»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.
00:02 rindolf left 00:03 mcmillhj left 00:05 cdg left 00:15 mcmillhj joined 00:18 wamba left 00:19 stmuk_ joined 00:21 stmuk left 00:22 mcmillhj left 00:37 troys is now known as troys_
lookatme o/ 00:37
Geth doc: 040ade60d2 | (Jeremy Studer)++ | xt/examples-compilation.t
Revert "Fix false positive on dd test due to yyyy-mm-dd"

This reverts commit a0ee6719c1b253b809c63c00e93f4c9f1c99cbb1.
00:41
doc: b73921bba5 | (Jeremy Studer)++ | 2 files
Revert dd test and okay yyyy-mm-dd examples

It was suggested that the yyyy-mm-dd exceptions be okayed and the modifications testing for it be reverted. In this case, it is better to manually ok the handful of exceptions than work the exceptions into the test, making it more cluttered and obscuring the intent.
00:44 reu_ left 00:45 aborazmeh joined, aborazmeh left, aborazmeh joined 00:52 adu joined 00:53 reu joined 00:54 Cabanossi left 00:55 Cabanossi joined 00:59 mcmillhj joined 01:00 lizmat joined, markong left, sena_kun left 01:04 lizmat left, mcmillhj left 01:12 mephinet joined 01:13 adu left 01:15 mcmillhj joined 01:20 mcmillhj left 01:22 adu joined 01:26 mcmillhj_ joined
AndChat|688961 Quiet night. 01:27
01:27 AndChat|688961 left, comborico1611 joined
comborico1611 Is it cold where you are? 01:28
01:30 mcmillhj_ left 01:31 cdg joined 01:36 cdg left, mcmillhj_ joined 01:38 llfourn_ left 01:39 jaush left
lookatme yeah 01:40
01:40 mcmillhj_ left 01:41 poohman joined 01:42 llfourn joined, pilne joined
comborico1611 How can it be cold in Vietnam? 01:42
poohman hello all - I am having a problem when trying to make HTTP get request for a particular website - it keeps giving a malformed UTF-8 error 01:43
timotimo poohman: then you may want to get the data as a binary blob instead of a string? entirely depends on what you're hoping to achieve 01:44
poohman LWP::Simple has an option to retieve it as binary
?
Can you please point me to some documentation 01:45
I did search becuase I had the same problem when reading PDF files
or any http client for that matter
timotimo you can LWP::Simple.new(:force_no_encode) i guess? 01:46
poohman ok let me give it a short - thanks
timotimo or if you really do want a string and mojibake is not a problem for you, you can :force_encoding<latin-1> instead
don't forget to grab the result of .new and call .get on that rather than on LWP::Simple itself 01:47
01:47 mcmillhj_ joined
poohman ill do that 01:47
01:52 mcmillhj_ left
poohman it defaults to utf-8 01:53
just created an LWP::Simple object with both the options 01:54
LWP::Simple.new(default_encoding => "utf-8")
getting this when I print it
timotimo that's the default default 01:56
poohman paste.gnome.org/p8rvaqxz4 01:57
get the same for :force_no_encode as well
timotimo gotta go, seeya! 01:58
poohman ok got to get ready for work - will check again in the evening - thanks for the help timotimo
02:00 adu left 02:01 mson left 02:02 kaare__ left, kaare_ joined 02:03 kaare_ left 02:04 kaare_ joined 02:06 mcmillhj joined 02:08 Cabanossi left 02:11 Cabanossi joined, mcmillhj left
buggable New CPAN upload: SemVer-0.1.1.tar.gz by TYIL cpan.metacpan.org/authors/id/T/TY/...1.1.tar.gz 02:13
02:14 stmuk joined 02:16 stmuk_ left 02:22 mcmillhj joined 02:27 mcmillhj left 02:28 itaipu joined 02:29 comborico1611 left 02:40 mcmillhj joined 02:44 mcmillhj left 02:46 ilbot3 left 02:55 ilbot3 joined, ChanServ sets mode: +v ilbot3 03:03 noganex_ joined 03:05 kitsunenokenja left 03:06 noganex left 03:09 Cabanossi left 03:10 mcmillhj joined 03:11 Cabanossi joined, ufobat___ joined 03:12 kaare_ left, kaare_ joined, jeek left 03:13 cdg joined 03:14 ufobat_ left 03:15 kaare_ left, mcmillhj left, kaare_ joined 03:17 kaare_ left 03:18 kaare_ joined, mson joined 03:22 aborazmeh left 03:24 mcmillhj joined 03:28 mcmillhj left 03:31 ryn1x_ joined
ugexe use HTTP::UserAgent; my $client = HTTP::UserAgent.new; say $client.get("httpbin.org", :bin).content 03:34
use Cro::HTTP::Client; my $client = Cro::HTTP::Client.new; say await $client.get("httpbin.org").result.body-blob
use Net::HTTP::GET; say Net::HTTP::GET("httpbin.org").body
03:34 mcmillhj joined
ugexe poohman: ^ there is how to get the raw content with 3 different http libs 03:34
japhb ugexe: Do you need to await in the second one? I thought .result already did that implicitly ...? 03:38
ugexe body-blob returns a promise and i didnt want to do .result again 03:39
japhb Actually, I'm assuming that $client.get() is returning a Promise ...
Ah, I see.
03:39 mcmillhj left
japhb So get() returns a Promise, and then .body-blob() does so again. Hmmm. 03:40
ugexe body blob might not be fully receieved with response promise 03:41
japhb I suppose so.
ugexe well, i guess the .result it should be
you can get just the header and start processing / making decisions 03:42
japhb I dunno, just feels a little redundant. But maybe the answer is just to have a convenience wrapper for when you want to pretend it's just a blocking call.
03:43 itaipu left 03:45 kaare_ left, kaare_ joined 03:55 mcmillhj joined 04:00 mcmillhj left 04:03 eliasr left, Herby_ joined
Herby_ o/ 04:03
04:10 mcmillhj joined 04:15 mcmillhj left 04:19 R0b0t1` left 04:21 R0b0t1` joined, ryn1x_ left 04:23 Herby_ left, wamba joined, Cabanossi left 04:25 Cabanossi joined 04:28 wander joined
wander m: say my %h is BagHash = BagHash.new([3,4,5]) 04:28
camelia BagHash(BagHash(3, 4, 5))
wander m: say my %h is BagHash = [3,4,5] 04:29
camelia BagHash(3, 4, 5)
04:29 mcmillhj_ joined
wander is it what we want? especially for the former case 04:29
m: say my %h = BagHash.new([3,4,5]) 04:30
camelia {3 => 1, 4 => 1, 5 => 1}
04:34 mcmillhj_ left 04:40 wander left 04:43 mcmillhj_ joined
lookatme m: say my %h := BagHash.new([3,4,5]) 04:44
camelia BagHash(3, 4, 5)
04:48 mcmillhj_ left 04:53 mcmillhj_ joined 04:54 Cabanossi left 04:55 Cabanossi joined 04:58 mcmillhj_ left 04:59 kaare_ left, ngaywood joined 05:00 kaare_ joined 05:04 mcmillhj_ joined 05:06 ryn1x_ joined 05:09 mcmillhj_ left 05:12 ryn1x_ left 05:18 mcmillhj joined 05:23 mcmillhj left 05:30 kaare_ left 05:31 kaare_ joined 05:32 mcmillhj joined 05:36 wamba left, ngaywood left 05:37 mcmillhj left 05:42 kaare_ left, kaare_ joined
perlawhirl bisectable6: say [3 ^... 1] 05:43
bisectable6 perlawhirl, On both starting points (old=2015.12 new=c967cb6) the exit code is 1 and the output is identical as well
perlawhirl, Output on both points: «04===SORRY!04=== Error while compiling /tmp/zUNDjlGgUn␤Unsupported use of . to concatenate strings; in Perl 6 please use ~␤at /tmp/zUNDjlGgUn:1␤------> 03say [3 ^...08⏏04 1]»
05:53 piojo joined, cdg left 06:02 mcmillhj joined
lookatme Seems like we not have `^...` operator 06:03
ugexe ^ and ... 06:06
06:06 mcmillhj left
piojo I want to add "Using () and [] to create a list or array" to the traps document-- because (x) won't create a one-element list, and [x] can create an array of any number of elements, depending on the value of x 06:14
Is that something the doc team would agree with?
06:15 mcmillhj joined 06:19 mcmillhj left 06:21 s0me0n3-unkn0wn joined 06:24 s0me0ne-unkn0wn left 06:31 mcmillhj joined 06:32 wamba joined 06:37 mcmillhj left 06:42 darutoko joined 06:47 ryn1x_ joined 06:48 mcmillhj joined 06:52 mcmillhj left 06:53 Cabanossi left
Geth perl6-most-wanted: f53829e0de | (David Warring)++ (committed using GitHub Web editor) | most-wanted/modules.md
Font::FreeType is released
06:54
06:55 Cabanossi joined
lookatme ls 06:58
06:58 kaare_ left, kaare_ joined 07:01 entonian joined 07:02 domidumont joined, entonian left 07:03 mcmillhj joined 07:06 domidumont left 07:07 domidumont joined 07:08 mcmillhj left 07:13 ryn1x_ left 07:15 konsolebox left 07:16 troys_ is now known as troys 07:17 konsolebox joined, geospeck joined 07:20 mcmillhj joined 07:24 mcmillhj left 07:27 khw left 07:30 konsolebox left 07:32 konsolebox joined 07:37 lowbro joined, lowbro left, lowbro joined 07:38 llfourn left 07:39 konsolebox left 07:40 konsolebox joined 07:43 abraxxa joined 07:44 kaare_ left, kaare_ joined 07:46 wamba left 07:52 mcmillhj joined 07:53 Cabanossi left 07:54 wamba joined, cdg joined 07:55 Cabanossi joined 07:56 mcmillhj left 07:58 cdg left
tyil can anyone check if SemVer:ver<0.1.1> installs for them? `zef install "SemVer:ver<0.1.1>"` fails for me, but the local installation with `zef install .` worked 08:05
08:05 mcmillhj joined
tyil and zef's error reporting as to why it fails is very much LTA 08:05
08:09 mcmillhj left
lookatme Is zef support module version installation? 08:14
I don't think so
m: sub postfix:<%>(Cool $n) { $n / 100; }; say 45% * 2; 08:16
camelia 0.9
08:18 mcmillhj joined 08:21 kaare_ left, kaare_ joined 08:22 mcmillhj left 08:24 Cabanossi left 08:25 Cabanossi joined, troys is now known as troys_ 08:32 dakkar joined, mcmillhj joined
buggable New CPAN upload: SemVer-0.1.3.tar.gz by TYIL cpan.metacpan.org/authors/id/T/TY/...1.3.tar.gz 08:33
08:37 mcmillhj left 08:38 parv joined 08:42 rindolf joined 08:45 itaipu joined 08:49 mcmillhj joined, ryn1x_ joined 08:54 mcmillhj left, ryn1x_ left 08:56 wander joined 08:58 scimon joined
wander [Coke]: what is 'xt/' stands for? I know 't/' folder holds test files. what's the different between them? 08:59
09:02 wamba left 09:03 mcmillhj joined, aborazmeh joined, aborazmeh left, aborazmeh joined 09:06 wamba joined, lookatme left 09:07 domidumont left, domidumont joined 09:08 mcmillhj left 09:09 sena_kun joined 09:11 gregf_ left
parv extra|experimental tests 09:16
... would be my guess.
09:18 steeznson joined 09:19 mcmillhj joined 09:22 itaipu left 09:23 mcmillhj left, Cabanossi left 09:25 Cabanossi joined
AlexDaniel wander: it's for tests that should ideally pass, but we don't want to bother random contributors with them 09:25
09:29 jeek joined
buggable New CPAN upload: App-Cpan6-0.17.3.tar.gz by TYIL cpan.metacpan.org/authors/id/T/TY/...7.3.tar.gz 09:33
09:36 rindolf left 09:39 mcmillhj joined 09:42 koalaaa joined
koalaaa m: BEGIN $?LANG.refine_slang: 'MAIN', role { token comment:sym<#> { '#' {say "IT WORKS"} \N* } }; say 42 # MEOW 09:43
camelia IT WORKS
42
koalaaa m: BEGIN $?LANG.refine_slang: 'MAIN', role { token comment:sym<#> { '#' {say "IT WORKS"} \N* } }; say 42; # MEOW
camelia 42
koalaaa And if I stick similar print statement into the very same token in src/Perl6/Grammar.nqp, then THAT print statement is printed. Any idea why it only works in this way and the redefined token only applies in some cases but not the other? 09:44
09:44 mcmillhj left
koalaaa And if I remove the token from the role in the first example, then the token in src/Perl6/Grammar.nqp gets triggered in its place 09:45
wander see 09:51
09:52 aborazmeh left 09:53 mcmillhj joined
koalaaa tyil: just use the --debug or --verbose switch to see more output. It fails because it can't find resources: gist.github.com/zoffixznet/99876a9...6bc66fa5be 09:54
tyil yes, I figured that out eventually
doesn't change the fact that the error is very much lta
koalaaa tyil: how can it be improved? It literally says "Failed to open file [...] SemVer-0.1.1/resources: No such file or directory". Looks PTG to me 09:55
tyil compared to an error message like No candidates found matching identity: SemVer:ver<0.1.3> 09:56
one line, says exactly whats wrong
09:57 zakharyas joined
koalaaa So what's the improvement? 09:57
tyil whereas the zef error on the resources is a big mess that happens to contain a line that refers to the real problem
09:58 mcmillhj left 10:01 mson left, zakharyas left 10:02 zakharyas joined 10:06 lizmat joined, mcmillhj joined
koalaaa Hm. Even if I stick the slang into a module, it still only works if the comment is inside some new scope 10:07
10:07 Cabanossi left
koalaaa Despite the fact that if I stick `BEGIN $?LANG.^name.say;` between the two comments, it tells me `Perl6::Grammar+{Recomment}`, indicating the role *is* mixed in at that point :S 10:09
10:10 Cabanossi joined, mcmillhj left 10:12 mcmillhj joined 10:16 mcmillhj left 10:18 _28_ria left
koalaaa And even if I just add an empty block after the `use` of slang module, then it works: `use Recomment; {}` 10:20
wtf :\
Oh wait, that's not even a block; it's a hash 10:21
Oh, it's the lack of semicolon that makes it work. If I change it to `use Recomment; {};` it stops working 10:22
10:28 _28_ria joined 10:29 zakharyas left
koalaaa No one knows why, eh? 10:30
ZofBot: *crickets* 10:32
10:32 parv left 10:35 koalaaa left 10:37 ryn1x_ joined 10:41 ryn1x_ left 10:42 wander left, AlexDaniel left 10:45 markong joined 10:47 steeznson left 10:50 lizmat left 10:54 cdg joined 10:59 cdg left 11:02 llfourn joined 11:07 Cabanossi left 11:09 AlexDaniel joined 11:10 char_var[buffer] left, Cabanossi joined 11:13 lowbro left
buggable New CPAN upload: App-Platform-0.1.0.tar.gz by KAJI cpan.metacpan.org/authors/id/K/KA/...1.0.tar.gz 11:23
11:30 kaare_ left 11:31 kaare_ joined
buggable New CPAN upload: SemVer-0.2.0.tar.gz by TYIL cpan.metacpan.org/authors/id/T/TY/...2.0.tar.gz 11:33
11:34 dct joined, dct left 11:54 Cabanossi left 11:55 Cabanossi joined 12:03 cdg joined 12:08 cdg left, raschipi joined 12:10 aeruder_ joined 12:14 koalaaa joined
koalaaa Figured it out: slang is loaded after we already started parsing and so it's too late to go back in time and affect how TOP, comp_unit, ws, statementlist, and stateenta 12:17
*statement get parsed.
In the mainline that is. Since comments are part of ws, they're affected too amd the change only works in nested scopes 'cause we start a brand new statementlist there. And the reason adding a clsemicolon-less block in mainline makes the following comment attain the change is 'cause that construct uses ENDSTMT, which has its own brand mew statementlist 12:19
So it's basically just a limitation in slangs. You can't affect a couple of tokens in mainline 12:20
q.e.d
12:20 rindolf joined 12:23 koalaaa_ joined
koalaaa_ Actually one mystery remains: I can affect tokens inside statementlist token, so how come tokens inside ws 12:24
12:24 koalaaa left, Cabanossi left 12:25 Cabanossi joined
koalaaa_ are not affected.... One guess is <ws> is part of the `rule` and not a literal <ws> in the regex, so maybe it's just outside the braid mutation and can't see it.... Wonder if sticking literal <ws> in statementlist would fix this.... 12:27
12:29 ryn1x_ joined 12:31 koalaaa_ left 12:34 ryn1x_ left
masak today's autopun spotting: twitter.com/znjp/status/933405548678021120 12:37
12:44 sjn left 12:45 sjn joined
moritz masak: nice one 12:46
12:52 Cabanossi left 12:55 Cabanossi joined 13:00 cdg joined 13:04 cdg left 13:17 eliasr joined 13:24 Cabanossi left 13:25 Cabanossi joined
ZzZombo m: my token identifier { <[\w \h]>* \w };say 'test' ~~ &identifier 13:25
camelia Nil
ZzZombo m: say '2 3' ~~ /(<[\w \h]>*?) (\w)/
camelia 「2」
0 => 「」
1 => 「2」
ZzZombo why the first doesn't match? 13:26
m: say '2 3' ~~ /(<[\w \h]>*) (\w)/
camelia 「2 3」
0 => 「2 」
1 => 「3」
ZzZombo err
I meant this.
13:29 kitsunenokenja joined 13:33 cdg joined
AlexDaniel ZzZombo: it did, but then it backtracked because you asked for \w also 13:38
or maybe I misunderstood the question
I probably did ;) 13:39
ZzZombo uh, sorry? how ca it match and not match at the same time?
jnthn Did you mean why the `token` case didn't match test? 13:40
m: my regex identifier { <[\w \h]>* \w };say 'test' ~~ &identifier
camelia 「test」
jnthn If so, you need regex 13:41
Or to rewrite it not to rely on backtracking
AlexDaniel relevant piece of docs: docs.perl6.org/language/regexes#Ratchet
13:43 cdg left, cdg joined 13:51 p000v joined
p000v What does nqp::rebless do? The docs say "Convert $obj to be an object of the new $type" but what does covert actually mean? Getting this error and trying to figure out why: "Incompatible MROs in P6opaque rebless for types Perl6::Grammar+{Recomment} and Perl6::Grammar" 13:52
jnthn It's the mechanism beneath infix:<does> 13:53
It does an in-place change of type
timotimo github.com/MoarVM/MoarVM/blob/mast...ue.c#L1165 - if you want the ugly details 13:54
p000v Thanks.
jnthn With the restriction that the MRO of the target type must have the MRO of the existing type as its suffix
13:54 Cabanossi left 13:55 Cabanossi joined
jnthn I strongly suggest using infix:<does> instead, or .^mixin 13:55
Unless doing something really unusual :) 13:56
13:56 dakkar left
p000v I'm trying to modify comment parsing :) 13:56
timotimo changing the type of something won't cause methods on the callstack to magically be replaced :) 13:57
p000v But that's the thing, the methods of, say, token that parses infix do get replaced, magically or otherwise, yet `token comment:<#>` inside ws doesn't get replaced until you start a new statementlist parse 13:58
timotimo maybe printing the stacktrace in a few places will be interesting?
p000v So I'm trying to figure out what is it that causes one to work and not the other. Seems the working ones use <EXPR> but the EXPR in grammar don't make anything special
jnthn m: sub infix:<omg>($a, $b = 1 omg 2) { } 13:59
camelia ( no output )
jnthn Hm, in that case it's put in place sufficiently to see it in the signature at least 14:00
An immediate mixin only affects the current Cursor, however, and the rebless in statement list is to make it take hold in the next statement
But there can certainly be "dead zones" where cursors further up the stack may not reflect the new language 14:01
14:02 zakharyas joined 14:03 zakharyas left, pmurias joined, zakharyas joined 14:05 zakharyas left, zakharyas1 joined
p000v timotimo: the stacktrace just corroborates that `EXPR` or `termish` do something that make the change be detectable in top-level scope. The `term:sym<bar>` is detected in top level, but `comment:sym<#>` isn't and only if I stick it inside curlies does it start to work and then the stacktrace shows EXPR/termish in there, but the top level wouldn't have them: gist.github.com/zoffixznet/5794fd7...aa386c86c7 14:07
m: BEGIN $*LANG.refine_slang: 'MAIN', role { token comment:sym<#> { '#' {say 'meow'} \N* } }; # bar 14:08
camelia ( no output )
p000v m: BEGIN $*LANG.refine_slang: 'MAIN', role { token comment:sym<#> { '#' {say 'meow'} \N* } }; {# bar␤} 14:09
camelia meow
p000v m: BEGIN $*LANG.refine_slang: 'MAIN', role { token termsym<bar> { 'b' {say 'meow'} \N* } }; b ar
camelia 5===SORRY!5=== Error while compiling <tmp>
Malformed regex
at <tmp>:1
------> 3fine_slang: 'MAIN', role { token termsym7⏏5<bar> { 'b' {say 'meow'} \N* } }; b ar
p000v m: BEGIN $*LANG.refine_slang: 'MAIN', role { token term:sym<bar> { 'b' {say 'meow'} \N* } }; b ar 14:10
camelia meow
p000v And I wanna steal that something to make the comment change work in top level too :) 14:11
jnthn Well, that starts a new statement
And iirc statement has a rebless in it 14:12
p000v Yeah it does
That the magic sauce? 14:13
p000v tries a few things
jnthn Yeah
14:14 dakkar joined 14:16 rindolf left 14:22 rindolf joined 14:25 cdg left
p000v Yeah. Sticking `<!!{ nqp::rebless($/, self.slang_grammar('MAIN')); 1 }>` into `token _ws` in grammar did the trick, but I guess there's no way to make it work from the outside, since it'd require reblessing things inside things that need to be reblessed 14:29
Geth: rakudo/master (p000v++ commited) "make `token _ws` 100x times slower; I need this feature for an article!!" 14:30
:)
Just in inner scope is good enough for my purpose. Glad to finally understand the behaviour tho 14:31
jnthn++
timotimo hm, spesh could eliminate rebless if the types are known to fit, and since reblessing anything will cause a deoptimization it might work out properly
moritz sounds like quite a speshial case 14:33
14:35 araraloren joined
araraloren HI 14:35
p000v \o 14:36
araraloren How to get Cro::HTTP::Client request content ?
It thrown an exception `Connection unexpectedly closed before response headers received` 14:37
I don't know what happened
jnthn araraloren: CRO_TRACE=1 in the environment 14:38
araraloren ok
ugexe title that exception from ComUnit::Repository but you seem so busy writing pissed off comments about zef in all your module commit messages that I can see why you would not have had the time to just look at the source 14:39
14:39 cdg joined
ugexe tyil ^ 14:39
tyil I wonder why 14:40
ugexe Me too
tyil I don't really feel like having to go through this again
p000v Then don't :)
tyil feel free to pm w/ otr if you truly wonder ;)
p000v There's a Steam sale. Go browse some cool games :)
araraloren jnthn, thanks, it works
14:43 p000v left, cdg left
ZzZombo <jnthn> Or to rewrite it not to rely on backtracking # how? 14:45
ugexe tyil you seem like an angry person 14:46
tyil I just invited you to discuss this in pm, and now instead you're going to make personal remarks against me
ZzZombo drama in perl6? I've seen everything... 14:47
ugexe You make them about “the author of zef” often
The racists stuff the other day isn’t helping 14:48
14:48 lizmat joined
tyil "racists stuff the other day"? 14:48
14:48 ryn1x_ joined
DrForr Drama? My... 14:49
huf awww, is the channel growing up? :)
timotimo having that "perl6-niggers" folder still in your distro 14:50
AlexDaniel huf: quite the opposite actually? :)
tyil "still"? I'm pretty sure I removed that
within a minute of spotting it
timotimo the moment you uploaded it, i mean
not sure why you'd call a folder that in the first place
tyil it was a test
DrForr Certainly seemed to be. 14:51
tyil but if you wish to call me a racist forever because of a mistake, I guess I can't stop you
even though I fixed it as soon as I could
timotimo there's literally millions of other test words you could have used …
tyil I'm not sure what else you want from me
yes, I could, but I didn't and made a mistake
should this be an issue forever now?
ugexe If you continue to be so angry, probably 14:52
tyil and now you're calling me angry for no reason
tadzik hah. I find myself wanting Perl 6 enums to be like Java enums: with constructors and associated values
AlexDaniel ugexe: ಠ_ಠ
tadzik ...they can't do that, can they
14:52 Zoffix joined
timotimo tadzik: you mean like my ADT module? ;) 14:52
14:52 ryn1x_ left
tyil again, I'm not sure what you want me to do, I invited you to discuss what issues we have in pm, but you don't want this 14:52
instead you're calling me racist and angry in here 14:53
Zoffix tyil: out of curiosity, what was it a test *of*?
tadzik timotimo: um, maybe %)
tyil Zoffix: `cpan6 new <repo>`, I should've removed the test dir but I forgot
I added functionality to show what is being packed up now, and tests happen automated outside of the module dir with a given set of names to avoid such things from happening 14:54
14:54 rindolf left
timotimo tyil: fwiw, you can still download the file in question from cpan with the same url 14:55
lizmat www.reddit.com/r/perl6/comments/7e...in_perl_6/ # nice, but very perl5ish
tyil isn't that the idea of cpan, to keep files available? 14:56
timotimo there's a deletion feature. it doesn't make the file disappear from the face of the earth, of course
lizmat tyil: modules that are past their due date, should be removed (manually) from CPAN 14:57
this is both needed to keep the total size down, *and* it allows you to make your mistakes less available
tyil timotimo: I removed the 0.8.0 release already though 14:58
timotimo hm, ok
tyil so again, I'm not sure what else you want me to do here
Zoffix It takes a bit of time (3 days I think) for scheduled deletion to occur
tyil or why I deserve to be called angry and racist without cause, and without the other party being willing to talk out the issue he or she has with me
Zoffix: those 3 days are over, they dont appear in my file overview anymore 14:59
I marked them for removal immediately after noticing
timotimo your apology is enough for me, fwiw
Zoffix Ah, I guess metacpan.org just keep them around.
15:00 geospeck left
tyil I don't think it allows for a inviting environment if you keep telling people they're angry for having made a mistake once either 15:00
Zoffix fwiw, I don't think you're angry, but I'm laughing at you tyil :) That's quite a faux pas
tyil so if anything, if you have a problem still, do bring it to pm
araraloren await $client.post( "{$ip}/goform/formLogin", headers => %{'Referer' => "{$ip}/dir_login.asp",}, body => ['username' => MIME::Base64.encode_base64('username'), 'password' => MIME::Base64.encode_base64('password'), 'Language' => 'Chinese', 'tokenid' => $tokenid ])
15:01 rindolf joined
tyil I'm trying my best here to work on modules I think are beneficial, and it's been going quite well and i've been having a ton of fun working in perl 6 15:01
araraloren The cro said No applicable body serializer could be found :(
tyil and I'd like it to stay that way
tadzik giggles github.com/tadzik/Net-Gopher
araraloren jnthn, any suggestion ? 15:02
ugexe tail I read your commit messages and they are often directed at me
AlexDaniel tyil++ # for handling the situation very well
ugexe I’m getting tired of it
Zoffix .hug everyone
huggable hugs everyone
timotimo i'll be AFK for an hour or so
tyil tadzik: heh, I was talking with some other people on Rizon earlier today about a perl 6 module to deal with gopher 15:03
jnthn araraloren: What are you expecting it to do with the body?
tadzik tyil: hah, I'm just in time :)
lizmat only did a gopher server once, but that was 23 years ago
tadzik this was fun, I want to do more
tyil more modules in general? 15:04
araraloren jnthn, I want login to the router
tadzik more gopher implementations
lizmat the client wanted a gopher server, I created a web server as well for the same price: they only ever used the web serer :-)
tyil or more gopher/networking stuff
tadzik I'm amused by silly and useless stuff :)
jnthn araraloren: Yeah, I mean how is the body meant to be serialized?
araraloren: The thing you're missing is a content-type argument to post
araraloren: Without one of those, it doesn't know whether it should turn the body into JSON, form-data/multipart, etc. 15:05
araraloren oh, jnthn
jnthn cro.services/docs/reference/cro-htt...quest_body
araraloren thanks, it was missing that headers
jnthn Examples there :)
araraloren: I don't mind answering cro questions here, but fwiw there's a 15:06
#cro channel now also
tadzik so, I should've been writing a gopher server for cro? %)
jnthn d'oh, hit enter at wrong time :)
araraloren Hmm, yeah I'm study with that example
jnthn tadzik: Reactive gopher microservies...why not? :D
araraloren hmm
tadzik Because we can[TM] 15:07
tyil
tadzik thanks. Because we can!™ 15:08
15:09 Cabanossi left
Zoffix 𝐁𝐞𝐜𝐚𝐮𝐬𝐞 𝐰𝐞 𝐜𝐚𝐧!™ 15:09
15:10 Zoffix left, Cabanossi joined
raschipi We do what we must because we can. 15:11
15:11 ugexe left 15:12 kitsunenokenja left
tadzik for the good of all of us. 15:13
raschipi Anyone here uses POWER for anything? 15:14
tyil I use a lot of power when compiling things 15:17
raschipi But do use as much power as people that use POWER?
15:18 araraloren left
tadzik Practically Open, Wrongly Executed Robotics? 15:19
ilmari assumes raschipi means the IBM CPU architecture 15:22
raschipi yeah, the one from IBM
ilmari debian has rakudo packages for ppc64el, at least
tyil thats more up to date than ubuntu 16.04 15:23
which still has parrot as vm for perl 6 15:24
ilmari uh, no 15:25
packages.ubuntu.com/xenial/rakudo
did you mean 14.04? packages.ubuntu.com/trusty/rakudo
tyil oh 15:27
we use 14.04 at work then 15:28
oh well 15:29
I use rakudobrew on all my machines
El_Che tyil: github.com/nxadm/rakudo-pkg <-- 14.04 pkgs 15:30
tyil I've learned about that as well yeah 15:31
El_Che I added 14.04 because travis runs that
tyil I was wondering if the travis perl6 builds could make use of that
but I havent found the right time to look at those possibilities yet
most tests now take 10 minutes, from which 9 minutes installing perl 6 :p 15:32
El_Che tyil: actually, I do: github.com/nxadm/rakudo-pkg/blob/m...travis.yml
tyil :o
El_Che If triggered, I could build historical versions and put them in a repo
so you could test against all post christmas rakudos 15:33
tyil install: wget $RAKUDO_DEB && sudo dpkg -i *.deb
I could use this
El_Che I haven't been triggered yet, I haven't had the time and I don't know if there is a need for it
15:35 starmarkd joined, starmarkd left 15:38 starmarkd joined
tyil El_Che: I'll look into updating App::Cpan6 to use that install script to lower test times 15:38
15:42 zakharyas1 left, ryn1x_ joined, zakharyas joined 15:43 poohman left 15:46 ryn1x_ left 15:50 khw joined 15:51 rocx joined 15:53 domidumont left 15:55 comborico1611 joined 15:56 ryn1x_ joined, geospeck joined 15:58 scimon left 15:59 scimon joined
El_Che tyil: ping me if several version of rakudo are needed. Add far low in my todo list, but you never know 16:00
tyil: I create 14.04 pkgs for every rakudo release.
tyil will do 16:01
:>
moritz if you need a repo software that can handle muliple version of the same package, use aptly
(though .deb only)
16:01 raschipi left 16:02 sjn left, Zoffix joined
Zoffix Does anyone have a script to convert Perl 6 code to HTML with syntax highlights? I know there's a Vim one in ecosystem, but I was thinking of the docs site's highlights, since they're probably more thorough... 16:04
moritz there's pygments also 16:05
El_Che moritz: I am experimenting with bintray
Zoffix Or more to the point: does anyone have a script to convert markdown to decent Advent articles? Last year I used gist scraper, but this year Perl 6 highlights are broken on github
timotimo you can probably gist the code and copy-paste the result into a "contentseditable" text entry thingie
El_Che moritz: it supports rpm, deb and other stuff
timotimo dangit
Zoffix Actually, if you gist as a Perl 6 code instead of markdown, it does highlight it. I think that'll work for my purpose. 16:07
16:10 ryn1x left, lizmat left
tyil El_Che: I was wondering why travis was complaining about a missing t directory now 16:10
but I just realized I branched off of master, and I have my tests in another branch 16:11
Geth mu: 2cb5712440 | (Itsuki Toyota)++ (committed using GitHub Web editor) | misc/perl6advent-2017/schedule
Add titsuki as a 13th writer
16:12
16:14 troys_ is now known as troys
Zoffix \o/ 16:14
Geth mu: 7517addb7e | (Moritz Lenz)++ | misc/perl6advent-2017/schedule
Move my advent post further down

I really want my Perl 6 Regexes book to be published by then, and it is being delayed a bit due to problems with the layout/typography
If this leaves a hole early on the in the schedule, I can try to fill it up with a different post.
16:15
Zoffix moritz: I can move "Zoffix: "Perl 6: Sigils, Variables, and Containers"" into the hole; it's ready already 16:16
If you wanty
El_Che tyil: that's a weird setup 16:17
tyil I was making the tests in a new branch
I didn't have tests before yet :(
once I have all enough tests they'll go back into master
but currently my tests cant pass because it requires a change in a dependency, and my PR there hasnt gotten a response yet :( 16:18
moritz Zoffix: that would be good, yes 16:20
Geth mu: 478c1adada | (Zoffix Znet)++ (committed using GitHub Web editor) | misc/perl6advent-2017/schedule
Mark zofspots as takable/fill 2nd hole
16:21
16:22 kaare_ left 16:23 Cabanossi left
El_Che tyil: ok, now it makes sense. A little :) 16:23
16:23 Ven`` joined
tyil I understand its not the best setup ever, but its also still in an early development stage 16:23
16:23 kaare_ joined
tyil preparing for a stable 1.0 release before the end of the year 16:24
El_Che remind me again what is does?
it's a cpan(m) like app? 16:25
geospeck What is the easiest way to do an http get request using the stdlib only?
16:25 Cabanossi joined
tyil it may one day become an actual cpan-like app, but I dont know enough about installing modules yet, so for now its just a tool to assist in building modules 16:25
setting up skeletons for new modules, dealing with creating new files and adding them to META6 16:26
Zoffix geospeck: stdlib is what? Whatever comes with Rakudo Star?
moritz geospeck: if you talk about the compiler only: I don't think there's a built-in HTTP client, so you'd have to open a socket and talk HTTP yourself
tyil I mostly very much disliked having to manually edit the META6.json for every change
geospeck yes, I mean without installing 3rd party libraries
16:26 zakharyas left
El_Che tyil: like p5's module-starter? 16:26
tyil yes
El_Che nice
tyil++
Zoffix star: use LWP::Simple 16:27
camelia ===SORRY!===
P6M Merging GLOBAL symbols failed: duplicate definition of symbol INET
Zoffix :/
tyil I have an unpublished tutorial to show basic use of it (and teach new perlers to create their own simple application with a gui)
geospeck moritz: docs.perl6.org/type/IO::Socket::INET is this the doc ?
tyil if you have time to spare, you're free to proof-run it, I want one or two more people to go through it before I publish it
Zoffix geospeck: well, LWP::Simple should be with the star. `use LWP::Simple; say LWP::Simple.get: 'blah blah'` is the simplest way 16:28
tyil I want to be very sure it works properly, as it is intended for new users
who usually don't know how to troubleshoot stuff yet
geospeck Zoffix thank you! that's great
Ven`` looking for advent topic ideas if someone's got one to throw my one. considering talking about macros... maybe just cro :). 16:29
Zoffix Ven``: Cro! 16:30
Ven``
.oO( Perl 6 advent day 3: How to use type shenanigans to crash MoarVM )
16:31 ryn1x_ left
tadzik . o O ( my Macro $cro ) 16:31
Zoffix
.oO( Macbook Cro )
16:32
16:32 Zoffix left 16:33 ChoHag left 16:34 _28_ria left
timotimo Crom Book Pixel™ 16:41
16:46 wamba left
teatime Happy Thanksgiving, Perl-people. o/ 16:47
El_Che s/(Perl-people)/$1 in the USA/ 16:48
:)
teatime I wished my Toronto friend a happy non-canadian thanksgiving.
El_Che the rest of the World is less sympathetic to the official story of the extinsion of native people :) 16:49
16:49 rocx left
El_Che but, Happy Thankgiving nevertheless for my friends in the US :) 16:49
16:49 _28_ria joined 16:51 poohman joined 16:52 ryn1x joined 16:54 Cabanossi left 16:55 Cabanossi joined 16:58 abraxxa left
ryn1x I needed some progress indicators for a project I am working on so I decided to make my own as a learning exercise. I am sure there is a conventional way to do this, but this is what I came up with. I would like any feedback on what I could/should do differently. github.com/ryn1x/Spinners 17:00
tyil are you looking to publish it as a module? 17:04
ryn1x I doubt it is worthy of being published... just trying to learn the basics so maybe one day I could write something to publish. 17:05
El_Che ryn1x: it sounds like a useful module to me 17:07
tyil ^ 17:08
the demo is well documented, so its easy to get started with it
I'd change `multi MAIN` to `sub MAIN` since you dont have multiple MAINs
ryn1x tyil: I did not notice it was still a multi. Thanks! I had a new mains while testing. 17:09
tyil I'd split up the class definition and the two subroutines you have into two seperate files
and while certainly not required, I prefer to add a `#! /usr/bin/env false` to files that should be loaded instead of executed 17:10
also, I see you're backspacing 80 chars to render the new bar 17:11
what happens if someone presses space while the bar is rendering?
ryn1x If I split them into two seperate files would I have to import them seperately or could they still be the same module?
tyil they can still be in the same module, but imported seperately 17:12
ryn1x I have not tried to debug that..I do not know of another way to render the progress bar in-place...
tyil me neither, so I don't know if there's a better solution for this 17:13
ryn1x I just tried... hitting the spacebar interferes with the spinners, but does not seem to affect the progress bars 17:14
tyil neat
ryn1x I suspect that might change with the speed they are updating though... 17:15
Nope... just slowed it way down... progress bar still is not affected by pressing space...interesting 17:17
tyil nice 17:18
have you ever released a module before? 17:19
ryn1x No. I am very new to Perl6, and programming in general for that matter. 17:20
tyil would you like to release a module?
17:21 sjn joined
ryn1x Yeah of course. I hope to try and start giving back to this community. I just feel a little inexperienced right now. 17:22
tyil everyone started at that point
if you want to release Spinners as a module, hit me (or anyone else in here) up, and we can assist you in the process 17:23
your first module is usually the scariest to release, so if you don't want to get started on it yet I understand too :)
that said, its time to get working on lunch/dinner, so I'll be afk for a while 17:25
17:25 mson joined
tyil I should be back in two hours if you need me 17:25
ryn1x Ok. Let me refactor it a bit with your suggestion to split into two files. I will also need to change the directory organization to match other modules right? I assume "use lib '.'" in in my demo file is incorrect too.
17:25 cdg joined
ryn1x ok thanks so much for your feedback 17:25
tyil here's a tiny module I made a little while ago, you can look at that for the structure: github.com/scriptkitties/perl6-SemVer 17:26
having a META6.json is mandatory
17:26 scimon left
ryn1x great. thanks! 17:27
tyil vanishes
17:28 ryn1x_ joined 17:29 cdg left 17:31 gregf_ joined 17:32 ryn1x_ left 17:33 starmarkd left
buggable New CPAN upload: JSON-Fast-0.9.7.tar.gz by TIMOTIMO cpan.metacpan.org/authors/id/T/TI/...9.7.tar.gz 17:33
New CPAN upload: Font-FreeType-0.0.3.tar.gz by WARRINGD cpan.metacpan.org/authors/id/W/WA/...0.3.tar.gz
timotimo it me 17:34
Ven`` IT-mp
mo*
17:36 char_var[buffer] joined 17:37 dakkar left
buggable New CPAN upload: JSON-Fast-0.9.8.tar.gz by TIMOTIMO cpan.metacpan.org/authors/id/T/TI/...9.8.tar.gz 17:43
17:47 wamba joined
timotimo "size: 13364 bytes" 17:50
:\ 17:51
17:52 poohman_ joined 17:56 poohman left, poohman_ is now known as poohman, poohman left 17:57 char_var[buffer] left 17:58 Ven`` left 18:00 char_var[buffer] joined 18:03 setty1 joined 18:05 domidumont joined
Geth mu: 6ab13a9edd | (Aleks-Daniel Jakimenko-Aleksejev)++ | misc/perl6advent-2017/schedule
Claim 23rd, moved other posts higher
18:07
18:07 troys is now known as troys_
Geth mu: 93d8065f86 | (Zoffix Znet)++ (committed using GitHub Web editor) | misc/perl6advent-2017/schedule
Make schedule look emptier
18:12
18:18 raschipi joined 18:29 lizmat joined 18:30 lizmat left 18:35 darutoko left 18:42 committable6 left, committable6 joined 18:57 pmurias left 18:58 ryn1x left 19:26 raschipi left 19:33 wamba left 19:34 wamba joined 19:38 TreyHarris joined 19:43 Xliff left
moritz so, I've been pondering more advent post topics 19:45
19:46 ryn1x_ joined
moritz and if there's interested, I could publish a chapter from the Perl 6 Regexes book 19:46
s/interested/interest/
tyil my interest has been tickled
19:49 quotable6 joined, ChanServ sets mode: +v quotable6, coverable6 joined, greppable6 joined, bloatable6 joined, nativecallable6 joined, tangible6 joined, benchable6 joined
moritz this is the table of contents: perlgeek.de/static/tmp/perl6regex-toc.pdf 19:49
19:49 statisfiable6 joined
moritz any preference for a particular chapter? 19:50
tyil chap5, it sounds like it will introduce a lot of practical niceness of regexes/grammars 19:51
9 also looks like a good fit to show some nice things and practical things 19:52
19:54 avalenn joined, Cabanossi left
moritz ok, thanks for the feedback 19:55
I'll wait for some second or third opinions
19:55 Cabanossi joined 19:57 comborico1611 left 20:03 domidumont left 20:17 mojca joined
mojca how I can figure out more details about this test failure? paste.macports.org/54b9a0914c09 20:18
tyil are you running prove? 20:19
20:20 ryn1x_ left
mojca no, just make test 20:20
should I run something else?
tyil oh, you're building perl6, not a simple module
20:20 geospeck left 20:21 xinming_ is now known as xinming
tyil best to wait for someone with more knowledge on that then 20:21
mojca yes, this is a failed test for rakudo
moritz mojca: run ./perl6-m -Ilib t/04-nativecall/17-libnames.t
probably doesn't even need -Ilib 20:22
mojca 1..7 20:23
Segmentation fault: 11
if I use llib: Grouped options '-llib' contain 'l', which is not a valid option
timotimo whee
it has to be capital i, not lowercase L 20:24
moritz mojca: no, -I (captial i), not -l
mojca: anyway, you can run ./perl6-gdb-m t/04-nativecall/17-libnames.t
and then "run"
and when it segfaults, "bt"
to get a backtrace
timotimo failures in native call aren't as terrible as other failures; it's not necessary for rakudo to run
mojca with -Ilib it's seven times: ok 1 - # SKIP <unknown> 20:25
timotimo you're not on linux? 20:26
gotta go afk
20:26 rindolf left
mojca how can I specify which gdb is used? 20:28
because it's called ggdb
moritz mojca: perl6-gdb-m is a shell script. You can edit it to your liking 20:29
mojca no, I'm not on linux, it's mac
ok, there seems to be one problem 20:33
the tests will use the installed version or rakudo rather than the one currently being built & tested
if I first install the new version and then test it, it works
I'm not too hapy about it 20:34
I used lldb and it refused to crash 20:35
20:39 Cabanossi left 20:40 Cabanossi joined 20:47 mojca left 20:48 espadrine joined 20:57 ryn1x joined 21:03 cdg joined, mudman joined 21:06 _28_ria left 21:07 cdg left 21:10 zakharyas joined 21:18 lizmat joined 21:24 wamba left 21:37 ryn1x left 21:39 ryn1x joined
ryn1x tyil: are you on? 21:43
tyil yes I am
what's up
ryn1x I made some changes to my module to get it ready to possibly publish. github.com/ryn1x/Terminal-Spinners 21:44
tyil looks nice 21:45
you can test if its installabe by just trying to install it 21:46
`zef install .`
ryn1x Did that and it works.. tried on two different computers, ubuntu and macos
tyil then you're good to go 21:47
modules are usually uploaded to CPAN nowadays
for this you need a PAUSE account, which you can request here: pause.perl.org/pause/authenquery?A...request_id
ryn1x Ok. I will register. I don't know anything about CPAN yet. 21:48
tyil its the Comprehensive Perl Archive Network 21:49
a network of servers serving Perl modules
this may be of interest to you: docs.perl6.org/language/modules.ht...le_to_CPAN 21:50
22:02 ChoHag joined 22:09 Cabanossi left 22:10 Cabanossi joined 22:11 mudman left
ryn1x Thanks for all the support tyil. I created the archive after reading the link you sent. Just waiting for my registration to be approved. 22:14
tyil thats usually within the hour 22:16
but I cant make any guarantees on that
once you have an account, you can upload manually via the webinterface, or use a module like App::Cpan6 (disclaimer, this is my own module, there may be alternatives available) 22:17
ryn1x Cool. Ill check it out.
tyil also, the convention name for those tarballs is Module-Name-version.tar.gz, which in your case would be Terminal-Spinners-0.0.3.tar.gz
22:18 setty1 left
tyil its also good to know that you cannot overwrite an upload 22:18
so once you upload that tarball, you can not re-upload it with fixes, and not upload the same version of Terminal::Spinners 22:19
you'd have to make fixes,increase the version number, and then upload a new tarball
ryn1x Perfect. That is what I named it with "git archive".
ok. that is good to know. 22:20
tyil also, consider any uploads to cpan permanent
even if you mark to remove them, they can (and usually will) still be available through cpan 22:21
ryn1x Ok. And you do recommend publishing my module through CPAN and not github right? Seems like a lot of the modules now are from github, but I think that CPAN support for perl6 is relatively new right? 22:22
22:23 mudman joined
tyil there are many still through github, but as far as I know, that will be deprecated at some point in time 22:23
it doesnt scale very well
as such I'd recommend using cpan, for it is the future
ryn1x ok 22:24
22:43 MasterDuke left 22:49 zakharyas left 22:52 Cabanossi left 22:55 Cabanossi joined 23:25 evalable6 left, evalable6 joined, ChanServ sets mode: +v evalable6 23:28 lizmat left 23:39 Cabanossi left 23:40 Cabanossi joined