»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!
Set by masak on 12 May 2015.
00:00 telex left 00:02 telex joined, mohij left
ugexe probably needs a more modern TAP::Harness 00:03
00:08 dolmen left 00:26 vendethiel joined 00:37 ir2ivps4 left
rjbs starts to bring perl v5.22.0 in for a landing. 00:39
That's the last v5 we need to make, right? :)
labster rjbs: yes :P 00:44
00:47 laouji joined, vendethiel left 00:56 adu joined 00:59 koo6 left 01:00 ssqq joined
timotimo i may do the weekly after sleep rather than before 01:00
01:02 mdinger left
ssqq How to get one subroutine name? sub get-it-name { say &?ROUTINE.^name } 01:02
m: get-it-name { say &?ROUTINE.^name }; get-it-name() 01:03
camelia rakudo-moar 00723c: OUTPUT«5===SORRY!5=== Error while compiling /tmp/PU8nGyN6Xh␤Undeclared routines:␤ &?ROUTINE used at line 1␤ get-it-name used at line 1. Did you mean 'gethostname'?␤␤»
ssqq m: sub get-it-name { say &?ROUTINE.^name }; get-it-name()
camelia rakudo-moar 00723c: OUTPUT«Sub␤»
timotimo m: sub get-it-name { say &?ROUTINE.name }; get-it-name()
camelia rakudo-moar 00723c: OUTPUT«get-it-name␤»
ssqq tomotimo: thanks 01:04
timotimo %) 01:05
01:07 rhr left, rhr joined 01:11 fhelmberger joined 01:12 kurahaupo joined 01:15 fhelmberger left 01:29 kurahaupo left 01:31 aborazmeh joined, aborazmeh left, aborazmeh joined
ssqq Editor `Atom` now support Perl 6 syntax highlight:) 01:37
timotimo has for a while; even the new visual studio has it now 01:38
because it was put into perl.tmbundle which is in upstream textmate 01:39
by visual studio i actually mean the new IDE microsoft put out there 01:40
ssqq Visual studio is too large ... 01:43
jdv79 how come rakudo isn't panda-able?
or rakudo star - i guess that's the only regular dist atm 01:44
timotimo Task::Star is panda-able
01:44 colomon left
jdv79 does that include rakudo? 01:44
01:44 colomon joined
jdv79 i don't see it in there 01:45
timotimo no
but how are you supposed to install rakudo with a perl6 program? :S
jdv79 what i'm getting at is cpan testers for perl6 which would include rakudo just as cpan testers smoke the perl dist
timotimo have you been working on a perl6 implementation in secret?
jdv79 no, just occurred to me 01:46
timotimo we have rakudobrew, which is most probably what you'd be using for something like that
01:46 laouji_ joined
jdv79 rakudobrew isn't cpan and cpantesters 01:46
01:46 raiph joined
timotimo well, we do have testers.perl6.org/ 01:47
01:49 laouji left 01:52 laouji_ left, laouji joined 02:00 aborazmeh left
jdv79 is there a regular report of test/pass of rakudo t, spectest, stresstest, etc? 02:00
timotimo yes, coke has that in his perl6-roast-data repository on github 02:01
it doesn't do stresstest and it doesn't do rakudo's own t, i believe 02:03
02:04 AlexDaniel left 02:12 tinyblak joined 02:19 ssqq left 02:20 ssqq joined 02:28 kurahaupo joined 02:29 rmgk is now known as Guest76179, rmgk_ joined, Guest76179 left, rmgk_ is now known as rmgk 02:38 laouji left 02:39 laouji joined 02:57 noganex joined 03:00 noganex_ left 03:01 ssqq left 03:03 bayprogrammer joined 03:10 koo6 joined
dalek kudo/nom: aadd29c | hoelzro++ | docs/running.pod:
Fix source reference for PERL6LIB/RAKUDOLIB
 03:36
kudo/nom: b30ef2e | hoelzro++ | docs/running.pod:
Add documentation for other env vars used by Rakudo
03:41 risou is now known as risou_awy, laouji left, risou_awy is now known as risou, laouji joined 03:44 sprinkles joined 03:45 sprinkles left 03:46 laouji left 03:48 Sqirrel left 04:10 jack_rabbit joined 04:13 BenGoldberg_ left 04:21 zcm left 04:25 quester joined 04:27 kaare_ joined 04:36 laouji joined 04:46 vendethiel joined 04:59 skids left 05:12 koo6 left 05:15 yubimusubi left 05:32 yubimusubi joined 05:45 Psyche^ joined 05:49 Patterner left 05:54 anaeem1 joined
adu pong 05:55
05:59 diana_olhovik_ joined, aborazmeh joined, aborazmeh left, aborazmeh joined 06:00 vendethiel left
jack_rabbit ping? 06:01
06:06 _mg_ joined
Ulti timotimo the Atom and VS Code syntax hilighting is fairly bad though :( nowhere near as good as vim. I think the main upside is that it means p6 is detected and you have a language name to hook plugins onto. 06:07
06:15 rindolf joined 06:17 espadrine joined
Ulti ooc is there a way to get Rakudo to output more machine readable error messages? 06:17
06:25 vendethiel joined 06:43 FROGGS joined
FROGGS o/ 06:45
06:47 pierrot joined 06:48 vendethiel left 06:49 domidumont joined
FROGGS Ulti: what do you meant by output? all exceptions have a type and inherit from Exception... if you want an error code-ish like thing, you might mess with Exception to always print the exception type upfront 06:49
Ulti I was thinking more like output JSON to make it easier to write IDE linters 06:51
just took a look at core/Exception.pm shouldnt be impossible to just abstract out the .gist in there a bit 06:52
maybe have a cli option like --error=json and default to --error=human or something?
06:53 domidumont left, domidumont joined
FROGGS hmmm, that sounds like a good idea 06:55
Ulti {line: x, pos: y, message: 'yay', type: 'PairMap::DoesNotExist'} 06:56
FROGGS yeah
Ulti --error=perl could just be the .perl of the exception which is nice 06:57
06:57 gfldex left
Ulti guess you might want file: in there too ;3 06:57
06:58 Ven joined
Ulti was looking at adding a rakudo p6 linter to Atom/Sublime 06:58
FROGGS IMO the exception should jsonify to: { "message" : "OHH NOES!", "data" : { "file" : ..., "line" : ..., "pos" : ..., "backtrace" : ... }
Ulti parsing the human readable error output feels wrong 06:59
FROGGS so you'd just need to set its "code" property, and use it in www.jsonrpc.org/specification
err, www.jsonrpc.org/specification#error_object
Ulti standards++ 07:00
ok I might have a play with that tonight since I only have to touch Perl 6 in Rakudo :)
FROGGS yeah :o) 07:01
07:02 raiph left, labster left, RabidGravy joined 07:07 zakharyas joined 07:08 Ven left
masak morning, #perl6 07:11
lizmat good *, #perl6! 07:13
only 2 more days until the next release!
sjn yay! 07:15
and good *, #perl6 07:16
07:16 dolmen joined
masak it will be the best Rakudo release so far! 07:18
top 'o the morning to ye, lizmat & sjn 07:19
lizmat masak sjn o/
07:19 bayprogrammer left
lizmat is enjoying the < 100 ms startup time 07:20
07:21 dolmen left 07:22 espadrine left, ribasushi left
dalek kudo/nom: a552010 | lizmat++ | docs/ChangeLog:
Add some more entries
 07:25
07:26 vendethiel joined 07:32 aborazmeh left 07:38 abraxxa joined, bjz joined 07:41 Sqirrel joined, Ven joined 07:44 quester left 07:45 tinyblak_ joined
FROGGS yes, only two more days :o( 07:46
07:48 tinyblak left
lizmat FROGGS: something I can do to help (apart from committing radical changes? :-) 07:49
*not
07:49 vendethiel left
FROGGS pheww 07:50
:P
Ven "Unhandled exception: ctxlexpad needs an MVMContext" trying to run perl6-debug-m (rakudo g72a187e)
07:51 ribasushi joined
FROGGS lizmat: hmmm, I need to get a jvm issue with wrt (de)serialization, dunno if you can help me there 07:51
fixed* 07:52
lizmat my JVM foo is limited...
FROGGS yeah, mine too
lizmat but if you have some pointers, I can do the grunt work of testing
FROGGS nah, it is not that easy... I think I can't even explain without understanding better what the issue is 07:53
I just know that bootstrapping panda does not work with CURLI and that it explodes in a strange way
lizmat in any case, your (de)serialization must have improved the past weeks :-)
07:53 aborazmeh joined, aborazmeh left, aborazmeh joined
FROGGS okay, I do know something: the string heap look very weird on the jvm, compared to moar 07:54
lizmat: yes, I learned a lot :o)
nwc10 en.wikipedia.org/wiki/ZingZillas -- Every Zingbop performed is of course "the best Zingbop ever".
however, I doubt that FROGGS has had massive exposure to CBeebies 07:55
er, or masak
masak: "so far" - but aren't they always? :-)
FROGGS nwc10: no contact whatsoever :o) 07:56
Ven can someone confirm debugger::ui::commandline works/doesn't work for them :)?
FROGGS Ven: I have a busted build, so I cannot help here 07:57
dalek Heuristic branch merge: pushed 24 commits to rakudo/newio by lizmat
07:59 laouji_ joined 08:02 laouji left 08:03 adu left, laouji_ left, rindolf left
masak nwc10: yes, that's the joke 08:05
nwc10: I watched a Guido talk, where he introduced Python 3.4 as "the best Python release ever" 08:06
08:07 tinyblak_ left, tinyblak joined
nwc10 I didn't know that. Does Guido use the phrase as a running gag? 08:07
08:09 laouji joined
masak not sure. I'm pretty sure there was some self-awareness as he said it. 08:10
just because it's always true doesn't mean it's not worth stating. 08:11
*especially* in the Python 3.4 vs Python 2.7 situation that's currently the fact.
08:17 FROGGS left 08:18 fhelmberger joined, darutoko joined 08:25 dakkar joined 08:33 average joined
average perl6, the language everyone's talking about 08:33
(along with 999 other languages)
[ptc] Ven: yes, I can confirm that it doesn't work for me as well
Ven thanks [ptc] 08:34
average modules.perl6.org/ <-- I just looked at this, pretty nice eco-system you guys have going on
[ptc] I even asked the same question in irc, but noone bit :-/
average how are threads doing in Perl6 ? how's async I/O doin' in Perl6 ?
[ptc] the issue has been around for a while too.
average coming along ?
08:34 mr-foobar left 08:35 laouji left
RabidGravy average, the concurrency piece is good enough to write code for right now 08:35
average RabidGravy: is the interface stable ? 08:36
to be clear, I'm refering to the API, and the parts of the language that deal with concurrency
RabidGravy yeah, largely I believe, to the extent that is documented and all 08:37
average is it stable from an API perspective ? (that's my perspective as an end-user)
RabidGravy yes
average ok, thank you
RabidGravy what is documented in docs.perl6.org/language/concurrency all works right now and is stable 08:38
08:40 laouji joined
dalek kudo/nom: 29a4089 | lizmat++ | src/core/Deprecations.pm:
Allow for :file/:line in DEPRECATED

This should make it easier to issue DEPRECATED messages at compile time		 08:41
kudo/nom: c53b603 | lizmat++ | src/Perl6/World.nqp:
DEPRECATE use MONKEY_TYPING
kudo/nom: a6e5e93 | lizmat++ | src/Perl6/Actions.nqp:
Use new World.current_file convenience method
08:42 bjz left
lizmat afk for a few hours& 08:42
08:43 bjz joined, espadrine joined
masak surely this list of examples breaks the rules set up in the paragraphs before? en.wikipedia.org/wiki/Unit_testing...ng_support 08:50
specifically, Go, Java, Python, Racket, Ruby, Rust don't have *language-level* testing support 08:51
(and probably more)
it's as if someone wanted to add their favorite language to that list but forgot to read the rules...
slightly off-topic, sorry. :) just happened to be on that page and saw it. 08:52
does anyone know if *any* language on that list merits its place there?
08:52 tinyblak left 08:53 tinyblak joined 09:00 FROGGS joined 09:05 khisanth__ joined, gargamel_ left 09:08 khisanth_ left, koo6 joined
masak oh, this is pointed out in the Talk page, too. 09:13
apparently D and Cobra have built-in testing support in the language.
09:14 tinyblak left 09:21 bjz left, tinyblak joined, Ven left 09:23 bjz joined 09:25 dayangkun joined
profan D has it's place at least, unittest and invariant blocks :) (very handy, especially the invariant blocks: dlang.org/contracts.html#Invariants) 09:26
masak I would make a distinction between unit tests and things like invariant blocks. 09:27
average hey masak , and everyone who's contributing
profan I suppose the second ones are closer to assertions
average to Perl6
masak average: hey! ÖD
:D
profan rather than just for testing purposes
09:27 tinyblak left
average I don't understand how you guys are pulling this off, but it's great.. 09:27
masak profan: well, it's a sliding scale, I guess.
profan masak: I'd guess perl6 doesn't have any plans for language level testing support? 09:28
masak average: 8 years ago, we figured out how to clone jnthn, and we keep 100 copies of him locked in the basement...
profan: no, I think we're fine with it in the Test.pm module :)
average what's your secret though ? 09:29
masak profan: but it's already possible to write an `assert` macro.
profan The language is already quite the beast, so that's probably sane :)
average you seem to have more tranction and more long-term commitment than many commercial projects that make loads of money
how does it work ?
*traction
masak average: passion.
average nice :)
09:29 dayangkun left
masak average: this channel is full of people who really enjoy Perl. 09:29
RabidGravy passion and sheer bloody mindedness 09:30
masak average: and who want to see it have a future through Perl 6.
average that's very nice :)
masak but yeah, we're feeling the second system syndrome vibes sometimes :P
guess that's unavoidable
profan looking forward to the language stabilizing around 1.0, bit hesitant to jump in until then :) 09:31
from the little perl6 i've written, i already miss many parts of perl6 whenever i write in other languages, so i wish you guys the best :D 09:32
itz_ jumping in now will help the language stablise quicker! :) 09:33
profan itz_: haha, my passion for p6 isn't quite as strong as yours ;) 09:37
09:39 bjz left
masak itz_ has a point, though. 09:44
the only thing that makes the language more stable is use.
(in my experience)
the "6.0" point will mostly be a ceremonial thing.
RabidGravy and the broadening of the contribution base to people who want to do things that nobody else wants to do is important IMO 09:47
masak yeah. 09:49
for the rest of 2015 at least, we can say this: we are looking for people who want to explore, get into trouble, and help fix the trouble :) 09:50
after the arbitrary "6.0" point, there will of course be no bugs :P
RabidGravy ;-O 09:51
masak pokes fun, but will probably be just as giddy as the rest of the channel once we reach those dates
DrForr To misappropriate Barbara Woodhouse, there are no bugs, only inappropriate behavior. 09:52
RabidGravy WALKIES!
09:55 bjz joined
FROGGS for me the "6.0" point is what happens after a lot of testing, fixing and cleaning up (the latter mostly for roast) 09:57
10:03 average left
masak *nod* 10:04
so that's what we focus on.
10:04 jack_rabbit left
FROGGS besides getting the last required bits in a good shape :S 10:05
profan i suppose that part about maturity is true, might poke at it a bit earlier than scheduled then :)
masak \o/ 10:08
10:08 bjz left
masak my part of the work is not a blocker for 6.0, but I'd still like to do a fair bit of the planning in 2015. 10:08
maybe even some implementation somewhere in a branch.
Real Actual Macros.
FROGGS masak: well, would be nice to know the final quasi quoting syntax later this year 10:09
so we can at least reserve it or so
10:10 bjz joined 10:15 bjz left 10:18 Sqirrel left, araujo left, telex left 10:19 Sqirrel joined 10:20 telex joined
masak aye. 10:23
there's no good answer to that particular one. 10:24
at least not yet.
mostly because unquoting/hole syntax is nasty business.
FROGGS aye 10:25
masak I mean, the most lightweight I can think of would be
my $ast = `say "OH HAI"`;
(that is, instead of 'quasi { ... }', we do '`...`' 10:26
)
but that doesn't address how to spell the holes.
FROGGS that feels like variabel interpolation 10:27
variable*
10:27 rindolf joined
FROGGS problem is, that this is code, not a string or regex :/ 10:27
10:28 jack_rabbit joined
masak well, you bring a nice point of view to it 10:28
(that I hadn't considered)
quasis are like a Perl 6 slang inside of Perl 6
FROGGS yeah 10:29
masak the only grammatical difference is that there are holes
FROGGS a macro can fill holes with lexicals from the outer scope? 10:30
masak oh yes. very yes. 10:31
FROGGS dang :P
DrForr We've already got twigils, what's one more character among friends? $`^x, $,^x, $@^x
FROGGS would be easy otherwise
DrForr: that was what I was thinking
which does not help when outer lexicals are involved
it is like you need an interpolation syntax like {{{ }}}, just shorter 10:32
though, I dunno if that is the biggest issue here 10:33
DrForr $x = `$a,$,^b,@@^b`; # remembering only now the fact that context now uses the type sigil.
literal make it lispy? `() :P
DrForr That was why I mentioned $`x, $,x, $@x inasmuch as they're the quasiquoted versions. 10:34
10:34 andreoss joined
masak interpolated ASTs don't have to come from variables. 10:34
they're arbitrary expressions.
FROGGS: simplest example of "lexicals from the outer scope" is... macro parameters. :) 10:35
masak .oO( Plan 9 From the Outer Scope ) 10:36
FROGGS masak: yes, but you have the naming of these params under control, so we could have interpolating twigils or so
10:36 vendethiel joined
FROGGS though, we usually have circumfixes for that, and it probably wants to be that 10:37
DrForr (defmacro foo(body) `(list 'a ,body)) (foo `(+ 1 55)) # and such. If we had to use {{{ }}} for delineating blocks it'd start to look like we added trigraphs from C.
There are some [[ .. ]] that look like doubled characters in Unicode, but I can't find them and imagine people would want alternatives in ASCII anyway to avoid copying from a UTF-8 text file. 10:39
10:39 aborazmeh left
DrForr ⋘$x + $y⋙ 10:40
(guillemots with one extra < > added, essentially)
10:42 jack_rabbit left 10:44 jack_rabbit joined 10:49 bjz joined
andreoss m: #`⦕ comment ⦖ say 1 10:50
camelia rakudo-moar a6e5e9: OUTPUT«1␤»
FROGGS not in my font :o(
andreoss xahlee.info/comp/unicode_matching_brackets.html 10:51
10:51 kurahaupo left 10:52 RabidGravy left
masak one thing I dislike about {{{ }}} is that it has very little to do with blocks. 10:52
also, quasi blocks are meant to be writable with other delimiters, like <> [] 〖〗 10:53
which -- contrariwise -- makes little sense for blocks. 10:54
dalek kudo/nom: 56e4733 | lizmat++ | src/Perl6/World.nqp:
Generalize a World.DEPRECATED method, and use it
masak I once thought that the great idea behind quasi blocks having different delimiters was that they could *nest*, and that at a certain depth you'll be interested in specifying a *certain* quasi block's hole 10:55
andreoss m: sub x {{{ 1 }}}; say x.perl;
camelia rakudo-moar a6e5e9: OUTPUT«1␤»
masak much in line with `next` et al. and loops/loop labels
andreoss: oh, that's another thing. {{{ }}} is *allowed* in non-quasi Perl 6. it's just that it doesn't mean much. 10:56
I'd much prefer if we found a hole syntax that wasn't already allowed in mainline Perl 6.
masak is glad about this discussion
at least I'm getting a clearer picture of my kvetches :)
11:02 vendethiel left 11:06 jack_rabbit left, berekuk joined
FROGGS damn, we are caching file stats, right? 11:07
11:08 _mg_ left 11:10 vendethiel joined 11:11 yqt joined, yqt left 11:12 yqt joined
FROGGS for example: 11:12
$ perl6 -e 'my $io = "foo".IO; mkdir $io unless $io.d; say $io.d'
Failed to find 'foo' while trying to do '.d'
lizmat: any idea?
lizmat: is there something in IO::Path that can clear $!e?
lizmat IO::Path doesn't guarantee something exists 11:13
FROGGS that's not the point
lizmat that's one of the things I tried to fix in newio branch
FROGGS problem is that it lies
lizmat until I realised that having symlinks work the way thay work now
FROGGS it created the directory but .d does not attempt to check it
lizmat means that a directory entry doesn't mean something exists 11:14
FROGGS hmm?
I don't get it
11:14 _mg_ joined
lizmat not important right now 11:14
FROGGS it is for me
lizmat well, not for your problem at hand
FROGGS because I cannot bootstrap panda
is there something that can reset $!e in IO::Path? 11:15
lizmat so, let me get straight what you want:
FROGGS k
lizmat 1. if there is a dir already by that name, do nothing 11:16
2. if there's nothing there, create directory
FROGGS correct
lizmat 3. if there is something there, and it's not a directory, fail
FROGGS no
3. bail out if there is still no dir
11:16 pauluu joined
lizmat then I would just do a mkdir unconditionally and check afterwards if it iexists and is a dir 11:17
mkdir $io; die unless $io.e && $io.d
FROGGS :/
lizmat checking for existence is a race condition anyway
(or non-existence) 11:18
FROGGS okay, so since panda is using Shell::Command my only choice is to convert between strings and IO::Paths it seems
quite smelly
lizmat many parts of panda are not the most hygienic :-) 11:19
FROGGS nah, that's not true
panda is quite well designed 11:20
lizmat pushing to @*INC ?
FROGGS that was your temp hack
lizmat huh?
did I?
FROGGS ups: github.com/tadzik/panda/commit/df8...7db78a57eb 11:21
I thought these @*INC pushed were not there at all
pushes* 11:22
lizmat :-)
*phew* :-)
FROGGS hmmm :o(
so these @*INC pushes are about the case when somebody installs to DESTDIR (and env var) 11:23
an*
lizmat I have no idea... and when I asked tadzik, he didn't know either 11:24
FROGGS the comment talks about the destdir not being in @*INC, and only DESTDIR can trigger that
bootstrap.pl sets up @*INC indirectly via PERL6LIB 11:25
lizmat yeah, not sure whether that is correct either... 11:27
FROGGS "==> Successfully installed panda" phew 11:28
lizmat because @*INC will still include the "normal" entries as well
11:32 vendethiel left
masak FROGGS++ # panda care and feeding 11:38
11:40 andreoss left 11:41 andreoss joined 11:44 colomon left 11:46 andreoss` joined 11:48 andreoss left 11:50 colomon joined, Ven joined
Ven masak: what counts as language-level testing support? 11:51
is java's "assert" considered enough for that list? haha. (D has "unittest" blocks)
masak Ven: my criterion when reading that list was "shouldn't have to import something" 11:52
Ven well, then that's true for, say, rust.
masak oh, ok.
Java's assert is close, I guess. but it doesn't behave like a unit testing construct. after an `ok` fails in Test.pm, more tests keep running after that.
Ven it has #[test] attribute 11:53
masak cute.
Ven Yeah, java's `assert` is pretty bad :-). Scala doesn't have anything, why is it in that list?
Ruby has something *included*, but not quite special syntax 11:54
> functional languages (Lisp, scheme, haskell) can be good for testing when their functions are side effect free -its very easy to test functions when their whole output is visible and testable. 11:56
Ven sighs
masak call that "TDD overconfidence" 11:57
well.
it *is* made easier by that, I guess.
Ven masak: it seems that list includes every language that includes a *module* for unit testing in its main distribution
then, is that true of c# which has a testing package *under visualstudio*?
masak yeah. which breaks the rules stated above the list.
11:57 FROGGS[mobile] left 11:58 FROGGS[mobile] joined
Ven (I mean, literally under the "visualstudio" namespace) 11:59
FROGGS lizmat: panda just wants to add entries to upfront, and an ENV var is quite handy when shelling out 12:00
12:00 vendethiel joined
masak gave in and fixed the list 12:01
12:01 Ven left 12:03 berekuk left 12:05 berekuk joined
dalek kudo/nom: 7803bf0 | lizmat++ | src/core/Deprecations.pm:
Generalize deprecation message re compile time use
 12:08
kudo/nom: 5278e45 | lizmat++ | src/Perl6/Grammar.nqp:
Make lack of unit a 'real' deprecation message
kudo/nom: 236b089 | lizmat++ | docs/ChangeLog:
Add some more ChangeLog entries
12:10 dolmen joined 12:11 mazeTemporal joined 12:13 Ven joined
Ven masak++ 12:13
12:13 yqt left
Ven thinks we're in a really good shape with 1) PRE/POST 2) subsets/where 12:14
(probably my favorite features of perl6... and sixcheck was a pleasure to write)
12:15 yqt joined
dalek ast: f6e598f | lizmat++ | S02-types/ (2 files):
Follow latest deprecation wording changes
 12:15
Ven looks at trigils in fear (and awe) 12:16
mazeTemporal Coke 2015-5-18 23:55 : Yes, with just perl6 but not prove, the tests worked as expected. The error is thrown when prove tries to execute: perl -x -S prove -e perl6 hellotest
Ven irclog.perlgeek.de/perl6/2015-05-19#i_10625744 masak, people used to think lisp's nested macros were behavior wrong. then someone came and (ab)used it, and everyone rejoiced 12:17
Ven thinks he's gotten this "history bit" from Let over lambda or On lisp
12:19 laouji left
mazeTemporal Does anyone know if there is a simple way to panda install a module but skip the testing? 12:19
masak Ven: well, TimToady has explained to me that quasis don't nest that way. 12:20
Ven masak: oh, seems I missed that in the backlog
masak Ven: ...so I really don't understand why quasis should have configurable delimiters.
I think I meant to arrive at that point, but I got derailed by the subsequent discussion :) 12:21
mazeTemporal: `panda install <module>` ?
oh wait
mazeTemporal: `panda --notests install <module>` ? 12:22
12:22 xfix joined
mazeTemporal masak: ah, I should have checked for flags. Excellent 12:22
12:24 vendethiel left 12:28 tinyblak joined
masak mazeTemporal: I found out by reading github.com/tadzik/panda/blob/maste.../panda#L18 12:29
12:31 tinyblak_ joined, tinyblak left 12:32 tinyblak joined
mazeTemporal masak: was there a way for you to know to pick bin/panda rather than any of the other items in bin/ ? 12:33
12:34 domidumont left
masak because `panda` is the command I (and everyone) run all day ;) 12:34
dalek kudo/nom: 8ab32b8 | lizmat++ | src/ (2 files):
Make 'winner' a compile time deprecation
12:35 tinyblak_ left 12:38 mazeTemporal left, domidumont joined, anaeem1 left
dalek kudo/nom: 4924f04 | lizmat++ | src/core/VM.pm:
Remove expired deprecations for 2015.05 release
 12:50
ast: db4b007 | lizmat++ | S02-types/deprecations.t:
Follow expired deprecation removal
 12:51
ecs/newio: 7a0cb02 | lizmat++ | S32-setting-library/Numeric.pod:
Concur with implementation of Complex.roots
 12:56
13:07 araujo joined, araujo left, araujo joined 13:08 vendethiel joined, araujo left 13:10 raiph joined 13:12 dayangkun joined 13:16 dayangkun left, dayangkun joined 13:17 rarara joined, dayangkun left 13:18 dayangkun joined, dayangkun left, dayangkun joined
dalek kudo/curly: 7b6caee | FROGGS++ | src/core/Distribution.pm:
restructure Distribution; do work once
 13:22
rakudo/curly: 74a03d0 | FROGGS++ | src/core/CompUnitRepo/Locally.pm:
rakudo/curly: do no create CURL directories "automagically"
13:22 dalek left, dalek joined, ChanServ sets mode: +v dalek
FROGGS what dalek did not show was this: " use CURLI on moar but not on jvm " - github.com/rakudo/rakudo/commit/2d5b07c 13:23
with this I'm considering merging rakudo/curly and panda/curly, to get feedback if there is any to get 13:24
masak I just realized that, as much as I like Moose and Perl 5, mentally I have these pegged as "more cumbersome/boilerplate-y ways to do Perl 6" 13:29
if that makes sense
probably most p5 people don't think of it this way :>
FROGGS I don't think it is boilerplately, I think it is only half awesome because they don't have jnthn\W6model 13:30
masak Perl 5 OO without Moose is definitely boilerplate-y. 13:31
PerlJam masak: only those that are familiar with Perl 6 :)
FROGGS masak: I just know it from code snippets... so don't listen to me :o)
13:31 skids joined
FROGGS masak: you've not yet seen Carrot :D 13:32
[Coke] timotimo: it runs everything in t/spectest.data, so it's effectively stresstest, yes. doesn't do t/
FROGGS masak: you need to look at that and tell me opinion: skim that: www.carrot-programming.org/ and scroll down this page: www.carrot-programming.org/perl/tutorial.html 13:33
[Coke]: star does t though, right? 13:34
masak FROGGS: invoking en.wikipedia.org/wiki/Poe%27s_law 13:37
FROGGS: I literally don't know whether to twitch the corners of my mouth into a smile, or to be very very disturbed by what I just read
[Coke] FROGGS: supposed to, but:
github.com/coke/rakudo-star-daily/...rakudo.log
masak which I guess means that the artist in question did a good job. 13:38
[Coke] apparently it's been failing since Mar 30th
FROGGS masak: the creator of carrot gave a talk at the german perl workshow this month...
masak: the audience was more than puzzled so to say
[Coke]: ewww
masak FROGGS: someone gave a talk about this? that's amazing. 13:39
moritz very good trolling 13:40
FROGGS masak: yes yes... and he could not explain why somebody should use it, or what it is for, or what it gives you... even after three question of the audience
masak FROGGS: but the SYNOPSIS explains it so clearly already! 13:41
I don't understand the source of your confusion
here, let me quote
"higher quality code through systematic expressiveness"
PerlJam FROGGS: it gives you a carrot!
masak what part of this don't you understand?
PerlJam or ... it doesn't give you a stick! 13:42
masak and if you understand it, what part would you *not* want? :P
moritz the consistent use of "Perl 5" instead of just "Perl" suggests that the author might be involved with Perl 6
"normal" Perl 5 people just write "Perl" when they mean Perl 5
FROGGS masak: the part where I write a huge block of code and don't know what *it* does with it :o)
s/code/comments/
moritz: no, he's a Perl outsider I think 13:43
like he asked: Is the magic constant in return(PERL_FILE_LOADED); still needed?
masak he's a good Samaritan 13:44
he saw the Perl 5 people suffering, and he swooped in to give them systematic expressiveness
kind of like someone decided to make "Perl 5 the good parts" 13:45
but without knowing much about Perl 6
Perl 5*
13:45 andreoss` left
masak or good parts 13:45
13:45 andreoss` joined
moritz I liked the PDF, with the first example being a class representing monadic names 13:46
masak he's mis-using the name "monad" for something that means "not quite a singleton". 13:51
looks like he's built a Perl 5 preprocessor. for some reason I don't understand. 13:52
dalek pan style="color: #395be5">perl6-examples: 66ad4a2 | andreoss++ | / (2 files):
[euler] solution for problem 80
FROGGS I think his goal is to have enough boilerplate in the source as some sort of documentation to state very very clearly what it does 13:54
though, it is not quite readable for me
masak I think I might be able to have an interesting conversation with him over beer or so. 13:55
seeing the examples, I don't see what this would give me over, say, Moose.
I already write nice and descriptive variable names, and I don't need a framework for that.
dalek kudo/nom: f4de5f3 | lizmat++ | src/core/ (2 files):
Eat our own unicode operator dogfood
 13:56
moritz thinks it's very artful trolling
FROGGS sad that we are not writing VB, we could have varying names instead :P
masak moritz: except it's been extracted from some production system he has called Mica. 13:57
FROGGS: you mean case-insensitive names?
*surely* there must be an Acme:: CPAN module for that?
otherwise we owe it to ourselves to produce a slang for it :P
FROGGS masak: aye
masak FROGGS: you know what to do... :>
FROGGS nooooo! /o\ 13:58
masak :)
FROGGS though, should be fairly easy to do btw :o)
damn 13:59
Program received signal SIGSEGV, Segmentation fault.
gc_mark (tc=0x6036f0, st=<optimized out>, data=<optimized out>, worklist=0x3984e30) at src/6model/reprs/SCRef.c:75
curly segfaults when used with EVALs
>.<
masak .oO( the three stoogefaults ) 14:00
FROGGS cries 14:04
lizmat resists slapping FROGGS repeatedly on the head 14:05
14:06 yqt left
andreoss` m: my @e := (2; gather take 1,$_,1 for 2,4 ... *); say @e[^20]; 14:06
14:06 anaeem1 joined
moritz masak: ... or so he claims 14:08
masak troo
moritz masak: "Not paranoid enough!" 14:09
14:09 newbie joined, newbie left
FROGGS I need some distance between and curly now 14:09
me* 14:10
masak moritz: ;) 14:11
14:11 anaeem1 left
FROGGS :P 14:11
lizmat: how kind of you :o)
m: say '(timeout)' 14:12
no timeout? 14:14
DrForr 'FROGGS' ~ "\n" xx 10e100 ~ 'DrForr' 14:17
s/DrForr/{/ :) 14:18
lizmat commute to NRPM meeting& 14:20
14:20 lizmat left
FROGGS *g* 14:20
14:23 Erendis42d joined
Erendis42d apa<3 14:23
masak Swedish for "monkey"
Erendis42d monkey<3
masak :)
Erendis42d: are you in .hu ? 14:24
Erendis42d oops
Erendis42d zipp
x)
masak ok, got it :) 14:25
camelia rakudo-moar 4924f0: OUTPUT«2 1 2 1 1 4 1 1 6 1 1 8 1 1 10 1 1 12 1 1 14 1 1 16 1 1 18 1 1 20 1 1 22 1 1 24 1 1 26 1 1 28 1 1 30 1 1 32 1 1 34 1 1 36 1 1 38 1 1 40 1 1 42 1 1 44 1 1 46 1 1 48 1 1 50 1 1 52 1 1 54 1 1 56 1 1 58 1 1 60 1 1 62 1 1 64 1 1 66 1 1 68 1 1 70 1 1 72 1 1 74 1…» 14:26
rakudo-moar 4924f0: OUTPUT«(timeout)␤»
masak andreoss`: interesting. what is that? 14:27
oh, it's the continued fraction for e. 14:31
14:31 yubimusubi is now known as Possum
masak oeis.org/A003417 14:31
oeis++
andreoss` it is.
masak now that Java 8 interfaces can have default method implementations, Java interfaces feel a lot more like Perl 6 roles. 14:32
andreoss` m: my @e := gather { take 2; take (1; $_; 1) for 2,4 ... * }; say @e[^10];
camelia rakudo-moar f4de5f: OUTPUT«2 1 2 1 1 4 1 1 6 1␤»
dalek kudo-star-daily: e816fe3 | coke++ | bin/git.p5:
Don't need to pull parrot at this time
 14:33
andreoss` is there a nice way to put something before gather without forcing to eager context?
m: my @e := 2, gather take (1; $_; 1) for 2,4 ... * ; say @e[^10]; #eager 14:34
camelia rakudo-moar f4de5f: OUTPUT«(timeout)»
14:36 andreoss` left, andreoss joined
andreoss *forcing it to 14:38
14:38 RabidGravy joined 14:40 vendethiel left
[Coke] rakudo-star-daily failure seems to begin with this: 14:40
perl6 tools/build/panda-state.p6 >>MANIFEST
Failed to connect: host is unreachable
14:41 anaeem1_ joined
[Coke] ... because it's hardcoded to go to feather.perl6.nl:3000/projects.json 14:42
When did feather die? 14:43
FROGGS end of march 14:44
[Coke]: use ecosystem-api.p6c.org/projects.json
skids glasses to glasses, rust to rust
FROGGS andreoss: wait for the GLR I'd say 14:46
m: my @e := 2, lazy gather take (1; $_; 1) for 2,4 ... * ; say @e[^10];
14:46 _mg_ left
camelia rakudo-moar f4de5f: OUTPUT«(timeout)» 14:46
14:47 lolisa joined 14:49 anaeem1_ left 14:50 anaeem1 joined
dalek ar: f704eda | coke++ | tools/build/panda-state.p6:
Update projects list, feather is no more.
 14:56
14:56 FROGGS left
[Coke] FROGGS++ # interestingly, we updated the changelog in star to mention it, but not the feather links themselves. 14:57
14:57 domidumont left 14:58 domidumont joined
dalek pan style="color: #395be5">perl6-examples: 4430880 | andreoss++ | / (3 files):
[euler] solution for problem 65
 14:59
14:59 berekuk left
[Coke] now getting: perl6 tools/build/panda-state.p6 >>MANIFEST 15:00
No such method 'ast' for invocant of type 'Any'
will dig more after dayjob
15:10 dolmen left 15:17 dolmen joined 15:23 gfldex joined 15:26 anaeem1 left
dalek pan style="color: #395be5">perl6-examples: 009e71b | andreoss++ | / (4 files):
[euler] solution for problem 66
 15:26
Ven if I use a proto in a grammar, will the modifiers be applied... "naturally"? What I'm thinking of: if I have several variants, but they're not ordered in any way, can I use a "proto regex" (with all actual regexes be "token"s) and will it backtrack upon failing to match the first one? 15:27
15:29 zakharyas left
Ven oh, obviously, <.ws> matches newlines as well... 15:30
skids token foo:<sym> is specced to be equivalent to an alternation I believe. 15:31
15:32 lolisa left
Ven really? I changed the order and some of my stuff suddenly matches 15:32
skids *foo:sym<bar> 15:33
Ven wow, ws handling is core is kind of crazy...
skids: ah, thanks
s05 doesn't talk about sym 15:34
(it *mentions* it without explaining if it means anything special) 15:35
skids The language says: "All of them will be matched in parallel when you try to match" based on longest token, with ties causing an actual dispatch using any formal parameters.
I think the :sym is just the name of the $sym inside the rule if you need to self-reference it. 15:36
And the magic is in the colon :-)
*<sym> not $sym apparently. It autmakes a subrule name. 15:37
Ven goes and writes "token sp { ' '* }"
skids: then I might have a LTM issue :) 15:39
15:40 TimToady left 15:42 TimToady joined, dayangkun left
hoelzro is there a way to add custom type adverbs? 15:50
instead of just Type:U and Type:D?
skids well, constraints. 15:51
[Coke] m: my $s = IO::Socket::INET.new(:host<ecosystem-api.p6c.org>, :port(80)); 15:52
$s.send("GET /projects.json HTTP/1.0\n\n");
my ($buf, $g) = '';
$buf ~= $g while $g = $s.get;
camelia rakudo-moar f4de5f: OUTPUT«IO::Socket::INET is disallowed in restricted setting␤ in sub restricted at src/RESTRICTED.setting:1␤ in method new at src/RESTRICTED.setting:32␤ in block <unit> at /tmp/v3WxcpXK4c:1␤ in any <unit-outer> at /tmp/v3WxcpXK4c:1␤␤»
[Coke] whoops.
gist.github.com/coke/822e10edcc7a89526294 - this gets a 404 from design.perl6.org 15:53
Guessing we need a host: line here as well.
15:54 dolmen left
[Coke] that doesn't seem to help. 15:54
Ven ...is it expected that "dba" makes a rule fail? 15:55
TimToady I would not expect it 15:56
[Coke] er, 400, not 404.
so, that gist is currently blocking anyone building a Star. 15:57
15:57 Ven left
tadzik oh hai 15:58
[Coke] ah, got it.
15:58 xfix left 15:59 lizmat joined
dalek ar: 6ed4ca0 | coke++ | tools/build/panda-state.p6:
Unify get request for proxy/non-proxy

ecosystem host confused by the http/1.0 attempt there
  (used to work fine against feather)		 16:01
16:02 xfix joined
japhb wierds out at servers failing to understand HTTP/1.0 16:02
[Coke] japhb: I have no idea what about that request was wonky. 16:04
but the 1.1 version worked fine, so why even have the 1.0 attempt. 16:05
16:06 khisanth__ is now known as Khisanth
[Coke] that fixed it. next rakudo-star-daily run should be fine again. 16:07
16:09 vendethiel joined
[Coke] (the NEXT one, not the one I just pushed.) 16:09
16:09 [Sno] left
andreoss m: say True ¿¿ 1 ¡¡ 2; 16:12
camelia rakudo-moar f4de5f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/aTsqg1zZXt␤Two terms in a row␤at /tmp/aTsqg1zZXt:1␤------> 3say True7⏏5 ¿¿ 1 ¡¡ 2;␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ …»
16:15 eli-se joined
eli-se hi 16:16
FROGGS[mobile] how wrong is it to merge a branch that one is confident enough when jnthn wanted to review it first?
TimToady std: say True ¿¿ 1 ¡¡ 2; 16:17
camelia std 28329a7: OUTPUT«5===SORRY!5===␤Confused at /tmp/saIuFP9Xbk line 1:␤------> 3say True 7⏏5¿¿ 1 ¡¡ 2;␤ expecting any of:␤ feed_separator␤ infix or meta-infix␤ infixed function␤ statement modifier loop␤Parse failed␤FAILED 00:00 136m␤»
TimToady I coulda sworn I fixed rakudo to say "Confused" there rather than TTIAR, which is misleading
hoelzro if someone wanted to modify smoke.perl6.org to include dists whose tests passed but raised warnings, what would be the best way to go about that? I saw last night that smoke.perl6.org uses emmentaler (sp?), which uses panda 16:18
16:20 abraxxa left, berekuk joined 16:25 berekuk left
FROGGS[mobile] I'd record the output of 'panda smoke' if it is meant for single use 16:26
16:27 anaeem1_ joined
dalek ast: a87fde4 | lizmat++ | / (95 files):
s/is_deeply/is-deeply/g
 16:28
16:38 domidumont left, isBEKaml joined
hoelzro FROGGS[mobile]: I didn't know about panda smoke, thanks! 16:40
16:40 mohij joined
hoelzro I think I would just need to capture $*ERR and see there's anything in there 16:41
FROGGS[mobile] in a subprocess? 16:42
hoelzro yeah 16:43
FROGGS[mobile] O.o
hoelzro FROGGS[mobile]: not a good idea? 16:44
FROGGS[mobile] tbh, I hack the deprecation printer in rakudo... 16:45
I'd*
16:46 lolisa joined
[Coke] if jnthn has requested to review something, I'd leave it until then. 16:48
16:48 diana_olhovik_ left 16:49 berekuk joined, mr-foobar joined
[Coke] looks like dalek missed a push to rakudo-star-daily. 16:50
16:50 berekuk left
lizmat FROGGS[mobile] hoelzro : RAKUDO_NO_DEPRECATIONS=1 will also kill any deprecation messages 16:51
16:51 [Sno] joined
andreoss how do i put math formulas in pod? 16:51
japhb lizmat: Does that also apply to the "compile time deprecations" you added recently? 16:53
lizmat yes
except for the nqp one
lemme check
japhb cool beans. lizmat++
Which is the nqp one?
lizmat m: nqp::say("hello")
camelia rakudo-moar f4de5f: OUTPUT«The use of nqp::operations has been deprecated for non-CORE code. Please␤change your code to not use these non-portable functions. If you really want␤to keep using nqp::operations in your Perl6 code, you must add a:␤␤ use nqp;␤␤to the outer …»
lizmat m: use nqp; nqp::say("hello")
camelia rakudo-moar f4de5f: OUTPUT«hello␤»
japhb Oh, gotcha 16:54
16:54 dakkar left, domidumont joined
japhb I'm just trying to get my local build-and-install-modules script to stop spitting a thousand warnings about "unit" during the install modules phase, just so I can concentrate on the actual *failures*. 16:55
16:56 vendethiel left 16:57 lichtkind joined
lizmat yeah, missing unit warning should be stoppable with RAKUDO_NO_DEPRECATIONS=1 now as well 16:57
working on the nqp one 16:58
16:58 vendethiel joined 16:59 anaeem1_ left 17:01 espadrine left 17:06 ShimmerFairy left 17:14 eli-se left 17:16 Hor|zon left 17:17 lichtkind left
hoelzro lizmat: I'd like to detect deprecation warnings rather than prevent them 17:18
FROGGS[mobile]: I would also like to detect nqp:: usage, and iirc, that goes through a different mechanism that deprecations
lizmat that's fine too... just don't do RAKUDO_NO_DEPRECATIONS=1 then 17:19
hoelzro noted =)
17:19 ShimmerFairy joined
lizmat yes, but it will be switch-offable in the same way soon 17:19
17:19 Hor|zon joined
hoelzro lizmat, FROGGS[mobile]: do either of you think there are other warnings I should try to detect? 17:21
17:21 vendethiel left
lizmat hoelzro: also, calling Deprecation.report in e.g. an END block, will give you all deprecations (and clean them out) 17:21
dalek kudo/nom: 27e531b | lizmat++ | src/Perl6/Actions.nqp:
nqp:: message stoppable by RAKUDO_NO_DEPRECATIONS
 17:24
17:25 _mg_ joined 17:30 vendethiel joined, diana_olhovik joined 17:33 woolfy joined 17:34 woolfy left 17:35 virtualsue joined 17:36 tinyblak left 17:37 tinyblak joined, anaeem1 joined 17:39 muraiki joined
dalek on: 532723d | paultcochrane++ | lib/JSON/Tiny (3 files):
Add unit declarator to module, class and grammar declarations

As of Rakudo 2015.05, the `unit` declarator is required before using
  `module`, `class` or `grammar` declarations (unless it uses a block). Code
still using the old blockless semicolon form will throw a warning. This commit stops the warning from appearing in the new Rakudo.		 17:39
on: 42df6e0 | moritz++ | lib/JSON/Tiny (3 files):
Merge pull request #15 from paultcochrane/pr/add-unit-declarator

Add unit declarator to module, class and grammar declarations
17:44 fhelmberger left 17:45 FROGGS joined
FROGGS o/ 17:45
lizmat FROGGS o/
FROGGS hoelzro: every warning is interesting 17:47
hoelzro FROGGS: works for me
17:50 tinyblak left 17:51 raiph left
Erendis42d ❤ 17:56
dalek p: 2844001 | FROGGS++ | src/vm/moar/QAST/QASTOperationsMAST.nqp:
map new scdisclaim op on moar
p: e7b8aef | FROGGS++ | src/vm/jvm/ (3 files):
first stab at implementing scdisclaim for jvm
p: 3135dfa | FROGGS++ | src/vm/jvm/runtime/org/perl6/nqp/runtime/Ops.java:
Check owning object wasn't disclaimed.

Port of github.com/MoarVM/MoarVM/commit/1a36596
p: 3776246 | FROGGS++ | tools/build/MOAR_REVISION:
bump moar rev for scdisclaim op
Erendis42d ❤
17:56 Erendis42d left
vendethiel o/ 17:56
FROGGS hi ven 17:57
18:00 ycaymanbo left
vendethiel irclog.perlgeek.de/perl6/2015-05-19#i_10627114 TimToady: I'll try to get a reproducible test case 18:07
I ended up using jnthn++'s Grammar::Debugger, as Debugger::UI::CommandLine was busted 18:08
.oO( duplication is GREAT )
18:08 eli-se joined 18:09 [Tux] left, telex left 18:10 telex joined
hoelzro dammit, I was using JSON::Tiny as a guinea pig for warnings =/ 18:11
18:11 andreoss left 18:14 raiph joined
timotimo oh hey everybody 18:15
FROGGS hi timotimo
hoelzro o/ timotimo 18:16
18:18 anaeem1 left 18:19 anaeem1 joined 18:20 [Tux] joined 18:21 _mg_ left, berekuk joined, berekuk left 18:22 berekuk joined 18:23 berekuk left 18:24 anaeem1 left 18:28 _mg_ joined, berekuk joined 18:37 Hor|zon left 18:44 eli-se left 18:46 Woodi left 18:47 rindolf left
dalek kudo/nom: f17b413 | lizmat++ | src/Perl6/World.nqp:
Don't load -M specs for each EVAL
 18:56
19:00 lolisa left 19:02 lolisa joined
itz_ is there no Czech or Polish Workshop this year? 19:05
19:06 anaeem1 joined
lizmat I've heard rumours about a Polish PW to be there one way or another 19:07
haven't heard anything about a 2nd Czech PW just yet 19:08
19:08 anaeem___ joined, anaeem1 left
itz_ k 19:09
19:12 domidumont left
colomon thinks he has a $work trip lined up that will allow him to attend the Swiss Perl Workshop 19:13
FROGGS colomon: \o/
19:13 _mg_ left
colomon nothing official yet 19:13
dalek pan style="color: #395be5">perl6-examples: 7c839b9 | paultcochrane++ | categories/cookbook/09directories/09-10-filename-splitting.pl:
[cookbook] remove extra semicolon
pan style="color: #395be5">perl6-examples: f496d51 | paultcochrane++ | / (4 files):
Merge branch 'master' of github.com:perl6/perl6-examples
pan style="color: #395be5">perl6-examples: 310934e | paultcochrane++ | t/categories/euler.t:
[euler] update test plan after added tests
timotimo i wonder if i should postpone the weekly until the release? 19:14
FROGGS aww
japhb timotimo: I know I would be sad, but I understand if you're feeling overloaded 19:17
timotimo "overloaded" sounds a bit too extreme :) 19:18
lizmat timotimo: please don't hesitate to call for backup :-) 19:20
timotimo hm, i wonder
19:26 kst`` is now known as kst
[Coke] is dwarring here? 19:28
(under an alias?)
timotimo he used to be "dwarring" 19:29
[Coke] roast tip: don't need to put a separate fudge for .moar and .jvm before the same test. just "rakudo" is fine in that case, it'll todo or skip both backends if you don't specify
timotimo correct 19:30
perhaps you want a different message, though?
[Coke] sure, but in the commit of his I saw, it was the same. 19:31
timotimo ah 19:32
dalek ast: 08b5de5 | lizmat++ | / (156 files):
s/throws_like/throws-like/g
 19:33
19:38 andreoss joined
timotimo .tell jnthn do we want to destroy/clean-up the sc reader when we do scdisclaim? since we demand every object in the sc anyway 19:40
yoleaux timotimo: I'll pass your message to jnthn.
dalek kudo/cunion: 67245d3 | FROGGS++ | / (4 files):
add support for unions in NativeCall using CUnion repr
kudo/cunion: bf3a3be | FROGGS++ | t/04-nativecall/13-union. (2 files):
add tests for C union support for NativeCall
kudo/cunion: 9186fdf | FROGGS++ | t/04-nativecall/13-union. (2 files):
test inlined CStructs in CUnions
kudo/cunion: 0d075f3 | FROGGS++ | t/04-nativecall/13-union. (2 files):
fix bogus test (wrong alignment)
vendethiel FROGGS++!
FROGGS just a rebase + a new commit :o) 19:41
(because I want to merge in a bit)
TimToady have you checked that out with the release manager :)
vendethiel ah, I thought you merged it all. misread!
FROGGS TimToady: ;o)
one last test and then it goes in 19:42
timotimo very cool
FROGGS would be nice if an apple device owner could test it afterwards 19:43
timotimo i'm not sure any of these people actually own their device
:P
FROGGS (I did get a confirmation of the branch on osx two days or so ago)
hehe
timotimo: maybe that is why it is often called 'user'
?
timotimo hehe
colomon FROGGS: you want confirmation that cunion branch works on OS X? 19:44
FROGGS colomon: that'd by cool, yes
colomon: MoarVM/cunion, nqp/master, rakudo/cunion
colomon just did rakudobrew build moar cunion, should that work or no? 19:45
timotimo i don't think that gives you moarvm/cunion as well as rakudo/cunion
dalek kudo/nom: 97b4db0 | lizmat++ | t/0 (5 files):
Some more kebabcasing test functions

Except is_approx, because it needs special attention?
FROGGS colomon: yeah, that wont pick up the right moarvm 19:46
colomon bother
19:46 brrt joined, eli-se joined
FROGGS colomon: np, you can test nom/master in a sec :o) 19:46
19:46 ollinator joined
colomon FROGGS: oh, sure, that’s easy enough. ;) 19:46
timotimo hehehe 19:47
TimToady we hope...
colomon is trying “rakudobrew triple cunion master cunion" in the meantime
timotimo you'll hardly break anything by merging that; it should only touch new code, right?
colomon and “triple” fails early on.
timotimo huh
dalek p: a56e469 | FROGGS++ | tools/build/MOAR_REVISION:
bump moar for CUnion repr
colomon I think it’s a triple bug, not cunion
FROGGS timotimo: the Wurst that could happen is that make test fails 19:48
ohh wait...
let's not merge the rakudo branch until I've implemented it on JVM O.O
timotimo mhm
FROGGS damn 19:49
I tend to forget that
lizmat FROGGS: well, at least you won't make the release manager angry ;-)
colomon the good news is that “rakudobrew build moar cunion” ought to get the correct nqp/moar now, right?
lizmat at least, I hope
FROGGS colomon: no, domt think so either 19:50
dont*
colomon well bother
FROGGS colomon: it is okay, it will get tested at some point 19:51
ollinator does anybody know how far perl6 is? can one use it for production scripts - or should you use perl5 instead?
brrt ehm
lizmat ollinator: yes 19:52
brrt we're releasing this christmas, so i'd guess you can use it already
but it depends really on what you mean by production
lizmat brrt++ for a more friendly answer
masak the answer depends on the person, and that person's uses/requirements 19:53
brrt amd what you consider 'production ready'
masak that's a less "fun" answer than just yes/no. but it's more accurate
ollinator okay, well i'm using perl5 to automate data transmission, e.g. sftp, and logging everything to a mysql db, that works perfectly. But I'd like to upgrade to perl6, don't know how compatible it is to version 5 19:54
FROGGS using it for on/off scripts, data mangling tasks etc: ohh yes! - using it for daemons: well, maybe not yet
ollinator Well okay, I have to read around about it, thanks anyway
timotimo FROGGS: i think the term you're looking for is "one-off" rather than "on/off" 19:55
masak ollinator: sounds like Perl 6 might be ready for you :)
FROGGS ahh, perhaps
brrt ollinator: i think perl6 can do that and more; but i would also point out that perl6 is not a dandified perl5 19:56
ollinator okay, but it's a daemon that i wrote, using net::sftp::foreign, dbi as modules, that kind of stuff
literal ollinator: it would not be an upgrade in the traditional sense, any more than "upgrading" from Javascript to Lua. Perl 5 and 6 are very separate languages, with a common root
ollinator okay, i see 19:57
thanks a lot for this introduction
brrt i'm not very familiar with net::sftp::foreign, is that using an external program?
literal especially performance-wise, at this stage. Perl 6 is not quite there yet
19:57 lolisa left
masak well, Perl 6 is very much an "upgrade" (just a non-backwards-compatible one) on Perl 5 from 2000. but, crucially, so is Perl 5 of today -- except it's backwards-compatible. 19:57
ollinator yes it is using the native openssh client that ships with the linux distro 19:58
dalek ast: d36a7d8 | usev6++ | S04-phasers/pre-post.t:
Make some tests more specific
FROGGS bartolin++
brrt ollinator: then I supsect we can get there already somewhat using Proc::Async
doc.perl6.org/type/Proc::Async is the default class for communicating with subprocesses 19:59
timotimo could still be nice to have someone build a port of that module 20:00
FROGGS .tell someone Please port that module 20:01
yoleaux FROGGS: I'll pass your message to someone.
20:01 Woodi joined
timotimo haha 20:01
FROGGS .oO( Owl post again. )
20:02 virtualsue_ joined, virtualsue left, virtualsue_ is now known as virtualsue
timotimo .. owl ... post? 20:03
ollinator okay i see perl6 is quite different from previous versions. but i'm interested in learning, so i'll have a closer look. thanks for the hint
thanks @everybody 20:04
timotimo cool
brrt good luck
timotimo keep us posted with your progress and questions :)
masak come back soon!
timotimo input about our learning material is always welcome
we know there's some rough edges
FROGGS timotimo: that's a Harry Potter chapter 20:05
timotimo ah!
eulenpost
dalek ast: 4005f4d | lizmat++ | S (12 files):
s/cmp_ok/cmp-ok/g
FROGGS aye
masak .oO( cat-head eagle post ) 20:06
jdv79 is all the testing stuff doc'd in a central place? 20:08
it would be nice if we tested all the things and it was easily knowable/findable
20:09 ollinator left
jdv79 basically like p5's cpantesters but with better reports maybe - idk 20:09
[Coke]: ^^^ (forgot to address that burst) 20:13
japhb jdv79: I was literally in the middle of suggesting you talk to [Coke] when you sent that last line. :-) 20:14
[Coke] I don't think such a thing exists, no. 20:16
dalek ast: b19b0bf | usev6++ | S04-phasers/pre-post.t:
Fix precedence for fudged test
 20:17
20:17 darutoko left 20:18 diana_olhovik left
FROGGS jdv79: we have something like cpantesters, testers.perl6.org - feel free to create feature request issues 20:18
20:18 raiph left
jdv79 docs 20:19
all i know about that is that if i go to that url i see reports in a semi-confusing UI
andreoss [ptc]: i want add a handler for =RESULT section to p6-examples, in order to place results of PE tasks under a spoiler tag, is it ok? 20:20
jdv79 who runs that stuff?
20:20 diana_olhovik joined
FROGGS jdv79: so you are missing information like these? wiki.cpantesters.org/ 20:20
jdv79: me 20:21
20:23 xfix left 20:24 anaeem___ left
[ptc] andreoss: go for it 20:24
20:24 anaeem1 joined
[ptc] andreoss: if you can add tests for the new feature, that'd be good too :-) 20:24
jdv79 the first thing that comes to mind is having all the major rakudo smokes (moar,jvm x t,spec,stress) somewhere 20:25
also where is there info about what happened here: testers.perl6.org/reports/47409.html
[ptc] andreoss: thanks for adding more examples and tests for them! However, could you update the plan in the tests as well? It was missed the last few times 20:26
20:26 diana_olhovik left, al2o3-cr joined
jdv79 where are the reports sourced? 20:27
i guess docs and rakudo smoking are my main interests 20:28
FROGGS jdv79: for report 47409, the report client (panda) did not record any data... dunno why, would be interesting to reproduce and fix
jdv79 is that what happens when panda was run with --notests? 20:29
[ptc] andreoss: also, I've been wondering if we should remove the expected results from the Euler examples, since the Euler people actually really don't like it having the results being published publically
dalek ast: 0cbc065 | lizmat++ | S (134 files):
s/eval_dies_ok/eval-dies-ok/g
jdv79 and does panda always report?
FROGGS rakudo smoking happens in [Coke]++'s repositories, and docs are at doc.perl6.org/language/testing
timotimo not always; you need to set an env var
FROGGS jdv79: might be about --notests... dunno offhand
20:29 anaeem1 left
FROGGS like in: PANDA_SUBMIT_TESTREPORTS=1 panda install Flubber 20:30
jdv79 is that doc'ed somewhere?
FROGGS I fear not
only mentioned in a blog post
bartolin jdv79: I also run daily rakudo smoke tests (on FreeBSD), based on [Coke]++'s scripts: github.com/usev6/perl6-roast-data
FROGGS I know I need to add pages to testers.perl6.org that explain that 20:31
jdv79 misses the maturity of the p5 world
FROGGS no time yet though :o(
jdv79 :(
much to be done!
hopefully not too much
FROGGS jdv79: yes, and we are just a handful of ppl
jdv79 i can work on this stuff
bartolin ++jdv79
FROGGS ++jdv79 20:32
gnight #perl6
brrt any help is highly appreciated :-)
20:32 FROGGS left
bartolin g'night, FROGGS 20:32
lizmat gnight FROGGS[mobile] 20:33
andreoss [ptc]: i guess removing them would be even better than hiding them under a spoiler tag
[ptc] andreoss: most certainly much easier :-)
andreoss: one could certainly look in the tests for the answers, however they are somewhat "hidden" there and then don't turn up on the website
andreoss: sorry if I spoiled your fun! 20:34
brrt goodnight, all
20:35 brrt left 20:38 kurahaupo1 joined
andreoss [ptc]: is the site updated automatically or manually? 20:39
TimToady another protip, you don't have to put #?USES on every call to a function; you're supposed to be able to put it onto the function itself, as long as it parses like one of the built-in tests 20:41
at least, I hacked that into fudge at one point, so maybe it still works 20:42
[ptc] andreoss: automatically. There's a cron job which checks for new updates every 5 min 20:43
20:44 woolfy joined, woolfy left
andreoss [ptc]: it's been almost 2 weeks since last updated 20:45
dalek ast: a4cac76 | usev6++ | S04-phasers/ (2 files):
Use unique ticket number for similiar test failures
 20:49
[ptc] andreoss: hrm, that's good to know 20:51
20:51 chansen joined
[Coke] (rakudo smoking) I need to be fix this so I can run the perl5 integration tests. :P 20:52
same issue as being able to run the linenoise repl out of the ./install dir
20:53 kbenson_ left
[ptc] andreoss: examples.perl6.org/build-log/ is a good place to look for errors. Wish I'd looked earlier... 20:53
20:54 atweiden joined, dolmen joined
dalek ast: 709e414 | lizmat++ | S (200 files):
s/dies_ok/dies-ok/g
 20:54
20:57 kurahaupo1 left
itz_ I bought this thinking it would be more reliable than wikipedia 20:59
www.britannica.com/EBchecked/topic/899428/Perl
Canadian-born programmer and linguist Larry Wall 21:00
eh
21:01 skids left 21:02 kurahaupo1 joined
[ptc] andreoss: great work with the lisp interpreter! 21:03
21:03 koo6 left
andreoss x 21:03
thanks
isBEKaml [ptc]: what? where's the lisp interpreter?
:-) 21:04
[ptc] isBEKaml: in the perl6-examples repo, in the categories/interpreters dir :-)
andreoss: it seems that the perl6 on hack.p6.org is too old for the perl6 code in the examples *sigh* 21:05
21:05 mephinet- is now known as mephinet, mephinet left, mephinet joined
isBEKaml [ptc]: Ah, it's been a while since I gpull'd anything.. Let me do this now :-) 21:05
21:06 raiph joined
[ptc] andreoss: looks like I'll have to do some testing on hack in order to get things back going again 21:06
andreoss [ptc]: though i guess implementing list interpreter as a slang would be more perl6 way
21:06 mephinet left, mephinet joined, raiph left, raiph joined
dalek pan style="color: #395be5">perl6-examples: 34504cc | paultcochrane++ | categories/interpreters/lisp.pl:
[interpreters] replace tabs with spaces

   - add vim coda
   - remove trailing whitespace		 21:07
pan style="color: #395be5">perl6-examples: ed9570f | paultcochrane++ | categories/euler/prob0 (3 files):
[euler] further whitespace, formatting cleanups
timotimo [ptc]: don't you just need to update the interpreter?
isBEKaml andreoss++ # 8 day old lisp interpreter!
timotimo or downgrade the Pod::To::HTML distro? 21:08
isBEKaml [ptc], andreoss: Not a problem really - just thought I'd let you know that lisp.pl has mixed tabs and spaces. Looks a bit funny in (old) vi 21:09
andreoss isBEKaml: oh, it's probably because i've edited with both vim and emacs (for indenting) 21:11
[ptc] timotimo: do you mean on hack? 21:12
timotimo yos
[ptc] ähm, I just use the one moritz++ installed there
isBEKaml andreoss: no worries - hack away on lispy!
[ptc] my guess is that it'll be updated after Thursday
timotimo well, rakudobrew is installed there, too :) 21:13
lizmat commute home&
[ptc] true
21:13 lizmat left
timotimo i would have petitioned to replace the "class Foobar;" form with braces for the time being 21:13
[ptc] maybe that's something for the morrow
timotimo because that will work on older and newer versions of rakudo
[ptc] the discussion with masak and FROGGS ended with a consensus for the blockless form... 21:14
at least I haven't updated all of the modules I'd planned to have submitted PRs for by now :-)
timotimo oh
OK
BBIAB
21:15 kurahaupo1 left, chansen left
jdv79 nine: around? 21:16
21:18 muraiki left
[ptc] andreoss: should I not have removed the tabs from lisp.pl? 21:19
andreoss: sorry if I've broken anything!
dalek pan style="color: #395be5">perl6-examples: 889bd65 | andreoss++ | categories/euler/prob06 (2 files):
[euler] pod formating
 21:37
RabidGravy does nqp::watchfile() poll or does it use some inotify thing if it's available
jdv79 .tell nine please see the PR on I::Py 21:43
yoleaux jdv79: I'll pass your message to nine.
flussence would like it if `panda smoke` also accepted --auth= so I could be super-lazy about keeping my own stuff un-bitrotted 21:44
timotimo RabidGravy: it uses whatever libuv offers
[ptc] hacked, hack. Now the examples build and deploy again. 21:48
unfortunately, it's a hack :-/
RabidGravy which is what I suspected, but it doesn't tie up with the cpu usage I am seeing, I'm wondering if it is built in such a way as not to use inotify 21:49
tony-o_ m: gist.github.com/tony-o/ad52076865a21edec940 21:50
camelia rakudo-moar 97b4db: OUTPUT«IO::Socket::INET is disallowed in restricted setting␤ in sub restricted at src/RESTRICTED.setting:1␤ in method new at src/RESTRICTED.setting:32␤ in block <unit> at /tmp/f1aLK_iZoo:5␤ in any <unit-outer> at /tmp/f1aLK_iZoo:1␤␤»
21:52 espadrine joined
RabidGravy flussence, you could parse the modules list and grab the ones with the appropriate auth 21:54
21:54 kaare_ left 21:55 [Sno] left
flussence oh, didn't notice --verbose there. That'll do fine. 21:56
.oO( next job: actually *finish* a few of these half-baked modules sitting on disk so I can put them up there... )
 21:57
timotimo RabidGravy: we currently have something in place that makes the event loop unable to not spin 22:00
RabidGravy right
timotimo RabidGravy: jnthn is going to solve it. on my local moarvm i have an usleep in the function that spins the most and it makes the cpu usage almost 0
22:01 berekuk left
RabidGravy with more than one core it isn't a problem 22:04
;-)
22:10 RabidGravy left 22:11 eli-se left, vendethiel left 22:15 labster joined 22:29 [Sno] joined 22:31 vendethiel joined 22:35 mohij left 22:40 andreoss left 22:43 skids joined 22:44 raiph left 22:46 zcm joined, bjz left 22:51 gfldex left
dalek kudo/nom: 03bf80f | TimToady++ | src/Perl6/World.nqp:
don't say TTIAR unless next thing looks like term
 22:55
22:59 dayangkun joined 23:03 zcm left 23:07 virtualsue left 23:16 vendethiel left 23:19 spider-mario left, spider-mario joined
TimToady m: say True ¿¿ 1 ¡¡ 2; 23:20
camelia rakudo-moar 03bf80: OUTPUT«5===SORRY!5=== Error while compiling /tmp/8Po7V7DShH␤Confused␤at /tmp/8Po7V7DShH:1␤------> 3say True7⏏5 ¿¿ 1 ¡¡ 2;␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ state…»
TimToady m: say True¿;
camelia rakudo-moar 03bf80: OUTPUT«5===SORRY!5=== Error while compiling /tmp/NX9jVCyVye␤Bogus postfix␤at /tmp/NX9jVCyVye:1␤------> 3say True7⏏5¿;␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement mo…»
TimToady those are both better than TTIAR
23:31 Vlavv left
dalek kudo/nom: 4ed69d2 | TimToady++ | src/Perl6/Grammar.nqp:
"parameterized" is misleading

When the parameters to block are optional and, in fact, not supplied by the user, it's rather confusing to say you were expecting a "parameterized block", which sounds as though the parameters are required.		 23:34
TimToady specifically, in regard to: 23:38
m: if 1.so? {...}
camelia rakudo-moar 03bf80: OUTPUT«5===SORRY!5=== Error while compiling /tmp/0FlafOJJYV␤Missing block␤at /tmp/0FlafOJJYV:1␤------> 3if 1.so7⏏5? {...}␤ expecting any of:␤ infix␤ infix stopper␤ parameterized block␤»
TimToady "Why does the block need parameters?"
23:42 zcm joined, zcm left 23:43 Vlavv joined 23:44 zcm joined, [Sno] left
TimToady anyway, will now say 'block or pointy block' 23:44
23:45 [Sno] joined 23:47 zcm left 23:49 zcm joined, zcm left 23:50 zcm joined, zcm left 23:51 zcm joined, zcm left 23:52 zcm joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!