Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2008-05-24

Perl 6 | Reference Documentation | Rakudo

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:01 fridim_ joined #perl6
00:18 cmarcelo joined #perl6
00:27 Limbic_Region joined #perl6
01:15 Caelum joined #perl6
01:27 Ched- joined #perl6
01:45 justatheory joined #perl6
01:53 penk joined #perl6
02:24 pen1 joined #perl6
03:09 alexe joined #perl6
03:22 penk joined #perl6
05:06 Ched- left #perl6
06:15 TimToady joined #perl6
06:18 meppl joined #perl6
07:00 alexe joined #perl6
07:31 _REPLeffect joined #perl6
07:35 rakudo_svn r27778 | pmichaud++ | [rakudo]:
07:35 rakudo_svn r27778 | pmichaud++ | * Refactor lots of stuff dealing with Pairs, argument handling, and lists.
07:35 rakudo_svn r27778 | pmichaud++ |   The resulting code is 127 lines shorter, more correct, gets rid
07:35 rakudo_svn r27778 | pmichaud++ |   of lots of unwanted references to "infix:,", and helps get us closer
07:35 rakudo_svn r27778 | pmichaud++ |   to being able to handle list assignment and list context.
07:47 iblechbot joined #perl6
08:08 pmurias joined #perl6
08:09 pmurias spinclad: lval-accessor idiom? what it is?
08:09 pmurias * what is it?
08:09 elmex joined #perl6
08:17 vixey joined #perl6
08:32 pmurias vixey: would you prefer the ast dumping to use named arguments
08:32 pmurias ?
08:33 vixey I'm not sure
08:36 vixey pmurias: is that a STD_red thing?
08:38 pmurias no, i mean the Ast elf prints out when -v is called
08:39 pmurias it's done in p6
08:40 pmurias by Ast i refered to IRx1
08:43 vixey I wasn't able to figure out if STD_red or IRx1_FromAST or whichever stage post if was lost
08:43 vixey but it doesn't seem to be around in the -v output
08:50 pmurias by named params i meant the ast being dumped as CompUnit(statements=>[...],filename=>undef) instead of CompUnit([...],undef)
08:50 vixey oh that sounds good
09:03 qmole joined #perl6
09:20 penk left #perl6
10:03 moritz_ rakudo: my $x = sub { say "foo" }; $x();
10:03 exp_evalbot OUTPUT[foo␤]
10:22 moritz_ rakudo: my &x = sub { say "foo" }; x()
10:22 exp_evalbot OUTPUT[foo␤]
10:22 moritz_ rakudo: my &x = 1; x()
10:22 exp_evalbot OUTPUT[invoke() not implemented in class 'Integer'␤current instr.: '_block11' pc 39 (EVAL_11:20)␤
10:23 moritz_ rakudo: my @x = 1; say "still there"
10:23 exp_evalbot OUTPUT[still there␤]
10:28 chris2 joined #perl6
10:38 jferrero joined #perl6
10:42 b_jonas joined #perl6
11:08 moritz_ @tell masak I made an article for my homepage, not a perlmonks meditation; anyway, here it is: http://perlgeek.de/en/articl​e/mutable-grammar-for-perl-6 - comments welcome
11:08 lambdabot Consider it noted.
11:13 ludan joined #perl6
11:14 ludan hi
11:14 moritz_ hi ludan ;)
11:22 fullermd moritz_: "<[A..Z> <[A..Z0..9_]>*"
11:22 fullermd Missing a ] there?
11:22 moritz_ fullermd: yes, thanks
11:22 Psyche^ joined #perl6
11:25 moritz_ fullermd: is the article understandable?
11:26 fullermd Well, I haven't finished it; just grabbed a minute or two away from work to glance at it.
11:26 moritz_ ok
11:26 fullermd The proto regex section is a little greek, but it feels like it'll yield to a more careful reading.
11:27 moritz_ perhaps I'll find a simpler/more verbose explanation
11:27 vixey moritz cool!
11:29 fullermd May have time later to go through it with more brain cells committed.
11:48 rindolf joined #perl6
11:53 pugs_svnbot r20515 | pmurias++ | [elf] removed a unused hack
11:53 pugs_svnbot r20515 | pmurias++ | the ast dump uses named arguments
12:03 ruoso joined #perl6
12:06 rakudo_svn r27779 | jonathan++ | [rakudo] A few updates to Mutable PMC, and subclass it for Scalar, Array and Hash containers.
12:06 rakudo_svn r27780 | jonathan++ | [rakudo] Tests for Mutable PMC.
12:28 iblechbot joined #perl6
12:38 Patterner <error revision="20515"> Generating precompiled Prelude, Math::Basic... pugs: Internal error: Invalid grammatical category: "Bool" </>
12:39 moritz_ Patterner: known error, nobody managed to fix it so far
12:53 jferrero left #perl6
13:00 lisppaste3 joined #perl6
13:31 kane_ joined #perl6
13:31 pmurias moritz_: why did you use a letter for a hypotetical new sigil?
13:32 b_jonas there was some old joke with internationalizing perl by replacing keywords with french variants and "$" with "Fr"
13:34 rindolf b_jonas: £
13:34 rindolf b_jonas: or ¥
13:34 rindolf There's also the ILS sign.
13:36 pmurias moritz_: it's really akin to using token sigil:sym<x> { <sym> } with x being easier to type
13:37 Patterner Freedom Sigil!
13:40 FurnaceBoy joined #perl6
13:49 pmurias vixey: is there anything in elf which particulary annoys you?
13:49 pmurias * particularly
13:51 vixey pmurias: The link from STD_red to IRx1 is not clear to me
13:52 pmurias STD_red emitts the match object as perl5 code which is eval'ed by Parser.pm
13:53 pmurias IRx1_FromAST (created by IRx1_FromAst_create) then turnes that into IRx1
13:54 pmurias * emits
14:00 pmurias vixey: clearer now?
14:00 b_jonas rindolf: it was before the introduction of EUR
14:00 rindolf
14:01 vixey pmurias: cool ty, I might try to get post if working when I'm back home
14:02 b_jonas let's use a funny space like as a silgil, e.g.  foo -- no-one would notice that
14:02 b_jonas or, as p6 already has @@ as a silgil, what about %%
14:02 b_jonas or &&
14:05 hsv-al joined #perl6
14:05 hsv-al left #perl6
14:06 alc joined #perl6
14:08 moritz_ pmurias: I wasn't aware that ¢ is a letter actually ;)
14:09 pmurias ł is a letter
14:09 moritz_ ah, that
14:10 moritz_ I randomly typed some charcates and took the weirdest-looking ;)
14:13 pmurias i have an ł in my name
14:18 pugs_svnbot r20516 | moritz++ | [fudge] add lives_ok and eval_lives_ok to fudge, bacek++
14:19 b_jonas cent certainly isn't a letter
14:23 pmurias it's much easier to pronounce too
14:25 jhorwitz joined #perl6
14:36 justatheory joined #perl6
14:46 demq joined #perl6
14:48 pugs_svnbot r20517 | moritz++ | [spec] fudged S29-list/{grep,sort}.t, patch curtesy of bacek++
15:44 alester_ joined #perl6
16:05 rindolf joined #perl6
16:14 wolverian hm. why is there /usr/bin/X11/perl on feather? just curious.
16:15 moritz_ what's the output of 'dpkg -S /usr/bin/X11/perl'?
16:16 wolverian not found
16:16 wolverian that's why I asked :)
16:16 wolverian someone *did* install 5.10 into /usr/local before debian got it in a package, perhaps it's a leftover from there (but why is it not in local?)
16:26 pmurias joined #perl6
16:27 pmurias wolverian: i have a perl in /usr/bin/X11/ on a gentoo box too
16:27 moritz_ that's really weird if you ask me
16:28 wolverian I don't have one on ubuntu 8.04 server
16:30 pmurias moritz_: it's a link to ../bin
16:33 pmurias on feather it's a symbolic link to /usr/bin
16:33 moritz_ ok
16:35 Juerd It doesn't appear to be installed by a debian package either
16:37 moritz_ systems tend to collect garbage unless cared for very thouroughly
16:37 wolverian it is? ls -l doesn't show it is..
16:38 pmurias wolverian: depends how you call it
16:38 pmurias ls -ld dosn't
16:38 wolverian how's that?
16:39 pmurias but ls -l /usr/bin/ | grep X11 does
16:39 moritz_ that's weird
16:39 moritz_ isn't that even a bug in ls?
16:39 pmurias likely
16:40 polettix joined #perl6
16:43 b_jonas no, how it quotes characters and how it columnates and colors CAN depend on whether the output is a terminal,
16:43 b_jonas but you can override all of these via options
16:43 moritz_ -d, --directory
16:43 moritz_ list directory entries instead of contents, and do not  derefer‐
16:43 moritz_ ence symbolic links
16:44 b_jonas uh
16:44 wolverian oh, *X11* is a symbolic link, not X11/perl
16:44 moritz_ it's all documented
16:44 b_jonas sure
16:44 moritz_ yes
16:44 wolverian that was confusing :)
16:44 b_jonas you added -d
16:45 moritz_ ah right, ls -l X11 shows that it's a link
16:45 moritz_ but ls -l $normal_dir shows $normal_dir's contents
16:46 moritz_ that's inconsistent
16:46 moritz_ (but documented)
16:46 pmurias it's not documented on my ls version
16:48 moritz_ uhm, you're right
16:48 pmurias so it could be a documented bug
16:51 ewilhelm joined #perl6
16:51 Psyche^ joined #perl6
17:47 alester_ joined #perl6
17:48 alester_ joined #perl6
17:50 rakudo_svn r27783 | jonathan++ | [rakudo] Add more tests to the known passing list. Patch courtesy of moritz++.
17:51 moritz_ YaY, karma in two channels ;)
17:51 moritz_ rakudo patches really pay off ;)
17:54 slavi3 joined #perl6
17:54 rakudo_svn r27784 | jonathan++ | [rakudo] Update makefile to build the scalar, array and hash PMCs that subclass Mutable.
17:54 slavi3 where can I go if I I am having trouble building pugs?
17:55 mncharity joined #perl6
17:55 slavi3 Generating precompiled Prelude, Math::Basic... pugs: Internal error:Invalid grammatical category: "Bool" Please file a bug report.
17:57 mncharity stevan_: re 6/3, neat.  lunch or somesuch?
17:59 moritz_ slavi3: pugs is known to be broken. Try an older version with ghc 6.6.1
17:59 moritz_ slavi3: more specifically r19915 (from svn) works for me
18:00 slavi3 I see, will have to bother later then
18:00 slavi3 ty
18:00 slavi3 left #perl6
18:04 rindolf joined #perl6
18:19 braceta joined #perl6
18:27 pmurias mncharity: hi
18:30 mncharity hi pmurias
18:30 mncharity have elf_f working again.  running regression tests.
18:31 rakudo_svn r27787 | jonathan++ | [rakudo] Turns out that calling a PMC Scalar isn't such a good idea. :-)
18:31 pmurias mncharity: is the hack pretty ;) ?
18:32 mncharity re 'what to do', how about attribute defaults for Faster?  has $.x = 3;   It's also used to initialize array/hash, eg, has @.a;
18:34 mncharity re pretty, not so much.  but needed things working again. ;)
18:34 pugs_svnbot r20518 | pmurias++ | [elf] -v uses Perl::Tidy
18:35 mncharity realized an lvalue sub could *also* support the (argument=>assignment) api.  with ~0 cost over lvalue.
18:35 mncharity re Perl::Tidy, I thought kp6 found PT was a major performance hit?
18:35 pmurias only on -v
18:37 pmurias if you prefer to get badly formated code faster i can make it dependend on an env var or something
18:38 mncharity re -v, ah, ok :)
18:40 mncharity err, probably want emitter1 rather than emitter0.
18:42 pmurias yes
18:43 pmurias have you seen the improved -v output?
18:43 mncharity not yet
18:46 pugs_svnbot r20519 | putter++ | [elf] elf_f works again.  run-tests.result_faster updated.
18:50 pmurias mncharity: (argument=>assignment) api, explain please
18:50 mncharity re -v output... ++avoids having to look at the node fields; --much more verbose, so harder to eyeball large chunks of code.  a good tradeoff.  might want both at some point.
18:52 spinclad pmurias: re lval-accessor:  accessor in lval position, $foo.part = newval  (as opposed to accessor-as-setter, $foo.part(newval) ).
18:53 spinclad (just names, no new content)
18:53 pmurias i know that, implemented that recently
18:54 mncharity re tidy, any needed p5 code should go into one of the preludes.  eval_perl5 should not appear elsewhere.  else it stops looking like a bootstrap.
18:54 mncharity re api,
18:56 spinclad moritz_: s/(changes )too( great care)/took/   # had to read this a few times to reparse and repair
18:57 moritz_ spinclad: thank, fixing now
19:00 cognominal_ joined #perl6
19:00 pmurias mncharity: could you continue please ;)
19:03 FurnaceBoy joined #perl6
19:05 pugs_svnbot r20520 | putter++ | [elf] elf_f_faster again uses lvalue subs.
19:05 mncharity sorry, got sucked into code.
19:05 pmurias np
19:06 Psyche^ joined #perl6
19:06 mncharity basically, Faster accessors can now use both the elf_f and elf_f_faster "assign to accessor method call" apis.
19:06 mncharity hmm.... now is that really needed anymore....
19:06 mncharity no
19:07 pmurias bo
19:07 pmurias * no
19:07 pmurias was working on the same thing
19:07 pmurias can i have a try?
19:08 mncharity re no, err, empirically yes.  not sure why.  my code is in.  re try, feel free.  I'll move on..
19:15 * pmurias would bet elf self-compilation is slower now
19:18 mncharity lvalues?
19:20 * pmurias hopes not
19:20 mncharity why slower?
19:20 mncharity or "observed to seem slower"?
19:21 spinclad not observed, expected?
19:22 mncharity re lvalues/accessor assignment, unless p5 is nonlocally pessimizing upon seing them, it doesn't seem likely.  my impression is they are not much used in elf.
19:22 pmurias used to self compile in 10sec in the old days, now 20
19:22 mncharity parser cache?
19:22 pmurias maybe just more code
19:22 moritz_ pmurias: well, but the code size increased as well, right?
19:23 pmurias it's 40 withough the cache
19:23 mncharity is the 20 repeatable?  ie, it's not just parser cache refresh?
19:24 mncharity re "old days", oh, not a recent step-function change.  hmm...
19:25 mncharity the elf_f startup times really tempt me to avoid regressing testing elf_f :/
19:26 mncharity hmm, maybe time to rename the elf_f executable elf_f_simple or some such, to avoid ambiguity.
19:26 moritz_ mncharity: can't you do many parses with one startup? (to get many tests parse)
19:26 mncharity the code size shouldn't have changed much.
19:27 moritz_ or is it the compied stuff that takes so long to execute?
19:27 pmurias maybe the machine slowdowned :(
19:29 mncharity moritz_: most tests still go boom.  hmm... might add a --avoid-going-boom-by-wrapping-file-in-eval argument... or, to help distinguish file results, perhaps a --ugly-speed-kludge-cooperating-with-run-tests argument... eep
19:31 mncharity pmurias: could choose a fixed test object which could be compiled the whole time, time "old days" and current elf_f_faster, and if there's a difference, go looking inbetween.
19:32 moritz_ bisecting?
19:34 mncharity hmm... not sure the new IR layout is actually a worthwhile tradeoff in practice...
19:34 mncharity maybe if it got split up onto separate lines like the Match output.
19:35 pmurias do you have Perl::Tidy?
19:36 pmurias it's supposed to formating
19:37 mncharity looking...
19:38 mncharity ah, no.
19:39 mncharity so when the eval_perl5 gets moved into the prelude, PT should be added to the dependencies so things die cleanly if its missing.
19:40 mncharity better simplicity of behavior than robustness in its absence.
19:40 mncharity s/its/tolerating its/
19:40 pmurias only speced things should be in the prelude
19:41 mncharity no.
19:41 mncharity the prelude is guided more by what degree of p5 independentness a bit of code has.
19:42 mncharity specific p5 api dependence -> Emit; generic p5 backend -> PrimitiveP5; p6 -> Prelude.pm.  spec'ness isn't an issue.
19:43 mncharity once rakudo can run p6 prelude, we may start working on a combined prelude.  but if so, that will be a separate thing.
19:44 mncharity ooooo, tidy... pmurias++
19:47 mncharity we aspire to specness... but basically, there is no existent compiler which can handle even a few lines of a real spec prelude.  that's been our long-term problem. :/
19:47 ting joined #perl6
19:48 agentz1 joined #perl6
19:50 pmurias while the PT adapter could be moved to a p5 stuff file it shouldn't be in the prelude
19:51 pmurias the prelude contains things which are part of the language
19:52 pmurias and does eval_perl5 hurt in the perl5 emitter?
19:52 pugs_svnbot r20521 | pmurias++ | [elf] the accessor hack is conditional
19:54 mncharity we can call it something other than prelude if you like.  but all p5 has to be localized or it becomes rather less a bootstrap.  the only two places with p5 at the moment are Emit's prelude string, and PrimitivesP5.
19:56 pmurias ok, it will matter when we have a non-p5 backend
19:57 mncharity indeed.  one of the stronger arguments for starting a js or CL backend now is it will help force us to stay honest.
19:58 pmurias do you now a good CL book?
19:58 mncharity http://www.lisp.org/HyperSpec/​FrontMatter/Chapter-Index.html ?
19:58 lambdabot Title: CLHS: Chapter Index
19:59 mncharity http://www.cs.cmu.edu/Groups​/AI/html/cltl/clm/node1.html
19:59 lambdabot Title: Contents
20:00 pmurias would non-spec things having a HACK prefix be a good idea?
20:00 mncharity http://www.gigamonkeys.com/book/
20:00 lambdabot Title: Practical Common Lisp
20:01 pmurias you have all of them?
20:01 mncharity they are all online :)
20:01 mncharity and free
20:01 mncharity CL++
20:02 mncharity haven't read Practical Common Lisp, but it's been well reviewed.
20:03 pmurias my only experience with a () style language is scripting the gimp
20:04 mncharity re HACK, basically I expect at most a stray fragment or two of code might survive from now until a spec-compliant p6.  even getting to pugs-level capability will involve quite a bit of churn.  so basically there's is a HACK prefix, attached to the entire codebase.  and things are just slowly made more test-passingly hackish.
20:05 mncharity javascript?
20:05 ewilhelm left #perl6
20:06 pmurias did an AJAX app
20:08 pmurias i can write js fairly fluently
20:09 pmurias the JavaScript: The Definitive Guide is a good book
20:10 mncharity be nifty to have a js backend.  with better analysis and increasingly optimized spidermonkey, it seems it would need not even be slow like PIL2JS was.
20:11 pmurias would a CL backend be much faster then the p5 one?
20:12 mncharity very much so.  at least for compilation, like elf's currently, which is focused more on performance than p6 correctness.
20:13 moritz_ maybe you could win avar for a CL backend? he said he'd work on one for kp6 as soon as it was bootstrapped
20:13 mncharity there was a cl backend for kp6.  might look at that.  i assume it has some of the "primitives p5 has but cl doesn't" code which could just be copied.
20:14 mncharity :)
20:15 mncharity re things to do, another is to replace cruft like  $.var.sigil eq '$'
20:16 mncharity in Emit with methods in/added-to Analysis.  eg, is_scalar.
20:17 mncharity basically Emit should move gradually towards being a clean explanation of how to compile p6 on a p5 like language.  with Analysis providing the analysis and vocabulary to let that happen.
20:17 pmurias should i remove non speced non used stuff as each_with_index?
20:18 mncharity perhaps comment it out, rather than removing?
20:18 * mncharity tries to remember what the p6 idiom is for each_with_index...
20:19 pmurias to tempt the next person into using non-speced stuff ? ;)
20:20 pmurias prototypes on methods don't make any sense
20:20 moritz_ mncharity: for @list Z 0 .. * -> $item, $index { ... }
20:21 moritz_ pugs: for <a b c>.kv -> $a, $b { say "$a|$b" }
20:21 exp_evalbot OUTPUT[*** Not a keyed value: VList [VStr "a",VStr "b",VStr "c"]␤    at /tmp/MTL59ZMmCL line 1, column 5-16␤]
20:21 cihhan joined #perl6
20:21 mncharity hmm... with a $500 multicore machine,  make test_faster would only take a couple of minutes...
20:22 mncharity re tempt, so if someone needs and goes looking for the capability, it's only an uncommenting away.
20:22 Ingmar joined #perl6
20:23 cihhan sorry I am new to Perl but I'd like to ask if tehre is any way to execute an exe and then read the outputs of this exe?
20:24 moritz_ cihhan: wrong channel, here is perl 6 compiler development
20:24 cihhan ooohh, so sorry
20:24 moritz_ cihhan: try #perl or #perlhelp or some channels on irc.perl.org
20:24 cihhan thank u very much
20:24 cihhan good luck on ur works
20:24 cihhan :)
20:24 pmurias ;)
20:24 moritz_ thanks ;)
20:25 mncharity for @list Z 0 .. * { ...$^each...$^index... }   err, that can't be the right placeholder variable format, can it?
20:25 cihhan left #perl6
20:27 pmurias that remindes me of $^whiteboard
20:28 moritz_ why not?
20:28 moritz_ for <a b c d e f> -> { say $^b, $^a; }
20:28 moritz_ pugs: for <a b c d e f> -> { say $^b, $^a; }
20:28 exp_evalbot OUTPUT[*** Undeclared variable: ("$^b",MkPad (padToList [("$_",PELexical {pe_type = (mkType "Scalar"), pe_proto = <Scalar:0xb679342c>, pe_flags = MkEntryFlags {ef_isContext = True}, pe_store = <ref:0xb67932f4>}),("@_",PELexical {pe_type = (mkType "Array"), pe_proto = <Array:0xb67932c8>, pe_flags =
20:28 exp_evalbot ..MkEntryFlags {ef_isContext = False}, pe_store = ...
20:28 pjcj joined #perl6
20:29 moritz_ rakudo: for (1, 2, 3, 4) -> { say $^b, $^a; }
20:29 exp_evalbot OUTPUT[21␤43␤]
20:29 moritz_ rakudo++
20:30 mncharity pmurias: todo: add positional analysis to Analysis. :)
20:30 mncharity *positionals
20:31 pmurias mncharity: each_with_index in p6 is @a:kv
20:34 pmurias mncharity: should i write the todo's down?
20:36 Eevee rakudo: for (1, 2, 3) -> { say $^b, $^a; }
20:36 exp_evalbot OUTPUT[21␤StopIteration␤current instr.: '_block11' pc 65 (EVAL_11:28)␤
20:38 mncharity writing todos would be great.  the "what are some easy things for people to work on" idea
20:38 mncharity README is almost empty in elf_f_src
20:39 mncharity @a:kv not @a.kv ?
20:39 pmurias yes
20:40 moritz_ pugs: for <a b c d>:kv -> $a, $b { say "$a|$b" }
20:40 exp_evalbot OUTPUT[*** ␤    Unexpected ":kv"␤    expecting term postfix, operator or block construct␤    at /tmp/87QA4HWkND line 1, column 14␤]
20:40 moritz_ pmurias: what's the syntax called? or where can I read more about it?
20:41 pmurias S09:527
20:41 pmurias right to using string as array indexes
20:41 moritz_ pmurias: thanks, looking
20:41 pmurias * next to
20:42 moritz_ ah, it's an adverb
20:42 moritz_ it'll take some time for me to get used to it ;)
20:44 mncharity btw, elf ignoring :todo's is intentional.  so some Test tests will fail indefinitely.
20:44 pmichaud in the suite I think we're planning to avoid :todo in favor of fudge anyway
20:44 mncharity re get used to, oh, yeah.  I keep on thinking "some day I should actually try to learn p6". ;)
20:45 pmichaud (assuming that was in reference to the suite)
20:45 mncharity was, thanks :)
20:47 mncharity also planning on avoiding fudge, at least until test coverage is pugs-like broad, and regression avoidance and test-loss is critical.
20:47 pmichaud that works.
20:47 mncharity * test-loss avoidance
20:50 mncharity pmurias: something for the todo list, range op ..
20:51 mncharity err, no, hmm, why didn't that work...
20:52 pmurias what?
20:54 pmurias mncharity: the match tree looks much more readable withought the text position and the matched text but looses some info
20:57 mncharity re what?, turned out to be a $n<x> to $n.x regression.  I should probably do a code walk to look for other such.  :/  ah well.
20:57 mncharity re tree readability, best bet is to do a string clipper.  ie, any string longer than n (say 30 chars) gets the middle clipped out of it and replaced by ... .
20:58 mncharity or have two different dumper formats
20:59 mncharity 60 chars?
20:59 mncharity also can do s/\n/\\n/ which makes it more compact
20:59 pmurias mncharity: two formats might be sensible
21:00 mncharity more compact, but less useful for debugging, so yeah, perhaps two formats.  we'll probably need to decompose -v at some point.  -v -v2.  -vSomeParticularThingToBeVerboseAbout.  somesuch
21:01 pmurias the char pos & orginal string is usefull for debugging parsing
21:02 pmurias otherwise it's a hindrance
21:05 mncharity err, when would one look at a Match tree except when debugging parsing, or looking a the Match result from a regex?
21:05 pmurias ast->ir
21:06 pmurias if the match is correct it's obvious which part belongs to what
21:07 pmurias i mainly look at the Match out of curiosity
21:08 mncharity ah, ok.
21:08 mncharity what did you think of showcode?
21:08 mncharity apropos ast->ir
21:09 pmurias i'll look at it again
21:10 pmurias i don't like pink
21:10 pmurias the left pane turnes left when i click on it
21:14 mncharity re pink, :)  could switch to external css. :)
21:14 mncharity "left pane turnes left", huh?
21:15 mncharity or just change the color :)
21:18 pmurias it turns red
21:19 mncharity I don't understand...
21:21 pmurias the text in the leftmost column turnes red permanently
21:22 mncharity oh, I see.  "Match's clicked on in the left window turn red, but isn't changed back when you cli..."  what you said.
21:22 pmurias yes
21:23 mncharity yeah.  never got as far as dealing with that.  I'm wondering what you think more "big picture" rather than "details'.
21:24 * mncharity crosses fingers, runs t/ yet again...
21:24 pmurias can't think of a real use for it
21:24 pmurias mncharity: have you changed anything significant?
21:25 mncharity you mean since it was created? no
21:25 mncharity re use, ah, ok
21:26 mncharity primary motivation was to have a way to show non-local analysis info... but it doesn't do that yet so...
21:26 pmurias it could used for exploring a file and seeing what it compiles to
21:28 pmurias shower&
21:28 mncharity yeah.  Emitter would have to add (lots of) position markers.
21:29 mncharity maybe could wrap Emitter methods.
21:40 pmurias the thing it you be usefull for is finding compiler errors in big pieces of code
21:41 pmurias checking if the compiler compiles the area near the assumed bug correctly
21:42 pmurias but it would need emitter integration
21:51 mncharity *nod
21:53 pmurias it could also be used as part of a "run elf in your browser and see how cool it is thingy"
21:54 pmurias but that sort of things don't seem a worthwhile time use
21:58 mncharity indeed
21:58 * mncharity fights with svn... svn winning... :/
22:02 mncharity weary yay.
22:03 Psyche^ joined #perl6
22:03 pugs_svnbot r20522 | putter++ | [elf] Better hash/array/pair literal support.  run-tests.result_faster updated.  And tweaks.
22:04 mncharity perl6: +"foo"
22:04 exp_evalbot kp6: OUTPUT[syntax error at position 0, line 1 column 0:␤+"foo␤^ HERE␤]
22:04 exp_evalbot ..pugs: RESULT[0.0]
22:04 exp_evalbot ..rakudo: RESULT[Method 'perl' not found for invocant of class 'Integer'␤current instr.: '_block11' pc 38 (EVAL_11:18)␤called from Sub 'parrot;PCT::HLLCompiler;eval' pc 783 (src/PCT/HLLCompiler.pir:458)␤called from Sub 'parrot;PCT::HLLCompiler;evalfiles' pc 1065 (src/PCT/HLLCompiler.pir:587)␤called
22:04 exp_evalbot ..from Sub 'parrot;PCT::HLLCompiler;command_line' pc 1244 (...
22:04 exp_evalbot ..elf: RESULT[syntax error at (eval 11) line 3, near "GLOBAL::statement_prefix:"␤ at ./elf_f_faster line 4371␤]
22:04 moritz_ perl6: say +"foo";
22:04 exp_evalbot kp6: OUTPUT[␤]
22:04 exp_evalbot ..pugs: OUTPUT[0␤]
22:04 exp_evalbot ..rakudo: OUTPUT[0␤]
22:04 exp_evalbot ..elf: OUTPUT[foo␤]
22:05 mncharity ah, tnx
22:11 mncharity nobody warned. :/
22:11 moritz_ mncharity: I guess that's because warnings aren't really specced all that much (and tested, for that matter)
22:13 pmurias mncharity: it might usefull to do seperate compilation before attempting the CL backend to cut compile times
22:17 mncharity because of sbcl?  it's fairly fast.
22:18 pmurias i'll get to sleep now, as i'm only rambling and yawing ;)
22:18 pmurias g'night
22:18 moritz_ good night
22:20 moritz_ rakudo: say 4.perl
22:20 exp_evalbot OUTPUT[4␤]
22:20 vixey joined #Perl6
22:20 moritz_ then why did it say Method 'perl' not found for invocant of  class 'Integer
22:20 moritz_ earlier on?
22:21 mncharity good night pmurias
22:21 moritz_ ah, something returned and Integer, not an Int
22:21 moritz_ rakudo: say  (+"foo").WHAT
22:21 exp_evalbot OUTPUT[Int␤]
22:21 mncharity Integer != Int ?
22:21 moritz_ intproto = p6meta.'new_class'('Int', 'parent'=>'Integer Any')
22:22 moritz_ rakudo: my Integer $x = 3; say $x.WHAT
22:22 exp_evalbot OUTPUT[Int␤]
22:22 moritz_ rakudo: my Integer $x = 3; say $x.perl
22:22 exp_evalbot OUTPUT[3␤]
22:22 mncharity rakudo: my Int $x = 3; say $x.WHAT
22:22 exp_evalbot OUTPUT[Int␤]
22:23 mncharity rakudo: my Int $x = 3; say $x.perl
22:23 exp_evalbot OUTPUT[3␤]
22:24 mncharity pugs: +{3=>4,5=>6}
22:24 exp_evalbot OUTPUT[*** ␤    Unexpected end of input␤    Closing hash curly may not terminate a line;␤    please add a comma or a semicolon to disambiguate␤    at /tmp/Y4YMUi0HGt line 2, column 1␤]
22:24 mncharity pugs: my %h = (3=>4,5=>6); say +%h
22:24 exp_evalbot OUTPUT[2␤]
22:27 mncharity /me so wants to do  multi prefix:<+> as a real multi... but just don't want a dependency on mmd working yet... :/
22:38 moritz_ mncharity: for now you can just make prefix:<+>($x) call $x.Num, that's also correct
22:52 jisom joined #perl6
22:57 mncharity re .Num, oh, neat.  tnx.
23:00 mncharity perl6: +(:a<3>)
23:00 exp_evalbot kp6: OUTPUT[no method 'APPLY' in Class 'Undef'␤ at compiled/perl5-kp6-mp6/lib/Ki​ndaPerl6/Runtime/Perl5/MOP.pm line 345␤   KindaPerl6::Runtime::Perl5::MO​P::__ANON__('HASH(0x824bdac)', 'APPLY', 'HASH(0x8e5335c)') called at compiled/perl5-kp6-mp6/lib/Ki​ndaPerl6/Runtime/Perl5/MOP.pm line 169␤
23:00 exp_evalbot ..main::DISPATCH('HASH(0x824bdac)', 'APPLY', 'HASH(0x8e5335c)') ca...
23:00 exp_evalbot ..pugs: RESULT[2]
23:00 exp_evalbot ..rakudo: OUTPUT[Null PMC access in type()␤current instr.: 'parrot;PAST::Compiler;keyed' pc 6062 (src/PAST/Compiler.pir:1694)␤called from Sub 'parrot;PAST::Compiler;post_children' pc 1671 (src/PAST/Compiler.pir:362)␤called from Sub 'parrot;PAST::Compiler;call' pc 2766
23:00 exp_evalbot ..(src/PAST/Compiler.pir:780)␤called from Sub 'parrot;PAST::Compiler;post_children' pc 167...
23:00 exp_evalbot ..elf: OUTPUT[Unquoted string "a" may clash with future reserved word at (eval 10) line 3.␤]
23:04 mncharity ok, let's see...
23:05 mncharity int is Int ?
23:05 mncharity err, no.  int does Int ?
23:06 mncharity and what is Integer ?
23:10 mncharity what is the relationship between Bit Int and Num?  Do they have a common role other than Any?
23:13 mncharity pugs: my $x = :a<3>; say $x.WHAT; say +$x;
23:13 exp_evalbot OUTPUT[Pair␤2␤]
23:13 mncharity rakudo: my $x = :a<3>; say $x.WHAT; say +$x;
23:13 exp_evalbot OUTPUT[Null PMC access in type()␤current instr.: 'parrot;PAST::Compiler;keyed' pc 6062 (src/PAST/Compiler.pir:1694)␤
23:22 mncharity sigh.  neither roles nor recursive inheritance in p5. :/
23:32 mncharity rakudo: Any.can('foo')
23:32 exp_evalbot OUTPUT[Method 'can' not found for invocant of class ''␤current instr.: '_block11' pc 35 (EVAL_11:18)␤
23:33 mncharity rakudo: Object.can('foo')
23:33 exp_evalbot OUTPUT[Method 'can' not found for non-object␤current instr.: '_block11' pc 35 (EVAL_11:18)␤
23:42 pugs_svnbot r20523 | putter++ | [elf] STRING now inherits from Str, etc.  .Num added to Prelude.  prefix:<+> calls .Num .  Emit of Pair() tweaked.

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

Perl 6 | Reference Documentation | Rakudo