»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend!
Set by sorear on 4 February 2011.
00:01 tokuhiro_ left 00:15 Krunch left, Krunch joined 00:17 Chillance left 00:23 tokuhiro_ joined 00:28 Krunch left, Krunch joined 00:32 sftp_ left 00:38 raiph left 00:45 dayangkun joined 00:52 fgomez left 00:55 dayangkun_ joined 00:56 dayangkun_ left, dayangkun left, tokuhiro_ left 00:57 am0c joined 01:01 anuby joined 01:07 hypolin joined 01:18 am0c left 01:19 am0c_ joined 01:24 atrodo left 01:27 cognominal left 01:36 shinobicl joined 01:43 bowtie left 01:46 am0c_ is now known as am0c, benabik left 01:58 grondilu left 02:09 benabik joined 02:17 Gothmog_ joined 02:23 am0c_ joined, am0c left 02:24 am0c_ is now known as am0c 02:25 fgomez joined 02:41 sftp_ joined, Zachary joined, Zachary is now known as Guest58684 02:44 anuby left 02:49 anuby joined 02:52 sftp_ left
Guest58684 should <alpha> match greek letters? 02:55
sorear it should, yes 02:59
Guest58684 does it depend on ICU? 03:00
sorear rn: say ?( "α" ~~ /<alpha>/ )
yes, you need ICU
p6eval rakudo c82d10, niecza v22-32-gee5dcf1: OUTPUT«True␤»
Guest58684 ok.
sorear looks like there is agreement
03:04 anuby left 03:05 sftp_ joined 03:06 anuby joined 03:07 gootle joined 03:11 jaldhar joined 03:23 jaldhar left 03:25 fgomez left 03:26 fgomez joined 03:27 jaldhar joined 03:36 BillySeth is now known as BillySeth|bbl 03:54 orafu left, orafu joined 04:09 Guest58684 left 04:21 sftp_ left 04:22 sftp_ joined 04:29 BillySeth|bbl is now known as BillySeth
shinobicl hi 04:34
what do you usually use to debug your perl6 programs?
04:46 Guest58684 joined 04:50 skids left
shinobicl how do i exit from a "for"? 04:55
r: for (1..10) -> $val{ if ($val == 3) { say "(found 3!.... exiting.)"; } else { say "-"} } 04:57
p6eval rakudo c82d10: OUTPUT«-␤-␤(found 3!.... exiting.)␤-␤-␤-␤-␤-␤-␤-␤»
shinobicl r: for (1..10) -> $val{ if ($val == 3) { say "(found 3!.... exiting.)"; break; } else { say "-"} }
p6eval rakudo c82d10: OUTPUT«===SORRY!===␤CHECK FAILED:␤Undefined routine '&break' called (line 1)␤»
doy r: for (1..10) -> $val{ if ($val == 3) { say "(found 3!.... exiting.)"; last; } else { say "-"} }
p6eval rakudo c82d10: OUTPUT«-␤-␤(found 3!.... exiting.)␤» 04:58
shinobicl "last" it was.. thanks!
doy and there's a debugger here github.com/jnthn/rakudo-debugger 04:59
sorear shinobicl: I mostly use print statements because I'm not cool enough for the debugger 05:02
05:02 mtymula left 05:04 xinming_ left 05:05 xinming joined
shinobicl r: my $x = -1; try { warn "Bad data" if $x < 0; CONTROL { say "x is now 0"; $x = 0 } }; say $x; 05:14
p6eval rakudo c82d10: OUTPUT«x is now 0␤Bad data in block at /tmp/vjUP4tk9x3:1␤␤0␤»
shinobicl i have a very similar structure in a module i wrote... but the "CONTROL" part is never reached... 05:15
the difference is that the "warn" is inside a "for" loop
r: my @data = <A B E D>; try { for (@data) -> $val { if ($val ne any(<A B C D>)) { warn "wrong char. assuming defaults"; last; }} CONTROL{ @data = <A B C D>; }};say @data.perl; 05:22
p6eval rakudo c82d10: OUTPUT«===SORRY!===␤Unable to parse expression in block; couldn't find final '}' at line 2, near "for (@data"␤»
05:22 jeffreykegler joined 05:23 telex left
shinobicl r: my @data = <A B E D>;try { for @data -> $val { if ($val ne any(<A B C D>)) { warn "wrong char. assuming defaults"; last; }; }; CONTROL{ @data = <A B C D>};}; say @data.perl; 05:29
p6eval rakudo c82d10: OUTPUT«wrong char. assuming defaults in block at /tmp/WciqV8ylOR:1␤␤Array.new("A", "B", "C", "D")␤»
shinobicl this same logic does not work inside a module :S 05:30
anyway, it's late.. good night everyone!
05:30 shinobicl left 05:36 telex joined 05:49 am0c left 05:50 jeffreykegler left 06:04 xdbr left 06:07 sorenso joined 06:09 Guest58684 left 06:10 kaleem joined 06:22 birdwindupbird joined 06:29 GlitchMr joined 06:31 kurahaupo left, BillySeth is now known as BillySleep 06:39 sorenso left 06:59 SHODAN left 07:00 SHODAN joined 07:02 sftp_ left, sftp_ joined 07:09 Guest58684 joined 07:15 _jaldhar joined, jaldhar left 07:18 GlitchMr left 07:25 am0c joined 07:27 flightrecorder joined 07:33 FROGGS joined 07:43 domidumont joined 07:44 SamuraiJack joined 07:46 Guest58684 left 08:09 brrt joined 08:12 domidumont left 08:13 awwaiid left, domidumont joined 08:16 Guest58684 joined 08:17 circlepuller left 08:22 fhelmberger joined, am0c left 08:23 circlepuller joined, circlepuller left, jlaire left, circlepuller joined, circlepuller left 08:24 circlepuller joined 08:29 jlaire joined 08:30 sorenso joined 08:32 brrt left, brrt joined 08:35 Su-Shee joined
Su-Shee good morning everyone.. 08:35
moritz \o
Su-Shee do I want star or the monthly release?
moritz star 08:37
that's monthly release + debugger + modules
Su-Shee yeah I know, I was asking more along the lines of "monthly is a little experimental these days because we added..." :)
ok, I'm just greedy for the bleeding edge. :) 08:38
nwc10 surely, if you want the bleeding edge you go for git://github.com/rakudo/rakudo.git ? 08:40
08:41 [hds] joined 08:42 [hds] left 08:43 [hds] joined
sorear good morning Su-Shee 08:45
moritz: is the categorical rewrite in a star yet? 08:46
good morning nwc10 too :D
good to see you about
hopefully soon niecza will have a bleeding edge again 08:47
moritz Su-Shee: star is no less experimental than the compiler release :-)
sorear: no, categorical rewrite is only in git right now 08:48
sorear what's the opposite of experimental? dogmatic?
theoretical?
nwc10 "special biologist word" ? 08:49
Su-Shee orthodox :)
moritz conservative
brrt what moritz says
sorear
.oO( Perl6 is experimental. It stopped being theoretical *years* ago. )
sorear is thinking physics
brrt applied! 08:50
08:52 [hds] left 08:53 wamba joined 09:00 kresike joined
kresike hello all you happy perl6 people 09:00
brrt hi kresike
kresike brrt, o/ 09:01
jnthn mornin' o/ 09:11
nwc10 jnthn: in multidispatch.c, I can't see anything that ever frees the memory allocated for Rakudo_md_cache *cache in Rakudo_md_candidate_info 09:16
jnthn The cache isn't referenced from Rakudo_md_candidate_info, no? 09:19
jnthn looks 09:20
No, it hangs off Rakudo_Code's dispatcher_cache slot, via a level of indirection. 09:21
nwc10 it returns cache->candidates, which (as best I can work out) is owned by that structure
but it will leak if the owner PMC is ever freed?
jnthn Yes, looks like. 09:23
sorear heya, jnthn
jnthn OTOH, most multis live "forever".
o/ sorear
nwc10 the code for sorting is only tightly coupled with the code for adding to the cache via the structure Rakudo_md_candidate_info 09:24
Rakudo_md_candidate_info.constraints appears not be used once sorting is done. 09:25
jnthn Correct. 09:26
nwc10 so, potentially, sort code that returned a massive flat list of PMCs could be called, and the caller process that flat list into the existing structure.
jnthn Once, a while ago, the multi-dispatcher did use that list of candidates at dispatch time. 09:27
Now just bind_check matters 09:28
One important detail on data structure is that the list of candidates is terminated by two NULLs, with an inner NULL delimiting a group of candidates that are equally narrow.
nwc10 I see accesses to all structure members except constraints in the rest of the code
jnthn Yeah. It could be moved into Rakudo_md_candidate_graph_node if it's only used in the sort. 09:29
nwc10 aha. good plan. I'd not thought of doing it like that. 09:30
09:35 am0c joined
FROGGS for the rakudo release: should I update it to use parrot 4.9.0 or should I stay at 4.4.0? 09:46
(I guess I know the answer already) 09:47
bonsaikitten FROGGS: both should work, I prefer new things as they have usually more bugfixes than regressions 09:48
FROGGS hmmm, I remember this IO-rewrite they did in parrot that made some trouble. I thought that we use 4.4.0 because that is right before these changes... 09:49
bonsaikitten yes, that sounds familiar 09:50
FROGGS and since I dont know how to test if these problems still exist... 09:51
brrt iirc, they had to do with readline not being nice
if we don't have tests for that it is un-awesomeness from all sides
09:53 dakkar joined
jnthn Yes, it was IO-rewrite related 09:58
nwc10 you'd be best asing Pm, I think. But I think it's issues github.com/parrot/parrot/issues/860 (now closed) and github.com/parrot/parrot/issues/847 (still open)
jnthn Ah 09:59
If there's tickets still open then that answers the question, I think.
09:59 domidumont left
FROGGS ya, and I have a testcase now to prove ;o) 10:00
thank you all ;o)
nwc10 jnthn: but from skimming #847 it's not clear to me whether it's closeable yet
jnthn nwc10: If it's not closed I'd rather be cautious and go with depending on a known good release. 10:01
nwc10 agree. I more meant - "it's not clear to me what the parrot folks still think that they need to do" 10:02
jnthn Ah, OK.
We can always bump to latest after this month's Rakudo release so those using bleeding edge Rakudo from Git will see the issues and we'll have time to know about them (or lack of them) well before a release.
FROGGS k 10:03
nwc10 that seems wise
10:06 hypolin left 10:07 anuby left
FROGGS jnthn: is that multi trait_mod:<is> merge problem solved yet? (using NativeCall and Inline::C, and just getting trait_mod:<is> from the last-imported module) 10:11
jnthn FROGGS: Not yet
FROGGS can I inspire you (via paypal) to have a look at it? ;o) 10:12
jnthn I'll get there, it's just that I want to understand another pre-comp/import issue first, since the solution for that may impact the solution for this... 10:14
FROGGS ahh, okay
jnthn Looking at that is already about top of my todo list. 10:15
FROGGS the offer still holds 10:16
10:16 domidumont joined
dalek rlito: e107c18 | (Flavio S. Glock)++ | / (5 files):
Perlito5 - parser - add "glob variable"
 10:16
10:16 domidumont left
jnthn FROGGS: The problem isn't a financial one, more that I'm currently in preparation for 2 solid weeks of travel/teaching/conference speaking, which is a small distraction from hacking on Rakudo. :) 10:17
FROGGS I dont mean that I want it done tomorrow ;o) 10:18
jnthn :-)
FROGGS its just that I got some spare time in december, and my modules (Inline::C, SDL, Games::BubbleBreaker) depend on that 10:21
10:21 kresike left
sorear sleep& 10:21
FROGGS gnight 10:22
10:23 domidumont joined
nwc10 wonders if the regulartory hurdles would be any less complex for online payment in BEER rather than cash. Different, yes. But "beer laundering"? 10:27
jnthn :D 10:29
What a concept...
nwc10 with rakudo, is there a way with Configure.pl to cause --gen-parrot *not* to build parrot with --optimize? I want -g for all the C files, to get debugging 10:35
jnthn --parrot-options lets you pass options through to Parrot's Configure 10:36
10:36 kresike joined
nwc10 Unknown option: parrot-options 10:37
whose Configure.pl takes that?
arnsholt NQP, I assume 10:38
nwc10 nope. not their either
gah. there.
where is my brain?
tadzik sorry, I took it yesterday, forgot to tell you :/ 10:39
FROGGS .oO( brainzzzz, eating brainzzzzz )
masak g'day, #perl6 10:49
jnthn good aftOH^Wmorning, masak 10:50
masak made it just before the meridian. :P
10:51 wk joined, wk is now known as Guest63538
FROGGS hi masak 10:55
11:01 gootle left
masak aloha. 11:02
11:05 Su-Shee_ joined 11:07 whiteknight left 11:08 Su-Shee left 11:10 am0c left
Woodi hallo today 11:22
masak good'di Woo'di. 11:23
Woodi I tried to make some bindings to C func like set_opt( void *v ) using CArray[int] but it errors with: Unknown type* Positional* used in native call 11:24
jnthn It'd be easiest if you could paste the signature 11:25
But my first guess is that you wrote CArray[int] @foo, not CArray[int] $foo
Woodi yyy, yes... 11:26
jnthn :)
Woodi I thinked I followed tests...
jnthn It's consistent with rest of Perl 6, fwiw
sub foo(Array @x) { } means "a Positional that holds Arrays"
Woodi checking 11:27
passed to next line ! :) however next line is error check... will see... 11:29
nwc10 jnthn: parrot-option (ie singular) 11:35
use the source, luke
jnthn nwc10: oh! Sorry.
nwc10 still not sure if it's actually adding -g, as I see lots of -O2 flying past 11:36
jnthn nwc10: (use the source) of perl Configure.pl -h :)
nwc10 CONFIG_ARGS = "--optimize=-g" "--optimize" "--prefix=/Volumes/Stuff/Perl/rakudo/install" 11:37
mmm. :-/
jnthn Did you try --parrot-option='--debugging' ?
er, or something like that :)
nwc10 nope, not yet
Woodi decisively things move forward :) thanx jnthn++ 11:41
jnthn np :) 11:42
11:46 fgomez left 11:58 Guest58684 left 12:01 Gothmog_ left 12:03 fgomez joined 12:09 awwaiid joined 12:11 clkao_ joined 12:13 Su-Shee_ is now known as Su-Shee 12:14 tokuhiro_ joined 12:21 domidumont left 12:24 am0c joined
moritz [Coke]++ #going through the perl6 RT queue 12:33
12:33 domidumont joined 12:34 Guest58684 joined 12:35 Psyche^ joined 12:39 Patterner left, Psyche^ is now known as Patterner 12:49 sisar joined
bbkr_ r: =pod 13:02
p6eval rakudo c82d10: OUTPUT«===SORRY!===␤Preceding context expects a term, but found infix = instead␤at /tmp/eqe30h6iti:1␤»
bbkr_ std: =pod
p6eval std 04216b1: OUTPUT«ok 00:00 40m␤»
bbkr_ which one is incorrect? should bare =pod (without closing =end tag) be allowed? 13:03
masak I don't think STD does any Pod parsing. 13:05
bbkr_ then it's LTA rakudo message, right? 13:06
13:07 spider-mario joined
masak possibly. 13:07
bbkr_ reports 13:08
masak I'd like to hear what tadzik has to say about it.
13:09 wamba left
masak the above should count as "abbreviated block" syntax, I think. as far as I see, it shouldn't need an ending delimiter, because abbreviated blocks auto-close. 13:09
bbkr_ so it should just work without failure? 13:11
13:12 pmurias joined
masak as far as I can see, yes. 13:14
13:15 domidumont left 13:17 bowtie joined 13:33 Gothmog_ joined
[Coke] yawns from the office. 13:35
13:40 isBEKaml joined
isBEKaml hola, #perl6! 13:40
gist.github.com/4078709 # any way I can improve this? 13:42
13:42 am0c left
masak indent looks weird on line 8. 13:43
isBEKaml yeah, hit tab accidentally. :) 13:44
not very comfortable with form entries that take tabs.
masak: I don't like the $^a in map:, anyway I can just say that once? 13:45
masak .comb(/./) better written .comb
isBEKaml: I was thinking of that. :)
@nums>>.chr.map( * xx 2 )
but it felt like a micro-optimization, so I didn't mention it.
isBEKaml r: my @nums = 1,2,3; @nums>>.chr.map(* xx 2); 13:46
p6eval rakudo c82d10: ( no output )
masak indent problems at line 19, too.
isBEKaml masak: now I know that it's my editor. notepad++ (not cool, np, not cool)
jnthn notepad++ doesn't screw up my indenting :P 13:47
masak $dict_size is not really necessary, since it's a loop invariant that it tracks +%dictionary
isBEKaml jnthn: smart indents seems to do it. I think it got confused between 8 and 4 char tabs. 13:48
jnthn Anything with smart in the name usually isn't... :)
masak any technology... what jnthn said.
isBEKaml jnthn: so I'll warn others about smartcards. :P 13:49
13:49 domidumont joined
masak we don't call them that here. :) 13:49
13:49 sisar left
masak we just call them "cards". 13:50
isBEKaml masak: hey, they carry that big label "SMART CARDS"!
calling them "cards" would confuse people. :) 13:51
masak ours don't.
guess we're too smart to label them that :P
13:52 wamba joined
jnthn Yeah. We'd never give them some funny name like yoyo card... :) 13:52
masak jnthn: that was an eerily prescient move, IMO. sometimes the system is up, sometimes down... 13:53
jnthn masak: That was *exactly* my first thought when I saw the name :D
masak (it's our local train and bus traffic here in southern Sweden)
www.skanetrafiken.se/templates/Info...x?id=26853
isBEKaml yoyo card for bus/train ticketing system in .se? 13:54
masak aye.
jnthn To be fair...the machines that update the cards, sell tickets etc actually *are* much more reliable, in my experience anyway, than the trains themselves :)
masak no doubt.
jnthn Which is quite amazing given they run Windows CE :P
masak funnest thing I've managed to do so far with the machines was to try to refill a limited-time summer offer card. 13:55
produced a number of "null" values in various fields in the machine.
jnthn :D
masak but it did survive.
13:55 [particle]1 left
moritz though when I travelled to Perl and missed the first connection, the vending machine told me "no connections found" 13:55
13:55 benabik left
jnthn d'oh 13:56
That's one of those times when you really hope it's wrong.
masak that's just a statement about the vending machine... :)
moritz so I went into the customer service office, and their computer found one
isBEKaml
.oO(moritz usually finds a connection. He worked on DBIsh, no? )
 13:57
masak :P
jnthn moritz: Still no home internets?
moritz jnthn: no
jnthn awww 13:58
masak hugs moritz
moritz it seems that the new internet connection uses VDSL, and our modem only supports ADSL.*
so, new router ordered
masak isBEKaml: nice to have you back. I hear there are 80-ish testneeded RT tickets... :)
moritz with a bit of luck it'll arrive on Saturday; with bad luck it'll be Monday
jnthn crosses fingers 13:59
13:59 tokuhiro_ left
FROGGS moritz: sad that you dont live neer to me, I have some modems lying around... 13:59
isBEKaml masak: testneeded? I'll look in.. :)
moritz FROGGS: and can they speak VDSL? 14:00
masak there's a handy testneeded url somewhere, but I don't know where... 14:01
moritz FROGGS: most can't, and those that do are expensive-ish
rakudo.de/testneeded iirc
jnthn rt.perl.org/rt3//Public/Search/Sim...needed%%27
er, moritz wins :)
isBEKaml jnthn: do you usually have this long url somewhere?
jnthn isBEKaml: Open in a browser tab that I've had open for like 5 months or something :)
isBEKaml jnthn: I just found out my muscle memory typed in rakudo.do than rakudo.de. :P 14:02
FROGGS moritz: I've got a dlink modem that does adsl2+ (up to 24mbit/s), it looks like VDSL2 is the same standard 14:03
masak isBEKaml: [Coke]++ has done a nice job closing some of these in the past few days. it's nice to see that queue shrink.
FROGGS moritz: so even if you ordered >24mbit, it should work at 24mbit/s
isBEKaml masak: some of them are NPMCA bugs, can we close them? 14:04
14:04 orafu left
masak don't think so. 14:05
jnthn NPMCA ones should even get a test to check they do what they're supposed to do or that they give a good error, possibly adding a typed exception along the way.
14:05 PacoAir joined
moritz FROGGS: oh, that's interesting 14:06
FROGGS: then tonight I'll call the telekom technicians and complain again :-)
jnthn takes another look at RT#112626
masak rt.perl.org/rt3/Ticket/Display.html?id=112626 14:08
moritz that's the one that blocks colomon++'s ABC module from running on rakudo, iirc
jnthn yeah
masak funny to see ufo in a bug report ;)
isBEKaml masak: the lzw entry with @nums>>.chr.map: { * xx 2 } led to too many Whatever<foo> entries. 14:09
Woodi ok, I have few (native)bindings and want to move them to module.pm. so module namename { ... } and then *always* namename::sub_name( ... ) ? or with class make methods to every native sub ?
isBEKaml masak: so I just reduced it to @nums>>.chr.map: { $^a xx 2 }
14:09 atrodo joined
masak isBEKaml: that's because you did both {} and * 14:09
isBEKaml: I did () and *
isBEKaml: * implies {} already. 14:10
jnthn Woodi: Or mark them "is export", so a "use" statement will import them.
Woodi ok. thanx again :)
isBEKaml masak: so * usable only in a blockoid. Whatever blockoid. I see...
14:12 kaleem_ joined
Woodi btw. there is some thing that could be put into FAQ. technical like: Data::Dumper answer and .base(x) and few more 14:12
masak isBEKaml: any expression containing * as a term turns into a closure, auto-creating {} around it, and turning the *s into $^a, $^b, etc. (conceptually) 14:13
Woodi: what FAQ about .base(x)? I haven't heard that one.
14:13 drbean left
isBEKaml the nascent faq.perl6.org, probably. 14:14
masak yes, but what is the question people frequently ask about .base(x)?
14:14 kaleem left
Woodi masak: I am sure you answered it few times probably :) 14:14
masak maybe so, but what is it? 14:15
isBEKaml Woodi: do you mean that we can do :8(x) :10(x) :16(x) and others?
14:15 drbean joined
masak rn: say Inf.base(16) 14:15
p6eval niecza v22-32-gee5dcf1: OUTPUT«Unhandled exception: Unable to resolve method base in type Num␤ at /tmp/ogVolikEst line 1 (mainline @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4211 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4212 (module-CORE @ 578) ␤ at /home/p…
..rakudo c82d10: OUTPUT«Type check failed in assignment to '$int_part'; expected 'Int' but got 'Failure'␤ in method base at src/gen/CORE.setting:2920␤ in block at /tmp/teecyNeIC5:1␤␤»
masak rn: say 16.base(Inf)
p6eval niecza v22-32-gee5dcf1: OUTPUT«Unhandled exception: base must be between 2 and 36, got Inf␤ at /home/p6eval/niecza/lib/CORE.setting line 1443 (die @ 5) ␤ at /home/p6eval/niecza/lib/CORE.setting line 842 (Int.base @ 6) ␤ at /tmp/_sApkhBW_U line 1 (mainline @ 3) ␤ at /home/p6eval/niec…
..rakudo c82d10: OUTPUT«base must be between 2 and 36, got Inf␤ in method gist at src/gen/CORE.setting:10149␤ in sub say at src/gen/CORE.setting:7460␤ in block at /tmp/CbpqvO0iRC:1␤␤»
Woodi isBEKaml: hmm, I was looking and found only faq with political q&a's 14:16
masak Woodi: what do people often ask about .base(x)?
I'm curious.
jnthn
.oO( who are it belong to? )
isBEKaml Woodi: political questions? I don't understand, I only found language and general questions there.
Woodi sec pls :) 14:17
colomon jnthn++ # RT#112626
masak r: my $pls = 1; say sec($pls)
p6eval rakudo c82d10: OUTPUT«1.85081571768093␤»
masak r: my $pls = 1; say sec $pls # :)
p6eval rakudo c82d10: OUTPUT«1.85081571768093␤»
isBEKaml groans
Woodi first result on google :) www.perl6.org/archive/faq.html
not quite political but not technical either 14:18
14:18 xdbr joined
masak Woodi: my question wasn't "where is the FAQ?". my question was "what is it that people are asking about .base(x)?" 14:18
Woodi sec pls
isBEKaml masak: no, no no.... :)
14:18 skids joined
masak r: my $pls = 1; say tan $pls # :) 14:19
p6eval rakudo c82d10: OUTPUT«1.5574077246549␤»
Woodi masak: from 2009 probably: 20:59 < LlamaRider> I couldn't find info on switching number systems, hex/oct/binary in Perl6 and what does mean for the types
and it is frequently asked question on that channel :)
masak if you say so.
isBEKaml do people read logs from the waybackmachine to judge language usability? 14:20
masak so the question wasn't about .base(x) (which I think is fairly straightforward, myself), the question was about how to represent numbers in binary, octal, and hexadecimal in Perl 6 source?
Woodi masak: 20:59 < moritz> nom: say 1234.base(16) :) 14:21
and another "is there Data::Dumper in Perl6 available maybe ?" :) 14:22
masak: 21:00 < LlamaRider> seriously now...
isBEKaml .oO (is Woodi now a collage artist? )
14:23 GlitchMr joined
Woodi isBEKaml: hmm, I collect jokes if it qualifies as art... 14:23
GlitchMr return 1 if $n < 2; 14:24
fact($n - 1) * $n;
oops
Accidental paste of wrong thing
"is there Data::Dumper in Perl6 available maybe ?"
Isn't it called .perl?
PerlJam or .pretty 14:25
(probably still conjectural, but I like it :)
Woodi yes GlitchMr :)
14:26 [particle] joined
masak PerlJam: I thought .pretty ended up being spelled .gist 14:27
14:27 spider-mario left
PerlJam oh, maybe so. 14:28
They occupy different areas of my brain if so.
isBEKaml Woodi: I was referring to how you paste content from backlogs to add to conversation. Sticky paste things. :) 14:33
masak: gist updated.
GlitchMr perl6: gist [1, 2, 3] 14:34
p6eval rakudo c82d10, niecza v22-32-gee5dcf1: ( no output )
GlitchMr perl6: say gist [1, 2, 3]
p6eval rakudo c82d10, niecza v22-32-gee5dcf1: OUTPUT«1 2 3␤»
14:34 nyuszika7h left
GlitchMr This is the reason I don't see reason for .gist 14:36
isBEKaml r: [1,2,3].pretty.say;
p6eval rakudo c82d10: OUTPUT«No such method 'pretty' for invocant of type 'Array'␤ in block at /tmp/7Olv2ElDS2:1␤␤» 14:37
GlitchMr Empty array looks identically to empty string
isBEKaml r: [1,2,3].perl.say;
p6eval rakudo c82d10: OUTPUT«[1, 2, 3]␤»
GlitchMr I really don't know what is .gist useful for
For user, the is a .Str method. For debugging, there is a .perl method.
dalek rlito: daea592 | (Flavio S. Glock)++ | / (13 files):
Perlito5 - add more "use strict"
 14:38
bbkr_ gist word is some kind of acronym?
moritz GlitchMr: if you have a large match object, you'll probably see that .Str isn't very informative, and .perl gives way too much information to be usful (most of the time)
bbkr_: no, a gist is some kind of summary
masak isBEKaml: (0 .. ^$dict_size) better written (^$dict_size) 14:39
Woodi NativeCall could have some sugar for C '&' oprator. Now this need to be used for passing **ptr : my OpaquePointer $ptr; my $array := CArray[OpaquePointer].new(); $array[0] = $ptr; ... here pass array to function; $ptr = $array[0]; 14:40
masak isBEKaml: and I still think you don't even need to maintain that variable, as all it does is track the number of entries in %dictionary, which you can easily get through +%dictionary.
Woodi naybe assigment to $array can be avoided...
GlitchMr Ok, regular expression case makes sense
masak r: %d = <foo 1 bar 2 baz 3>; say +%d; %d<blah> = 4; say +%d
p6eval rakudo c82d10: OUTPUT«===SORRY!===␤Variable %d is not declared␤at /tmp/bZHQaqquVE:1␤»
masak r: my %d = <foo 1 bar 2 baz 3>; say +%d; %d<blah> = 4; say +%d 14:41
p6eval rakudo c82d10: OUTPUT«3␤4␤»
masak isBEKaml: so ditch $dict_size, and just do +%dictionary on line 16.
isBEKaml masak: hmm, I knew I didn't have to do 0..^dict_size. :)
14:42 nyuszika7h joined, nyuszika7h left
isBEKaml masak: does, %dictionary.elems do the same thing as +%dictionary? 14:43
moritz aye
14:43 nyuszika7h joined
isBEKaml moritz: danke 14:44
[Coke] I am pretty sure that .perl is for computers, .gist is for people, and ~ is the default. (not sure which market the default is for.) 14:47
masak stringy people.
isBEKaml or stingy ones.
Woodi isBEKaml: pastebin.com/JDe4icdc :) 14:48
dalek rlito: dcde2e3 | (Flavio S. Glock)++ | / (4 files):
Perlito5 - add more "use strict"
 14:50
rlito: 1395266 | (Flavio S. Glock)++ | TODO-perlito5:
Perlito5 - TODO update
 14:53
[Coke] Is there a reason aside from inertia for rakudo to stick with RT? 14:54
(note: this is a HUGE reason that should absolutely be respected.)
isBEKaml [Coke]: I know I asked this question too. The answer(back then) was: I know it sucks, but I can live with it.
PerlJam [Coke]: what would you propose as an alternative?
moritz [Coke]: apart from inertia, just the desire to open tickets by email 14:55
isBEKaml why not github?
moritz [Coke]: which other bug trackers can do, but github can't, and parrot's what-was-it-name installation couldn't do it well either
isBEKaml trac
moritz right
masak an intricate autopun: twitter.com/jacksonh/status/269046172239417345 14:56
14:57 perigrin left
masak [Coke]: I know others hate RT search, but I actually love it. I'd call it codependency if it weren't that I actually reliably find stuff with it. and as far as I know, I can explain how to others. 14:57
[Coke]: so for me (a frequent user of RT), the reason to stick with it is "it works". 14:58
moritz masak: is it really an autopun? it's not obvious to me
arnsholt Your relationship with RT sounds like me and TeX/LaTeX =)
masak moritz: yes. it combines use and mention.
14:59 kaare_ joined
moritz masak: what does it use? 14:59
(sorry, I know it sucks to explain jokes)
masak no prob. let's test this.
moritz: mention: ".mobi". use: "people who use this TLD on mobile platforms -- the target group of the .mobi TLD -- will actually have a harder time[, sort of defying the purpose]" 15:00
PerlJam isn't that more of a meta-use?
moritz masak: I don't see how it's a use
[Coke] The search is an issue. the fact that you can comment on tickets and no email is generated is painful. (we're not using RT as an internal/external tool like some do). I often have issues where I can only have one search going at a time. (testneeded versus all of perl6).
moritz masak: it would be a use if he mis-typed .mobi, or something 15:01
masak moritz, PerlJam: see your point.
15:01 Guest63538 left
masak "I meant to type .mobi, but it was too long and my mobile devide didn't accept a four-letter TLD..." 15:01
[Coke] I'm not pushing for a switch, I'm just wondering if there are reasons to stay other than that we've always been at war with Eurasia. 15:02
masak s/see your point/I see your point/
[Coke] but "it works and we use it" seems to be the prevailing thought. No worries.
PerlJam [Coke]: This war is the least-suffering war :) 15:03
masak [Coke]: I use a lot of other issue trackers besides RT. none of them give me the warm fuzzies like RT does. I think the warm fuzzies come from email reporting and flexible search.
PerlJam masak: either you've found the zen of RT or you have more RT-fu than most. I always find RT awkward. 15:04
[Coke] I would argue with your use of the word flexible, but understand your meaning. 15:05
masak note that I didn't say "obvious" or "simple". :)
PerlJam: fwiw, how to find a Perl 6 ticket: (1) search for the most relevant/precise term that's bound to be in the ticket title, (2) switch to "Advanced search", and add "queue: perl6" to the search, (3) profit! 15:07
it also helps if (0) masak submitted it, because then the ticket will actually have whatever term in the subject line. :) 15:08
Woodi arnsholt: maybe you have idea how to push NULL in place of CArray[type] ? some sugar for this or just CArray[type].new() ?
isBEKaml masak: RT feels like a duck barking at me. :P
PerlJam isBEKaml: hahaha
masak :D
PerlJam masak: ah ... (0) is, I think, the important bit.
jnthn Woodi: Just pass some type object, I think
Woodi: As in, pass CArray[type] itself as the argument 15:09
Woodi k
jnthn Type objects should become nulls.
15:11 stopbit joined
isBEKaml r: my @foo = 101..107; my %props = @foo>>.chr.map(* xx 2); %props.say; 15:17
p6eval rakudo c82d10: OUTPUT«No such method 'count' for invocant of type 'List'␤ in method reify at src/gen/CORE.setting:5467␤ in method reify at src/gen/CORE.setting:5390␤ in method reify at src/gen/CORE.setting:5390␤ in method gimme at src/gen/CORE.setting:5768␤ in method eager at src/g…
15:20 kaleem_ left
isBEKaml r: my @foo = 101..107; my %props = @foo>>.chr.map: { $^a xx 2}; %props.say; 15:21
p6eval rakudo c82d10: OUTPUT«("e" => "e", "f" => "f", "g" => "g", "h" => "h", "i" => "i", "j" => "j", "k" => "k").hash␤»
isBEKaml Was the first approach wrong in any way?
jnthn Looks like xx doesn't auto-curry 15:23
moritz no, because xx * means the same as xx Inf
jnthn ah, that's a good reason 15:24
also xx is a bit special 'cus it thunks its LHS
isBEKaml then, surely the error message is misleading. 15:25
jnthn Maybe map should only take a Callable. 15:26
isBEKaml maybe something like "no term found to repeat" is more useful given the special nature of xx.
15:28 domidumont left
jnthn The problem wasn't with xx 15:28
The xx did its thing fine, then map got a List.
isBEKaml do we still distinguish List and Array? I thought we gave that up... 15:29
jnthn No, we never gave that up. Seq is what went away. 15:30
dalek p: b571803 | jnthn++ | src/6model/serialization.c:
Detect object conflicts during deserialization.

If two pre-compiled modules both tried to replace an object or STable with their own updated view of it, the latest one would always just win. This creates the "disappearing symbol" issue. This patch detects that it's going on and complains, rather than leaving a weird failure later on (the actual fix will be to replace the exception thrown with a hook to allow the module loader to do some smarter merging).
Breaks no NQP tests or Rakudo spectests; does correctly pick up the RT#112626 case.		 15:35
15:36 shinobicl joined
isBEKaml jnthn: I vaguely remember Pm's docs talking about losing the differences between List and Array. I see that things have changed. :) 15:36
isBEKaml is happy reading rakudo source after a long gap. 15:37
moritz isBEKaml: an Array ist just a LIst where every element is a Scalar container 15:38
(unless you bind to an array element, in which case it might not be a container anymore)
isBEKaml moritz: Is that the mutable/immutable difference ?
moritz: List -> immutable, Array -> mutable. 15:39
moritz isBEKaml: well, Array elements are always replacable. List elements only if they happen to be containers 15:40
isBEKaml moritz: I see, thank you. 15:43
PerlJam reminds everyone of the advent calendar again. 15:44
github.com/perl6/mu/tree/master/mi...dvent-2012
GlitchMr Now, where is dalek? 15:45
dalek : 38786a1 | (Конрад Боровски)++ | misc/perl6advent-2012/topic-brainstorming:
Change "Rosettacode" to "Rosetta Code"
 15:46
GlitchMr oh, ok
jnthn eek, I'm the only person who signed up for any slots! 15:47
moritz jnthn: don't worry, I will too
jnthn phew :)
15:47 Su-Shee left
jnthn But I hope there'll be more than two of us. :) 15:47
PerlJam jnthn: I'll probably do a couple too, I'm just more of a last-minute kind of guy :)
jnthn OK :)
GlitchMr Perhaps I could actually write an article about sub MAIN
jnthn +! 15:48
er, +1
moritz I think we had that in the previous years
GlitchMr We did? 15:49
I will look, ok
moritz perl6advent.wordpress.com/2010/12/0...main-subs/
GlitchMr Well, too bad
PerlJam Surely there's something new to cover in that arena? 15:50
15:50 Guest58684 left
GlitchMr I actually think that sub MAIN is one of best Perl 6 features 15:50
It allows you to make scripts quickly.
15:50 domidumont joined 15:51 domidumont left
PerlJam could talk about sub MAIN ( :foo($bar) ) { ... } 15:53
masak sure. 15:54
15:57 domidumont joined
dalek : c1f455c | (Carl Mäsak)++ | misc/perl6advent-2012/schedule:
[perl6advent-2012/schedule] claim four spots
 15:58
jnthn \o/ 16:02
r: say 25 - 7
p6eval rakudo c82d10: OUTPUT«18␤»
GlitchMr glitchmr@feather ~> perl6 -e 'sub MAIN (*@blah) {say "ok"}' a 16:04
ok
glitchmr@feather ~> perl6 -e 'sub MAIN (*@blah) {say "ok"}' --
Usage:
-e '...' [<blah> ...]
Is it a bug?
-- should work as end of argument list, shouldn't it? 16:05
glitchmr@feather ~> perl6 -e 'sub MAIN { %_.perl.say }' -- 16:06
("-" => Bool::True).hash
- option? Why?
tadzik say about what?@:) 16:07
jnthn r: =pod 16:08
p6eval rakudo c82d10: OUTPUT«===SORRY!===␤Preceding context expects a term, but found infix = instead␤at /tmp/PuDUJrM8wj:1␤»
jnthn tadzik: That :)
tadzik yeah, bare =pod is amroght 16:09
a'right *
like =head1 for example
r: =begin pod 16:10
p6eval rakudo c82d10: OUTPUT«===SORRY!===␤Preceding context expects a term, but found infix = instead␤at /tmp/kcKlWjHGkq:1␤»
tadzik hm, I thought that dies in a nicer way
moritz r: =begin pod␤
p6eval rakudo c82d10: OUTPUT«===SORRY!===␤=begin without matching =end␤at /tmp/06PvqPrGAg:1␤»
tadzik maybe p6eval's fault
ah, there we go
moritz++
moritz it's the missing newline, not p6eval ;-)
kresike bye folks 16:11
16:11 kresike left
masak r: =pod␤ 16:11
p6eval rakudo c82d10: ( no output )
masak so, no bug.
masak closes ticket
tadzik jnthn: is today's nqp commit possibly fixine the Bailador bug I sent you once? 16:12
jnthn tadzik: Very possibly.
tadzik \o/
can't wait to get back home
the commit message is promising :) 16:13
isBEKaml r: say $_!X::a;
p6eval rakudo c82d10: OUTPUT«===SORRY!===␤Cannot call private method 'a' on package X because it does not trust GLOBAL␤at /tmp/GrO7Ezc_sX:1␤»
pmurias jnthn: should contextuals be renamed in nqp to dynamic variables? 16:14
jnthn tadzik: oh, wait...that commit won't fix it
tadzik: I'm working on a fix.
tadzik: That commit is just about detecting the underlying problem. Now I can reliably detect it, I'm working on the stuff to deal with it. 16:15
pmurias: Probably
[Coke] gets an idea for a perl6 module and sees it was done 2 years ago.
masak GlitchMr: feel free to submit a bug report about --
[Coke] will catch up eventually.
masak [Coke]: what module, ooc?
jnthn pmurias: Just dates back to when we called them contextuals ;)
[Coke] masak: one of yours. 16:16
16:16 shinobicl left
masak thought so ;) 16:16
then the chance is high that it needs refreshing or de-dusting.
tadzik what modue, ooc? :) 16:17
jnthn: aha, I see
so atm it just complains. Still good :) 16:18
[Coke] masak: is yapsi dead? 16:20
masak very dormant, at least.
I've learned enough about compilers since last touching it that it would probably benefit from another complete rewrite. :) 16:21
16:21 spider-mario joined
[Coke] if I clone it, run "ufo && make", I see no yapsi. :( 16:22
masak hm.
there should be one in bin/yapsi
[Coke] er, sorry: Could not find Yapsi in any of: /home/coke/.perl6/lib, /home/coke/sandbox/installed_rakudo/install/lib/parrot/4.7.0-devel/languages/perl6/lib
masak you probably still have to set PERL6LIB to ./lib 16:23
I usually do.
isBEKaml [Coke]: IIRC, there was a make install step or atleast one step that would copy over Yapsi.pm to your PERL6LIB.
[Coke] masak: yup, just tried that; nice if that was in the readme. 16:24
masak I'd apply that patch.
[Coke] isBEKaml: I'd rather run it locally if I'm going to do something evil with it.
16:25 perigrin joined
GlitchMr rt.perl.org/rt3/Public/Bug/Display...?id=115744 16:25
isBEKaml [Coke]: I don't understand, read your msg like 5 times and see no relation to what I said. :)
dalek rlito: 5f528ed | (Flavio S. Glock)++ | TODO-perlito5:
Perlito5 - parser - more TODO special cases
 16:26
masak [Coke]: Yapsi was very much -Ofun. this is one of the most tongue-in-cheek things I've ever written: github.com/masak/yapsi/blob/master...ce/2011.03
isBEKaml masak: Yeah, I went over your announce messages and sort of wondered at the striking references to HHGTTG. 16:29
masak :)
isBEKaml masak: At first, it read like the announce made to destroy earth for the super highway on the notice board at the end of the galaxy or a longer vogon poetry. :P 16:30
16:31 brrt left
masak I wonder if I wrote this before or after I had read "Quicksilver" by Neal Stephenson. 16:31
isBEKaml Where would tests for #115308 go? S32-exception, I believe? 16:34
dalek ast: b5f8ab8 | GlitchMr++ | S06-other/main-usage.t:
Add tests for RT #115744
 16:35
moritz isBEKaml: S32-exception/misc.t, or (preferrably) S12-methods/private.t
isBEKaml moritz: okay 16:37
[Coke] points jnthn at the mailing list for a patch.
nwc10++ # ooh, 2 patches today. 16:38
jnthn yeah, just looking at it 16:39
nwc10: about?
[Coke] +1 on the optimize patch from me, as ISTR that's bitten me at one point. 16:40
jnthn yeah, I'm +1 to that too
[Coke] seen au 16:41
aloha au was last seen in #perl6 2 days 6 hours ago joining the channel.
16:48 gdey joined 16:49 fgomez left
dalek p: 7a24af6 | jnthn++ | src/ (6 files):
Make handlers for repossession conflicts possible.

This allows a language where such issues may arise to provide its own handler for them. If it doesn't, we just flag up the error as before.		 16:53
ast: 1ef7dda | isBEKaml++ | S12-methods/private.t:
Added test for RT #115308
 16:57
isBEKaml moritz: please check my last commit. I'm not sure that's correct.
16:59 fgomez joined
dalek kudo/nom: 55912f8 | jnthn++ | tools/build/NQP_REVISION:
Bump NQP_REVISION for deserialization improvements.
 17:05
kudo/nom: bfa208d | jnthn++ | src/Perl6/ (2 files):
Handle Stash deserialization conflicts.

This fixes RT#112626, and probably a bunch of other disappearing symbol in pre-compilation issues.
17:06 birdwindupbird left 17:07 lumi_ left
isBEKaml rn: class A { state $b; }; 17:08
p6eval niecza v22-32-gee5dcf1: OUTPUT«Potential difficulties:␤ $b is declared but not used at /tmp/fAs9b7g2pS line 1:␤------> class A { state ⏏$b; };␤␤»
..rakudo c82d10: ( no output )
dalek ast: d0b57ce | isBEKaml++ | S12-class/attributes.t:
Added test for RT #114234
 17:10
17:10 Guest58684 joined
isBEKaml r: :(::T $x); 17:11
p6eval rakudo c82d10: ( no output )
dalek ast: d20870d | isBEKaml++ | S12- (2 files):
Incremented plan count for two RT issues in previous commits.
 17:15
jnthn std: sub infix:<> ( ){}; 17:19
p6eval std 04216b1: OUTPUT«===SORRY!===␤Null operator is not allowed at /tmp/p0nC7uP2Sy line 1:␤------> sub infix:<>⏏ ( ){};␤Other potential difficulties:␤ Pair with <> really means a Nil value, not null string; use :('') to represent the null string,␤ or…
jnthn ah...that's how it gets the better error... 17:20
std: :<>
p6eval std 04216b1: OUTPUT«Potential difficulties:␤ Pair with <> really means a Nil value, not null string; use :('') to represent the null string,␤ or :() to represent Nil more accurately at /tmp/zK3fncqXkh line 1:␤------> :<>⏏<EOL>␤ok 00:00 41m␤»
masak isBEKaml++ # writing tests
jnthn masak: bfa208d may well fix the pre-comp issue you ran into with Adventure::Engine. 17:21
phenny: suggest to tadzik that a module test run may be an idea, now I've fixed are pre-comp bug (hopefully without breaking anything else) 17:22
phenny: tell tadzik that a module test run may be a good idea, now I've fixed are pre-comp bug (hopefully without breaking anything else)
phenny jnthn: I'll pass that on when tadzik is around.
tadzik jnthn: I'm on it
phenny tadzik: 17:22Z <jnthn> tell tadzik that a module test run may be a good idea, now I've fixed are pre-comp bug (hopefully without breaking anything else)
jnthn \o/
tadzik phenny: why, thank you
good that I have fresh test results from ~monday 17:23
jnthn phenny is sufficiently smart, sometimes I forget I can't put <anything> <name> :)
masak jnthn: yay 17:24
jnthn: I'll try later tonight.
felher Any comments on gist.github.com/4079921 as a translation of/for rosettacode.org/wiki/Bulls_and_cows/Player#Perl ? 17:26
17:27 MayDaniel joined
tadzik wow, I just got a segfault running panda 17:27
awesome :)
masak reproducible? 17:28
tadzik yes
but seems to have stopped after I removed /home/tadzik/src/parrot/install/lib/4.8.0-devel/languages/perl6/site
masak felher: looks very nice. 17:29
my $loose = +$a.grep: { $_ == any @$b }; 17:30
tadzik felher: that looks nice :)
masak could be shortened to +$a.grep: any @$b
tadzik ok, reproducible again 17:31
after bootstraping panda on rakudo HEAD, it always segfaults
jnthn fuck.
tadzik: It certainly isn't always segfaulting?
felher masak: of course, smartmatching. Thanks a lot :)
tadzik gist.github.com/4079963
jnthn tadzik: As in, simple perl6 -e "say 42" works?
tadzik panda bootstraped ok
that's a lot of perl6 already ran 17:32
jnthn Hm
tadzik the segfault is in the new-ish code
Serialization_deserialize
jnthn shit.
:/
tadzik :(
masak felher: the rest looks really nice.
jnthn The backtrace doesn't even have line numbers...guess it's not a debug build? 17:33
tadzik no :/
debug builds are awfully slow on this machine
I can do it though
it'll take ~30 minutes I think
do want?
felher masak: happy to hear that :)
dalek ast: 0b90fb4 | isBEKaml++ | S06-signature/type-capture.t:
Added test for RT #114216
 17:34
jnthn tadzik: Seeing if I can reproduce it here
tadzik ok
steps: bootstrap panda, run 'panda help' or whatever
17:34 sorenso left
masak felher: I tend to inline functions in their innermost relevant block nowadays. people's tastes vary, though. 17:35
17:36 Vlavv` joined
masak felher: I note that if you do that with score-correct (put it in the repeat while loop), it only needs one parameter, not four. 17:36
felher masak: this is an awsome idea :)
masak and read-score would need none.
jnthn *sigh* This is why I put off looking at pre-comp bugs. It's always hassle. 17:37
masak .oO( first he breaks rakudo, now he complains? sheesh. ) :P
17:37 Chillance joined
tadzik hehe, that's about my stance about panda's bootstrap.pl 17:37
No! Don't touch this! Please! :(
masak hammer time. 17:38
tadzik Panda Stahp
isBEKaml
.oO(alpha code don't play well with fire?)
17:39 Vlavv left
masak "at least it works." oh wait. 17:39
jnthn how the hell is that thing NULL... 17:41
17:42 xilo joined
isBEKaml jnthn: just like my days back last year, "Of all the exceptions/errors out there, why did I have to get a friggin' NPE????" 17:42
17:43 wamba left 17:44 grondilu joined
grondilu rn: my $x = 3; $x [R/]= 2; # meaning $x = $x R/ 2; 17:44
p6eval niecza v22-32-gee5dcf1: ( no output ) 17:45
..rakudo bfa208: OUTPUT«===SORRY!===␤Preceding context expects a term, but found infix = instead␤at /tmp/peySrlC51n:1␤»
grondilu r: my $x = 3; $x [R/]= 2; say $x;
p6eval rakudo bfa208: OUTPUT«===SORRY!===␤Preceding context expects a term, but found infix = instead␤at /tmp/e9_RzXEwIm:1␤»
grondilu n: my $x = 3; $x [R/]= 2; say $x;
p6eval niecza v22-32-gee5dcf1: OUTPUT«0.66666666666666663␤»
grondilu n: my $x = 3; $x [R/]= 2 += 1; say $x;
p6eval niecza v22-32-gee5dcf1: OUTPUT«Unhandled exception: Writing to readonly scalar␤ at /tmp/iVbUO53CVY line 1 (mainline @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4211 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4212 (module-CORE @ 578) ␤ at /home/p6eval/niecza/li… 17:46
isBEKaml r: my $x = 3; $x R/= 2; say $x;
p6eval rakudo bfa208: OUTPUT«Cannot assign to a non-container␤ in block at src/gen/CORE.setting:12308␤ in block at src/gen/CORE.setting:12320␤ in block at /tmp/3EQ00lUCDT:1␤␤»
grondilu n: my $x = 3; ($x [R/]= 2) += 1; say $x;
p6eval niecza v22-32-gee5dcf1: OUTPUT«1.6666666666666667␤»
jnthn Well, the bothersome thing is that it deferences the same thing later on, right after the code I added...
felher masak: I updated the version to declare the subs within the scope of repeat :) 17:48
masak: I think you are right. This looks neater :)
17:49 wk joined, wk is now known as Guest3370
masak felher: nice! 17:51
felher masak: ah, and since we have smartmatching and score-correct only has one argument left, one can shrink the ".= grep"-line :) 17:53
isBEKaml #114134 is an error originating in binder. does dies_ok fit the bill?
17:54 Guest58684 left
masak yeah :) 17:54
@candidates .= grep(&score-correct)
isBEKaml: should do, if the binding happens at runtime.
isBEKaml masak: nominal type check, unbound variable - not compile time? 17:55
masak looks 17:56
oh, that one.
17:57 dakkar left
masak no, that should be runtime, and should be perfectly testable with dies_ok. 17:58
[Coke] do we need "dies_with_isa" ? 17:59
(or something much better named)
isBEKaml masak: I added two anyway. Hopefully, you or others can update it with a better test. 18:05
dalek ast: edf27d4 | isBEKaml++ | S32-exceptions/misc.t:
Added test for #114134
18:08 benabik joined
isBEKaml g'night, #perl6! :) 18:10
18:10 isBEKaml left, fgomez left
[Coke] (or perhaps a multi of dies that takes an optional class) 18:11
er, of dies_ok
pmurias jnthn: when do you plan to start porting nqp/rakudo to the jvm? 18:14
felher rosettacode.org/wiki/Bulls_and_cows...yer#Perl_6 18:15
skids Is there a settled solution for using data file from t/ during tests -- e.g. how to find the path, given the cwd could be anywhere? 18:16
18:16 wtw left
jnthn pmurias: I'm slowly chipping away at some initial design work. 18:19
18:22 wtw joined 18:23 wamba joined
pmurias works on rakudo-js in a TDD manner, making things saner when it's required to pass tests ;) 18:29
jnthn :) 18:30
tadzik: OK, got it
tadzik \o/
jnthn Well
Got a fix.
dalek p: a827893 | jnthn++ | src/6model/serialization.c:
Fix double-repossession.

Not completely sure why that can happen just yet, but this should be enough to make it not be a problem.
pmurias jnthn: that means we need to expand the nqp test suit at some point :)
18:30 grondilu left
jnthn Still wouldn't mind figuring out exactly why it's needed... 18:30
pmurias: I'm most certainly fine with expanding the NQP test suite. I know we're probably falling somewhat short on having good test coverage. 18:31
nwc10 jnthn: I'm sort of about, but I'm not sure when I'm going to fall asleep 18:38
jnthn nwc10: OK. The patches look good; on the multi one I'd have been tempted to pass the graph nodes rather than pass info and the constraints, but it's not a big thing. 18:39
nwc10: Do you have a commit bit?
nwc10 no
jnthn Hm
Do I have to worry about CLA stuff? :) 18:40
As in, does the thing you already have with TPF for Perl 5 mean we're good?
nwc10 I don't know and I don't know.
pmurias .WHAT returns the type object? 18:41
jnthn nwc10: Did you ever sign a CLA for TPF, e.g. www.perlfoundation.org/contributor_..._agreement ?
pmurias: yes
nwc10 no, I have never signed a CLA for TPF.
jnthn Ah.
Oh, maybe the Perl 5 and Perl 6 copyright work different... 18:42
nwc10 well, TPF doesn't hold the copyright on Perl 5
jnthn That'd explain why you've not done a TPF CLA...
PerlJam nwc10: in any case, it can't hurt to submit a CLA :) www.perlfoundation.org/contributor_..._agreement 18:43
diakopter yes, it can hurt
PerlJam diakopter: how?
diakopter if the person doesn't want to
nwc10 it can hurt.
PerlJam oh, then don't. that's a "doctor it hurts when I do this" situation. 18:44
diakopter no, it isn't; he wasn't complaining about something that already happened or is happening
he was trying to ascertain whether it would hurt. 18:45
18:46 am0c joined
ggoebel Uniqueness and Reference Immutability for Safe Parallelism: robjsoftware.org/2012/11/02/a-publi...in-public/ 18:46
diakopter skimmed it; it was interesting to me
nwc10 jnthn: I'd not considered passing the nodes. It would drop the parameters to that function, which seems like a better idea 18:47
as the long-ish parameter list was bugging me a bit
looked ugly
jnthn nwc10: Yes, keeping it to the two was my thought.
nwc10 will (try to) do that tomorrow 18:48
jnthn OK. I can apply the other one. 18:49
nwc10 cool, thanks
dalek kudo/nom: ac77f28 | jnthn++ | tools/build/NQP_REVISION:
Bump NQP_REVISION to get SC fix.
jnthn Thank *you*. :)
tadzik: Give it another go :) 18:50
nwc10 more general thought was that the C code in multidispatch.c seems to be doing three things:
1) converting Rakudo (well Parrot) data structures descripting subroutines into the C structs
2) sorting them 18:51
3) managing the cache
and right now it has to be in that order, because the sorting is done in C using the C structs
jnthn There's a 4) based on given arguments, choosing the candidate 18:52
and 5) doing compile-time analysis of a dispatch
:)
nwc10 and a fanatical loyalty to the pope?
masak :) 18:53
jnthn No, just ruthless efficiency :P
pmurias subclassing now works in nqp-js :) 18:54
nwc10 :-)
tadzik jnthn: yep, doesn't segfault anymore :)
jnthn phew
nwc10: 1, 2 and 5 are not performance sensitive
nwc10: 4 sorta is BUT if a cache hit happens can be dealt with very swiftly 18:55
nwc10 I'd not identified what is doing (5). But that doesn't matter (yet).
I can see how (2) and (1) have to happen before (3), and (3) before (4) 18:56
jnthn (5) is Rakudo_md_ct_dispatch
nwc10 aha. As it says in its docs - *at compile time*
jnthn Need to pop out for 10 minutes...bbs 18:59
19:01 HarryS left 19:02 pmurias left, HarryS joined 19:05 sorenso joined 19:07 HarryS left 19:08 fhelmberger left
jnthn back 19:11
19:20 benabik left 19:32 sizz_ joined 19:33 sizz left
sorear good * #perl6 19:34
19:38 hash_table joined 19:39 Guest58684 joined 19:44 fgomez joined
masak sorear! \o/ 19:47
19:47 BillySleep is now known as BillySeth
sorear masak! \o/ 19:52
19:54 benabik joined 19:58 am0c left
[Coke] skids: when running files in t/ your cwd is pretty much always "the directory above t" 20:00
20:02 HarryS joined 20:17 domidumont left 20:19 benabik left 20:24 benabik joined 20:30 cognominal joined 20:32 pnu_ left, pnu joined
jnthn tadzik: Did it do much beyond "not segfault"? :) 20:34
tadzik jnthn: emmentaler is still running :)
patience, that's just good ol' core2duo :)
jnthn tadzik++ 20:36
tadzik I'll probably make it to the shop before it finishes :) 20:37
. o O ( I hope something breaks, so I can show off difftool )
jnthn What if it fixes something?
Then you can also show off difftool :P
tadzik oh, right, right :) 20:38
dalek kudo/nom: c58f4ce | nicholas++ | tools/lib/NQP/Configure.pm:
Using --parrot-option should override the default option of '--optimize'

Otherwise, it's possible neither to call Parrot's Configure.pl without enabling optimizations, nor to change the value of the --optimize option.		 20:41
kudo/nom: a41a483 | jnthn++ | docs/ChangeLog:
Update ChangeLog.
GlitchMr I wonder how fast Perl 6 would be on CPU like... let's say... Pentium 133MHz
20:41 SamuraiJack left 20:42 Guest58684 left
xilo would take ages just to compile it 20:44
masak it would still need at least 1 GB of RAM. 20:45
preferably 1.5 GB.
sorear rakudo today would, sure 20:50
howefully be the time the P133 comes out we'll be able to make a leaner implementation 20:51
20:55 eternaleye left, daxim left
tadzik :D 21:02
jnthn: breakages :( gist.github.com/4081246 21:04
oh, wait
sorear o/ tadzik
tadzik sorear: o/
21:04 GlitchMr left
tadzik jnthn: actually, gist.github.com/4081251 21:04
may be unrelated to new rakudo commits, but still a green light :) 21:05
jnthn tadzik: hah, diff'd the wrong way round :P 21:08
.koń? horse? :)
tadzik: Thanks for testing 21:09
I think Game::Crypt was one of the modules with issues
tadzik yes, .koń is a horse :)
jnthn Close to the Slovak word :)
tadzik .dupa was taken :P 21:10
21:10 Guest58684 joined
masak hey! watch the language! :) 21:10
masak .oO( Polish that turd ) 21:11
tadzik :P
jnthn Well, I did for a moment worry that it wasn't horse, but a loan from French that would be a good match for dupa :P 21:12
tadzik ;)
masak the perils of knowing a bunch of languages... :) 21:13
tadzik cross-language puns 21:14
jnthn "knowing" :)
(in my case)
tadzik Quidquid latine dictum sit, altum videtur
masak indeed.
wow, I can still parse that, after all these years.
the passive should be a verb form in more languages... :) 21:15
phenny: "ĉio dirita en Esperanto aspektas amuza"? 21:16
phenny masak: "everything said in English looks fun" (eo to en, translate.google.com)
masak ahaha.
statistical translation. gotta love it. 21:17
tipdbmp How to match a non word character (\W), that is not a "space" character \S?
tadzik :D
my second favourite sentence now
masak tipdbmp: <-[\w\s]>
tadzik masak: is the pronounciation easy? 21:18
tipdbmp Hm, tnx.
masak tadzik: was just gonna say. I don't even have to tell you how to pronounce that.
tadzik hm
masak tadzik: penultimate-syllable stress, just like in Polish.
tadzik well, I'm the one who turns "ci" into "ć" and stuff
masak troo
tadzik tró 21:19
21:19 benabik left
Tene disappointed by recent nqp issue, "a strange loop" 21:19
diakopter :)
tadzik but well, I only once met a guy who knew latin and had to fix my pronounciation of the former sentence
diakopter Tene: what's disappointing about it? do you have any ideas what could be causing it? 21:20
diakopter has no clue
21:21 kaare_ left
jnthn It's because it doesn't do any optimization for while loops in sink context at code-gen time. 21:21
21:21 wamba left, eternaleye joined
sorear phenny: fr en "Paris"? 21:21
phenny sorear: "Paris" (fr to en, translate.google.com)
jnthn And thus it needs to find a common register type between the condition and the body, and in the case without the line commented out that comes out as int. 21:22
masak tadzik: how to pronounce Latin phrases depends if you ask a Swede, a German, or an Italian.
Tene diakopter: There's a Douglas Hofstadter book titled "I Am A Strange Loop", which is sort of a followup to Godel, Escher, Bach
21:22 wamba joined, FROGGS left, flightrecorder left
masak good book. 21:22
diakopter Tene: oh, yes, intentional allusion
masak phenny: la en "quid pro quo"? 21:23
phenny masak: "Grindhouse" (la to en, translate.google.com)
masak :D
at one point the translation was "What happens in Vegas"...
tadzik masak: I asked a student (or former student) of antique culture
diakopter Tene: but what's disappointing 21:25
Tene diakopter: that it wasn't about self-referential loops 21:26
sorear despite the efforts of conservatives, Latin pronunciation has changed quite a bit in the last 2,000 years 21:27
tadzik dead language, eh? :)
sorear there's been some interesting progress in the last couple decades based on using a comparative study of loanwords to tease out phonology from the old documents themselves 21:28
masak contemporary Latin isn't dead.
tadzik Latin Punks Not Dead! 21:30
masak ;) 21:31
21:37 benabik joined 21:42 marloshouse left 21:51 bbkr__ left 21:53 marloshouse joined, Guest58684 left 21:54 hash_table left, sorenso left
masak phenny: en la "Latin Punks Not Dead!"? 21:57
phenny masak: "Latine Punks non est mortuus !" (en to la, translate.google.com)
21:57 spider-mario left
tadzik I'll write that on some wall 21:58
22:00 Guest58684 joined, benabik left
jnthn I thought you were meant to write ROMANES EUNT DOMUS on walls :P 22:02
22:03 benabik joined
tadzik :D:D:D 22:03
100 times!
22:04 skids left
sorear phenny: "ROMANES EUNT DOMUS"? 22:05
phenny sorear: "ROMANESQUE EUNT DOMUS" (fr to en, translate.google.com)
sorear phenny: "romanes eunt domus"?
phenny sorear: "Roman domus eunt" (fr to en, translate.google.com)
sorear phenny: la en "romanes eunt domus"?
phenny sorear: "Romans go home" (la to en, translate.google.com)
22:08 MayDaniel left
masak 'night, #perl6 22:25
cotto ''ight, masak 22:30
22:32 cognominal left 22:40 xilo left 22:42 cognominal joined 22:43 Guest3370 left 22:53 xilo joined, wamba left 22:55 xilo left 22:58 hash_table joined 23:02 PacoAir left, grondilu joined
grondilu rn: my %h = 0 => 1, 1 => 0; say max(:by(*.value), %h); say %h.max(:by(*.value)); 23:03
p6eval rakudo a41a48: OUTPUT«"0" => 1␤"1" => 0␤»
..niecza v22-32-gee5dcf1: OUTPUT«"0" => 1␤Unhandled exception: Excess arguments to Any.max, unused named by␤ at /home/p6eval/niecza/lib/CORE.setting line 0 (Any.max @ 1) ␤ at /tmp/Ns7P8fgveK line 1 (mainline @ 7) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4211 (ANON @ 3) ␤ at /hom…
grondilu was expecting "0" => 1n"0" => 1 23:04
dalek rl6-roast-data: fb84488 | coke++ | / (4 files):
today (automated commit)
 23:07
23:09 BillySeth is now known as BillySeth|bbl 23:16 stopbit left 23:17 hash_table left 23:19 ifim joined 23:21 ifim left 23:24 cognominal left 23:25 ifim joined 23:28 cognominal joined
jnthn 'night, #perl6 23:33
23:45 alester left 23:49 skids joined
dalek ok: 397fd9c | timo++ | src/regexes.pod:
fix #78, thanks hbaragar (s/use/lose/)
 23:55

Powered by Cro and the Raku® Programming Language.

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

Thank you!