»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, std:, or /msg camelia p6: ... | irclog: irc.perl6.org | UTF-8 is our friend! | feather will shut down permanently on 2015-03-31
Set by jnthn on 28 February 2015.
virtualsue timotimo: you'd have to ask ilmari. i gather he barely saw any of it even so. 00:01
00:02 raiph left
timotimo holy wow. 00:03
apparently just using at_pos instead of AT-POS (which has the burden of making sure the program will warn about its impending deprecation) makes farabi serving pages A WHOLE LOT slower
that's a way to discourage deprecated methods
00:08 gcole joined 00:09 eli-se left
ugexe i keep getting the following error during gen/jvm/BOOTSTRAP.nqp on win32 now: java.lang.LinkageError: loader (instance of org/perl6/nqp/runtime/ByteClassLoader): attempted duplicate class definition for name: "2298F07D3EE83B7B2FB42666D8F5903870D8701D" 00:20
the hash(?) changes each time though 00:22
00:27 laouji joined 00:37 mr-foobar left 00:38 mr-foobar joined 00:39 laouji left 00:40 laouji joined 00:41 larion joined
dalek kudo-star-daily: 5a32219 | coke++ | log/ (9 files):
today (automated commit)
 00:43
rl6-roast-data: c3b41a1 | coke++ | / (9 files):
today (automated commit)
00:44 laouji left 00:45 C0rruptor left 00:48 larion left 00:50 lichtkind left, laouji joined
dalek kudo/tab-completion: fdb8ed9 | hoelzro++ | / (2 files):
Decouple completion from linenoise
 00:55
00:59 Foxcool left 01:00 Foxcool joined
b2gills nsharrok: I have my doubts that anyone has attempted that 01:05
01:06 pecastro left 01:08 raiph joined
Mouq hoelzro: Is there some way to avoid doing self.eval("use Linenoise;…") over and over? maybe bind to a list of functions and then assign each variable to each position of the returned list? 01:08
01:08 pecastro joined 01:09 aborazmeh joined, aborazmeh left, aborazmeh joined 01:10 thilp left
Mouq wonders if we could allow "use Linenoise:from<Perl6>" in NQP… 01:10
01:10 aborazmeh left
[Coke] Wouldn't you need to load Perl 6 to do that? 01:13
01:21 larion joined
hoelzro Mouq: I would like to find a way; I'll be refining it over time 01:22
I don't think NQP supports use ...:from<...>
01:27 larion left 01:35 mr-foobar left 01:53 kaare_ left
nsharrok Thanks Brad. I hadn't seen anything on IIS setup. I'm having a go at setting up a handler mapping. I'll let you know 01:57
02:04 larion joined
dalek k-simple/dllbundle: 3908285 | retupmoca++ | / (28 files):
Bundle .dll files for windows

Note that this introduces a 300 line NativeLib.pm6 file to properly load the .dll files when they aren't in the system path. See the comment at the top of that file for details.		 02:04
retupmoca ^ can a couple of people look at that GTK::Simple branch and let me know if we want something like that? 02:06
or if it isn't worth it and we should just point people to the GTK download page
02:08 larion left 02:09 larion joined
retupmoca (use case: I have a toy windows program that uses GTK::Simple. If I can distribute it without mucking with $PATH and 3rd-party installs, I would be very happy) 02:10
02:11 virtualsue left 02:14 larion left 02:15 virtualsue joined
skids m: my uint32 @s; my uint32 @t; @s »+« @t; 02:17
camelia rakudo-moar af6bd0: OUTPUT«Cannot find method 'array[uint32]'␤ in sub hyper at src/gen/m-CORE.setting:22228␤ in block at src/gen/m-CORE.setting:22094␤ in block <unit> at /tmp/hblT4N5yQN:1␤␤»
02:20 azawawi joined
azawawi hi 02:20
skids o/ 02:21
02:25 adu joined 02:28 raiph left 02:33 virtualsue left 02:39 kaare_ joined, kaleem joined 02:46 larion joined 02:48 chenryn joined 02:49 anaeem1 joined 02:52 larion left, C0rruptor joined
skids m: my $i = 0; my @A = 1,2,3; @A.push(@A.map: { $_ }); @A.say; my uint32 @a = 1,2,3; @a.push(@a.map: { die(@a.gist) if $i++ > 10; $_ }); @a.say; 02:52
camelia rakudo-moar af6bd0: OUTPUT«1 2 3 1 2 3␤1 2 3 1 2 3 1 2 3 1 2 3 1 2␤ in block <unit> at /tmp/kxA4JxTEJb:1␤␤»
02:53 anaeem1 left 02:58 karniel joined
adu hi skarn 02:59
hi skids
skids o/ 03:00
adu how goes?
03:00 C0rruptor left
skids Having fun working some native sized types into Sum. 03:00
adu skids: ooo the crypto digest thing?
skids Not seeing much speedup, because I think my slowness is in the marshalling. 03:01
adu skids: have you tried a disasm?
japhb .botsnack
yoleaux 21 Mar 2015 09:37Z <nine_> japhb: I cannot reproduce your MUTEX_LOCK Inline::Perl5 failure. Can you give me some information like which exact Perl 5 version with which configuration options you're using and maybe a backtrace?
:D
22 Mar 2015 12:07Z <Ulti> japhb: let me know if there is anything I can do to make this worthy of perl-bench github.com/japhb/perl6-bench/pull/23
japhb .tell Ulti OK, I replied to your perl6-bench pull request. 03:02
yoleaux japhb: I'll pass your message to Ulti.
skids adu: no, I did try to make sense of a profile run butI don't know enough guts to figure out what to look for.
03:02 chenryn left
japhb nine_: I'll run a fresh build now, and see if I can recreate the Inline::Perl5 problem with Rakudo HEAD 03:03
03:04 karniel left
adu skids: what's the difference between buf8 and blob8? 03:04
03:04 C0rruptor joined
skids blob8 ir ro, buf8 rw 03:04
03:04 chenryn joined
skids *is 03:04
adu ah
skids m: say buf8 ~~ blob8; say blob8 ~~ buf8 03:05
camelia rakudo-moar af6bd0: OUTPUT«False␤False␤»
skids hrm. works for MMD.
adu skids: hmm, I tried building it and got "lib/Sum/Recourse.pm6" 03:06
skids: "Failed building" ...
skids Yeah, jvm is messed up right now.
adu I'm on moar 03:07
skids Sure? that's where jvm fails.
adu I'm on moar, for sure
skids I just did a panda install Sum and got through build stage just fine. How old is your rakudo? 03:08
adu pretty old
647fbd1
I'm updating now
03:10 noganex_ joined 03:12 noganex left 03:13 aborazmeh joined, aborazmeh left, aborazmeh joined
skids Oh well, I don't have time to hunt down a "Must parameterize array[T]" mystery precomp problem before bed, so a push will have to wait. 03:20
adu skids: so I just updated rakudo to latest, and rebuilt Sum, and no luck 03:23
skids Is this Win? 03:24
adu skids: Mac
skids Ah.
adu pastie.org/10046356
skids Well, at least that is more helpful than the JVM. Wonder why OSX has different nativecall. 03:25
adu skids: macho? 03:26
there are actually two major differences, linux uses ELF, mac uses Mach-O, and ELF files only contain 1 arch, whereas Mach-O files can contain x86-32, x86-64, ppc, ppc64, etc. 03:27
skids Maybe Mach-O-specific code is lagging behind. 03:28
I would expect if you were accidentally loading the out-of-tree older zavolaj stuff that it would blow up sooner than that, but might be worth checking. 03:30
adu skids: I dunno
skids: but userspace mac stuff is usually pretty portable to/from linux 03:32
but the linker generally messes up any plans of portability
skids Unfortunately no bsd or mac reports at testers.p6c.org lately. 03:33
adu hmm
[Coke] How does one submit reports to testers.p6c.org ? 03:35
(there's no docs on the home page)
skids Yeah it's a magic env var when running panda. It really needs to be put on the testers.p6c.org front page.
adu what's the var? 03:36
skids It's on a blog somewhere, looking.
[Coke] or in the README on the fork-able site, that would help too
skids PANDA_SUBMIT_TESTREPORTS 03:37
It was in weekly.
[Coke] IWBNI if rakudo-brew also dealt with panda's path. 03:38
adu woah 03:39
this might have something to do with it: pastie.org/10046387
[Coke] I am just built Sum using rakudobrew on OS X with no errors
s/am //
skids adu: it shouldn't fetch nativecall anymore. 03:40
adu wtf is my problem
[Coke] why are you fetching nativecall?
skids adu: probably an old out of date panda cache.
[Coke] have you upgraded your rakudobrew/panda recently?
adu [Coke]: yes, about 5 mins ago
how do I panda uninstall? 03:41
skids [Coke]: yeah I was just about to script the necessary steps to refresh panda when you --configure-opts="--prefix=*" when it changed from languages/site to site so I punted today. 03:42
03:44 telex left
skids adu: it can get a bit messy, especially since the paths just changed. I would first try mv'ing json/state out of site/panda 03:44
adu I'll try deleting rakudobrew and starting again
skids You'll probably have to do something in theinstall path as well as that.
adu skids: install path? 03:45
you mean ~/.rakudobrew/bin?
skids Well, if that is where you installed, you might be OK.
adu yeah I was going to remove ~/.rakudobrew and start over 03:46
03:46 larion joined, telex joined
[Coke] building sum the first time failed. building it a second time with the special env var set for submitting reports, it worked and installed. 03:46
skids but /usr/share/perl6 might exist or /usr/share/languages/perl6 from older stuff.
adu nope 03:47
it's all in ~/.rakudobrew
skids [Coke] hah well at least I got the ood side of the coin in the stats :-)
03:47 kaleem left
[Coke] how long does it take for submittals to be recorded? 03:47
er, reported is a better word.
skids I didn't remember there eing much delay. 03:48
[Coke] and is "PANDA_SUBMIT_TESTREPORTS=1 panda install Sum" sufficient?
skids should be I think.
[Coke] ok, don't see anything at testers.p6c.org/dist/S/Sum/Brian%20...Julin.html
nor any indication in the run that it sent anything
03:49 laouji left
skids I'm running the same now. 03:50
adu oh no testers.p6c.org/dist/C/C::Parser/An...bbins.html
does that red mean that the perl6-j screwed up, or does that mean I need to write a fix for jvm? 03:51
03:51 laouji joined
azawawi how can one force panda into installing into core lib instead of site/lib? 03:51
03:52 davido_ left
azawawi e.g. panda install HTTP::Easy 03:52
skids azawawi: I haven't figured that out yet myself.
03:53 chenryn left, davido_ joined, chenryn joined
azawawi skids: dual life modules pain... 03:54
adu azawawi: symlinks? 03:55
azawawi adu: what about windows? 03:56
japhb .tell nine_ Yep, I get 'panic: MUTEX_LOCK (22) [op.c:697].' installing Inline::Perl5 on rakudo-moar-af6bd025a5e1028b4de047fdbf76da833e92d992 using Ubuntu LTS system perl5: 'This is perl 5, version 18, subversion 2 (v5.18.2) built for x86_64-linux-gnu-thread-multi'
yoleaux japhb: I'll pass your message to nine_.
adu azawawi: msdn.microsoft.com/en-us/library/w...85%29.aspx
azawawi adu: i know about mklink 03:57
03:57 chenryn left
adu azawawi: ah, I didn't 03:58
I don't use windows
azawawi adu: the real problem is with installing a newer version of HTTP::Easy on rakudo. You need to nuke lib/HTTP/*
adu: im talking about rakudo star ofcourse 03:59
skids tried with "export PANDA_SUBMIT_TESTREPORTS=1; panda install" to no avail.
azawawi skids: tried PANDA_SUBMIT_TESTREPORTS=1 panda install ?
skids Tried that as well, but the latter would be more definitive if anything. 04:00
04:01 tinyblak left
skids I do note that is is 2/3 the way through march, and there are few 2015.3 entries compared to 2015.2. So I'm guessing submissions are not being taken maybe. 04:02
azawawi takes a look at panda 04:03
skids Anyway, bedtime. 04:05
04:06 Sqirrel left
adu the mac bug still hasn't gotten fixed 04:07
pastie.org/10046425
04:08 b2gills left 04:13 BenGoldberg left
adu o wait 04:13
it has been fixed
04:17 skids left 04:20 azawawi left
adu should I be using ./moar-nom/install/bin or ./moar-nom/install/share/perl6/site/bin or ./moar-nom/panda/bin? 04:21
04:27 laouji left 04:30 nsharrok left 04:34 b2gills joined 04:36 tinyblak joined 04:41 laouji joined 04:44 Akagi201 joined 04:46 jack_rabbit left 04:54 chenryn joined 04:58 nige joined 04:59 chenryn left 05:05 aborazmeh left 05:06 aborazmeh_ joined, aborazmeh_ left, aborazmeh_ joined, aborazmeh_ is now known as aborazmeh 05:25 diana_olhovik_ joined 05:28 adu left, chenryn joined 05:30 aborazmeh left 05:31 Sqirrel joined
Mouq Heh. The profiler page doesn't much like hundred-ly recursed functions :P 05:32
05:46 kaleem joined 06:01 xinming_ joined 06:04 jack_rabbit joined, xinming left 06:08 mr-foobar joined
Mouq just realized ideone.com has a Perl 6 option :O 06:09
"say $*PERL.compiler" => "rakudo (2014.7)"; "say $*VM" => "parrot (6.6.0)" 06:13
Could be worse
06:14 mr-foobar left 06:16 mr-foobar joined 06:20 mr-foobar left, bjz joined 06:21 mr-foobar joined 06:23 C0rruptor left 06:28 diana_olhovik_ left 06:33 rurban joined 06:34 yeahnoob joined 06:36 bjz left 06:41 jack_rabbit left 06:42 MadcapJake left 06:46 espadrine joined 06:48 jack_rabbit joined 06:55 mr-foobar left 06:58 mr-foobar joined, jack_rabbit left 07:02 mr-foobar left 07:05 xprime left, diana_olhovik joined 07:06 aborazmeh joined, aborazmeh left, aborazmeh joined 07:07 mr-foobar joined 07:08 prime joined, prime left, prime joined 07:10 jack_rabbit joined 07:20 MadcapJake joined 07:21 anaeem1 joined 07:24 jack_rabbit left 07:25 Rounin joined 07:28 rindolf joined 07:29 jack_rabbit joined 07:36 bjz joined 07:37 anaeem1 left 07:46 andreoss joined 07:49 Akagi201_ joined, nige left 07:50 bjz left 07:51 Akagi201 left 07:58 Ven joined 08:01 jack_rabbit left, Ven left 08:03 Ven joined
nine_ .tell japhb Cannot repor using the same perl and rakudo version. Will get me an Ubuntu 14.04 LTS VM and try it there. 08:06
yoleaux 03:56Z <japhb> nine_: Yep, I get 'panic: MUTEX_LOCK (22) [op.c:697].' installing Inline::Perl5 on rakudo-moar-af6bd025a5e1028b4de047fdbf76da833e92d992 using Ubuntu LTS system perl5: 'This is perl 5, version 18, subversion 2 (v5.18.2) built for x86_64-linux-gnu-thread-multi'
nine_: I'll pass your message to japhb.
08:06 larion left 08:07 fhelmberger joined 08:11 darutoko joined
Ven o/, #perl 08:13
08:13 FROGGS joined
Ven o/, #perl6*. 08:13
08:13 abraxxa joined 08:14 aborazmeh left
sergot morning \o 08:16
08:16 mr-foobar left 08:17 bjz joined 08:18 azawawi joined
azawawi hi 08:18
FROGGS morning
azawawi im looking for the msi for rakudo starr 2015.03
s/starr/star 08:19
moritz if it's not on rakudo.org/download/star, it doesn't exit
*exist
08:19 grondilu joined
FROGGS I'll do that this evening 08:19
azawawi FROGGS++ 08:20
how does one become a win32 release maintainer? :) 08:21
moritz azawawi: by simply doing it
azawawi: when you have something to upload, I can upload it for you, or give you access directly 08:22
azawawi moritz: i meant the release procedure :)
FROGGS azawawi: there is an undocumented 10 step instruction list
I probably should put the steps I do online... 08:23
azawawi listens
moritz FROGGS: please document it
abraxxa morning 08:24
FROGGS: just saw that my git push halted at the github password prompt ;( 08:25
FROGGS abraxxa: a push to your DBIish?
nwc10 good *, *
FROGGS hi nwc10
abraxxa FROGGS: yes
08:26 Ven left
FROGGS azawawi: gist.github.com/FROGGS/6753acf3fe110f100782 08:26
masak morning, #perl6
FROGGS azawawi: we can get rid of one step or the other by creating the .bat files automagically... 08:27
azawawi FROGGS: sure
FROGGS azawawi: and I also need to document how your build environment has to look like
08:28 zjmarlow joined 08:31 brrt joined 08:32 zjmarlow__ left 08:33 zjmarlow left, zjmarlow joined 08:34 Akagi201_ left, Akagi201 joined 08:39 sivoais left 08:40 Akagi201_ joined 08:43 pdcawley joined, sivoais joined, coffee` joined, Akagi201 left 08:45 sirsleepalot left 08:46 yeahnoob left 08:48 sivoais left, Alina-malina left 08:49 Alina-malina joined 08:51 sirsleepalot joined
abraxxa great, the demo oci c apps all don't use floats... 08:53
azawawi FROGGS: could you please send me the msi rakudo star build environment via .tell once you're finished with it? 08:54
moritz abraxxa: have you looked on github for "example" code? 08:55
FROGGS azawawi: will do
08:55 espadrine left
abraxxa moritz: not directly, duckduckgo didn't really bring up anything useful for SQLT_FLT. the datatype docs mention binding as string, but that lead to a segfault 08:56
08:56 kjs_ joined
abraxxa i've tried DBD::Oracle as template but as I'm no C programmer that code is way over my head 08:56
moritz github.com/search?q=SQLT_FLT&t...=%E2%9C%93
abraxxa i only say that it also binds using SQLT_FLT
moritz no idea how useful that is 08:57
FROGGS "bind_unsupported_1.phpt "
azawawi FROGGS: thanks :)
abraxxa moritz: thanks
FROGGS abraxxa: github.com/nonrational/qt-everywhe...i.cpp#L324 08:58
08:59 kurahaupo left 09:00 sivoais joined, zjmarlow left, zjmarlow joined
abraxxa looks like what I'm doing 09:00
i'll try to encode some form of OCI tracing 09:01
is it possible that signed/unsigned is a problem?
as far as I've seen NativeCall can't distinguish between the two
the error i'm getting is ORA-01024: invalid datatype in OCI call which according to the docs means that I've passed an illegal value to the datatype identifier 09:02
which is 1..7 and SQLT_FLT is 4
09:02 kurahaupo joined
nwc10 lostinjit.blogspot.co.at/2015/03/hi...fails.html -- lostinjit.blogspot.co.at/2015/03/hi...fails.html 09:02
whole post and comments are interesting.
abraxxa i've defined ub2 as int16
FROGGS no, signedness is about the value, not the type constant it complains about me thinks
moritz abraxxa: 4 is the same bit pattern in int16 and uint16 09:03
so that doens't matter here
abraxxa ok, thanks
dalek ast: 307c450 | lizmat++ | S32-num/rat.t:
Unfudge passing Rat tests
 09:05
09:06 nige joined
dalek Heuristic branch merge: pushed 45 commits to rakudo/newio by lizmat 09:07
abraxxa is there debugging/tracing available in nativecall? 09:10
09:10 Ven joined 09:11 cschwenz joined
FROGGS abraxxa: no 09:11
09:12 larion joined 09:13 Akagi201_ left 09:14 Akagi201 joined 09:25 zjmarlow left 09:26 zjmarlow joined
dalek ast: ff518a8 | lizmat++ | S02-types/WHICH.t:
Add some native WHICH testing
 09:29
moritz interesting python precedence WTF 09:30
python recycles the + operator on strings for string concatenation 09:31
nwc10 is python sufficiently typed that that works? Or does it have the same mess as JS? 09:32
moritz and the % operator with a string LHS for fmt-like string templating
nwc10: it never implicitly coerces, so it's less of a wtf
but
precedence of % is tighter than +
so if you have a longish string, and line-break it
you write that as "first half " + "second half" 09:33
and now you add templating
"%s half %s half" % ("first", "second")
splits to
"%s half " + "%s half" % ("first", "second")
and BOOM
because % is tighter than % 09:34
09:34 sivoais left, sivoais joined
moritz so you have to write 09:35
09:35 andreoss left
moritz ("%s half " + "%s half") % ("first", "second") 09:35
09:36 laouji left 09:37 dakkar joined 09:42 Akagi201 left 09:43 Akagi201 joined
abraxxa what's the correct method to check for 'not null' on a p6 var of any datatype? 09:44
09:45 kaleem left
lizmat m: say Int.DEFINITE; say 42.DEFINITE 09:46
camelia rakudo-moar af6bd0: OUTPUT«False␤True␤»
lizmat S12:1616
synopsebot Link: design.perl6.org/S12.html#line_1616
09:46 cdc joined
lizmat abraxxa: ^^^ 09:47
09:47 kaleem joined
abraxxa lizmat: thanks 09:47
the variable should allow undefined values too, i just need to check it 09:48
09:48 laouji joined
Timbus that seems awfully special, why not just `defined $var` ? 09:49
abraxxa m: my Int $v = ''; say $v.chars; say $v.DEFINITE;
camelia rakudo-moar af6bd0: OUTPUT«Type check failed in assignment to '$v'; expected 'Int' but got 'Str'␤ in block <unit> at /tmp/y0hxmWLSrU:1␤␤»
09:50 espadrine joined
moritz usually .defined works fine 09:51
no need to scream :-)
09:51 brrt left
Timbus a blank string though, is defined 09:52
dalek ast: 04af9e7 | lizmat++ | S02-types/WHICH.t:
Add Buf/Blob to WHICH type checking
 10:01
Heuristic branch merge: pushed 101 commits to roast/newio by lizmat 10:03
10:04 lolisa joined
lizmat afk& 10:05
10:07 chenryn left, azawawi left 10:08 virtualsue joined
abraxxa m: my Str $v = ''; say $v.chars; say $v.defined; 10:13
camelia rakudo-moar af6bd0: OUTPUT«0␤True␤»
10:13 lolisa left
abraxxa in case of Oracle I think I need to stay with .chars > 0 10:13
should a NativeCall sub be defined with { * } now instead of { ... }? 10:15
nwc10 moritz: (delayed) thanks
but of course, your tests will catch that :-/
10:17 zjmarlow left, zjmarlow joined
FROGGS m: my Str $v = ''; say $v.chars; say so $v 10:18
camelia rakudo-moar af6bd0: OUTPUT«0␤False␤»
FROGGS m: my Str $v = ''; say $v.chars; say ?$v 10:19
camelia rakudo-moar af6bd0: OUTPUT«0␤False␤»
FROGGS abraxxa: an empty string boolifies to false
abraxxa: so if it boolifies to true (in e.g. 'if $v {'), then it is a Str instance that has chars in it 10:20
abraxxa: and { * } vs. { ... } is up to your taste 10:21
moritz nwc10: huh, what did I do? :-) 10:22
nwc10 moritz: answered my question about + and % in python about an hour ago 10:23
moritz ah 10:24
dalek kudo/nom: bc08619 | usev6++ | src/core/operators.pm:
Do not return the previous value when generator function of sequence calls 'last'

fixes things like 'say 10,9,8, { diff - 1 || last } ... *'		 10:27
kudo/nom: 9b5d4e2 | FROGGS++ | src/core/operators.pm:
Merge pull request #393 from usev6/nom

Don't return previous value when generator function of sequence calls 'last'
10:29 kjs_ left
bartolin m: say 10,9,8, { $_ - 1 || last } ... * # s/diff/$_/ in commit message 10:30
camelia rakudo-moar af6bd0: OUTPUT«10 9 8 7 6 5 4 3 2 1 1␤»
10:31 laouji left 10:33 sqirrel_ joined 10:37 kjs_ joined
dalek ast: 311cf9d | usev6++ | S03-sequence/basic.t:
Unfudge passing tests for 'sequence + last'
 10:37
grondilu "diff"? 10:45
oh I see the s/diff/$_/ now
abraxxa FROGGS: thanks 10:46
grondilu not that it matter but I think you can use the amend git comment to change it
s/comment/command/ 10:47
abraxxa FROGGS: i guess .chars > 0 better carries the information what I tried to achieve
bartolin grondilu: that was a copy-and-pasto with github :/ 10:48
grondilu still, it should be possible to edit it with git commit --amend 10:49
and the push again
s/the/then/
bartolin grondilu: I'd have to make a PR. do you think it's worth it? 10:50
grondilu probably not
bartolin I'll at least add a comment to the commit
Ven grondilu: no, if you --amend you'll need to force-push 11:03
you shouldn't force push a repo so big
11:04 lumimies left
grondilu ok, noted 11:06
11:07 kurahaupo left
Ven m: say [].WHAT.perl 11:08
camelia rakudo-moar 9b5d4e: OUTPUT«Array␤»
11:14 sqirrel_ left 11:17 lolisa joined 11:19 rurban left
psch \o 11:26
is there a reason HLL::Grammar.integer doesn't parse negative integers? 11:27
i've been looking at #123682, and that seems the most sensible spot to fix
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=123682
11:27 zakharyas joined
psch m: say :10<-1> 11:27
camelia rakudo-moar 9b5d4e: OUTPUT«5===SORRY!5=== Error while compiling /tmp/HB848L3x9H␤Malformed radix number␤at /tmp/HB848L3x9H:1␤------> 3say :10<7⏏5-1>␤ expecting any of:␤ number in radix notation␤»
psch m: say :10<1*2**-64>
camelia rakudo-moar 9b5d4e: OUTPUT«5===SORRY!5=== Error while compiling /tmp/AvAC1jYzTC␤Malformed radix number␤at /tmp/AvAC1jYzTC:1␤------> 3say :10<1*2**7⏏5-64>␤ expecting any of:␤ number in radix notation␤»
11:33 lumimies joined 11:34 Akagi201_ joined
grondilu m:say :10(-1) 11:35
m: say :10(-1)
camelia rakudo-moar 9b5d4e: OUTPUT«You have confused the number -1 with the textual representation "-1";␤if you wanted to render the number in the given base, use $number.base($radix)␤ in block <unit> at /tmp/iwpn8jZGp1:1␤␤»
11:35 Akagi201 left
cdc Hello #perl6 11:36
grondilu that's quite a talkative error message (which is good)
cdc I'd like to make yada exceptions distinguishable from user exceptions, ex. X::Stub::Executed vs. X::AdHoc. 11:37
I tried the following change, but it doesn't work: gist.github.com/anonymous/27da128ae72198ceb182 (warning: cargo-cult programming).
Any usggestion is welcomed! 11:38
*suggestion
psch cdc: Perl6/World.nqp has a method that gives you a typed exception
cdc: src/Perl6/World.nqp:2908
cdc: using that instead of building the ast yourself should work 11:39
cdc psch: thanks, I'll give it a try.
11:43 chenryn joined 11:46 chenryn left 11:47 chenryn joined
psch m: say 0xdeadbeef * 16 ** -1 11:48
camelia rakudo-moar 9b5d4e: OUTPUT«233495534.9375␤»
psch m: say :16<deadbeef> * 16 ** -1
camelia rakudo-moar 9b5d4e: OUTPUT«233495534.9375␤»
nwc10 I'm curious about "Nuitka Python", which seems to be (yet another) Python re-implementation (and seemingly pretty cool): nuitka.net/posts/nuitka-release-0511.html 11:50
anyone know why it's being written - just fun?
because supporting everything from 2.6 to 3.4 seems pretty awesome
11:50 zjmarlow left 11:51 zjmarlow joined
bartolin I've created another PR for sequences (394). it would be great if someone could take a look at it, since I'm not absolutely sure about it 12:08
m: sub f (Int $n) { $n > 3 ?? "liftoff!" !! $n + 1 }; say 1, &f ... * 12:09
camelia rakudo-moar 9b5d4e: OUTPUT«===SORRY!===␤Type check failed in binding $n; expected 'Int' but got 'Str'␤»
bartolin as I understand S03 that should give '1 2 3 4 liftoff!'
12:10 eli-se joined
psch m: 1, { $^a.WHAT.say }, ... 6 12:10
camelia rakudo-moar 9b5d4e: OUTPUT«5===SORRY!5=== Error while compiling /tmp/uPROEWWVjI␤Comma found before apparent series operator; please remove comma (or put parens␤ around the ... listop, or use 'fail' instead of ...)␤at /tmp/uPROEWWVjI:1␤------> 031, { $^a.WHAT.say },7…»
psch m: 1, { $^a.WHAT.say } ... 6 12:11
eli-se good day
psch o/ eli-se
camelia rakudo-moar 9b5d4e: OUTPUT«(timeout)(Int)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool…»
eli-se how does resource cleanup work in Perl 6?
e.g. closing files. is there FINALLY?
psch m: 1, 2, { $^a.WHAT.say } ... 6
ah meh
eli-se: there's LEAVE and UNDO 12:12
eli-se nice
camelia rakudo-moar 9b5d4e: OUTPUT«(timeout)(Int)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool)␤(Bool…»
eli-se why not refcounting like in Perl 5?
psch eli-se: i think because backends are free to do their GC stuff however they want
s/backends/implementations/
eli-se: consider that the test suite is the specification, and testing for GC behavior might not belong into the language 12:13
cdc eli-se: for MoarVM -> jnthn.net/papers/2013-bs-secret-life-of-gc.pdf 12:16
eli-se psch: I mean from a design PoV, since refcounting is typically deterministic whereas tracing GCs typically aren't
psch eli-se: i don't see how that change my argument. Perl 6 the language doesn't dictate GC behavior 12:17
eli-se I'm asking why
psch i don't know, but i think it's because the language isn't supposed to be tied to an implementation 12:18
which is what some of Perl 5's problems came from, i.e. "if perl parses it, it's Perl" 12:19
12:19 zjmarlow_ joined
cdc psch: $*W.typed_exception($/, ['X', 'Stub', 'Executed']) complains during the parse stage: "Could not locate compile-time value for symbol X::Stub::Executed" 12:20
(this does not work either with ['X', 'Comp'], for instance)
I don't understand why
psch eli-se: TimToady++ probably can go into more detail on the "why is GC behavior not part of the language", while jnthn++ will likely be able to tell you why moar does what it does
eli-se ok 12:21
arnsholt The problem with refcounting is that it fails horribly when there are cycles
psch cdc: can you update your gist? 12:22
12:22 eli-se left 12:23 zjmarlow left
arnsholt And if you mandate timely/predictable collection, you're basically requiring some kind of refcounting 12:23
12:25 chenryn left 12:27 ggoebel left 12:29 kjs_ left 12:32 chenryn joined
cdc psch: gist.github.com/anonymous/2d5160e65ef9c12a5cd2 (I wasn't able to update the previous gist) 12:33
12:35 Akagi201_ left 12:36 Akagi201 joined 12:38 Akagi201_ joined 12:40 gfldex joined 12:42 Akagi201 left
psch cdc: actually, that's interesting 12:46
cdc: we're stubbing things right at the start of CORE.setting
cdc: and during compilation of that, your patch builds an X::Stub::Executed for every stub it encounters 12:47
cdc: but the symbol doesn't exist yet, and you can't stub it either, because circularity
12:48 Akagi201_ left, Akagi201 joined
psch cdc: i think it could work if you stub X::Stub::Executed in src/Metamodel/BOOTSTRAP.nqp, but i'm not sure it belongs there 12:50
cdc psch: thanks for your help! 12:51
psch: I should ask to moritz if it belongs there?
12:52 Akagi201_ joined
moritz if what belongs where? 12:53
12:53 kjs_ joined
moritz no, it doesn't belong there 12:53
you can emit a symbolic lookup 12:54
and it'll exist by the time you throw such an error
12:54 chenryn left 12:55 sirsleepalot left
cdc moritz: could you show me some examples of symbolic lookups, please? 12:55
I never wrote anything in NQP before. 12:56
12:56 Akagi201 left 12:57 colomon left
psch oh 12:59
yeah, the Actions code is wrong 13:00
i think :P
$*W.find_symbol looks during compile time, but make_yada should emit something that looks during runtime 13:01
cdc: ^^^
cdc psch: ok, how I can defer this action? 13:02
psch cdc: $*W.symbol_lookup, defined at Perl6/World.nqp:2666 13:03
cdc thanks, I'm looking at it
psch cdc: the bit you had before with :op<call> should take the result of symbol_lookup as argument i think 13:04
13:04 telex left
cdc ok, I give it a try 13:04
13:06 telex joined 13:10 nige left
moritz for lexicals you can also do a QAST:Var.new(:name<...>) 13:10
but I fear that doesn't work with multi-joined class names
13:11 rurban joined
Ven .u        13:12
yoleaux U+0020 SPACE [Zs] ( )
U+2007 FIGURE SPACE [Zs] ( )
13:13 Akagi201_ left, PerlJam joined, Akagi201 joined 13:14 Akagi201_ joined
dalek kudo/no-linenoise: f04bcb6 | hoelzro++ | tools/build/NQP_REVISION:
DON'T MERGE ME Bump NQP revision
 13:15
13:15 Akagi201 left
dalek p/no-readlineint: 5cc28f9 | hoelzro++ | tools/build/MOAR_REVISION:
DON'T MERGE ME Bump MoarVM version
 13:16
kudo/tab-completion: 7418f0a | hoelzro++ | / (2 files):
Add rudimentary tab completion
kudo/tab-completion: 886d00b | hoelzro++ | / (2 files):
Decouple completion from linenoise
13:16 zakharyas left 13:21 zakharyas joined 13:23 coffee` left, sivoais left, lolisa left 13:24 Akagi201_ left, sivoais joined, Akagi201 joined 13:25 Akagi201_ joined, fhelmberger left 13:29 Akagi201 left 13:31 ibo2perl joined
ibo2perl m: my $x; say $x; $x = 1/0; say "alive and " ~ ? $x; say $x 13:32
camelia rakudo-moar 9b5d4e: OUTPUT«(Any)␤alive and True␤Divide by zero␤ in method Numeric at src/gen/m-CORE.setting:15820␤ in sub infix:<-> at src/gen/m-CORE.setting:5279␤ in method Str at src/gen/m-CORE.setting:13115␤ in method gist at src/gen/m-CORE.setting:5116␤ in sub…»
ibo2perl it is not clear to me why the ? $x evaluates to true 13:33
psch m: say ?1; say ?0 13:35
camelia rakudo-moar 9b5d4e: OUTPUT«True␤False␤»
13:35 sqirrel_ joined
psch ibo2perl: prefix:<?> is prefix:<so> with looser precedence 13:35
ibo2perl: it's the opposite of prefix:<!>, like prefix:<not> for prefix:<so> 13:36
ibo2perl psch: yes, sure. But $x has 1/0 an exception
psch m: say so 1/0
camelia rakudo-moar 9b5d4e: OUTPUT«True␤»
psch m: say so X::DivideByZero.new
camelia rakudo-moar 9b5d4e: OUTPUT«Could not find symbol '&DivideByZero'␤ in method <anon> at src/gen/m-CORE.setting:15826␤ in any find_method_fallback at src/gen/m-Metamodel.nqp:2908␤ in any find_method at src/gen/m-Metamodel.nqp:1052␤ in block <unit> at /tmp/ymmgZb_AwB:1␤␤»
psch m: say so X::Numeric::DivideByZero.new 13:37
camelia rakudo-moar 9b5d4e: OUTPUT«True␤»
psch m: say so die "foo"
camelia rakudo-moar 9b5d4e: OUTPUT«foo␤ in block <unit> at /tmp/XL9ibnNhSB:1␤␤»
psch m: say so fail "foo"
camelia rakudo-moar 9b5d4e: OUTPUT«Unhandled exception: foo␤ at <unknown>:1 (/home/camelia/rakudo-inst-1/share/perl6/runtime/CORE.setting.moarvm:throw:4294967295)␤ from src/gen/m-CORE.setting:15830 (/home/camelia/rakudo-inst-1/share/perl6/runtime/CORE.setting.moarvm:sink:29)␤ from…»
psch i see
ibo2perl: i missed that, sorry
13:38 colomon joined
ibo2perl psch: np, may be it is like schrödingers cat. without looking you dont know whether it's alive or not 13:39
psch ibo2perl: Mu.Bool calls self.defined, and Exceptions are marked as handled with a call to defined 13:40
ibo2perl: it's definitely surprising though, and might want some kind of rethinking
maybe Exception.Bool shouldn't call defined, but throw again, i'm not sure
actually, no, that'd would likely be horrible 13:41
13:43 ibo2perl left 13:46 ibo2perl joined
ibo2perl m: say (1/0).WHAT 13:46
camelia rakudo-moar 9b5d4e: OUTPUT«(Rat)␤»
ibo2perl supporting my like schrödingers cat theory 13:47
psch schrödinger's Rat, surely :)
13:48 ibo2perl left 13:49 nige joined, rurban left 13:52 kaleem left 13:55 xfix joined 13:57 coffee` joined 13:59 tinyblak left 14:11 skids joined 14:13 sqirrel_ left 14:14 kaare_ left 14:15 C0rruptor joined 14:25 tinyblak joined
masak m: say 1/0 > 5 14:41
camelia rakudo-moar 9b5d4e: OUTPUT«True␤»
masak m: say 1/0 < 5
camelia rakudo-moar 9b5d4e: OUTPUT«False␤»
masak m: say 1/0 == Inf
camelia rakudo-moar 9b5d4e: OUTPUT«True␤»
masak m: say Inf + 7 == Inf 14:42
camelia rakudo-moar 9b5d4e: OUTPUT«True␤»
14:45 zakharyas left 14:46 Akagi201_ left, Akagi201 joined
hoelzro o/ #perl6 14:47
masak \o 14:48
Tux__ m: sub foo (Any :$x){$x.WHAT.say;if ($x.WHAT eq "Array"){my@x=$x;@x[0].WHAT.say;}};foo(x=>[[1,2][3,4]]); 14:52
camelia rakudo-moar 9b5d4e: OUTPUT«5===SORRY!5=== Error while compiling /tmp/pIvmvXOCKv␤Missing block␤at /tmp/pIvmvXOCKv:1␤------> 3HAT eq "Array"){my@x=$x;@x[0].WHAT.say;}7⏏5};foo(x=>[[1,2][3,4]]);␤ expecting any of:␤ parameterized block␤»
14:52 Tux__ is now known as |Tux|
masak |Tux|: .WHAT gives back a type object 14:53
dunno if type objects enjoy being string-comapred
|Tux| I know most of you would not use parens there
masak I don't care so much about that
that's just superstitious, it's up to you ;)
|Tux| masak: HAT vs WHAT
hoelzro one of the next steps for my tab completion branch is going to be how Linenoise.pm is to be distributed; I was thinking that it belongs out of core, but in Star
any thoughts on that?
masak |Tux|: that's just cutoff 14:54
14:54 Ven left
|Tux| ah, so I am being misguided there. will havbe to dig deeper 14:54
Chasing (and thus simplifying) this message: "Cannot look up attributes in a type object" 14:55
masak, what is a "better" way to check WHAT? WHICH? 14:56
masak m: my @a; say @a.WHAT === Array
camelia rakudo-moar 9b5d4e: OUTPUT«True␤»
|Tux| of course!
masak m: my @a; say @a ~~ Array
camelia rakudo-moar 9b5d4e: OUTPUT«True␤»
|Tux| still has to get used to === 14:57
masak both are better, IMO
depends if you tolerate subtypes of Array
~~ does
|Tux| much better now. thanks (again)
masak my pleasure :) 14:58
14:59 Ven joined 15:01 kaleem joined 15:06 aborazmeh joined, aborazmeh left, aborazmeh joined 15:13 rindolf left 15:14 eli-se joined 15:17 eli-se left 15:19 rmgk_ joined, rmgk left, rmgk_ is now known as rmgk 15:22 kaleem left 15:23 kaleem joined 15:26 rurban joined, C0rruptor left 15:31 C0rruptor joined 15:35 risou is now known as risou_awy 15:39 risou_awy is now known as risou 15:40 kaleem left
TimToady .bothealthymeal 15:41
moritz wonders whether that begins with 'bot' or 'both' :-) 15:42
Ven .botsnack 15:43
yoleaux :D
Ven he prefers that :p
15:46 vendethiel left 15:48 colomon left, vendethiel joined 15:49 Akagi201_ joined 15:50 aborazmeh left 15:51 Akagi201 left 15:54 fhelmberger joined
raydiak g'mornin #perl6 15:55
masak g'raydiak, mornin
TimToady o/
15:55 Woodi joined 15:57 timotimo is now known as p6weekly
TimToady
.oO(G'raydiak: fantasy town in Alaska that never has good weather.)
 15:57
p6weekly .botsnack
yoleaux :D
21 Mar 2015 21:17Z <raiph> p6weekly: Tux++ RFC7111 finished irclog.perlgeek.de/perl6/2015-03-20#i_10312704
15:57 p6weekly is now known as timotimo
moritz also, I updated en.wikipedia.org/wiki/Perl_6 15:58
|Tux| :)
15:59 fhelmberger left
raydiak looks out the window and sees...dreary weather 16:00
moritz (removed some out-of-date mentions of parrot, and made it clearer which compilers aren't actively developed anymore)
16:00 aborazmeh joined, aborazmeh left, aborazmeh joined
masak moritz++ 16:00
decommute &
moritz also en.wikipedia.org/wiki/Rakudo_Perl_6 16:01
timotimo moritz: you left "The Synopses (The official language specification)." intact?
moritz timotimo: there's still lots to clean up 16:02
timotimo fair enough :)
jnthn yawns 16:05
moritz 17:05: jnthn slowly stands up, makes his morning coffee :-)
16:05 donaldh joined
iv Hm, should the release link on the wikipedia page link to rakudo.org/2015/03/21/announce-raku...e-2015-03/ instead of the tar.gz listing? 16:07
jnthn moritz: This is yawn from having completed the day's work :) 16:08
TimToady yawns thinking about about starting the day's work...
16:11 fhelmberger joined 16:14 dombo joined 16:17 Vlavv_ left 16:29 diana_olhovik left 16:30 Vlavv_ joined 16:32 Ven left, colomon joined 16:33 sirsleepalot joined, dombo left
timotimo hmm 16:33
16:35 Perl6_newbee joined, fhelmberger left 16:39 donaldh left, rurban left
|Tux| does "given" when still exist in perl6 (it cannot be found on doc.perl6.org) or does a new syntax exist for swutch (which also shows no hits) 16:39
switch
16:40 Ven joined
masak |Tux|: heh. 16:40
16:40 molaf joined
masak |Tux|: given/when comes from Perl 6 ;) 16:40
m: given 42 { when Int { say "it's an Int! yaaay!" } }
camelia rakudo-moar 9b5d4e: OUTPUT«it's an Int! yaaay!␤»
|Tux| but you can change your minds! it happened before :P
masak troo. 16:41
in Perl 6, given/when doesn't suck, because smartmatching doesn't suck, because Perl 6 has types in a much deeper way than Perl 5.
|Tux| so the "bug" is that I cannot use "given" in the search box on doc.perl6.org
16:41 fhelmberger joined
masak yeah. 16:41
|Tux| does «when Routine» also match Sub ? 16:45
yes, I can try
yes it does 16:46
Ulti given '1.0' { when 1 {say "smart match";}; when 1.0 {say "doesn't suck";} }
yoleaux 03:02Z <japhb> Ulti: OK, I replied to your perl6-bench pull request.
Ulti m: given '1.0' { when 1 {say "smart match";}; when 1.0 {say "doesn't suck";} }
camelia rakudo-moar 9b5d4e: OUTPUT«smart match␤»
Ulti even
Ven would expect none to match, probably 16:47
it does try to be smart, tho..
TimToady smartmatching is all about dwimmery, and trying to put most of the dwimmery in one spot 16:48
timotimo m: given '1.0' { when 1 {say "smart match"; proceed;}; when 1.0 {say "doesn't suck";} }
camelia rakudo-moar 9b5d4e: OUTPUT«smart match␤doesn't suck␤»
TimToady there you go
timotimo the type of matcher deciding how a match should be done makes a whole lot of sense to me in understanding how things go in smart matching 16:49
Ulti oh so given has break by default I didn't even realise that :S
TimToady more than that, relying primarily on the one type means such switches are optimizable 16:50
why we had to break the symmetrical smartmatch good and hard
Ulti wait so why does that work and my example a few days ago didnt work
TimToady link?
Ulti using first-index
16:50 sirsleepalot is now known as houseboat
retupmoca timotimo: can you look at irclog.perlgeek.de/perl6/2015-03-23#i_10323954 ? Let me know if it looks like it makes sense or if it isn't worth it? 16:52
jnthn: or you; since you're on windows? 16:53
Ulti m: given 1.0 { when 1 {say "smart match"; proceed;}; when '1.0' {say "doesn't suck";} }
camelia rakudo-moar 9b5d4e: OUTPUT«smart match␤»
Ulti there we go
it's that the match isn't commutative 16:54
16:54 Ven left
Ulti so its even less predictable 16:54
TimToady wrong
Ulti yeah I know why
TimToady it's more predictable
Ulti its that it isnt even doing what I think
16:54 fhelmberger_ joined
TimToady it's asking a string whether it matches something that can be coerced to a string 16:55
but string coercion of a Rat doesn't not automatically add .0
masak Henrik van Oers going "What part of 'private' did you mis[s]?" on p6u felt like a very non-#perl6 way to hit someone over the head with new juicy knowledge. :)
Ulti TimToady which is fine because I could have written an arbitrary number of .0000s
16:56 fhelmberger left
masak I like how most of #perl6's attitude is more "you're one of today's lucky 10,000" -- xkcd.com/1053/ 16:56
Ulti the issue is I hadn't properly learnt smart match
timotimo masak: i thought these mails were a bit unfriendly in their tone
masak m: given '1.0' { say "smart match" when 1; say "doesn't suck" when 1.0 } # also works, because (in Perl 6) statement_mod when doesn't break/succeed 16:57
camelia rakudo-moar 9b5d4e: OUTPUT«smart match␤doesn't suck␤» 16:58
masak timotimo: oh, so it wasn't just me
timotimo: I was wondering whether to call it out, but I didn't want to come across as hyper-sensitive
16:59 Rounin left
Ulti masak that would have held if I hadn't turned 30 ;___; or is the age of Perl6 mastery about 70 16:59
16:59 vendethiel left
masak Ulti: well, Perl 6 *is* meant to be a last language... 16:59
16:59 larion left
Ulti I doubt any of you have used my slang so you don't know all of Perl 6 :P 17:00
which is going to be a fun future
I can just see Perl 6 running into the same sorts of problems human languages have 17:01
I guess the spec is like a dictionary
masak Ulti: back in 2010, I tweeted about being a Perl 6 user for five years already. TimToady tweeted back, saying he'd been using Perl 6 for twice as long. :P
Ulti anything else is just pidgin perl 6
masak & 17:02
17:02 vendethiel joined
Ulti I've been doing a bit of Perl 6 since the first R* so 5-6 years? 17:03
the problem is it's easy to go wide in Perl 6 rather than deep in your learning
unless you are actually helping with the implementation I guess 17:05
|Tux| FROGGS, HELP! gist.github.com/Tux/90b22dd241ab896cc392
TimToady we need both approaches
17:06 diana_olhovik joined
jdv79 isn't that called a "T shaped dev" or something these days? 17:06
TimToady thinks that leaves out a lot of hypotenuses 17:07
▼ is more like 17:09
we also need people who go half-down and half-wide :)
such as Yours Truly
|Tux| FROGGS, ignore. I'll use parens on given 17:13
FROGGS |Tux|: I'll look into it in a few hou... okay 17:14
|Tux| it is your last reverted change (reserved words like if, sub, method, and given)
somehow for does not need the parens 17:15
FROGGS gtg now, might look into it later
17:15 FROGGS left
|Tux| thanks in advance 17:15
17:15 sqirrel_ joined
TimToady m: say .so ?? ｢\o/｣ !! ｢/o\｣ for <1/3> ~~ Rat, <+1-2i> ~~ Complex 17:18
camelia rakudo-moar 9b5d4e: OUTPUT«\o/␤/o\␤»
17:19 fhelmberger_ left
TimToady m: my $state = ｢\o/｣; $state .= flip for (^10).pick; say $state; 17:20
camelia rakudo-moar 9b5d4e: OUTPUT«/o\␤»
17:20 fhelmberger joined
TimToady m: my $state = ｢\o/｣; $state .= flip xx 10.rand; say $state; 17:21
camelia rakudo-moar 9b5d4e: OUTPUT«===SORRY!===␤Method call node requires at least one child␤»
TimToady m: my $state = ｢\o/｣; $state.=flip xx 10.rand; say $state;
camelia rakudo-moar 9b5d4e: OUTPUT«/o\␤»
TimToady m: my $state = ｢\o/｣; $state.=flip xx 10.rand; say $state;
camelia rakudo-moar 9b5d4e: OUTPUT«\o/␤»
17:23 abraxxa left, vendethiel left 17:24 mr-foobar joined
iv Hm, why doesn't `42.WHAT` print anything on the REPL, unlike `say 42.WHAT`? 17:25
17:25 vendethiel joined
iv Where `` are not part of the code… :P 17:27
TimToady probably trying not to report a result from 'my $x;' and such
iv Ah, okay. 17:29
Maybe un-perlish to report a result (Any) from 'my $x'?
Asking because my expectation for a REPL is that it tries to print whatever I give it 17:34
learned from Other Languages
17:35 eli-se joined 17:36 muraiki joined
timotimo be cautious, our REPL is very bad at a few things, especially involving the := operator :( 17:36
eli-se hola 17:39
cola
timotimo çola?
eli-se Cool 17:40
|Tux| m: sub foo(Any :$in){$in.list[0] ~="Boo";};my Str $s="";$s.say;foo(in=>\$s);$s.say;
camelia rakudo-moar 9b5d4e: OUTPUT«␤Boo␤»
|Tux| NICE!
is this to stay? can I rely on that?
17:42 zjmarlow joined
timotimo m: sub foo(Any :$in is rw) { $in ~= "Boo";}; my Str $s = ""; $s.say; foo(in => $s); $s.say; 17:44
camelia rakudo-moar 9b5d4e: OUTPUT«5===SORRY!5=== Error while compiling /tmp/V5OKvXpRRe␤Cannot use 'is rw' on an optional parameter␤at /tmp/V5OKvXpRRe:1␤------> ␤»
timotimo ah, that's why you did that 17:45
17:45 zjmarlow_ left
|Tux| not really :) 17:45
17:45 colomon left
timotimo no? 17:47
|Tux| have a sec. will give you a link in a few secs
TimToady m: sub foo(Any :$in! is rw) { $in ~= "Boo";}; my Str $s = ""; $s.say; foo(in => $s); $s.say;
camelia rakudo-moar 9b5d4e: OUTPUT«␤Boo␤»
timotimo ah, much cleaner
|Tux| github.com/Tux/CSV/blob/master/test-t.pl#L1194 17:49
17:51 grondilu left 17:52 FROGGS joined, dakkar left 17:53 virtualsue left
timotimo that code doesn't speak for itself, at least not to me 17:54
|Tux| in => "file.csv" should open that file and parse the data in it, in => \"some data" should parse the string itself 17:55
in => $io parses from an already opened IO etc etc 17:56
timotimo yeah
i got that much
why would you rely on mutating arguments passed in via named parameters for that? 17:57
TimToady that's P5-think fighting P6-think, I think; why not just multi-dispatch to IO vs Str?
|Tux| Ah, that comes inportant for out
TimToady, because that would explode 17:58
out => "file.csv" write data to the file, out => \$s appends data to the string
TimToady, in can alse be Array, in which case it can be an Array of arrays or an array of hashes 17:59
used for out
TimToady so why not out => "file.csv".IO vs out => $s vs out => @array
18:00 C0rruptor left
TimToady IO is short for a reason 18:00
|Tux| I will ponder on that. Thanks
TimToady Capture is long for a similar reason :)
timotimo retupmoca: bundling the .dll files in the git repository feels a bit icky to me; i'd prefer not having such binary files in there. on the other hand, making dll files easily attainable during the build phase of gtk::simple would be nice to have for windows users 18:01
|Tux| out => $s ... can that be said to *append* instead of just out?
timotimo we can re-host the dll files on www.p6c.org and just download them via curl or HTTP::Client or what it's called
|Tux| or just have Any :$out is rw (which doesn't works as just shown) 18:02
timotimo Any :$out! is rw does work 18:03
TimToady showed a working example of that a few lines above
|Tux| but out is optional (defaults to $*OUT)
timotimo oh
when you said "multi dispatch would explode" you meant a combinatoric explosion? 18:04
|Tux| yes
18:05 kjs_ left
timotimo you don't have to have a candidate for every single combination, though 18:06
TimToady m: sub foo(\x) { say x.WHAT }; foo $_; foo VAR($_)
camelia rakudo-moar 9b5d4e: OUTPUT«Nil␤(Scalar)␤»
TimToady there's another approach
timotimo ah, pass in the VAR of something if you want it to be changed 18:07
TimToady out => VAR($foo) is even kinda self-documenting
timotimo that's very close to what tux was doing with the $out.list[0] ~= ... trick
18:07 avuserow joined
|Tux| yes, but that puts the burdon on the caller instead of on me :) 18:07
18:07 sqirrel_ left
TimToady you were already putting the burden there with \ 18:08
timotimo imagine the fun when you pass in a variable that has a where clause attached to its type %)
|Tux| and you are right: I still have to make the gap wider between perl5 think and perl6 think
but I'm just doing perl6 for 50 days or so and I think I got way further than imagined at the start 18:10
18:10 aborazmeh left
timotimo that's good to hear :) 18:11
18:11 fhelmberger left
|Tux| anyway, sorry to have stolen some time again. Now off to play badminton 18:12
avuserow moritz: in Python, what you're "supposed" to do rather than `("a %d" + "b %d") % (1,2)` is use `"a %d" "b %d" % (1, 2)` (that is, no operator between string literals makes them concat with higher precedence, assuming you are only dealing with literals) 18:13
avuserow is glad that Perl 6 has no such implicit "operator" 18:14
TimToady it's a low-leverage feature with many negative consequences
18:15 espadrine left
TimToady funny how languages cargo-cult these low-leverage features from other languages. concat with + also comes to mind 18:15
18:18 mj41 joined
masak I don't think there's a good understanding of the scope of the negative consequences most of the time. 18:20
eli-se Woo, fixed six issues in a single commit. 18:26
masak ok, I wrote an answer to the "Object Introspection" thread. I am literally unable to judge whether it'd be a net good to post it.
therefore I'm posting it here first for general comment. gist.githubusercontent.com/masak/a.../reply.txt
eli-se: o.O 18:27
eli-se: you should come work on the RT queue :>
eli-se what is the RT queue?
masak all the rakudobugs.
eli-se Are retweets processed to slowly?
Oh, I see.
I'm bad at C.
masak ok, I'm going for dinner now. feel free to leave comments here or on the gist. gist.github.com/masak/a788a742b70c47e46565 18:28
eli-se I'll probably introduce buffer overflows and memory leaks.
masak eli-se: Rakudog is mostly Perl 6, actually.
Rakudo*
&
eli-se oooot
arnsholt Rakudog =D 18:30
18:30 colomon joined
tadzik <3 18:31
arnsholt eli-se: There used to be a bit more C for compiler hackery, but even then you had to dig pretty deep before you ran into that
Heck, on the JVM backend there's no C 18:32
[Coke] masak: +1; last paragraph is perhaps slightly too direct. Perhaps instead ask him to keep it in mind when he is representing the community.
arnsholt You should probably know a decent amount of C if you're gonna work on the NativeCall bits, even on JVM, but outside of that there's no C
mj41 timotimo lizmat: hi. perl6-star 2015.03 docker image available registry.hub.docker.com/u/mj41/perl6-star/ 18:33
~ $ docker run -i -t --rm --name my-perl6-star-2015.03 mj41/perl6-star:2015.03 perl6 -e'say "{$*PERL.compiler.gist} on {$*VM.gist} for {$*DISTRO.gist}"'
rakudo (2015.3) on moar (2015.3) for fedora (21.Twenty.One)
~ $
18:33 _edwin joined
eli-se Alright, time to start working on my JIT compiler. 18:35
18:36 nige left
_edwin hello perl6 hackers! I have a problem with Rakudo Star 2015.2 on Windows: "AltGr" combinations on my (German) keyboard are not working in the REPL. Is this a known problem? 18:38
18:40 Hor|zon left
vendethiel _edwin: accents refusing to print is a known problem :) 18:40
_edwin vendethiel, typing the accents at least works for me, I have the more basic problem that I cannot type {, [, etc. in the REPL :/ 18:43
jdv79 nine_: around? Inline::Perl6 failed to build for e.
vendethiel ah yes, it was brackets not accents
jdv79 *me
vendethiel _edwin: can you type accents and submit, tho?
it should be crashing
_edwin I can type 'say "á"' for example, but executing it crashes with "Malformed UTF-8" 18:44
vendethiel okay, so that's the known bug indeed. 18:45
_edwin but the AltGr + * combinations are just completely dead, they do not even produce a wrong character 18:46
vendethiel yep :)
18:49 kjs_ joined
_edwin does it make sense to submit this problem somewhere -- I did'nt find it in rt.perl.org or the rakudo/star github issues 18:49
vendethiel i believe it's in RT 18:57
18:58 Perl6_newbee left 19:00 telex left 19:02 telex joined 19:04 Bucciarati left, haroldwu left, ingy left, Bucciarati joined, Celelibi left, captain-adequate left, petercommand left, lestrrat left, betterwo1ld left, sftp left, haroldwu joined, ingy joined, betterworld joined, petercommand joined, captain-adequate joined, sftp joined, Celelibi joined 19:05 virtualsue joined 19:07 lestrrat joined 19:09 rindolf joined 19:13 kjs_ left, raiph joined 19:23 kaare_ joined 19:33 abraxxa joined 19:36 abraxxa left 19:37 Hor|zon joined 19:39 abraxxa joined 19:42 Hor|zon left, bayprogrammer joined 19:49 zjmarlow_ joined, Hor|zon joined 19:51 zjmarlow left 19:58 danaj left
lizmat mj41: I believe timotimo will be doing the P6W this week 19:59
mj41: thanks for reminding nonetheless :-)
19:59 bayprogrammer left, dombo joined 20:01 raiph left 20:02 zjmarlow_ left, zjmarlow_ joined 20:03 kjs_ joined 20:04 kjs_ left, molaf left
hoelzro mj41: I didn't know someone else had made a docker image for R* 20:04
20:05 darutoko left 20:06 larion joined
dalek ar: 3df9a6f | FROGGS++ | tools/build/Makefile.in:
install panda's state and projects list

We do that so one can query what is installed by doing: panda --installed list This also prevents panda to refetch and rebuild already installed stuff, which can even break because the module in the ecosystem might already have changed, triggered by a rakudo compiler change.		 20:07
ar: da6cb95 | FROGGS++ | tools/build/bin-install.pl:
create windows .bat files for each script in bin/
nine_ jdv79: I'm here 20:08
20:09 spider-mario joined
mj41 hoelzro: yes, I saw you image, but I prefer automated build 20:10
hoelzro mj41: I changed mine to an automated build recently
cut down on my headaches a bit =) 20:11
mj41 :-) ... any chance to merge them?
jdv79 nine_: paste.scsys.co.uk/470160 20:12
hoelzro mj41: sure! afaict, they're the same instead of distribution 20:13
your README is much better, though!
nine_ jdv79: is your perl compiled with multiplicity?
mj41 or maybe it's time to create official one under github.com/perl6
nine_ jdv79: oh, I just noticed that I did fix building on a perl without multiplicity, but have not yet done a release containing the fix. 20:14
hoelzro mj41: good idea
TimToady "module in the ecosystem might already have changed" seems to violate the immutability principle of official modules, though I suppose what's really going on is loose versioning
jdv79 usemultiplicity=undef - i guess not
hoelzro I'm not sure where packaging stuff is stored atm
probably in rakudo/star?
mj41 hoelzro: btw you are on collaborators list from the start github.com/mj41/docker-perl6-star/...laboration
jdv79 i just installed the way the Inline::Perl5 docs said 20:15
hoelzro oh, cool
20:15 kjs_ joined
jdv79 perlbrew install 5.20.2 -Duseshrplib 20:15
TimToady and, of course, immutable modules will depend somewhat on a stable VM, which we don't quite have yet 20:16
hoelzro mj41: we should create an official perl6/rakudo repo on Docker hub, and push images there
mj41 hoelzro: there is a issue with automated build ... old images (tags) are rebuilded every time I add new tag :-(
hoelzro I see 20:17
jdv79 oops - i could read a few lines farther - i'll try that. thanks nine_.
hoelzro I just have the Dockerfile, update it post-release, and docker build
vendethiel .u star 20:19
yoleaux U+0001 START OF HEADING [Cc] (␁)
U+0002 START OF TEXT [Cc] (␂)
U+0086 START OF SELECTED AREA [Cc] (<control>)
mj41 hobbs: our target should be official repo docs.docker.com/docker-hub/official_repos/
20:19 anaeem1 joined
hoelzro heh, I was just looking at that =) 20:19
20:19 bjz left
mj41 s/hobbs/hoelzro/ 20:19
vendethiel .u crying cat 20:22
yoleaux U+1F63F CRYING CAT FACE [So] (😿)
20:27 larion left 20:28 kurahaupo1 joined
mj41 hoelzro: maybe workaround is to use separate directory/dockerfile for each version/tag? github.com/docker-library/postgres 20:28
20:29 larion joined
hoelzro ah, interesting 20:29
mj41: which image should we use as the base?
mj41 hoelzro: not sure, I prefer Fedora :-) 20:30
hoelzro I figured ;)
FROGGS _edwin: I can confirm that AltGr+[ is a noop :o(
hoelzro I picked Ubuntu because it seems to be pretty standard for Docker images
dalek c: 87bdef4 | moritz++ | lib/Type/Pair.pod:
Pair.antipair
 20:33
20:33 cschwenz left
dalek line-Perl6: 8149f4d | (Stefan Seifert)++ | Perl6.xs:
Be compatible with rakudo 2015.03

Rakudo installs in install/share instead of install/languages since commit 7920c16b8d81f0feaf16ed5ce9620c26ca67e63e Date: Mon Mar 16 08:44:19 2015 +0100		 20:34
line-Perl6: 0d16f8e | (Stefan Seifert)++ | / (2 files):
Version 0.04
20:34 _edwin left
nine_ jdv79: I just uploaded Inline::Perl6 version 0.04. This contains the fix for perl without multiplicity and for compatibility with Rakduo 2015.03 20:35
jdv79 cool, i'll try it in a bit.
moritz nine_++ 20:36
20:37 larion left
moritz TimToady: I can't find any description in the design docs about what .antipairs is supposed to do. The only hint seems to be in the section about invert: "Produces a backmapping of values to keys, but unlike C<.antipairs>, expands list values 20:38
into multiple pairs."
but that seems pretty obscure to me
20:39 larion joined, nige joined
mj41 hoelzro: I'm not familiar with Ubuntu, so feel free to switch Dockerfile in 'develop' branch to Ubuntu/Debian if you think it's good idea :-) 20:39
hoelzro ok 20:40
mj41: do you mind if I copy your README for the docker repo on GH? 20:41
jdv79 nine_: seems good to go
20:42 bayprogrammer joined 20:43 pdcawley left
nine_ jdv79: yeah :) 20:43
20:44 zacts joined 20:45 larion left, larion joined
mj41 hoelzro: IMHO it's going to mystify users. I prefer to start new GitHub repo. We can transfer it under perl6 project/namespace later. 20:46
hoelzro mj41: I'll create a new GH repo later tonight
20:47 [Sno] left
mj41 hoelzro: ok, thx, good luck. I should get some sleep. See you/backlog tomorrow. 20:48
hoelzro mj41: thanks, good night! 20:49
20:50 mj41 left 20:52 [Sno] joined
FROGGS rakudo star 2015.03 for windows: rakudo.org/downloads/star/ 20:53
20:53 raiph joined
masak [Coke]: thank you. exactly the kind of -- literally -- moderation I was looking for. 20:54
[Coke]: is this last paragraph better? gist.github.com/masak/a788a742b70c47e46565
eli-se wooo my JIT compiler works! 20:57
nwc10 FROGGS++ 20:58
vendethiel FROGGS++!
eli-se: gg!
nine_ Ok, first Ok, first attempt to install Ubuntu in a VM failed...
jdv79 nine_: so i have 5.20.1 without multiplcity and your 0.04 seems to work. but i just installed 5.20.2 with multiplicity and your 0.04 seems to not: paste.scsys.co.uk/470161 20:59
nine_ jdv79: did you install Inline::Perl5 for this new perl with multiplicity? You have to use the same perl for both Inline::Perl5 and Inline::Perl6
jdv79 different perls 21:00
that's interesting
i didn't see that in the docs but i probably missed it 21:01
21:01 Peter_R left
nine_ jdv79: it's probably missing in the docs :/ 21:01
21:02 skids left
dalek line-Perl6: d31cf6c | (Stefan Seifert)++ | lib/Inline/Perl6.pm:
Document the need for using the same perl for Inline::Perl5 and 6
 21:03
nine_ jdv79: corrected that
jdv79 thanks! rebuilding rakudo takes a long time for me so i'll stick with the older p5 for now. 21:05
21:05 yqt joined
nine_ jdv79: you don't have to re-install rakudo when switching perl 5 versions. Rakudo needs perl 5 only during the build. But you have to re-install Inline::Perl5. 21:06
21:06 abraxxa left
jdv79 oh, duh. that makes more sense. 21:07
21:08 larion left 21:10 larion joined
eli-se vendethiel: github.com/mill-lang/mill/blob/fea...rc/jit.hpp :D 21:11
jdv79 and that worked as advertised. very nice:)
21:11 rindolf left 21:14 dombo left 21:15 larion left 21:16 larion joined
[Coke] masak: thanks, that seems nicer. 21:18
21:19 virtualsue left
masak goes ahead and posts 21:19
[Coke]++ # feedback 21:20
PerlJam masak++
dalek kudo-star-daily: 49ba47c | coke++ | log/ (9 files):
today (automated commit)
rl6-roast-data: 46202f5 | coke++ | / (9 files):
today (automated commit)
[Coke] heads hoe 21:23
*home
21:23 Rounin joined
vendethiel eli-se: I read it :P 21:24
eli-se woot
21:25 anaeem1 left, anaeem1 joined
retupmoca modules.perl6.org/ looks a little sparse from here 21:27
timotimo so it does :\
21:28 ggoebel joined 21:29 anaeem1 left 21:34 kjs_ left 21:36 kurahaupo1 left
pmichaud re star issue #43.... if I precompile a module using a non-installed copy of rakudo-m, can I still use the precompiled module once rakudo-m has been installed? 21:36
21:36 xfix left 21:38 lolisa joined 21:39 kjs_ joined 21:40 muraiki left
[ptc] masak++ # p6u post 21:40
21:43 jschulz joined
masak puts on a weary smile 21:44
I had a very good day teaching JavaScript today, by the way.
my best kind of JavaScript class is the kind that goes "oh dear, this language is *weird*!" with me
eli-se A day involving JavaScript is never a good day.
dalek pan style="color: #395be5">perl6-examples: 5f21112 | paultcochrane++ | t/004-website-basic.t:
Correct spelling of "receiver"

  (duh)		 21:45
pan style="color: #395be5">perl6-examples: 0e0a49e | paultcochrane++ | lib/Pod/Htmlify.pm6:
Add base-dir argument to files-in-category()
pan style="color: #395be5">perl6-examples: ce1370c | paultcochrane++ | / (3 files):
Convert write-example-files() into a method

This took a lot of work just to get the code tested. It seems I'd written legacy code... *sigh* Nevertheless, it's now tested, which means it can be more safely refactored into something better.
pan style="color: #395be5">perl6-examples: bd1b98b | paultcochrane++ | t/004-website-basic.t:
Extract repeated examples mock data into single variable
21:45 yqt left
pan style="color: #395be5">perl6-examples: 5d2b38c | paultcochrane++ | / (2 files):
Convert collect-example-metadata into a method
pan style="color: #395be5">perl6-examples: 98ff1a8 | paultcochrane++ | / (2 files):
Use $categories attr in collect-example-metadata
masak eli-se: I couldn't agree less. but you're entitled to feel that way.
21:47 larion left
masak quiz: anyone ever used `$.new` in a method? would you consider it a good idea? (as opposed to doing `CurrentClass.new`) 21:48
pmichaud masak: I think I would prefer 'self.new' to '$.new' 21:50
eli-se I prefer not to confuse types and values of types.
masak pmichaud: fair point.
pmichaud well, except that CurrentClass and self.WHAT might not be exactly the same.
vendethiel I'd definitely use self.new :-)
but that'd mean I'd "new" a subclass if I'm really inside a subclass, right? 21:51
eli-se $?CLASS.new
pmichaud to me, self.new means "make a new instance of whatever type I am"
21:51 woolfy joined
pmichaud and that's what "$.new" would mean. 21:51
21:51 woolfy left
masak yeah. 21:52
`$.` basically means `self.`, modulo itemization
pmichaud and as far as "confusing types and values of types".... that's not exactly what is happening here. :)
$?CLASS just happens to be an undefined value of the type.
it's not the "type" 21:53
masak conclusion so far: some things are better written `$.foo`, and some things `self.foo`, and I'm not really sure what the rule is :) 21:57
value-y things clearly look better as `$.foo`
things that do things clearly look better as `self.foo` (like .new and .clone)
pmichaud I tend to think of $.foo as being primarily for accessors. 21:58
masak right.
pmichaud I haven't wrapped my head around the idea of using it for general method calls. It's obviously okay to do it that way... but it still looks like an accessor to me
masak I think I'm basically using it for accessors.
and slowly coming to terms with it not being all that appropriate/intuitive for other things :) 21:59
pmichaud well, at one time there was a long design discussion about what ".methodcall" (no invocant) should mean/do.
i.e., should it default to "self.methodcall" or should it be "$_.methodcall".
it was finally decided to go to "self.methodcall", and there was also "$.methodcall" as an option. 22:00
masak that's one of the oldest discussions I remember. must've been around 2005.
pmichaud in the case of "$.new", "self.new", "CurrentClass.new", and "$?CLASS.new", I tend to think of them as actually doing different things. 22:01
well, "$.new" and "self.new" are equivalent.
masak my strangelyconsistent.org/blog/the-do...test-sigil is basically a celebration of the distinction we ended up with between `self` and `$_`
pmichaud some people would prefer to think of "self.new" as being more along the lines of "self.WHAT.new" (ensuring that .new is being handled by the type instance), but I think "self.new" is sufficient. 22:02
lizmat MMD is your friend if you really want to distinguish between .new on a type object or an instantiation :-)
masak "keeping $_ and self separate from each other in Perl 6 was a really good idea" 22:03
pmichaud I figure that anyone choosing between "self.new", "CurrentClass.new", and "$?CLASS.new" ought to be sure they realize how these can be subtly different. :)
lizmat m: my %h := Hash[].new # should this fail? if so, we need to handle the LTAness of the error message
camelia rakudo-moar 9b5d4e: OUTPUT«===SORRY!===␤Too few positionals passed; expected at least 3 arguments but got only 2␤»
jdv79 m: my @a = 1; for @a -> $a is rw { $a := 2; } # that error isn't clear as to why; at least to me 22:04
camelia rakudo-moar 9b5d4e: OUTPUT«5===SORRY!5=== Error while compiling /tmp/QHdJhkWUo0␤Cannot use bind operator with this left-hand side␤at /tmp/QHdJhkWUo0:1␤------> 3my @a = 1; for @a -> $a is rw { $a := 27⏏5; } # that error isn't clear as to why; ␤»
pmichaud jdv79: good point... I don't know if the "error" here is because it's disallowed in Perl 6 or because Rakudo doesn't know how to do it yet. 22:05
I suspect the error message should read "We don't know how to do this yet."
lizmat m: Hash[] # shorter way for same error 22:06
camelia rakudo-moar 9b5d4e: OUTPUT«===SORRY!===␤Too few positionals passed; expected at least 3 arguments but got only 2␤»
jdv79 is that a ticketable thing? or just ignore it for now?
dalek ar: 4fb39e4 | FROGGS++ | tools/star/windows-msi.pod:
add guide for MSI package creation
lizmat I assume, for simplicity sake, that Hash[] should be the same as Hash[Str] aka Hash
pmichaud jdv79: it's ticketable.
jdv79 ok, thanks
FROGGS please, review: github.com/rakudo/star/blob/master...ws-msi.pod 22:07
pmichaud FROGGS: reading 22:08
retupmoca FROGGS: I have strawberry perl already installed; is that fine?
FROGGS .tell azawawi github.com/rakudo/star/blob/master...ws-msi.pod
yoleaux FROGGS: I'll pass your message to azawawi.
22:08 Foxcool left
FROGGS retupmoca: I think were issues with strawberry, at least once there were IIRC 22:09
retupmoca ah
22:09 nige left
retupmoca I shall have to test later 22:09
22:10 kjs_ left
pmichaud FROGGS: that's interesting, I used to do all of my .msi releases using Strawberry Perl 22:10
FROGGS ahh, hmmmm
pmichaud since I didn't have MS Visual C++
and I also wanted some potential compatibility with Strawberry 22:11
FROGGS then my brane tricks me :o)
DrForr Just a brief doucmentation note - I don't see a .pod on doc.perl6.org where looping constructs are mentioned explicitly. (They're used in sample source) 22:12
22:12 virtualsue joined
pmichaud FROGGS: the guide looks really good to me. I suspect someone going through the document will find a couple of things overlooked, but it has all of the pieces I can remember. 22:12
FROGGS cool 22:14
yes, I also suspect that someone trying it will see glitches
gnight #perl6 22:15
22:15 FROGGS left 22:18 work_op joined, ingy left, ingy joined, raiph left
pmichaud I'm also afk... be back tomorrow 22:23
dalek kudo/nom: a57affe | lizmat++ | src/Perl6/World.nqp:
Don't curry if there's nothing to curry

This fixes the LTA error on e.g. Hash[]:
   Too few positionals passed; expected at least 3 arguments but got only 2		 22:24
22:26 spider-mario left 22:28 Woodi left 22:33 Woodi joined 22:37 work_op left, eli-se left 22:38 Woodi left 22:41 diana_olhovik left 22:42 skids joined 22:44 Woodi joined
TimToady moritz: see Pair.antipair vs Pair.invert 22:48
and note that the antipair definition contains no hyper 22:49
22:50 gfldex left
TimToady Oh, nevermind, you already found it 22:53
TimToady slinks back to the log
23:02 lolisa left 23:03 larion joined 23:05 Woodi left 23:07 hitchens joined 23:08 larion left 23:12 xprime joined 23:13 prime left
dalek ecs: c340ab6 | TimToady++ | S32-setting-library/Containers.pod:
document Hash.antipairs for moritz++
 23:13
rjbs Woah. 23:17
japhb .ask jnthn How scalable are supplies, in terms of sheer volume of Supplies and taps that I can have active before performance falls over or RAM gets prohibitive? My use case is an entity-relationship model for which I'm considering having an "updates" supply on every Entity and Relation, tapped by all clients who care about those nodes. At least thousands x dozens.
yoleaux 08:06Z <nine_> japhb: Cannot repor using the same perl and rakudo version. Will get me an Ubuntu 14.04 LTS VM and try it there.
japhb: I'll pass your message to jnthn.
rjbs I typed "s" at the REPL and it kept waiting for more input.
What happened?
japhb nine_: Dangit.
TimToady rjbs: given my fan sped up when I did that, I'll bet it's infinitely looping somehow 23:22
does if you feed a newline to perl6 -e too 23:23
well "s\n"
same for m
rjbs My work here is done! 23:24
(I tried things like a few /'s on subsequent lines, but only ^C seemed to help.)
TimToady guesses there's a problem in <.qok> 23:25
or triggered by .qok
rjbs Oh, now *that's* a useful thing to find: doc.perl6.org/images/type-graph-Associative.png 23:26
japhb rjbs: Available in SVG as well.
rjbs Yeah, but I couldn't link to that as well, so I chose the inferior png. :)
japhb Should just be s/png/svg/
Yeah, WFM
rjbs So it is! 23:27
japhb wonders how TypedHash gets a double role connection to Associative
rjbs There must be a good talk to be built out of the list of core roles.
japhb Some other core Roles/Classes like Numeric and Cool are interesting too. 23:28
rjbs Yeah. The good find was more "this kind of graph" than just that one.
japhb Yeah, understood 23:29
masak 'night, #perl6
japhb wonders why the type graph got moved to the bottom of the doc pages
Used to be right after the TOC
rjbs Huh. So a List is Cool, so I can use it as a string or number? 23:36
I see!
23:38 raiph joined
TimToady that was not part of the original intent of Cool, actually 23:39
but obviously it works out to some extent... 23:40
I wonder where the WAT is though 23:41
b2gills m: sub f (Int $n = 0) { $n > 3 ?? "liftoff!" !! $n + 1 }; say &f ... Str; 23:42
TimToady testing a fix for rjbs++'s looper
camelia rakudo-moar a57aff: OUTPUT«1 2 3 4 liftoff!␤»
dalek kudo/nom: 9746e88 | TimToady++ | src/Perl6/Grammar.nqp:
require existing character before peeking delims
 23:44
TimToady oh, and removed a <.ws> that is always redundant 23:47
(the problem was actually down in babble) 23:48
rjbs TimToady++ Thanks! 23:50
23:50 Rounin left 23:51 Woodi joined
TimToady m: m 23:53
camelia rakudo-moar a57aff: OUTPUT«5===SORRY!5=== Error while compiling /tmp/A1tbxsIt2D␤Undeclared routine:␤ m used at line 1␤␤»
TimToady arguably this message is LTA though
well, probably not worth the effort to be awesomer, since it can only occur at the end of the program 23:54
m: m 42
camelia rakudo-moar a57aff: OUTPUT«5===SORRY!5=== Error while compiling /tmp/oDyTyruIxQ␤Alphanumeric character is not allowed as a delimiter␤at /tmp/oDyTyruIxQ:1␤------> 3m7⏏5 42␤»
TimToady m: m;
camelia rakudo-moar a57aff: OUTPUT«5===SORRY!5===␤Regex not terminated.␤at /tmp/Zvjm9i6dQi:1␤------> 3m;7⏏5<EOL>␤Regex not terminated.␤at /tmp/Zvjm9i6dQi:1␤------> 3m;7⏏5<EOL>␤Couldn't find terminator ;␤at /tmp/Zvjm9i6dQi:1␤------> 3m;7⏏5<EOL>␤ e…»
TimToady it gets much more upset if there's anything after it
.oO(I get upset where there's anything after me...)
 23:56
*when
23:56 virtualsue left

Powered by Cro and the Raku® Programming Language.

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

Thank you!