»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg camelia perl6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by sorear on 25 June 2013.
TimToady slurp-rest is good; it shows relationship with slurp while indicating it is a lower-level operation by being longerly huffmanized 00:00
BenGoldberg I'd also suggest an optional parameter, :close-after-eof, which causes slurp-rest to close the handle when it's done. 00:01
TimToady as usual, the presence of a named option is a design smell
Tekk_ BenGoldberg: why would you need that? 00:02
BenGoldberg shrugs. It was just an idea.
Tekk_ you've already read the whole file
TimToady let's use language composition, not add options that reinvent things 00:02
this is not a common enough operation to warrant the mental overhead of another option 00:03
BenGoldberg One might have just read the whole file; however, one might want to seek backwards and re-read parts of the file.
TimToady or append lines 00:04
BenGoldberg That, too.
Or perhaps another program might be appending lines, and you want to read the new lines whenever the file changes size. 00:06
lizmat sleep& 01:00
leont lizmat: it's one of the first things to get back-ported to CPAN, but IMO it should be core 01:08
timotimo hey 01:38
hey folks
hey
leont Did some ground-work for that in Salzburg, didn't quite finish it :-/
timotimo i made a thing
t.h8.lv/add_core_op.svg
look at my thing
like my thing!
it's version 1 of my thing 01:39
next step: connect multiple versions of the same register 01:43
then: get PHI to appear properly
and lastly: figure something out for before/after comparisons
raydiak timotimo++ awesome thing! 01:54
_sri is that serious? fosdem.org/2015/schedule/event/get..._to_party/ 02:35
raydiak 2015 is the officially stated goal, yes 02:37
_sri faints
timotimo t.h8.lv/add_core_op_with_dataflow.svg
raydiak perhaps "goal" is a bit strong..."stated likelihood"? :) 02:39
irclog.perlgeek.de/perl6/2014-11-06#i_9622354
timotimo: ooh even more awesome
japhb timotimo: Why do some nodes have a black successor arrow and others have a grey one? 02:41
japhb feels like _sri 02:42
(Get ready to party!)++
timotimo japhb: because i forgot to put the grey in some other places :) 02:45
japhb guesses that the answer to his previous question is "arrows in and out of exit_* are black, other successors are grey"
timotimo it's a glorious mess of special cases :D
japhb Heh
timotimo also, arrows out from jumping instructions are black as well
(which i think i'm going to keep)
raydiak added a bin/symbolic script to Math::Symbolic for manipulating equations from the command line 02:48
timotimo t.h8.lv/add_core_op_extops_and_dataflow.svg ← now has a limited understanding of how our extops work 03:04
(it just assumes they all write their first register arg and read all others)
ventica 'o' ~ '/' 03:48
camelia: 'o' ~ '/'
ventica lost my touch 03:49
timotimo t.h8.lv/graph_of_gimme.svg - latest version i've built - this time with a much bigger spesh graph
raydiak m: say 'o' ~ '/'
camelia rakudo-moar 315ec6: OUTPUT«o/␤»
ventica ty raydiak lol
raydiak yw ventica :)
ventica masak: Congrats on your GOTO talk... I found it truly awesome! 03:51
raydiak ugh another precomp bug 04:22
[Tux] Wow, «rakudobrew build moar» automatically also fetches/updates the modules installed by panda. Less to watch :) 07:47
Ven o/, #perl6 09:44
looks like that fosdem "announcement" (or "ambiguity") made quite some noise..
moritz sighs 09:55
yakudza hello 09:59
Ven, what announcement 10:00
?
nwc10 how is NSA?
Ven helo yakudza
yakudza: no announcement, only misreadings :)
Ulti fosdem.org/2015/schedule/event/get..._to_party/ this?
yakudza Very nice. For long time watching at perl6 I cant find use cases for it. Is there any articles where some one describes where perl6 is better that perl5 and maybe some use cases. 10:06
moritz yakudza: for one thing, parsing is much nicer in Perl 6. See for example github.com/moritz/json/blob/master...Grammar.pm for a complete JSON parser in Perl 6 10:09
nwc10 yakudza: perl 6 has proper multi-core concurrency
moritz yakudza: another thing that Perl 5 is very bad is is concurrency; Perl 6 (at least rakudo on MoarVM and on JVM) are already better here
nwc10 no ithreads, no GIL, etc 10:10
moritz also Perl 5 makes it far too easy to mix text and byte strings, causing trouble with encodings to be prevalent 10:13
yakudza github.com/koorchik/formula-evalua...-benchmark 10:15
Why perl 6 is so slow ? because of startup time ? 10:16
perl5 version is not optimized there. 10:17
anyway perl6 takes much more time
moritz yakudza: currently list iteration is still very slow 10:26
there's a project under way to fix that
FROGGS yakudza: startup time (for rakudo on moarvm) is very good, though as moritz said lists are quite slow, but also string operations 10:48
yakudza: lists will be refactored soonish to improve performance
azawawi hi everyone 10:51
how do i delete a key from a hash?
%in-stock.delete($name); # which is mentioned in the Perl 6 book is not working any more 10:52
and looking at the Hash class documentation does not help also
moritz azawawi: %in-stock{$name}:delete 10:53
azawawi very confusing
the question is why does not the hash class has a remove or delete method to it? 10:54
lizmat azawawi: it used to have 11:02
but Perl6 allows something to be both accessible with [] *and {}
the term "delete" was ambiguous, as it was used both for [] as well as {}
so they got separated into "delete_pos" for [] and "delete_key" for {} 11:03
so you still *can* do %hash.delete_key('onlyonekey') 11:04
but that should probably be considered an unspecced rakudo implementation detail
azawawi i see
but it is still confusing :)
lizmat please note that you can combine adverbs 11:06
and make them conditional
and do it with more than one key
%hash<please delete all of these keys>:delete($really)
lizmat would only delete if $really were true 11:07
azawawi im working with proc::async to drive the perl6-debug-m process to step in
in farabi6, so far it is working well
lizmat cool, azawawi++
azawawi now im trying to keep promises from ending :)
lizmat ?
azawawi A cool comment # Add a user friendly message to signify the demise of a promise :)
dalek kudo/nom: a0304c3 | (Elizabeth Mattijsen)++ | src/core/ (2 files):
Deprecate $handle.slurp -> slurp-rest

As discussed on IRC between
   irclog.perlgeek.de/perl6/2014-11-06#i_9621920
and
   irclog.perlgeek.de/perl6/2014-11-07#i_9622383		 11:12
lizmat azawawi: when you break a promise, you can state why that is 11:15
are you saying that somewhere in the bowels of Proc::Async promises are broken with less then user friendly messages? 11:16
azawawi no at all
lizmat or are you looking something more qmail like, like "Sorry it didn't work out" ? :-)
azawawi im working on github.com/azawawi/farabi6/issues/16 11:17
so far step in till end is working as promised :)
with async output
im building rakudo on feather to demo it :)
lizmat again, cool!
azawawi im parsing jnthn++ debugger output and rendering it on the web... until we can figure out a good json format 11:19
if you think of it... the core debugger should not use ANSI output... it should work as an API for other UIs 11:20
azawawi maybe exchanging JSON packets over a TCP port or something 11:20
so that stdout and stderr is pristine
azawawi installs Panda and friends :) 11:21
azawawi feather.perl6.nl:3030/ # open debug_01.pl and click step in... It is not much but it is a beginning 11:26
lizmat looks 11:29
azawawi step in takes time like 1 sec 11:30
because of it is polled on a 1 sec timer
because of = because
do not try prompt... because it will stall ofcourse
lizmat how do I open debug_01.pl? 11:31
azawawi alt-o and then type debug_01.pl 11:32
you can write simple stdout programs and they will work :) 11:33
lizmat did that, and then step in: I only get "debugging finished"
azawawi older session
lizmat well, I'm about to go out the door for the rest of the day: see some of London :-)
azawawi just refresh :)
have fun 11:34
lizmat away& 11:39
brrt wtf: fosdem.org/2015/schedule/event/get..._to_party/ 12:07
TimToady++ for guts 12:08
espadrine_ note the first comment of news.ycombinator.com/item?id=8571837 12:18
brrt oh.. so it's not really happening?
TimToady is not speaking as FOSDEM? 12:19
espadrine_ yes he is
brrt that's good enough for me :-) 12:21
pmurias vendethiel: the only sensible way I can see Perl 6 being handled by an IDE is reusing the compiler for parsing 13:18
vendethiel: and one the code is parsed by a real compiler and the BEGIN blocks are executed the metamodel stuff should be that harmfull 13:19
azawawi hmmm 14:15
is the step out functionality in perl6-debug not implemented or has a bug? 14:16
azawawi it is telling me that "im not in a routine" even though im in a sub :) 14:16
azawawi m: sub sum($a, $b) { my $sum = $a + $b; return $sum; }; say sum($a, $b); 14:17
camelia rakudo-moar 315ec6: OUTPUT«===SORRY!=== Error while compiling /tmp/4w02g7DXEG␤Variable '$a' is not declared␤at /tmp/4w02g7DXEG:1␤------> m = $a + $b; return $sum; }; say sum($a⏏, $b);␤ expecting any of:␤ postfix␤»
azawawi m: sub sum($a, $b) { my $sum = $a + $b; return $sum; }; say sum(1,2);
camelia rakudo-moar 315ec6: OUTPUT«3␤»
azawawi is it a known issue that perl6-debug-m doesnt step out? 14:20
timotimo i wrote that "we should nominate random outsiders to be spokespersons" many hours before i actually saw the fosdem talk abstract 14:21
azawawi timotimo: hi there
timotimo greetings
t.h8.lv/add_core_op_extops_and_dataflow.svg ← have you seen this? i'm quite proud of it :) 14:22
azawawi timotimo: i just implemented step out functionality in farabi6... it is buggy but shows you async stdout/stderr...
timotimo: debug step in... that is
timotimo that's very good!
azawawi and an answer to your previous question, markText basically marks it for ever... so if you move it... not a big deal... it is basically a CSS class.. 14:23
timotimo thought so. that's good! 14:23
azawawi im implementing codemirror.net/demo/lint.html atm
perl6 async linting (syntax checking) 14:24
timotimo oooooh
gtodd nice 14:25
gtodd it would be cool if something like that could work in the regular REPL :-D 14:27
gtodd I'd like to have my REPL flashing red and green as I typed ... as long as it was an option heheh 14:29
timotimo kind of like fish's prompt?
gtodd oh yeah
async perl6critic .... 14:30
... "you're not going to type that without a putting a ":" in front of it are you??"
azawawi well that's doable... :) 14:31
gtodd :-D 14:32
azawawi gtodd: care to add an issue for it github.com/azawawi/farabi6/issues ? :)
gtodd some yak shaving would be involved ... i.e. there is no perl6critic 14:33
or P6BP :-)
timotimo "BP"? 14:34
PerlJam Best PRactices 14:35
timotimo oh
of course
gtodd Damien's Perl Best Practices
timotimo PEP6 :P 14:36
moritz gtodd: there's github.com/moritz/perl6-wtop 14:42
gtodd moritz: ++ :-) 15:01
alexghacker lizmat++ for slurp-rest, tyvm 15:07
gtodd moritz: I think raiph had some ideas for either a bot or a filter that would automatically create interesting and documented .gists or examples from this channel ... just by adding tags or .tell-ing the bot to grab a section of text (or maybe marking it for later in the logs?) maybe that would work perl6-WTOP :-) 15:25
moritz: WTOP #1 create a WTOP and advertise it on IRC ... (this WTOP will self destruct once the N of WTOP > $somenumber) 15:27
brrt i should know better than to read hn comment threads too 15:28
moritz gtodd: huh? we have a WTOP 15:29
github.com/moritz/perl6-wtop/blob/...ctices.pod 15:32
two, in fact :-)
gtodd :-D 15:33
I guess the WTO part is important 15:34
timotimo brrt: they don't call it "hater news" for no reason.
brrt: have you seen my svgs? :)
brrt i have
very nice
timotimo++ 15:35
timotimo :3
something i should add?
i was trying to figure out a way to annotate register versions with facts flags
without cluttering things too much
and i'm not entirely sure what PHI nodes should look like
brrt ehm, phi nodes should look like they are the joining of other registers, which is what they are 15:36
i thought they were clear enough
[Coke] moar non jit failed a few more tests today: 32089 total. 15:37
parrot, still clean. 15:38
brrt that looks a bit bad
colomon how many tests are there total/ 15:40
?
timotimo i know what phi nodes do
i just don't want them to cause a gigantic swarm of arrows between blocks
brrt i don't think you have a way arround that 15:42
is what i mean
brrt i'd expect them to look the way they do :-) 15:42
timotimo i'd kind of like to have a row of registers at the bottom of each block and one at the top that would then "gather" the arrows for the registers up 15:43
Ovid oi 15:43
timotimo but the problem with that is that i don't get to see the phi nodes until later
hello ovid!
Ovid Hm, where did “oi” come from?
Hello, timotimo.
hoelzro maybe shortened from ahoj? 15:45
[Coke] Hey, greybeard.
moritz m: say 'Ovid'.lc.comb(/<[aeiou]/).join
camelia rakudo-moar 315ec6: OUTPUT«===SORRY!=== Error while compiling /tmp/vOW1YGG4W1␤Unable to parse expression in metachar:sym<assert>; couldn't find final '>' ␤at /tmp/vOW1YGG4W1:1␤------> say 'Ovid'.lc.comb(/<[aeiou]⏏/).join␤ expecting a…»
moritz m: say 'Ovid'.lc.comb(/<[aeiou]</).join
camelia rakudo-moar 315ec6: OUTPUT«===SORRY!=== Error while compiling /tmp/ntL730wMkW␤Unable to parse expression in metachar:sym<assert>; couldn't find final '>' ␤at /tmp/ntL730wMkW:1␤------> say 'Ovid'.lc.comb(/<[aeiou]⏏</).join␤ expecting …»
moritz m: say 'Ovid'.lc.comb(/<[aeiou]>/).join 15:46
camelia rakudo-moar 315ec6: OUTPUT«oi␤»
moritz being able to type sure helps a lot :-) 15:46
brrt hmm i see timotimo
arnsholt curtispoe: Wikipedia claims "oi" is h-dropped "hoi", which dates at leastd back to Middle English apparently =)
arnsholt Potentially related to "hei" in Scandinavian, I guess 15:47
And English "hi", for that matter
dalek rl6-roast-data: 43ad309 | coke++ | / (4 files):
today (automated commit)
curtispoe Heh.
brrt and dutch 'hoi' 15:50
[Coke] curtispoe, what brings you to #perl6? Anything we can help with? 15:51
lucas_ hello 15:51
is perlcabal down?
moritz lucas_: nope
curtispoe I was just curious what people thought about the FOSDEM annoucement. I had known about it, but I thought it was being kept under wraps.
lucas_ moritz: strange... 15:52
must be a local problem here
[Coke] what FOSDEM announcement? 15:53
moritz lucas_: www.isup.me/perlcabal.org
curtispoe This has been making the rounds on Facebook, Ycombinator, and Reddit: fosdem.org/2015/schedule/event/get..._to_party/ 15:54
[Coke] Huh. someone actually said that?
curtispoe Well, it has Larry’s name on it :)
[Coke] I note that it's not attributed.
he's listed as a speaker there.
brrt i'm excited for one :-)
curtispoe I’ve never spoken at a conference where the description of my talk wasn’t written by me or approved by me. 15:55
[Coke] It's not like we had a cabal meeting and discussed it. looks to me like whoever wrote that announcement took some liberties. 15:55
timotimo curtispoe: i only saw someone quote the text (not who it was from) on irc and jokingly said "we should nominate a random person as spokesperson" and people linked to that part of the irc logs, too %) 15:56
[Coke] ok, so it looks like larry took some liberties. seems like a reasonable guess, though.
curtispoe: it'll go faster if you help, though .grab a shovel willya? 15:57
brrt :-)
curtispoe Veure will go faster if you help. Can you grab a shovel? :)
Believe me, if I had the time, I would dearly, dearly love to do more to help Perl 6 besides writing blog entries. 15:58
colomon curtispoe: I hear you. :) 15:59
curtispoe (And for those who didn’t get the “Veure” reference: I’m writing a ctext-based MMORPG in Perl blogs.perl.org/mt/mt-search.fcgi?bl...;limit=20)
brrt nice 15:59
moritz has fun with rabbitmq 16:00
curtispoe And it’s a business I’m pouring money into, not just a hobby. Hence, my inability to spend too much free time on other stuff.
timotimo what is ctext?
brrt well, i'll be sure to be there @fosdem :-) 16:07
curtispoe I should be too. 16:09
curtispoe Now I just need to figure the topic I’ll be speaking on. 16:09
brrt how you've become a perl6 convert? :-P 16:10
regards 16:11
timotimo FROGGS[mobile]: i see a fail for GTK::Simple, but the test stage is kind of not very helpful. it just says "failed tests: 1, 3"; maybe panda could be made to re-run the tests with --verbose when they fail and report the result of that? 16:21
FROGGS[mobile]: and maybe we could have a different grade for "this thing has no tests at all" rather than just failing 16:22
colomon timotimo: agreed on "this thing has no tests", though up until recently I had a module that had tests but no actual module! ;) 16:37
timotimo hah 16:38
colomon ABC module is failing tests, but only when compiled. :\ 16:39
colomon looks like it might be a Rakudo bug? 16:41
line is my $match = ABC::Grammar.parse($test-case, :rule<element>, :actions(ABC::Actions.new));
error is
Could not find symbol '&Actions' 16:42
in method <anon> at src/gen/m-CORE.setting:13707
in any find_method_fallback at src/gen/m-Metamodel.nqp:2725
in any find_method at src/gen/m-Metamodel.nqp:988
in block <unit> at t/07-stringify.t:35
timotimo oh, that is not good
colomon but I've done "use ABC::Actions", which defines the class ABC::Actions
timotimo hm. how do we debug something like that?
colomon well, based on the smoke logs it looks like it started three days ago
that ought to help narrow down what rakudo commit introduced the bug 16:43
timotimo sure, that's good
hm, i wonder if you could make emmentaler also report to cpandatesters?
colomon probably? 16:44
timotimo dunno if emmentaler would accept just changing the panda branch to "report" or what it's called 16:45
(and setting that environment variable)
colomon I'll try to remember to give it a try sometime soon. :) 16:47
timotimo cool, froggs would probably like that :)
colomon is having the kind of week where bugs accumulate rather than getting fixed. 16:49
okay, 2b6e9e75cf272934acf42bd046705c515a6e345f is good
timotimo i'm having the kind of week where motivation seem to fluctuate heavily from one moment to the next 16:55
colomon moar-d2e301e24ad7ed51484a16ed81db81df9f724e66 fails 16:55
liztormato [Coke]: Fwiw, I'm pretty sure Larry wrote that description himself
timotimo hey liz :) 16:56
[Coke] I wish I could get a week where I could hack on p6 stuff. I suspect I'd be a lot less stressed out. :)
liztormato: I'm kind of pissed that there wasn't a discussion about it.
[Coke] ah well. c'est le larry. 16:56
liztormato why? 16:57
dalek rl6-bench/stress: 3e2d044 | (Geoffrey Broadwell)++ | minibenchmarks.pl:
Fix scaling parameters for divide-and-conquer test
rl6-bench/stress: fa5cc93 | (Geoffrey Broadwell)++ | / (2 files):
Allow better testing of flappy compilers by specifying max failures in a single SCALE before stopping that test for that compiler
rl6-bench/stress: a876e9a | (Geoffrey Broadwell)++ | / (2 files):
Support timeouts in timeall and bench
rl6-bench/stress: fba2e90 | (Geoffrey Broadwell)++ | README.md:
Update README.md: Mention quickstress, s/Capture::Tiny/IPC::Run/
liztormato People have been giving presentations about Perl 6 for years without community discussion
liztormato So why can't Larry? 16:59
[Coke] liztormato: there is a difference between me giving a presentation about cool bits of perl6 or saying "look how much faster we are" and -larry- saying we're shipping by this date. 16:59
timotimo oh 17:00
japhb Speaking of stress, the perl6-bench/stress branch could use a little cross-platform testing, because I had to change a rather finicky bit of platform independence from using Capture::Tiny to IPC::Run.
timotimo no he isn't saying we'll be shipping by that date 17:01
[Coke] so, two bits. 1) things larry say are going to have an impact outside of our echo chamber. (which they obviously did, see ovid's appearence today), and 2) talk about ship dates is something that we, as a community, should probably be careful of in general.
timotimo: squint, and pretend you're outside the echo chamber. I think you can read it that way.
timotimo irony of today: a script i had lying around called "gcd_bench" only spends a fraction of time actually calculating gdc and much, much, much, much more time doing a loop over two ranges X'd 17:01
timotimo [Coke]: well, yeah, you could if you wanted to ... which is probably what most people outside the echo chamber do want 17:02
japhb The commands to use are `bench stress` and `bench diagnose` after you have built your compilers. rakudo-moar and rakudo-jvm should be testable.
timotimo thank you, japhb, i'll try it out :)
japhb html_plot output doesn't work yet, but all the other outputs should work.
timotimo japhb: suggest how we should handle moar-jit? 17:03
because moar is now jit by default :\
i used to have a rakudo-moar-jit, now it's gone from the compilers.json because it was just a local patch i had
japhb Erm, hmmm. This comes back to your age old request for custom build options.
*sigh*
timotimo kind of :\ 17:04
liztormato [Coke]: I remember working with Perl 5.0. It was not without bugs. Some people I know are saying we should ship perl 6.0 as it is now.
timotimo well, it's even more complicated now, because the flag "--enable-jit" is deprecated now (softly) and --disable-jit would upset a configure.pl from way back when
japhb OK, I recognize that as a key next feature. However, I want to get the stress branch merged before I deal with that feature (in a new branch, most likely, unless I can find a not-invasive solution I'm happy with.)
japhb Frankly I want stress merged as a "Welcome Home!" to jnthn. ;-) 17:05
timotimo hm. don't prioritise that too high 17:06
there's probably more interesting stuff
japhb Note that the html output has very simple rollover text now, I'd like to extend that. And --verbose on output matters; it determines whether all 10 runs at a given scale are summarized in one icon, or all shown separately. Personally, I prefer the latter right now. 17:07
timotimo OK
japhb Bus stop & 17:08
Tekk_ why does nobody here know how shells work? bus stop & will most certainly not get it to wait for you :) 17:09
peteretep What does the shell say? 17:09
Tekk_ [1] 2353
in my case :P
timotimo oh wow
perl6-bench is *very* unhappy 17:10
Tekk_ and then "command not found: Bus"
timotimo pages upon pages upon pages upon pages of use of uninitialized value @parts of type Any in string context in sub MAKE-CLEAN-PARTS at /home/timo/perl6/rakudo/../install/languages/perl6/runtime/CORE.setting.moarvm:1
Tekk_ that's not good
liztormato Yuck
And probably my fault
timotimo as well as use of uninitialized value @parts of type Any in string context in block at ./bench:256 17:11
that's in "EXTRACT CHECKOUTS" after "nom already cloned"
colomon timotimo: okay, commit d2e301e24ad7ed51484a16ed81db81df9f724e66 is the one that breaks ABC
timotimo ah, bumps the nqp/moar revision 17:12
liztormato I can't do anything about it until much later tody
timotimo liztormato: s'okay
timotimo AFK 17:13
liztormato Aflk for LPW pre-event social 17:16
japhb Tekk_: You would prefer '& # Bus stop' instead? :-) 17:23
timotimo: OK, I've been running benchmarks with a build from a few days ago. Try running with 2014.10? 17:24
ugator I have installation problems for rakudo under ubuntu (14.04) when I'm not in the rakudo dir, only perl6-p works, perl6-m(==perl6) and perl6-j only work in the dir where they're installed (thx for any help ) I did everything according to the manual (at least I triedt wice^^) 17:25
japhb ugator: Which Rakudo release? 17:26
ugator I downloaded it Nov 6th 17:27
japhb ugator: But are you using Rakudo Star, or a tarball of Rakudo itself, or building using rakudobrew, or building from a raw git clone ...? 17:28
masak howdy, #perl6
japhb o/ masak
ugator :D not sure^^, I did: git clone git://github.com/rakudo/rakudo.git 17:29
TimToady masak: \o and do you know offhand which rt is about the loop vs $_ problem?
japhb OK, then you should be building straight from head AKA 'nom', ugator
TimToady didn't fudge for.t because he didn't know that...
japhb What configure/make commands did you use, ugator? 17:30
ugator perl Configure.pl --gen-parrot --gen-moar --gen-nqp --backends=moar,parrot,jvm
make; make install;
make test suceeds 17:31
masak TimToady: hold on -- looking. 17:31
japhb OK, and when you ran them from another directory, did you use a relative or absolute path?
TimToady it's also possible it's a different bug 17:32
ugator for the $PATH-variable I used absolute paths
japhb Oh, you ran them as just 'perl6-j' for instance, and trusted $PATH 17:33
ugator it does find the commands 17:34
I just tested to directly use abs/rel paths, doesnt work either
japhb ugator: error message? 17:35
masak TimToady: not finding anything concrete. which loop vs $_ problem are we talking about?
ugator for perl6-m: unhandled exception: failed to load library 'dynext/libperl6_ops_moar.so' 17:36
TimToady last test in S04-statement-modifiers/for.t
but someone was mentioning a problem earlier that seemed like it was related, so I thought maybe it had an RT
but maybe not
ugator for perl6-j: Fehler: Hauptklasse perl6 konnte nicht gefunden oder geladen werden, translated: Mainclass perl6 couldn't be found 17:37
japhb Well that's odd
ugator so I just stick to perl6-p for now^^
masak looks
masak TimToady: think it's this one: rt.perl.org/Ticket/Display.html?id=100746 17:38
japhb ugator: what happens if you start from a fresh git clone and *only* build perl6-m? 17:39
ugator ok, I will try that 17:40
japhb has separate clones for each backend as a side effect of his build scripts, so wouldn't notice if we broke multi-backend builds.
ugator no it's not about the mult-backend, same error 17:46
alexghacker if I were to write a Bencode module for perl6, should I just name it Bencode or should I put into a namespace? e.g. Encoding::Bencode 17:52
BenGoldberg Besides being something with a wondeful name, what's Bencode? 17:52
BenGoldberg grins. 17:53
alexghacker it's a binary data interchange format first used by BitTorrent
BenGoldberg Putting it in a namespace is a good idea. Whether Encoding:: or Data:: or something else is best, though, I'm not entirely sure. 17:54
alexghacker I'm looking into it because it's one of the few data interchange formats I've seen where there's only one valid encoding for any given value (which can be arbitrarily complex)
dalek ast: eee416a | TimToady++ | S04-statement-modifiers/for.t:
fudge "{$_}" for ...
alexghacker that's a useful property for computing cryptographic hashes of content
BenGoldberg Unless the main purpose of your module is reading from or writing to streams, Encoding *probably* isn't the right place. 17:57
azawawi pbrd.co/10F1nbM # Perl 6 debugging + Error checking in farabi6 :) 17:58
alexghacker I'm probably going to be exchanging JSON5 messages, then bencoding them only for (the optional) validation of the signatures. 17:59
BenGoldberg I would guess Data::Bencoding is likely the best module name. 18:01
alexghacker BenGoldberg++ thanks. I'll get started using that.
There should be plenty of time before I've finished for others to chime in. :) 18:02
BenGoldberg Err, Data::Bencode :) 18:02
alexghacker hmm... I could have sworn that was what you said in the first place... 18:03
azawawi .tell tadzik it would be nice to have bin/script => bin/script.bat in win32 environment. This is one of the things that prevents the inclusion of farabi6 in Task::Star 18:04
yoleaux azawawi: I'll pass your message to tadzik.
regreg what is faster moarvm or parrot? 18:05
dalek ast: bdf4af4 | TimToady++ | S04-statement-modifiers/given.t:
test scoping of "{$_}" given ...
 18:06
BenGoldberg It depends.
azawawi BenGoldberg: on? 18:07
BenGoldberg Hmm, actually, iirc, moarvm is faster, but parrot is more feature-complete. But I could be wrong.
azawawi i think the real question is moarvm vs jvm backend
jvm startup is slower but on the long run jvm is gonna be faster for server apps 18:08
regreg i see..
is the jvm support mature enough to use in a small commercial project?
BenGoldberg Plus, you could, theoretically, run perl6 in a user's web browser, using jvm.
Tekk_ BenGoldberg: not so much practically since oracle has disabled jvm plugins by default now 18:09
regreg right
Tekk_ for the security problems
azawawi you mean java applets 18:09
Tekk_ mhm
BenGoldberg azawawi, yes.
TimToady moar has fewer fudges than parrot; jvm has the most
Tekk_ not surprising 18:10
I'm honestly stunned that a perl can run on jvm in the first place
BenGoldberg . o O (Chocolate fudges?)
Tekk_ BenGoldberg: it's an english expression, not sure where it comes from
arnsholt regreg: Potentially viable. Probably depends a lot on what exactly you want to do, though
Tekk_ little coverups
like you get a 78 on a test and say you got an 80
regreg arnsholt a 2d game
Tekk_ you're fudging your grade
an american expression in english rather :) 18:11
azawawi benchmarks farabi6 on jvm... rebuilding rakudo
FROGGS o/ 18:12
timotimo: yes, rerunning failed test verbosely would be cool :o) 18:14
colomon +1
FROGGS is happy 18:18
arnsholt regreg: Hmm. Maybe. You'll probably have to hack on the JVM interop libraries to get full interaction with Java libraries, but it's not a priori impossible I think 18:19
[Coke] anyway, TimToady++ for lighting a fire under our collective butts. 18:23
azawawi FROGGS: Perl 6 debugging now works in farabi6 along with syntax checker gutter hints pbrd.co/10F1nbM :)
FROGGS ohh wow!
azawawi++
arnsholt Do we have a tool that takes a .moarvm and dumps the bytecode to something human-readable? 18:24
FROGGS arnsholt: moar
moar --dump foo.moarvm
arnsholt Yaay!
Cheers!
FROGGS :D
azawawi FROGGS: thanks 18:25
FROGGS azawawi: thank you :D
azawawi feather.perl6.nl:3030/ # alt-o and open debug-01.pl or debug-02.pl
it is still buggy... 18:26
but i noticed a strange thing today...
step outside a routine seems not to work in perl6-debug-m
the "so" command that is 18:27
FROGGS what does it do? nothing?
azawawi says it is not in routine lol when in a sub... 18:28
azawawi arnsholt: do you have some documentation for moarvm file format? 18:29
arnsholt I know next to nothing about MoarVM internals, actually 18:30
But I'm the dump looks mostly like NQP op names (which makes sense, since the Moar ops are mostly the same as the NQP ops)
s/I'm//
[Coke] see also #moarvm 18:31
dalek osystem: 9c3baf7 | tony-o++ | META.list:
Bench[mark] module

More similar to P5's Benchmark, includes cmpthese, timethese, timethis		 18:36
pmurias curtispoe: re the production date, a lot of the confusion is a troll misreading the talk summary and posting a mislabeled link 18:40
masak public notice: if you announce a date when Perl 6 will be ready, that attracts trolls. thank you. 18:44
tony-o so, February? That's what TimToady was personally guaranteeing me yesterday ;) 18:45
TimToady silly dilly, Christmas is not in February
tony-o haha
pmurias masak: the guy who posted in on reddit seems to be a dedicated Perl 6 hater
nine Someone at the LPW tomorrow? 18:46
tony-o trying to figure out how i could take a crack at Perl6::Tidy
if i should use the grammar or .. 18:47
pmurias Perl6::Tidy would be awesome
tony-o: is there a alternative that isn't batshit crazy?
tony-o i'm thinking no but i'm not very clear thinking today 18:48
raydiak imo Christmas is whenever Perl 6 is ready regardless of what the calendar says :) 18:49
pmurias using the grammar is the way I thought about doing it 18:50
tony-o i think it would be a waste to do it another way 18:51
TimToady some great quotes about PR on: www.contentfac.com/damn-i-wish-id-s...pr-quotes/ 18:52
TimToady “There is only one thing in the world worse than being talked about, and that is not being talked about." – Oscar Wilde 18:53
tony-o i like the S H Simmons quote 18:54
masak pmurias: oh. reddit. then disregard what I said. 18:57
curtispoe pmurias: that’s some of the confusion, but there are other discussions which have seen beyond the mislabeliing. 18:58
For example: news.ycombinator.com/item?id=8571837
timotimo i wonder what dreamweapon on HN refers to when saying that perl 6 adds "a lot of visual clutter" 19:11
TimToady probably doesn't like hypers 19:12
TimToady twigils are also arguably additional clutter 19:13
timotimo and this kamaal person goes a long way to say "cool libraries and framework ideas from perl 5 are going to seep into perl 6" and completely fails to mention that perl 6 has inspired a bunch of changes to perl 5, too
someone else says "Perl 6 is a modest advance not a revolution", which ... well ... what do i know. 19:14
timotimo d'oh, someone said "One of the strongest parts of perl, the regular expressions, have received further improvements", someone else retorts with "the regular expressions [...] have been copied to Nimrod (now called "Nim") -- even with native C performance and native Perl syntax" 19:19
that is missing the point by so many kilometers it's not even funny
TimToady well, not everyone appreciates Martha Stewart either, "the woman who does everything better than you". :) 19:22
vanila hi guys 19:23
perl6 this christmas?
curtispoe Well, I was specifically referring to the title not being a misunderstanding. There’s still going to be lots and lots of confusion about the language. 19:24
And that
And that “modest advance” comment through me off guard too.
s/through/threw/ (and I don’t know where my enter key is)
TimToady vanila: no, we're still just getting ready to get ready to party this christmas 19:25
timotimo that still sounds like "party this christmas" :P
TimToady vanila: no, this christmas we're still just getting ready to get ready to party
next year we're getting ready to party :) 19:26
bartolin Hi, #perl6 19:27
TimToady o/
[Coke] hio 19:28
curtispoe Is there any discussion of work on DBI for Perl 6? 19:30
raydiak I think irrational attitudes are a normal reaction to p6 being ahead of its time, or at least very powerful...makes huge waves relative to the size of its user base...therefore, many more people talking about it than actually experiencing it 19:32
and from the perspective of knowing about it without having the experience of having used it, parts of it scared the heck out of me too tbh :)
timotimo raydiak: you're laying it on a bit thick there ;) 19:33
computers have been able to do what perl6 does for a very long time. it's just you had to use assembler or lisp :P 19:34
raydiak heh
hardly equivalent from the perspective of a human being (like programmers, for instance)
ajr_ Many of the discussions on the Internet appear to have more to do with the participants' internal states than anything objective about the topic of the discussion. 19:36
raydiak I certainly don't mean it'll make your computer magic and make breakfast for you, it'll still only do what they do...but how hard is it to accomplish a given task, for a given individual? that's one of the things I mean when I say "power" wrt a tool 19:37
ajr_ People who are severely wound up internally project the feelings onto some external cause, whether it be political, religious, or technical religion, (computer language, &c.)
arnsholt curtispoe: There's a DBIish 19:38
raydiak ajr_: agree 100%, that was kinda my initial point...people are saying things that stretch reason abotu p6 b/c of their feelings about it 19:38
curtispoe arnsholt: I’ve seen it, but real database support is needed.
arnsholt The DBI namespace is currently unoccipied, as I believe Tim Bunce has been working on an API modelled on JDBC
ajr_ raydiak - your definition of "power" is very good. 19:39
curtispoe Night all! 19:39
raydiak thanks ajr_ :) I like to play armchair philosopher once in a while 19:40
timotimo raydiak: can i interest you in improving the html app our profiler emits? 19:41
raydiak timotimo: sure, I'd be happy to lend a hand 19:43
timotimo cool
raydiak what does it need? what is this thing I'm even talking about?
timotimo have you seen the data format that moarvm spits out?
raydiak don't think so
timotimo t.h8.lv/profile-1412458985.39989.html - here's an old-ish one
m: say DateTime.new(1412458985.39989)
camelia rakudo-moar 315ec6: OUTPUT«Default constructor for 'DateTime' only takes named arguments␤ in method new at src/gen/m-CORE.setting:926␤ in block <unit> at /tmp/11nQLwMqSU:1␤␤»
timotimo hum. 19:44
timotimo that may be a bad example, it seems to be pretty big 19:44
just run anything with perl6 --profile; should best be quite quick
raydiak oh wait, yes...I tried this thing on pray
timotimo good
one of the thing i find most sorely missing is getting a list of callees "globally" for a given routine in the routines tab 19:45
i don't know if the data format we emit can easily offer that, though
and if you could figure something out that'd make the app more responsive even if the amount of data is overwhelming, that would be crazy cool
raydiak no kidding, glad I'm not running my laptop on battery :P 19:46
timotimo also, the exclusive times of routines quite often has a very long tail; it'd be cool if the routines list had some flags at the side saying things like "from here, the rest of exclusive times add up to 30%"/"20%"/"10%" OSLT
like, 1/4 of the way down the percentage numbers reach 0% 19:47
timotimo oh and something purely cosmetic: when you view the details of any given allocation thingie, the progress bars quite often extend outside the "pop-up" that comes up 19:48
i'll be AFK for a bit again now :)
ooh
but before i go, i shall document something else:
timotimo IMO if there's a single call to a routine or something, the red bar basically says "oh no! look at me! all of my executions were not spesh'd and not jitted!", which is kind of uninteresting since it's only called once 19:49
timotimo so i'd love for the red bar to be somewhat desaturated or in a less alarming color in that case 19:49
TimToady maybe you should make it pink 19:50
masak encouraging comment: news.ycombinator.com/item?id=8573337 19:50
[Coke] tim bunce mentioned something years ago; I'm not sure any work on DBI is currently being done. 19:51
timotimo he talked to us about that idea during APW, too
o/
raydiak timotimo: cool, I've made a list from your suggestions, will dig in some time in the next several days \o 19:53
gtodd timotimo: the colours could come from perl6.org website ... there's a chance for the profiler's HTML to become trend setting ... like Bootstrap :-)
except !Bootstrap ;-)
[Coke] timotimo: still, haven't seen any code. 19:54
masak I can't believe how much constructive discussion is going on in that HN thread. that's very encouraging. 19:56
TimToady most people want to believe 19:57
vendethiel o/ again 19:58
masak ok, then let's please make this one count. 20:01
FROGGS hi vendethiel
vendethiel masak: I can guess which one... But which one?
masak vendethiel: 2015 as the Year of Perl 6.
vendethiel guessed :) 20:01
rjbs Where's the HN thread? And, more importantly and first, do I want to read it? :) 20:02
Oh, found link.
masak maybe we need to think a little bit more like a business, for once. a business with paying customers.
even though we're not.
vendethiel
.oO( what does carlosdelrey wants us to think )
hahahaha 20:03
masak life is to short to spend on carlosdelrey.
too*
vendethiel
.oO( Mu is too short )
 20:04
masak m: subset Too::Short of Mu; say Mu ~~ Too::Short 20:05
camelia rakudo-moar 315ec6: OUTPUT«True␤»
vendethiel masak: I've been considering perl6 production ready for a few months :P
masak vendethiel: I've been using it in production for years. get off my lawn. :P 20:06
vanila I wanna part y w ith you all
vendethiel masak: yera hipster, harry :P
masak vanila: \o/
vendethiel vanila: for me, the party's already over, but I'm fine with opening antoher bottle :)
vanila :D
masak vanila: you're just in time! apparently, we are just about to release Perl 6! \o/
vendethiel good thing I was told to keep quiet about this announcement :P 20:07
tony-o im already migrating off of perl6 for production to perl7
vendethiel oh wow, that looks cool tony-o. keep us in!
FROGGS does somebody use NativeCall on JVM?
vendethiel FROGGS: I've used the java-binding feature, but that's not nativecall :P 20:08
masak tony-o: I know for a fact that Perl 7 does not exist yet, because I still haven't designed it together with sorear in our generation ark.
FROGGS vendethiel: that does not help here :o)
vendethiel knows
tony-o tony-o: production dreams :-)
oops
masak: ^^
FROGGS vendethiel: I want to know if my java installation is to blame or if something recentish in rakudo/nqp breaks it 20:09
vendethiel masak: so, what would you think about having a MATCH in the language :P
Story is -- We have a great match feature, called MMD. But, why don't we have it available *outside* of subs?
masak vendethiel: I like where that thought is going. but would need more detauls. 20:10
details*
vendethiel 1) variable binding 2) variable comparing (mmd doesn't have that) 3) multiple combinations
masak: natefaubion's sparkler (sweet.js macros for pattern matching) show a good idea of how this looks in practice
(that is, having both MMD and a "match" keyword working the same underneath)
masak ok. 20:12
sounds quite interesting.
and occasionally useful.
vendethiel thinks sweet.js macros can be very good examples of what good can be done with macros "like that" :-)
masak if you're also telling me that someone could use this to implement Prolog-like unification in Perl 6, then I'm sold :)
vendethiel well, most definitely, yes 20:13
but you need backtracking.
masak for unification?
vendethiel ... so you need continuations :P
for prolog-like unification at least, yes, you need backtracking
TimToady or lazy lists of closures
TimToady (that's how STD does it internally) 20:14
vendethiel well, you can always find "another way", that's for sure. I'd say it "doesn't look as good" 20:15
(moslty because it's more transparent)
TimToady transparency cuts both ways :)
TimToady but yes, it fits more with the FP idea of sweeping all the state off into the part of the implementation you can't easily see 20:16
TimToady Abstraction is a good thing, except when it isn't. 20:18
vendethiel But it's not really an "abstraction"
cosimo_ hn thread is definitely interesting 20:20
:)
hello, all :)
vendethiel helo!
masak m: print "$_o " for <hell cosim>; say "!" 20:21
camelia rakudo-moar 315ec6: OUTPUT«===SORRY!=== Error while compiling /tmp/DLrObmZRIF␤Variable '$_o' is not declared␤at /tmp/DLrObmZRIF:1␤------> print "$_o⏏ " for <hell cosim>; say "!"␤ expecting any of:␤ postfix␤»
masak m: print "{$_}o " for <hell cosim>; say "!" 20:22
cosimo_ i missed all the fun, what happened that triggered this great announcement? :-)
camelia rakudo-moar 315ec6: OUTPUT«o o !␤»
vendethiel tss, tss :)
masak m: print "$($_)o " for <hell cosim>; say "!"
camelia rakudo-moar 315ec6: OUTPUT«hello cosimo !␤»
masak grumbles
vendethiel never knows when to use $() vs {}
cosimo_ masak: hi Carl 20:23
TimToady vendethiel: the first one is supposed to work, but doesn't yet
vendethiel ohhh. my bad :-)
psch hi #perl6 20:23
masak hi psch! 20:24
FROGGS hi psch
TimToady everyone's dropping in today for some reason :)
psch masak, FROGGS o/
FROGGS *g*
cosimo_ TimToady: for some definition of "some"
FROGGS TimToady++ # bringing us together
TimToady quick, barricade the doors 20:25
FROGGS |\o\
cosimo_ in my defense, I've always been here, just very quiet :)
raydiak wait you're not making us drink the kool-aid yet are you? 20:26
FROGGS cosimo_: I've got work for you :P github.com/cosimo/perl6-digest-md5/pull/7
cosimo_: hopefully it will turn green then: testers.perl6.org/dist/Digest::MD5 20:27
tony-o can i include Perl6::Grammar somehow or do I need to copy it and actually make it a module, i'm assuming that .msi and future precompiled versions will be available and that the src stuff isn't going to be available everywhere
vendethiel "one of us, one of us" or something 20:28
FROGGS tony-o: the problem with the Perl6::Grammar in rakudo is that it is an nqp module
psch $~MAIN should be Perl6::Grammar i think? i.e. getting to Perl6::Grammar is still NYI 20:29
tony-o: ^
FROGGS correct, we have a language barrier here
tony-o i see, it will eventually be there or should i just port the grammar? 20:30
psch tony-o: well, if youre porting it you might as well stuff it into $~MAIN... :)
although that might be wonky wrt keeping them synced 20:31
FROGGS it will be available at some point, but that won't happen in the next 6+ months I guess
TimToady for now I'd just port it; we originally were hoping to become fully bootstrapped, but it's easier to port nqp than full Perl 6, so it's a useful breakage to have the current Perl 6 in nqp
cosimo_ FROGGS: thanks for bringing it up, and sorry for being very sloppy. Had very little time for spare time programming lately
FROGGS cosimo_: np
tony-o okay
cosimo_ there is a perl6 testers? OMG
FROGGS cosimo_: but I'd also take a commit bit in exchange :o)
tony-o psch: not sure what $~MAIN is 20:32
FROGGS cosimo_: since a few days, aye
TimToady it's short for %*LANG<MAIN> in the parser, which contains the current top-level language at any point 20:33
whenever you derive a new language during the parse, that goes into %*LANG for the remainder of the current lexical scope
psch perlcabal.org/syn/S02.html#Slangs some more context 20:35
tony-o ahh gotcha 20:37
alexghacker Does anyone know if specifying the normal form in Str.encode not yet fully implemented?
I'm getting the same buffer contents no matter what values I send in for nf 20:38
FROGGS alexghacker: :nf was specced this year IIRC and is not yet implemented, aye 20:39
alexghacker FROGGS++ thanks
PerlJam alexghacker: also, github.com/rakudo/rakudo/blob/nom/...r.pm#L1035 20:40
S02:Slangs # test
FROGGS S02/Slangs 20:41
S02#Slangs
:/
PerlJam if someone would merge my last PR to the bot, S02:Slangs would work
FROGGS I have no rights to do so :o( 20:42
cosimo_ FROGGS: got you commit bit on p6-digest-md5, and moritz too 20:43
FROGGS cosimo_: thank you! :D
hah, moritz++ almost made the same PR 20:45
alexghacker 2nd question: should I expect «$b = "abc".encode; say [~] $b.bytes, ':', $b» to consume all available memory and CPU (hint: it does for me) 20:47
FROGGS alexghacker: no, it should not 20:48
FROGGS alexghacker: though that happens right now when you concat a buf to something else 20:48
alexghacker ah, so known issue at least 20:49
FROGGS concating bufs works, you just cannot mix...
yes, had no time yet to fix or even analyze it :o(
FROGGS right now I am bisecting a problem with NativeCall on the JVM... which is also very important, because many modules depend on NativeCall 20:50
arnsholt FROGGS: What kind of problem? 20:53
tony-o FROGGS: i can test nativecall on jvm now that i'm home
FROGGS nqp::nativecallbuild explodes
arnsholt Oh. That's no good 20:54
What kind of explosion?
FROGGS because the first arg is not recognized as a NativeCallInstance and explodes when trying to unbox it
arnsholt Segfault or JVM?
alexghacker FROGGS++ don't worry too much, there are other ways I can accomplish the same thing. This is rt.perl.org/Public/Bug/Display.html?id=122600 , yes?
FROGGS no, just an exception
alexghacker: yes
arnsholt Hmm. Strange 20:54
alexghacker I'll just document my work-around as being tied to that ticket
FROGGS arnsholt: the problem was introduced between Aug 1st and Sep 5th or so 20:55
alexghacker: I bet it is not even hard to fix... it might be something silly 20:56
arnsholt Right. Potentially, something dealing with P6Opaque
alexghacker «($b.bytes, ':', $b).join('') works just as well for now
arnsholt IIRC there're some shenanigans that deal with how the NativeCallInstance is inlined into the P6Opaque 20:57
alexghacker sighs about his dangling »
FROGGS arnsholt: I still keep on bisecting... because I dunno where to start looking
arnsholt Sounds like a good plan
Incidentally, this is why I'd like to find a way to move NAtiveCall into the Rakudo NQP or repo. It's way too tied to the internals 20:58
FROGGS yeah, just takes its time
ohh yes, you get my +1 for that 20:59
PerlJam arnsholt++
(er, +1 from me too :)
arnsholt Main problem ATM, I think, is that the NativeCall test suite is a bit annoying and not very portable 21:00
vendethiel *g* 21:01
FROGGS but the idea was that it should be portable, no? :o)
arnsholt The code is (mostly) good 21:04
But compiling the C part of the tests is a bit finnicky
colomon +1 to NativeCall in core 21:08
vendethiel oh, carlos. www.reddit.com/r/programming/commen...date_says/ <3 21:15
masak carlosdelray must be Perl 6's most reluctant bug fan, ever. 21:24
big*
FROGGS I still wonder if he is/was also educated_poo 21:25
masak it's like he simply cannot stop paying attention -- but he's fighting it at every step. 21:25
frankly, I don't care what kind of poo he is.
FROGGS hehe
masak I'm just wishing him well, and thanking him for all the abject bile he spills over Perl 6, hopefully making people think a bit for themselves. 21:27
vendethiel :-) 21:29
masak and now I'm hoping we can talk about something that matters instead. 21:29
vendethiel masak: bijections are just injections that happen to be interjections as well! 21:31
sur
FROGGS masak: have you seen that btw? github.com/FROGGS/p6-Slang-Tuxic 21:33
masak: in case you want to mess with the Perl6::Grammar and inject/replace stuff
masak vendethiel: that is true, but in no way noteworthy. :) 21:38
FROGGS: oooh!
vendethiel masak: oh, I thought it could matter :P
FROGGS hmmm, chdir is broken on windows... chdir 'C:\panda' fails while chdir 'C:/panda' is working
masak FROGGS++ # for Tuxifying Perl 6 in a slang! 21:39
PerlJam vendethiel: it might make a nice schoolhouse rock type song ;)
vendethiel could it :P
masak FROGGS: any plans for making it also work for methods? 21:42
FROGGS masak: yes, will look into that tomorrow (evening) 21:45
masak ++FROGGS
FROGGS that's the bug I am facing:
C:\panda>perl6 -e "say MAKE-ABSOLUTE-PATH('C:\panda',$*CWD)"
C:\panda\panda
psch m: say MAKE-ABSOLUTE-PATH(%*ENV<HOME>, $*CWD) 21:46
camelia rakudo-moar 315ec6: OUTPUT«===SORRY!=== Error while compiling /tmp/fCqIM0ICj7␤Undeclared name:␤ MAKE-ABSOLUTE-PATH used at line 1␤␤»
FROGGS psch: camelia here is from october 22nd 21:47
psch FROGGS: yeah, i ran it locally and it looks fine on linux 21:48
FROGGS yes, because it checks for forward slash only 21:49
moritz ok, manually triggering a rakudo rebuild on camelia 21:57
FROGGS moritz++ 21:58
moritz now I wonder why the cron jobs don't work :( 21:59
alexghacker is there a way to make reduction operators hyper? e.g. to individually sum up lists inside of a list? I tried «[+] << $list-of-lists» and it's clear that the parser and I are on completely different pages. 22:02
psch m: say [[1, 2], [3, 4]]>>.map(&[+]) 22:03
camelia rakudo-moar 315ec6: OUTPUT«3 7␤»
psch alexghacker: if you know your dimension that's it
alexghacker psch++ I'll accept that 22:04
hoelzro github.com/github/linguist/issues/...t-62165770 22:09
"We are actually moving away from pygments and using TextMate/SublimeText/Atom syntax bundles instead"
='(
FROGGS :o( 22:12
[Coke] hoelzro: that sucks. :| 22:14
hoelzro yeah, I know!
oh well, I might just make a textmate "port" of my highlighter 22:15
tbf, it might do a better job than pygments did
FROGGS hoelzro: please tell me if you need any help then
hoelzro FROGGS: are you a TM/ST/Atom user?
FROGGS hoelzro: no 22:16
hoelzro oh, just pitching in then =)
thanks
I'll let you know
I built up a little pile of tuits for perl6/syntax
so I think I might be spending them tonight/tomorrow
FROGGS arnsholt: funnily, now after bisecting for hours, I remember the workaround >.<
ohh, cool :o) 22:17
FROGGS arnsholt: it is even still open in my browser gist.github.com/FROGGS/0d4b5c771c7d9eb4f5b6 22:18
ugator diff perl6 ./install/bin/perl6 in rakudo gives me different lib paths. Why is the libpath for the perl6 in the main dir just "." and not the absolute path? (like in ./install/bin/perl6) 22:19
alexghacker psch, I'm not sure that actually does the reduce properly 22:20
m: say [[1,2,3],[4,5,6],[7,8,9]]>>.map(&[+])
camelia rakudo-moar a0304c: OUTPUT«3 3 9 6 15 9␤»
alexghacker I think that should give «6 15 24»
psch alexghacker: right, apparently i generalized wrongly. &[+] is still 2-ary 22:21
m: say [[1, 2, 3], [4, 5, 6], [7, 8, 9]]>>.map([+] *) 22:22
camelia rakudo-moar a0304c: OUTPUT«Cannot call 'Numeric'; none of these signatures match:␤:(Mu:U \v: *%_)␤ in sub infix:<+> at src/gen/m-CORE.setting:4492␤ in sub at src/gen/m-CORE.setting:19556␤ in block <unit> at /tmp/Pf4LTSobKq:1␤␤»
psch m: say [[1, 2, 3], [4, 5, 6], [7, 8, 9]].map([+] *)
camelia rakudo-moar a0304c: OUTPUT«Cannot call 'Numeric'; none of these signatures match:␤:(Mu:U \v: *%_)␤ in sub infix:<+> at src/gen/m-CORE.setting:4492␤ in sub at src/gen/m-CORE.setting:19556␤ in block <unit> at /tmp/AKCCteSMW9:1␤␤»
psch m: say [[1, 2, 3], [4, 5, 6], [7, 8, 9]].map({ [+] @^a }) # this works
camelia rakudo-moar a0304c: OUTPUT«6 15 24␤»
psch oooh 22:23
m: say [[1, 2, 3], [4, 5, 6], [7, 8, 9]].map([+] ** ) # this too?
camelia rakudo-moar a0304c: OUTPUT«Cannot call 'Numeric'; none of these signatures match:␤:(Mu:U \v: *%_)␤ in sub infix:<+> at src/gen/m-CORE.setting:4492␤ in sub at src/gen/m-CORE.setting:19556␤ in block <unit> at /tmp/szkzcsbLdI:1␤␤»
psch aw shucks
Mouq++ didn't finish HyperWhatever :)
masak heh, `[+] *` is one attempted use of WhateverCode I would not have come up with myself.
psch masak: with ** it should work i think 22:24
FROGGS ugator: the perl6 script in your build directory is just something is meant to work exactly in that dir, so . is enough
psch but maybe i have a wrong model of HyperWhatever
ugator so that one is not to be accessed from outside? ok,that wasnt clear^^
FROGGS ugator: exactly 22:25
ugator: only the one in install/bin should be used from elsewhere
psch hm, actually i'm most likely wrong. HyperWhatever stills builds a WhateverCode but does something flatten-y to its arguments..? 22:28
m: say (** - 1)(1, 2, 3) 22:30
camelia rakudo-moar a0304c: OUTPUT«0 1 2␤»
psch m: say (** + *)(1, 2, 3) 22:31
camelia rakudo-moar a0304c: OUTPUT«===SORRY!=== Error while compiling /tmp/UmMNA5h3hJ␤Multiple HyperWhatevers and Whatevers may not be used together␤at /tmp/UmMNA5h3hJ:1␤------> say (** + *⏏)(1, 2, 3)␤ expecting any of:␤ postfix␤»
dalek kudo/nom: a3d18ca | (Tobias Leich)++ | tools/build/Makefile- (3 files):
allow fudging of src/Perl6/Actions.nqp
 22:36
psch maybe i'm not wrong about HyperWhatever... 22:39
m: say (**.[0])([1, 2], [3, 4])
camelia rakudo-moar a0304c: OUTPUT«1 3␤»
masak ...fudging of Actions.nqp...?
psch it's too late to think about this, actually o/ 22:40
psch goes to sleep
FROGGS masak: for a single line workaround for the jvm
masak FROGGS: one that we expect to go away as soon as possible? 22:42
dalek kudo/nom: 393acf0 | (Tobias Leich)++ | src/Perl6/Actions.nqp:
unbreak NativeCall for the JVM

The call to nqp::buildnativecall fails because it does not recognize the first arg as an NativeCallInstance. Doing the clone always like before e167669a498b91543144c7657e5cdb8d0cb40e65 makes it work, see also: gist.github.com/FROGGS/0d4b5c771c7d9eb4f5b6
FROGGS masak: as soon as jnthn or something like that 22:43
I don't like it either but it is now broken for about two months 22:44
masak has a new least favorite SQLite error code: SQLITE_ERROR 23:08
Diederich masak: thanks for the notice, Re: news.ycombinator.com/item?id=8573337 23:11
masak Diederich: well, thank you for your patience. 23:12
Diederich hah!
Diederich what makes you think I've been patient :) 23:12
masak :P 23:13
Diederich mostly I feel bad that I haven't been able to take time to contribute to P6
I've always been so blessed with the ability to do very creative and innovative work at work,
masak no need to feel bad -- even writing stuff on HN is a contribution.
Diederich that I have little 'creative energy' left after :) 23:14
masak sometimes it's nice for some of us to see that there are people out there who care.
Diederich I have been creating more P5 CPAN modules at home, and then using them extensively at work :)
masak that's excellent.
Diederich when I taught Perl,
I actually had my students read and do a brief writeup of one of Larry's talks 23:15
specifically, Perl, the first postmodern computer language...really changed my personal outlook
masak :) 23:15
Diederich so Perl has resonated with me more and more deeply over the years.
masak I guess Perl 6 is post-post-modern... 23:16
Diederich as I was getting ready to leave my last job, I learned Ruby and Python (again...for each) and ported some of my CPAN modules tothem
assuming that there was No Way I was going to get another Perl job in 'The Valley'
but here I am, once again, in a very heavy Perl shop. go figure
and we have a real need to run a 'scripting language' inside of our big Java apps, so in 2015, I plan on pushing Perl 6 into that role here 23:17
masak woot
all I can say is, please stay in touch
Diederich while I have your attention, I was wondering if SystemV IPC was implemented/available in P6?
masak we're happy to ease that journey in any way we can. 23:17
I know nothing about SystemV IPC, unfortunately. 23:18
Diederich that's always been the basis for the systems I develop
uh
message queues
masak I haven't *heard* anyone working on it, but that doesn't mean no-one has.
Diederich well,
it's C functions
so as I understand it,
one can call native C functions from Rakudo JVM n ow?
masak yes. 23:19
Diederich cool
and Moar?
I need to look into making those bindings then
vendethiel you can :)
Diederich specifically, metacpan.org/source/IPC::Msg 23:20
more specifically, for example, metacpan.org/source/IPC::Msg#L97
P5 has those 'bindings' natively
linux.die.net/man/2/msgrcv 23:21
timotimo o/ 23:21
wow, so much backlog tonight
Diederich unrelated, I'm sure, to The Big Posting :) 23:22
vendethiel *g*
or the thought-to-be one
TimToady www.bmsi.com/java/posix/ might also be of some help
Diederich indeed
TimToady++ indeed, thank you 23:23
jdv79 is that fosdem talk summary misleading or is xmas '15 really it? 23:26
leont Diederich: you're the first person I've ever heard use SysV message queues in production! :-o 23:35
leont POSIX message queues really are a nicer interface though 23:35
Diederich leont: I have kept my eye on POSIX queues since 'the beginning'. The big delta is thread safety, but I never ever write threaded code :) 23:37
thank you 23:38
Tekk_ Diederich: imo it's better to use processes and ipc unless you're on windows 23:41
so the parallelism in most languages works fine for me :)
leont The difference is bigger for semaphores and shared memory IMO, because those are kind of crazy in SysV. 23:42
timotimo it's always super frustrating to read backlog and know the answer to some people's questions and they leave in the mean time >_<
Diederich leont: definitely. 23:43
timotimo m: say [[1, 2, 3], [4, 5, 6], [7, 8, 9]].map(&[[+]])
camelia rakudo-moar a0304c: OUTPUT«===SORRY!=== Error while compiling /tmp/lw5aG29PvT␤Variable '&infix:<[+]>' is not declared␤at /tmp/lw5aG29PvT:1␤------> 2, 3], [4, 5, 6], [7, 8, 9]].map(&[[+]]⏏)␤ expecting any of:␤ postfix␤»
timotimo m: say [[+]] [[1, 2, 3], [4, 5, 6], [7, 8, 9]]
camelia rakudo-moar a0304c: OUTPUT«3␤»
timotimo oops :3
m: say [+]«[[1, 2, 3], [4, 5, 6], [7, 8, 9]] 23:44
camelia rakudo-moar a0304c: OUTPUT«===SORRY!=== Error while compiling /tmp/TfhvWt3MQ6␤Unable to parse quote-words subscript; couldn't find right double-angle quote␤at /tmp/TfhvWt3MQ6:1␤------> ay [+]«[[1, 2, 3], [4, 5, 6], [7, 8, 9]]⏏<EOL>␤ …»
timotimo m: say Z[+] [[1, 2, 3], [4, 5, 6], [7, 8, 9]]
camelia rakudo-moar a0304c: OUTPUT«===SORRY!=== Error while compiling /tmp/lwtIXp5dnQ␤Preceding context expects a term, but found infix + instead␤at /tmp/lwtIXp5dnQ:1␤------> say Z[+⏏] [[1, 2, 3], [4, 5, 6], [7, 8, 9]]␤»
timotimo mhm
Diederich leont: 'back in the day', I initially developed what would become IPC::Transit using shared memory and semaphores, but they had all kinds of platform specific problems. message queues worked everywhere (in 'UNIX' land)
timotimo alexghacker: the concatenation of bufs with strings is broken because Buf does the Stringy role *and* coercing it to Stringy just calls .Stringy on self 23:45
so the Stringy method on a buf will call Stringy on self, which will call Stringy on self which will call Stringy on self et cetera ad finitum
Diederich ouch
timotimo i once tried to find a good fix to that. nobody tuned in to the discussion :( 23:46
timotimo FROGGS: can we get tooltip texts for the bars on perl6testers that tell not only which backend that is, but also how many passes/fails it has? 23:48
leont Shared memory and semaphores can be really powerful (and fast), but it tricky to get right :-/ 23:50
Diederich leont: yup. but I was dealing with some serious bugs in the way semaphores worked on HPUX 9.x at the time :( sank the ship 23:51
timotimo Diederich: with NativeCall you can access Sys5 IPC 23:54
Diederich timotimo: that's actually what I was thinking, but I haven't worked on it yet. 23:54
timotimo: are there any good docs/examples with NativeCall available now? 23:55
dalek ast: 1dae23b | TimToady++ | S02-lexical-conventions/comments.t:
comments do NOT nest, but brackets do
timotimo should be interesting :)
we have a few examples
why don't we have something like "search for things that depend on X"?
github.com/timo/cairo-p6/ - a not too bad binding to cairo with NativeCall
timotimo github.com/perl6/gtk-simple/ - a mostly acceptable binding to GTK3 with NativeCall 23:56
Diederich thanks so much 23:57
timotimo you're so welcome :) 23:58
Diederich and thank you all, once again!
timotimo ( i'm just blatantly promoting my own work :P ) 23:58
masak timotimo++ 23:59

Powered by Cro and the Raku® Programming Language.

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

Thank you!