»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
00:03 cdg_ joined 00:04 cdg left 00:09 cdg_ left 00:10 jnthn joined 00:12 cdg joined
sjn reads the weekly changes blogpost 00:14
where's the squashaton announced?
timotimo squashable6: status 00:16
squashable6 timotimo, Next SQUASHathon in 31 days and ≈9 hours (2018-01-06 UTC-12⌁UTC+14). See github.com/rakudo/rakudo/wiki/Mont...Squash-Day
00:16 rmusial left, mcmillhj left
sjn ah, cool 00:17
timotimo not sure where else we announce it
since not everybody's on irc
sjn sounds like a blog post about community activit 00:18
activities* 00:19
doesn't seem like it's for newbies though (low-hanging/simple bugs aren't mentioned) 00:21
timotimo it's meant to be for newbies; people can get as much assistance as they need from irc at least
sjn Hm. The february squashathon is at the same time as FOSDEM 00:24
Maybe an idea to highjack a table in the cantina and have a physical squashathon there? 00:25
AlexDaniel that's a wonderful idea. In fact, there's a small chance that I'd be able to get my ass there 00:33
sjn: this month it kinda fell through the cracks and wasn't advertized much 00:34
00:40 xcm left 00:42 xcm joined
sjn It might be a nice way to get people to try out Perl 6 00:44
AlexDaniel yes 00:45
sjn although, I guess the Cantina staff and/or FOSDEM orgas probably don't like the thought of using the cantina for organizing a hackathon
(not to mention issues around power outletes & networking) 00:46
AlexDaniel if you can help with anything squashathon-related, please do
this month I even had a problem coming up with the topic :) 00:47
00:48 itaipu left 00:49 mson joined
sjn Hm. Didn't the the FOSDEM orgas at some point have a dedicated hackathon room? 00:51
AlexDaniel: nothing wrong in recycling topics :) 00:52
AlexDaniel sjn: sure, but doing a doc squashathon two times in a row is a bit boring 00:53
sjn AlexDaniel: as for helping, I don't know yet. I'll keep it in mind.
AlexDaniel 👍
00:54 pat_js left
sjn Maybe the thing is to find ways to make hackathons more interesting? :) (e.g. inchstones, competitions, beer or a combination of these) 00:55
00:55 patrickz_ joined
sjn Having a hackathon progrss dashboard might help 00:56
sjn imagines a track with differently-coloured larvae competing to become first over a goal line (and turn into a camelia) 00:58
00:58 patrickz left
sjn and progress on the track is made by the amount of karma messages (e.g. AlexDaniel++ ) on #perl6 for that person within 5 minutes of an accepted pull-request 00:59
00:59 evalable6 left, evalable6 joined
AlexDaniel well, squashable6 already creates a log like this: gist.github.com/Whateverable/fea65...df5cd397d8 01:00
sjn that looks like a good start :) 01:01
01:02 margeas left 01:06 kaare__ joined 01:08 cdg left, kaare_ left 01:10 ctilmes joined 01:20 lookatme joined
lookatme hi 01:23
morning o/
sjn it's good night in Oslo :) 01:36
lookatme :)
sjn in #perl6, we often say "good *, #perl6" for timezone reasons :) 01:38
lookatme good morning #perl6 :) 01:39
Hah
I am reading the advent calendar
sjn yeah, lots of good stuff there 01:43
01:43 chenyf joined 01:56 ccntrq joined 01:57 Cabanossi left 01:58 NEveD joined, mcmillhj joined 01:59 chenyf left, Cabanossi joined 02:01 comborico1611 left 02:04 perlpilot_ joined 02:10 kitsunenokenja joined 02:16 TimToady joined 02:20 perigrin joined, perigrin left 02:21 perigrin joined 02:25 mcmillhj left 02:28 troys_ is now known as troys 02:44 Voldenet joined, Voldenet left, Voldenet joined 02:45 perlpilot_ left 02:46 ilbot3 left
SmokeMachine hi there! 02:52
02:52 pavel_ joined
SmokeMachine m: my %a = a => (1,2,3); class B {has @.a}; my $b = B.new: |%a; dd $b.a # How can I make the attr a do not be an array of a list, but be only the list or an array 02:53
camelia Array @!a = [(1, 2, 3),]
SmokeMachine ?
pavel_ Quick question: Where is there a changelog posted? I'm considering building the latest version of perl6 since the version in my distro's repo is outdated, and I'd like to know if it's worth the effort. 02:54
AlexDaniel pavel_: github.com/rakudo/rakudo/blob/mast.../ChangeLog 02:59
02:59 ilbot3 joined, ChanServ sets mode: +v ilbot3
pavel_ Thanks much! 02:59
AlexDaniel pavel_: and MoarVM has its own changelog: github.com/MoarVM/MoarVM/blob/mast.../ChangeLog
pavel_: what version is in your distro? And what's the distro? 03:01
pavel_ Fedora, and 2017.08
It's not that out of date, but it's for an online interpreter, and people want the latest features. 03:02
AlexDaniel huggable: deb
huggable AlexDaniel, CentOS, Debian, Fedora and Ubuntu Rakudo packages: github.com/nxadm/rakudo-pkg/releases
AlexDaniel pavel_: what about taking an rpm from there ↑ ?
pavel_ Oh that's really neat! I think I'll do that. 03:03
Thanks!
AlexDaniel pavel_: may I also ask what online interpreter it is? :)
pavel_ tio.run/#perl6 03:04
It's not *actually* maintained by me, but someone I know. I'm just helping out.
tio.run supports several hundred programming languages, and new ones are added basically as fast as the maintainer finds out about their existance :P 03:05
lookatme m: my %a = a => [1,2,3]; class B {has @.a}; my $b = B.new: |%a; dd $b.a
camelia Array @!a = [[1, 2, 3],]
AlexDaniel pavel_: that's pretty cool 03:06
lookatme m: my %a = a => [1,2,3]; class B {has @.a; method TWEAK(:$a) { @!a := $a; }}; my $b = B.new: |%a; dd $b.a 03:07
camelia [1, 2, 3]
lookatme SmokeMachine, ^^ works but IDK if there exist more convenient way to do it 03:08
03:08 mson left
AlexDaniel huggable: try :is: f.perlbot.pl/#perl6 tio.run/#perl6 glot.io/new/perl6 ideone.com/ 03:10
huggable AlexDaniel, Added try as f.perlbot.pl/#perl6 tio.run/#perl6 glot.io/new/perl6 ideone.com/
AlexDaniel ↑ if anyone knows more let me know 03:11
I mean, don't let me know, just edit the entry
lookatme :) use above command ? 03:12
AlexDaniel yea
lol ideone lists perl 6 as just “perl”. Whoops. 03:13
lookatme I remember it was perl6 long time ago, maybe something bug in that site 03:14
It has two perl
SmokeMachine lookatme: thanks... but I need to use that for a class that wasn't created by me... :( 03:15
but is that expected?
m: my %a = a => (1,2,3); class B {has @.a}; my $b = B.new: |%a; dd $b.a # is this expected?
camelia Array @!a = [(1, 2, 3),]
03:21 Voldenet left 03:24 ctilmes left
pavel_ huggable try 03:25
huggable: try
huggable pavel_, f.perlbot.pl/#perl6 tio.run/#perl6 glot.io/new/perl6 ideone.com/
pavel_ Neato
sjn m: my %a := a => (1,2,3); class B {has @.a}; my $b = B.new: |%a; dd $b.a
camelia Array @!a = [1, 2, 3]
sjn SmokeMachine: is that what you're looking for?
03:25 Voldenet joined, Voldenet left, Voldenet joined 03:27 Cabanossi left 03:29 Cabanossi joined
SmokeMachine sjn: yes! I think that’s it! 03:35
sjn SmokeMachine: check out this blog zoffix++ wrote about the topic: perl6advent.wordpress.com/2017/12/...ontainers/ 03:36
SmokeMachine m: my %a := a => (1,2,3); dd %a
camelia :a((1, 2, 3))
SmokeMachine sjn: my real problem is here: github.com/FCO/p6-react/blob/cb302...nt.pm6#L31 03:42
I’ll see if I can use your suggestion to fix that tomorrow... thanks 03:43
sjn I have no idea what I'm looking at there 03:44
03:45 troys is now known as troys_ 03:52 pavel_ left 03:53 xinming joined 04:11 Cabanossi left 04:12 Cabanossi joined 04:19 mson joined 04:30 samcv joined 04:38 kitsunenokenja left 04:42 Cabanossi left 04:43 Cabanossi joined 04:50 BinGOs left, tyil[m] joined, wander joined 04:51 BinGOs joined 04:55 troys_ is now known as troys
lookatme m: my %a = a => (1,2,3); class B {has @.a}; my $b = B.new: %a<>; dd $b.a 05:01
camelia Default constructor for 'B' only takes named arguments
in block <unit> at <tmp> line 1
lookatme m: my %a = a => (1,2,3); class B {has @.a}; my $b = B.new: %a>><>; dd $b.a
camelia Default constructor for 'B' only takes named arguments
in block <unit> at <tmp> line 1
Geth doc: c75c186d1c | (Alex Chen)++ | doc/Language/concurrency.pod6
Fix #1697
 05:04
synopsebot Link: doc.perl6.org/language/concurrency
05:09 eliasr left 05:12 daxim joined 05:23 AlexDaniel left 05:25 ChoHag joined
SmokeMachine m: my %a = a => (1,2,3); class B {has @.a}; my $b = B.new: |%a>><>; dd $b.a 05:30
camelia Array @!a = [1]
SmokeMachine m: my %a = a => (1,2,3); class B {has @.a}; my $b = B.new: |%a<>; dd $b.a 05:31
camelia Array @!a = [(1, 2, 3),]
05:35 ivans left
piojo wander: thanks for fixing my bug just now :) 05:37
The same problem appears in doc/Type/Supply.pod6
I'll fix it if you don't see this 05:39
Geth doc: 4054e408cd | (Alex Chen)++ | doc/Type/Supply.pod6
Fix broken link
synopsebot Link: doc.perl6.org/type/Supply
wander just see it :) 05:40
piojo wander: great, thanks! 05:41
05:46 daxim left 05:47 daxim joined 05:48 khw left 05:50 troys left 06:04 releasable6 left, releasable6 joined, greppable6 left, greppable6 joined, squashable6 left 06:05 squashable6 joined 06:17 AlexDaniel joined 06:26 ChoHag left 06:29 mson left
tyil I'm trying to use the alpine-perl6:latest docker image, but installing OpenSSL fails its test, saying libssl.so cant be found. I've tried adding an `apk add libssl1.0` but that didn't seem to fix the issue. Anyone here who might have any idea? 06:32
06:32 konsolebox left 06:35 konsolebox joined
geekosaur alpine has dev packages 06:53
libsso.so without a version is a dev package (and, boo hiss, whatever is doing that really shouldn't; it will likely break if the library ABI changes) 06:54
pkgs.alpinelinux.org/contents?file...&arch= 06:55
(ssl is also a special case since openssl and libressl can be built using (mostly) the same ABI and can both provide the dev package)
unfortunately I suspect that's also why it uses the dev package :/ 06:57
tyil ah 07:11
07:11 domidumont joined
tyil thats an odd choice for a convention, but if it works for them 07:13
07:16 parv joined, geospeck joined, domidumont left, domidumont joined
tyil it passes more tests now, but OpenSSL seems to not work as expected in this docker :( 07:27
El_Che my experience with alpine that it's fine for small self-contained projects 07:28
everything bigger ends in wtfs and lost time 07:29
tyil: use ubuntu or debian-slim and wget rekudo from this site in your image: github.com/nxadm/rakudo-pkg/releases ? 07:30
tyil that'll have to wait until I'm done at work I think, but if that's the suggested solution I'll use that instead
El_Che it's not *the* suggested solution, just mine 07:31
if you tinker enough with alpine you'll get things working
El_Che remember the tinkering time spent to get alpine Rakudo packages built 07:32
s
tyil :p
07:32 aindilis left
El_Che I spent more time on alpine than on all other distros I support together 07:32
07:34 AlexDaniel left 07:35 reportable6 left, reportable6 joined, ChanServ sets mode: +v reportable6
tyil I havent tried getting a p6 app into a docker container, this was just the first I found on dockerhub 07:36
El_Che I am working on that as well, but I haven't released it 07:38
the idea is to use my binary packages instead of using rakudobrew
the image build time would be seconds
tyil I dont mind having a master image that compiles it tbh
El_Che I do :)
tyil :p 07:39
El_Che I use CI heavily
every commit generates a new docker image
tyil as a funtoo user it would be weird to complain on compiling a single master image :p
El_Che hehe
tyil but yeah, for CI I prefer non-compiled images, since compiling takes longer than testing
and its just a drag thats not necessary there 07:40
El_Che also, if you're not testing rakudo itself, compiling has no added value in the chain
tyil making sure that you get the original, non-modified source
07:40 darutoko joined
tyil perl 6 doesnt have much compile-time options afaik, so there's not much to win in customization there 07:41
El_Che original source: that's the reasonthe only upstream images I use are the base images from the distro's and I build everything else on top myself 07:42
(being part of a security team makes you paranoid :) ) 07:43
tyil you dont need to be part of a security team for that
these days all you need is to read the news
El_Che :) 07:44
07:47 parv left 07:48 |oLa| joined 07:53 ChoHag joined 07:55 Cabanossi left 07:58 Cabanossi joined 08:04 sena_kun joined 08:05 xcm left, |oLa| left 08:07 xcm joined, lizmat left 08:08 lizmat joined
moritz is it paranoia if you know they are after you? 08:09
08:10 |oLa| joined, |oLa| left
lizmat no, that's common::sense :-) 08:11
lizmat clickbaits p6weekly.wordpress.com/2017/12/04/...ieventing/ 08:15
tyil lizmat: old news! 08:23
but still lizmat++ :> 08:24
08:30 abraxxa joined 08:31 aindilis joined
lookatme m: multi sub trait_mod:<is>(Attribute \a, :$constructor) { say a; a; }; class A { has Int $.a is constructor; }; 08:38
camelia Int $!a
08:38 lowbro joined, lowbro left, lowbro joined
lookatme m: multi sub trait_mod:<is>(Attribute \a, :$constructor) { say a.package; a; }; class A { has Int $.a is constructor; }; 08:38
camelia 5===SORRY!5=== Error while compiling <tmp>
No such method 'gist' for invocant of type 'A'
at <tmp>:1
lookatme What the problem here ? ^
08:40 abraxxa left
lookatme s/What/What's/ :) 08:40
08:41 dakkar joined, abraxxa joined
moritz lookatme: the trait runs as soon as it's parsed, so class A isn't fully composed yet 08:41
in particular, the implicit parent class Any (which provides method gist) isn't added yet
lookatme hmm, that make sense
oh 08:42
moritz btw you should make :$constructor a required param by adding a ! after it
multi sub trait_mod:<is>(Attribute \a, :$constructor!)
lookatme oh, you are right 08:43
It should be
tyil I was already wondering what that ! did in one of the advent articles 08:45
moritz tyil: which one?
tyil day 4 08:46
it has `multi sub MAIN(Bool :$test!) {`
moritz ah :-) 08:47
09:06 jeek joined 09:08 cdg joined 09:09 lookatme left
tyil news.ycombinator.com/item?id=15850497 09:16
09:25 Cabanossi left
piojo Is there a way to look up a class's method using ::?CLASS without calling it? 09:25
moritz m: say Int.^lookup('sqrt').^name 09:26
camelia Method
moritz piojo: ^^ like this
09:28 Cabanossi joined
piojo moritz: thanks! To be clear, the name is static. Is that still the right way? 09:28
It's just to set up an event handler (err, a tap) 09:29
moritz piojo: sound more like you want *.yourmethod 09:30
piojo moritz: I'm just going for: INIT { $supply.tap(::?CLASS.handler); } 09:31
but the above doesn't work since it calls the handler, and I can't seem to reference it with &::?CLASS.handler 09:32
and $?CLASS seems to not be defined in INIT
moritz so, use the class name? 09:36
piojo moritz: the class name is long, and the fact that this is early code means it'll definitely need to be refactored 09:39
Seems like the best solution is to call it:
INIT { $supply.tap: { ::?CLASS.handler($_); } }
lizmat m: class A::B::C { }; constant S = A::B::C; dd S.new # piojo: you can shorten class names 09:42
camelia A::B::C.new
piojo lizmat: That's extremely helpful. Thanks!
09:45 wander left
piojo I don't think using the class name actually works for this 09:47
m: my $method = &Int.Str; say $method;
camelia 5===SORRY!5=== Error while compiling <tmp>
Illegally post-declared type:
Int used at line 1
piojo But using a block works well enough, and .^lookup would also work 09:48
09:55 eliasr joined 10:00 virtualsue joined 10:11 virtualsue left 10:21 cdg left, jnthn left, cdg joined, jnthn joined 10:26 cdg left
Geth doc: 42344f80f2 | (Alex Chen)++ | 2 files
Fix #1689, rewrite stuff about redispatch
 10:33
10:34 wander joined
wander Not quite good, if we want, as discuss in perl6/roast/issues/357, regard our doc site as textual specs of perl6, we have more things TODO 10:38
Recur, we have 2 language pages about grammar, 3(if includes MOP) pages about OOP
Besides, re-dispatching and return statements about intersection of Control Flow and Function, we might doc them in Function and set links in CF 10:40
10:44 donaldh_ joined 10:50 pmurias joined 10:52 margeas joined 10:54 donaldh_ left, donaldh joined 10:57 Cabanossi left 10:58 Cabanossi joined
SmokeMachine m: my %a = a => (1,2,3); my %b; for %a.kv -> \k, \v {%b{k} := v<>}; class B {has @.a}; my $b = B.new: |%b; dd $b.a # :) 10:59
camelia Array @!a = [1, 2, 3]
11:01 TEttinger left
tyil lizmat: neat, I didnt know one could shorten class names like that 11:08
wish there was a more concise form to do that though, like `use My::Long::Class::Name as ShortClass;`
lizmat I think there was some speculation about that 11:12
wander tyil: github.com/rakudo/rakudo/blob/mast.../Cursor.pm 11:13
tyil wander: dont I need to have a `use` for that somewhere? 11:14
wander aha, only I show you we use it in rakudo :) 11:15
DrForr Pastebin::Gist failed while installing the Advent thing. I'll force it because it looks like a network timeout issue, but someone should know.
tyil ah
wander not reply to "[19:09] <tyil> wish there was a more concise form to do that though, like `use My::Long::Class::Name as ShortClass;`"
DrForr tyil: I've meant to add a shadow/alias thing for quite a while, especially as my packages use long class names internally which tend to make tracebacks messy. 11:16
tyil now that all my questions are being answered at random anyway, is there a standard way to bundle manpages with a module, so the module manager will install them at the right location once the module gets installed?
would be nice if we could specify pod documents in a man entry in the META6 for instance 11:17
and have zef compile those to manpages and put them in /usr/share/man (or wherever your man page directory resides)
if this is currently not possible, should I make a pr or drop this question in #perl6-toolchain? 11:18
to at least get input in some form, and potentially put it on a roadmap? 11:19
lizmat DrForr tyil there's actually syntax for that in S11:528
synopsebot Link: design.perl6.org/S11.html#line_528
lizmat use Short:name<Long::Name::Of::MOdule>
tyil oh
snap
thanks lizmat
you the best
lizmat alas, it appears NYI 11:20
DrForr Yeah, that would be why I stalled.
tyil and you shattered my dreams within 10 seconds!
lizmat but nine ugexe should be able to point where that logic lives
DrForr Too many things to do, too little time.
I've got a long Xmas vacation with scant network access, maybe while I'm writing a book outline. 11:21
(yes, bowing to the will of the majority and resurrecting my \LaTeX knowledge.) 11:22
tyil nice
write it in POD6
DrForr *don't* get me started :)
tyil hehe
DrForr (as another project alongside Pod::To::BlogspotHTML is Pod::To::LaTeX.)
tyil I currently use adoc for most things, I've done a little pod6, but that doesnt seem to be parsed by github 11:23
I was considering to rewrite App::Cpan6's adoc files to pod6 to get better at it, but I also use the adoc format to render to a manpage
DrForr I'd prefer LaTeX because I can't stand most editors.
tyil latex, pod and adoc all go through the same editor for me 11:24
11:25 Cabanossi left
moritz writes in markdown these days 11:26
DrForr I'm just something of a typographical purist, I suppose. My copy of Bringhurst is still in storage...
moritz there are a few extensions that make it book-capable
the thing I don't like about latex is that there is no shortcut for literal code
\verb|$c| is just too much overhead, compared to `$c` 11:27
DrForr Use the \listing package, and add it to a vim register.
11:28 Cabanossi joined 11:36 ctilmes joined
DrForr I suppose a markdown <-> LaTeX converter would meet my spe... oh god another project. 11:38
I wonder if it's feasible to write a Bringhurst LaTeX template... 11:42
11:43 mscha joined
mscha m: my Int @a = (1,2,3); say @a > 2; # fine 11:43
camelia True
mscha m: my int @a = (1,2,3); say @a > 2; # why not fine? 11:44
camelia Cannot resolve caller Real(array[int]: ); none of these signatures match:
(Mu:U \v: *%_)
in block <unit> at <tmp> line 1
mscha m: my int @a = (1,2,3); say +@a > 2; # workaround
camelia True
11:55 Cabanossi left
jnthn m: say Array ~~ Cool 11:56
camelia True
moritz has anybody looked at github.com/Microsoft/language-server-protocol ?
jnthn m: say array[int] ~~ Cool
camelia False
moritz my impression is that folks put a lot of work into it
sjn moritz: that one has been mentioned here at least two times the last year :) 11:57
I think it would be great to support it 11:58
11:58 Cabanossi joined
moritz sjn: ok, great to know I'm not the only one :-) 11:59
DrForr Another idea to add to the list... bugger. 12:00
It kind of screams "enhanced revenue stream" to me... 12:01
pmurias "enhanced revenue stream"? 12:05
yoleaux 03:41Z <samcv> pmurias: it's <:InBlockName> if you see how it is in roast
pmurias DrForr: ahh, you mean it sound like some corporate buzzword 12:06
moritz: there was some "proof of concept" attempt for that
DrForr No, not that. I just meant that it could be added to VStudio and with some "secret sauce" libraries require your editor to be hooked to the 'net and access the server for "special" content. I'm probably being entirely too paranoid there, but just remembering how M$ used to be. 12:07
tyil moritz: I did markdown before getting to adoc, but I like adoc better for writing technical docs 12:08
pmurias DrForr: it's just a protocol so unless we plan to raise funds for the TPF that way there is no reason for our implementation to do that ;) 12:11
DrForr Heh.
I know, I've just seen lock-in occur. That being said I'd love to have a portable way to let people use a remote Perl 6 compiler. 12:12
lizmat perhaps a Cro plugin should be enough ? 12:14
pmurias lizmat: for the language-protocol-server? 12:15
lizmat yeah?
lizmat has only been half listening so may have been off the bat completely
pmurias lizmat: I haven't looked into the technical details closely yet 12:17
DrForr It appears to be JSON-RPC on the inside. 12:18
12:21 geekosaur joined
buggable New CPAN upload: Sparky-0.0.23.tar.gz by MELEZHIK cpan.metacpan.org/authors/id/M/ME/....23.tar.gz 12:23
12:25 scimon joined, natrys joined
scimon Afternoon all. 12:25
Phlogistique good afternoon 12:28
moritz tyil: adoc eq asciidoc?
tyil moritz: yes
buggable New CPAN upload: Game-Sudoku-0.1.0.tar.gz by SCIMON cpan.metacpan.org/authors/id/S/SC/...1.0.tar.gz 12:33
scimon So. It's not finished yet but it does some fun stuff and I had a lot of fun with Sets and Junctions. 12:34
12:35 mcmillhj joined 12:39 mcmillhj left 12:47 cdg joined
scimon Now to write up the Advent calender. 12:47
12:51 cdg left 12:59 kannan joined
kannan good day all, feels like home , back on the irc chat after months. 13:00
lizmat kannan o/
13:03 char_var[buffer] left
scimon Hi 13:06
13:12 mscha left 13:13 wander left, leah2 left 13:14 leah2 joined 13:18 ctilmes left, arnsholt joined 13:24 cdg joined
El_Che If I am using NativeCall and I receive a Pointer (in order to hold data, don't care about the contents during execution), is there a way to print its contents while debugging without going through the effort of writing C structs representations of the data? 13:25
13:26 Cabanossi left 13:27 Cabanossi joined 13:28 cdg left 13:31 jeromelanteri joined 13:38 jeromelanteri left 13:39 scimon left, mcmillhj joined 13:40 virtualsue joined 13:54 cdg joined 13:55 comborico1611 joined 13:57 cdg_ joined 14:00 cdg left 14:05 dakkar left 14:06 dakkar joined 14:09 virtualsue left 14:25 Cabanossi left 14:28 Cabanossi joined 14:31 kannan_ joined 14:35 kannan left, kannan_ is now known as kannan 14:38 Morfent left 14:46 wander joined 14:47 konsolebox left 14:48 cpage_ joined, konsolebox joined 14:50 pmurias left, khw joined 14:51 pmurias joined 14:55 mson joined 15:00 epony left, epony joined 15:01 mcmillhj left, mcmillhj joined 15:04 ctilmes joined 15:07 raiph left 15:13 mcmillhj left 15:21 kannan_ joined, kannan left 15:22 kannan_ is now known as kannan 15:26 troys joined 15:42 zakharyas joined 15:43 domidumont left 15:44 zakharyas left 15:49 zakharyas joined 15:51 zakharyas left 15:54 zakharyas joined 16:02 zakharyas left 16:03 zakharyas joined, parv joined 16:04 parv left 16:06 zakharyas left, domidumont joined 16:10 mcmillhj joined 16:11 zakharyas joined, ExtraCrispy joined 16:12 Cabanossi left 16:13 Cabanossi joined 16:15 zakharyas left 16:20 ExtraCrispy left 16:23 gdonald joined 16:31 zakharyas joined 16:43 lowbro left 16:59 abraxxa left 17:02 zakharyas left, scimon joined 17:03 ChoHag left
Voldenet I might be rehashing the question from long ago, but is there any fast and production-ready http server in p6? 17:05
17:06 zakharyas joined
Voldenet I've tried 'cro' which looks very nice, but isn't very fast yet, maybe I'm not doing stuff correctly 17:06
perlpilot Voldenet: have you seen cro? cro.services
oops, I'm too slow
modules.perl6.org/search/?q=http+server maybe one of those will do you well? 17:07
jnthn I've benchmarked Cro doing 600+ requests per second using ab 17:08
That was doing a "hello world" style thing 17:09
El_Che is there an easy way to print the contents of a NativeCall C-Pointer of C struct without writing a struct for them. Only for debugging 17:10
jnthn found the file where he tracks the best numbers he's seen...seems 660 req/sec. Apparently if there is no matching route it can do 816 req/sec handing back the 404 :P
17:11 Cabanossi left 17:12 Morfent joined 17:13 Cabanossi joined
jnthn (And yes, I've also benchmarked it being beaten by other things by a factor of 2-3, so there's some way to go...) 17:14
moritz fail fast, eh? :-)
jnthn Apparently :)
Voldenet Yeah, 600rq/s isn't too fast indeed - I'm not expecting nginx-like performance, but something around 2k would be decent 17:15
ctilmes Put nginx in front of cro, caching frequent requests
17:15 ChoHag joined
Voldenet but that's kind of cheating ;) 17:16
ctilmes Depends on what you are trying to do. Pretty frequent production set up.
Voldenet also, this nodejs snippet is 10x faster: ix.io/CNW
and it does more or less what cro's example 17:17
ctilmes I'm not saying a faster cro would be nice, but I really like cro, so I'd rather write cro code and front-end cache/run multiple servers, etc. for now 17:18
jnthn About the only optimization effort that's gone on so far with Cro has been improving a handful of hotspots inside of Rakudo and MoarVM that it showed up. Given that, I suspect there's plenty of room for more wins. 17:19
Voldenet I'm kind of tempted to try benchmarking IO::Socket::Async 17:20
17:22 ufobat joined 17:23 scimon left
jnthn Those'd be interesting numbers to see 17:24
Voldenet > 540.43 [#/sec] 17:28
but that's very naive approach: ix.io/CNZ 17:29
17:29 mson left 17:32 ChoHag left 17:34 dakkar left 17:37 Levex joined 17:38 ChoHag joined
Voldenet bam, 3k rq/s 17:39
that's a bit warmer ix.io/CO1 :)
jnthn Yeah, the first one forces all the processing onto a single thread, the second doesn't 17:40
Voldenet Ah, so there's no need for the second schedule-on 17:41
jnthn No
Voldenet so, 4k rqs: ix.io/CO3 17:43
I'm kind of wondering why I can't set initial_threads to 16
ilmari -> $ { ... } since you're not using $l 17:44
Voldenet Well, /I kind of need to use that if I actually want to parse anything/ ;P
ilmari well yes, but not for this trivial benchmark
jnthn Home time & 17:45
Voldenet I guess the socket performance isn't the big issue here 17:48
17:53 ChoHag left 18:00 ChoHag joined
[Coke] is in the middle of apache hell and would love to be able to replace the config here with a few lines of perl 6. :P 18:02
lizmat reminds me of my first Apache config generator, written about 20 years ago? 18:03
18:05 ctilmes left 18:09 kannan left 18:16 ChoHag left 18:18 ChoHag joined 18:29 Ven`` joined 18:31 ChoHag left 18:34 mcmillhj left
mspo [Coke]: apache hell? 18:35
[Coke] mspo: Using apache 2.4 and mod_perl2; know just enough to get the basics working, but rapidly am reaching points where things I try to do don't work, and I don't have enough knowledge to determine why not. 18:37
Is it a bug? am I expecting something to work that shouldn't?
mspo [Coke]: any reason you're using mod_perl?
[Coke] is it related to the mod_perl2?
I'm not having this conversation again. :)
mspo [Coke]: beyond the code cache stuff, I mean; are you using mod_perl for its apache-configuring-ability?
did we already?
18:38 zakharyas left
[Coke] Having already had it beaten out of me in #perl. 18:38
mspo oic
[Coke] no, not you, you're fine. :)
mspo [Coke]: for actual apache stuff mod_lua (the built-in one) is probably the way to go
[Coke]: that is, if you need something bananas that you can't do with regular httpd.conf
[Coke] we already had apache, doing 90% of what we needed. to do the one last thing, mod_perl was an easy fit. did that 10%. now there's "one more thing", which may or may not be a reasonable thing to do try to do. 18:39
mspo [Coke]: apache works by execution phases and you can use mod_(perl|lua) to hook into them
[Coke]: but it isn't always obvious what context you are in
18:44 darutoko left, raiph joined 18:45 setty1 joined 18:48 Ven`` left 18:51 raiph left 18:52 raiph joined 18:53 ChoHag joined 19:00 mcmillhj joined 19:01 raiph left 19:05 mcmillhj left 19:06 mcmillhj joined 19:10 cdg_ left 19:11 cdg joined 19:12 mcmillhj left 19:14 mcmillhj joined 19:15 cdg left
donaldh m: my @list = 0..5; '2'.match(/(\d+)/) and say @list[$0] 19:23
camelia 2
donaldh m: my @matrix[2;2]; '0,0'.match(/(\d+) ',' (\d+)/) and @matrix[$0; $1] = True
camelia Partially dimensioned views of shaped arrays not yet implemented. Sorry.
in block <unit> at <tmp> line 1
donaldh m: my @matrix[2;2]; '0,0'.match(/(\d+) ',' (\d+)/) and @matrix[+$0; +$1] = True 19:24
camelia ( no output )
19:24 pmurias left
donaldh indices for shaped arrays need manually coerced to int 19:24
Normal arrays just DWIM 19:25
okay, that actually golfs down to 19:26
m: my @matrix[2;2]; @matrix['0'; '0'] = True
camelia Partially dimensioned views of shaped arrays not yet implemented. Sorry.
in block <unit> at <tmp> line 1
19:26 Cabanossi left
donaldh m: my @array; @array['2'] = True 19:26
camelia ( no output )
19:28 Cabanossi joined 19:31 troys is now known as troys_, cdg joined 19:35 bwisti left 19:36 cdg left, unicodable6 left, unicodable6 joined, ChanServ sets mode: +v unicodable6, bisectable6 left, committable6 left, bisectable6 joined, ChanServ sets mode: +v bisectable6, committable6 joined, ChanServ sets mode: +v committable6 19:41 natrys left
lizmat donaldh: please make a GH issue :-) 19:41
19:49 itaipu joined 20:01 cdg joined 20:03 xinming left 20:06 cdg left 20:08 domidumont left 20:21 ctilmes joined 20:27 mcmillhj left 20:32 benchable6 left 20:33 benchable6 joined 20:34 mcmillhj joined 20:47 troys_ is now known as troys 20:54 pochi joined 20:55 Cabanossi left 20:57 geospeck left 20:58 Cabanossi joined 21:05 zakharyas joined, xinming joined 21:28 rindolf left 21:30 TimToady left 21:31 cdg joined 21:32 TimToady joined 21:36 cdg left 21:38 cdg joined 21:39 mson joined 21:40 sena_kun left 21:47 pmurias joined 21:56 Cabanossi left 21:58 Cabanossi joined 22:03 TEttinger joined, mcmillhj left 22:08 lostinfog joined 22:10 m0ltar left 22:16 zakharyas left
Geth ecosystem: ace48d620c | (Tony O'Dell)++ (committed using GitHub Web editor) | META.list
InterProcess Event::Emitter functionality

Now you can send data [multiplexed] between processes with `named` events with ease		 22:16
22:20 m0ltar joined 22:41 cpage_ left 22:48 AlexDaniel joined
AlexDaniel u: bicycle 22:49
yoleaux 16:38Z <tbrowder> AlexDaniel: what do you think of my last comment on last Rakudo PR-
unicodable6 AlexDaniel, U+1F6B2 BICYCLE [So] (🚲)
AlexDaniel, U+1F6B3 NO BICYCLES [So] (🚳)
yoleaux 16:45Z <tbrowder> AlexDaniel: PR? I am getting close to a good solution for the table config key values but I think the last PR is a good interim solution. Note also the previous PR must also be merged to get the new tests to work. So I’m respectfully requesting merge of PR #1287 and PR #1291.
22:52 comborico joined
comborico Moritz, you around? 22:52
22:55 Cabanossi left 22:56 lostinfog left 22:58 Cabanossi joined 23:06 cpage_ joined 23:14 rindolf joined
DrForr .tell moritz Scheduled Day 7 for Thursday 11:59 AM so we've got time for last minute critiquing. 23:16
yoleaux DrForr: I'll pass your message to moritz.
23:19 troys is now known as troys_ 23:23 mcmillhj joined, comborico left 23:25 Cabanossi left
DrForr Gotta get to bed. Also sent a note to the OP, I'll be around at prob. 9:30am CET. 23:26
23:27 mcmillhj left 23:28 Cabanossi joined 23:29 troys_ is now known as troys 23:30 dayangkun joined 23:32 mcmillhj joined 23:33 pmurias left 23:39 mcmillhj left 23:48 mson left
donaldh lizmat: GH issue added 23:51
23:55 rindolf left 23:56 donaldh left 23:57 Cabanossi left 23:58 Cabanossi joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!