»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.
00:00 kivutar left 00:02 berekuk left 00:03 berekuk joined 00:11 btyler left
clsn Whyfor aren't ⸨⸩ considered a bracketing pair? They're Ps/Pe and all. But they aren't in the lists of openers/closers in Rakudo or in Std. I think there are others too. Std does mention in a comment that the list should be based directly on Unicode tables, but isn't yet. 00:15
00:15 Vlavv_ left 00:16 aindilis left, aindilis joined 00:20 lue left 00:22 jeffreykegler joined 00:27 rurban1 left
TimToady clsn: it was not added till Unicode 5.1, so probably after we generated that list 00:27
the STD list was just copied out of pugs, which came even ealier 00:29
*earlier
feel free to automate that now that we actually have Unicode props :)
smls Does P6 support P5-style subpattern recursion within a single regex (i.e. without defining a full-fledged grammar with multiple tokens)? 00:31
TimToady in theory 00:32
jnthn Well, you can always define it as my token foo { ... <&foo> ... } or so to lighten it 00:33
Otherwise there's <~~>
TimToady I think smls is looking more for <~~1> and such to reuse whatever matched $1
jnthn Ah
I'm not sure that's implemented
smls see rosettacode.org/wiki/Balanced_brackets#Perl for example
jnthn 'cus it's...hard to do. 00:34
smls subpattern recursion allows the P5 solution to be nice and pretty :)
jnthn Like many other things, but this one is rarely asked for :)
smls sub balanced { shift =~ /^ (\[ (?1)* \])* $/x; } 00:35
timotimo hard to do? :(
timotimo has a hard-to-do thing already on his plate
smls well ok, in my editor with syntax highlighting it looks nice :)
00:36 daniel-s left
jnthn timotimo: It's not greedy to have more than one... :P 00:38
00:38 Vlavv_ joined
smls in this example, P5 subpattern-recursion vs P6 grammars is just a stylistic difference 00:39
but the P5 recursion allows backtracking across subpatterns
and P6 tokens prevent that, right?
jnthn token does, regex don't 00:42
smls ah ok
TimToady but you don't really want backtracking here anyway 00:44
the current P6 solution using a grammar is busted, fixing
smls yeah, that's why I said in this example there's no difference
TimToady see new rosettacode.org/wiki/Balanced_brack..._a_grammar 00:46
reduced the grammar to a one-liner 00:47
since TOP now implicitly checks ^ and $
or rather, .parse does
smls nifty :) 00:48
TimToady in fact, wouldn't work if TOP implicitly checked, since we call TOP recursively 00:50
00:52 berekuk left
smls Of course the main problem with P5 recursive regexes, is that they discard all information about what they matched - except for the *last* *top-level* match of each capture group. 00:52
grondilu TimToady: isn't there a way to shorten brace expansion as well? I suspect the "meta" rule is not really necessary. 00:55
TimToady grondilu: I couldn't make it work; the different terminators mess up the levels 00:56
and doing things like passing a $level arg didn't seem to pan out 00:57
also, didn't work with the '{' ~ '}' idiom
00:58 rurban1 joined
TimToady because that commits in a way that seems detrimental to backtracking 00:58
I tried a *lot* of things before the initial entry...
though most of that was trying to build up the X during the parse, which didn't work out well at all, because I think rakudo is biased towards action routines that are not inlined 00:59
so I ended up with a separate walker
grondilu well, if there's room for improvement, someone will find them eventually. 01:00
jnthn TimToady: If you tried it on JVM, until a couple of weeks ago there was a serious bug with closures inside of regexes.
TimToady the walker could probably be simplified/obfuscated too to make it "pointless"
jnthn TimToady: Which could well have gotten in the way of inline actions
TimToady but I didn't want to explain [X~]
I tried it on all three, in fact 01:01
dunno what I was running into
but { make ... } would not pass an ast up
and all my attempts to set a $<foo> directly met with failure
it's probably better to do it outside with a separate walker anyway, because then you don't have to throw stuff away on backtracking 01:02
timotimo oh my
"not" xx not $balanced
TimToady smls: yes, we notice that problem with P5, which is why P6 does it differently :)
TimToady grins at timotimo 01:03
smls :)
TimToady you can do the same trick, kinda, in P5
only of course it's x there
konnjuta how many types of encoding does rakudo currently have? 01:04
TimToady three: the Good, the Bad, and the Ugly
01:05 jnap left
TimToady seriously, it's probably highly platform dependent at the moment 01:05
jnthn Yeah, I think it'll depend by backend somewhat 01:06
I imagine latin-1, ascii and utf-8 can re relied upon to exist
konnjuta is it possible to create custom encoding? 01:07
say come up with my own charmps
to read legacy data
01:07 arnsholt left
TimToady r: say "foo".encode('ISO-8858-15').decode('UTF-8') 01:08
timotimo that's backend-specific, but there should be a way; even if not yet
camelia rakudo-jvm d6a93b: OUTPUT«Unknown encoding 'iso-8858-15'␤ in method encode at gen/jvm/CORE.setting:6275␤ in block at /tmp/tmpfile:1␤␤»
..rakudo-parrot d6a93b: OUTPUT«encoding #-1 not found␤ in method encode at gen/parrot/CORE.setting:6279␤ in block at /tmp/tmpfile:1␤␤»
..rakudo-moar d6a93b: OUTPUT«Unknown string encoding: 'iso-8858-15'␤ in method encode at src/gen/m-CORE.setting:6275␤ in block at /tmp/tmpfile:1␤␤»
TimToady not "euro"-centric :) 01:09
r: say "foo".encode('ISO-8858-1').decode('UTF-8')
camelia rakudo-moar d6a93b: OUTPUT«Unknown string encoding: 'iso-8858-1'␤ in method encode at src/gen/m-CORE.setting:6275␤ in block at /tmp/tmpfile:1␤␤»
..rakudo-jvm d6a93b: OUTPUT«Unknown encoding 'iso-8858-1'␤ in method encode at gen/jvm/CORE.setting:6275␤ in block at /tmp/tmpfile:1␤␤»
..rakudo-parrot d6a93b: OUTPUT«encoding #-1 not found␤ in method encode at gen/parrot/CORE.setting:6279␤ in block at /tmp/tmpfile:1␤␤»
TimToady oops
r: say "foo".encode('ISO-8859-1').decode('UTF-8')
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«foo␤»
TimToady r: say "foo".encode('ISO-8859-15').decode('UTF-8')
camelia rakudo-jvm d6a93b: OUTPUT«Unknown encoding 'iso-8859-15'␤ in method encode at gen/jvm/CORE.setting:6275␤ in block at /tmp/tmpfile:1␤␤»
..rakudo-moar d6a93b: OUTPUT«Unknown string encoding: 'iso-8859-15'␤ in method encode at src/gen/m-CORE.setting:6275␤ in block at /tmp/tmpfile:1␤␤»
..rakudo-parrot d6a93b: OUTPUT«encoding #-1 not found␤ in method encode at gen/parrot/CORE.setting:6279␤ in block at /tmp/tmpfile:1␤␤»
TimToady n: say "foo".encode('ISO-8859-15').decode('UTF-8') 01:10
camelia niecza v24-109-g48a8de3: OUTPUT«foo␤»
TimToady niecza++
n: say "foo".encode('UTF-32') 01:11
camelia niecza v24-109-g48a8de3: OUTPUT«Buf.new(...)␤»
TimToady n: say "foo".encode('UTF-32').chrs 01:12
camelia niecza v24-109-g48a8de3: OUTPUT«Unhandled exception: Cannot unbox a BigInteger from an object of repr P6box[Byte[]]␤ at /home/p6eval/niecza/lib/CORE.setting line 1496 (chr @ 4) ␤ at /home/p6eval/niecza/lib/CORE.setting line 1498 (ANON @ 4) ␤ at <unknown> line 0 (KERNEL map …»
TimToady huh
n: say "foo".encode('UTF-32')[]
camelia niecza v24-109-g48a8de3: OUTPUT«Buf.new(...)␤»
TimToady n: say "foo".encode('UTF-32')[*]
camelia niecza v24-109-g48a8de3: OUTPUT«Buf.new(...)␤» 01:13
konnjuta perl6 -e 'open("-",:!chomp).lines>>.encode("ascii").say' < /tmp/utffile
TimToady sigh
konnjuta Lossy conversion to single byte encoding in method encode at gen/parrot/CORE.setting:6277 in method dispatch:<hyper> at gen/parrot/CORE.setting:1245 in block at -e:1
can't lossy conversion be done with a subchar?
TimToady n: say "foo".encode('UTF-32')[^3]
camelia niecza v24-109-g48a8de3: OUTPUT«Buf.new(...) (Any) (Any)␤»
TimToady p: say "foo".encode('UTF-32')[^3]
camelia rakudo-parrot d6a93b: OUTPUT«encoding #-1 not found␤ in method encode at gen/parrot/CORE.setting:6279␤ in block at /tmp/zo6MU4Vysi:1␤␤»
TimToady that message is LTA
jnthn r: say "foo".encode('UTF-32')[^3] 01:14
camelia rakudo-moar d6a93b: OUTPUT«Unknown string encoding: 'utf32'␤ in method encode at src/gen/m-CORE.setting:6275␤ in block at /tmp/tmpfile:1␤␤»
..rakudo-jvm d6a93b: OUTPUT«102 111 111␤»
..rakudo-parrot d6a93b: OUTPUT«encoding #-1 not found␤ in method encode at gen/parrot/CORE.setting:6279␤ in block at /tmp/tmpfile:1␤␤»
jnthn Moar message seems better, and it works on JVM
TimToady j: say "foo".encode('UTF-32').chrs 01:15
camelia rakudo-jvm d6a93b: OUTPUT«No such method 'chrs' for invocant of type 'utf32'␤ in block at /tmp/2Vm7XH8XTP:1␤␤»
jnthn What does .chrs on a Buf mean?
TimToady nothin'
jnthn OK, so that's correct... :)
TimToady j: say "foo".encode('UTF-32').list 01:16
camelia rakudo-jvm d6a93b: OUTPUT«102 111 111␤»
konnjuta lta = long term abuse?
TimToady less than awesome 01:17
jnthn haha :)
TimToady perlcabal.org/syn/S99.html#LTA
hah: 17:17 < TimToady> less than awesome 01:18
17:17 < jnthn> haha :)
17:17 < TimToady> perlcabal.org/syn/S99.html#LTA
er, wrong paste
jnthn haha :)
TimToady but if you follow that link, you'll find an autopun 01:19
01:19 arnsholt joined
timotimo hm. i wonder if :!chomp would have helped me ~1 year ago when i tried to handle $a_socket.lines on top of parrot 01:20
TimToady though as definitions go, it's a bit LTA
timotimo (the problem i ran into was not being able to differentiate "a full buffer's read, or just one line that ended on a newline" 01:22
and since i was reading some json data that was sent to me via the wire, i ran into some trouble trying to parse it ... 01:23
01:29 rurban2 joined 01:32 rurban1 left
konnjuta r: "\x41".print 01:35
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«A»
konnjuta r: "\x41\xE2\x82\xAC".print 01:36
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«A⬻
01:44 lue joined 01:46 jnap joined
TimToady grondilu: I added a point-free version of crosswalk at the end of the rosettacode.org/wiki/Brace_expansion entry for Perl 6 01:55
TimToady wonders why it doesn't build a TOC till there's four languages...
02:02 rurban2 left
konnjuta perl -e 'print "\x41\xE2\x82\xAC"' | perl -le 'print unpack("H*",<>)' 02:05
41e282ac
perl6 -e 'print "\x41\xE2\x82\xAC"' | perl -le 'print unpack("H*",<>)'
41c3a2c282c2ac 02:06
whats happening here?
skids wide unicode characters? 02:07
jnthn \xE2 in a string refers to Unicode code-point E2, not a byte E2 02:08
If you want to work with bytes, use Buf, not strings.
smls TimToady: yikes, what happened to this solution: rosettacode.org/wiki/First-class_fu...ons#Perl_6
konnjuta thank you 02:09
jnthn yay, I think I'm an hour so so's more work of having a deliverable FOSDEM talk. 02:10
I guess that means I should get some sleep before the flight.. :)
'night, #perl6
skids sleep tight. Don't let the unicode byte 02:11
smls TimToady: never mind, I fixed it. 02:17
02:20 FROGGS left 02:21 smls left 02:25 kbaker joined 02:34 FROGGS joined 02:36 Alula left 02:38 Alula joined
TimToady smash: yeah, someone was using a bad browser and/or editor that thought it would helpfully decode to ASCII 02:38
I saw them getting told in another discussion page somewhere, so it shouldn't happen again soon 02:39
but we'll need to keep our eyes peeled for other such damage
didn't ingy++ have a download stashed somewhere? 02:42
or maybe I can figure out whose edits did it and look at those pages 02:44
02:44 regreg joined 02:54 dayangkun_ left 02:57 thou left 03:04 grondilu left 03:06 rurban1 joined
konnjuta r: (<[ ]> xx 5).pick(*).join.say 03:22
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«[][][][][]␤»
konnjuta r: (<[ ]> xx 5).pick(*).join.say
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«[][][][][]␤»
konnjuta r: (<[ ]> xx 5).pick(*).elem
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«No such method 'elem' for invocant of type 'List'␤ in block at /tmp/tmpfile:1␤␤»
konnjuta r: (<[ ]> xx 5).pick(*).elems 03:23
camelia ( no output )
konnjuta r: (<[ ]> xx 5).pick(*).elems.say
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«5␤»
lue Apparently I signed up for RosettaCode way back in 2010 :P 03:24
konnjuta seems like list is not flatten? 03:28
r: (<[ ]> xx 5).pick(*).list.elems.say 03:29
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«5␤»
konnjuta r: (<[ ]> xx 5).list.pick(*).elems.say
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«5␤»
konnjuta r: (<[ ]> xx 5).list.elems.say 03:30
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«5␤»
konnjuta r: @(<[ ]> xx 5).list.elems.say 03:31
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«5␤»
konnjuta r: (<[ ]> xx 5).flat.elems.say 03:36
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«10␤»
konnjuta r: my @a=(<[ ]> xx 5); 03:37
camelia ( no output )
konnjuta @a.elems.say
r: my @a=(<[ ]> xx 5); @a.elems.say
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«10␤»
konnjuta r: (<[ ]> xx 5).elems.say
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«5␤»
raydiak r: my @a=(<[ ]> xx 5); @a.perl.say; 03:40
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«Array.new("[", "]", "[", "]", "[", "]", "[", "]", "[", "]")␤»
raydiak r: (<[ ]> xx 5).perl.say 03:41
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«(("[", "]"), ("[", "]"), ("[", "]"), ("[", "]"), ("[", "]")).list␤»
konnjuta seems like (<[ ]> xx 5) does not flatten 03:42
r: (<[ ]> xx 5).flat.perl.say
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«("[", "]", "[", "]", "[", "]", "[", "]", "[", "]").list␤»
raydiak it's not supposed to flatten, it's a parcel until you assign it to something else, like an array
TimToady it's actually a parcel of 5 parcels 03:43
lue konnjuta: it does flatten when you use .flat
TimToady r: (<[ ]> xx 5).perl.say
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«(("[", "]"), ("[", "]"), ("[", "]"), ("[", "]"), ("[", "]")).list␤»
TimToady which is why I changed it for the RC entry 03:44
I note that the .roll($n*2) entries don't have this problem
otoh, they don't guarantee an even number of opening/closing brackets
which is okay for that problem 03:45
r: say pick, *, <[ ]> xx 5 03:46
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«===SORRY!=== Error while compiling /tmp/tmpfile␤Calling 'pick' requires arguments␤ Expected any of: :(Any $n, *@values)␤at /tmp/tmpfile:1␤------> say ⏏pick, *, <[ ]>…»
TimToady r: say pick *, <[ ]> xx 5
camelia rakudo-parrot d6a93b: OUTPUT«[ ] [ ] ] [ ] [ ] [␤»
..rakudo-moar d6a93b: OUTPUT«[ ] [ ] ] ] ] [ [ [␤»
..rakudo-jvm d6a93b: OUTPUT«[ ] ] ] [ ] [ [ ] [␤»
TimToady that works too
since a slurpy flattens
konnjuta under what condition in addition to (<[ ]> xx 5) does it no flatten? 03:49
TimToady parcels never flatten until bound into something that flattens it 03:50
konnjuta ok
TimToady and xx is defined to return a parcel for each iteration
the outer parens have nothing to do with it
r: given <[ ]> xx 5 { .perl.say } 03:51
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«(("[", "]"), ("[", "]"), ("[", "]"), ("[", "]"), ("[", "]")).list␤»
TimToady see, no outer parens, still not flat
r: given (((<[ ]> xx 5))) { .perl.say } 03:52
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«(("[", "]"), ("[", "]"), ("[", "]"), ("[", "]"), ("[", "]")).list␤»
TimToady extra parens don't matter either, unless there's a comma
r: given (((<[ ]> xx 5),),) { .perl.say }
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«(((("[", "]"), ("[", "]"), ("[", "]"), ("[", "]"), ("[", "]")).list,),)␤»
03:56 jnap left 04:02 BenGoldberg left 04:11 regreg left 04:19 colomon left 04:21 colomon joined 04:26 jnap joined 04:31 jnap left 04:48 SamuraiJack joined 05:02 btyler joined
TimToady okay, turned in my SOTO application for OSCON, though now I see that Geoffrey has also proposed it with the Lightning Talks 05:18
05:23 [Sno] left, kbaker left 05:27 jnap joined 05:31 jnap left 05:36 jeffreykegler left 05:44 mcglk left 05:55 Alina-malina joined
konnjuta how to convert string to buf? 06:04
moritz with .encode
with .encode($encoding) to be precise
06:17 sqirrel_ joined 06:27 [Sno] joined 06:28 jnap joined
lue grumbles in the direction of all things list-like in Perl 6 06:29
06:29 zby_home_ joined 06:32 jnap left
raydiak contemplates putting together a list behaviour matrix which crosses parcel, list, array, capture, etc, with mutable, lazy, flattening, context-imposing, and so forth 06:35
06:38 kaleem joined 06:39 darutoko joined
TimToady revised rosettacode.org/wiki/ABC_Problem#Perl_6 to be correct in the face of different blocksets 06:41
06:41 rurban1 left
lue I can make Nil disappear anytime I want to, so long as it's not the code I'm trying to fix >_< 06:41
konnjuta r: (<1 2>, <1 2>).WHAT 06:42
camelia ( no output )
konnjuta r: (<1 2>, <1 2>).WHAT.say
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«(Parcel)␤»
konnjuta r: (<1 2>, <1 2>).elems.say
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«4␤»
konnjuta r: (<1 2>, <1 2>).list.WHAT.say 06:43
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«(List)␤»
konnjuta r: (<1 2>, <1 2>).list.elems.say
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«2␤»
konnjuta r: (<1 2> xx 2 ).WHAT.say
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«(List)␤»
konnjuta r: (<1 2> xx 2 ).elems.say
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«2␤»
konnjuta r: (<1 2> xx 2, ).WHAT.say 06:44
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«(Parcel)␤»
konnjuta r: (<1 2> xx 2, ).elems.say
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«4␤»
konnjuta So parcel flattens but list don't?
lue If someone can tell me why I can't make the Nils disappear, that'd be great: gist.github.com/lue/2761b448d04219e23cc4 06:45
moritz how about not using 'make Nil' if you don't want Nils? :-) 06:48
TimToady what moritz sez 06:50
06:51 sqirrel_ left, bibifuc_ is now known as bibifuc
TimToady neither parcels nor lists are interested in flattening 06:51
it's how you use them
lue moritz: I haven't experimented as much with not doing make Nil, though I didn't have much success there either. 06:52
TimToady what do you want instead of the Nil?
lue nothing, that's why I was counting on Nil's "disappear from the array" trick 06:53
TimToady Nils have not disappeared from lists for most of a year now...
except as fossilized bugs
lue (the list will be turned into a Bag later, and I don't want (Any)=> to be an entry)
r: say (1, Nil, 3).Array
TimToady did you try () instead?
camelia rakudo-parrot d6a93b, rakudo-jvm d6a93b, rakudo-moar d6a93b: OUTPUT«1 3␤»
lue TimToady: no, testing... 06:54
TimToady interesting, something is itemizing it
I'm guessing there's an itemize somewhere in make, when maybe there shouldn't be 06:57
lue TimToady: that might explain it, $() and things refusing to flatten seem to be plaguing me at the moment. 06:58
It could be I'm using not-the-absolute-latest builds of rakudo, my p6 is 2014.01-25-gfd8316f 06:59
TimToady sub make(Mu $ast) { <-- the $ is itemizing what you make, so () turns into $() 07:01
that should probably be using a \ instead of a $ 07:02
07:02 regreg joined
lue TimToady: replacing 'make Nil' with 'make ()' and using 'make $<word>».ast».Array.flat.Bag;' works for me. 07:03
TimToady testing a patch 07:05
07:07 FROGGS left 07:08 konnjuta left
lue goes for the night 07:09
raydiak goodnight lue
07:26 crab2313 joined 07:29 jnap joined 07:33 jnap left 07:35 aindilis left 07:41 kaare_ left 07:42 rurban1 joined 07:46 kaare_ joined 07:51 dmol joined 07:52 FROGGS joined 07:56 kaare_ left 08:01 kaare_ joined 08:05 zakharyas joined
dalek kudo/nom: 8128f6a | larry++ | src/core/Match.pm:
Don't itemize make's ast argument
08:10
FROGGS morning 08:12
raydiak \o morning FROGGS
TimToady o/ 08:14
08:15 berekuk joined, rurban1 left
dalek kudo/nom: 1d6accc | moritz++ | CREDITS:
larry is typically known as TimToady++
08:21
moritz 07:58 < lue> TimToady: that might explain it, $() and things refusing to flatten 08:22
of course it doesn't flatten, it begins with a $ !
08:25 tgt joined 08:29 jnap joined 08:33 BooK left 08:34 jnap left, BooK joined 08:37 pdcawley joined 08:39 rindolf joined 08:41 dmol left 08:42 fhelmberger joined 08:44 berekuk left 08:46 berekuk joined 08:50 tgt left
hoelzro morning #perl6! 08:58
raydiak good morning hoelzro 08:59
hoelzro moritz: raydiak
09:00 berekuk left
timotimo i can sense making make not itemize by default is going to chang lots of user code 09:03
but i agree that it's more sensible to use \foo rather than $foo for make.
hoelzro oops 09:04
that was supposed to be "morning raydiak" =P
timotimo why didn't it autocomplete to "mroing" or something? :P 09:05
09:05 bjz left, bjz_ joined
raydiak b/c it's not a cellphone? :) 09:05
09:08 bjz_ left, bjz joined
moritz because my nick isn't 'MRoing' 09:12
timotimo :) 09:14
can't you just interpolate @foobar into a regex? 09:20
rather than evaling "/" ~ .comb.join('|') ~ "/"?
09:20 rurban1 joined
FROGGS timotimo: you can 09:21
who says the opposite?
09:24 lestrrat left
timotimo the ABC problem code from RC does that :) 09:25
gotta run now.
09:25 lestrrat joined 09:27 sqirrel_ joined 09:29 tgt joined 09:30 jnap joined 09:35 jnap left 09:37 dmol joined 09:38 dmol left 09:39 zby_home_ left
masak mroing, #perl6 09:39
09:40 dmol joined
FROGGS mroing mrask! 09:40
raydiak good method resolution order, masak
masak (masak) (RT-bot) (bot) (Mu) 09:42
09:45 rurban1 left
masak "Blocks are stored as precompiled regexes." -- the code seems to have obsoleted the description here. rosettacode.org/wiki/ABC_Problem#Perl_6 09:48
oh wait.
no, there it is, EVAL and slashes.
never mind. :)
masak ponders if macros could do that
diakopter macros can do anarthing 09:49
masak except when they can't 09:57
diakopter .. but you can do anything within a macro :) 09:58
masak hehe 09:59
by that argument, subs and class declarations can do anarthing too 10:00
diakopter classes can do subs!
subs can do classes!
such do
masak p: class C1 { has $.x = sub { my class C2 { has $.y = sub { say "OH HAI" } }; return C2 } }; C1.new.x()().new.y()() 10:05
camelia rakudo-parrot 1d6acc: OUTPUT«OH HAI␤»
masak \o/
diakopter lolz 10:06
moritz subs in attributes in clases in subs in attributes in classes
such inception
10:13 btyler left
jnthn waves from l'airport 10:28
10:28 tgt left
FROGGS waves back from l'office 10:29
jnthn Late flight is late
Not much, thankfully
masak hi, jnthn from l'airport 10:30
10:31 jnap joined
moritz L'ate flight is late :-) 10:31
\o jnthn, FROGGS, *
10:31 crab2313 left
jnthn J'ai l'ate un cookie 10:32
10:33 zby_home_ joined
masak .oO( how l'ame ) 10:33
10:33 zby_home_ left 10:35 jnap left 10:47 daniel-s joined 10:50 rurban1 joined 10:56 daniel-s_ joined, daniel-s left
masak sometimes wishes he could declare named lexical rules and invoke them from a // regex 10:57
10:57 fhelmberger left
FROGGS masak: what do you mean? 10:58
jnthn /<&foo>/ ?
dalek kudo/nom: 1a44ca6 | jonathan++ | tools/build/Makefile-Moar.in:
Install the debugger on MoarVM.
10:59
FROGGS p: my &foo = regex { \w+ }; say "ohh?" ~~ / <&foo> /
camelia rakudo-parrot 1d6acc: OUTPUT«「ohh」␤␤»
jnthn r: my regex foo { \w+ }; say "ohh?" ~~ / <&foo> / 11:00
camelia rakudo-parrot 1d6acc, rakudo-jvm 1d6acc, rakudo-moar 1d6acc: OUTPUT«「ohh」␤␤»
diakopter masak: I wanted that so badly, I thought we already had it
11:01 Alula left
jnthn OK, time to wander gate-wards I guess... 11:01
bbl
masak jnthn: I suspected there was a syntax for it, thanks. 11:02
FROGGS resists to ask "Wie gates?"
masak haha
masak .oO( look to gate-wards )
FROGGS r: my regex foo { \w+ }; say "ohh?" ~~ / <foo> /
camelia rakudo-parrot 1d6acc, rakudo-jvm 1d6acc, rakudo-moar 1d6acc: OUTPUT«「ohh」␤ foo => 「ohh」␤␤»
masak FROGGS++
FROGGS :o)
masak wow, I don't even need the & 11:03
FROGGS ENOSPECIALSYNTAX
the hardest thing about it is the "my"
though I think the errmsg would point you in the right direction nowadays
r: regex foo { \w+ }; say "ohh?" ~~ / <foo> /
camelia rakudo-moar 1d6acc: OUTPUT«Useless declaration of a has-scoped method in mainline (did you mean 'my regex foo'?)␤No such method 'foo' for invocant of type 'Cursor'␤ in method ACCEPTS at src/gen/m-CORE.setting:12552␤ in method ACCEPTS at src/gen/m-CORE.setting:782␤ in bloc…»
..rakudo-parrot 1d6acc: OUTPUT«Useless declaration of a has-scoped method in mainline (did you mean 'my regex foo'?)␤No such method 'foo' for invocant of type 'Cursor'␤ in regex at /tmp/tmpfile:1␤ in method ACCEPTS at gen/parrot/CORE.setting:12575␤ in method ACCEPTS at gen…»
..rakudo-jvm 1d6acc: OUTPUT«Useless declaration of a has-scoped method in mainline (did you mean 'my regex foo'?)␤No such method 'foo' for invocant of type 'Cursor'␤ in method ACCEPTS at gen/jvm/CORE.setting:12554␤ in method ACCEPTS at gen/jvm/CORE.setting:782␤ in block at…»
FROGGS yeah :o) 11:04
masak wow, that made the code lots more readable. nice! 11:05
FROGGS yeah, I mean named captures already are nice in a sense of self-documenting code (/$<date>=.../), but declaring these rules somewhere and use these as captures is nice in several ways 11:07
masak indeed. 11:08
11:09 tgt joined
diakopter thinks the camelia smudger could make responses equivalent if they're errors and the first two lines of the error/backtrace are the same 11:12
masak aye. 11:13
p: my regex t { \s+ <(foo|bar)> }; say " bar" ~~ / <t> /
camelia rakudo-parrot 1d6acc: OUTPUT«「bar」␤ t => 「bar」␤␤»
masak p: my regex t { \s+ <(foo|bar)> }; say " bar" ~~ / <t>? / 11:14
camelia rakudo-parrot 1d6acc: OUTPUT«「」␤␤»
masak I'm... wondering if the that second one is correct?
FROGGS does not feel right 11:15
n: my regex t { \s+ <(foo|bar)> }; say " bar" ~~ / <t>? /
camelia niecza v24-109-g48a8de3: OUTPUT«「」␤␤»
FROGGS n: my regex t { \s+ <(foo|bar)> }; say " bar" ~~ / <t> /
camelia niecza v24-109-g48a8de3: OUTPUT«「bar」␤ t => 「bar」␤␤»
FROGGS hmmmm
rn: my regex t { \s+ <(foo|bar)> }; say " bar" ~~ / <t>* /
masak consistent across implementations.
11:15 rurban1 left
camelia rakudo-parrot 1d6acc, rakudo-jvm 1d6acc, rakudo-moar 1d6acc, niecza v24-109-g48a8de3: OUTPUT«「」␤␤» 11:15
FROGGS rn: my regex t { \s+ <(foo|bar)> }; say " bar" ~~ / <t>+ /
masak I'm feeling we're missing something.
camelia rakudo-parrot 1d6acc, rakudo-jvm 1d6acc, rakudo-moar 1d6acc, niecza v24-109-g48a8de3: OUTPUT«「bar」␤ t => 「bar」␤␤»
FROGGS rn: my regex t { <(foo|bar)> }; say "bar" ~~ / <t>? / 11:16
camelia rakudo-parrot 1d6acc, rakudo-jvm 1d6acc, rakudo-moar 1d6acc, niecza v24-109-g48a8de3: OUTPUT«「bar」␤ t => 「bar」␤␤»
FROGGS rn: my token t { \s+ <(foo|bar)> }; say " bar" ~~ / <t>? /
camelia rakudo-parrot 1d6acc, rakudo-jvm 1d6acc, rakudo-moar 1d6acc, niecza v24-109-g48a8de3: OUTPUT«「」␤␤»
FROGGS I dunno 11:17
rn: my token t { \s+ }; say " bar" ~~ / <t>? /
camelia rakudo-parrot 1d6acc, rakudo-jvm 1d6acc, rakudo-moar 1d6acc, niecza v24-109-g48a8de3: OUTPUT«「 」␤ t => 「 」␤␤»
FROGGS at least that makes sense to me :o)
masak then please explain it to me :) 11:18
rn: my token t { \s+ <(foo|bar)> }; say " bar" ~~ / <t>? /
camelia rakudo-parrot 1d6acc, rakudo-jvm 1d6acc, rakudo-moar 1d6acc, niecza v24-109-g48a8de3: OUTPUT«「」␤␤»
masak in the above, why isn't 'bar' captured?
11:19 tgt left
FROGGS rn: my token t { \s+ (foo|bar) }; say " bar" ~~ / <t>? / 11:19
camelia rakudo-parrot 1d6acc, rakudo-jvm 1d6acc, rakudo-moar 1d6acc, niecza v24-109-g48a8de3: OUTPUT«「 bar」␤ t => 「 bar」␤ 0 => 「bar」␤␤»
masak is thinking how this relates to github.com/perl6/specs/commit/4128...3fb4260f43
11:20 sqirrel_ left
FROGGS rn: my token t { \s+ <(foo|bar) }; say " bar" ~~ / <t>? / 11:21
camelia rakudo-parrot 1d6acc, rakudo-jvm 1d6acc, rakudo-moar 1d6acc: OUTPUT«===SORRY!=== Error while compiling /tmp/tmpfile␤Malformed regex␤at /tmp/tmpfile:1␤------> my token t { \s+ <(foo|bar⏏) }; say " bar" ~~ / <t>? /␤ expecting an…»
..niecza v24-109-g48a8de3: OUTPUT«===SORRY!===␤␤Unable to parse regex; couldn't find right brace at /tmp/tmpfile line 1:␤------> my token t { \s+ <(foo|bar⏏) }; say " bar" ~~ / <t>? /␤␤Parse failed␤␤»
FROGGS rn: my token t { \s+ <((foo|bar) }; say " bar" ~~ / <t>? /
camelia rakudo-parrot 1d6acc, rakudo-jvm 1d6acc, rakudo-moar 1d6acc, niecza v24-109-g48a8de3: OUTPUT«「 bar」␤ t => 「bar」␤ 0 => 「bar」␤␤»
FROGGS rn: my token t { \s+ <((foo|bar))> }; say " bar" ~~ / <t>? / 11:22
camelia rakudo-parrot 1d6acc, rakudo-jvm 1d6acc, rakudo-moar 1d6acc, niecza v24-109-g48a8de3: OUTPUT«「 bar」␤ t => 「bar」␤ 0 => 「bar」␤␤»
FROGGS interesting
rn: my token t { \s+ <(foo|bar)> }; say " bar" ~~ / <t>? /; say $/.perl 11:23
camelia rakudo-parrot 1d6acc, rakudo-moar 1d6acc: OUTPUT«「」␤␤Match.new(orig => " bar", from => 0, to => 0, ast => Any, list => ().list, hash => EnumMap.new())␤»
..niecza v24-109-g48a8de3: OUTPUT«「」␤␤#<match from(0) to(0) text() pos([].list) named({}.hash)>␤»
..rakudo-jvm 1d6acc: OUTPUT«「」␤␤Match.new(to => 0, hash => EnumMap.new(), ast => Any, list => ().list, orig => " bar", from => 0)␤»
masak niecza is wrong there. <( and )> are supposed to be independent.
FROGGS but <( is a thing on its own 11:24
so <(foo) is invalid
masak ah, troo.
anyway, I think there was a bug in there, but I'm too fuzzy-headed to analyze it. 11:25
FROGGS yeah, me too
11:29 tgt joined
daniel-s_ perl6: say 'hi' 11:30
camelia rakudo-parrot 1d6acc, rakudo-jvm 1d6acc, rakudo-moar 1d6acc, niecza v24-109-g48a8de3: OUTPUT«hi␤»
daniel-s_ :-) 11:31
11:32 jnap joined
FROGGS hi daniel-s_ 11:32
daniel-s_ Hey
I figure, camelia is always reliable about saying hi... even if it's only because she's running the code that tells her to do it. 11:33
11:34 tgt left 11:36 jnap left
masak oh hai daniel-s_ 11:36
some of us humans are fairly reliable too ;) 11:37
albeit admittedly less so
smash TimToady: huh?! ah, that wasn't for me, roger
11:41 daniel-s_ left 11:42 daniel-s_ joined
Ulti [Coke] what did you get up to re. bioperl people? 11:43
last time I spoke to them they said they weren't going to push for bioperl6 yet but would get back to me 11:44
11:58 tgt joined 12:02 tgt left 12:10 atroxaper joined 12:13 beastd joined 12:22 tgt joined
masak , for perhaps the first time ever, uses TDD purely to map out a domain 12:27
12:31 darutoko- joined 12:32 jnap joined 12:33 darutoko left 12:37 jnap left
timotimo hm, it would appear that's one of its strengths, no? 12:38
masak yes, it is.
but I usually do it mainly to implement something.
here I do it simply to *understand* what the domain *is*.
as in, I primarily expect the TDD process to *tell* me things.
(and it does)
timotimo as in "what is the exact size and shape of the domain"?
masak more like "how the heck does the domain work" 12:39
note, "domain" as in "problem domain", not as in "domain of a function".
timotimo that's what i thought, too
12:41 zacts left
lizmat is off to FOSDEM 12:47
see you there or be square :-)
12:47 lizmat left
masak .oO( □ ) 12:48
moritz
.oO( is that square fair? )
FROGGS is now known as froggsbob nottherepants
12:54 kaare_ left
colomon is a well-known square 13:03
13:07 kaare_ joined 13:15 tgt left, tgt joined 13:19 ajr joined 13:20 ajr is now known as Guest47646, Guest47646 is now known as ajr_ 13:25 denis_boyun joined 13:28 kaleem left 13:29 Ven joined 13:30 tgt left 13:33 jnap joined 13:38 jnap left 13:40 tgt joined, kbaker joined
Ven Hi ! I wanted to know if there was a "special variable" that contains the (lexically scoped) current loop iteration (like in awk) 13:42
masak Ven: no, we usually do @array.kv for that 13:44
p: my @array = <A B C D>; for @array.kv -> $i, $val { say "On index $i we have $val" } 13:45
camelia rakudo-parrot 1a44ca: OUTPUT«On index 0 we have A␤On index 1 we have B␤On index 2 we have C␤On index 3 we have D␤»
masak the "key" in an array is simply the index, as if it were a (non-sparse) Int-keyed hash. 13:46
it's a nicer solution, in my opinion. a hidden special variable would affect everyone's performance, not just those who used the feature. 13:48
13:49 woolfy1 left
moritz brings up the sufficiently smart compiler straw-man 13:50
Ven "a SSC would statically determine that the variable isn't used and not generate it" ? 13:51
moritz yes
masak yes, sure. 13:52
but even considering that, there's something to be said for being explicit.
moritz but we still hate special variables enough to reduce their number, not increase it
it's down to 4
masak if someone really liked the idea of a special variable, they could module it.
moritz ... with a macro
moritz ducks
masak ooh, that's a nice example.
yes, because they'd essentially be creating their own for loop, so it's legit. 13:53
moritz++
moritz "rebaes linearizes history" # 3 words 13:56
one of them missspelled
masak ww? 13:58
moritz yes 13:59
14:01 telex left
[Coke] Ulti: nothing happened, I am just in that channel. no substantial discussion. 14:01
14:02 telex joined 14:04 tgt left 14:05 denis_boyun left
[Coke] I did hear that a lot of people doing bio in perl5 are rolling their own instead of using BioPerl. 14:09
(because NIH, not for good reasons)
moritz because perl is so expressive, that in the beginning it's less of a hurdle to roll something on your own
instead of reading docs, and finding the stuff in the first place 14:10
14:10 kaare_ left
moritz and getting acquainted with the module author's terminology 14:10
masak indeed. 14:12
just because there's a module out there doesn't mean it's the right fit for everyone.
14:13 faceblaster joined 14:14 tgt joined
ajr_ In a field close the experimental edge, the terminology might not have settled down to the point that people discussing the same subject would recognise the fact. 14:17
moritz most bioperl tasks don't seem to apply to that
ajr_ (Not to mention professional competition generating a struggle over definitions.
14:24 SamuraiJack left, SamuraiJack joined 14:25 bjz left, bjz joined
Ven mmh, `class` syntax is a new perl6-feature, right ? It's kinda hard to google things sometimes haha, I got onto Moo's page 14:29
[Coke] if you're looking for perl6 docs, the advent calendar, doc.perl6.org and perlcabal.org/syn are good places to poke about. 14:30
(plus also the roast test suite)
(but to navigate that, you need the ../syn)
[Coke] hasn't done moo programming in a loooong time. 14:31
Ven [Coke]: yeah, that's there I navigate, roast test suite + github finder usually brings good result. I have doc.perl6 and perlcabal open and I get often on the advent calendar. (is it up-to-date with new standarts etc?)
moritz yes, 'class' syntax is a new feature in Perl 6 14:32
geekosaur is somewhat amused, considering that if you follow Moo backwards you end up back here 14:33
14:34 jnap joined 14:35 SamuraiJack left
masak geekosaur: I'm so old on this channel, I remember when stevan_++ was active around here, cooking up Moose. :) 14:37
geekosaur yep :)
masak Ven: the advent calendar is up-to-date, except in very rare cases.
Ven: if you think you've found such a case, we're interested to hear about it.
[Coke] is reminded that we have some roast tests for older advents, butnot newere. 14:38
14:38 tgt left
[Coke] *but not newer 14:38
14:39 jnap left
Ven moritz: what are the "4 special variables" still left ? Do you still count $^ variables ? $\d, $`, $' etc ? 14:47
moritz Ven: there are $`, $' etc. variables in Perl 6
Ven: $!, $_, $/ and $¢
the latter one is usually only used inside the regex engine
Ven uuh. :D 14:49
moritz sorry 14:50
there are *no* $`, $' etc. variables in Perl 6
missed that one, unimportant word
masak those two are simply attributes on $/ these days.
forgotten what exactly...
I just remember there was quite some discussion over the names :) 14:51
FROGGS $¢ might be $/.CURSOR
moritz $/.prematch and $/.postmatch iirc
FROGGS ahh, these 14:52
moritz m: 'abcdef' ~~ /c/; say $/.prematch
camelia rakudo-moar 1a44ca: OUTPUT«ab␤»
Ven very interesting :)
FROGGS p: 'abcdef' ~~ /c/; say $`
camelia rakudo-parrot 1a44ca: OUTPUT«===SORRY!=== Error while compiling /tmp/i5f432KmM2␤Unsupported use of $` variable; in Perl 6 please use explicit pattern before <(␤at /tmp/i5f432KmM2:1␤------> 'abcdef' ~~ /c/; say ⏏$`␤ expecting any of:…»
masak right .prematch and .postmatch -- moritz++
FROGGS std: $'
camelia std 09dda5b: OUTPUT«===SORRY!===␤Unsupported use of $'variable; in Perl 6 please use explicit pattern after )> at /tmp/W9f_2P1c5X line 1:␤------> $'⏏<EOL>␤Parse failed␤FAILED 00:01 121m␤»
masak Ven: Perl 6 is very thought-through. it is extremely earnest in trying to improve on and fix Perl 5. 14:53
FROGGS and I try to get all the nice P5 things back :o) 14:55
Ven masak: I'm afraid the only knowledge of perl is a IRC bot I did when I was 12 ;). Coming to perl 6 directly without experience from perl 5 (except some snippets I've read here and there)
tadzik ok, time for Star release :)
Ven But sometimes I think it's a hard and steep choice to do (like python did)
masak Ven: you are right. 14:56
Ven: and there's a long way for Perl 6 to go still.
Ven: what you see here is a group of enthusiasts who really believe in the idea, and carry it forward day by day.
moritz and the python step from 2 to 3 isn't nearly as steep as p5 to p6
masak and yet people seem embittered and disappointed even by Python's small step. 14:57
dalek ar: 9ed36f1 | tadzik++ | tools/star/release-guide.pod:
Fix release guide

git submodule sync doesn't have a --force option
14:59
masak Ven: I think the best way to approach Perl 6 is with a curiosity of what's possible. Perl 5 is awesome, but it has some known limitations. Perl 6 is all about thinking bigger and further than that, and to create a language which transcends Perl 5 while still remaining Perl.
15:00 skids left
moritz tadzik: oh, 'git submodule update' has the --force option 15:00
masak Ven: if that ends up sounding interesting, feel free to use Perl 6. if it sounds *very* interesting, feel free to get involved. :>
tadzik moritz: not here
moritz git 1.7.9.5
tadzik oh, update
Ven masak: definitely why I approached it
tadzik should it be 'update --force' then?
masak Ven: \o/ 15:01
Ven: let me know if I can be of some assistance. I was once where you are right now.
Ven: experience tells is that newcomers/volunteers have the best change of being helpful/useful by (a) doing what they like and (b) asking questions freely. 15:02
Ven masak: seems like a lot of fun. also used an altJS with a perl-like syntax, so several things feel home. I'm also interested in a lot of programming languages (generally), so generally I just think "oh that's like X in language Y", which helps a lot. But there are a *lot* of features. Read 4 synapses yet, I'll get some as PDFs. 15:03
FROGGS Ven: it feels like as if you belong to us :o) 15:04
15:04 btyler joined
moritz you have been 6ieted. Resistance is normal. 15:05
15:06 treehug88 joined
TimToady r: my regex foo { \w+ }; say "ohh?" ~~ / <foo> / 15:07
15:07 tgt joined
Ven One thing for sure, this community's reputation isnt stolen 15:07
camelia rakudo-parrot 1a44ca, rakudo-jvm 1a44ca, rakudo-moar 1a44ca: OUTPUT«「ohh」␤ foo => 「ohh」␤␤»
TimToady masak: don't need the & either
unless you want to force it to not mean a grammar rule, but here it's not ambiguous in any case
15:08 jnap joined
FROGGS just ooc, are dynamic subs specced? like &*foo? 15:08
dalek ar: 35aca2d | tadzik++ | tools/star/Makefile:
Update Makefile for 2014.01
ar: ebaa7e5 | tadzik++ | docs/announce/2014.01.md:
Add the new release announcement
tadzik proofreading will be welcome
FROGGS not that I'd need it, just wanted to know
15:08 thou joined
TimToady well, you have to call it using &*foo() 15:09
FROGGS yes yes
I am aware of that
TimToady to me it just falls out of everything else; dunno if it's specced
FROGGS p: my &*foo = sub { 42 }; &*foo()
camelia ( no output )
FROGGS p: my &*foo = sub { 42 }; say &*foo()
camelia rakudo-parrot 1a44ca: OUTPUT«42␤»
FROGGS mkay 15:10
(fall out)++
moritz tadzik: I've skimmed it, looks good
tadzik moritz: thank you
TimToady in S06 it has things like &*saith($next_id); # print the lexical $next_id;
dalek ar: 9f13743 | tadzik++ | / (2 files):
Update version numbers
15:11 raiph joined 15:12 tgt left
TimToady oh, I see you discovered the lack of need for &, nevermind :) 15:12
TimToady never knows whether to respond to the backlog now, or later after he's forgotten to...
FROGGS hehe
better now than never
moritz redundant replies don't hurt.
FROGGS except when they do 15:13
FROGGS .oO( just wanted to say a Toadism )
timotimo hi Ven
what does "this community's reputation isn't stolen" mean?
TimToady the problem is, if your pride takes a hit point, and it was already negative, you start taking the absolute value as a source of pride
(re redundant replies) 15:14
tadzik urgh, my star doesn't like my ICU, it seems
auto::icu - Is ICU installed...error: icushared not defined 15:15
does that ring a bell?
I also had to give it --icuheaders so it'd find them
FROGGS tadzik: you've got the /usr/include/unicode symlink?
ahh, hmmm
tadzik FROGGS: I don't think so
timotimo tadzik++ # star release 2014.01 15:16
tadzik FROGGS: but adding it helpde
might be worth telling people that
I'll check if it's in the readme
15:17 tgt joined
TimToady I do think a lot of our community's reputation was stolen at the beginning, but all new communities do that... 15:17
timotimo i don't know what "stolen" means in this context
TimToady not earned, I suspect
timotimo oh, of course
15:17 kaleem joined
TimToady at the start, Perl 6 was riding on the Perl 5 community's reputation, for instance 15:18
timotimo geekosaur is somewhat amused, considering that if you follow Moo backwards you end up back here -- is that because Moo backwards reads OOM and that's what perl6 tends to do for bigger programs? :P
moritz :-)
timotimo Ven: btw, i've come from python directly to perl6 and i'm quite happy here
TimToady and Perl 1 was riding on a whole heap of borrowed Unix memes
moritz timotimo++
TimToady not to mention Henry Spencer's regex engine :)
dalek ar: c2982ee | tadzik++ | README:
Note potential Parrot ICU problems in the README
15:19
timotimo next month's rakudo star may be the first rakudo jvm star and i'm very excited about that possibility
masak TimToady: re "stolen at the beginning" -- I disagree, to the extent that au++ single-handedly bootstrapped a community around Pugs. 15:20
tadzik in the long run it would be nice to have fixed on Parrot side
moritz masak: afiact Perl 6 started way before au++ was involved
masak oh, of course. 15:21
moritz sadly I wasn't there
masak me neither.
but up until Pugs, it was basically the Perl 5 and/or Parrot community building Perl 6.
moritz masak: when did you join, roughly?
raiph re perl6 advent bitrot: perlmonks.org/?node_id=1067940
TimToady well, au borrowed heavily from Haskell culture at that point too :)
15:21 tgt left
moritz for me it must have been in 2006 or 2007 15:21
masak moritz: I was aware of Perl 6 already sometime around 2003-2004. I was reading apocalypses in 2004. I joined the #perl6 channel in 2005. 15:22
TimToady and certainly we've been avoiding success at all costs since then :)
moritz the first rakudo commit mentioning me is from 2008
15:22 REPLeffect left
masak TimToady: was gonna say. people self-identified as "lambdacamels"; half Perlers, half Haskellers. 15:22
moritz first mu patch mentioned me in 207-02 15:23
TimToady wow, that was a long time ago :P
moritz erm
2007-02 :-)
tadzik :D
moritz and first roast patches in 2008-01
FROGGS I started recognizing Perl 6 at the YAPC in frankfurt 2012 :o) 15:25
(and fell in love)
15:26 atroxaper left
TimToady you probably had heatstroke 15:26
15:26 atroxaper joined 15:27 aindilis joined
masak that would explain the warm feelings 15:28
PerlJam blogs.perl.org/users/damian_conway/...urich.html
I see Damian will be giving an intro to perl6 course. 15:29
15:29 daniel-s_ left
PerlJam Maybe that will garner some new contributors :) 15:29
15:31 kaleem left
FROGGS well, my projects before involved hacking threaded stuff in xs (SDL game library), and so the talk about parrot's threads by seven++ where kinde sweet :o) 15:31
I either had ignored Perl 6 before or I have missed its existence for other reasons 15:32
15:35 denis_boyun joined 15:43 denis_boyun left
dalek kudo-star-daily: 064a2c7 | coke++ | log/ (5 files):
today (automated commit)
15:43
[Coke] lots of failures. :( 15:44
moritz :( 15:45
[Coke] panda, lwp-simple, dbiish... json? 15:47
15:47 census joined
timotimo how did that happen? 15:47
[Coke] I just report 'em.
timotimo ah 15:48
yeah, we changed make
to not automatically itemize any more
moritz oh dammit
that'll break json
[Coke] note: i'm testing with the latest rakudo, not the last rakudo-compiler tag. 15:49
timotimo irclog.perlgeek.de/perl6/2014-01-31#i_8202738
[Coke] so perhaps this doesn't matter for the star release. 15:50
moritz likely not
timotimo it doesn't, luckily
moritz builds a parrot again
Ven timotimo: Sorry, just got back. That's a french phrase I tried to translate, what I meant is that you didn't usurp your reputation, if that makes more sense. ya'all are nice and fun folks 15:51
(also hi and thanks for the help yesterday :P)
timotimo that's a part of what kept me here :P
15:51 dmol left
PerlJam Ven: #perl6 was founded on fun :) 15:52
tadzik uhh, it doesn't seem like modules got updated in the star release I prepared
json, Bailador and others still warn about EVAL
timotimo ah, you were vendethiel yesterday?
tadzik while it was clearly fixed recently
timotimo tadzik: i personally recommend "git submodule foreach gitk --all"
er, actually 15:53
"git submodule foreach git fetch\; gitk --all"
or ... some kind of quoting or something?
that'll give you a gui that shows where the checkout is, what's available, and you can right-click on any of the commits and select "check out this commit"
tadzik but well, at least the old ones all work :)
15:53 kbaker left
FROGGS hoelzro: ping 15:54
hoelzro: I palyed around with your openpipe... and I was wondering if using the macros from procops.c would help
hoelzro: so the result is this, which even seems to work: gist.github.com/FROGGS/998bacca7b3d3abab644
palyed*
played*
-.-
tadzik ah, the instructions are RONG agian 15:56
FROGGS aww, not agian
timotimo toldya its tyop day!
FROGGS \o{ 15:57
dalek ar: f269c3f | tadzik++ | modules/ (11 files):
Update modules, all but MIME::Base64
15:58
ar: 89ae451 | tadzik++ | tools/star/release-guide.pod:
Fix release guide to not tell releaser to rollback module updates
15:58 Psyche^ joined
tadzik I think I need to take a break and go buy some foods :o 16:01
16:02 faceblaster left, Psyche^_ left 16:13 panchiniak joined, FROGGS left, kbaker joined
dalek ecs: fcaf02c | larry++ | S05-regex.pod:
make uses parcel semantics
16:16
tadzik rakudo.org/downloads/star/rakudo-st....01.tar.gz is out! 16:17
moritz tadzik++
TimToady \o/
moritz it has escaped!
tadzik I'll quickly go and grab something to eat; I'll post all the announcements after the shoptrip :)
so there's still time to find a critical bug that'll make me crazy :P 16:18
shop &
TimToady testing 16:23
(on Linux Mint)
16:23 [Sno] left
moritz I have pushed a json fix; it should not even break on old rakudo 16:28
16:29 jeffreykegler joined
timotimo i expected as much 16:29
but i also expect some more breakage to show up in the smoker 16:30
[Coke] moritz++ tadzik++ TimToady++ 16:36
TimToady I am very glad we found this one before 6.0.0 :) 16:40
[Coke] is there a test?
TimToady not yet
but it's specced now :)
[Coke] not without a test it's not. :P 16:41
TimToady hopes at some point some minions will go through the specs and figure out what doesn't have a test yet :P
[Coke] you've only synned. :)
tadzik . o O ( without a test it's snot ) 16:42
timotimo got some good noms, tadzik? :)
tadzik TimToady: does it seem to work well?
timotimo: yeah :)
timotimo yay
tadzik timotimo: also, it's so freezing that I think I'll design a plan to always have a 2-weeks-worth of food at home
canned or whatever
TimToady tadzik: you mean the 'make' parcel patch? 16:43
tadzik TimToady: um, I mean the new star :)
TimToady it spectested before I pushed it
timotimo hah 16:44
i hope the weather will nicen up soon :)
TimToady it seems to have finished 'make install' successfully
tadzik okay 16:45
I'll send out the good enws
not sure if I have wordpress' rakudo.org/ account tho 16:46
TimToady trying a make modules-test
16:48 kaleem joined
tadzik moritz: do you have a writeaccess to rakudo.org? 16:49
TimToady passes modules-test 16:50
16:51 panchiniak left
PerlJam tadzik: do you need an invite? 16:51
16:52 FROGGS[mobile] joined
stevan_ masak: I am so old in this channel I remember when masak++ first arrived ;) 16:52
tadzik PerlJam: I think so 16:53
vendethiel timotimo: yes, that's me :P 16:55
dalek ar: a48e619 | tadzik++ | tools/star/release-guide.pod:
Update release guide
16:56
16:56 FROGGS[mobile]2 joined
tadzik ok, now rakudo.org/2014/01/31/rakudo-star-2...-released/ is out too :) 16:58
16:59 FROGGS[mobile] left
[Coke] Much of synopsis 9/11 17:00
timotimo =o 17:01
masak .oO( never forget ) 17:02
[Coke] m: say 9/11
camelia rakudo-moar 1a44ca: OUTPUT«0.818182␤»
17:02 rindolf left 17:03 rindolf joined
tadzik
.oO( miecza )
17:03
diakopter m: say e - 2 17:04
camelia rakudo-moar 1a44ca: OUTPUT«0.718281828459046␤»
diakopter m: say e - 1.9
camelia rakudo-moar 1a44ca: OUTPUT«0.818281828459046␤»
tadzik m: say e - (9/11 + 1) 17:06
camelia rakudo-moar 1a44ca: OUTPUT«0.900100010277227␤»
tadzik m: say e - (9/11 + 1.9)
camelia rakudo-moar 1a44ca: OUTPUT«0.00010001027722728␤»
17:07 panchiniak joined
timotimo m: say chr(0b0_000100010) 17:07
camelia rakudo-moar 1a44ca: OUTPUT«"␤»
PerlJam Who does the R* MSIs typically?
17:07 kaleem left 17:08 rindolf left
FROGGS[mobile]2 I am able to 17:08
17:08 rindolf joined
PerlJam The last one appears to be from September 2013. It would be nice to have one for the first R* of the new year. 17:09
tadzik any volunteers for Star 2014.02, or should I pick one? :> 17:10
FROGGS[mobile]2 I think I can do that this weekend
tadzik thanks FROGGS[mobile]2!
:P
FROGGS[mobile]2 nooooo!!!!
>.<
tadzik that'd be a daily R* release 17:11
timotimo :D 17:12
17:15 beastd left 17:16 skids joined 17:18 denisboyun joined 17:30 ajr_ left 17:33 zakharyas left 17:34 denisboyun left 17:37 Vlavv_ left 17:39 erkan left 17:42 atroxaper left
TimToady the current rosettacode.org/wiki/Brace_expansion#Perl_6 solution makes the Haskell solution look rather cluttered :) 17:45
17:45 denisboyun joined 17:49 Vlavv_ joined
moritz TimToady++ 17:49
though the use of try is a bit hacky :-) 17:50
diakopter rosettacode-driven idiom/language design
17:50 grondilu joined
grondilu perl6: say ++(my$ ) Ror 2 17:51
camelia niecza v24-109-g48a8de3: OUTPUT«===SORRY!===␤␤This macro cannot be used as a function at /tmp/tmpfile line 1:␤------> say ++(my$ ) Ror⏏ 2␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1502…»
..rakudo-parrot 1a44ca, rakudo-jvm 1a44ca, rakudo-moar 1a44ca: OUTPUT«1␤»
TimToady well, hey, Haskell uses try :)
grondilu was expecting 2 17:52
TimToady besides, the whole brace substitution meme is hacky in the first place :)
in this case, the try saves the redundancy of evaluating $<meta><alts> twice, and makes it read better, so I judged it to be worth it 17:54
17:54 denisboyun left
timotimo TimToady: what's the try needed for, ooc? 17:54
TimToady was just noticing that it works without the try too :) 17:55
moritz \o/
timotimo ah, ok
TimToady since it cascades Nil, I presume, even through the »
anyway, seems to work 17:56
timotimo i wonder, can we ».?&crosswalk
17:57 SamuraiJack joined
TimToady what, you think crosswalk might not be defined? :P 17:57
timotimo er ... hah
TimToady I can see some use for ».?&*dynasub 17:58
timotimo :D
TimToady but I'll be surprised if that parses currently
maybe it will
timotimo here we go again with the line noise :)
FROGGS[mobile]2 *g*
feels home
TimToady okay, try removed, and now it really is quite pretty 18:00
18:01 dmol joined 18:04 FROGGS joined 18:06 bjz left, bjz joined 18:07 benabik joined 18:09 [Sno] joined 18:10 tgt joined
PerlJam reading the description for brace expansion ... "$<meta> is likewise short for $/<meta>" while entirely true, saying it's "short for" when it's only one character seems weird. :) 18:11
18:16 arnsholt left, arnsholt joined 18:17 ajr joined
grondilu TimToady: have you tried using || instead of | in the grammar in order enforce resolution order? 18:17
18:17 ajr is now known as Guest5682 18:18 Guest5682 is now known as ajr_
TimToady I decided I don't care if it does LTM instead in this case 18:18
in fact, probably does the 2nd bracket matcher first
but it gets a pretty quick reject if there's something in the middle 18:19
also, need to propagate the LTMs to the other rules so | works right there; if I put || into meta, I'd have to put || everywhere 18:20
with LTM I don't actualy have to try meta if the dfa won't match, it goes straight to the . (in TOP) 18:21
well, nfa, but yeah 18:22
18:23 panchiniak left, kivutar joined
lue TimToady: I'm going to hazard a guess and say that code I plan on putting on RC probably shouldn't bother using ANSI colors to make things prettier :) 18:23
TimToady
.oO(A character here, a character there, pretty soon you're talking real bloat.)
18:24
18:28 SamuraiJack left, denisboyun joined
TimToady I wanna test the Haskell version, but I've forgotten how to download Text.Parsec (have ghc) 18:29
oh wait, aptitude found it
geekosaur normally you'd want to install haskell-platform but ubuntu broke it 18:30
diakopter it's a conspiracy.
18:31 jeffreykegler1 joined, rindolf left 18:32 berekuk joined, rindolf joined, darutoko- left, berekuk left, berekuk joined, jeffreykegler left 18:33 denisboyun left 18:37 rindolf left 18:38 rindolf joined 18:40 arnsholt left 18:41 jeffreykegler joined, vendethiel left 18:42 rindolf left 18:43 Alina-malina left, rindolf joined, jeffreykegler1 left 18:44 Alina-malina joined 18:49 erkan joined 18:52 arnsholt joined 18:56 rindolf left, rindolf joined 18:57 arnsholt left 19:00 denisboyun joined 19:08 arnsholt joined 19:10 zakharyas joined, rindolf left 19:11 rindolf joined 19:12 arnsholt left 19:15 erkan left 19:17 rindolf left 19:18 rindolf joined, araujo joined, jeffreykegler left 19:20 araujo left 19:24 arnsholt joined 19:33 smls joined 19:35 census left 19:37 jeffreykegler joined 19:38 jeffreykegler left 19:43 jeffreykegler joined, jeffreykegler left 19:47 arnsholt left 19:48 tgt left 19:49 tgt joined 19:50 rindolf left 19:51 arnsholt joined
raiph 10 python idioms: news.ycombinator.com/item?id=7151433 19:51
19:52 bluescreen10 joined
PerlJam raiph: btw, thanks for fighting the good fight on perlmonks 19:53
timotimo PerlJam, raiph, was that about the post on the advent calendar uptodateness? 19:55
19:57 arnsholt left
PerlJam timotimo: well, the original was about "Perl 7", but yeah. 19:57
timotimo i remember seeing that 19:58
that was pretty good, indeed
i kind of thought the reply by the anonymous monk was kind of lackluster
but oh well.
19:59 arnsholt joined
PerlJam Sometimes I think "Just ignore the anonymonk" but then I think "It's good that the anonymonk wasn't ignored because others might think his word true" 19:59
(but I tend to ignore anonymonks anyway because I don't frequent perlmonks much these days :)
lue has never really been to perlmonks, barring the occasional passing-by. 20:00
timotimo i was refering to the anomonk that answered raiphs stuff
i don't think that person understood the concept of the advent calendar; although it would probably be a good idea to put a short snippet of text in a colored box on top of every advent calendar post if there's something outdated in it and explain (as exactly as seems prudent) what's the new status 20:01
PerlJam aye. 20:03
lue timotimo: and/or a general site-wide disclaimer at the top of each page saying "Posts are contemporaneous, not current." :) 20:04
timotimo right.
but that does seem like an LTA to me 20:05
lue timotimo: it's one of those "It should be obvious" things
lue likes how make -j3 in the nqp repo builds all three nqps at once :D
20:05 arnsholt left
timotimo that's very enjoyable, indeed 20:05
PerlJam I think the anonymonk's point was that "Perl 6 isn't usable because the language isn't stable (enough)" or something.
timotimo re-reads 20:06
lue Could I have a link to this thing we're discussing? :)
timotimo perlmonks.org/?node_id=1067940
lue danke
timotimo okay, i don't think i actually know what that anonymonk referred to with "a significant portion of the important documentation you keep claiming that people will be working on in 2013"
20:08 pdcawley left
PerlJam I think that was a mistake. The anonymonk conflated the Synopses with the Advent articles. (that's just my guess anyway) 20:09
timotimo wow, that'd cause a pretty grave misunderstanding
[Coke] no change in rakudo/moar log files on daily roast.
timotimo "these crazy fools are only working on the specification on 24 days in the whole year!"
20:10 tgt left
lue timotimo: that "sig portion of imp. doc" thing is likely in response to "I'm not currently willing to go thru all 100+ articles." 20:10
20:11 tgt joined
[Coke] points at github.com/perl6/roast/tree/master/integration //advent* 20:11
we have at least partial coverage in the test suite for 2009&2010 20:12
lue As far as I can tell, Perl 7 is an idea by people who want a shiny new major version on their favorite language, but are peeved that the next major version of a language is taking so long :)
[Coke] LHF to add the rest (and we probably should)
20:13 kaare_ joined
PerlJam [Coke]++ 20:13
[Coke] but: If I go ahead and do that (which needs doing) that will not satsify anonymonk.
so anyone who does it, do it because it's right, not because he's wrong.
PerlJam cares not for anonymonk's desires
lue [Coke]: my senses tell me anonymonk does not want to be satisfied.
FROGGS [Coke]: pretty much nothing satisfies haters
just go on and deliver an awesome product 20:14
[Coke] FROGGS: exactly.
[Coke] sets up some quests for the various advent stuff.
timotimo lue: Anonymonk is probably in THERE IS SOMEONE WRONG ON THE INTERNET mode 20:16
[Coke] hurls questhub.io/player/coke
lue timotimo: agreed
[Coke] feel free to join in any of those.
timotimo some stuff to +1 \o/
masak 'night, #perl6
[Coke] +1 ALL MY THINGS!
timotimo FWIW, i looked through a bunch of advent calendar posts for testable code to copypaste and came up with not much 20:17
[Coke] *\[oo]/
timotimo i'd say whoever goes with that will have to do a more thorough reading and thinking than i did :)
lue [Coke]: should advent tests be contemporaneous or current? (e.g. leave eval($stuff) or change to EVAL($stuff))
[Coke] I had a bunch of code in 2013 that's testable.
lue: IMO, the code in the article should be updated to match current code.
that's a lot of work, which is why it isn't done. (ditto RC code) 20:18
but we shoudl either update the article in situ, or at least have a box at the top saying something like "This article displays an older style of code. <link to roast tests showing current style>" 20:19
lue [Coke]: by "RC code" you mean "script that downloads all completed tasks from RC one-by-one and tests them", right?
I think updated articles should update code + a box somewhere detailing the changes. 20:20
PerlJam lue: +!
er, +1
[Coke] lue: something like that. ingy has work in that direction. also see integration/rose* in roast.
moritz turns 2013 day 02 into tests 20:21
[Coke] moritz++
sweet, now I get credit for doing less work! ;)
lue
.oO(We clearly need gitblog for our advent posts, so we don't have to NIH changelogs)
PerlJam day 24 is going to be a little harder to turn into a test :)
FROGGS is somebody willing to test this? froggs.de/perl6/rakudo-star-2014.01-RC0.msi
e.g., test panda
colomon isn't it released already? 20:22
FROGGS colomon: not the msi :o)
lue colomon: the MSI is specifically released "sometime after the main release"
FROGGS colomon: it is the compiled version for windows
[Coke] FROGGS++ trying now. 20:23
FROGGS awesome!
[Coke]: thank you
[Coke] (don't tell anyone I have a sometimes win box)
timotimo i'd like to see the test that verifies "Clever dogs can do up to 10 words a minute, with surprisingly few typos." :P
FROGGS *g*
[Coke] I'd love for us to get a publisher cert.
FROGGS yeah
colomon is also downloading it
dalek ast: 1ff6d63 | moritz++ | integration/advent2013-02.t:
turn advent 2013 day 02 into a test
lue Rakudo doesn't do the same "all three rakudos at once with -j3" though :( 20:24
moritz lue: it doesn't?
lue oh, nevermind, I just wasn't patient enough for when it starts :)
FROGGS lue: my does
hehe
lue: yeah, you have to wait for parrot kicking in :o)
[Coke] FROGGS: installed it, C:\rakudo\bin\perl6.exe seems to work ok. 20:25
FROGGS cool, I hope the repl and panda is fine... then we can ship it
lue Well, it wavers between "3 jvm", "all three", and "2 jvm + 1 parrot" for me
PerlJam Can we fudge tests for just one rakudo implementation?
FROGGS PerlJam: yes
moritz yes
lue PerlJam: rakudo.parrot (etc.)
moritz #?rakudo.parrot 'OH NOEZ'
FROGGS #?rakudo.jvm skip
PerlJam nice.
FROGGS or todo or emit... 20:26
raiph timotimo: the anonymonk in question is almost certainly educated_{f,p}oo
[Coke] panda fails here, but EFIREWALL
FROGGS hmmm
[Coke] moritz++ is missing a todo or skip there.
moritz aye 20:27
timotimo raiph: well, if he aint got any other hobbies, let's share a moment of silence for his pitiful existence ...
raiph timotimo: he's a dedicated and unreasonable P6 troll
[Coke] FROGGS: but the REPL and the command line seem fine.
20:27 vendethiel joined
lue Funny thing, I still don't care about the troll :) 20:27
FROGGS [Coke]++ # thank you very much
[Coke] r: say [*] 2..1000000
^^ returns 0 on win7 20:28
FROGGS panda seems to work here
20:28 denisboyun left
camelia rakudo-parrot 1a44ca, rakudo-jvm 1a44ca, rakudo-moar 1a44ca: OUTPUT«(timeout)» 20:28
raiph lue: right, I try (and generally succeed I think) in following au++to go with
[Coke] wonders if questhub.io/realm/perl/quest/511a56...060700005f is worth doing these days - is there another smolder-alike that doesn't have the maintenance issues smolder does?
lue raiph: what does au say on the matter? 20:29
colomon can confirm I get working REPL on * 2014.1 on Windows Vista
20:29 treehug88 left
raiph lue: hug trolls: answer the substance, acknowledge any authentic feelings, ignore the rest 20:29
FROGGS colomon++ # thanks for testing
colomon FROGGS: trying to use it to install ABC module at the moment. 20:30
FROGGS moritz: can you please upload froggs.de/perl6/rakudo-star-2014.01-RC0.msi with s/-RC0// ? 20:31
Please pre-compile C:\rakudo\.work\1391200250_1\lib/ABC/Header.pm
build stage failed for ABC: Failed building lib\ABC\Actions.pm
20:32 cooper_ joined, arnsholt joined 20:33 treehug88 joined
tadzik my dream: when rakudo says "Please pre-compile", it's followed by "because I tried to do it myself and failed" 20:34
FROGGS tadzik: I want to implement proper cache invalidation some day... 20:35
tadzik: I've made the msi btw 20:36
tadzik FROGGS: that's awesome :)
it's like windows is almost a first-class citizen
FROGGS *g*
raiph perlmonks.org/?node_id=1072014 # chromatic "secretly" "publicly" aplogogizes to educated_ 20:38
PerlJam FROGGS: When you put it on rakudo.org, add a short article that says it's there with a link to it.
colomon FROGGS: ABC installation via panda failed, I suspect because the perl6 command is not in the path
is it suppose to be.
FROGGS PerlJam: I can't upload it, but I can add a link to the post
colomon: it is in my path 20:39
panda does not get the deps right for some reason
colomon give me a moment
20:39 cooper_ is now known as cooper, cooper left, cooper joined
tadzik oh! 20:39
dalek ast: 90ed0c3 | moritz++ | integration/advent2013-04.t:
Add tests for advent 2013 day 04 (incomplete)
20:40
PerlJam FROGGS: let me know the URL of the final MSI and I'll upload it to rakudo.org
FROGGS PerlJam: that is the final IMO 20:41
froggs.de/perl6/rakudo-star-2014.01-RC0.msi with the -RC0 stripped
PerlJam okay.
FROGGS because I can't fix panda and/or ABC in a reasonable timeframe 20:42
dalek kudo/nom: 8ea845b | moritz++ | t/spectest.data:
run new advent tests
PerlJam FROGGS: uploaded. See rakudo.org/downloads/star/ 20:43
FROGGS PerlJam++ # :o)
20:44 psch joined, arnsholt left
psch hi #perl6 20:44
PerlJam hello psch
20:45 arnsholt joined
dalek ast: dacd397 | moritz++ | integration/advent2013-06.t:
Add test for advent 2013 day 06
20:46
kudo/nom: 2f948dd | moritz++ | t/spectest.data:
run another advent test
20:47
20:48 rindolf joined 20:50 arnsholt left, arnsholt joined
dalek ast: a78a516 | (Pepe Schwarz)++ | S32-exceptions/misc.t:
Added test for RT #98854
20:50
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...l?id=98854
colomon FROGGS: I don't see any sign that c:\rakudo\bin has been added to my oath 20:52
path
(obviously I can do it manually….) 20:53
FROGGS ohh, I might have done that in the past
20:54 kbaker left
hoelzro FROGGS: nice, I like it! 20:57
I had just started to look at procops =/
tadzik++ # release
FROGGS hoelzro: cool!
colomon tadzik++
# indeed
FROGGS hoelzro: though the output of "perl6-p -e 'use v5; say `ls`'" and "perl6-m -e 'use v5; say `ls`'" looks a bit different 20:58
maybe perl6-p's openpipe does not spawn a shell?
hoelzro try echo $$
FROGGS an btw, I did no test except v5 stuff so far 20:59
hoelzro ok
psch lines 135ff in S14-roles/parameterized-basic.t look like they kinda-sorta test for RT #84492. can someone explain what's different if there is anything?
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...l?id=84492
hoelzro I'm hoping to finish it up at FOSDEM
FROGGS hoelzro: I get the pid on both backends 21:00
hmmm, then I dunno why `ls` displays in column on parrot but not on moar
hoelzro hmm 21:01
well
ls will columnize in a tty
21:02 arnsholt left
clsn r: my \φ; say "alive" 21:04
camelia rakudo-moar 1a44ca: OUTPUT«===SORRY!===␤Cannot find method 'ast'␤»
..rakudo-parrot 1a44ca, rakudo-jvm 1a44ca: OUTPUT«===SORRY!===␤No such method 'ast' for invocant of type 'NQPMu'␤»
dalek rl6-roast-data: 45afbfa | coke++ | / (3 files):
today (automated commit)
clsn Hm, when I tried earlier it worked for moar and jvm, but failed for parrot.
FROGGS clsn: huh
clsn Maybe I'm wrong; moar just SEGVd when I tried it now. 21:05
[Coke] looks like additionally, moar has a LTA LTA error.
(those 3 errors should be identical)
lue somehow, the +x on perl6-debug-m wasn't carried over when I did 'make install' 21:06
clsn huh, I thought it worked for me for -j and -m before... Guess not. 21:07
21:07 zakharyas left
FROGGS $<term_init> in token declarator is null 21:07
21:10 havenwood joined
FROGGS tries a fix 21:11
21:13 arnsholt joined
FROGGS r: my \φ 21:13
camelia rakudo-moar 1a44ca: OUTPUT«===SORRY!===␤Cannot find method 'ast'␤»
..rakudo-parrot 1a44ca, rakudo-jvm 1a44ca: OUTPUT«===SORRY!===␤No such method 'ast' for invocant of type 'NQPMu'␤»
FROGGS r: my \φ = 7
camelia ( no output )
FROGGS scratches his head
21:13 DarthGandalf left
clsn Oh right, maybe I was trying it with a value... 21:13
FROGGS std: my \φ
camelia std 09dda5b: OUTPUT«===SORRY!===␤Term definition requires an initializer at /tmp/Zn96AjnSG9 line 1 (EOF):␤------> my \φ⏏<EOL>␤Check failed␤FAILED 00:01 121m␤»
clsn Yeah, I guess all looks well... I could have sworn I was getting weird errors from parrot... 21:15
FROGGS still weird that it does not complain about the missing initializer
clsn And I was getting different behavior for my \x vs my \φ, that was the cool thing.... Wonder what I was doing differently. 21:16
lue $ ./bootstrap.pl 21:17
==> Bootstrapping Panda
Cannot call 'postcircumfix:<{ }>'; none of these signatures match:
Anyone else get this?
dalek kudo/nom: d5f7c71 | (Tobias Leich)++ | src/Perl6/Grammar.nqp:
need to panic now!

Otherwise it would continue and complain about missing method NQPMu.ast because of lacking $<term_init>.
21:19
FROGGS clsn++
moritz FROGGS++ 21:20
clsn moritz++
moritz isn't a big fan of continuing to parse after errors
FROGGS well, it *is* nice to get a list of error, like "unknown sub foo used in line 223, 448 and 976" 21:21
because you can fix all in one go
timotimo agreed. as long as there are no BEGIN blocks in between :P
moritz FROGGS: aye. But I can also fix them all in one go by using the search function in my editor 21:22
lue I am I the only one getting that random "cannot call postcircumfix" error with panda?
FROGGS lue: possible, and I don't even use panda O.o 21:23
21:23 DarthGandalf joined
lue r: class Foo { }; say Foo ~~ Any 21:25
camelia rakudo-parrot 1a44ca, rakudo-jvm 1a44ca, rakudo-moar 1a44ca: OUTPUT«True␤»
clsn r: my @all=(1...*); my @ones:=(1,@all) Z- @all; say @ones[^10]
camelia rakudo-parrot 1a44ca, rakudo-jvm 1a44ca, rakudo-moar 1a44ca: OUTPUT«0 -1 -1 -1 -1 -1 -1 -1 -1 -1␤»
clsn r: my @all=(1...*); my @ones=(1,@all) Z- @all; say @ones[^10] 21:26
But it's Bad if you don't use :=. Which indicates that I don't really understand = vs :=.
camelia rakudo-parrot 1a44ca, rakudo-jvm 1a44ca, rakudo-moar 1a44ca: OUTPUT«(timeout)»
timotimo using = with a list on the LHS will eagerly assign 21:27
:= replaces the listy thing with the lazy thing 21:28
clsn Apparently. And I guess := doesn't project eagerness.
timotimo rather than asking the lazy thing to fill up the listy thing
clsn Can only use := in declarations though, right? as in my @foo :=... not with things already declared.
timotimo nope. you can do := as often as you want 21:29
NQP only has :=, no = at all
dalek kudo/nom: aa72afa | moritz++ | CREDITS:
fix typo in credits
21:30
clsn I knew that about NQP. Yes, I just tried out and see what you mean about :=. Once again, I was misled by REPL weirdness, since I do most of my tinkering in the REPL. The REPL won't let you use := most places.
timotimo yeah, the repl is usually trouble :(
clsn Yeah, I'm starting to understand that. I really need to test out anything anomalous non-REPLly more carefully. 21:31
timotimo yes :( 21:34
i've started using a vim buffer and having a :!perl6 % on hand
21:34 raiph left
PerlJam timotimo: do you map that to one of the function keys or type it out by hand each time? 21:34
timotimo neither 21:36
i type :!<Up>
well, actually i usually type :w<CR>:!<Up>
which is pretty bad as far as huffmanization goes :)
grondilu TimToady: trying to do brace-expansion with a simpler grammar. Doesn't quite work yet, but works for a few examples at least. gist.github.com/grondilu/8743717 21:39
21:41 rurban1 joined
PerlJam grondilu: did the "redundancy" in the original grammar inspire you? :) 21:41
grondilu not sure what you mean 21:42
diakopter repetition of the < >
er <[ ]>
grondilu well, there are definitely a few things I don't like in the current entry. It's great that we have something that works, but the logic behind it is hard to grasp, imho. 21:44
21:44 arnsholt left
grondilu I really have the feeling it's possible to write something more elegant. 21:44
(thus my modest attempt) 21:45
21:45 arnsholt joined
diakopter grondilu: well, idiomacy trades off with elegance.. 21:45
21:46 smls left
diakopter sometimes. 21:46
timotimo since i can't decide between organisation and organization, i'll write organißation 21:48
clsn TIL: (just now, really) It's faster to say sub blah($a, $b) { if $a > $b ... else ... } than it is to say multi sub blah($a, $b where $a>b) { ...} and multi sub blah($a, $b) { ... }.
i.e. it executes faster.
timotimo get it? because ß is "sz"
FROGGS Missing test file: t/spec/integration/advent2013-day02.t 21:49
Missing test file: t/spec/integration/advent2013-day04.t
Missing test file: t/spec/integration/advent2013-day06.t
the files don't have a "day" in their names
colomon clsn: yes, definitely
FROGGS moritz: ^^
clsn Makes sense. But it had to be experimented with. 21:50
lue That panda error in more detail ==> gist.github.com/lue/1029c4d379bb8143b8f3
dalek kudo/nom: 257e61a | moritz++ | t/spectest.data:
fix file names, FROGGS++
21:54
moritz I shouldn't hack this late in the week anymore
timotimo ... but it's friday? 21:56
friday night isn't in the week :)
21:56 rindolf left
moritz well, that's the problem, aint it? :-) 21:56
21:56 rindolf joined 21:57 treehug88 left 22:02 benabik left 22:06 smls joined 22:10 berekuk left
smls moritz: re "continuing to parse after errors", why not do it like Perl: Continue parsing, but stop at the first thing that's not safe after an error. 22:13
--> "line 5: Error X; line 17: Error Y; line 33: BEGIN not safe after errors; aborting." 22:14
22:17 BenGoldberg joined
vendethiel is this good ? 22:17
subset FilledArray of Array where *.elems > 0;
tadzik why don't you try it out? :) 22:19
[Coke] can we pick a consistant way to name the advent files? 22:21
vendethiel tadzik: yeah that works but I'm looking for a way to convert an Array to FilledArray
[Coke] (rather than some day, some not?)
vendethiel I tried .FilledArray, FilledArray([1,2])
[Coke] if no objections, I'll unify them later this even. 22:22
smls vendethiel: I don't think subsets are types that you can instantiate, but merely meant as restrictions for parameters/attributes 22:23
vendethiel smls: ah, okay :). thanks !
smls r: subset FilledArray of Array where *.elems > 0; sub foo (FilledArray $x) { say $x }; foo [1, 2]; foo [] # like this
camelia rakudo-parrot 2f948d, rakudo-jvm 2f948d, rakudo-moar 2f948d: OUTPUT«1 2␤Constraint type check failed for parameter '$x'␤ in sub foo at /tmp/tmpfile:1␤ in block at /tmp/tmpfile:1␤␤» 22:24
smls I'm new to Perl 6 as well, though :)
vendethiel smls: thanks 22:25
22:26 arnsholt left 22:29 mcglk joined 22:38 btyler left 22:47 FROGGS[mobile]2 left 22:48 FROGGS[mobile] joined 22:54 arnsholt joined 22:56 kaiz joined 22:57 kaiz left 22:58 grondilu left, dmol left, tgt left
Ulti [Coke] yeah the only part of bioperl that I use is the parsers, the more analytical things I don't trust at all 23:00
lue I really have no clue why panda suddenly breaks on postcircumfix:<{ }> :( 23:01
Ulti which is why Perl6 is neat because a module filled with well tested grammars for different file formats would be really great
23:12 bluescreen10 left 23:14 btyler joined
dalek ast: 1ccbe07 | coke++ | integration/advent2013- (6 files):
follow naming convention of other advents.
23:15
vendethiel is ==>> not implemented yet ?
dalek kudo/nom: dc42ad8 | coke++ | t/spectest.data:
track file rename
[Coke] easy enough to check with the evalbot. 23:16
vendethiel (did not mean "yet", sry)
[Coke] looks up the syntax
vendethiel r: my @a; 1..2 ==> @a; 3..4 ==>> @a;
camelia rakudo-parrot 257e61, rakudo-jvm 257e61, rakudo-moar 257e61: OUTPUT«===SORRY!=== Error while compiling /tmp/tmpfile␤==>> feed operator not yet implemented. Sorry. ␤at /tmp/tmpfile:1␤------> my @a; 1..2 ==> @a; 3..4 ==>> @a⏏;␤ expecti…»
vendethiel oh okay :)
diakopter n: my @a; 1..2 ==> @a; 3..4 ==>> @a;
camelia niecza v24-109-g48a8de3: OUTPUT«Unhandled exception: Feed ops NYI␤ at /home/p6eval/niecza/lib/CORE.setting line 1536 (die @ 5) ␤ at /home/p6eval/niecza/lib/CORE.setting line 3571 (infix:<==>> @ 4) ␤ at /tmp/_dW_Xtvh9G line 1 (mainline @ 4) ␤ at /home/p6eval/niecza/lib/CO…»
[Coke] there ya go. 23:17
vendethiel++
vendethiel [Coke]++
(now I know where au got this habit of ++'ing people :p
23:17 arnsholt left
[Coke] ;) 23:18
23:18 jeffreykegler joined
TimToady r: 1..2 ==> my @a; 3..4 ==> @a.push; say @a 23:18
camelia rakudo-parrot 257e61, rakudo-jvm 257e61, rakudo-moar 257e61: OUTPUT«===SORRY!=== Error while compiling /tmp/tmpfile␤Sorry, do not know how to handle this case of a feed operator yet.␤at /tmp/tmpfile:1␤------> 1..2 ==> my @a; 3..4 ==> @a.push⏏…»
TimToady r: 1..2 ==> my @a; 3..4 ==> push @a; say @a
camelia rakudo-parrot 257e61, rakudo-jvm 257e61, rakudo-moar 257e61: OUTPUT«1 2 3 4␤»
TimToady there ya go
smls www.dlugosz.com/Perl6/web/info-model-TOC.html 23:19
^^ How up-to-date is this?
vendethiel I don't even understand what ==>> does compared to ==> :P
r: 1..2 ==> my @a; 3..4 ==> @a; say @a
TimToady it may be a syntax in search of a feature
[Coke] smls: it says right there :)
camelia rakudo-parrot 257e61, rakudo-jvm 257e61, rakudo-moar 257e61: OUTPUT«1 2 3 4␤»
vendethiel r: 1..2 ==> my @a; 3..4 ==> @a; say @a.perl
camelia rakudo-parrot 257e61, rakudo-jvm 257e61, rakudo-moar 257e61: OUTPUT«Array.new(1, 2, 3, 4)␤» 23:20
smls [Coke]: I mean, does it still apply to today's Rakudo more or less
vendethiel nah, this is not a LoL. So I'm not sure why it's there
TimToady vendethiel: we might decide that implicit "push" semantics are abug
[Coke] smls: yah, I know, I'm just being snarky.
23:20 Pleiades` left
vendethiel TimToady: and make my last case [1, 2, [3, 4]] ? 23:20
[Coke] (and I don't know the answer)
23:21 rurban1 left
vendethiel (==> actually is like F#'s |>, I like it) 23:21
TimToady vendethiel: depends on how we implement it; mind end up just 3,4
*might 23:22
vendethiel oh yeah, just overwrite, makes sense !
TimToady but I shouldn't do design while I'm sick :)
vendethiel while it could just be `1..2 ==> my @a; 3..4 ==> (@a = [])`, couldn't it ?
to "just end up 3,4". (I don't say that's practical, I'm just asking if how I see evaluation order is correct) 23:23
TimToady we haven't really nailed down the most desired semantics of ==> yet, specwise
especially since ==> may end up representing channels
and behave even more like Unix pipes that way
vendethiel TimToady: I see. Does it append to the arglist of the following call, kind of like a macro, or does the right operand's function needs to be curried ?
like 1..2 ==> map({$_ * 2}) 23:24
TimToady right, that's the debate :)
23:24 arnsholt joined
TimToady and to what extent do we recognize different targets at compile time vs run time 23:24
vendethiel not sure I'm getting that last bit
you mean, semantics to decide if it should just call the function or append to arglist ? 23:25
TimToady to what extent to we limit visibility of data structures outside that might need locking or lock-free consideration?
vendethiel oh, okay.
23:25 Pleiades` joined
TimToady the original idea was just appending to an arglist 23:25
the ==>> spec-ese came from groping our way toward inter-thread communication, which jnthn++'s more recent stuff serves as a good basis for 23:26
so we'll need to revisit whether ==> represents channels, and how we target multiple feeds to the same target 23:27
and whether push semantics even make sense in that context
or whether push semantics are in some sense mandatory
also whether we specifically recognize declarations like my @a, or need something around that 23:28
exactly at what point do we figure out the nature of the function or object that we're passing values to... 23:29
and what are the optimization tradeoffs for each approach? 23:30
23:30 arnsholt left, arnsholt_ joined
TimToady I can come up with lots of questions while I'm running a fever, but I'd better not come up with any answers... :) 23:30
vendethiel TimToady: you're still interesting to read tho :) 23:31
TimToady thanks, and welcome to the research project
23:35 ponbiki left
dalek ast: 2162736 | (Pepe Schwarz)++ | S12-methods/delegation.t:
Added tests for RT #95514.
23:40
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...l?id=95514
23:41 xinming left 23:42 ajr_ left, xinming joined, ponbiki joined 23:43 ponbiki is now known as Guest62859, colomon left 23:50 havenwood left 23:54 arnsholt joined, arnsholt_ left 23:55 jnap left
dalek ast: 84c4742 | (Pepe Schwarz)++ | S14-roles/parameterized-type.t:
Added test for RT #84492.
23:58
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...l?id=84492
psch has run out of things he understands how to write tests for 23:59