»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 Zoffix on 25 May 2018.
lookatme Is it fine if I put the deallocate operator in a END block ? 04:07
lookatme such as ` sub f() {　state $var; END { $var.close() } }; f()` 04:08
Xliff .seen sergot 04:40
yoleaux I saw sergot 14 Mar 2018 21:46Z in #perl6: <sergot> ufobat_: ping
Xliff Does anyone know of a decent replacement for HTTP::UserAgent? 04:41
Or do I have to Inline::Perl5?
moritz .tell Xliff you might be interested in the cro HTTP client: cro.services/docs/intro/http-client 05:31
yoleaux moritz: I'll pass your message to Xliff.
jmerelo Hi 06:47
yoleaux 03:59Z <AlexDaniel> jmerelo: I looked at shareable issue and it seems like it can only serve one file, then it just hangs or something like that (some time later the watchdog will restart it and it'll be able to serve another file, but it takes time for the watchdog to kick in). Looks very similar to R#1595
03:59Z <AlexDaniel> jmerelo: I'll have to figure something out for this before the squashathon :S
04:03Z <AlexDaniel> jmerelo: the workaround mentioned in github.com/perl6/whateverable/issu...-371292860 is in place, but it doesn't help for some reason. Moreover, bots have some restrictions to which paths they can write, and *that* is what affects it, I think.
04:08Z <AlexDaniel> jmerelo: oh actually, I think I ““fixed”” it. Now every service will use an existing precomp file instead of generating a new one on the fly… that fixes shareable 🤦 No idea how that can possibly help, but it works now
jmerelo .tell AlexDaniel thanks!
yoleaux jmerelo: I'll pass your message to AlexDaniel.
jmerelo Hum. Getch has stopped working? 06:48
*Geth
Yep, it has 07:09
andrzejku_ do you think if I know Perl6 but not Perl can I write in CV that I know Perl 07:20
to not recreuiter told me that I am learning useless language?
or it is bad idea 07:21
geekosaur on the one hand, perl 6 is not perl 5
on the other you're talking about recruiters
the sort who insist on sending me python jobs, ruby jobs, various other things not even on my resume/cv 07:22
not to mention insisting on treating "openafs" as a financial reference 07:23
basically, recruiters are bald-faced liars out of the box, and you can assume you are being manipulated not for your own good 07:24
andrzejku_ geekosaur: ya thats right 07:25
geekosaur: it even happend in my company, they told one dev that he is hired then he terminated his contract and comes to assign the new one but they changed conditions which they promised him 07:27
Tison \o 08:46
lookatme o/ 09:06
Tison,
Tison We get $=pod contains Pods only by load source code, don't we? 09:22
Is there a safe way to get Pods of source code without load(means, execute) it? 09:23
jmerelo Tison: right 09:24
jmerelo Tison: second question: what do you mean? Extracting just the pods from a file? Also: pods might include source code too in :preamble or :config sections. 09:25
Tison maybe i can say acting like `javadoc` 09:27
github.com/perl6/Pod-To-HTML/blob/...pm#L25-L28
IIRC there was a discuss about retrieving pods by `EVAL` 09:28
Tison Modules and Pod 6 files often have no bare code though 09:29
jmerelo Tison: perl6 --doc is a bit like Javadoc. 09:30
Tison: you can use it to generate markdown or HTML if the modules are installed. 09:31
Tison at then timo show me a example like `=pod ... \n INIT say 42` which `perl6 --doc` execute code inside INIT parse 09:43
Tison I'm ok with Pod 6, here is some little wart IMO 09:44
AlexDaniel . 10:43
yoleaux 06:47Z <jmerelo> AlexDaniel: thanks!
AlexDaniel squashable6: next
squashable6 AlexDaniel, ⚠🍕 Next SQUASHathon in 1 day and ≈23 hours (2018-06-02 UTC-12⌁UTC+14). See github.com/rakudo/rakudo/wiki/Mont...Squash-Day
throwaway321 hello * 10:49
moritz \o throwaway321
throwaway321 hi moritz 10:50
any idea how nativecall works with varagrs? 10:51
throwaway321 I am having trouble with a function similar to this: int add(int number, ...) 10:52
how do I take care of those varargs in my signature? 10:54
moritz throwaway321: I'm not really the export on nativecall :(
throwaway321: one thing you can do is add multis for different arities
throwaway321 I tried seems multis won't work with is native 10:55
moritz multi sub a(int) is native(...); multi sub a(int, int) is native(...);
:(
maybe ask on stackoverflow?
I've seen some good answers to nativecall questions there
unless timotimo or so shows up earlier :-)
jmerelo throwaway321: there's this question in StackOverflow along those lines stackoverflow.com/questions/495860...nativecall 10:56
Unfortunately, no answer so far... So we might have to ask timotimo to answer it there too.
throwaway321 oh thanks 10:57
timotimo oh no 10:58
throwaway321 timo already answered and said there is no support for varargs yet
throwaway321 hello timotimo 10:59
timotimo you can have multiple subs refer to the same C symbol with "is symbol"
that can help with the multi problem, though i didn't know until now that multis aren't supported by NativeCall
throwaway321 this is what I get when I define a multi: "Cannot invoke object with invocation handler in this context"
lizmat how would NativeCall's support for multi's help ? 11:00
jmerelo timotimo: you might want to expand your comment to an answer there. Comments are easily missed there...
timotimo it would mean you don't have to do your own multi dispatch manually
like, you could have something like
sub fprint_two_ints(Str, int32, int32) is native { }; sub fprint_two_doubles(Str, num64, num64) is native {}; ... 11:01
and then multi sub fprint(Str, int32, int32) { fprint_two_ints(...) }; multi sub fprint(Str, num64, num64) { fprint_two_doubles(...) }; ..
that's obviously annoying
lizmat yeah, I guess I need to get into NativeCall innards to fix that 11:02
but I do get your point now 11:03
timotimo i think we probably have a problem with the way we compile nativecall subs today
timotimo if i understand correctly, no nativecall code has to be changed in moarvm 11:03
throwaway321 would support for varargs be doable?
timotimo because we pass in objects already, they know their type, so we don't even have to have something like format strings to tell what types are to come in the argument list 11:04
timotimo i might be wrong about that 11:04
throwaway321 thanks everyone 11:05
timotimo of course you can build something with EVAL that generates the right subs on the fly as they are needed for the first time 11:21
moritz do you even need EVAL? I mean, all you need for nativecall is a code object and a signature, no? 12:33
and then call trait_mod:<is> with the right arguments
[ptc]_ timotimo: just wondering: is there a community-run MoarVM Coverity-Scan account?
timotimo yeah, yours :P 12:34
[ptc]_ hrm, that's what I thought :-/
oh well
I still wonder if it would be better linking the actual MoarVM repo to Coverity rather than my fork 12:35
especially since I don't have so much time to spend on P6 stuff atm :-(
timotimo i have no idea how to do that :D 12:38
[ptc]_ are you an admin of the MoarVM repo? I think that'd be a necessary requirement 12:40
other than that, I can find out how one would do things from the main repo, so that it's in central location
somehow that feels a cleaner solution... 12:41
timotimo i'm not an admin there, no 12:42
[ptc]_ ok. I'll find out what's likely to be required and will let you know 12:43
timotimo thanks 12:44
[ptc]_ no worries :-)
[ptc]_ re-enters lurk-mode...
lizmat [ptc]_ : did you do anything to Travis config? my module testing now breaks with: The program 'perl6' is currently not installed. To run 'perl6' please ask your administrator to install the package 'rakudo' 13:10
[ptc]_ lizmat: haven't done anything to Travis in ages. That's something I definitely need to look into again 13:13
lizmat hmmm... weird
lizmat tries again
[ptc]_ come to think of it, I think brian d foy has implemented an improved Travis config for himself, since the one I developed ages ago hasn't really stood the test of time... 13:14
[ptc]_ really needs to dive back into this community 13:15
El_Che I build ubuntu 14.04 packages for all releases and they are in a repo. Maybe it could be a faster and better alternative 13:17
[ptc]_ El_Che: I agree totally, that would be much faster, since at present everything is built from source 13:18
El_Che I have the repo up since last release, but if needed I can build packages retroactively 13:20
although I think versions from there onwards is ok 13:21
I don't see much the point of testing against 2015.12
[ptc]_ yeah, 2015.12 is much too old. Actually, it'd probably be a good idea to drop the oldest from the list of possible releases at the time of each new release 13:22
El_Che yeah, the starting point is arbitrary anyway, and everything from last release is automatically build and uploaded to the repo 13:23
built 13:24
[ptc]_ The script one needs to edit is in the travis-build repo: github.com/travis-ci/travis-build/...t/perl6.rb 13:26
at present, only "blead" is built by default. If one specifies another version as part of the 'perl6:' spec in the yaml, then it gets built as well 13:27
quick hack idea: add a switch in the setup() method which then installs one of your Ubuntu packages. Otherwise everything gets built from scratch 13:31
El_Che I want something like in go, support for a list like this: 13:32
perl6:
evalable6
El_Che - 2018.01.x
- 2018.02.x
- 2018.03.x
where x is the latest dot release 13:33
I haven't implemented myself because I am not sure it's something everyone wants
timotimo i wonder if travis users would be upset if we pinged our server with "this perl6 version was just built on travis" so we can see if anyone is actually using super old versions. probably not a cool thing to do; that's surely a thing people want an opt-in for, rather than an opt-out 13:34
El_Che timotimo: well, if the requested version if not found, it could be built on the fly 13:36
e.g. for building from a commit
however, I think it's an anti-pattern 13:37
and if you need that, it's easy enough to do it yourself in your own travis setup
[ptc]_ there is support for a list, one can specify multiple perl6 release versions and a the module is built and tested against each version 13:47
the main issue is that each perl6 version is built from scratch; using the packages would drop module build times drastically 13:48
El_Che yes, that's the idea
[ptc]_ interestingly enough, with the current travis-build code for perl6, one could specify a minimum perl6 version required for packages to be used, and if one wants 2015.12, then that gets built from scratch 13:49
it's sort of the best of both worlds
lizmat my, Travis is finciky: modules pass again now 13:51
El_Che rakudo tests are finicky
if you build them (and test) on the fly you'll have some failures 13:52
[ptc]_ El_Che: do you know Dominique Dumont? He's on the Debian Rakudo team and organises a lot of the stuff for the official Debian packages 13:53
El_Che I know of hime 13:53
I know of him
raschipi Please convince him to package zef... 13:54
El_Che e.g. that debian testing is very up to date
(I only provide debian stable packages)
considering adding testing and unstable, but it's kind of a moving target
raschipi: yeah, the first releases of my packages didn't include zef, but I decided to add it 13:55
rakudo + zef is a better base
raschipi zef should be a CORE module
El_Che zef or something similar 13:56
I love the functionality, not the interface
with perl6 being a moving target, kudos to the dev 13:57
[ptc]_ I believe creating packages for perl6 on Debian is an "interesting" problem. It seems that all modules need to be rebuilt each time rakudo is updated, which makes things hard for the package maintainers 14:11
AlexDaniel bugs.debian.org/cgi-bin/bugreport....=867237#10
.seen robertle
yoleaux I saw robertle 22 May 2018 11:59Z in #perl6: <robertle> looks like a correct british salute from behind to me... ;)
AlexDaniel ↑ that's another Debian hero
timotimo the compunitrepo staging thing is for that purpose, [ptc]_ 14:12
[ptc]_ unfortunately, I'm waaaay behind with perl6 developments atm... :-/ 14:13
raschipi They refuse to allow zef in the archive because they are going to package all the modules. But don't have time for it.
lizmat Debian should perhaps make their own repository storage module :) 14:14
AlexDaniel there's also some semi-related discussion here: github.com/rakudo/rakudo/issues/12...-380552818
[ptc]_ raschipi: they mention that help is welcome :-)
raschipi To package the modules. I want to package zef, which they say they don't want. 14:17
AlexDaniel maybe also this: github.com/ugexe/zef/issues/117 14:18
El_Che not adding zef makes no sense 14:41
being not happy about the sha1 rebuilt is very understandable, though 14:42
cpanminus is packages for perl 5
zo it's not about zef installing modules elsewhere 14:43
[ptc]_ agreed: not adding zef makes no sense. 14:44
although I do wonder what would happen when a user does `zef install My::Module` and then later does `apt install perl6-my-module` 14:46
timotimo i suppose you'd have one auth:<apt> and one auth:<cpan:username> or something 14:47
El_Che [ptc]_: I "solve" that problem by having the bundled zef install ins /opt/rakudo-pkg/share and supply a script for users to install in ~/.perl6 14:47
timotimo: You don't want to mix directories 14:48
[ptc]_ iirc when python is upgraded on Debian, all modules are rebuilt as part of the upgrade process 14:49
maybe that's a good pattern to follow
for the Debian packages, at least. Then the zef stuff can go in ~/.perl6 14:50
raschipi Yes, the package manager owns the directories where it installs things. Other things need to install into their own.
El_Che [ptc]_: a global local site dir is fine 14:54
sometimes ~/.perl6 is not a good match for the use case at hand 14:55
raschipi Many processes don't even have a $HOME. 14:59
AlexDaniel domidumont: irclog.perlgeek.de/perl6/2018-05-30#i_16225092 15:05
buggable New CPAN upload: Memoize-0.0.1.tar.gz by ELIZABETH modules.perl6.org/dist/Memoize:cpan:ELIZABETH 15:12
jmerelo o/ 15:21
jmerelo Geth is still busted 15:43
AlexDaniel undersightable6: status 15:52
undersightable6 AlexDaniel, OK! Working on it…
AlexDaniel, 0 errors, 5 warnings: gist.github.com/648e1659e26c7189db...a34c4afc1f 15:54
AlexDaniel oh lol look at that 15:55
I fixed it xD
raschipi botsnack 16:00
.botsnack
yoleaux :D
Geth doc: 35df3e17b7 | (JJ Merelo)++ | doc/Language/variables.pod6
Eliminates another category, refs #1410

And chekcs that #1358 is well closed.		 16:04
synopsebot Link: doc.perl6.org/language/variables
jmerelo AlexDaniel: thanks! 16:09
rindolf Hi all 16:26
raschipi oi 16:27
rindolf I am beginning to think that pypy really is alicorn magic - it runs my program at 7s vs. cpython's 2m56.689s 16:28
and the perl 6 ver is still running :( 16:29
raschipi TimToady wants p6 to run as fast as C someday. 16:30
El_Che you got to geve it to the pyhton guys, they mean it when they say that speed isn'nt everything 16:30
raschipi rindolf: you sure you didn't enter an infinite loop somehow? 16:31
timotimo raschipi: it's quite possible to get perl6 code to run that slow
rindolf raschipi: let me paste site it
timotimo rindolf: is that the same code you showed recently?
raschipi Oh yes, but it's also quite easy to enter an infinite loop by mistake too. 16:32
buggable New CPAN upload: Env-0.0.2.tar.gz by ELIZABETH modules.perl6.org/dist/Env:cpan:ELIZABETH
rindolf raschipi: paste.debian.net/1027294/ 16:33
timotimo: a different one
Geth doc: d00b6468af | (JJ Merelo)++ | doc/Language/iterating.pod6
Continues with #2069, with iterating constructs
synopsebot Link: doc.perl6.org/language/iterating
rindolf raschipi: this is the py ver - github.com/shlomif/project-euler/b..._287_v1.py 16:34
timotimo how deep does it recurse? 16:36
rindolf timotimo: 24 16:38
timotimo: log 2
timotimo we'd reduce $N to make the work load smaller?
rindolf timotimo: we can 16:38
timotimo so 23 takes half as long as 24?
rindolf timotimo: or a quarter as long 16:40
timotimo ok, with a load of 20 it takes 3:10 on my machine 17:06
Geth doc: 458e6ffe05 | (JJ Merelo)++ | doc/Language/iterating.pod6
Finished iterating data structures.

It's not perfect, but it's there. Much of it was already covered in the control flow document. This closes #2069.		 17:09
synopsebot Link: doc.perl6.org/language/iterating
timotimo the timing isn't all that stable at the moment, my system is far from idel 17:13
idle
mrdside hi 17:14
i have "method get-updates(%params? (:$offset, :$limit, :$timeout))" 17:15
how can i pass %params?
timotimo just pass a hash
my %foo = offset => 1; get-updates(%foo) 17:16
m: say 9.3 / 10.3
camelia 0.902913
timotimo rindolf: want an easy 10% speed improvement? get rid of "return" :)
mrdside doesn't work 17:20
error
wrong argument
timotimo m: class Test { method get-updates(%params? (:$offset, :$limit, :$timeout)) { say $offset; } }; my %options = offset => 1; Test.get-updates(%options) 17:21
camelia 1
timotimo could something else be different somewhere? 17:23
rindolf timotimo: ah 17:28
Geth doc: bbfb2ec17d | (JJ Merelo)++ | 2 files
Deletes nativeInt

This file actually did not have anything, causing more confusion than benefit, as indicated by @jnthn in #1534. The native types are right now described in nativecall, but they will be moved somewhere else, as per #1512.
Also changes redirection to the new page that includes all these types.		 17:30
timotimo rindolf: i know how to make the program a whole lot faster, but give incorrect output. does that count? 17:32
rindolf timotimo: no
timotimo can you verify the right output value for N == 16?
mrdside > my %foo = offset => 1; $bot.get-updates(%foo) 17:33
Cannot convert string to number: base-10 number must begin with valid digits or '.' in '⏏offset=1' (indicated by ⏏)
in block at /home/ymak/.perl6/sources/1A204C32E7D911DE0864A9BF3E06FA10C387892F (TelegramBot) line 32
in block at /home/ymak/.perl6/sources/1A204C32E7D911DE0864A9BF3E06FA10C387892F (TelegramBot) line 28
in method send-request at /home/ymak/.perl6/sources/1A204C32E7D911DE0864A9BF3E06FA10C387892F (TelegramBot) line 27
in method get-updates at /home/ymak/.perl6/sources/1A204C32E7D911DE0864A9BF3E06FA10C387892F (TelegramBot) line 80
in block <unit> at <unknown file> line 1
github.com/GildedHonour/TelegramBot 17:34
timotimo looks like the value was passed correctly, but you're trying to intify a string that doesn't want to be intified
oh, yes 17:35
github.com/GildedHonour/TelegramBo...am.pm6#L32
you can't use + for this purpose, you probably meant to use ~ for string concat, but there's probably something that has to go in between parameters, too 17:36
mrdside ok 17:36
i have working bot with this module - github.com/Robertof/perl-www-telegram-botapi 17:37
i think i should use it instead
mrdside if i have not enough experience in Perl yet? 17:38
timotimo ah, that module doesn't belong to you?
that explains things %) 17:39
mrdside no
)
i just playing
in perl 5 i don't understand how work with ref()
but i think this is not big deal 17:40
there is too much '#todo' and '#todo refactor' in github.com/GildedHonour/TelegramBot 17:41
timotimo i submitted an issue
mrdside thanks a lot!
timotimo no problem
mrdside any advise about v5 or v6 for this purpose? 17:42
timotimo i've never programmed anything in perl 5 :) 17:43
mrdside me too :) just some oneliners 17:44
i think i should try both 17:47
i need advise about perl6 REPL
"You may want to `zef install Readline` or `zef install Linenoise` or use rlwrap for a line editor
"
what's this?
rindolf timotimo: the C versions runs in under 1s
mrdside ok 17:48
AlexDaniel squashable6: status
squashable6 AlexDaniel, ⚠🍕 Next SQUASHathon in 1 day and ≈16 hours (2018-06-02 UTC-12⌁UTC+14). See github.com/rakudo/rakudo/wiki/Mont...Squash-Day
mrdside both?
jmerelo mrdside: no, you can use one or the other 17:50
mrdside: it will help you to use history and other nice features with the perl6 CLI 17:51
Geth doc: 3067d8ce67 | (JJ Merelo)++ | doc/Language/nativetypes.pod6
First attempt at documenting nativetypes

I have moved the table to this page and split it in two, eliminating nonspec types. It will clarify what you can use in native call (those declared with nativesize) and which ones you can't. Refs #1512.		 17:54
synopsebot Link: doc.perl6.org/language/nativetypes
timotimo rindolf: i think your code exposed an interesting bug in the perl6 optimizer 17:57
rindolf timotimo: ah 17:58
timotimo: nice
timotimo m: sub foo($x, $y) { True }; sub bar($x, $y, $w) { { my int $x2 = 1; my int $y2 = 2; if foo($x, $y) == foo($x, $y2) == foo($x2, $y2) == foo($x2, $y) { 2; } else { 1 } } }; say bar(1, 2, 25); 18:08
camelia Cannot resolve caller Numeric(Sub+{is-pure}+{Precedence}: ); none of these signatures match:
(Mu:U \v: *%_)
in sub bar at <tmp> line 1
in block <unit> at <tmp> line 1
timotimo it's trying to numify the sub instead of the result of the sub when it does the == operation 18:09
BBIAB
raschipi jmerelo: did you see zoffix has a branch that documents the native numeric types? 18:11
timotimo mrdside: honestly, i'd almost recommend using the jupyter notebook instead of the perl6 repl to figure stuff out :)
mybinder.org/v2/gh/bduggan/p6-jupy...orld.ipynb 18:12
jmerelo raschipi: it's the numeric types, in general 18:20
raschipi just making sure you're aware, no need to redo those.
mrdside timotimo: maybe, after Emacs )))
I can use C-p C-n C-b C-f to navigate 18:21
jmerelo raschipi: there's some overlap, probably. We can rewrite stuff when it's eventually merged if needed.
mrdside i think Perl 6 Grammar will help a lot with messages processing. 18:22
btw how can i define KIO8-R text encoding for read text files? 18:24
jmerelo mrdside: not sure you can... github.com/rakudo/rakudo/search?q=...d_q=kio8-r this returns nothing. 18:28
KOI8-R returns nothing either 18:29
Geth doc/rasch-readme: e7ab3cd32e | (Rafael Schipiura)++ (committed using GitHub Web editor) | 2 files
README in Portuguese

Tradução do README para português.		 18:30
doc: rafaelschipiura++ created pull request #2081:
README in Portuguese		 18:31
mrdside i asked this before and got advise to try write code for this purpose. but i don't now how 18:34
lucasb raschipi: ^^ does your commit accidentaly deletes README.es.md? 18:35
raschipi yes, I'm fixing it 18:36
Geth doc/rasch-readme: f88f0391f5 | (Rafael Schipiura)++ (committed using GitHub Web editor) | README.es.md
commit accidentaly deletes README.es.md
jmerelo raschipi: I wouldn't forgive you for that :-) 18:37
mrdside: Far as I can tell, you can't do it out of the box. You can probably create some easy translation table from KOI8 to Unicode, but you'll have to do it. 18:38
raschipi I know. That's why I sent a pull request instead of committing directly.
jmerelo mrdside: This table does the conversion, for instance koi8.pp.ru/utf-8.koi8-r.html 18:39
mrdside nice, and how use it with "my $fh = open('text.txt', enc => 'kio8-r');"? 18:41
mrdside jmerelo: . 18:43
jmerelo mrdside: Not really, I'm afraid... You'll have to turn that table into a Hash, read the KOI8 into a buffer, and do a one-by-one character conversion from one to the other. 18:52
mrdside: it would make a nice new module :-)
Geth doc: ronaldxs++ created pull request #2082:
Update Language/regexes.pod6. Fix confusing sentence about zero-width assertions		 18:55
doc: e3d825ac85 | (Ronald Schmidt)++ | doc/Language/regexes.pod6
Update Language/regexes.pod6. Fix confusing sentence about zero-width

assertions included with character classes and remove likely pasto right after.
synopsebot Link: doc.perl6.org/language/regexes
Geth doc: 0bfe6e6dba | (Juan Julián Merelo Guervós)++ (committed using GitHub Web editor) | doc/Language/regexes.pod6
Merge pull request #2082 from ronaldxs/fix-regex-pre-defined-character-classes

Update Language/regexes.pod6. Fix confusing sentence about zero-width assertions
mrdside jmerelo: ok. i'll try 18:57
jmerelo mrdside: good luck! 18:58
Geth doc: 0e6e6d2faf | (JJ Merelo)++ | 2 files
Reflows and link nativetype → nativecall

And the other way round. Refs #1512		 19:02
doc: 6268c1fafb | (JJ Merelo)++ | doc/Language/nativetypes.pod6
Adds explanations on native types refs #1512
synopsebot Link: doc.perl6.org/language/nativetypes
doc: 7ffae42f08 | (JJ Merelo)++ | doc/Language/nativetypes.pod6
Adds documentation for nativesize

Only in the way of explanation of the difference between native types and native types with native size; added warning that it's not spec. But this would close #2051.
I am going to close #1512 too. It does document all native types, and clarifies the difference between them. This would be the last document in #114, which was already closed, so I can't close it, but this would basically be it. As the rest of the documents I have created, it's not perfect, but it's there. It can be incrementally improved from now on.
Geth doc/rasch-readme: 51a1a7a0f8 | (Rafael Schipiura)++ (committed using GitHub Web editor) | README.pt.md
Update README.pt.md
 19:04
timotimo rindolf: did you send the right result value for N = 16? 19:07
Geth doc/rasch-readme: d245445f42 | (Rafael Schipiura)++ (committed using GitHub Web editor) | README.pt.md
Update README.pt.md
rindolf timotimo: no, wait
timotimo: 1221292 19:08
timotimo OK, that's what i normally get 19:09
rindolf timotimo: ah
timotimo cool, i got it from about 11:30 to 1.8
rindolf timotimo: 1.8s? 19:10
rindolf timotimo: 1.8second? 19:10
timotimo yes
er 19:11
those numbers are not right
11.83s vs 1.8s
not 11 minutes
Geth doc/raschi-readme2: 21931518aa | (Rafael Schipiura)++ (committed using GitHub Web editor) | README.pt.md
Create README.pt.md
 19:12
timotimo 1.33s 19:13
Geth doc: rafaelschipiura++ created pull request #2083:
Create README.pt.md closes #1984		 19:14
timotimo 0.8s 19:15
Geth doc: 21931518aa | (Rafael Schipiura)++ (committed using GitHub Web editor) | README.pt.md
Create README.pt.md
 19:16
doc: e38a6194df | (Rafael Schipiura)++ (committed using GitHub Web editor) | README.pt.md
Merge pull request #2083 from perl6/raschi-readme2 closes #1984

Create README.pt.md closes #1984
jmerelo And thanks to this last closed issue by raschipi we're down to 180 github.com/perl6/doc/issues
El_Che \o/ 19:18
raschipi 🎉🎉🎉
jmerelo jmerelo: tomorrow is the last official day of the grant, let's see if we can bring it down a bit further. 19:19
Now I'm AFK. See you tomorrow!
timotimo oh, dangit, i missed when it went from right result to wrong result 19:20
timotimo rindolf: is it wrong to assume all involved numbers fit into 64bit? 19:32
rindolf timotimo: no
timotimo turning the optimizer off makes things work again. cool. not.
chained comparisons might be broken at the moment 19:35
bisectable: sub foo($x, $y) { True }; sub bar($x, $y, $w) { { my int $x2 = 1; my int $y2 = 2; if foo($x, $y) == foo($x, $y2) == foo($x2, $y2) == foo($x2, $y) { 2; } else { 1 } } }; say bar(1, 2, 25); 19:39
bisectable6 timotimo, Bisecting by exit code (old=2015.12 new=9472fdf). Old exit code: 0
timotimo, bisect log: gist.github.com/b17966f8408b98e631...9a1ba06f28
timotimo, (2018-01-08) github.com/rakudo/rakudo/commit/14...09dcea845a
timotimo bisectable6: gist.github.com/timo/11b3dc224bb75...8c3e3a0fc2 19:50
bisectable6 timotimo, It looks like a URL, but mime type is ‘text/html; charset=utf-8’ while I was expecting something with ‘text/plain’ or ‘perl’ in it. I can only understand raw links, sorry.
timotimo bisectable6: gist.githubusercontent.com/timo/11...uler287.p6
bisectable6 timotimo, Successfully fetched the code from the provided URL.
timotimo, Bisecting by exit code (old=2015.12 new=9472fdf). Old exit code: 0
timotimo, bisect log: gist.github.com/49664966566ab8996f...f89cece80c 19:51
timotimo, (2018-01-08) github.com/rakudo/rakudo/commit/14...09dcea845a
timotimo rindolf: anyway, i got it down to 0.78s 20:05
rindolf timotimo: the full problem? 20:06
timotimo only for $N == 16
rindolf timotimo: ah 20:12
timotimo rindolf: what's the right result for $N = 24? 20:30
is it 313135496? 20:33
rindolf timotimo: 313135496
timotimo that's what i get after 1:56m of run time
rindolf timotimo: yes
timotimo the first lines of main in the python version don't have an equivalent in the perl6 version? 20:35
anyway, the python version takes 2:10m, which is just a little slower than the perl6 version ... but the perl6 version is ugly to read because i optimized it partially by putting nqp code in 20:38
El_Che hallelluyah to the day nqp can be removed 20:39
lizmat I would say: hallelujah to the day when our code becomes the most optimized nqp automagically :-) 20:40
El_Che $*AUTOMAGICALLY_MAKE_EVERYTHING_FAST = 1; 20:41
lizmat m: $*AUTOMAGICALLY_MAKE_EVERYTHING_FAST = 1; 20:42
camelia Dynamic variable $*AUTOMAGICALLY_MAKE_EVERYTHING_FAST not found
in block <unit> at <tmp> line 1
lizmat m: my $*AUTOMAGICALLY_MAKE_EVERYTHING_FAST = 1; # voila
camelia ( no output )
mrdside jmerelo: i found this - docs.perl6.org/type/Encoding 20:43
but i see only 'nqp:: ... ' calls in *.pm6 on github.com/rakudo/rakudo 20:45
El_Che .tell jmerelo < mrdside> jmerelo: i found this - docs.perl6.org/type/Encoding
yoleaux El_Che: I'll pass your message to jmerelo.
Spot_ Cheers from Amsterdam :) ... lizmat & others!
mrdside .tell El_Che thanks
yoleaux mrdside: I'll pass your message to El_Che.
El_Che mrdside: he's not here, use the .tell command to keep the message until he returns
yoleaux 20:45Z <mrdside> El_Che: thanks
El_Che lol
lizmat Spot_ o/ 20:46
Spot_ Good quality angus beef and Heineken makes the day :) 20:47
mrdside good Banana, maybe $)
El_Che Spot_: you're new to the Low Countries? 20:48
I haven't met locals that like Heineken
mrdside Where i can read about 'Encoding::Decoder'? 20:49
El_Che Spot_: the opinion of Heineken in Belgium varies from water to horse piss :)
mrdside what i should put in to "method decoder(Encoding:D: *%options --> Encoding::Decoder) { ... }"?
what is ':$replacement' here - "method decoder(:$replacement, :$translate-nl, :$strict --> Encoding::Decoder)"? 20:51
just byte array?
tbrowder_ El_Che: what do locals think about Grolsch? 20:54
El_Che: ^^
lizmat tbrowder_ Grolsch > Heineken but that's not saying much 20:55
tbrowder_ so what is the local’s favorite beer?
El_Che tbrowder_: not well known in the Southern Netherlands 20:56
tbrowder_: we have about 800 of them, so difficult to say
tbrowder_: I like this one www.achouffe.be/en 20:57
tbrowder_: it use to be a small obscure beer and got popular overnight
tbrowder_: snobish enough to say I liked it before i was popular :)
tbrowder_ well, I lived near Brunssum while stationed with NATO and most of us US folks were introduced to Grolsch by the Brits. It was the cheapest beer around and, in my neighborhood, we could get home delivery in the refillable bottles—not such a bad deal, but maybe not to everyone’s taste. 21:01
El_Che tbrowder_: when you're young everything that's cheap and not too bad is great :) 21:02
lizmat tbrowder_: I live about 20km away from Brunssum :-) 21:04
tbrowder_ roger! i could never find a belgian beer i liked (but i didn’t try to many), the ones i tried were too sweet. but my favorite beer, which i haven’t had for many many years, is an english bitter. 21:05
lizmat: what town?
lizmat Echt
El_Che tbrowder_: there are plenty of non-sweet beers 21:06
tbrowder_: try Orval
lizmat tbrowder_: just on the other side of Selfkant
El_Che lizmat: too bad this conversation isn't in dutch
Echt == Really in English
lizmat El_Che: really ? :-)
El_Che :)
tbrowder_ i’m sure, but i kept going back to Grolsch—just too set in my ways. 21:07
El_Che sure, don't change a winning team
I live in the village where Duvel is brewed
when the wind is just right we get drunk for free! 21:08
:)
tbrowder_ I’m going to have to look at the maps of the Dutch “alpine” area...
El_Che and I work on the city where Stella is brewed
and on my train ride passed the Haacht brewery 21:09
:)
tbrowder_ i have tried stella artois, very drinkable!
El_Che we may have an alcohol problem :)
as a region
moritz as a culture 21:14
tbrowder_ lizmat: could i have passed Echt on the train from Brussum to Eindhoven? 21:15
El_Che we don't even have the excuse that it was safer than water like in the Middle Ages :) 21:16
lizmat tbrowder_: yes... although you would probably be going ~ 100km / hour :-) 21:17
tbrowder_ no stop there, huh? i remember several stops enroute, but not what towns 21:19
Spot_ El_Che: Grolsch is far better 21:45

Powered by Cro and the Raku® Programming Language.

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

Thank you!