»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.
lizmat .tell jnthn apperently your commit 442a465a947fdad4d4d has borked rakudo build on at least OS X 00:00
yoleaux lizmat: I'll pass your message to jnthn.
lizmat Stage start : 0.000
Stage parse : moar(46847,0x7fff7c1a8310) malloc: *** error for object 0x7f94858bfda0: pointer being freed was not allocated
*** set a breakpoint in malloc_error_break to debug
make: *** [CORE.setting.moarvm] Abort trap: 6
lizmat catches some sleep & 00:01
[Coke] I have a build directory for rakudo that I've used to install nqp, moar, perl6-m 00:29
How can I force a rebuild of nqp if the installed nqp is recent enough. 00:30
BenGoldberg j: say 'alive'; # what's wrong with jakudo? 00:44
camelia rakudo-jvm 442a46: OUTPUT«Can't call method "syswrite" on an undefined value at /home/p6eval/jvm-rakudo/eval-client.pl line 32.␤»
tony-o [Coke]: i believe moving the directory nqp builds in would let you rebuild nqp 04:54
rakudo/nqp iirc 04:55
FROGGS_ TimToady: in German I'd say: "Der Vergleich hinkt." :o) 05:10
FROGGS[mobile] BenGoldberg: something about the JVM or environment camelia is running on 06:00
leont My grammar works in the debugger, but not outside of it􏿽x85 06:02
Debugging this is difficult :-s
nwc10 leont: are you in the right place to catch your ride? 06:03
leont I'm on the train :-) 06:04
nwc10 excellent
FROGGS[mobile] leont: and it just does not match outside of the debugger? 06:05
lizmat leont++ :-)
nwc10 should set off
FROGGS[mobile] do I also get karma for sitting in a train? :o)
leont Almost halfway to lizmat and woolfy
nwc10 & 06:06
leont FROGGS[mobile]: nope. See github.com/Leont/toml6 06:08
FROGGS[mobile] also, I've got three accu packs and should have a power socket in the next train... hell am I prepared *g*
leont: and subparse also does not match? 06:09
leont Doesn't make a difference 06:11
FROGGS[mobile] leont: I can look at it in about 30min, but I'd need a test file
leont Oh, right. I forgot I hadn't pushed that! :-o 06:12
leont (train internet)-- 06:23
A test is pushed now
It seems the datetime rule is the issue, probably because it prefix-overlaps with number 06:24
leont Getting rid of the datetime in the example makes it pass again. 06:24
FROGGS[mobile] ohh, that's a start 06:27
I guess you want some look-ahead then 06:28
leont I have some lookahead, but I'm not sure it's enough 06:30
lizmat .tell jnthn your last patch seems without problems on jvm and parrot 06:31
yoleaux lizmat: I'll pass your message to jnthn.
[Tux] FYI: Metadata Vocabulary for Tabular Data: www.w3.org/TR/tabular-metadata/ 06:36
FROGGS[mobile] leont: you can additionally do: <!datetime> as the first thing in value:sym<number> 06:37
lizmat commute commences& 06:38
leont This gets stranger 06:51
Adding an empty code block fixes it too
Feels like some kind of optimization bug to me
moritz en empty code block terminates LTM 06:52
leont LTM should pick datetime over number, so I don't understand 06:54
FROGGS leont: are you sure? 06:58
in the number token you have literals at the beginning, in datetime you only have character classes 06:59
FROGGS LTM prefers literals over cclasses 06:59
leont & # train switchover 07:00
moritz if the length is the same, yes 07:06
FROGGS in this case here the '-'? probably matters most I think 07:08
jnthn is at CPH, sipping coffee and waiting for his flight 07:12
yoleaux 00:00Z <lizmat> jnthn: apperently your commit 442a465a947fdad4d4d has borked rakudo build on at least OS X
06:31Z <lizmat> jnthn: your last patch seems without problems on jvm and parrot
FROGGS jnthn: I guess you wanna grab lizmat laptop this evening? :o) 07:13
jnthn .tell lizmat well, good timing...I've not had thse happen on Windows at all, and so haven't had chance to hunt the thing down yet...provided there's decent debugging stuff on your laptop I might finally stand a chance :) 07:14
yoleaux jnthn: I'll pass your message to lizmat.
jnthn Linux doesn't seem to be so sensitive to whatever it is either... 07:15
FROGGS true
jnthn should get some OSX box some day...
FROGGS jnthn: but other ppl reported it too
jnthn FROGGS: on OSX? Yes.
FROGGS yeah, colomon for example
jnthn FROGGS: But lizmat will be in the same place as me AND have an OSX machine that exhibits the issue :)
FROGGS jnthn: I hope that I get also these results on my PowerPC hardware, provided I can get my hand on an OS X 10.4.11 07:16
all true :o)
jnthn steadily wakes up 07:23
By the time I'm awake enough to do anything it'll be boarding time...
timotimo o/ 07:33
jnthn o/ timotimo 07:36
timotimo in my sleepy head while waking up i thought about the interaction of shaped arrays and hyper operators 07:37
jnthn I just thought "oh yeah, I should go to the airport like now" :) 07:38
jnthn should probably go find his gate... :)
&
woolfy On our way to Salzburg in 10... 9... 8... (minutes)
jnthn: good flight! 07:39
exit
timotimo aye, good flight!
timotimo and good drive, woolfy and lizmat 07:39
(i know for sure part of it will be good, because i'll be on board! :P )
azawawi hi everyone 07:40
moritz hi all
FROGGS hi
timotimo ohai 07:42
azawawi moritz: is there a way to get a documentation index of all keywords that p6doc knows about? 07:43
moritz: like p6doc --list-keywords or something
moritz: im integrating p6doc with farabi6
moritz azawawi: not that I'm aware of
TimToady m: my $a = 1; my $b = 2; say $a + $b\i
camelia rakudo-moar 442a46: OUTPUT«1+2i␤»
TimToady timotimo: looking for that ^^ maybe? 07:44
timotimo oh, cute
TimToady and to the first approximation, hypers on shaped arrays can ignore the shape, except to reproduce the shape on the result end 07:45
but the distribution of the operations can view the elements as flat
timotimo right, the flatness makes it quite a bit less complicated i think
TimToady as long as none of the dimensions are extensible 07:46
then we're back to pointerland
timotimo yesterday i had code like "my @yoffs = ((100, 80, 45, 15) >>*>> nqp::time_n()) >>->> H;" or something similar
it was oodles of times slower than the same code with an explicit for loop over that parcel
and i'm just saying: in this particular case, we could have done much better - with a bunch of intelligence in place, yeah, but still ... 07:48
TimToady I tend to use X* and X- in those cases 07:49
timotimo ah, that's a good point
TimToady then you can at least lose the parens around the list 07:50
not that it would necessarily run any faster
timotimo i could probably check the performance of that
TimToady any metaop will have some setup and teardown overhead 07:51
FROGGS and that won't go away once we multi-thread these
TimToady in fact, deciding not to multithread will also be an overhead :) 07:52
FROGGS true :o)
timotimo hopefully we'll be able to decide those things at something similar to "spesh time" in many cases
TimToady most of the GLR can be summarized as "wait till you have a known source and destination before you decide how to do it" 07:55
timotimo i'd like to TL;DR that as "procrastinate! it's more efficient!"
FROGGS yeah
TimToady including such information as how long the list is known to be (finite or infinite too)
TimToady is developing a crossbar of how various source and dest policies interact 07:56
TimToady probably the src will specify a list of acceptable policies in order of preference, and the dest will then pick one or more to negotiate with its downstream dest 08:02
such that you cannot try to feed an infinite list to an "all" policy, even indirectly
TimToady this relates to jdv79++'s question about ranges as well 08:05
currently a range will autobatch itself because that's much more efficient on average than doing the lazy range 08:06
timotimo yeah, no matter how small the overhead is, simply generating "the next number" is probably always going to be darn cheap 08:07
TimToady my current list api poliices are: lazy, batch, silo, all, sink, hyper and race
the difference betwen batch and silo basically being sync vs async 08:08
timotimo so silo behaves a bit like CSP? 08:10
TimToady I think of it as a circular buffer with both a reader and a writer 08:10
timotimo that's fair 08:11
TimToady whereas with batch, it's "please produce as many as you can, up to N, then get back to me"
either of batch or silo could have slow-start semantics too, adjusting the batch size as we know more about the data stream 08:12
but hopefully, the GLR will also make pure lazy faster as well, since it can reduce to function calls just returning the next value, with no buffering overhead 08:13
assuming we generate functions with sufficient state to track the interation of whatever we're iterating 08:14
timotimo sounds just a tiny bit over-engineered :P
TimToady well, but once we've got functions rather than methods, we get more chance of speshing, I suspect
and inlining 08:15
mostly, the GLR is trying to get rid of asking the same question over and over
or from a different angle, hoisting the negotiation out of the datastream, and making .gimme and .reify go away 08:16
timotimo that does sound quite good.
dalek k-simple: 8fd7a35 | (Timo Paulssen)++ | lib/GTK/Simple.pm6:
now the timeout actually works

and it gives both the t and dt.		 08:17
FROGGS ohh, getting rid of .gimme and .reify sounds very nice, these were quite expensive in the few profiles I've seen 08:18
TimToady and basically trying to get the OO overhead out of data streaming
figuring out where we can cheat on immutable iterators is also a big part of it 08:19
but that takes src/dest knowledge, again 08:20
nwc10 FROGGS: The t-shirts have reached the geizhals office. 08:26
FROGGS \o/
timotimo r: my num $foo = 0; # imo this should work just fine 08:29
camelia rakudo-moar 442a46: OUTPUT«This type cannot unbox to a native number␤ in block <unit> at /tmp/tmpfile:1␤␤»
( no output )
timotimo well, it does on parrot
i can look into that. maybe on the car ride if we get into a traffic jam and the car moves slow enough for me to not get sick :|
TimToady r: my num $foo = "foo" 08:30
camelia rakudo-moar 442a46: OUTPUT«This type cannot unbox to a native number␤ in block <unit> at /tmp/tmpfile:1␤␤»
..rakudo-parrot 442a46: OUTPUT«Cannot convert string to number: base-10 number must begin with valid digits or '.' in '⏏foo' (indicated by ⏏)␤ in method Num at gen/parrot/CORE.setting:13608␤ in method Num at gen/parrot/CORE.setting:6103␤ in any at gen/parrot/BOOTSTRAP.nq…»
timotimo seems like we're not even trying
TimToady arguably, natives are already coercion types, where the coercion in question is unboxing 08:31
but ought to be a little broader than that, maybe 08:32
at mininum, we ought to be able to say my num() $foo = 0 08:33
timotimo new troll strategy: pass a tr or s regex to functions that expect some kind of smart matcher ... 08:43
m: my @test = <hello goodbye>; say @test.grep(tr/e/a/); say @test; 08:44
camelia rakudo-moar 442a46: OUTPUT«␤hello goodbye␤»
timotimo m: my @test = <hello goodbye>; say @test.grep(s/e/a/); say @test; 08:45
camelia rakudo-moar 442a46: OUTPUT«␤hello goodbye␤»
timotimo fair enough.
psch hi #perl6 08:50
timotimo hey peə 08:51
psch i recognize that character! timotimo o/ 08:51
fwiw, 'grep: 1' on a line alone segfaults my repl 08:52
FROGGS psch: also 'FOO: 1' ? 08:54
psch FROGGS: yup
this is on moar
FROGGS then maybe labels are not repl compatible?
I can't test that right now... have no perl6-m here 08:55
timotimo it does the same thing on my machine fwiw 08:55
bartolin there is an old ticket (RT #116695) describing that .delete (now :delete) on the last element of an array removed preceding elements which where "manually" set to Any. lizmat++ explained that setting an array element to Any is not the same as deleting an element. would you say that the following code demonstrates the current (and expected) behaviour in a good way? (the Any in $a[1] was created by autovivification and did not 'exist', so it gets delet
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=116695
timotimo could it be it's trying to figure out the $?FILE and explodes?
bartolin m: my $a; $a[0,2]=(Any,"foo"); say $a; my $e; for ^$a.elems { $e[$_] = $a[$_] :exists }; say $e; say "... deleting"; $a[2] :delete; say $a; say $a[0] :exists;
camelia rakudo-moar 442a46: OUTPUT«(Any) (Any) foo␤True False True␤... deleting␤(Any)␤True␤»
timotimo nope, that works fine 08:55
brrt \o 08:56
bugs?
FROGGS no, only people here 08:57
timotimo: it could very well be related to $?FILES
timotimo no rabbits here 08:59
brrt any rodents then? 09:05
TimToady well, elephant shrews are really elephants, not rodents 09:18
brrt very true 09:19
laben hello #perl6, has APW begun? 09:22
nwc10 laben: no, schedule starts at 10:00 CEST tomorrow 09:23
the talks will be recorded, but not live streamed
laben oh, right it was 10/10
TimToady CEST la vie 09:24
laben can someone try to test XML::Query? i would like to know if it's me or the module that fail 09:26
dalek kudo/nom: d3a2841 | (Tobias Leich)++ | src/core/Label.pm:
fix usage of Labels in REPL, discovered by psch++
 09:29
FROGGS timotimo++ # forgot to mention that 09:30
timotimo ah, not bad :)
very good "report to fix" time on that one, FROGGS++ 09:31
TimToady
.oO(Elephants Of Unusual Size)
 09:32
FROGGS ohh, U.S. is actually for "Unusual Size"? 09:33
BSc_tadzik :D 09:33
FROGGS that would explain my GSoC shirt :o)
timotimo :D
mine too!
FROGGS surely, it has an "L" in it, but I meant "large", not "LARGE!!" 09:34
timotimo if i had super low hanging pants to go with that shirt, that would be another story %)
FROGGS Low Hanging Pants.... I love it :o)
BSc_tadzik yeah, my gsoc t-skirt is huge too 09:34
timotimo what about your BSc t-shirt? 09:35
moritz laben: what do you mean by "test"? Run the unit tests?
t/basic.t .. Default constructor for 'Query' only takes named arguments
laben morits: yep, that's it 09:35
timotimo that doesn't seem like something that'd be a regression from changes in rakudo 09:36
laben moritz: i too get that, but i was warrying since tony-o said he didn't get that error
moritz laben: I just did a very fresh rakudo build, bootstrapped panda, and got that error too
laben: so not just you
laben timotimo: i don't think it's recent, much more probably it's a change in syntax or semantic some time ago
timotimo mhm 09:37
we do have a smoker now, but we're not watching the results closely enough ...
laben i was just trying to understand why did it give that error in particular and how to fix it
but i still cannot make it out
timotimo someone was XML::Query.new(1, 2, 3) 09:38
but it should have been XML::Query.new(:foo(1), :bar(2), :baz(3))
laben i tried changing the method new in Query.pm6 but to no avail 09:39
timotimo oh, it has a new method already? 09:40
something ends up calling "right past" that constructor
laben timotimo: yes and it's strange since it's basically new ($args) { self.new(:$args)} which doesn't make much sense to me 09:41
timotimo is that a multi method new?
laben i changed that to new ($args) {self.bless(...)}
timotimo: yes 09:42
timotimo in that case, it must be ignoring that particular candidate
try putting a "say" in front
i suspect that say wouldn't get triggered
laben in front of what?
timotimo at the very beginning of the method new you changed 09:43
timotimo brrt: do the jit ops "usecapture" and "savecapture" as well as "capturehasnameds" and "captureposelems" seem tricky to jit? or are they due for a rewrite when the args passing changes happen? 09:43
brrt lemmesee 09:44
timotimo interestingly, my heavily NativeCall using program here only bails 3x on nativecallinvoke
perhaps that means many of the postcircumfix:<( )> are inlined into a big frame that happen to abort earlier due to whatever
brrt usecapture is easy, savecapture can be functified 09:45
timotimo do we have a jit argument shorthand constant thingie for "pass the current capture here" for our "emit c function call" thingie? 09:46
brrt .. no
timotimo do we need one?
brrt not that i can tell 09:47
timotimo cool 09:48
a tiny something for the hackathon, then :) 09:49
hm, also a few bails on getobjsc, that should be jittable easily
just look into the header, where the pointer to the sc is stored, right?
laben timotimo: there is something strange going on there
timotimo or use MVM_sc_get_obj_sc as a C function 09:50
ah, needs to be that C function, as it may want to demand the object to be deserialized first
i'm not sure we can just say "at spesh time, that code has been run often enough that that doesn't matter any more!" 09:51
brrt i'll go get lunch now :-)
timotimo good lunch!
laben timotimo: without modifying anything, the result changes if i use the compiled version in blib/lib or the textual version in lib 09:52
like this, "env PERL6LIB=lib perl6 t/basic.t" works, but "env PERL6LIB=blib/lib perl6 t/basic.t" doesn't
brrt timotimo: pushed 09:53
timotimo i'm getting excited :3 10:30
FROGGS about what? :o) 10:31
timotimo i'm going to APW!
FROGGS and you are with awesome ppl in a car 10:32
timotimo aye
it kind of sounded like there'd be another person besides me in that car with liz and woolfy
is that right?
FROGGS leont, aye
timotimo cool
FROGGS yepp 10:33
ohh damn...
I wanted to take a book with me so that TimToady could sign it :/
timotimo i don't have any perl-related books :S 10:34
FROGGS GET OUTTA HERE! *g*
I have one or two books, but just skimmed them 10:36
timotimo perhaps at some point i could write a perl-related book :P
FROGGS do you think you could be/are a good writer?
I know that I'm not 10:37
timotimo i don't think i could :\
FROGGS well, I'm not sad about it... I know where my strengths are... and where my strengths are not that well distributed 10:38
so I try to do the things I can do :o) 10:39
these are usually the things that are fun anyway
though, I enjoyed writing the stubs for the story of my game... but that was more about writing down thoughts instead of writing a proper story 10:40
what do I do for the next four hours train? I should probably read about ::, :::, cut and commit 10:41
ewww, or I don't 10:43
timotimo what's ewww about that? :(
FROGGS "Evaluating a triple colon throws away all saved choice points since
the current regex was entered. Backtracking to (or past) this point
will fail the rule outright (no matter where in the regex it occurs)"
like this throw away thing... how do I do that? 10:44
timotimo empty the backtrack stack of all the cursors or something ... no clue :\
oh, "since the current regex was entered"
that makes it more doable
FROGGS so ::: is commit? 10:45
ahh, kinda
<cut> might be the one that is the easiest of that bunch 10:50
timotimo could be, yeah
FROGGS hmmmm, STD's moreinfo does not look that hard at a first glance 10:53
timotimo i don't know what that is :S 10:58
FROGGS the hook that will be called when the grammar likes to have more input :o)
like, you write a few lines in a repl and it will trigger at the final } 10:59
(or so)
so we would eval when that happens, and not per line as we do
timotimo ah 11:02
FROGGS err, s/moreinfo/moreinput/, sorry
timotimo well, i tried to port that and it turned out to not work, but that was long long ago
FROGGS yeah, many things have changes 11:03
changed*
timotimo OK, i should get off IRC and make sure i'm ready to go 11:04
FROGGS see you in a bit :o) 11:05
timotimo "bit" :\
looking forward to it :)
FROGGS yay, I can leave that train in less then two hours /o/ 11:06
timotimo yay 11:40
and I will get into the car in five minutes
FROGGS :o) 11:41
timotimo does cut look like something you can work out? or perhaps morelines ? 11:42
FROGGS I'm working on moreinput right now
timotimo cool 11:44
FROGGS I currently try to tweak it so that it fetches more input in rule ws (by doing an interactive readline), and in rule ENDSTMT I'm going to evaluate the code we got 11:45
so that we eval multiline stuff when we hit an ENDSTMT only 11:46
moritz attention, <ws> is *very* performance sensitive 11:48
FROGGS I know :o)
FROGGS stage parse has not changed yet, so that is a not too bad indication 11:50
pepl nine: location for pre-conf dinner has changed to Imlauer/Stiegl Bräu (near central station) 11:57
peteretep Anyone feel they have any particular insight as to why CPAN6 stalled? 12:29
[Coke] tony-o: I removed nqp. nope. it's checking the -installed- copy, not hte build copy.
lichtkind feel like hulk hogan 12:57
FROGGS[mobile] why? 12:58
lichtkind just ment out of oral exam: 1,0 :)
went
BSc_tadzik is 1.0 the highest you can get? :) 12:58
lichtkind yes
BSc_tadzik I went out with 5.0 from my defense :P
congratulations :) 12:59
FROGGS[mobile] O.o
lichtkind thanks :) but it was a 5 point course not a final one :)
FROGGS[mobile] wow, congrats
FROGGS yay, I'm sitting in the last train for today 13:00
lichtkind to austria? , thanks man
FROGGS to Salzburg, aye
Ulti BSc_tadzik congrats :3 I'm now PhD_Ulti too, picked up my hardbound thesis just now! .o/
BSc_tadzik Ulti: yay, awesome :) 13:01
FROGGS damn, what's happening here? :D
Ulti haha
BSc_tadzik FROGGS: education!
FROGGS Ulti: congratulations!
BSc_tadzik: I am just employed :/
Ulti thanks, it was a PhD mostly in Perl 5 hacking though :3
lichtkind yea this what i plan too
Ulti I'll try for the nobel in Perl 6 :P
BSc_tadzik my project was also Perl
lichtkind my thesis will be a perl 5 program
BSc_tadzik I promised to opensource it :) 13:02
colomon lichtkind++ BSc_tadzik++
FROGGS hehe, nice
lichtkind for linguists and other social studies :)
Ulti oh I notice you already had the title prefix/postfix debate 13:02
lichtkind what you mean? 13:03
FROGGS there was a discussion about titles yesterday 13:03
Ulti in the UK degree letters go after your name
lichtkind my most import title i would consider is DJ :) 13:04
Ulti postfix it to be edgy
scumdog1 Can I ask why
sub bar(*@args) { @args.elems.say; }; bar(1, Nil, 3); # => 2
says 2 and not 3?
lichtkind but my favorite dj's dont have it in their names :) 13:05
FROGGS scumdog1: because Nil was meant to vanish in lists
scumdog1: but it is supposed not to anymore, so that needs to be fixed
FROGGS scumdog1: this will one of the topics of the next days actually 13:06
(not directly fixing it, but discussing the implementation)
scumdog1 thankyou, so 13:07
sub bar2(|) { nqp::elems(nqp::p6argvmarray()).say; }; bar2(1, Nil, 3); # => 3
is the best way for now?
FROGGS scumdog1: probably, yeah 13:07
jnthn waves from Salzburg
FROGGS waves from Munich 13:08
FROGGS jnthn: we're meeting here: www.imlauer.at/de/restaurant/index....rg-anreise 13:09
FROGGS jnthn: ohh, even in english: www.imlauer.at/en/restaurant/index....rg-anreise 13:09
the title has changed at least *g* 13:10
lichtkind FROGGS hava a great time there, monday i have my next exam which is my very good exuse not to be in salzburg 13:11
jnthn FROGGS: Yes, that's the pre-conf meet this evening?
FROGGS lichtkind: yeah, that's a good reason
jnthn: yes 13:12
We want pre-conf meat!
jnthn FROGGS: Cool. I plan to join this evening. Pm plans to come along with me at same time. :)
(We were on same flight from Frankfurt :))
FROGGS yay!
coincidence?
jnthn Yes :)
FROGGS hehe
even nicer 13:13
lichtkind but now that i have moar and panda running i can at least post my goto tool to perl 6
jnthn japhb was on it too! :)
Very Perl 6 flight :)
But it went from gate 4. Not gate 6. Dang.
FROGGS good that it came down properly :o)
jnthn Aye. *and* I made it
FROGGS ... and you were three, not four
jnthn Which at one point was starting to look questionable...
FROGGS why that? 13:14
jnthn Headwind + long wait for a plane parking place + the bus from the plane to the terminal ended up in a traffic jam at the airport entrance too! 13:14
FROGGS >.< 13:15
what the heck
jnthn Yeah. German organisation at work!
FROGGS bah
jnthn To be fair, I did make my flight :P
FROGGS Germans are usually better at this
moritz
.oO( the probably use Gearman :-)
FROGGS *g* 13:16
jnthn Anyway, I think I'm going to find a quick nom, and then hide away to write talk, code for talk, etc... :) 13:16
FROGGS yeah, have a good nom then :o)
azawawi feather.perl6.nl:8080/ # Farabi6, Perl 6 editing in browser. May break under pressure :) 13:17
lichtkind a perl 6 binding for reaper would be heaven 13:18
FROGGS azawawi: sadly my internet connection here seems not good enough to test it :/ 13:19
azawawi FROGGS: panda install Farabi6 # and run locally :)
lichtkind i wanted test farabi 6 yesterday
at least it installed :)
azawawi lichtkind: hi there
lichtkind cheers 13:20
azawawi btw, feather.perl6.nl is under pressure from a starman process
lichtkind azawawi wx::scintilla doesnt compiles :)
BSc_tadzik i.imgur.com/e3Blgj6.jpg
FROGGS *g* 13:21
azawawi lichtkind: you caught me :)
lichtkind bad joke *g*
azawawi i wanted test it because im again do a lot of stuff with wx 13:22
just look into kephra repo there are a lot of demo programs that might be also useful to you
azawawi lichtkind: will take a look at it over the weekend 13:23
lichtkind azawawi: actually i looked at it because std wx stc still screws on defined or op but now that finally 3.0 got wrapped we might finally breath without wx::scintilla :) 13:24
azawawi lichtkind: hopefully :)
lichtkind: this latest farabi6 ui proof of concept tries to be like ipython notebook 13:25
lichtkind: opening files vertically instead of tabs with output pane for each one of them
lichtkind yes multi view in different design 13:31
also working on that with kephra
azawawi the pieces you might found most interesting are in the proto/function dir bitbucket.org/lichtkind/kephra/src...ion/?at=cp 13:32
azawawi lichtkind: this is another design that i tried feather.perl6.nl/~azawawi/
lichtkind how you do highlighting? 13:33
azawawi JS-based cheat highlight perl5 + perl6 keywords 13:34
:)
lichtkind so nothing fancy
azawawi not at the moment, need to focus on bigger stuff :)
lichtkind im also close before dip stc becasue most of parts i have to redo anyway for the features i need
carlin is it possible to `require 'somescript.pl6'` and then call a sub that is declared in somescript.pl6?
moritz carlin: nope
azawawi watchs "How the Camel is de-cocooning" 13:35
carlin darn
azawawi watches "How the Camel is de-cocooning" :)
lichtkind: Remind me again please. What happenned to the Perl 6 tablets that you were working on? 13:36
FROGGS m: require Test <&plan>; plan(1) # carlin 13:37
camelia rakudo-moar d3a284: OUTPUT«1..1␤# Looks like you planned 1 tests, but ran 0␤»
FROGGS moritz: see^
moritz FROGGS: tht'as not a .pl6 script
FROGGS this should also work with a script in case the sub are exported
perl6 -e 'require "lib/Test.pm" <&plan>; plan(1)' 13:38
Trying to import symbols &plan from 'lib/Test.pm', but it does not export anything
in sub REQUIRE_IMPORT at src/gen/m-CORE.setting:18815
:o(
lichtkind azawawi: they are online and well but not complete but when im in good spirits i update or even enhance parts
tablets.perl6.org
FROGGS perl6 -e 'require Test:file<lib/Test.pm> <&plan>; plan(1)' 13:39
1..1
# Looks like you planned 1 tests, but ran 0
moritz: ^
that's script-ish enough for me
moritz woah 13:40
FROGGS++
I didn't know that worked
moritz maybe lizmat++ and timotimo++ too (the usual suspects on that front :-) 13:40
FROGGS I just remembered that there was a test about that
(S11-modules/require.t)
carlin aha that works 13:43
FROGGS++
FROGGS nice :o)
azawawi p6doc -f say # Returns <In order to use unqualified sub and method names like "p6doc -f say" you will need to run "p6doc-index build" to build index.data otherwise use "p6doc -f Str.split" instead of "p6doc -f split" for now> 13:46
same with p6doc -f Str.split with p6doc -f Str.say 13:47
confusing
lichtkind cool gearge carlin is da house 13:48
cant spell george
PerlJam
.oO( Why would anyone bring the corpse of a dead comedian into the house?!? )
lichtkind carlin just reminded me 13:49
leedo PerlJam: ever seen Weekend At Bernie's? 13:52
:P
PerlJam Carlin has been dead a bit longer than Bernie was at the time :) 13:53
FROGGS do they stack dead comedians in a house in that movie?
leedo almost! 13:54
FROGGS almost? they arn't actually dead? 13:54
:o)
carlin isn't actually dead 13:55
timotimo ohai
FROGGS carlin: very true :o)
timotimo: ohai
timotimo woolfy is caring us now
FROGGS timotimo: when you stretch your arms, will leont do the keyboard cat then? :D 13:56
timotimo wat? :)
FROGGS nvm *g*
be nice :P
BSc_tadzik wat :D 13:57
timotimo trying 13:58
being nice is in my nature
FROGGS :/
leont gives no talk 13:59
FROGGS ohh, I can also see the Alps properly 14:04
azawawi How do i read and write to an external process in Perl 6? 14:19
lichtkind azawawi: last rant , i hope for the day than we have full runtime introspection int whol semantics of anyl language through perl 6 this will be a new class of editors 14:21
azawawi true 14:22
PerlJam azawawi: Proc::Async maybe? That's where I'd start anyway. 14:26
(I'm fairly sure there are unimplemented bits, though) 14:27
carlin this is what I did with the magic that FROGGS++ showed earlier: github.com/carbin/stomp/commit/830...c48f3a1455 14:29
moritz you have file names that contain double colons? 14:33
that won't work on windows, iirc
FROGGS moritz: that interpolates 14:34
m: require Test:<<$foo::bar>>
camelia rakudo-moar d3a284: OUTPUT«use of uninitialized value $v of type Any in string context in code at /tmp/99pBGZmPkt:1␤␤use of uninitialized value $v of type Any in string context in block <unit> at /tmp/99pBGZmPkt:1␤␤No such method 'value' for invocant of type 'Parcel'␤ …»
moritz FROGGS: yes, but into :file
FROGGS why not? 14:35
ahh
wait
moritz that's a file name, not a module name, right?
FROGGS m: require Test:file<<$foo::bar>>
camelia rakudo-moar d3a284: OUTPUT«use of uninitialized value $v of type Any in string context in block <unit> at /tmp/WT4e1FxLdl:1␤␤»
FROGGS file name, aye
Test would be the module name alias 14:36
carlin m: my $x = 'lib/Test.pm'; require Test:file<<$x>> <&plan>; plan(1) 14:37
camelia rakudo-moar d3a284: OUTPUT«1..1␤# Looks like you planned 1 tests, but ran 0␤»
carlin $Stomp::Config::Hooks is a variable, with the location of the file 14:40
carlin it seems to work, it's running the code from ~/.stomp/hooks, is that not supposed to work? 14:45
azawawi PerlJam: im looking at it github.com/rakudo/rakudo/blob/nom/...c/Async.pm . Thanks :) 14:47
carlin m: module Stomp::Config { our $Hooks = 'lib/Test.pm6' }; require Test:file<<$Stomp::Config::Hooks>> <&plan>; plan(1); 14:54
camelia rakudo-moar d3a284: OUTPUT«1..1␤# Looks like you planned 1 tests, but ran 0␤»
azawawi & 14:55
timotimo where is the social going to be today? 16:03
timotimo are you all assembled somewhere yet? 16:04
jnthn timotimo: www.imlauer.at/de/restaurant/index....rg-anreise 16:09
timotimo: I'm planning to head over there in a bit...
timotimo: Not exactly sure when.
timotimo thanks 16:12
FROGGS yays, I've got internet again... 16:13
jnthn yays 16:14
FROGGS okay, I am walking to Imlauer now, and hope to be there in about 30min
pmichaud good evening, #perl6 16:15
jnthn o/ pmichaud
FROGGS and since I've got internet on my laptop only, I can only check here every ten minutes or so :o)
jnthn FROGGS: You're going to stop walking, and check IRC on your laptop? :)
isBEKaml FROGGS: you walk for 30 minutes? 16:15
timotimo will be there soon 16:16
TimToady the wiki sez it starts at 8:00 16:19
but maybe we can have an early shift of dinner? 16:20
dalek p-js: 466c150 | (Pawel Murias)++ | src/vm/js/ (5 files):
Hacky support for knowhow instances attributes (will be flesh out in future commits).
TimToady btw, I couldn't get in on the hotel wifi, had to hijack someone's netgear 16:22
pmichaud TimToady: o/
TimToady well, I could get in through bluetooth to my telephone and thence to HOTELPLAINBRUEKE, but then port 22 seems to be blocked 16:23
pmichaud: \o
are you here? 16:24
FROGGS TimToady: I'm (was) at the same hotel 16:25
now I walked in some random direction :S 16:26
ohh, direction seems fine
TimToady: if you go to the shop right nearby, you can by a mobile phone card with 7gigs download volume for 15eur 16:27
(there is only one card left)
the shop closes in 33mins though
jnthn TimToady: pmichaud and I are planning to set off for the dinner place in 10 mins or so 16:28
uh, 5 now :)
FROGGS jnthn: aye, I'm sitting on the pavement currently :o)
TimToady we'll join you
jnthn TimToady: So will be there a bit earlier :)
FROGGS isBEKaml: what's wrong with walking?
jnthn FROGGS: So comfort!
FROGGS hehe
I'll have beer in a bit, it will be all fine by then 16:29
:o)
isBEKaml FROGGS: nothing wrong. Just found it odd that you're willing to walk that much to have a beer. :P
FROGGS hey! I'll also meet ppl and will have something to eat :o)
isBEKaml :o) 16:30
pmichaud yes, I'm here... and on my way to dinner. see you soon 16:33
afk
jnthn on way too & 16:34
Ven Well hello #perl6 16:55
FROGGS we're sitting in the garden of that hotel/restaurant Imlauer
Ven Waving from timo and me's room o/
timotimo o/ 16:56
carlin waves from the other side of the world 17:02
timotimo we didn't tell leont when we'd be showing up at the reception desk again to go to the social ... m) 17:10
brrt wishes the #perl6 people lots of fun on APW 17:11
FROGGS thank you :o) 17:12
jeffreykegler Question: I'm looking for the Perl 6 doc which describes the notation for pattern repetitions, particular with separators ... 17:13
jeffreykegler words or numbers separated by commas, that sort of thing ... 17:13
I recall reading some very cool stuff and now I can't find it.
brrt also the regular perl people, of course
timotimo jeffreykegler: do you mean for matching in regexes? 17:14
jeffreykegler YEs
timotimo like / (\d+)+ %% "," /
that should be syn 05
jeffreykegler Yes!!!
Thanks much!
timotimo YW :)
one of my fav features of p6 regexes
timotimo BBL 17:15
jeffreykegler On another subject, while I'm here, we'd talked plans for doing a BNF grammar of Perl 6, as a start toward a prototype Marpa parser, and even got a volunteer to do it ... 17:16
leon__ arrived safely in Salzburg :-) 17:17
jeffreykegler Unfortunately, due to press of time, that's not likely to happen very soon. Just too much other stuff has come up. Just thought I'd put it on the log in case anyone was watching this particular effort.
leon__ And now that he picked the right network, he might even have usable internet 17:18
nine anyone already at the pre conf meeting? 17:25
lizmat and woolfy are now heading to the pre-conf as well 17:40
yoleaux 07:14Z <jnthn> lizmat: well, good timing...I've not had thse happen on Windows at all, and so haven't had chance to hunt the thing down yet...provided there's decent debugging stuff on your laptop I might finally stand a chance :)
oskie hello, is nqp intended as a programming language for humans, or is it only an intermediate language? 18:13
brrt nqp is a programming language all right 18:14
PIR is an intermediate language
brrt it can be tricky though (only binding, no assignment) 18:15
but in general i like it well enough :-)
oskie hmm, interesting 18:18
dalek ast: 6e02af8 | usev6++ | S04-phasers/enter-leave.t:
Add test for RT #118387
 18:24
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=118387
dalek ast: f386c96 | usev6++ | S32-array/delete.t:
Add test for RT #116695
 20:02
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=116695
dalek kudo-star-daily: d207c97 | coke++ | log/MoarVM-version.log:
today (automated commit)
 20:40
rl6-roast-data: 1acdc9d | coke++ | / (4 files):
today (automated commit)
[Coke] moar clean; parrot still has two failures! 20:40
LHF: please create a parrot-specific skip/RT for those tests. 20:41
(or, better, fix 'em.)
dalek ast: 67583ed | usev6++ | S32-exceptions/misc.t:
Add test for RT #116607
 20:44
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=116607
tony-o [Coke]: for future ref, Configure checks ./install/bin/nqp-[m|j] before building 20:51
attempting to build p6 in cygwin 20:52
[Coke] tony-o: Yes; it checkes the INSTALLED version. 20:58
this is suboptimal when you have installed in a system location.
I should be able to force a rebuild from rakudo with a config option.
[Coke] (and not by removing the installed version which someone might be using) 20:59
[Coke] RT: down to 775 tickets. 21:05
usev6++ 21:06
bartolin thanks! to me it feels useful to go through tickets and try to grok the different bits of the language. 21:07
[Coke] usev6 == bartolin? 21:08
bartolin if someone sees mistakes, please point them out (but I think you do that anyway)
yes, it's me ;-)
[Coke] bartolin++
m: say +Failure; 21:09
camelia rakudo-moar d3a284: OUTPUT«Invocant requires an instance, but a type object was passed␤ in method Numeric at src/gen/m-CORE.setting:13672␤ in block <unit> at /tmp/nxvAUTiTCs:1␤␤»
[Coke] m: say "abcd".index("xyz")==0
camelia rakudo-moar d3a284: OUTPUT«Invocant requires an instance, but a type object was passed␤ in method Bridge at src/gen/m-CORE.setting:4767␤ in sub infix:<==> at src/gen/m-CORE.setting:4688␤ in block <unit> at /tmp/hcu3GZPd2U:1␤␤»
[Coke] m: say Failure == 0
camelia rakudo-moar d3a284: OUTPUT«Invocant requires an instance, but a type object was passed␤ in method Numeric at src/gen/m-CORE.setting:13672␤ in sub infix:<==> at src/gen/m-CORE.setting:4530␤ in block <unit> at /tmp/RZ60dqSCYI:1␤␤»
[Coke] bartolin: ^^ RT #76294 updates.
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...l?id=76294
[Coke] Unsure what the desired behavior is anymore, though. 21:10
bartolin [Coke]: I think I didn't read that one yet. 21:11
dalek ast: f15d9ae | usev6++ | S05-metachars/tilde.t:
Add fudged test for RT #77616
 21:13
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...l?id=77616
[Coke] bartolin: commented on 112988 for you - you want "is_run" 21:14
bartolin * reading 21:15
bartolin [Coke]: about 112988: I already thought about is_run. since the original request was to omit excessive errors, I have to do something additional to looking for "Could not open <file>". 21:19
[Coke]: (that was in the excessive error as well.) But I could check for number of newlines or something.
[Coke]: that just feels a bit crude ...
bartolin [Coke]: btw, is_run is a nice thing to have. used it just two hours ago for 118387 where I wanted to check stderr and stdout separately 21:23
hagiri ops 21:52
sorry!
[Coke] bartolin: "error starts with the text we're testing for, and doesn't have more than one newline"? 21:56
bartolin [Coke]: oh! but how to make sure error is not excessive? I'll have a look at that later. but thanks for the pointers! 22:00
bartolin good night, #perl6 22:12
Ven well, that was a nice dinner 22:28
next to timotimo, in front of nine, with leon__ aside
just on the side*, er
azawawi finally initial ANSI color support in farabi6 github.com/azawawi/farabi6/commit/...7718c1c051 23:23
tony-o has anyone successfully built perl6 in cygwni? 23:25
tony-o guess not, there is no pthread_barrier 23:31
azawawi feather.perl6.nl:3030/ # mounted over perl6-examples with initial ANSI color support for error messages... Dont use Firefox.. it will die lol 23:36
tony-o the link to feather.perl6.nl/~foo/ points to feather.perl6.nl/~/ 23:39
azawawi: 23:40
azawawi tony-o: ? 23:42
tony-o azawawi: the link to feather.perl6.nl/~foo/ points to feather.perl6.nl/~/
azawawi tony-o: no idea why it does that but i remember there is an Apache mapping somewhere (e.g. feather.perl6.nl/~azawawi/ 23:44
tony-o oic - it wasn't meant to be a hardcoded link to somewhere,just depends on who published whatever 23:47
azawawi btw why is perl6 warning about deprecated IO::Spec => $*SPEC 23:50
Any reason behind that change?

Powered by Cro and the Raku® Programming Language.

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

Thank you!