniel oh my, github keeps changing the interface all the freaking time :) 12:26
sena_kun and the new one is awful. :/
AlexDaniel yeah, I'm trying out the beta 12:28
don't know if it's better or worse, it's just completely different, I feel disoriented
“Discover repositories” what, why 12:29
12:30 Extirpator joined
sena_kun it feels like they don't have a designer that actually studied how UI should look and work. too much elements, element positioning doesn't align with how human eye browse pages, too wide, ugh. 12:30
*too many 12:31
Extirpator Hi, friends! I've confused with var interpolation at regexes with P5 modifier. I need some help. This code fails: perl6 -e 'my $r="[23]+"; my Regex $rx=rx:P5/$r/; if "232323" ~~ $rx {"matched".say};', but this works: perl6 -e 'my Regex $rx=rx:P5/[23]+/; if "232323" ~~ $rx {"matched".say};' 12:32
AlexDaniel hmm, with p6 regexes that'd be this: 12:35
Extirpator perl6 -e 'my $r="[23]+"; my Regex $rx=rx/<$r>/; if "232323" ~~ $rx {"matched".say};'
AlexDaniel m: my $r="[23]+"; my Regex $rx=rx/<{$r}>/; if "232323" ~~ $rx {"matched".say}; 12:36
camelia matched
Extirpator yes it works!
AlexDaniel or that, yeah
Extirpator but i need p5 :)
MasterDuke m: my $r="[23]+"; my Regex $rx=rx/<$r>/; if "232323" ~~ $rx {"matched".say};
yoleaux 11:18Z <Zoffix> MasterDuke: don't know if you get notifications, but would you know the cause of github.com/rakudo/rakudo/issues/2157 ? Also, would you add comments documenting the magic numbers? Why would nqp::iseq_i(0x1ffffffff,8589934591) be false on 32bit for example, or what's the significance of 16 characters (~53bit number)?
camelia matched
yoleaux 11:21Z <Zoffix> MasterDuke: nm the `iseq_i`, I see on HEAD it's not an `add_i`, but what's special about 53-bit numbers?
12:02Z <Zoffix> MasterDuke: nm, the bisect is bogus it seems as the commit is in Rakudo, but the problem exists in nqp as well
AlexDaniel MasterDuke: now with :P5 :)
MasterDuke yeah, that doesn't seem to work 12:37
12:37 Zoffix joined
Zoffix MasterDuke: never mind that stuff. I found the problem 12:37
MasterDuke Extirpator: but fwiw, you'd need to put <> around the variable to interpolate it as a regex, not just text
Zoffix (along with specialness of 53-bit numbers)
MasterDuke Zoffix: i just commented on the issue
Zoffix MasterDuke: yup, that is it. It's using doubles for ints 12:38
And numbers up to 2**52 can be represented exactly in double
m: dd 9930972392403501e0.Int
camelia 9930972392403500
AlexDaniel oh my
Zoffix m: dd 9930972392403503e0.Int # <--- that's ...504 12:39
camelia 9930972392403504
Zoffix m: dd 9930972392403503 + 1 # <--- so it follows that this would be ...505
camelia 9930972392403505
Zoffix :)
Well, I'm just gonna dial it down in Rakudo to be within representable limit to fix the bug and I guess N#363 covers the NQP side of the problem 12:40
synopsebot N#363 [open]: github.com/perl6/nqp/issues/363 NQP converts large integers to floats when it shouldn't/doesn't need to
Extirpator Masterduke: here's P5 code - perl -e 'my $r="(?:[23]+)"; if ("232323" =~ /$r/) {print "matched\n"};', i need to exec P5-like regex in P6: perl6 -e 'my $r="(?:[23]+)"; my Regex $rx=rx:P5/$r/; if "232323" ~~ $rx {"matched".say};', <$r>, <{$r}> are not work too 12:44
timotimo maybe you'll have to go through EVAL for now, could be a bug or NYI 12:45
MasterDuke Extirpator: there does seem to be a bug, but you'd definitely have to use <$r>, just $r by itself doesn't work like that 12:46
timotimo MasterDuke: i'm pretty sure it does work like that in perl5, though?
MasterDuke docs.perl6.org/language/regexes#in...erpolation
Zoffix Yeah, in P5 the default is to treat text as regex 12:47
MasterDuke hm, good point. don't know if :P5 should work that way or not
timotimo it should, i think
Zoffix it shouldn't, I think
timotimo if you want to put text in a variable and match it directly, put \Q and \E
Zoffix What happens with @foo?
timotimo whatever perl5 does with it
12:47 sftp_ joined
Zoffix OK 12:48
What happens with $r := [<a b c>] ?
What happens with $r := [<a b c>] => rx:P5/$r/
12:48 domidumont left, aindilis left, andrzejku left, ExtraCrispy left, sftp left, SyrupThinker left, sergot left, Praise left, cognominal left, cxreg left 12:49 renormalist left, tinita left, freeze left
timotimo :P5 is there to support perl5 style regexes 12:49
12:49 sftp_ is now known as sftp
timotimo if you use something perl5 just can't do, throw an exception or something 12:49
Zoffix timotimo: right, but you're crossing over Perl 6 objects into the space of Perl 5 regexes. There aren't Str or Array objects in Perl 5. `$r` cannot contain an Array in P5, but it can in P6; `$r` can contain a non-P5 regex in Perl 6 12:50
So you either have to define semantics for all of that stuff, or just draw the line at "variables work like in P6" :)
timotimo what if you put an arrayref into $r in perl5?
12:51 ccntrq left, yoleaux left
Zoffix timotimo: it gets stringified... literally: /(?^u:ARRAY(0x14b5f30))/ 12:51
12:51 reportable6 left, squashable6 left, undersightable6 left, statisfiable6 left, greppable6 left, benchable6 left, nativecallable6 left
timotimo i bet someone on darkpan is relying on that 12:52
Zoffix heh
timotimo what is ?^u: do?
12:53 tinita joined
Zoffix locally enables /u modifier 12:53
timotimo ah, just unicode 12:54
so, don't you think someone will use /$r/ to check if $r is the same arrayref as something else by going through regex? :)
Zoffix "These modifiers, all new in 5.14, affect which character-set rules [...] /u" sets the character set to Unicode."
timotimo like, $r ~= /$r/ would give True
Zoffix timotimo: dunno, there aren't no arrayrefs in p6 12:55
MasterDuke +1 for "variables work like in P6"
12:55 Altreus joined 12:56 Altreus is now known as Guest57458 13:00 domidumont joined
Extirpator timotimo: "put \Q and \E", you mean: my Regex $rx=rx:P5/\Q$r\E? 13:00
Zoffix Extirpator: no, we were just talking about how to make Perl 5 regex behave with variables like Perl 6 does
Extirpator :) 13:01
Zoffix Extirpator: just use Perl 6 regex, if you can.
13:01 SyrupThinker joined
Zoffix Extirpator: there was a recent intro talk about them, if you have any interest: rakudo.party/post/Intro-Into-Perl6...-Recording 13:01
13:01 vrurg joined
Extirpator i'm porting a P5 module, everything is done but regexes 13:01
Zoffix I see 13:02
Extirpator i want to leave them with P5 style
13:04 vrurg left, vrurg joined
Extirpator MasterDuke, timotimo, Zoffix: thank you guys, P6 rules! 13:05
Zoffix \o/ 13:06
.oO( but sometimes P6 tokens! )
13:07 vrurg left, vrurg joined 13:09 vrurg left
jkramer Does anyone know why this is so incredibly slow (and how to fix it)? bpaste.net/show/0ee15ed17b8e 13:09
Input is this word list: raw.githubusercontent.com/dolph/di...nable1.txt
13:10 Zoffix left, vrurg joined
jkramer Is it just the size of the lists to grep? 13:12
lizmat jkramer: have you tried running it with --profile ?
13:13 KotH joined
lizmat also: why isn't &works an ordinary sub ? 13:13
in any case, --profile should tell you a lot 13:14
jkramer lizmat: I thought making @chars a temp might be cheaper than calling a sub with @chars as "is copy" 13:15
lizmat I'm still in the middle of something else, will look at it after that 13:16
Extirpator lizmat: I've just used your P5quotemeta module. It fails sometimes at test stage during zef install. Failure is not repeat constantly. I have to create issue, but maybe you know about it... Here's capture from docker: gitlab.com/pheix/router-right-perl...s/85819394 13:19
lizmat Extirpator: do you have the latest rakudo ?
it was one of the issues that jnthn looked at / fixed before he went on holiday, afaik
tyil according to the build output, rakudo-star:Latest docker image 13:20
so that would not be latest rakudo, but 2018.04 13:21
lizmat hmmm... strange
Extirpator 2018.04.1
lizmat fwiw, I can't repro it on HEAD
tyil ah yes, that one had a point release 13:22
lizmat oddly enough, the breakage dates from post 2017.06, afaik
and the breakage was in MoarVM
Extirpator: could it be that you're somehow using a more recent MoarVM ?
tyil I'll pull it on docker and see what perl6 --version replies in that container 13:23
13:23 Xliff left
tyil I'm on the train though, so responses may be delayed 13:23
lizmat tyil++ 13:24
13:24 mcmillhj joined
tyil root@7568b54b732d:/# perl6 --version 13:24
This is Rakudo Star version 2018.04.1 built on MoarVM version 2018.04.1
Extirpator lizmat: yes, i've made a few installations before 13:25
tyil updates his default test script to run perl6 --version first
that output is always useful when you're trying to get help :p
lizmat Extirpator: well, I'm not sure how I can help you atm, as I can't reproduce it on HEAD
Extirpator: I think you're fine if you don't use the quotemeta() without parameters, which relies on the outer $_ 13:26
tyil rakudo star is still planned to get a new release soonish, right?
lizmat tyil: there's a 2018.06.RC0
tyil oh, I'll test that soon-ish then
lizmat not valid in sprintf 13:27
argh, pasto
tyil I don't see any such release on the rakudo.org site :( 13:28
timotimo Zoffix, i wonder if it would be prudent to ask before taking "another happy perl6 user" screenshots and posting them on social media; it's one thing to express your joy in an irc channel, it's another - to me at least - to post it to social media where it can have infinite range, plus the weekly usually links to such tweets as well 13:29
jkramer: i wonder if it's worth to convert @chars to a Set once, since the grep would coerce it over and over to use (<) on it, i think? 13:30
lizmat tyil: it's a Release Candidate, no official release yet
mst lizmat: I've +r'ed -dev and moarvm
looks like they've pre-seeded the bots with channels this time so our network level mitigations aren't as effective
lizmat mst++
AlexDaniel heh what a waste of time 13:31
mst also your bots are going to need to log in to nickserv before joining channels
sorry about that
AlexDaniel mst: they can't, unfortunately. There's a limit for 5 nickserv idents
13:32 Zoffix joined
Zoffix huggable: test 13:32
huggable Zoffix, Test failed
mst oh, hang on, I can I-line them I think
13:32 psychoslave joined
Zoffix timotimo: OK 13:32
lizmat timotimo: perhaps blot out the nick ? 13:33
13:34 Zoffix left 13:35 nativecallable6 joined, ChanServ sets mode: +v nativecallable6
timotimo i don't think anybody has complained yet about the "happy perl6 user" tweets, but i don't know if they were aware of them in the first place 13:39
lizmat but blotting out the name wouldn't detract from the PR value, would it? 13:41
Zoffix: don't you agree ?
timotimo i'd think it's enough 13:42
to blot it out, i mean
13:43 ChanServ sets mode: +o mst, mst sets mode: -r 13:44 mst sets mode: +b $~a
jkramer timotimo: I tried that initially but it didn't change anything. Actually the (<) is not even used anymore (notice line is commented out) because it didn't seem to help speed it up. 13:44
timotimo oh, right
i'll have a look with the profiler
13:45 aindilis joined 13:46 greppable6 joined, reportable6 joined, statisfiable6 joined
timotimo how long does it run? 13:46
jkramer pste.eu/p/GQRo.html 13:47
13:47 notable6 joined
jkramer This is just for the first 10 words (however using the whole dictionary). Takes ~25 seconds 13:47
timotimo oh, good
13:47 squashable6 joined
timotimo oh, perlseen takes a whole lot of time, sounds like .perl on something is taking a long time 13:47
jkramer Well, on my server, but the HW is not the worst :)
13:47 benchable6 joined
timotimo oof 13:48
.gist on capture, eh?
throw out the "is cached"
tyil lizmat: I understand, but where can I download it to test it out?
13:49 undersightable6 joined
jkramer timotimo: I suspected that too but without cached it's still really slow and when operating on the full dictionary it might actually be useful later 13:49
W/o cached it's still 25-26 seconds
timotimo OK 13:50
lizmat jkramer: perhaps modules.perl6.org/dist/Sub::Memoize...:ELIZABETH could be of help?
don't look at the tag line, I just realized that suffers from a copy-pasto
timotimo i forgot to give the filename on the commandline %) 13:51
AlexDaniel dpk: yoleaux seems to be down?
MasterDuke there's a lot of printing. with a reduced size input, directing output to /dev/null was quite a bit faster
re jkramer's script 13:52
AlexDaniel dpk: it should be identified to join, but I thought that's not an issue?
jkramer timotimo: Actually removing 'is cached' DOES make it faster, I had changed the code in the meantime and added it to another sub :) 13:54
It's now ~6 seconds for the first 10 words
timotimo instead of 25?
jkramer Yes :) 13:55
timotimo TBH, cutting down enable1.txt with head -n 10 it finishes in 0.35s
oh, surely you didn't do it like that?
jkramer Well I made the &works block a sub and added 'is cached', which was not a good idea obviously :)
13:56 yoleaux joined, ChanServ sets mode: +v yoleaux
dpk sorted 13:56
it just crashed and didn't come back up for some reason
jkramer timotimo: No, I use the whole dictionary, just in the last line I
... I changed the last line to: .&longest-chain.Str.say for @words[^10];
13:57 AlexDaniel` left
timotimo What? 13:58
jkramer I hit enter too early :)
timotimo yeah, 5 seconds on my end, too
AlexDaniel I… didn't mean that… oops :)
timotimo though i kept "is cached" on longest-chain
buggable New CPAN upload: Sub-Memoized-0.0.3.tar.gz by ELIZABETH modules.perl6.org/dist/Sub::Memoize...:ELIZABETH
AlexDaniel just ghosted his matrix self
jkramer Yes, I just added that one back too, now it's ~5.5 instead of ~6.
timotimo how did we have 25s before? :) 13:59
jkramer timotimo: I added 'is cached' to the &works thing :)
(And also outsourced it into a sub) 14:00
MasterDuke jkramer: can you gist what you have now? 14:03
jkramer Yup, one second
bpaste.net/show/f3d52cd5fe18 14:04
timotimo something i see is that the machinery for temp is taking 12% of total time 14:05
jkramer timotimo: Check out the new paste, I made it its own sub without temp.