6.2.7 released! | pugscode.org/ <Overview Journal Logs> | pugs.kwiki.org | paste: sial.org/pbot/perl6
Set by autrijus on 13 June 2005.
mugwump works far too late at a job that jumped the shark about 6 months ago 07:20
svnbot6 r5127, pdcawley++ | Added test for nested, named subroutines. 07:31
kungfuftr morning all 09:37
mugwump hi scott 09:38
mugwump ducks off & # ... damn pre-launch evenings 09:39
gaal|work ping -1 12:35
pugs -CPugs produces output that contains things like 12:36
envGlobal = <ref>
how is this supposed to work? 12:37
Limbic_Region any LiveJournal users out there?
gaal|work raises his hand
need anything, L~R? 12:38
The <ref> thing breaks compilation, because it gives
# /home/roo/src/pugs/MainCC.hs:434:252: parse error on input `<' 12:39
Limbic_Region do you follow the perl communit journal there?
gaal|work no, i think i only visitied once or twice.
Limbic_Region I dropped my LJ some time ago for personal reasons
but I have been reading the perl community journal there off and on for a while 12:40
gaal|work k
Limbic_Region I have noticed quite a bit of linking to O'Reilly copyrighted material
as well as some flat out lies about p6
www.livejournal.com/community/perl/102202.html
I did what I could as Anonymous Coward - but think that registered members might have more weight in their responses 12:41
gaal|work FWIW, LJ's origianl author, Brad, is pushing hard a new decentralized identity system, so you don't strictly need an LJ account to make identified contributions. 12:42
openid.net
12:44 calvin__ is now known as calvin_
gaal|work what lies do you see there? i think you may be taking it a little harshly. misinformed opinion maybe, but where's the lie? 12:44
Limbic_Region gaal - not necessarily in that post - general trend over time 12:45
gaal|work *shrug* i can't make everyone enlightened. 12:46
Limbic_Region nor can I
gaal|work (that was an understatement)
Limbic_Region I was just wondering how many people here were LJ users
and either didn't bother or were unaware with the community 12:47
gaal|work L~R, does pugscc work for you? script/pugscc --pugs -e "say <hi>" --run 12:50
I think it broke, but I want to know where
we need this for the precompiled prelude thing 12:51
Limbic_Region I have been compiling since I got in this morning - 55 minutes so far (and that was only taking into account changes in the last 24 hrs) 12:56
when it finishes, I will let you know
pdcawley ?eval my %hash = ( if => 'then') 12:59
evalbot6 {('if' => 'then')}
pdcawley coo... id've thought that would have been a pita to parse.
QtPlatypus It will be, at the moment => stringafies its LHS. 13:00
gaal|work anyone around with a working pugscc? 13:02
(thanks, l~r!)
Limbic_Region gaal - turns out that pugscc isn't recognized by Win32 (though I distinctly remember running it in the past)??? 13:25
gaal|work beh. can you try to run it with perl explicitly? (it's a p5 script) 13:26
i have to leave now for a few hours
but really i think the compiler itself is broken. the problem needs to be backtraced 13:27
anyway - thanks - bbl
stevan_ jabbot: seen autrijus 13:31
jabbot stevan_: autrijus was seen 7 hours 24 minutes 57 seconds ago
coke__ skips the huge review and hopes he didn't miss anything. 13:53
svnbot6 r5128, Stevan++ | Perl6::MetaModel - More refactoring before the plane back to the states
r5129, Stevan++ | Perl6::MetaModel - The last vestiges of the old republic has been swept away! The meta model is now clean (well except for roles, I will do those on the plane home) 14:12
autrijus gaal: yes, pugscc is broken like that. will fix 14:40
pdcawley Have we got any design/tests on querying variables for their traits (or do I mean properties)? 15:03
kolibrie I think those are in t/var 15:16
Liquid hello 15:44
kungfuftr castaway_: OI!
Liquid I'm trying to get mortgage rates from a website to put on to my website and I dont know how to go about it, I need it so it updates automaticly 15:45
coke__ and you want to write it in perl6?
Liquid Well
I'm willing to write in what ever works
*it
coke__ "this is not the channel you're looking for" 15:46
Liquid Thanks
Any suggestions?
osfameron perlmonks.org # not irc, but possibly more helpful 15:47
Liquid Thank you
Y'all have a good day
Limbic_Region also #perl 15:50
osfameron but that's more of a social clubi
people seem to get a bit bad tempered about technical questions
(sometimes)
Limbic_Region happens at the Monastery too - though less often 15:51
osfameron really? I've not looked at anything except weekly best for quite a while now
wolverian efnet #perlhelp is quite helpful at times 15:52
Limbic_Region osfameron - heh - you should see some of the flamewars 15:53
QtPlatypus keeps thinking of the Monastry as the 'Scary devil Monastery' and they dislike technicall questions.
Limbic_Region but for the most part, the noise:signal ratio at the Monastery is quite good
coke__ pmi? 15:54
will perl6 support recursive grammars? 15:56
(rules defined in terms of themselves?)
QtPlatypus coke__: I beleve they should.
coke__ which variant of should is that? =-) 15:57
QtPlatypus SHOULD
nothingmuch morning
coke__ I'll clarify my request for clarification. =-)
is it in the p6bible, do you know? are you saying they should because you think it's nifty? because @larry said so? 15:58
or, more simply, I suppose, are you speaking from a position of authority in such matters.
kungfuftr coke__: a gramme is a rule... and rules are recursive.
so... all things being... i'd have thought so 15:59
QtPlatypus I'm saying that it should because it's nifty and without recurvive rules you wouldn't be able to define alot of usefull grammers.
coke__ danke. now to figure out if patrick's actually made it work yet. 16:00
Limbic_Region is finally at a point to start working on condensing his ball balancing code but needs to go to lunch
food &
nothingmuch seen autriju 16:01
jabbot nothingmuch: I havn't seen autriju, nothingmuch
nothingmuch s
seen autrijus
jabbot nothingmuch: autrijus was seen 1 hours 21 minutes 36 seconds ago
nothingmuch hmm... i wonder if stevan & luqui gave autrijus my tripod. 16:02
nothingmuch forgot it in their room
QtPlatypus Isn't that the diffrence between regular expression and a rule anyway?
coke__ I'm trying to get at this from inside parrot, not from pugs, so I'm not entirely up to speed on the p6 aspect of things. 16:04
Anyone interested in writing a perl6 grammar for tcl expressions? =-) 16:06
kungfuftr sicko 16:07
coke__ yes?
Odin- Is that sick? :p 16:08
kungfuftr then again... i was thinking about introducing some ecmascript stuff into perl6.. but anyhows 16:09
Odin- Again ... is that sick? ;p 16:10
kungfuftr probably... but i'm biased
coke__ well, if anyone *is* feeling sick, the "spec" is at www.tcl.tk/man/tcl8.5/TclCmd/expr.htm . =-0
Khisanth not sick, bored maybe ... 16:13
kungfuftr coke__: might be more intresting to auto codegen using lex/yacc files (can't remember the name of the process) 16:15
nothingmuch oh crap, poor autrijus 16:16
kungfuftr bnf even 16:17
BNF grammar -> perl6 grammar 16:18
coke__ I don't want to require lex/yacc. 16:20
QtPlatypus BTW do you know if there is any plans to back port PGN into perl5 so we can avoid the cost of Parse::RecDecent 16:21
PGE
kungfuftr coke__: it would be a one-time operation with lex/yacc, etc 16:22
nothingmuch if autrijus comes in, tell him i left to pick him iup
integral PGE's PIR -> Perl6 -> PIL -> Perl5 ... 16:29
Odin- Wouldn't PGE run through Ponie? 16:33
kolibrie should, since it's all in parrot, and not tied to perl6 16:34
coke__ I'm tempted to implement a wrapper for this in partcl. 16:46
geoffb appears fuzzily 17:07
oy, almost a full day of backlog to read
ackpth
gaal rehi 17:15
geoffb Hey gaal, what's up? 17:23
gaal all good. waiting for pugs to build so i can see if pugscc is fixed
geoffb :-) 17:24
God, it's like a badly run PBS telethon 17:25
gaal how so?
geoffb The last message from lilo -- why would the message need to be repeated to our server only?
oh wait -- you may not have seen it. 17:26
Duh.
gaal true
geoffb goes back to reading backlog on various channels
coke__ if I run pugs in interactive mode, type some crap and *tap* delete, it proceeds to delete the entire line one char at a time. 17:32
gaal OS?
oh, i forgot, it doesn't run in win32 anyway 17:33
coke__ os x
gaal but it looks like your readline lib is crazy
does, eg, the CPAN shell do the same? 17:34
coke__ works fine for everything else. =-)
gaal assuming you have the same readline there
coke__ I grabbed a pre-built GHC. that'd be my first guess as to borkage.
gaal FW(vl)IW, this worksforme on linux
coke__ is there a nopaste for this channel? 17:36
gaal perlbot nopaste
perlbot Paste your code here and #<channel> will be able to view it: sial.org/pbot/<channel>
pasteling "Coke" at 199.33.156.133 pasted "simple grammar" (9 lines, 165B) at sial.org/pbot/11455 17:37
coke__ that kills PGE. (using the demo.pir handler, which doesn't exactly use perl6 to define a grammar). Does that seem like a reasonable grammar? 17:38
gaal no idea. are mutually recursive rules allowed? 17:39
coke__ this is why I asked earlier about recursive grammars. =-)
gaal is rule-ignorant
gaal is backlog-ignorant too :)
coke__ alternatively, is there a non recursive way to write that.
gaal dunno :) 17:43
coke__ I don't see any tests yet for grammar. hurm.
gaal oh, too bad, pugscc is still broken :( 17:49
integral in what way is pugscc broke? 17:53
Limbic_Region gaal - on the bright side, I finally got my ball balancing code working and condensed
gaal integral: the haskell target produces code that doesn't compile
it has things like <ref> in the generated globC
can you fix that?
integral oh, I'd expect that that's known since Pugs.Compile.Haskell isn't using PIL (and the new Pugs.CodeGen system) 17:54
gaal L~R, the puzzle? cool
integral not something I really know how to fix
Limbic_Region gaal - it isn't a solution to the challenge - it is a solution to the puzzle 17:55
gaal Is PIL mature enough to compile everything no?
now
integral doesn't know
Limbic_Region perlmonks.org/index.pl?node_id=471405 # if you are interested
gaal thanks, L~R 17:56
hmm, well fwiw pugs -CPIL on Prelude.pm doesn't *croak*, but that doesn't say much
jdv79 what does it do? :) 17:57
gaal oh, my bad
on -e 'say <hi>' it doesn't croak
on Prelude.pm it does :)
grrrr, no, i'm completely dyslectic it seems! 17:58
diskless
or something :)O
but it doesn't emit anything remotely like an AST for the prelude (only four lines :) 17:59
hey, weren't semicolons at the end of where blocks once required? 18:00
when /moose/ { say <hi> };
(*when* blocks, eep)
brb 18:02
spinclad coke__, aside from mutual recursion, i see two issues with your grammar, and wonder if it's saying what you intend: 18:22
1. <expr>*<expr> should be the same as <expr>+. Should you excape the *? 18:23
wolverian gaal: not in perl6. when is a statement 18:24
gaal: dunno about pugs :)
spinclad 2. if you do, then A*A*A*A has five parses. Do you intend to parse uniquely?
coke__ bah. yes, I mean to escape the * 18:26
please explain #2, pretending I don't know anything. =-09
gaal wolverian: thanks
coke__ (e,r =-)
(changing that, I get a match in demo.pir, and can "next" until I exceed max recus depth in parrot) 18:27
spinclad A*(A*(A*A)) A*((A*A)*A) (A*A)*(A*A) (A*(A*A))*A ((A*A)*A)*A
gaal anyone know what autrijus' schedule is? i mean both for flights and for, eh, flights. 18:29
coke__ My ultimate goal would be to have that to the last one of your examples.
er, "have that translate"
gaal wow! $obj.perl is real neat 18:33
spinclad your grammar is ambiguous. fraid i can't speak about PGE here (haven't dug into it);
coke__ how can I disambiguate it?
(a pointer to a non perl6 way of doing so would also be fine)
spinclad some grammar engines have constraints on the grammars they'll accept 18:34
yacc and bison are the two i've used; antlr i expect; plenty of others i don't know 18:35
key buzzwords: LL(n), LR(n), LALR(n) grammars 18:36
as soon as you research these you'
coke__ pltplp.net/lex-yacc/example.html.en appears to have the sort of thing I want.
spinclad ll probably know more than me
kolibrie gaal: I think nothingmuch knows something about autrijus' flight 18:40
spinclad rule expr { <binary> }; rule binary { <atom> | <expr> \* <atom> } should be unambiguous
kolibrie but he disappeared in a hurry a few hours ago
gaal: luqui and stevan may also have known, but they left the conference four hours ago 18:41
gaal kolibrie, are there still talks on now?
coke__ the con ended yesterday, no more talks.
(i'm guessing people were hanging out and getting stuff done,though)
gaal oh, i thought for some reason this was day 4 18:42
3
kolibrie nope, yesterday was it
gaal guess i'm a little mixed up :)
well in that case he's probably asleep! :)
kolibrie possibly
or in transit 18:43
gaal yeah, that happens more often
coke__ spinddf~!~ 18:47
spinclad he slept 16 hours this morning
gaal long morning!
spinclad coke_ae76?
(or was that yestermorn? i've lost track) 18:48
coke__ my screen session went south, sorry.
hurm. with your suggested change, I can get PGE to bus error much more frequently. =-) 18:49
spinclad but at least unambiguously :) 18:50
gaal bbiab
spinclad (i gotta finish building everything so i can talk more than theory!) 18:51
svnbot6 r5130, iblech++ | * Usual svn properties added to various new files. 19:48
r5130, iblech++ | * Added "#ifndef HADDOCK" directive to src/Emit/PIR/ParrotObject.hs so Haddock
r5130, iblech++ | runs again.
eric256 returns after nearly two week without pugs.....soo long!!! ;)
Darren_Duncan especially in internet time
eric256 especially in pugs internet time 19:49
Darren_Duncan that's like a year in Pugs time
eric256 just finished reading the current page of journal....now to travel back....back...way back....to the page before the current page. ;)
coke__ bah. does autrijus's journal have its own rss feed? 19:55
gaal coke__, use.perl.org offers feeds. eg. syndicated on livejournal.com/users/autrijus_rss 19:56
eric256 it is are to read the journal, find last entry, read, scroll up and find previous entry, read down, scroll up agian....okay not hard, just weird. lol 19:57
s/are/hard/
coke__ goal - that looks like HTML. is there no actual feed ? 19:58
gaal that's a known quirk with onilne journals. some hosts have day pages you can go to a certain date and navigate forward from.
coke__ ah. that points back to feed://use.perl.org/~autrijus/journal/rss
thank you. =-)
gaal coke__, use.perl.org/~autrijus/journal/rss
yup
coke__ why that's not aggregated into planet parrot, I'llnever know. =-) 19:59
wolverian coke__: planet.perl.org 20:06
and planetsix.perl.org
coke__ ah. 20:07
ah well, I've grabbed autriju's, that's good enough for now. =-) 20:08
nothingmuch evening 20:09
gaal hey nm!
nothingmuch hi ho
what's up?
gaal dunno, but pugscc is down :)
nothingmuch will hack up consolidated smoke viewing as soon as he goes to the bathroom.
i'll tell autrijus when he wakes up 20:10
gaal kewl
when are flights scheduled?
wolverian www.geeksunite.net # help chip
nothingmuch mine arrives on June 3, 3:45 20:11
israel time
autrijus's leaves at around the same time, july 2nd, canada time
err, s/June/July/;
eric256 june 3? ehhh perhaps you mean July?
gaal when's pugs planned to fly? 20:12
kolibrie nothingmuch: did you find your tripod? 20:25
putter coke__: ping? 20:35
coke__ pong 20:36
putter hi, recursive rules are supported,
rule binary { <expr>*<expr>} should probably be rule binary :w { <expr> \* <expr>} 20:37
the :w means whitespace in pattern matches optional(? hmm, I better check that) ws in target. 20:38
there are grammar tests, some never enabled, some disabled recently when PGE was broken by the new parrot calling conventions. 20:39
coke__ and how do you specify the :w at the parrot level? 20:41
(since that's outside the rule itself)
putter I have drafts of non-pge rx engines in p6 and p5. the p6 ones are rather crufty due to current pugs limitations. there hasnt been much interest in grammars (your's is the first I've seen), so I have them all backburnered. if you'd like to push on something, like a tcl expr grammar, that could be fun...
coke__ putter: that's exactly what I'm writing. it's failing wonderfully! =-)
putter re :w, you can put it in the rule, but dollars to donuts, its not yet implemented by pge... 20:42
coke__ eh. I can make the whitespace explicit for now, that's not an issue. 20:43
the fact that I can't seem to make demo.pir do any kind of nontrivial recursion is, however.
nothingmuch kolibrie: yes I did, it was with autrijus after all 20:44
putter I don't know the state of pge re recursion. You can recurse in pcre (using names), and in p5 (using code returning regexes). 20:45
kolibrie nothingmuch: awesome
coke__ eh. PGE or bust.
putter re pge or bust, well...
If you have a grammar you would like to have work in p6 (as opposed to parrot), I can dust of enough of a p6 engine to drive it. Especially if namespaces have started working better. Won't get done today though. 20:47
s/dust of/dust off/
nothingmuch ugh.... Method::Alias uses ExtUtils::AutoInstall, which needs CPANPLUS 20:48
putter PGE is currently limited enough that it is straightforward to use try/die backtracking to do an re match.
err, limited enough ... to do an _equavalent_ ... 20:49
nothingmuch that's so overkill for something that simly does adds some subs to the caller of import
*sigh*
oh, CPANPLUS is optional
coke__ putter: I appreciate the offer, but I'm working in parrot, not perl6. I'll just keep bugging pmi 20:50
putter coke__: oh, and you can "use" : :: ::: cuts to prune grammars. Though I suspect pge hasnt implemented them yet. 20:51
kolibrie putter: I'm interested in grammars, but no hurry 20:52
coke__ putter: I'm enough of a noob that I don't get how that helps me.
ah. prevent recursion where possible? 20:53
putter re parrot v p6, ah, ok. let me know if you want a grammar, or a :parsetree, for regexps (to replace pge parse). that was the original objective of my... puttering. 20:54
kolibrie: :) Well, let me know if I can assist. 20:55
putter thinks grammars are going to end up being _very_ nifty...
coke__: re cut, managing backtracking, either for efficiency or control, is the biggy... so yes, you can use it to control recursion. 20:57
coke__ of course, being a noob, I still don't know where I'd put such things. 20:58
putter the cuts are basically "commits" with assorted "so, if you fail after reaching me, instead of just backtracking normally, fail back mumble far". 20:59
Dont worry about them to start with. When you find things that take horribly long to (not/)match, or you get multiple match results and you want to get rid of some of them, or..., then you can worry about them. 21:01
s/When/If/ # see also, premature optimization ;) 21:02
coke__: well, if you have any questions at all... I'll try to find answers... 21:04
coke__ hurm. I should write something that does what demo.pir does so I can keep my grammar as a perl6-looking thing, and import it into parrot automatically. Should make getting help from y'all easier. =-)
kolibrie putter: thanks, I'll holler when I'm floundering 21:05
putter its easy to create a string containg the code for a couple of rules, and yank them out with a regexp, generating whatever form you need. thats how I dealt with grammars not yet existing. 21:06
putter thinks of several fish puns, several involving dogs/pugs
I have to run, but do backread irc. 21:07
&
kolibrie pmichaud had nice slides of the current state of rules/grammars/pge, which he promised to put online 21:08
so someone can update S05
Darren_Duncan following wolverian's posting ... that's unfortunate what happened to Chip Salzenberg 21:13
nothingmuch irssix crashes too often 21:34
autrijus says: "i have a dream, i dream of languages judged not by their runtime but by their charectaristics" 21:37
and now he's going back to sleep
coke__ ah. pge seems to respect the :w modifier if you put it in as the first part of the rule. 23:07
coke__ wonders how to weld perl6 grammars into a tcl syntax. 23:18
Limbic_Region I saw your journal 23:19
you are looking at using PGE from within Parrot correct - and not Pugs?
coke__ <nod<> 23:21
coke__ decides to let the tcl people figure out the syntax they want.
kungfuftr 6.2.7 released! | pugscode.org/ <Overview Journal Logs> | pugs.kwiki.org | paste: sial.org/pbot/perl6 | Chip needs help!!! www.geeksunite.net 23:22
Limbic_Region kungfuftr - I am having trouble accessing the main site (it's been slashdotted) - perhaps include a mirror? 23:25
kungfuftr phoenixrising.ws/mirror/www.geeksunite.net/ 23:34
Limbic_Region I meant in the topic 23:36
but I think it will all settle down in a bit 23:37
getting slashdotted probably killed the main site
kungfuftr yar... i'm _sorta_ getting it 23:38
Juerd I wonder - is all the weird stuff with names *coindidentally* happening much in the perl community 23:49
Or does Perl attract weird people
coke__ this was discussed on lily earlier today. Exactly. 23:50
Juerd I can think of at least 4 examles, including myself
(chromatic, abigail, brian d foy, juerd (I use that as my first name, but it is not))
There are probably more.
Perhaps people even with normal looking names, who are really someone entirely different 23:51
Limbic_Region Dan Sugalski goes by Elian at the Monastery
Juerd that's just a nick, isn't it?
Limbic_Region oh - you mean beyond online handles
Juerd I'm more referring to people who use something weird as their rYea 23:52
wolverian that last word really freaked out here
Juerd oh goddamn, it's so annoying if irssi interprets typing-faster-than-the-connection-can-handle as a paste
wolverian set the autopaste thing off :)
Juerd I first wanted to type "I'm more referring to people who use something weird as their real name"
Then tried to erase that with ^U
wolverian how do you define weird in a multicultural society though? 23:53
Juerd Then acknowledged L~R's "oh - you mean beyond online handles" with Yea.
wolverian I usually use the arrow keys, in case I want to recall the line after all
that is slower, though.
Juerd wolverian: Well, using a nickname as your real name (to the extent of having it printed on/in books is rather extreme) or having a guide for capitalisation, or the combination of that, I find weird :) 23:54
Even though I do that myself.
wolverian: ^Y recalls a ^U-deleted line.
I must admit that when I read in scrotties book my own name as "Juerd Waalboer" I found it weird 23:55
Even though I myself chose to use that name, even on CPAN.
wolverian oh, well. yes. that is extreme. 23:56
I do prever my nick in all lc, so I guess I'm weird too, in a small way.
er. s,v,f,
Juerd I wonder if Perl does this to people, or whether Perl attracts these people, or if it's just all coincidence
Nah, just wishing for proper capitalization of a handle|nick is normal I think 23:57
wolverian oh, right. only if you wish it on your real name.
Juerd Although wishing that something is lc even at the beginning of a sentence is a bit weird :)
wolverian (or nick as real name, or whatever the other way around, maybe.)
Juerd: I like chromatic because of that. :) 23:58
Juerd Oh, I do like it.
But even though I myself do a similar thing, I find it weird
(With me, it just slowly got to be like this, up to the point where I sort of decided to keep it that way) 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!