pugscode.org/ planetsix.perl.org/ | nopaste: sial.org/pbot/perl6 | evalbot: perl6: say 3; (or rakudo:, pugs:, elf:, etc) | irclog: irc.pugscode.org/
Set by mncharity on 5 January 2009.
00:00 vixey left
eric256 mberends: sorry work called 00:03
i thought it was single line ;)
okay that will take me a while to understand but thanks! ;) 00:04
later 00:05
00:05 eric256 left 00:08 mbaro joined 00:12 justatheory left, Student left 00:13 spx2 joined 00:16 Jack joined, Jack left 00:18 spx2 left 00:19 spx2 joined 00:22 r0bby left 00:23 dwhipp left 00:27 justatheory joined 00:50 Student joined
pugs_svn r24789 | particle++ | [S19] rip out option categories and services until they prove useful 00:53
00:55 |jedai| left, |jedai| joined
pugs_svn r24790 | particle++ | [S19] address concerns about mixing -e and -e6 00:56
01:02 hercynium joined 01:06 perl6examples left 01:07 xinming joined, r0bby joined 01:09 cspencer joined 01:14 DemoFreak left 01:17 japhb joined
pugs_svn r24791 | lwall++ | [S11] clarify the intent of the -e6 remark not to be special syntax, but 01:20
r24791 | lwall++ | a natural outcome of -e and '6;'
01:24 xinming_ left 01:26 cspencer left 01:30 |jedai| left 01:31 |jedai| joined 01:36 Limbic_Region left 01:44 stephens left 01:46 jhorwitz left 01:52 cspencer joined 01:59 eric256 joined 02:00 perl6examples joined, perl6examples left, perl6examples joined 02:03 alester_ joined 02:04 perl6examples left 02:07 xinming_ joined
cspencer perl6: class Foo { method list { say "bar" } }; my $f = Foo.new; $f.list 02:08
p6eval rakudo 35086: OUTPUT«too few arguments passed (0) - 1 params expected␤current instr.: 'parrot;Foo;list' pc 187 (EVAL_13:91)␤»
..elf 24791, pugs: OUTPUT«bar␤»
02:08 mbaro left 02:15 aindilis` joined 02:17 alc joined 02:23 eternaleye joined 02:24 alester_ left 02:25 xinming left 02:30 aindilis left 02:32 fridim_ joined 02:34 |jedai| left 02:35 |jedai| joined 02:37 perl6examples joined 02:40 ChrisDavaz joined
eric256 abandoned twitter in favor of a pure HTTP + IRC bot 02:44
02:46 cspencer left
pmichaud rakudo: class Foo { method list { say "bar" } }; Foo.new.list; 02:49
p6eval rakudo 35086: OUTPUT«too few arguments passed (0) - 1 params expected␤current instr.: 'parrot;Foo;list' pc 164 (EVAL_12:84)␤»
pmichaud interesting. Will fix after rvar.
02:54 perl6examples left 03:00 |jedai| left, alc left 03:01 |jedai| joined 03:02 ChrisDavaz left 03:03 wknight8111 left 03:04 ChrisDavaz joined 03:07 perl6examples joined
eric256 the bot can now add people as collaborators..... 03:10
perl6examples: add eric256
hmmm it should probably announce that you've been added, but this is nice and quite ;)
literal you mean quiet? 03:12
mberends are there security concerns with adding collaborators via irc? 03:15
literal not if the bot only obeys commands from a certain hostmask, or requires authentication...
mberends perl6examples: add xxxx # only joking... 03:18
eric256 mberends: i don't think there realy are security issues 03:19
someone could abuse it, but then they could come on here and ask me to add them and then abuse it too
the barrier to entry is the same either way ;)
this is just easier on me ;)
mberends that's fair enough
03:20 ChrisDavaz left, ChrisDavaz joined
eric256 so feel free to add people and/or edumacate people on adding themselves if i'm not around 03:20
mberends certainably 03:21
s1n eric256: have you run the euler code recently? 03:23
eric256 no 03:25
with our naming convention i'm hoping to automate nightly smokes of it though ;)
why? is there a problem with it? 03:26
03:26 xinming joined
s1n eric256: try it out, i'm getting segfaults on some 03:31
eric256: for example, prob004 segfaults for me
eric256 yea i get a segfault on 001-unobe and i'm waiting for a reply on prob004 03:32
s1n backtraces don't seem consistant 03:33
eric256 i've been getting segaults on large ranges lately
03:34 xinming_ left
s1n i tried some of the new prefix operators pmichaud blogged about and it didn't crash from the RE(P)L 03:34
perl6: my $sum = [+] 1..5; 03:35
p6eval rakudo 35086: RESULT«15»
..pugs: RESULT«\15»
..elf 24791: OUTPUT«/home/evalenv/pugs/misc/elf/elf_f_src/STD_red/match.rb:117:in `block in to_dump0': undefined method `to_dump0' for nil:NilClass (NoMethodError)␤ from /home/evalenv/pugs/misc/elf/elf_f_src/STD_red/match.rb:117:in `map'␤ from
../home/evalenv/pugs/misc/elf/elf_f_src/STD_red/match.rb:1...
s1n see, it works there, but if i put it into a script and run it, it segfaults
03:39 azawawi_ joined
eric256 are you running the same version? that works fine on my rakudo 03:42
03:43 alc joined
eric256 rakudo: my $sum = [+] 1..100; 03:43
p6eval rakudo 35086: RESULT«5050»
eric256 rakudo: my $sum = [+] 1..1000;
p6eval rakudo 35086: RESULT«500500»
eric256 rakudo: my $sum = [+] 1..10000;
p6eval rakudo 35086: RESULT«50005000»
eric256 rakudo: my $sum = [+] 1..100000;
p6eval rakudo 35086: No output (you need to produce output to STDOUT) 03:44
eric256 run a perl6 -v and see what revision you are on
mberends numeric overflow because Num is not (fully) implented. you get full precision to 1 million. 03:45
*implemented
s1n 35086 03:46
revision 35086 built on parrot 0.8.2-devel, for x86_64-linux-thread-multi
03:49 azawawi left 04:01 mberends left 04:02 elmex_ joined
s1n ugh, spectest passes but i'm still getting segfaults 04:02
04:17 elmex left, elmex_ is now known as elmex 04:18 xinming_ joined 04:23 fridim_ left 04:25 kisu_ joined, Student left 04:27 AzureStone_ joined, AzureStone left 04:29 kisu left, Ardonel|work is now known as Ardonel|sleep
s1n is there any way of converting very large numbers (1e50) to something like the _full_ number 04:35
for example, "(5 ** 50).say" gives scientific notation, but i want the full number 04:36
04:36 xinming left 04:37 pdcawley joined
s1n without going the grunt work of expanding the number myself that is 04:39
rakudo: (5**50).say 04:43
p6eval rakudo 35087: OUTPUT«8.88178419700125e+34␤»
s1n in other words, how can i get that full number in either a string or some other means for processing (i.e. printing)?
04:46 plainhao joined 04:54 alc left 05:00 |jedai| left 05:02 |jedai| joined 05:11 plainhao left, alester left 05:19 |jedai| left 05:20 |jedai| joined, masak joined
azawawi_ hi masak 05:28
05:28 japhb left
masak azawawi_: hi 05:28
azawawi_ how r u? i hope you're doing fine/great ;-)
masak is happy because he got up early today 05:29
azawawi_: how about you?
azawawi_ is happy since he has like 25 mins to hack before going to work ;-)
05:29 azawawi_ is now known as azawawi
masak sounds like we're in approximately the same time zone :) 05:29
azawawi yup
05:30 xinming joined
azawawi 7:30am 05:30
05:30 japhb joined
masak 6:33 here. 05:30
azawawi run rakudo script in Padre win32 worked today...Yay!
masak cool! 05:31
azawawi i cant find 'say' in S29
masak I see use.perl.org is back up.
azawawi: it's in S16.
azawawi: I don't know what to do about those yet -- maybe duplicate outright. 05:32
but it feels partly wrong.
05:32 alester joined
azawawi i cant understand S29 is about functions 05:32
and say is a function right?
masak :) 05:33
yes.
you're right, you're right.
tell you what, I'll add it right away.
azawawi remove obsolete functions to another section head? 05:34
masak azawawi: say again? I didn't get that.
azawawi use the script you created, and take a look
i need a keyword or a list of keywords in =item -s 05:35
05:36 xinming_ left
azawawi so S29 should be more parse friendly 05:36
masak azawawi: ah, so your problem is that the 'Obsolete' section contains items with multiple words?
azawawi so i type 'Obsolete' i get a keyword calltip on time for instance 05:37
s/so/so when/
masak azawawi: suggestion. add a special case for 'Obsolete', so that it's ignored when parsing S29.
azawawi or we can think a bit more generic and add a tag or something 05:38
to mark it as a function
masak sure, that works too.
azawawi so we can extract info from S29 or S16 or S06...
what do u suggest for the tag? 05:39
something like smart links but let us call it dumblinks ;-)
masak 'dumblinks'. I like that. :) 05:40
05:40 |jedai| left
rakudo_svn r35091 | infinoid++ | [cage] Fix up some codetest failures. 05:40
05:41 |jedai| joined
pugs_svn r24792 | masak++ | [S29] put the C<polar> function inside a list, for consistency 05:43
masak ok, so IO currently doesn't have a =head2 in S29.
I'll add one.
I also expect some people will protest; this is duplication of information, after all. 05:44
we'll see what happens. the revert button is always close by.
I'll add print, say and printf for the time being. 05:47
oh, and getc. 05:48
we should probably add a readline eventually too. 05:49
azawawi if you add for instance a link to other Synopsis-es in =item...
05:50 kisu_ is now known as kisu 05:51 alc joined
masak azawawi: yes, that's what's done in some places. I find that unsatisfactory, however, if S29 is to function as a sort of 'perldoc perlfunc'. 05:52
hm.
azawawi perldoc perl6s29func ;-) 05:53
masak I'm just about to commit. maybe I should abort and just add 'see S16' for these four functions?
what do you think?
azawawi i think we need to think about it a bit
masak ok. aborting commit.
azawawi since i have like 5 min before going work ;-) 05:54
let ask moritz_ about that today
masak oh, we can fix it before you go :)
let's be conservative and add 'see S16' for the time being.
and then later, if it's a good idea, we'll copy the content. 05:55
azawawi and when the script sees S\d\d it should jump there and index its =item 05:56
masak azawawi: I can't blame you for lack of ambition :)
pugs_svn r24793 | masak++ | [S29] added mentions of getc, print, say and printf, with references to S16 05:59
r24793 | masak++ | for details
r24794 | masak++ | [S16] slight disambiguation 06:00
masak purrs 06:01
azawawi lex cache for parsing std.pm => 1660 file
lex cache for parsing hello.p6 => 840 files 06:02
masak sounds impressive. in a slightly sick way. 06:03
azawawi work & 06:07
thx masak++! cya in the evening...
06:07 azawawi left
masak buubot: spack succ 06:12
buubot masak: S02-bits.pod:1 S03-operators.pod:11 S04-control.pod:9 S05-regex.pod:25 S06-routines.pod:4 S07-iterators.pod:1 S09-data.pod:3 S12-objects.pod:7 S16-io.pod:9 S26-documentation.pod:1 S29-functions.pod:4
masak buubot: spack pred
buubot masak: S02-bits.pod:11 S03-operators.pod:6 S05-regex.pod:2 S06-routines.pod:4 S09-data.pod:2 S10-packages.pod:1 S11-modules.pod:1 S12-objects.pod:1 S26-documentation.pod:3
06:13 xinming_ joined
masak buubot: spack \bsucc\b 06:13
buubot masak: S03-operators.pod:3
masak buubot: spack \bpred\b
buubot masak: S03-operators.pod:2
masak that's better. 06:14
ok, so .succ and .pred currently have one behaviour on Num, and another on Bool. 06:15
adding that to S29.
06:18 xinming left
pugs_svn r24795 | masak++ | [S29] added .succ/.pred for Num, Int and Bool 06:25
06:38 eternaleye left 06:53 ashizawa joined 06:57 xinming joined 07:01 pbuetow joined 07:10 DemoFreak joined 07:14 mberends joined 07:15 xinming_ left 07:16 mtnviewmark joined
mtnviewmark ping masak.... 07:17
masak mtnviewmark: I'm here.
mtnviewmark do you ever sleep?
:-)
masak mtnviewmark: actually, yes. sometimes.
but today I decided to get up early.
I'm glad you have the impression that I'm always here.
mtnviewmark so - would you like to take a gander at the new chart and would you be willing to proof the factual content (operator names, precedences, etc...)? 07:18
masak yes, of course.
URL?
mtnviewmark I actually think you are the oracle behind std: !
masak mtnviewmark: I'm not denying that.
07:19 |jedai| left
mtnviewmark I could e-mail it to you 07:19
assuming that e-mail I ahve for you works (I sent you e-mail the other day)
masak std: $*oracle_behind === 'masak' ?? 'I knew it' !! 'oh well'
mtnviewmark: it works 07:20
p6eval std 24795: OUTPUT«00:05 86m␤»
07:20 |jedai| joined
masak mtnviewmark: I didn't reply because you came here instead. 07:20
mtnviewmark np
07:22 jan_ left
mtnviewmark anyone else I should mail this to? 07:24
07:24 jan joined
masak mtnviewmark: that's a difficult question to answer. I'm sure many would be interested/eager to help. 07:24
but we work mainly through IRC.
mtnviewmark true - but I'm a bit wary of putting it out there in this "almost done" state -- it is useful enough to get wide circulation -- and, as you'll see, the design part isn't done 07:25
okay - well, I'll start with you
sent
comin' at ya on the inter-toobs 07:26
masak I'm sure moritz_ would be interested. probably jnthn and pmichaud also.
ah! it's here.
masak looks
mtnviewmark hopes masak has a good PDF reader 07:27
masak it's sufficiently good.
ooh -- very different layout.
07:28 xinming_ joined
masak I agree with you on the non-operators -- only the feed thingies feel like ops 07:28
mtnviewmark yes - well, things changed radically since 2004
masak I can imagine... although for me, it has seemed very gradual. 07:29
kudos for doing this -- it's both useful and nice to look at. 07:30
mtnviewmark you should see: i have a printout of the old table with circles and arrows sketching how things changed - it is like a crazy "all change positions" dance
masak :) 07:31
mtnviewmark welcome - there is still a few days worth of work to do on it -- now that I have the factual data in place and laid-out, then there is endless tweaking and enhancing the design --- My guess is it will be done in a week or two
any how -- I'm off to bed - if you find problems in the text, e-mail me please! If you want to pass it on to a few others please do, though I don't think it is ready to be posted on a web site.. 07:32
thanks again
masak np 07:33
07:33 ejs joined
masak mtnviewmark: are you sure there should be 8 variants of ff and fff? 07:33
mtnviewmark they are all in Syn03, no?
masak I'd have to check S03 again.
I can't imagine that they're all different.
but then again, it might just be my lack of imagination...
mtnviewmark: ok, with your permission, I'll fwd it to moritz_, jnthn and pmichaud. 07:34
mtnviewmark please
masak ah, you're right about the ff and fff variants. 07:35
mtnviewmark well - there is no description of ff vs. fff except that fff is "sed style"
masak mtnviewmark: I think they correspond to Perl 5 .. and ... , respectively
mtnviewmark hmmmm, okay
masak so, 'perldoc perlop'. 07:36
mtnviewmark yes- I was off reading it... mind you, it is sort of a numbing thing to read the description of 07:38
and "awk style" vs. "sed style" only helps those who came from there -- and while I *did* come from there (I don't want to tell you how many years ago) -- I long sinced GC'd by brain of awk and sed in favor of perl 07:39
masak mtnviewmark: Steve Yegge once used that part of the documentation as evidence that Perl is a very odd language.
uh, here: steve.yegge.googlepages.com/ancient...uages-perl 07:40
07:41 xinming__ joined
masak Stevey has since repented somewhat. 07:41
07:45 xinming left
mtnviewmark heading to bed - later and thanks 07:46
07:46 mtnviewmark left 07:59 xinming_ left 08:01 maerzhase joined 08:02 eternaleye joined 08:07 ejs left 08:11 xinming joined 08:15 |jedai| left, |jedai| joined 08:22 alester left 08:29 xinming__ left 08:31 pbuetow left
pugs_svn r24796 | masak++ | [S29] added any, all, one and none as subs and methods on List and Hash 08:34
08:41 |jedai| left 08:43 |jedai| joined
masak is there a way to define new metaoperators in Perl 6? 09:10
Matt-W I hope so 09:11
Although I've never seen any syntax for it
masak I'm not so sure it's possible. 09:14
09:15 _jedai_ joined 09:16 |jedai| left
masak (Dave Whipp)++ # dave.whipp.name/sw/perl6/perl6_xmas_2008.html 09:23
mberends Hmm, currently enjoying that as well 09:26
Matt-W that's good 09:27
it was that post on the scripting games that got me trying out perl 6 again, but I did my own thing rather than trying those problems
mberends discovers that Block is already registered as a class name 09:34
avar masak: S03 L</Meta operators
> ... would seem to suggest that metaoperators are just named by convention 09:35
lambdabot <no location info>: parse error on input `...'
avar i.e. ?+ ~+ etc are manually defined, so if you wanted a new metaop you'd have to define one FOO+ op, one BAR+ op etc
09:39 _jedai_ left 09:40 _jedai_ joined 09:41 iblechbot joined 09:42 donaldh joined 09:48 ft joined 09:49 ejs joined
eternaleye masak: Perhaps it could be defined in a similar way to ??!!, with it looking for an operator between the two parts rather than an expression, and using apply() with that captured op 09:57
10:06 _jedai_ left 10:07 _jedai_ joined 10:09 ejs left 10:16 kidd` joined 10:18 spx2_ joined 10:25 _jedai_ left 10:26 DemoFreak left, _jedai_ joined 10:29 spx2__ joined, spx2 left 10:30 spx2_ left 10:33 silenius joined 10:51 _jedai_ left 10:52 _jedai_ joined 11:01 ChrisDavaz left 11:08 mj41 left 11:15 _jedai_ left 11:16 _jedai_ joined, Ardonel|sleep is now known as Ardonel 11:28 silenius left, japhb left, Ehtyar left, zev left, frodwith left, Grrrr left, Southen_ left, szabgab left, Caelum left, SubStack left, cls_bsd left, Tene left, obra left 11:29 Tene joined, silenius joined, japhb joined, Ehtyar joined, Caelum joined, Southen_ joined, SubStack joined, obra joined, Grrrr joined, frodwith joined, zev joined, cls_bsd joined, szabgab joined, irc.freenode.net sets mode: +o Tene 11:42 jferrero joined 11:48 brunoV joined 11:58 ashizawa left, alc left 11:59 _jedai_ left 12:00 _jedai_ joined 12:02 DemoFreak joined 12:03 sri_kraih joined 12:06 jan left 12:07 jan_ joined 12:10 _jedai_ left, _jedai_ joined 12:12 araujo left 12:20 Ardonel left 12:25 kisu_ joined
masak S16:409 defines C<prompt> as an only sub. is there a particular reason why C<prompt> is C<only> while virtually everything else in S16 and S29 is C<multi>? 12:32
mberends there is probably only one sensible way to call it, namely with a Str 12:35
12:35 Lorn joined
masak mberends: so if I'd like to do `prompt(42)`, I should forget about it and define my own wrapper instead? 12:36
12:38 kisu left
mberends how would `prompt( ~ 42 )` go? 12:40
masak that'd work. just unusual to have to do the casting manually in Perl.
wolverian huh. surely coercions are taken into account?
(implicitly) 12:41
masak wolverian: not that I know of.
Str is Str and Num is Num.
at least in parameter passing.
12:41 plainhao joined
mberends I get type mismatch errors regularly ( with typed parameters ) 12:41
wolverian I thought otherwise, but I don't know why. 12:42
masak wolverian: there's been talk about relaxing many of the type signatures in S29 from e.g. 'Str' to 'of Str' or something like that. 12:43
mberends casting manually is also necessary from Match to Str assignment
masak mberends: that's because sometimes you'd want one and sometimes the other. 12:44
mberends yes, but when the lvalue is Str it's clear that coercion is required 12:45
masak haven't tried that 12:46
masak tries that
rakudo: my Str $s = "foo" ~~ /foo/; say $s
p6eval rakudo 35117: OUTPUT«Type mismatch in assignment.␤current instr.: 'die' pc 14950 (src/builtins/control.pir:204)␤»
mberends rakudo: my Str $s = ~ ( "foo" ~~ /foo/ ); say $s 12:48
p6eval rakudo 35117: OUTPUT«foo␤»
masak works as advertised, I would say.
mberends rakudo: my $s = "foo" ~~ /foo/; say "but this is a match: $s" 12:49
p6eval rakudo 35117: OUTPUT«but this is a match: foo␤»
masak mberends: aye, a Match that gets stringified in the string interpolation :) 12:50
mberends a Match can act like a Str, but it doesn't quack like one
masak more like it can turn into a Str when requested to. 12:51
mberends it's knocked my head at least ten times, so now I look out (and duck) 12:53
masak I think it's a matter of having the right expectations in this case. 12:54
I use the Match about half of the time and the Str about half.
mberends btw, I got into Match and so on thanks to the November code. love it! 12:55
masak mberends: glad to hear that. :)
Matt-W Hmm yes, one would expect stringification to be implicit in a string interpolation :) 13:00
masak it sure helps, aye.
mberends otherwise it would have to be "blah {~$foo}". yuk 13:01
Matt-W yes indeed 13:02
uck
what's the prefix operator for numify?
mberends +
Matt-W ah of course
thanks
masak and ? is for boolify.
mberends oh, will have to try that somewhere 13:03
masak often there's no need.
Perl 6 DTRT in most cases.
Matt-W I've not had to do much 13:04
and only explicit string conversion
usually with Match objects
mberends ditto 13:05
13:08 sri_kraih left 13:09 kisu_ left
rakudo_svn r35120 | jkeenan++ | Applying patch submitted by Geraud Continsouzas in trac.parrot.org/parrot/ticket/135 (plus one small fix to CREDITS). 13:10
13:10 DemoFreak left 13:15 _jedai_ left 13:16 _jedai_ joined 13:20 alester joined 13:25 mj41 joined 13:32 kisu joined 13:33 alester left 14:01 kisu_ joined 14:02 kisu left
rakudo_svn r35122 | pmichaud++ | [rakudo]: spectest-progress.csv update: 279 files, 6171 passing, 0 failing 14:10
14:11 meppl joined 14:12 cj joined
cj Trey: hey. I went to tag you in a photo on facebook, but it doesn't look like you've got an account. :) 14:13
14:16 _jedai_ left 14:18 _jedai_ joined 15:00 ejs joined 15:02 eric257 joined
mberends hey eric256++ 15:03
15:03 vixey joined 15:05 Exodist joined
eric257 thanks....dunno what for but i'll take it ;) 15:06
mberends 256 -> 257 15:07
eric257 lmao ahh
yea apparently left IRC on at home...
eric257 tries to get motivated to work 15:10
masak eric257: do you have 256 homes?
eric257 hehe who do you think i am? John McCain? 15:11
15:29 mj41_ joined 15:36 |jedai| joined 15:38 _jedai_ left 15:39 ejs left 15:48 mj41 left 15:50 meppl left 15:51 ft left 15:52 jhorwitz joined 15:56 donaldh left 16:00 aindilis` left 16:10 eternaleye left 16:19 meppl joined 16:20 dwhipp joined, |jedai| left 16:21 |jedai| joined 16:27 spx2__ left 16:28 spx2 joined 16:33 stephens joined
[particle] std: sub perl6(:M($use|$no)) {...} 16:35
p6eval std 24796: OUTPUT«############# PARSE FAILED #############␤Unable to parse named parameter; couldn't find right parenthesis at /tmp/iy0EY3M7LQ line 1:␤------> sub perl6(:M($use|$no)) {...}␤ expecting whitespace␤00:05 83m␤»
16:39 icwiener joined
[particle] std: sub perl6(:M($use)) {...} 16:39
p6eval std 24796: OUTPUT«00:05 84m␤» 16:40
16:41 cspencer_ joined 16:45 |jedai| left 16:47 |jedai| joined 16:50 eternaleye joined, maerzhase left 16:59 kidd` left 17:01 eternaleye left
masak anyone have an opinion on Str.warn? 17:02
if not, I'll consider the issue stalled for now, and move on. 17:04
cspencer_ perl6: class Foo { method list { say "bar" } }; my $f = Foo.new; $f.list 17:07
p6eval rakudo 35131: OUTPUT«too few arguments passed (0) - 1 params expected␤current instr.: 'parrot;Foo;list' pc 187 (EVAL_13:91)␤»
..elf 24796, pugs: OUTPUT«bar␤»
dwhipp Carl was concerned that Object.warn was "too much OO"
masak ooh, is that a bug I see?
cspencer_ it looks bugish
dwhipp I disagree, because p6 builtins don't default to $_ 17:08
[particle] dwhipp: carl is masak
masak dwhipp: I'm just not very eager to pollute the common classes with useless convenience methods.
dwhipp so "for @a { .warn if xxx }"
masak dwhipp: and I don't see the advantage of "message".warn over warn("message")
dwhipp: ah, hadn't thought of that. 17:09
dwhipp that's obviously true for literal strings
masak dwhipp: but in that case, it's even more important not to put .warn only on Str, because the elements of @a might be anything.
dwhipp yes, putting it on .Object seems right 17:10
[particle] yep, should be Object.warn i think
however, { 'foo' => 'bar' }.warn is funky
masak very well, I'll add it to S29 under "Object". 17:11
[particle]: why? just stringify.
[particle] well, i guess i'd write {...}.perl.warn in real code anyway
masak aye 17:13
TimToady Object.warn should just use warn($self)
and let MD handle it 17:14
17:16 arthurium2 joined
masak but that's an implementation detail, no? the Perl user just sees the method .warn and the sub warn, without suspecting that one calls the other. 17:16
TimToady yes, but if .warn doesn't have the MD semantics that warn() does, it's also a matter of interface 17:17
masak TimToady: I'm writing the signature now. does `our multi method warn (Object $o:) is export` look about right?
17:17 alester joined 17:18 arthurium2 left
masak must go, will backlog 17:19
17:19 masak left
TimToady is wondering what happens if $o is a Junction... 17:19
17:20 icwiener left
rakudo_svn r35135 | particle++ | [rakudo] fix typos in makefile template 17:20
TimToady arguably, some of these things should be in Any rather than Object
17:21 maerzhase joined
pmichaud (Any vs. Object) agreed. 17:22
TimToady we also need to carefully evaluate which forms are more basic, so that we don't end up doing SD -> MD -> SD -> MD ... 17:23
in general SD seems more basic than MD, so having SD redispatch to MD seems a bit strange, but if we make sure MD doesn't redispatch to Object/Any types, we're probably safe from cycles 17:24
17:24 kisu__ joined 17:26 kisu_ left
rakudo_svn r35136 | pmichaud++ | [rakudo]: Change find_global -> get_global. 17:30
17:32 silenius left 17:37 _smg joined 17:39 smg left
rakudo_svn r35140 | particle++ | [rakudo] refactor makefile template to create spectest_checkout and spectest_update targets; fix logic errors 17:40
r35144 | particle++ | [rakudo] add 'testable' make target to create perl6 executable, compile the test library, and update the official test suite
17:40 jferrero left 17:44 icwiener joined, cspencer_ left 17:45 icwiener left 17:46 |jedai| left 17:47 icwiener joined 17:48 |jedai| joined 17:52 mberends left 17:53 pdcawley left 17:56 azawawi joined 17:57 mberends joined
azawawi hi 17:57
18:00 simcop2387 left 18:02 masak joined
masak should I put .perl, .print, .say, .printf and .warn in Any? 18:03
pugs_svn r24797 | moritz++ | [t/spec] correct some OO tests that I broke before, masak++
TimToady masak: yes, I think so 18:05
moritz_ I find it very un-intuitive that junctions autothread through .perl 18:06
TimToady if any(@x.perl) eq '42'
moritz_ TimToady: testing for the contents of .perl is never a good idea
TimToady nevertheless, it's sufficient if a Junction stringifies to whatever .perl would have produced 18:07
[particle] timtoady: is there a syntax for me to express something like: sub perl6(:M($use|$no)) {...} 18:08
TimToady :M(:use($no))
if you just want another alias
[particle] std: sub perl6(:M(:use($no))) {...}
p6eval std 24797: OUTPUT«00:05 84m␤»
[particle] as long as i can tell which alias was used, it'll work 18:09
TimToady you can't
[particle] feh.
i was hoping i could make -M and :use work, and -/M and :no
rakudo_svn r35153 | particle++ | [rakudo] rearrange makefile for maintainability 18:10
TimToady you just want :$M, :$use, :$no in that case
[particle] yes, okay. it's a bit ugly, but i think it'll work. any objection to that syntax?
TimToady I'm not terribly attached to -M 18:12
[particle] it's hardwired in many fingers, though
-U and -N could work 18:13
TimToady what is -u now?
and why not -/u instead of -n
[particle] that could work, but --/use is ugly 18:14
-u is free
TimToady I don't think I've ever seen someone use 'no' on the command line 18:15
[particle] here's current: perlcabal.org/syn/S19.html#Synopsis
TimToady and you can always say -e 'no whatever';
[particle] -e 'no strict;'
TimToady -e '6; ...' 18:16
[particle] i agree, it's not commonly used, but somebody's gonna want it.
i'll proceed with use, and leave no for later
TimToady in general I don't think we need to add shortcuts for things that can easily be worked around by actually using the language :) 18:17
[particle] the big thing rattling around it my head now is the metasyntactic options
++COMPILER -e 'CHECK{...}' ++/COMPILER
...this means every subsystem must accept -e 18:18
TimToady note that bare -e 'CHECK {...}' laready means that
*already
-e gives you the full power of Perl 6
[particle] troo
i wonder, does ++CMD...++/CMD execute immediately? 18:19
things like that are rattling around
should crystalize (hopefully) as i write about it
dwhipp I was looking into writing tests for S09 fixed size arrays: is "for @input -> *@a[ ++$n ] { ... } legal; and will it have the semantics of iterating increasing sized chunks from @input? 18:20
TimToady anything post compile could be handled with some kind of compiler directive like CHECK
CMD is pre though
also thinking about p5 errors like "Too late for -T"... 18:21
[particle] yes
i'm used to that error on windows :)
18:21 Psyche^ joined
[particle] #!perl -T #too late! 18:21
TimToady dwhipp: I doubt it, the ++$n would be evaluated at compile time 18:22
18:22 cspencer joined
TimToady and we haven't specced that arrays can behave like slices either 18:23
(in binding) 18:24
dwhipp My ++$n was a little extreme, but I was thinking one might write code that (e.g) calculated a number of columns to format the output; and used a fixed-size array to slurp them
TimToady might be an argument for treating the type as generic and instantiated at binding time, I suppose 18:25
not sure what the performance would be though
probably abysmal...
shower & # thinking 18:26
dwhipp if it's calculated once at the start of the loop, then probably not as bad as fully dynamic
TimToady seems ill defined as to scope 18:27
bbl &
dwhipp in what way?
I guess a similar Q relates to the simple case: " my $n = 42; my @a[$n];" -- hopefully legal to sue $n in that context without needing "BEGIN" block 18:29
18:32 jferrero joined 18:33 Patterner left, Psyche^ is now known as Patterner 18:36 aindilis joined
masak dwhipp: I don't see why not. 18:45
dwhipp That was my thought ... but Larry seems concerned when we apply that same reasoning to point-block signatures 18:46
18:49 Lorn left 18:50 masak left
dwhipp I guess the limit case of the question is if the signature :( $size, *@array[$size] ) is legal 18:52
moritz_ a slurpy array with fixed size?
that looks *very* weird
dwhipp My use case was:
my $columns = 8; 18:53
for @input -> *@values[$columns] { say @values }
## prints 8 values per line
moritz_ while @input { my @cols = @input.splice(0, 8); say @cols }; # or something like that 18:54
dwhipp Yes, there's always MTOWTDI
moritz_ and I don't think there has to be a way to do everything with signatures 18:55
actually we discourage large numbers of positional arguments generally, because they are seldom useful in real world subroutines 18:56
dwhipp It's only one positional argument: it just happens to take up 8 positional slots :)
And this was a block, not a subroutine 18:57
azawawi moritz_: hi
moritz_ hi azawawi
dwhipp so no nasty semantics-at-a-distance issues
19:00 dwhipp left
TimToady declarations are evaluated at compile time, including traits 19:03
but possibly we can do some indirection if a closure is found 19:04
my @a[{$n}]
but my concern is forcing run-time instantiation of generic types, which could get expensive, esp when accidental
bbl & 19:05
azawawi moritz_: i need s// for removing ansi escape sequences... 19:11
cspencer am i able to define infix operators written in perl6 in rakudo yet? 19:15
i'm not quite sure whether it's unsupported or i've got the syntax wrong :)
also, if i define a class in perl 6 that's already been defined, is it possible to replace the original definition? 19:17
19:18 pbuetow joined
PerlJam cspencer: I don't think so (re: operators) 19:19
cspencer perljam: thanks, i'll work around it for the moment 19:20
PerlJam cspencer: wrt replacing a class, my gut says you should be able to do it, but I wouldn't know how :)
19:20 |jedai| left 19:21 |jedai| joined
moritz_ azawawi: $string.subst() might work for you 19:22
19:23 ft joined 19:31 jferrero left 19:34 fridim_ joined 19:36 DemoFreak joined
pugs_svn r24798 | particle++ | [S19] flesh out most of option reference 19:42
19:43 cspencer_ joined 19:46 |jedai| left, |jedai| joined 19:52 km2 joined
pugs_svn r24799 | particle++ | [S19] describe unchanged syntax 19:53
20:00 cspencer left 20:04 eric256 left 20:13 cspencer_ left 20:14 cspencer joined 20:19 DemoFreak left 20:24 mberends1 joined 20:25 mberends left 20:34 km2 left, km2 joined 20:36 |jedai| left, |jedai| joined 20:40 mberends1 left 20:44 spx2_ joined 20:50 spx2 left 20:51 cspencer left, alester_ joined 20:53 mberends joined 20:58 donaldh joined 20:59 |jedai| left 21:00 DemoFreak joined, |jedai| joined 21:01 DemoFreak left 21:06 km2 left, km2 joined 21:08 alester_ left 21:14 _smg is now known as smg 21:15 eternaleye joined 21:26 araujo joined 21:30 km3 joined 21:32 simcop2387 joined 21:37 jhorwitz left 21:42 km3 left 21:46 aindilis left, aindilis joined 21:47 eternaleye left 21:51 |jedai| left, |jedai| joined 21:57 km2 left 22:00 iblechbot left 22:04 kisu__ is now known as kisu 22:06 maerzhase left 22:26 wknight8111 joined 22:35 Helios- joined 22:39 |jedai| left 22:40 |jedai| joined 22:43 Limbic_Region joined 22:47 donaldh left 23:00 pbuetow left 23:01 REPLeffect left 23:07 eternaleye joined 23:09 finger joined, plainhao left, finger left 23:11 pdcawley joined
pugs_svn r24800 | azawawi++ | [S:H:P6] Fixes hilitep6 binmode STDIN issue for utf8 23:12
r24800 | azawawi++ | [S:H:P6] Updated to latest STD.pm; releasing 0.30 to CPAN
23:21 aindilis` joined 23:26 |jedai| left 23:27 |jedai| joined 23:29 brunoV left 23:32 ft left 23:35 icwiener left, aindilis left 23:36 LimbicRegion joined, Limbic_Region left, LimbicRegion is now known as Limbic_Region 23:38 eternaleye left 23:39 LimbicRegion joined 23:44 Limbic_Region left, LimbicRegion is now known as Limbic_Region
azawawi TimToady: STD.pm finally parses in Padre... Yay! Next step outline mode for comp_unit visualization (methods/subs/tokens...etc). 23:46
23:46 alester left 23:50 |jedai| left 23:51 |jedai| joined 23:55 meppl left 23:56 Exodist left 23:58 ruoso joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!