pugs.blogs.com | pugscode.org | pugs.kwiki.org | paste: sial.org/pbot/perl6 | <stevan> Moose... it's the new Camel ":P | .pmc == PPI source filters!
Set by Alias_ on 16 March 2006.
buu go go +R powers! 00:07
00:21 mako132_ joined 00:40 macroron joined 00:44 macroron joined 00:52 macroron joined 00:54 macroron joined 00:57 macroron joined 01:00 amnesiac joined, macroron joined 01:13 nirgle joined 01:14 frederico joined 01:36 EyeCue joined 01:37 EyeCue left 01:41 LCamel joined 01:56 tcliou joined 02:02 KingDiamond joined 02:07 bradleym_ joined 02:08 bradleym_ is now known as bradleym 02:30 aufrank joined
aufrank hello.\(room.all) 02:32
audreyt yo 02:34
aufrank has been reading the logs audreyt linked to in her last journal 02:35
neat stuff goes on when I'm not here!
I should be here more often
maybe I'll catch it ;)
audreyt: have you checked in anything on your way to doing opp all with rules? 02:36
audreyt aufrank: src/Pugs/Rule/Expr.hs? ;)
no, there is no code yet. 02:37
aufrank so, my first post ever to p6l didn't actually get there because I posted using the google groups interface 02:40
this was probably a year ago
but I was thinking about the question again the other day
does anyone have any ideas about how to use p6 rules to generate patterns rather than parse them? 02:41
I'd like to be able to generate samples from a grammar 02:42
02:43 macroron joined 02:44 DesreveR joined
aufrank something like $rule = { (g|b)[1] (a|i)[1] (p|d)[1] (o|u)[1] } 02:46
$rule.pick # gapo or gido or badu or ... 02:47
audreyt why, yes 02:57
also see misc/pX/audreyt/rules-are-templates
related but not the same idea
(where (g|b) becomes validators instead of generators) 02:58
aufrank goes to look 03:01
03:04 nirgle left 03:06 macroron joined
aufrank [% GET foo %] <=$foo> 03:06
you think the TT2 GET returns an iterator in p6 world? 03:07
audreyt this is rules
not expr
so <=$foo> is just... imaginative syntax
for
$foo := .*?
probably
aufrank oh oh
I see
03:09 Quell joined 03:29 nirgle joined 03:50 qu1j0t3 joined 03:53 FurnaceBoy joined, Daveman joined 03:56 macroron joined 04:01 Soga joined 04:49 spo0nman joined 04:53 Southen_ joined 05:01 kakos joined 05:17 nirgle left 05:36 justatheory joined 05:38 marmic joined
GeJ good localtime moosefolks 05:56
06:23 Frances joined
lypie morning * 06:26
arcady evening 06:29
06:52 Cryptic_K joined 06:54 pdcawley joined 07:36 Southen joined 07:38 dakkar joined 07:39 Aankhen`` joined 07:50 iblechbot joined 07:53 nnunley joined 08:27 hugues23 joined 08:37 pdcawley joined 08:46 f0rth joined 08:55 elmex joined, ruz joined 09:11 drrho joined 09:30 lypie left 09:45 Maxiii joined 09:51 gaal_ joined
xinming img.photobucket.com/albums/v715/ray...mputer.gif :-) 10:11
dakkar is away: pranzo 11:01
11:04 frederico joined, TimToady joined
xinming Larry landed.... :-) 11:06
TimToady: hmm, what is the role of control statements? objects? or it built-in... 11:07
xinming is thinking.... if the control statements can be object... as &?ROUTINE
11:11 ghenry joined 11:29 nnunley joined 11:55 mako132_ joined, lichtkind joined 12:03 fglock joined 12:09 mako132_ joined
ruz eval? sub foo($arg!){say $arg}; sub bar(*@rest){say @rest;foo(*@rest)}; foo arg=>2; bar arg=>3; foo() 12:15
lichtkind ?eval sub foo($arg!){say $arg}; sub bar(*@rest){say @rest;foo(*@rest)}; foo arg=>2; bar arg=>3; foo() 12:17
12:17 evalbot_10038 is now known as evalbot_10042
evalbot_10042 OUTPUT[arg2 ] Error: No such sub: "&foo" 12:17
ruz oh :) I have to recompile 12:18
12:18 kolibrie joined
lichtkind :) 12:18
ruz ?eval sub foo($arg!){say $arg}; sub bar(*%rest){say %rest;foo(*%rest)}; foo arg=>2; bar arg=>3; foo()
12:18 Limbic_Region joined
evalbot_10042 OUTPUT[arg2 arg3 arg3 ] Error: No such sub: "&foo" 12:18
ruz the first thing is wrong too 12:19
?eval sub foo($arg!){say $arg}; sub bar(*@rest){foo(*@rest)}; bar arg=>3; # should fail or output something sane 12:20
evalbot_10042 OUTPUT[ ] bool::true
ruz is it correct syntax? 12:21
fglock putter: a p5 implementation of PGE would be very useful - re: colabti.de/irclogger/irclogger_log/...=670#l1116 12:23
but an Inline::C version would be even better :)
I'm on YAPC::Brasil - the hosting conference is 5200 people and still counting! 12:25
audreyt, TimToady - if you'd likely come to YAPC::SA in november please let me know 12:27
everyone else too :)
kolibrie fglock: did you see the p6grammar to PCR module compiler I wrote yesterday? 12:31
svn.openfoundry.org/pugs/misc/pX/Co...grammar.pl 12:36
dakkar is back (gone 01:35:19)
fglock kolibrie: looking 12:39
wow - very short :) 12:42
kolibrie fglock: thanks to PCR and 'return' in rules
fglock cool. A possible usage is also "... > Perl6GrammarFile.pmc" 12:44
kolibrie yes
or I was wondering if the pmc should be even further compiled Perl 5 code 12:45
fglock yes - but then you have to recompile when you upgrade PCR 12:46
kolibrie ahh
does PCR depend on pugs? I'd like to be able to use grammars on systems without pugs 12:47
fglock recompiling the PCR internal mini-language is fast, and it only happens at load time
no it doesn't
kolibrie okay, cool 12:48
fglock it only depends on PadWalker, for looking up lexical variables
kolibrie I thought it depends on v6.pm and Inline, too 12:49
fglock v6.pm doesn't exist physically - it is there just for dieing in case you try to execute the p6 module with p5 directly - Inline is not used either 12:53
kolibrie ok 12:54
kolibrie is very excited to start using grammars in production
fglock: is there a way to specify modifiers with PCR? :perl5 :i and such 12:56
fglock you can add { p => 1 } at compile time, for matching at position 0 - but this is not in CPAN yet 13:01
:perl5 will be { perl5 => 1 } - but this is not implemented yet 13:02
going to merlyn's talk - dunno if there is wifi there... & 13:03
kolibrie so those go in Pugs::Compiler::Rule->compile(q( { perl5 => 1 } <rule_stuff> { return ... } ))
fglock ...->compile( $rule, %opts ) 13:05
kolibrie ok 13:06
fglock wifi is ok - still here :) 13:07
kolibrie (wifi)++
13:08 Aragone is now known as Arathorn 13:36 vel joined 13:54 KingDiamond joined
audreyt TimToady: how does /x++/ differ from /x+/? 13:54
also, "parse" is a rather common prefix sub name
and not a noun 13:55
using it as a declarator is very, very weird.
Juerd has plenty parse_foo's
audreyt I mean, that will deny people of saying: sub parse { ... } ; parse foo(); 13:56
which is rather strange
Juerd It's the same with log. 13:57
People get used to having to choose alternatives.
audreyt Juerd: log is not a declarator.
Juerd Though I'd prefer to be able to use log myself, and yea, parse too.
audreyt TimToady's last S05 change put "parse" in same category as
"sub" and "method".
and "rule". 13:58
my $x = parse { ... };
grammar Foo{ parse moose { ... }; }
I'm very uncomfortable with a verb at that position.
Juerd Your discomfort appears sane. 13:59
pmichaud audreyt: I agree "parse" is not so good; what's a good alternative?
audreyt if people can't use token:w{...}
then "term" is second best I suppose. 14:00
pmichaud oh, I think token:w is probably perfect
audreyt but I think token:w{...} is very nice already
pmichaud yes, I agree
audreyt it feels like overhuffmanizing.
pmichaud <audreyt> TimToady: how does /x++/ differ from /x+/?
token { /x+/ } and rule { /x+/ } are different 14:01
token { /x++/ } and rule { /x++/ } are the same
(i.e., it forces backtracking)
audreyt gotcha.
pmichaud is composing a reply to S05 now 14:02
14:03 bradleym joined
audreyt oh 14:04
should I abort my edit? :) 14:05
pmichaud no, we can both reply, I think :)
audreyt k :)
pmichaud depends on what you're replying to
14:05 r0nny joined
theorbtwo Forces backtrackability, surely. 14:06
audreyt just the "parse".
pmichaud okay, I'll just put a "me too" in mind about the parse
14:08 justatheory joined
audreyt actually, I listed you in the mail :) 14:09
sent
14:18 amnesiac joined 14:24 bsb joined
bsb hello from hong kong 14:26
gugod hello bsb 14:28
bsb hi gugod 14:30
gugod :) 14:31
14:33 PolettiX joined
bsb gugod, you're a .tw hacker right? I'm there tomorrow for a few days 14:33
gugod bsb: heh, really ?
bsb: yes I'm from .tw
14:34 PolettiX is now known as froh-doh
bsb I'm taking the long way home 14:34
gugod bsb: where will you be ? taipei ?
bsb I fly to taipei
clkao bsb: you should call for an emergency socail meet for taipei.pm
gugod exactly
and poke hcchien 14:35
bsb put up the camel spotlight in the sky...
that'd be great
gugod are you staying maybe to the end of April ? you can probably meet ingy and miyagaywa ;) 14:37
bsb only until the 26th, I offered ingy beer|coffee to coax him out
gugod ha 14:38
bsb I miss miyagawa by a day I think :(
gugod bsb: right... well it's still cool, so will you be staying in taipei for the most of time ? 14:39
bsb I don't know, I have no sense of Taiwan's scale
I just bought a 2nd hand travel guide 14:40
gugod bsb: ingy and I are living in hsinchu city, which is 1 hour driving from taipei
bsb: oh so you are still planning ?
bsb "still" would suggest that I'd started... 14:41
gugod alright, I get it
bsb if there's public transport, I'd make a trip out some time 14:42
gugod public transport in Taipei city is good enough, and bad enough for mostly all other cities 14:44
14:44 FurnaceBoy joined
ingy hola 14:45
gugod!
gugod I can almost hear you typing that excalmation mark, ingy
ingy :) 14:46
bsb "Hsinchu is a medium sized city that is famous in Taiwan for it windy winter weather..." 14:47
ingy soothes himself with an excalmation mark
bsb: how many days are you here? 14:48
bsb "...if Hsinchu has anything (besides computers) that would appeal to foreigners, it's the nightlife"
5 days
ingy where are you staying? 14:49
bsb not booked yet, taipei somewhere I guess
ingy cool, maybe I'll come to taipei on Monday 14:50
audreyt ooh
audreyt tries to reschedule $job away
azuroth yo, I'm really cool
ingy :)
gugod we could meet on monday, I will be in taipei on monay 14:51
ingy pets azuroth
bsb took a while to work out what day it is (jetlag)
gugod monday
azuroth I'm really, really supercool. :-)
ingy azuroth: yes, we know
audreyt azuroth: so cool you triggered fusion? 14:52
bsb so how do we meet up?
azuroth no, but I can try 14:53
bsb gugod, ingy: email me any plans or tips at bereft döt net 15:04
15:07 weinig joined 15:09 hexmode joined 15:23 Yumma joined
Yumma bigtitsroundasses.bangbros1.com/gal...videopost/ www.sog10.com/gallery01/c2100k/index.html 15:23
15:23 Yumma left
bsb good night all 15:56
16:08 hexmode joined 16:10 froh-doh is now known as |froh-doh|, |froh-doh| is now known as froh-doh 16:21 zgh joined 16:22 xinming1983 joined 16:32 hlen joined 16:34 xinming1983 joined
audreyt by request of #parrot folks - currently in #perl6 emulation mode there - 16:39
I did a sketch port of
perlcabal.org/~autrijus/tmp/tg.html
from TGE syntax
to
perlcabal.org/~autrijus/tmp/p6.html 16:40
in Perl6 syntax.
16:40 PerlJam joined
ingy hi audreyt 16:40
audreyt hi 16:41
16:42 vel joined
xinming1983 audreyt: hmm, I wonder, what the roles are control statements in perl 6. is it built in, or an object... 16:44
audreyt xinming1983: they are functions 16:45
16:45 xinming1983 is now known as xinming
xinming sorry, bbl 16:45
16:46 xinming joined
xinming hmm, what's the proto of them? I mean, if they are functions, how to "implement" if... :-) 16:46
I know loop statements can be implemented using goto... How about if... :-/ 16:47
theorbtwo xinming: Can be emulated with a form of if that only does if (boolean) {goto foo}. 16:51
xinming theorbtwo: but the problem is... does the if call itself recursively? 16:53
Limbic_Region xinming - sure (I think) 16:54
as long as the condition can be translated to a bunch of booleans 16:55
if ($foo eq 'bar' && $blah == 1) { ... } = $foo eq 'bar' && $blah == 1 && goto somewhere
xinming kisses Limbic_Region :-) 16:56
right... I know now...
Limbic_Region but that is only if - not if/else or if/elsif/etc
as theorbtwo was saying above 16:57
xinming given when can be emulated with match pattern... So It's not a problem...
Limbic_Region although - I think they fall out quite easily
if/else is a no brainer because
$cond1 && $cond2 && goto somewhere 16:58
can unconditionally be followed by
goto alternative
xinming hmm, Maybe I still ask about how to implement the infix && ?
:-)
PerlJam Limbic_Region: did you happen to read Allison's use.perl journal on implementing if/elsif/else for punie?
xinming It's really made me insane... as I don't understand how can these be implemented...
PerlJam: url please... :-)
Limbic_Region and if/elsif/else is easy too because they are just a series of ifs with an unconditional goto at the end
PerlJam - yes
PerlJam xinming: um ... use.perl.org/~allison/journal should get you close enough 16:59
theorbtwo Limbic: Don't goto for the else clause, just follow it.
Limbic_Region use.perl.org/~Allison/journal/29358
xinming how about infix && ?
PerlJam What L~R said
Limbic_Region theorbtwo - I thought the game was to implement as much as possible with gotos :-)
theorbtwo Ah. 17:00
Limbic_Region PerlJam - I am not sure I agree with Allison's conclusions though
theorbtwo BTW, the infocom compiler uses the idiom that Allison liked.
xinming hmm, infix && is handcoded into parrot assembler, right?
Limbic_Region wanders off for a meeting 17:01
pmichaud the perl6.pbc compiler handles if/elsif/else
and infix:&&
(infix:&& is a special form of if/elsif/else)
PerlJam pm: But it doesn't do variables :( 17:08
pmichaud pj: it doesn't?
PerlJam well, it didn't the last time I looked :)
xinming wonders if we can make these control statements objects, so that we can access the target address in VM. and backtrack the address where it starts... :-P
a crazy idea though...
pmichaud (it doesn't understand real lexicals yet, but it does do variables) 17:09
scalars, at least
PerlJam pmichaud: what about looping contstructs?
pmichaud those are coming soon -- I may unify those with the conditionals
PerlJam if it groks goto, then I could already write BASIC-like programs :) 17:10
xinming do you mean, every thing in perl 6 can be objects?
even include the control statements?
pmichaud control statements are exceptions :-)
depending on the statement
PerlJam xinming: everything in perl 6 can be thought of as an "object" :) 17:11
Whether you'd want to do that or not is another matter however
xinming PerlJam: yes, as audreyt ever said that control statements are functions... So, If they are functions, can we make it real object... :-P
hmm, I mean, it is specified.... 17:12
which means, we can ask the current "if" object something interesting, eg: test statements, and where the target in source will it go, and where the target after it compiled into parrot assembly... 17:14
and even where it belongs to...
etc etc...
is it crazy or it is insane? :-) 17:15
PerlJam xinming: well ... I don't know that we'll be able to ask "if objects" much about how they're compiled, but all of that other stuff sounds reasonable.
17:47 FurnaceBoy_ joined 17:51 hlen joined 17:54 hexmode joined 17:58 particle_ joined 18:04 DaGo joined
svnbot6 r10043 | audreyt++ | * Upgrade Artistic license to the latest revision (2.0beta14), 18:28
r10043 | audreyt++ | made availble yesterday, under:
r10043 | audreyt++ | www.perlfoundation.org/legal/licens...c-2_0.html
TimToady pmichaud: ping 18:29
pmichaud TimToady: pong 18:31
TimToady I was just wondering offhand if <$foo> captures by default in PGE, and whether the default should be to capture to $<foo>, and suppress with <?$foo>, or default to not capture, and force capture with <:$foo>.
likewise for <@foo> and <%foo>
pmichaud I had been thinking that <$foo> would capture by default
it seems the most natural 18:32
we can also let it be non-capturing by default, and then use aliases
$<foo>:=<$foo>
TimToady <:$foo> would be shorthand for that
as in sigs
pmichaud at some point I feel like ETOOMANYSHORTHANDS :-)
TimToady we already have that shorthand. 18:33
pmichaud at one point we had also discussed using the leading : for intra-rule modifiers
i.e., <:words(0)>
18:33 fglock joined
pmichaud (to answer your original question -- at the moment PGE doesn't handle lexical variables other than the ones in the rule itself, such as backreferences) 18:34
18:34 aufrank joined
TimToady it would only work on :<sigil> presumably, but I think I like capture by default anyway. 18:34
pmichaud (being that lexicals in Parrot weren't completely spec'd)
audreyt oh? 18:35
pmichaud capture by default "feels" right to me
svnbot6 r10044 | audreyt++ | * Back out the previous revision, restore Artistic 2 to 2.0b5,
r10044 | audreyt++ | as the new b14 asserts itself to derivatives to each files under
r10044 | audreyt++ | the collection, not only the collection as a whole -- the meaning of
TimToady phone
svnbot6 r10044 | audreyt++ | "Package" is changed from the one used Artistic1/2.0b5.
r10044 | audreyt++ | (as such, I can't upgrade the license as we have parts under the
r10044 | audreyt++ | BSD/GHC license.)
audreyt (what part of pdd20 is missing?)
pmichaud audreyt: sorry, I mis-stated. At the time I wrote that portion of PGE, lexicals weren't completely implemented
audreyt ah. ok :)
pmichaud even then, it requires a bit more scaffolding in place to implement, since we have to know the lexical environment at the time the rule is compiled 18:36
xinming TimToady: is it possible make control statements objects? as I asked an hour ago. :-P
fglock wonders if he will ever catch up after the YAPC
pmichaud audreyt: I'm one of those who doesn't believe in a specification until there's an implementation for it :-)
audreyt pmichaud: tis also me :) 18:37
TimToady unphone
18:38 hugues23 joined
TimToady "everything is an object", but some things are things... 18:38
and some things aren't things...
xinming .doit .() # okay, no arguments, same as .doit() <--- doesn't this follow the "long dot" rule? I mean, .doit. .()
TimToady certainly some control structures are just subs. 18:39
pmichaud ISTR that we had said in an earlier spec (perhaps one of the drafts of S05) that <$foo> would capture to $<$foo>
TimToady where?
xinming in S12
TimToady will fix.
xinming Parentheses (or a colon) are required on the dot notation if there are any arguments (not counting adverbial arguments). There may be no space between the method name and the left parenthesis unless you use the dot form of parentheses:
TimToady or you can fix 18:40
I'm thinking maybe we want it to capture to $<foo> instead, to be more like named params. 18:41
pmichaud so, <$foo> and <@foo> both capture $<foo>?
xinming TimToady: I don't have commit bit. >_<
pmichaud er, to $<foo>
xinming -- I'll fix
TimToady Seems unlikely that you'd want <$foo> and <@foo> in the same scope. 18:42
pmichaud agreed
TimToady and you can always $<$foo>:= if you like.
pmichaud We already capture <OtherGrammar::rule> to $<OtherGrammar::rule>, so it might be nice to be consistent 18:43
avar uses $foo and @foo in the same scope in p5
pmichaud looking at it the other way, one can always $<foo>:=<$foo> :-)
Is this still true...? >> .doit. .(1,2): 3 # okay, same as .doit(1,2,3) 18:44
(just verifying)
btw, audreyt, I like 'lexeme' 18:45
aufrank does research on lexemes and lemmas 18:46
xinming I agree, that something are things, but if we make if objects, maybe we can access the control statement object inside the code... hmm, eg: get the test_statements, and modify it(in loop), since &?BLOCK can become object, why can't the control statements?
TimToady avar: I meant the same rule scope. Don't care if the surrounding scope has both variables 18:47
aufrank lexeme means "form-based representation of a word in the mental lexicon" to me
TimToady pmichaud: yes, that's still correct
pmichaud committing
aufrank but I recognize that I'm in the minority there
PerlJam greetings programs. 18:48
pmichaud watched Tron a few nights ago also :-)
theorbtwo ?eval "Hello, PerlJam"
TimToady xinming: if the compiler can detect the desire to objectify "if" and pessimize, it's okay.
PerlJam aufrank: lexeme doesn't work for me either btw :-)
18:48 evalbot_10042 is now known as evalbot_10044
evalbot_10044 "Hello, PerlJam" 18:48
FurnaceBoy_ hehe@PerlJam
TimToady But we have to give the optimizer room to work by default.
FurnaceBoy_ end of line. 18:49
TimToady to me lexeme is synonymous with token.
PerlJam mistakes FurnaceBoy_ for the MCP and starts blasting
TimToady: me too.
TimToady END OF LINE!!!
lemma is one I hadn't heard. 18:50
aufrank 's fine by me. I'm used to typing it anyway ;)
PerlJam phrase seems to work best as the intent of "parse"
or how about "idiom" :)
FurnaceBoy_ "who is your user, program?!"
PerlJam or even "words" (a little vague probably)
pmichaud I came across "phrase" while thinking of "sequence of words", since it's essentially "rule :ratchet :words" 18:51
aufrank fragment is possible, as well
pmichaud shorten it to "frag", and we'll be playing FPS games :-)
xinming TimToady: Ok, I know that we can objectify control statements. then perl 6 can be as dynamic as possible. :-P
lichtkind labels are still made: "label:" ? 18:52
TimToady lichtkind: yes 18:53
must be at statement boundary though.
(including after do)
I rather like "phrase" myself, though that kind of drives s/token/word/ 18:54
pmichaud I was also trying to think of what word would work well in the grammar spec instead of "rule" (and "parse" doesn't work)
lichtkind on left boundary?
aufrank re: lemma www.smithsrisca.demon.co.uk/psychol...ssary.html
TimToady where parser is expecting a statement.
doesn't have to be leftmost
pmichaud I think I still prefer "token" over word 18:55
since rules (that aren't phrases) will use tokens
although "atom" comes to mind
TimToady It's not badly overloaded, true.
atom tends to be badly overloaded.
pmichaud yup
TimToady and our tokens tend to be more like molecules anyway. 18:56
pmichaud well, I vote for phrase (for whatever my vote counts)
TimToady I kind of like that the fancier things get longer: rule/token/phrase 18:57
but not too much longer.
the intent of phrase as a keyword is not really huffmanization of typing, but huffmanization of thinking,
because I want something that maps almost exactly to the concept of a BNF production, if not exactly the syntax 18:58
aufrank in probabilistic context free grammars we talk about conditioning on the head of a rule
lichtkind thanks TimToady maybe for your amusement the perl6tut im writing on opens with an quate of paul graham "There are some stunningly novel ideas in Perl, for example. Many are stunningly bad, but that's always true of ambitious efforts. At its current rate of mutation, God knows what Perl might evolve into in a hundred years.", but chapter 2 and 4 are yours :)
aufrank TimToady: me too 18:59
lichtkind qoute
TimToady also, having a declarator naturally give a hook for pragmas
use phrase :myopt;
aufrank nice 19:00
TimToady use token :foo;
use rule :exhaustive; # :-/
of course, if we define one in terms of the other, the question arises whether a pragma on rule also applies to the others... 19:01
sort of the "is deep" thing.
pmichaud I could see it going either way 19:02
TimToady If we follow "is deep", then default is syntactic, and you use :deep or some such for semantic mods.
optimized that way obviously because :deep is shorter than :shallow :) 19:03
but also because part of the reason for "parse" in the first place is so that you can think of it as its own thing. 19:04
PerlJam hands the parse to TimToady
TimToady *phrase
pmichaud I think I need to disappear from IRC for a while to get some actual coding done -- anything else for me before I /away? 19:05
PerlJam token/phrase doesn't quite mesh for me. Those two words live in different mental buckets.
aufrank one nice thing about phrase is that it corresponds to tree or subtree in syntax
FurnaceBoy_ true 19:06
aufrank (in that case I would think of tokens as either words or terminals or nodes)
PerlJam I know, let's go with word/stanza/poem instead of token/phrase/grammar ;-) 19:08
audreyt phrase sounds good.
aufrank PerlJam: then we get sub-word distinctions as well (foot, iamb, spondee)...
xinming role A does A { }; class B does A; B.new 19:09
what's the result ? :-)
PerlJam xinming: an error because "role A does A" doesn't make much sense. 19:10
or perhaps a warning and it works anyway depending on your level of stricture.
xinming shakes hands with PerlJam
TimToady token/syntax maybe...
xinming PerlJam: It should be an error... or compiler automatically ignore this.
audreyt syntax error { ... }
xinming hmm, I don't know though...
audreyt "syntax" means "phrase structure", right? 19:11
(in natural, instead of computer, languages)
PerlJam audreyt: right.
TimToady It means "BNF production" in computerese
but yes, syntax can mean a set of stuctures, not just one 19:12
audreyt but we have alternation etc
xinming audreyt: which part should I write test? I mean, test `role A does A;` part, or B.new part?
audreyt so we can say it already is representing syntax
TimToady maybe we live with token/phrase, but they're in different mental buckets for me too.
audreyt instead of just one possible phrase arragement
I think "syntax" reads better 19:13
as a plus, it has no verb form.
pmichaud syntax if_statement { if <expression> <block> }
audreyt while "phrase" is both vt and vi
and suffers from the same problem as "parse"
TimToady and the multilevelness of it doesn't really bother me, since
PerlJam audreyt: you know us computer geeks though, we can verb anything ;)
TimToady the top refers to the bits of it anyway.
syntax top {...}
audreyt 21:13 < audreyt> but we have alternation etc 19:14
TimToady alternation doesn't backtrack either
audreyt and subrules kinda means named explicit structure
TimToady in syntax
audreyt I was agreeing with you :)
(in saying that the syntax refers to the structure not just one phrase)
pmichaud syntax would work for me
TimToady that's what you get for typing in the wee hours of the morning.
audreyt indeed
I should sleep :) 19:15
TimToady going, going, gone...SOLD!
audreyt spent the whole night parsing Artistic 2.0b14.
19:15 Arathorn is now known as Aragone
audreyt it... reminds me of Perl 6 in many ways, compared to Artistic1. 19:15
pmichaud could we have :syntax instead of :ratchet, then? Or is that too much aliasing?
PerlJam would like :syntax over :ratchet 19:16
particle_ yes, nurse ratchet.
19:16 asdddd joined
audreyt I have no idea what ratchet means. 19:16
19:16 asdddd left
PerlJam (in fact, anything would work better than ratchet I t hink) 19:16
pmichaud "ratchet" in general means that once you take a step, there's no going back
audreyt pmichaud: webster doesn't list that... 19:17
wordnet lists that meaning, under the verb slot
which is even more confusing.
aufrank wordnet++
particle_ :oneway
audreyt oh, as "rachet up" compound.
TimToady :syntax doesn't work for that, since token also ratchets.
pmichaud google "define:ratchet"
PerlJam aufrank: have you ever seen the innards of an old-fashioned clock?
TimToady and I wanted something very attention getting.
PerlJam er, audreyt
audreyt PerlJam: yeah. but I lack the English vocab :) 19:18
(or rather, the cultural metaphoric context)
TimToady likewise for :panic instead of "try"
pmichaud okay, I'm not strongly opposed to :ratchet -- just wanted to (briefly) explore the possibility
I think :panic is appropriate :-)
audreyt TimToady: so maybe :Legacy instead of :Perl5 ;)
TimToady it seems like a lively metaphor.
pmichaud audreyt observed on #parrot that ":panic" is likely to be the $& of Perl 6 :-) 19:19
TimToady Thats falling into Pythons .repr trap.
*which* legacy?
svnbot6 r10045 | yiyihu++ | role A does A { }; class B does A; B.new;
r10045 | yiyihu++ | This will cause infinite loop in pugs, a partial test...
audreyt true
aufrank the contextually relevant one, of course
pmichaud away again
TimToady all the more reason to make :panic something with strikingly negative connotations.
pmichaud yes, that's why I think it's appropriate :-) 19:20
TreyHarris TimToady: I'm confused about "rule ident { [ <alpha>: | _ ]: \w+: }". Why is there a colon after <alpha>, but not after _?
pmichaud TreyHarris: the colon prevents backtracking into the <alpha> subrule
TimToady cause you can't backtrack a literal anyway.
TreyHarris yes, but wouldn't token do the equivalent of putting a colon there, even though it's useless? 19:21
TimToady and if you pass the literal, the []: prevents backtrack to _.
pmichaud TreyHarris: we don't know that it's useless :-)
audreyt I think the jargon is :predictive. 19:22
TimToady but yes, following the ":"-anything-that-progresses rule might be clearer.
audreyt (as the antonym to "backtracking")
pmichaud I was thinking "committed"
PerlJam audreyt: It would almost work as :backtrack(0) and :backtrack except that usually you want the opposite sense.
TreyHarris apologies if i'm being pedantic, but I was comparing the "token ident" line and the "rule ident" line looking for the mechanistic behavior of "token", not the semantic result 19:23
TimToady that's why I said it might be clearer.
pmichaud TreyHarris: np -- I understand your question better now
yes, it probably would make more sense w/a :
TimToady it's the "is deep" thang again...
audreyt :predictive and :committed both sound like computerese :) 19:24
just like "syntax"
but unlike "parse", "phrase", or ":ratchet"
TreyHarris audreyt: lots of "syntax" in the non-computer world. there's a syntax on liquor, syntax or tobacco.... :-) 19:25
particle_ :committed and :panic have other meanings, to asylum-seekers
pmichaud perhaps just :commit and :panic
PerlJam :commit plays well with <commit> :)
svnbot6 r10046 | yiyihu++ | Rename role_does_itself.t to self_inheritance.t, as I forgot the
r10046 | yiyihu++ | class A is A { }; A.new; test++
PerlJam though if we do too much of that people will derive a general rule that just isn't true. 19:27
audreyt . o O ( :chill and :panic )
PerlJam heh
TimToady ooh.
PerlJam :hhgttg 19:28
TimToady that works on so many levels...
PerlJam :chill ? 19:29
TimToady As you chill toward 0C, you're reducing your possible states.
pmichaud (0K)
TimToady that too. :)
till all you have left are the one's allowed quantumly. 19:30
which is like *+ and friends.
pmichaud oooh
Juerd eh, *+?
audreyt "formal; distant; lack of enthusiasm or activity" 19:31
Juerd Either 0 or 1, or more? :)
pmichaud *+ forces * to be a backtracking quantifier
Juerd Ah
TimToady not been following p6l have we?
PerlJam was thinking that :relax (the not so literal meaning of chill) would work too
pmichaud as opposed to *: which is greedy but doesn't backtrack
audreyt PerlJam: but it doesn't relax the parse...
Juerd TimToady: 402 Tuits required
aufrank wonders if someone will come up with a connection b/w chill and freeze
audreyt it makes it stricter
PerlJam audreyt: no, it's a command ot the user :)
audreyt heh ok :)
TimToady except :relax has connotations of allowing more freedom, not less.
audreyt :chill is like, discourage your freedom 19:32
TimToady :chillout
particle_ :veg
audreyt :kewl
TimToady use ingy;
particle_ :peace
audreyt no way;
PerlJam Just use a parameterized :entropy
audreyt (I like the fact that :predictive is very long :)) 19:33
pmichaud btw, :backtrack works, since the default is backtrack and :backtrack(0) means "don't backtrack"
TimToady :blackhole(•)
pmichaud then token is rule :backtrack(0)
audreyt but that reads to me asid you are saying backtrack over 0
pmichaud sorry rule:backtrack(false) (or however we define "false") 19:34
audreyt rule :backtrack('')
TimToady hmm, :blackhole would never do...either in Russia, or on slashdot...
particle_ rule :-backtrack
audreyt that feels like double negatives...
unless... ooh 19:35
"token" is the base
"rule" is token :backtrack
"syntax" is token :words
TimToady i dunno. I kinda like that rule is simple like rx. 19:36
PerlJam Could what particle said work as a general syntax? :thingy == :thingy(1), :-thingy == :thingy(0)
TimToady so far we've consistently gone with :foo(0) for the negative form.
pmichaud perhaps :!thingy (but I still prefer :thingy(0) ) 19:37
19:37 froh-doh joined
audreyt in two places: :w(0) and :c(0), both in S02. 19:37
particle_ :-foo or :!foo could mean flip-flop the previous state 19:38
audreyt but (0) is fine... curiously, :w<> works also.
PerlJam the parentheses are good for general parameterization, but when you're talking about just "on" and "off", it seems like a simpler syntax would be better.
particle_ s/flip-flop/toggle/
PerlJam Of course, if :-thingy were valid syntax, it begs the question of what :-thingy(1) would mean (if anything) 19:39
s/-/!/ if you like
TimToady I think :!foo reads much better.
audreyt :!w is natural if that's the case...
TimToady then :foo is the same as :?foo I suppose...
:)
audreyt still, feels strange to have boolean named arguments that default to true if you don't pass in false 19:41
but maybe that's just because I havn't got used to :!foo.
TimToady I avoided the Unix shell -foo syntax precisly because of its negative polarity, but it almost seems to be reinventing half the mistake if we don't have a negative form.
particle_ well, p6 defaults to 'use strict'
audreyt so what does :!foo(3) mean? 19:42
TimToady would be illegal, I presume.
audreyt :foo(!3) ?
PerlJam audreyt: I'd say it's a mistake.
audreyt cool
TimToady In fact, prevents scanning for an argument entirely.
PerlJam sounds perfect 19:43
particle_ so, warning if it finds one?
PerlJam particle_: error!
TimToady Does :?foo prevent :?foo(0) then?
particle_ yes
audreyt so they pass in boolean true and false
particle_ then + and - may be more appropriate 19:44
TimToady what I was thinking.
particle_ ? doesn't look true to me
TimToady me either
audreyt but + has twigil meaning
and I don't think we need a :?foo form
unless you want to use that to solve the space-bracket problem
q:?w(...)
TimToady yeah, if you're gonna add an extra character anyway
just add a space after :foo
audreyt right. 19:45
if you want to print 1
just print 1
it's !1 that takes typing :)
ingy feels used
TimToady no;
ingy thx
audreyt tested, implemented, committing. 19:46
ingy o/~ there's no ingy like show ... o/~
TimToady have to run off for a bit. I think things are pretty settled, at least on the surface...
biab& 19:47
audreyt foo(:!a :b :!c) 19:48
legal syntax, yay
xinming audreyt: what does that mean? 19:49
19:49 heng joined
xinming foo(:!a :b :!c) *_* 19:49
audreyt foo(a=>0, b=>1, c=>0)
heng may I have a perl 5 socket question in this channel ? I asked already in #perl , but no one answered yet, hehe
merlyn please dont't ask in #perl6
heng sure, thanks, got ur msg in #perl, thanks, :-) 19:51
svnbot6 r10047 | audreyt++ | * Implement :!foo form, which is like :foo except the value is false.
r10048 | audreyt++ | * 01-sanity tests: back port test renumbering fix from parrot. 19:54
19:57 Barry joined
audreyt reads on Artistic 2.0b14 more, and is filled with the desire to relicense Pugs to the SQLite 3-clause blessing. 20:00
(as well as all my CPAN modules.) 20:01
20:02 pmurias joined
pmurias hi 20:02
xinming audreyt: why? :-)
audreyt xinming: license compatibility issues
pugs currently include a lot of 3rd party code 20:04
and each one needs to be rechecked against Artistic 2.0b14 for compatibility
plus, I'm pondering adding LGPL'ed code
in particular Judy
and allison in #parrotsketch warned that it will conflict with placing Artistic license restriction on the entire pugs tree. 20:05
xinming audreyt: but, the will the problem gone after perl 6 can be self-hosting?
PerlJam xinming: pugs != perl6
xinming hmm, I should say pugs...
PerlJam: sorry. :-)
audreyt xinming: I doubt we will write all runtime parts without any C bits...
not any time soon
and I'm not really so skilled in C that I can reimplement judy, or syck, or other C libraries we use, from scratch 20:06
PerlJam goes back to figuring out why his RT setup isn't working.
audreyt and also because I really like SQLite :) 20:07
aufrank can't wait for the SQLite backend in the Mozilla suite
audreyt it's already in Firefox2 preview
so, in short, the SQLite noncopyright is the only way that's always compatible with whatever license of contrib/ tree uses 20:08
and there cannot be copyright disputes where there is no copyright :)
(also, Taiwan is not part of WIPO, so I'm entirely operating outside the native jurisdiction anyway)
xinming audreyt: what does syck do? 20:09
aufrank xinming: yaml parser
audreyt xinming: it parses and dumps YAML.
gaal_ meows
20:09 saorge joined, gaal_ is now known as gaal
aufrank audreyt: are you using the ff2 preview as your day-to-day browser? 20:10
audreyt aufrank: no, not yet 20:11
xinming audreyt: Is perl 5 version of yaml parser based on syck?
audreyt xinming: I wrote YAML::Syck.
ingy wrote a pure perl5 YAML.pm long before that.
(which I believe is the first YAML implementation in existence... or something close to that)
xinming audreyt: what's the license of ingy's version? 20:12
audreyt xinming: Perl
Artistic1/GPL 20:13
xinming wonders if ingy's version isn't as powerful as C's version except the speed...
audreyt it is more featureful 20:14
certainly has more OO APIs.
xinming why not use ingy's version? :-P
obra It's slower ;)
lots slower 20:15
audreyt you need to embed perl5 for it :)
so takes a lot of memory as well.
also, you'd need to coerce your perl6 data into SVs
which adds another layer of overhead.
clkao it was also broken for lots of edge cases
xinming look forwards perl 6's version... 20:16
s/forwards/forward to/
xinming thinks it's a pain that there are so many licenses... 20:17
ingy use YAML::3000; 20:18
lichtkind ?eval imkreis: 20:20
$zaehler++;
$zaehler > 10 and goto frei;
goto imkreis;
frei:
20:20 evalbot_10044 is now known as evalbot_10048
evalbot_10048 Error: unexpected "i" expecting program not a class name or Only one invocant allowed 20:20
xinming lol... I think you should write them in one line 20:21
lichtkind ?eval imkreis: $zaehler++; $zaehler > 10 and goto frei; goto imkreis; frei: 20:23
evalbot_10048 Error: unexpected "i" expecting program not a class name or Only one invocant allowed
xinming $zaehler not declars...
lichtkind ?eval my $zaehler; imkreis: $zaehler++; $zaehler > 10 and goto frei; goto imkreis; frei: 20:24
evalbot_10048 Error: unexpected "i" expecting comment, ";", Doc block, block declaration, declaration, construct, expression or end of input not a class name or Only one invocant allowed
lichtkind ?eval my $zaehler; imkreis:; $zaehler++; $zaehler > 10 and goto frei; goto imkreis; frei:
xinming lichtkind: I don't think goto has implemented in pugs...
evalbot_10048 Error: unexpected "i" expecting comment, ";", Doc block, block declaration, declaration, construct, expression or end of input not a class name or Only one invocant allowed
xinming Loading Prelude... done.
pugs> goto
*** No such sub: "&goto"
at <interactive> line 1, column 1-5
pugs>
lichtkind ok
20:30 DaGo joined
lichtkind but after goto i dont need to set ; ? 20:33
hust like in perl5
just
20:36 qu1j0t3 joined
xinming lichtkind: I think so, but It's not specified yet. IMO. :-) 20:36
lichtkind and block naming is without : ?
thanks xinming
xinming lichtkind: no, I think it might be.... `BLKNAME: { .... };`
lichtkind but if goto blk... than the block will be executed? 20:37
xinming otherwise, base words are considered a function call.
lichtkind: I think so.
as I asked earilier... how to implement if... if $a == $b && $c == $d { ... }; will become.... 20:38
$a == $b && $c == $d && goto &BLK_passed to if...
lichtkind: but I'm not sure... you'd better ask someone who is sure about this. :-) I'm still a newbie... :-P 20:39
lichtkind but you seem longer here than i :) 20:40
xinming you are chinese?
xinming longer time does prove anything...
lichtkind: yes, I am, In mainland.
lichtkind xinming cool im big fan of chinese culture 20:41
and have some years of tai chi practice
xinming you hack on pugs? 20:42
xinming lichtkind: In fact, if you learn perl 6, I'd say.... perl 6 is a bit like Chinese, and C is like English. :-)
lichtkind: No, I just write some testing.
BTW, `xinming otherwise, base words are considered a function call.` <-- here I'm wrong, not base words, In fact, I wish to type bare words. 20:45
lichtkind xinming i hope your not scared but im also big fan of falun dafa
xinming lichtkind: I'm not, as I don't have religon... :-) 20:46
lichtkind: are you american?
lichtkind xinming its no religion just a sort of qi gong that i even dont practice but i like their teachings 20:47
theorbtwo xinming: I think Larry has mentioned Japanese as an influence on p6, and I should expect they're very similar, at least on the level that p6 can be similar to a human language.
lichtkind xinming no im east german
partly czech
xinming theorbtwo: In fact, Japanese Characters are learnt from Chinese. :-) 20:48
theorbtwo xinming: I know.
lichtkind: How old are you?
lichtkind 30
in a month
but im looking like 24 :)
and in the heart im 5
theorbtwo So it's not been east germany for slightly over half your lifetime. 20:49
theorbtwo is only 25.
lichtkind www.perlmonks.org/index.pl?node_id=275692
theorbtwo My sister remembers the wall falling; I don't.
lichtkind theorbtwo its me giving my first talk
xinming theorbtwo: why I have that kind of feeling is all because the context... but this feeling isn't so strong when I learn programming in perl 5. 20:50
lichtkind theorbtwo your german too?
xinming is only 23
20:50 fglock joined
lichtkind xinming what you mean? 20:51
xinming doesn't know if he is the youngest here.. :-) as Juerd is 5 month older than xinming
lichtkind: I'm 23 years old...
lichtkind theorbtwo i was 13 when wall came down
theorbtwo No. My sister watched it live on TV from her German class. I'm 25, and from Lancaster, PA, USA.
lichtkind xinming no i meant your previous sentence
theorbtwo I did live in Munich for 2 years, though.
rw-rw-r-- is also 23.
pmurias i'm probably the youngest pugs commiter
gaal unlikely. 20:52
theorbtwo nothingmuch is quite young.
gaal don't we have a commiter who got their bit on their birthday? 20:53
like, the 0th one
xinming lichtkind: perl is a context sensive language, which means, what it returns, It is all about what the caller want(eg, assignment), if you read more about synopsis... you will know the want function, :-)
pmurias i'm 16, nothingmuch is 2X 20:54
lichtkind uuuh i dont know the want function ..should read this
xinming My birth is 1983.11.05 :-)
....
theorbtwo BTW, you might want to change easter(n) to east(ern) Germany.
pmurias xinming:1989.12.24
xinming young enough. :-) 20:55
buu So who is writing the builtin christmasy_day function?
xinming needs to go bed... It's 05:00 now... 20:57
xinming waves...
bye all.
gaal night
theorbtwo G'night, xinming.
pmurias night
buu: i can write the christmas_eve function 20:58
:)
buu pmurias: Excellent.
pmurias: We have to compete some how.
lichtkind xinming im also not much into religion but i believ in god 20:59
xinming or rather to say i know he is there
xinming needs to finish talking about the god... 21:00
I believe we are the God... No one else... :-) 21:01
If our technolegy reaches a level which is high enough, we can make something we never see... and we can also make another planet full of creatures, and let themself do the rest of evolution... :-) 21:02
amv i'm constantly getting this spam which says: "she needs a better sex" 21:03
xinming maybe sound crazy... :-)
amv: you can tell him(it) "Fucking too much is bad for your(her) health" ;-) 21:04
lichtkind xinming its true god is within us
pmurias amv: always ask the spammers to phone back :) 21:05
amv but i'm a bit puzzled with it because i always thought female _is_ the beter sex ;)
but they are selling viagra, so probably they just need more customers and thus would like all females to turn into males.. 21:06
lichtkind amv no im glad to be male
xinming amv: I think you can ask his(her) phone... and advertise it on msn-space... :-) Kind of DDOS. ;-) 21:07
amv lichtkind: but you can't give birth.. doens't that make you kind of inferior? soon they will be able to produce semsn from female dna and males become completely useless.. 21:08
Juerd Useless? 21:09
Or redundant?
lichtkind amv no because you need both
Juerd Some redundancy is good.
Juerd is redundant
amv Hmm.. why don't partnerships have redundancy? 21:10
they always have only two single points of failure
xinming amv: I don't think so, If there is a possibility to make people through the lab... I'd prefer the natural one...
Juerd If there are two, they're no longer single ;)
21:10 Ann joined
amv Juerd: a point there ;) 21:11
Juerd Hello, Ann
Ann Hi there
xinming really goes to bed...
Ann: Hi, have a good time here...
xinming sinks...
amv xinming: but why would you prefer a person born from the dna of a male and a female over a person born from the dna of two females?
theorbtwo Juerd: It's even worse then a single point of failure -- it's two points, either of which can cause catastrophic failure. 21:12
Ann Thx, just listening in... Only about in the middle of the Alpaca :-)
pmurias amv: actually genders developed for a pourpous it avoids mitochondria wastage 21:13
amv pmurias: could i have that in english? =)
pmurias purpose 21:14
amv what is mitochondrial wastage and how having two genders avoids it?
pmurias you only take mitochondrials from your mother 21:15
if you took them from both parents, they would evolve arms and engage into bloody conflict
Barry Could someone explain what the purpose of perbot's phone function was when it worked?
obra does phone msg people on your behalf, barry? 21:16
pmurias there for semen don't have so much mitochondrias, and are cheaper to produce
obra freenode has crazy /msg policies to prevent spammers
Barry perlbot help
perlbot (fact)::tell (who) about (what)::(what) > (who)::learn (what) as (info)::relearn (fact) as (info)::phone (phone number)::shorten (url)::shorten it::search (module)::docs (module)::perldoc -f (function)::jargon (term)::math::fortune::flip::host (type) (record)::rot13::roll (die)::tempconv (temp)::scramble (foo)::8ball::slap::diss::what time is it::top/bottom (number) karma::geoip (ip)
obra oh. wacky 21:17
Barry You feed it a telephone number, so...I don't know
pmurias once you have to times of gamets you have gender specialisations
theorbtwo perlbot, phone +18005551212
perlbot Bad number: +18005551212
amv pmurias: but why couldn't a female produce also semen so that two females could impregnate each other?
theorbtwo perlbot, phone 18005551212
perlbot Bad number: 18005551212
theorbtwo perlbot, phone 8005551212
perlbot That function is broken for now, try back later
Barry Good number though 21:18
pmurias s/times/types
21:18 ribo joined
obra barry: chrisangell.com/incoming/chrisbot/v.../Common.pm 21:18
does reverse lookups
theorbtwo I suspect it's for reverse-look... ah.
Barry Oh. 21:19
21:19 ribo left
amv pmurias: (if we are thinking of an evolutional counterpart for the act of artificially extracting dna from a female to impregnate an another female) 21:19
Barry Defunct functions should have better documentation.
21:19 fglock joined
pmurias we could put the extrac mitochondrias to some other use :) 21:19
amv is there a need for cromosome y? =)
pmurias s/extra 21:20
amv: it contains mostly junk, but it activates other genes
amv but i think i'll stop talking because it seems that all i can do is derail conversation ;)
SamB so is y chromosome like key file? 21:22
pmurias i'm not really an expert, but it's sort of
actually most of the genome are comments
SamB well, might be 21:23
pmurias containing dead code
SamB it could be that they are padding or something
or unneeded library functions
pmurias evoultion is a patchwork job
it uses often cut & paste 21:24
SamB they should add a DNA mode to IDA
fglock re: a word for "token": "production" "subrule"
theorbtwo production and subrule are more general.
In purticular, anything that you can write in rule syntax is a production.
pmurias that could be a argument against Inteligent Design, God would have used source control, instead of keeping old code in comments 21:25
SamB even with source control you sometimes keep old code in comments...
21:26 Barry left
fglock theorbtwo: colabti.de/irclogger/irclogger_log/...l=325#l510 21:26
pmurias he surly wrote a code generator, and that's why there's so much redundancy
SamB I thought it was to guard against damage
and encourage microevolution and stuff 21:27
pmurias it actually can help in adaptation
amv pmurias, intelligent designers would counterargue that all source control systems of that time were proprietary since free software movement didin't exist yet
SamB hahah
pmurias *bang*, archaic genome sequence for long fangs activates (*let the suckers run away*) 21:28
21:28 ruoso joined
pmurias night 21:28
21:29 larsen joined
SamB anyway, God seems to have used some sort of distributed source control system 21:29
pmurias sleep& 21:30
21:30 aufrank joined
amv i'd liket o get my hands on that - it must be quite good since all branches of human are still easily integratable 21:30
aufrank did TimToady react to damian's regex/token/rule suggestion at all in the channel? 21:38
pmichaud aufrank: I didn't see anything 21:41
aufrank is keeping score at home 21:43
cognominal audreyt: I am studying if siva and parrot objet can be merged, The problem is that implementation wise, a parrotobject is already an array. It seems that an special pmc object could be baked with a siva pmc as unique attribute so that an array access would be rerouted to the siva pmc. 21:47
I am not sure yet if this can be done. I would like to avoid to touch anything beyond that parrotobjectwithuniqueattribute pmc and the siva pmc 21:50
PerlJam What is siva?
theorbtwo Siva is the implementation of captures in parrot.
PerlJam Why is it called siva?
cognominal that a pmc I have created that has the hash, the array, int and string natures 21:51
PerlJam Oh, I see.
cognominal ...because it juggles with many things
audreyt would like to use it for Captures, I an not sure it is possible.
PerlJam aufrank: who is winning? :-)
cognominal I designed it for nodes of trees, not as real objects 21:52
so I try to see if we can get both
PerlJam cognominal: don't captures potentially have an invocant slot? where would that fit in siva?
cognominal PerlJam: yes, that's what I say about siva not being object 21:53
21:57 larsen_ joined
cognominal btw: in the parrotobject, the array is used to hold attributes 21:57
22:02 stclare` joined 22:05 soisoisoi joined
meppl gute nacht 22:11
cognominal bonne nuits les petits :)
Juerd "hello i feather please i not perl 6 but need server" 22:30
clkao lol 22:31
Juerd The best thing I can do is make fun of it here, because the email address in the From: header bounces -- over quota.
aufrank Juerd, I not perl 6 either but need dinner please i 22:33
clkao 22:34
Juerd Hehe :)
clkao dccs trotters to aufrank
22:48 Limbic_Region joined 22:55 Southen joined 23:01 Quell joined 23:05 azuroth joined 23:10 ruoso left 23:20 Khisanth joined 23:34 cognominal_ joined
pmichaud TimToady: ping 23:43
pmichaud wanders off 23:59

Powered by Cro and the Raku® Programming Language.

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

Thank you!