Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2007-09-17

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:05 theorb joined #perl6
00:24 justatheory joined #perl6
00:33 dduncan joined #perl6
00:36 thestarslookdown joined #perl6
01:41 agentzh joined #perl6
01:51 araujo joined #perl6
02:13 weinig joined #perl6
02:14 weinig joined #perl6
02:39 tflorez joined #perl6
02:41 Helios- joined #perl6
02:42 nelhage joined #perl6
02:43 spinclad joined #perl6
03:03 toshinori joined #perl6
03:05 elmex_ joined #perl6
03:24 idiotmax joined #perl6
03:25 [particle] joined #perl6
03:26 jisom joined #perl6
03:38 c9s joined #perl6
04:48 dduncan joined #perl6
05:02 daxim joined #perl6
05:12 xinming_1 joined #perl6
05:37 elmex joined #perl6
05:40 amnesiac joined #perl6
05:44 moritz_ moin
05:44 pugs_svnbot r17874 | moritz++ | cross-metaop.t: removed redundant, broken smartlink
05:44 pugs_svnbot diff: http://dev.pugscode.org/changeset/17874
05:44 lambdabot Title: Changeset 17874 - Pugs - Trac
05:47 pugs_svnbot r17875 | moritz++ | fixed smartlink in t/regex/from_perl6_rules/qinterp.t
05:47 pugs_svnbot diff: http://dev.pugscode.org/changeset/17875
05:47 lambdabot Title: Changeset 17875 - Pugs - Trac
05:47 moritz_ the last two tests with broken smartlinks need more looking at
05:53 pyer joined #perl6
05:54 pyer left #perl6
05:55 IllvilJa joined #perl6
05:59 devogon joined #perl6
06:10 pdcawley joined #perl6
06:12 pdcawley joined #perl6
06:12 Aankhen`` joined #perl6
06:40 BinGOs joined #perl6
07:00 moritz_ @seen ferreira
07:00 lambdabot I saw ferreira leaving #perl6 2d 10h 41m 58s ago, and .
07:39 dduncan left #perl6
07:41 the_dormant joined #perl6
07:46 iblechbot joined #perl6
08:01 cognominal_ joined #perl6
08:09 franck___ joined #perl6
08:11 jisom joined #perl6
08:14 the_dormant left #perl6
08:37 autark joined #perl6
08:45 pugs_svnbot r17876 | moritz++ | ext/Recurrence: removed unneccessary skip in t/recurrence-grep.t
08:45 pugs_svnbot diff: http://dev.pugscode.org/changeset/17876
08:45 lambdabot Title: Changeset 17876 - Pugs - Trac
08:51 andara joined #perl6
09:04 ruoso joined #perl6
09:10 drupek12102 joined #perl6
09:18 riffraff joined #perl6
09:26 drrho joined #perl6
09:43 Psyche^ joined #perl6
09:48 IllvilJa joined #perl6
09:48 cognominal_ joined #perl6
09:49 cognominal__ joined #perl6
10:00 Blwood joined #perl6
10:18 ahmadz joined #perl6
10:19 ahmadz hi all
10:19 moritz_ hi ahmadz
10:20 ahmadz http://feather3.perl6.nl/runpugs/ ;-)
10:20 lambdabot Title: Run Perl 6 Now -- in your browser!
10:21 moritz_ Sorry, the pugs server is not running.
10:21 ahmadz try again... i restarted it
10:22 moritz_ it appears to have problems with control chars
10:22 moritz_ I entered 'my @a = <foo bar 1 2 3>' and hit return...
10:23 moritz_ what I got is this:
10:23 moritz_ pugs>
10:23 moritz_ pugs> my
10:23 moritz_ pugs>
10:23 moritz_ even the input line disappeared from the terminal
10:23 ahmadz hmmm
10:23 moritz_ tested with firefox 2.0.0.something on linux
10:23 ahmadz let me check
10:24 ahmadz http://feather.perl6.nl:8096/runpugs/
10:24 pugs_svnbot r17877 | fglock++ | [kp6] token-sym clean up
10:24 pugs_svnbot diff: http://dev.pugscode.org/changeset/17877
10:24 lambdabot Title: Run Perl 6 Now -- in your browser!
10:24 lambdabot Title: Changeset 17877 - Pugs - Trac
10:24 ahmadz feather3 seems to have a configuration issue with IO::Pty
10:25 ahmadz feather.perl6.nl is working correctly
10:25 pugs_svnbot r17878 | fglock++ | [kp6] added multi token test
10:25 pugs_svnbot diff: http://dev.pugscode.org/changeset/17878
10:25 lambdabot Title: Changeset 17878 - Pugs - Trac
10:25 ahmadz moritz: ah html escaping problem
10:27 ahmadz moritz: after ironing out these bugs, i should work on the tutorial part...
10:30 moritz_ ahmadz: perl 6 or runpugs tutorial? or combined?
10:30 ahmadz im thinking of using http://jnthn.net/papers/2007-ukuug-perl6programming-slides.pdf as a basis for the examples for the tutorial
10:30 ahmadz moritz: perl6 tutorial
10:31 ahmadz moritz: along with a brief runpugs/pugs command features...
10:31 moritz_ ahmadz: great idea
10:32 cognominal_ joined #perl6
10:32 ahmadz any ideas are welcome on runpugs
10:32 moritz_ kp6 backend - but you know that idea ;)
10:32 Juerd ahmadz: Shall I ask Audrey to point run.pugscode.org to feather3?
10:34 ahmadz juerd: hi, no not yet... we're having problems running it on feather3
10:35 ahmadz moritz_: cool, we have been discussing it with andara, and i think we can extend runpugs to run kp6
10:35 Juerd ahmadz:
10:35 Juerd ahmadz: Anything I could help with?
10:36 ahmadz juerd: it seems to a bug in IO::Pty, andara mentioned it was a race condition
10:36 ahmadz juerd: slow down the machine? ;-)
10:36 Juerd Haha :)
10:36 moritz_ LOL
10:37 Juerd I could give it 64 MB of RAM so it'd have to swap all the time
10:37 Juerd But it's already down to 1 out of 2 CPUs, and I can't make it any slower cpu-wise :)
10:38 ahmadz andara will look at it meanwhile i will iron out the remaining ui bugs... im targeting thursday as a final release date.
10:58 pugs_svnbot r17879 | azawawi++ | [runpugs] added 'fix html escaping' reported by moritz my @a = <foo 1 2 3>
10:58 pugs_svnbot diff: http://dev.pugscode.org/changeset/17879
10:58 lambdabot Title: Changeset 17879 - Pugs - Trac
11:10 devogon_ joined #perl6
11:10 riffraff joined #perl6
11:11 rutlov joined #perl6
11:17 pdcawley joined #perl6
11:20 rutlov left #perl6
11:30 pugs_svnbot r17880 | fglock++ | [kp6] multi token ident:string {...}
11:30 pugs_svnbot diff: http://dev.pugscode.org/changeset/17880
11:35 fglock joined #perl6
11:36 Blwood joined #perl6
11:38 ahmadz joined #perl6
11:49 drupek12102 joined #perl6
12:00 franck___ joined #perl6
12:01 theorbtwo joined #perl6
12:12 kanru joined #perl6
12:13 pdcawley joined #perl6
12:18 Blwood_ joined #perl6
12:21 Blwood__ joined #perl6
12:32 Psyche^ joined #perl6
12:34 cognominal_ joined #perl6
12:34 pdcawley joined #perl6
12:35 HiTech69 joined #perl6
12:43 devogon joined #perl6
12:57 cmarcelo joined #perl6
13:19 drrho joined #perl6
13:22 fglock i'm confused with the meanings of variables within a regex
13:23 fglock how do i get to the string being matched
13:23 polettix joined #perl6
13:27 TJCRI joined #perl6
13:27 pdcawley joined #perl6
13:34 agentzh left #perl6
13:35 jhorwitz joined #perl6
13:41 fglock Cursor.pm seems to imply that the string being matched is $_.orig
13:43 pdcawley joined #perl6
13:53 [particle] joined #perl6
14:06 amnesiac joined #perl6
14:11 integral joined #perl6
14:16 pugs_svnbot r17881 | fglock++ | [kp6] more Pair syntax sugar
14:16 pugs_svnbot diff: http://dev.pugscode.org/changeset/17881
14:16 cognominal_ joined #perl6
14:20 buetow joined #perl6
14:35 pugs_svnbot r17882 | fglock++ | [kp6] dollar-cent is a valid identifier
14:35 pugs_svnbot diff: http://dev.pugscode.org/changeset/17882
14:36 c9s joined #perl6
14:44 kanru joined #perl6
15:00 pugs_svnbot r17883 | fglock++ | [kp6] multi token supports french quotes
15:00 pugs_svnbot diff: http://dev.pugscode.org/changeset/17883
15:01 moritz_ fglock: so are you parsing the inpt files consistently as unicode?
15:08 fglock moritz_: it seems to work
15:08 moritz_ cool
15:09 moritz_ and from the sym multis in grammars I gather that you are slowly adapting STD code?
15:09 fglock see t/kp6/56-token-sym.t
15:10 * moritz_ looks
15:10 fglock i'm trying to - there are some things in the new regexes that i don't understand yet
15:11 moritz_ like what?
15:12 fglock the calling convention - like, where is the starting position specified
15:13 vel joined #perl6
15:13 moritz_ you mean in subrules, or when TOP is matched?
15:13 moritz_ TOP is a token, so implicitly anchored to ^
15:22 fglock_ joined #perl6
15:23 fglock_ moritz_: in subrules
15:24 moritz_ fglock_: they are always implicitly anchored to where the calling rule left off
15:24 moritz_ so m/foo<bar>/ will match foo first, set .pos on $_ and call subrule bar
15:24 moritz_ that is implicitly anchored to .pos
15:25 fglock_ there was some talk about .pos being deprecated
15:27 moritz_ fglock_: the biggest change iirc is that the regex remembers the position, not the string
15:28 moritz_ or in terms of variables, it's $/.to
15:28 TimToady the fundamental is that there is no single $_ that is continually being mutated
15:28 moritz_ TimToady to the rescue ;)
15:28 TimToady vaguely
15:29 TimToady coffee is still brewing so ignore half of what I say
15:29 amnesiac hah
15:29 moritz_ "Within a closure, the instantaneous position within the search is denoted by the .pos method on that object."
15:29 moritz_ is this true for subrules as well?
15:29 Kattana which half?
15:29 TimToady this half
15:30 fglock_ i'm translating regexes to plain methods
15:30 TimToady when you have a partial result in a logic program, if you want your logic program to be parallelizable, you can't store that partial result in a mutating variable
15:31 TimToady I am also tranlating regexes to plain methods
15:31 fglock_ how does a method that matches a simple constant looks like
15:31 fglock_ TimToady: should i reuse your code? :)
15:31 TimToady it's still mutating
15:32 TimToady but hang on and I'll send an example
15:32 fglock_ TimToady++
15:32 pasteling "TimToady" at 71.139.49.56 pasted "token { 'a' }" (10 lines, 191B) at http://sial.org/pbot/27526
15:33 TimToady at the moment current "cursor" is in $¢
15:33 TimToady that happens to be self at the top of the method, but each closure within the method gets its own copy of $¢
15:34 TimToady because in theory different parts of it can be "in flight" simultaneously
15:34 TimToady (though generally not in a ratchet grammar)
15:35 TimToady if you want to see more complicated things, "cheat STD.pm | metholate >std"
15:35 fglock_ i've been using self as the current grammar
15:36 TimToady it's also that, but I decided that the current grammar could derive from Cursor, which is the current parser state
15:36 TimToady so a method call can be calling either into Perl or into Curso
15:36 TimToady it seemed to simplify things
15:37 TimToady so I don't have to pass parser state separately from the parser
15:37 TimToady before that my self was pretty much just a static object that never changed
15:37 TimToady (unless you changed the grammar)
15:38 TimToady my current goal is to work out the longest token semantics
15:38 TimToady and make each of these methods actually return a list of longest tokens and actions to take on them.
15:38 TimToady they can then be composed at any point into a lexer
15:39 TimToady but that means the methods themselves are really metamethods part of the time
15:40 TimToady and at the moment I'm looking at swapping in a different PreCursor class to actually do the meta calculations with the same exact body
15:41 TimToady this seems to be the part that nobody else is working on, so I should work on it
15:43 TimToady it's also arguably the core of what makes Perl 6 Perl 6
15:44 TimToady (the ability to mutate your grammar cleanly and efficiently on the fly)
15:44 pdcawley joined #perl6
15:44 pugs_svnbot r17884 | azawawi++ | [runpugs] fixed html escaping problem (my @a = <foo 1 2 3>)
15:44 pugs_svnbot r17884 | azawawi++ | [runpugs] fixed prompt not being taken from server
15:44 pugs_svnbot r17884 | azawawi++ | [runpugs] fixed duplicate prompt
15:44 pugs_svnbot r17884 | azawawi++ | [runpugs] updated html/css for cross-browser compatibility
15:44 pugs_svnbot diff: http://dev.pugscode.org/changeset/17884
15:44 fglock_ http://svn.pugscode.org/pugs/v6/v6-KindaPerl6/lib/KindaPerl6/Runtime/Perl6/Multi.pm
15:45 fglock_ my impl of "multi token"
15:45 [particle1 joined #perl6
15:45 dwave joined #perl6
15:46 fglock_ add_token_variant() adds "sym" routines
15:46 TimToady seems to assume you know the token length in advance
15:47 TimToady S05 does not define tokens such that you can always know their length in advance
15:47 fglock_ i see
15:48 TimToady usually you can, so one could fake DFA semantics by keeping two tables
15:48 TimToady one of static lengths and one of dynamic
15:48 TimToady but you'd have to run all the dynamic ones to see how they compare to the static
15:48 TimToady depending on the mix that might be more space efficient than DFA
15:49 TimToady but probably not more time efficient
15:50 TimToady perhaps one could traverse a trie of static tokens in parallel with stepping through a DFA
15:52 TimToady any of these ways would probably work okay, given that most dynamic tokens would fail near the beginning, not matching <alpha> for instance in the case of identifiers
15:53 TimToady anyway, that's why I'm working on that part of it.  :)
15:54 TimToady it's not something to be terribly worried about in a bootstrap, but when we actually get to where we're doing it all, this needs to be part of it
15:56 fglock_ i'm working on merging STD into kp6 - is there a particular feature that you'd recommend implementing first?
16:00 fglock_ re two tables, that seems doable
16:02 TimToady can you do gather/take lazily?
16:02 fglock_ S05 seems to imply that sigil:sym<$> is fixed-length
16:02 fglock_ i can try :)
16:02 TimToady if so, then you can do nonratchet intermixed with ratchet
16:02 TimToady kept running into pugs bugs when I tried prototyping that
16:03 fglock_ would you have a pointer to a small test case
16:03 TimToady sym is a way of specifying the identifier of some tokens and later some operator names, but it's not the definition of longest token
16:04 TimToady if the rule says "<sym> 'bar'" then the token would be $bar (at least)
16:04 TimToady it's basically everything DFAable up to <ws> or ::
16:05 TimToady sec...
16:07 TimToady it basically includes any assertions even if they might have to be checked after the DFA is done, but leaves out anything with side effects, such as {*} and friends
16:09 TimToady so the DFA yields a list of candidates from longest to shortest, and the closure we call may say "I didn't really match", and then we have to try the next-longest DFA match instead.
16:09 pugs_svnbot r17885 | fglock++ | [kp6] added "fixme" to Multi.add_token_variant()
16:09 pugs_svnbot diff: http://dev.pugscode.org/changeset/17885
16:10 TimToady most of those action closures can just start up where the DFA left off
16:10 TimToady but some of them will need to reparse the token
16:11 TimToady to apply any additional constraints that the DFA can't check
16:11 TimToady or to capture bits of it
16:11 TimToady that's all part of what I'm trying to figure out
16:14 TimToady by the way, the current metholate does not assume ratcheting at all, except to apply .[0] to the gathered list of candidates.  much more could be done there, but I think it would be a premature optimization
16:14 andara left #perl6
16:14 fglock_ re lazy gather, i think it's doable
16:15 TimToady in any case, having bounced several attempts at ratcheting off of pugs, I've given it up as a bad job till we implement "andthen" :)
16:15 fglock_ it would be good to have a test - i'm looking at t/statements/gather.t
16:16 TimToady run a simple sequence through metholate and you'll get a bunch of nested gather/takes; might make a good test
16:17 fglock_ k
16:17 wolverian hm.. what will we call the stm orElse now that it's taken? :)
16:17 TimToady it's already named "maybe"
16:17 TimToady and it's a block prefix, not a separator
16:17 TimToady you can string a bunch of maybe blocks in a row
16:18 wolverian oh... dammit, I knew that! :)
16:18 wolverian is it specced?
16:18 drupek12102 joined #perl6
16:19 TimToady don't think so, though "contend" is mentioned in S04
16:19 wolverian okay, thanks.
16:19 TimToady it's one of those things that should find its way into the concurrency spec
16:20 wolverian yeah, it's just not on spec.pugscode.org, which is what I check first
16:20 * wolverian needs more time for pugs
16:20 * TimToady needs more time for everything
16:22 TimToady actually, it's already in Concurrency.pod
16:22 TimToady so it's kinda specced
16:22 * kolibrie looks in Concurrency.pod for more time
16:23 TimToady keep handing out those commit bits...
16:23 TimToady biab &
16:26 fglock_ hmm - is there a way to kill a perl5 thread? i miht need this in case @a = gather {...} and @a is destroyed
16:26 fglock_ might
16:27 ruoso fglock_, does the ast in kp6 contains line number information?
16:28 fglock_ no
16:28 ruoso it's starting to be KindaHard to debug kp6 without that
16:40 ruoso fglock_, what's needed to implement { return ::Foo( a => b) } inside rules
16:40 ruoso ?
16:41 fglock_ ruoso: it should work already; needs some debugging
16:41 ruoso fglock_, the tests still fail
16:42 fglock_ kp6 already has named parameters, it should be possible to fix .new
16:42 rindolf joined #perl6
16:42 fglock_ and use Foo.new( a => b )
16:44 pmurias joined #perl6
16:44 pmurias hi
16:44 justatheory joined #perl6
16:45 hoelzro joined #perl6
16:46 ruoso fglock_, tests grammar/08 and 09 tests that
16:46 ruoso and both fail
16:46 ruoso pmurias, hi
16:46 ruoso pmurias, echo 'die' | perl -Ilib-kp6-kp6-p5 kp6-kp6.pl
16:46 ruoso some clue?
16:47 ruoso some value of some scalar have a perl5 1 instead of a ::Int 1
16:47 ruoso the problem is how to find out where?
16:47 ruoso fglock_, what is the auto_deref thing?
16:48 fglock_ it tells for() whether to interpolate @array, and not $scalar
16:48 ruoso who sets it?
16:48 ahmadz joined #perl6
16:49 fglock_ it is a role in the Array/Hash class
16:49 fglock_ in MOP.pm
16:49 shriphani joined #perl6
16:50 weinig joined #perl6
16:50 ruoso hmhmm
16:50 shriphani left #perl6
16:50 ruoso for some reason a plain Scalar with a value of "1" has this
16:50 ruoso in kp6-kp6
16:52 pmurias maybe some perl5 component "leaks" perl5 values
16:54 ruoso pmurias, I think it's in the closure of Rule::Block
16:55 pugs_svnbot r17886 | fglock++ | [kp6] investigating lazy gather/take
16:55 pugs_svnbot diff: http://dev.pugscode.org/changeset/17886
16:56 ruoso pmurias, it's probably using the last statement from some closure
16:56 ruoso which returns 1
16:56 ruoso gah...
16:58 ruoso ::DISPATCH( $COMPILER::Code_add_pad, 'APPLY', $Class_name );
16:58 ruoso this is the line that returns a p5 1
17:01 barney joined #perl6
17:02 ruoso hmmm
17:03 ruoso Perl5/Array.pm unshift was returning 1 instead of ::Int 1
17:05 pmurias push makes the same mistake
17:06 * pmurias is thinking of making a Carp which would serialise perl6 values properly
17:06 ruoso pmurias, I've just fixed both
17:07 ruoso interesting is that now I get a Syntax Error...
17:07 ruoso :)
17:08 ruoso that should be a good sign
17:09 pmurias ruoso: do you think perl5 values are a common enough problem to add (optional) automatic detection of them?
17:09 ruoso pmurias, it already detects it
17:09 pugs_svnbot r17887 | fglock++ | [kp6] gather TODO list
17:09 pugs_svnbot r17888 | ruoso++ | [kp6] Perl5::Array unshift/push were returning plain p5 scalar. fixed
17:09 pugs_svnbot diff: http://dev.pugscode.org/changeset/17888
17:09 hoelzro left #perl6
17:10 ruoso pmurias, the message isn't that helpfull
17:10 ruoso but I'm not sure it can ebe
17:11 dakkar joined #perl6
17:17 pmurias ruoso: if we can extract the helpfull information
17:18 pmurias or if any information croak could return would be usefull?
17:19 ruoso pmurias, fglock_, take a look at this new test I've just submitted
17:21 jjore-w left #perl6
17:21 pugs_svnbot r17889 | ruoso++ | [kp6] new failing test: code inside rule always return, even if it did not explicitly return
17:21 pugs_svnbot diff: http://dev.pugscode.org/changeset/17889
17:22 lambdabot joined #perl6
17:24 pmurias ruoso: i know the problem
17:24 pmurias there used to be a return 'stange value' prepended inside code clousres in rules
17:25 TimToady fglock_: actually, I doubt that we'll get a viable implementation of gather/take from ithreads--too heavy
17:26 ruoso pmurias, yep... I remember that
17:26 TimToady note that we changed "return" to "make" last week, and the semantics just now set the return value, but don't halt parsing
17:26 fglock_ hmm
17:26 ruoso pmurias, are you putting that again?
17:26 ruoso fglock_, TimToady, I'm not sure you have that much to gain from parallel executing right now... we can already use make -j to speed as possible
17:27 ruoso and we're talking mostly about 2 cpu machines at most
17:27 ruoso for the currenct scenario
17:27 TimToady I'm not talking about practice, but about theory :)
17:27 ruoso ah
17:27 ruoso ok
17:28 ruoso TimToady, about the s/return/make/... I think we probably will stick with return... at least until we bootstrap kp6....
17:28 fglock_ ruoso: laziness is only meant for non-infinite loops, mostly
17:28 ruoso or else we would need to change mp6 too...
17:29 TimToady don't you have to handle "return" specially anyway?
17:29 TimToady it can't really be a return
17:29 ruoso TimToady, yep... but we use mp6 to compile kp6's grammar right now
17:30 ruoso we would need to modify mp6 to support that change
17:30 ruoso when we bootstrap kp6
17:30 ruoso we will refactor the entire compiler
17:30 TimToady *nod*
17:30 ruoso using the features that are already available in kp6, but are not available in mp6
17:30 fglock_ ruoso: we can tell mp6 to use [ return | make ]
17:31 pmurias make works differently
17:31 ruoso actually, that's why I think it's so important to bootstrap ASAP...
17:33 pmurias although i don't think mp6 depends on return ending the rule
17:33 ruoso kp6 depends on mp6 doing that
17:33 ruoso and that's why kp6-kp6 is failing right now
17:34 ruoso because mp6 does, but kp6 doesn't
17:34 ruoso . o O (geez... that kpmsp6 is confusing)
17:35 pmurias ruoso: it's the other way around
17:35 ruoso nope...
17:35 ruoso mp6 is the compiler for Grammar
17:36 ruoso when the grammar is compiled by kp6
17:36 pmurias yes
17:36 ruoso it fails because of that
17:36 pmurias kp6 depends on mp6 not ending the rule on a  closure without a return
17:36 pmurias is what i mean
17:37 ruoso ah... ok
17:37 fglock_ i've got an old implementation of yield() somewhere, that doesn't use threads - i'll try to find it
17:38 ruoso pmurias, so, do you know how to fix that?
17:38 pugs_svnbot r17890 | azawawi++ | [runpugs] show "Please wait" message now works on FF2,IE7,Opera
17:38 pugs_svnbot r17890 | azawawi++ | [runpugs] hidden iframe is loaded after document is ready (faster page load)
17:38 pugs_svnbot diff: http://dev.pugscode.org/changeset/17890
17:38 lambdabot Title: Changeset 17890 - Pugs - Trac
17:39 pmurias ruoso: doing it right now ;)
17:39 ruoso :)
17:40 pmurias the long compiles help to catch up with irc ;)
17:40 pmurias *catching
17:40 ruoso home &
17:41 ruoso try to run kp6-kp6 when you finish
17:41 ruoso pmurias, ^
17:41 TimToady fglock_: finally backlogged, did your question about current position get answered?
17:41 ruoso pmurias, like: echo 'say "Hello World"' | perl -Ilib-kp6-kp6-p5 kp6-kp6.pl
17:42 pmurias make is overzelous
17:42 fglock_ TimToady: i think so, current position is stored in an instance of Cursor, which inherits from the current grammar
17:42 pmurias * zealous
17:43 TimToady basically, all that information (both current position, and which string it points into) must come through Cursor or we lose any hope of multithreading
17:43 fglock_ TimToady: ok
17:43 fglock_ home too &
17:43 TimToady bye &
17:44 PerlJam joined #perl6
17:45 TimToady actually, current grammar inherits from Cursor because Perl inherits from Cursor
17:52 kyrbe joined #perl6
17:52 kyrbe left #perl6
17:55 Psyche^ joined #perl6
17:55 ahmadz_ joined #perl6
17:57 daxim joined #perl6
17:58 ferreira joined #perl6
17:59 ferreira moritz_: hi
17:59 lambdabot ferreira: You have 1 new message. '/msg lambdabot @messages' to read it.
17:59 ferreira @messages
17:59 lambdabot TimToady said 6d 20h 8m 24s ago: pugs does not yet implement Range objects, nor does it really treat * as anything other than Inf, so 0..* is pretty much doomed at the moment
18:01 ferreira I just wrote something about the microgrant to write micro-articles on Perl 6 operators: http://use.perl.org/~ferreira/journal/34473
18:01 lambdabot Title: Journal of ferreira (5993)
18:02 ferreira Is there someone willing to read the one about the zip operator: http://ferreira.nfshost.com/perl6/zip.html
18:02 lambdabot Title: Yet Another Perl 6 Operator: Zip
18:03 PerlJam ferreira++ btw
18:03 pugs_svnbot r17891 | pmurias++ | [kp6]: fixed closures in rules withought return
18:03 pugs_svnbot diff: http://dev.pugscode.org/changeset/17891
18:03 lambdabot Title: Changeset 17891 - Pugs - Trac
18:04 PerlJam ferreira: How much did you ask for in your grant?  I was thinking of submitting an article on perl 6 regex as a TPF microgrant but I really have no clue what to ask for.
18:05 ferreira Well, I did not say how much. I believe each microgrant is $500 and that's what I am assuming.
18:06 PerlJam oh really?
18:06 ferreira That's what I infered from: http://www.nntp.perl.org/group/perl.perl5.porters/2007/03/msg122448.html
18:06 lambdabot Title: Perl 6 Microgrants. Now accepting proposals. - nntp.perl.org, http://tinyurl.com/2bjsdl
18:06 ferreira and the microgrants that were accepted until now.
18:07 PerlJam I thought it was like any other TPF grant where you *must* specify the amount or it would be rejected.
18:07 PerlJam no having to make that decision makes the idea of proposing something much easier  :)
18:07 PerlJam s/no/not/
18:08 ferreira I don't think so. In the presentation of the project, it says: "We'll be making USD500 grants"
18:09 [particle1 yep, there's 10 USD500 microgrants
18:09 PerlJam Well, I should write my proposal then :-)
18:12 ferreira Go for it! See the last paragraph of http://www.perlfoundation.org/july_5th_2007_a_trio_of_perl_6_microgrants
18:12 lambdabot http://tinyurl.com/37fnew
18:13 PerlJam ferreira: thanks.
18:13 IllvilJa joined #perl6
18:13 PerlJam ferreira: did you need reviewers of your articles or you just want people to read them in general?  :)
18:14 ferreira Reviewers indeed ;-) I need to learn a lot for catching up with Perl 6
18:14 ahmadz_ http://feather3.perl6.nl/runpugs/
18:15 lambdabot Title: Run Perl 6 Now -- in your browser!
18:15 ferreira For example, by now I want a statement to use (from Perl 6) a Perl 5 module and import a sub from it.
18:17 PerlJam ferreira: your eq_array() subroutine it buggy.
18:17 ferreira Ok. What's wrong?
18:17 PerlJam "return False if $x eq $y"
18:18 PerlJam s/eq/ne/
18:18 PerlJam :)
18:18 PerlJam also, that's just string comparison.
18:18 PerlJam you may want to use eqv
18:18 PerlJam (or whatever it's called)
18:18 c9s joined #perl6
18:18 ferreira ha ha. Maybe I should rename it to neq_array ;-)
18:19 kjwcode joined #perl6
18:19 PerlJam Equivalence operators should get an article of their own early on IMHO
18:20 PerlJam ferreira: And I think it's just  use "perl5:Lingua::EN::Numbers";  But that may just a be a pugsism rather than a perl6ism.
18:20 jisom joined #perl6
18:21 PerlJam ferreira: I *Really* like your disclaimers
18:21 ferreira Well, I am not seriously concerned about not using non-introduced operators. For example, ~~ will be used
18:21 sonmez joined #perl6
18:21 PerlJam ferreira: I realize that, I just think that there will be confusion about ==, =:=, eq, eqv, and whatever other equivalence operators there are  :)
18:22 PerlJam (and I tend to think that those things with a high chance of confusion should be covered early and well)
18:23 ferreira eqv is the general comparison operator? The one that works kind of Java equals()?
18:23 sonmez left #perl6
18:28 ferreira Well, I fixed 'eq' by '!eqv' and used the 'use perl5:' thing
18:39 franck___ joined #perl6
18:39 TimToady ferreira: "perl5" used to be a perl6ism but is now just a pugsism because Perl 6 is now specced as "use LingENNum:from<perl5 Lingua::EN::Numbers>" or some such.  but pugs doesn't do that yet
18:40 TimToady that in S11:390 and thereabouts
18:42 ferreira TimToady: thanks. Reading S11
18:50 ferreira So what I want is "use Lingua::EN::Numbers:from<perl> <num2en>" or do I need the Perl 5 module name and an alias?
18:50 PerlJam TimToady: could that have been "use Foo:from<perl5 Lingua::EN::Numbers>"  (where Foo is essentially a made-up name that acts as an alias)  ?
18:51 TimToady the intent is the latter, but the either form could probably be made to work
18:53 ferreira Ok, this is clear from the example at the synopsis: " use Whiteness:from<perl5 Acme::Bleach 1.12 cpan:DCONWAY>; "
18:53 TimToady presumably the syntax inside the :from<> corresponds to the source language's view of the module name
18:53 TimToady which might not map into the Perl 6 namespace cleanly
18:53 TimToady plus it's useful to have a ::-less alias
18:54 PerlJam I guess, I'm wondering if there are any "hidden" (to me) constraints on the perl 6 name.
18:54 PerlJam s/,//
18:55 ferreira In the example, the only thing needed is a way to use num2en from Perl5 Lingua::EN::Numbers.
18:56 TimToady presumably one could just say "use :from<perl5 Lingua::EN::Numbers> <num2len>" to just import that, and perhaps generate an alias
18:56 TimToady but since imports are lexical by default, we don't actually need to have an alias
18:57 TimToady well, except maybe for decent error reporting
18:58 TimToady I guess we really need to track a name and version and authority even for foreign modules, if we want to pull in multiple of the same name...
18:59 TimToady but that might be impractical depending on the foreign language
19:00 ferreira Even for Perl 5 it would depend on an extension mechanism that does not exist today. That is an enhanced toolchain that saved that info (name, version, authority)
19:01 theorbtwo For example, I'm betting .net CLRage will have something similar in spirit but quite different in details already that we'll want to pull from when loading C# stuff.
19:02 TimToady indeed, there are similarities there
19:02 TimToady we won't say who borrowed the idea from whom...  :)
19:05 ferreira Any implementation of lexical imports by now? Sorry to keep asking: as I said I need to catch up yet with current Perl 6.
19:06 TimToady pugs, presumably
19:23 moritz_ ferreira: I looked at zip.html, looks fine at first sight
19:24 theorbtwo Hm, I wonder if you can do lexical imports in 5.10 with ours?
19:27 TimToady only by also doing a package import at the same time, but the point of lexical import is to reserve packages for export lists, not imports
19:27 TimToady well, that's one of the points, anyway
19:28 jjore-w joined #perl6
19:32 vel joined #perl6
20:06 pmurias CLRage?
20:07 pmurias goole dosn't find anythinh
20:07 pmurias *anything
20:14 pmurias or is it something like a ice age?
20:15 autark joined #perl6
20:33 dduncan joined #perl6
20:42 ferreira left #perl6
20:43 [particle] CLR age, maybe?
20:47 rhr @tell ferreira in 'my @E = map ...' there needs to be a comma after the block
20:47 lambdabot Consider it noted.
20:53 TimToady I took it to be the -age suffix, as in postage is what goes along with the post
20:53 TimToady CLRage is then just the stuff associated with CLR
20:54 TimToady a package is something you pack, etc.
20:54 TimToady it's not a terribly productive suffix in modern English
20:55 TimToady amperage is the amount of amperes, voltage is the number of volts
20:57 vel joined #perl6
20:57 TimToady @tell pmurias on CLRage see http://moritz.faui2k3.org/irclog/out.pl?channel=perl6#i_104982
20:57 lambdabot Consider it noted.
21:01 kane_ joined #perl6
21:12 cmarcelo joined #perl6
21:17 ashleyb joined #perl6
21:56 justatheory joined #perl6
21:57 vel joined #perl6
22:06 polettix joined #perl6
22:16 thoughtpolice joined #perl6
22:36 pugs_svnbot r17892 | avar++ |  r63477@t:  avar | 2007-09-17 22:35:14 +0000
22:36 pugs_svnbot r17892 | avar++ |  Replaced the old Makefile with a custom Makefile.PL system based on
22:36 pugs_svnbot r17892 | avar++ |  ExtUtils::MakeMaker, `make test', `make dist', `make disttest' and all
22:36 pugs_svnbot r17892 | avar++ |  the other usual suspects now work as with other perl dists. It works
22:36 pugs_svnbot r17892 | avar++ |  for all the stuff the old Makefile worked for (unless I screwed up).
22:36 pugs_svnbot r17892 | avar++ |  However `make install' does not yet work, but that's just a matter of
22:36 pugs_svnbot r17892 | avar++ |  tweaking the right parameters and maybe modifying a couple of .pl files.
22:36 pugs_svnbot diff: http://dev.pugscode.org/changeset/17892
22:36 lambdabot Title: Changeset 17892 - Pugs - Trac
22:37 Limbic_Region joined #perl6
22:37 avar fglock_: (and other kp6 guys), don't be afraid to revert that change if it screws stuff up
22:38 avar but it's a nice start for the stuff I was whining about at yapc::eu, getting kp6 on cpan and having it smoked:)
23:10 polettix joined #perl6
23:20 [particle] avar: was that kp6 specific? i didn't see a [kp6] tag in the commit message
23:20 pugs_svnbot r17893 | avar++ | make test TEST_VERBOSE=1 now works as in other EU::MM dists
23:20 pugs_svnbot diff: http://dev.pugscode.org/changeset/17893
23:20 lambdabot Title: Changeset 17893 - Pugs - Trac
23:21 avar [particle]: Yeah it was, I don't know why [kp6] doesn't appear
23:21 avar Maybe because svk really commits to the whole repository rather than a subdir
23:22 avar [particle]: The last one was too:)
23:22 [particle] nice, hopefully it'll build on windows :) i'll try later, time for a meeting &
23:22 avar It requires GNU specific makefile features still, but that should be hackable
23:23 avar Just need to make a glob-thingy in perl instead of the makefile stuff it has now
23:47 luqui joined #perl6
23:52 rfordinal_ joined #perl6

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

Perl 6 | Reference Documentation | Rakudo