Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2005-09-29

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
07:58 ilogger starts logging #perl6 at Thu Sep 29 07:58:44 2005
07:58 ilogger has joined #perl6
07:59 rantanplan_ has joined #perl6
08:00 buu http://erxz.com/log.pl
08:00 buu or something
08:02 buu Can't make the first day work, but is that what you had in mind?
08:03 spinclad good enuf in a flood
08:04 spinclad ^F is my friend
08:04 * nothingmuch forgot what ^F is
08:04 nothingmuch form feed?
08:04 buu Find in firefox?
08:04 nothingmuch oh
08:04 geoffb Mozilla "find in page"
08:04 spinclad yeh, that
08:04 buu Ahah, fixy first page.
08:04 * nothingmuch thought terminal
08:04 buu s/page/day/
08:04 nothingmuch http://colabti.de/irclogger​/irclogger_log_search/perl6
08:05 buu Interesting
08:05 autrijus rehi
08:05 nothingmuch hi ho
08:05 autrijus so I was hacking here in the hallways
08:06 autrijus and brendon just showed up suddenly
08:06 autrijus and I gave him a T-shirt and talked for half an hour :-))
08:06 nothingmuch brendon?
08:06 autrijus the one guilty for javascript and spidermonkey
08:06 nothingmuch ah
08:07 spinclad hi autrijus-san
08:07 autrijus he mentioned they are investigating the Rhino continuations class for spidermonkey
08:07 autrijus which would be lovely
08:07 nothingmuch woot
08:07 autrijus and also toying with the idea of having firefox work with system perl python ruby etc as XUL scripting
08:08 autrijus ...and I updated him on parrot -- turns out their impression was still two yearsD ago
08:08 nothingmuch that is local UI bindings, e.g. GTK, but based on the FF implementation
08:08 autrijus nothingmuch: right
08:08 nothingmuch s/$/?/
08:09 nothingmuch ∆ $parrot.at(Time::Piece->new("two years ago")), $parrot.now
08:09 geoffb hi, autrijus, nothingmuch
08:09 nothingmuch hi ho geoffb
08:09 dada has joined #perl6
08:10 nothingmuch there's a unicode char called 'n-ary circled dot operator'
08:10 nothingmuch perhaps that's hyper method call on boxed types?
08:10 geoffb nothingmuch, I'd be interested in that delta as well
08:10 geoffb autrijus, what did you tell them?
08:11 coral ooo, XUL scripting :)  autrijus++
08:11 spinclad (brendon++?)
08:12 * nothingmuch was actually wrong, i inserted "Increment" instead of "Delta" because i found the triangle under "
08:12 nothingmuch Mathematical Symbols" instead of the greek script
08:12 geoffb looked close enough in my window . . . .
08:13 spinclad i think you got it right by accident then
08:13 spinclad difference == increment in a way
08:13 nothingmuch really? why is it "increment"? I would have expected it to be named otherwise
08:14 * nothingmuch should have gone with |$parrot.now - $parrot.then|
08:14 buu Hrm. Can we have a 'then' operator?
08:14 spinclad x1-x0 is what you increment x0 by to get the next
08:14 nothingmuch btw, do we have circumfix:<| |>
08:14 nothingmuch woh
08:14 nothingmuch oh
08:14 nothingmuch buu: please explain what it does
08:15 buu nothingmuch: scalar comma operator
08:15 nothingmuch ?
08:15 buu || is to 'or' as , is to 'then'
08:15 nothingmuch sub &infix:<then> ($lhs, $rhs) { ... }
08:16 buu Excellent
08:16 nothingmuch is this a cause and effect then, or a this after that then?
08:16 buu 'this after that'
08:16 nothingmuch oh, i see why i got confused =)
08:16 nothingmuch ?eval my $x = 1, 2, 3; $x
08:16 evalbot_7198 [1, 2, 3]
08:16 spinclad i was thinking if..then, and thinking either ?? or !!
08:16 nothingmuch errm
08:16 nothingmuch ?eval my $x = scalar 1, 2, 3; $x
08:16 evalbot_7198 Error: No compatible subroutine found: "&scalar"
08:17 buu Dear god.
08:17 nothingmuch ?eval my $x = +(1, 2, 3); $x
08:17 evalbot_7198 \3
08:17 buu ?eval $x=42,$y=53;
08:17 nothingmuch why is it useful, btw?
08:17 evalbot_7198 Error: Undeclared variable: "$x"
08:17 buu nothingmuch: The same places we use scalar commas
08:17 nothingmuch i don't know where we use scalar commas
08:18 geoffb nothingmuch, '$broken++, next if /broken/;'
08:18 nothingmuch that's void comma
08:18 buu Same difference.
08:18 nothingmuch does it behave the same?
08:18 spinclad if do this then do that then do theother else do something else
08:18 nothingmuch list and scalar comma behave the same in void context, don't they?
08:19 spinclad EAMBIG
08:19 nothingmuch ?eval sub &infix:<then> ($lhs, $rhs) { $rhs }; my $x = 4 then 2; $x
08:19 evalbot_7198 Error:  unexpected "&" expecting subroutine parameters, trait or block
08:19 geoffb Um, I think it's not so much that they behave the same but rather there would be no way to tell the difference
08:19 nothingmuch ?eval sub infix:<then> ($lhs, $rhs) { $rhs }; my $x = 4 then 2; $x
08:19 evalbot_7198 \2
08:19 geoffb (so same by default)
08:20 nothingmuch brb
08:20 buu Apparently scalar commas behave differently now
08:20 buu The main difference was the return values.
08:21 geoffb (count of items v. last item, yes?)
08:23 buu Er, no?
08:23 buu I'm not sure what commas are doing in p6, apparently they create annonymous arrays
08:26 nothingmuch buu: no
08:27 nothingmuch it's that arrayish things, when assigned to a scalar become an array ref
08:27 nothingmuch i think
08:27 nothingmuch wait
08:27 nothingmuch ?eval my $x = +(5, 6, 7); $x
08:27 evalbot_7198 \3
08:27 nothingmuch okay
08:27 nothingmuch quick answer: there is no scalar comma
08:27 nothingmuch maybe it's a pugs bug
08:27 nothingmuch maybe not
08:28 nothingmuch The scalar comma , now constructs a list reference of its operands. You have to use a [-1] subscript to get the last one.
08:28 nothingmuch http://dev.perl.org/perl6/doc/design/syn/S03.html
08:28 nothingmuch first hit for "comma"
08:28 nothingmuch you were right
08:28 nothingmuch ?eval my $x = (5, 6, 7)[-1]; $x
08:28 evalbot_7198 \7
08:28 nothingmuch anyway, for 99% of the cases it doesn't really matter
08:28 nothingmuch I don't know of any use of the scalar comma as something you want to get a void out of
08:29 nothingmuch for sequencing operations you can still use it (scalar or list comma - it doesn't matter) in void context (maybe there's even the void context comma, implemented exactly like the then operator i just gave above)
08:30 nothingmuch but mose '$foo++, next if cond' can be redone as given/when
08:30 nothingmuch for @list -> $x { given $x { when cond { $counter++ } } }
08:34 svnbot6 r7199 | Darren_Duncan++ | /ext/Locale-KeyedText : resync with final Perl 5 version 1.6.2, which is now on CPAN ... the next anticipated edits to this perl 6 version is to ensure it still compiles/executes correctly under the current Pugs
08:34 evalbot_7198 has quit IRC (Remote closed the connection)
08:35 evalbot_7199 has joined #perl6
08:35 nothingmuch amazing: http://www.perlmonks.org/?node_id=494663
08:36 rantanplan_ has quit IRC ("Leaving")
08:39 scook0 has joined #perl6
08:41 ruz has joined #perl6
08:49 dduncan has left
08:50 buu Perlmonks is powered by the souls of damned orphans.
08:52 Juerd has quit IRC (Read error: 110 (Connection timed out))
08:52 geoffb OK, I've been up way too damn long
08:52 geoffb time to attempt sleep again
08:53 geoffb g'night
08:53 spinclad night geoff
08:53 spinclad me too, soon
08:55 autrijus geoffb: g'nite!
08:55 autrijus sorry I was abducted by aliens
08:55 autrijus or at least an earthling with an alien mind :)
08:56 spinclad that tends to happen at some of these conferences
08:56 autrijus geoffb: I told brendon, quite truthfully, about Dans' absence and the recent revival attempts
08:56 spinclad but only the best ones
08:56 autrijus shared runtime is still a win and they are still not at all comfortable with mono
08:57 autrijus and python is not targetting spidermonkey2 any time soon
08:57 autrijus so still a chance for parrot if pirate can somehow show concrete results.
09:13 * nothingmuch somehow doubts parrot will get stuff soon
09:13 nothingmuch it seems like it's doing maintenance jobs
09:13 nothingmuch not really developing
09:17 spinclad leo's branch has finally made trunk.  does that mean pugs could start targeting it again?
09:17 spinclad (as if nobody had enough work already)
09:19 broquaint Will parrot ever be finished?
09:20 pdcawley Not if it keeps chasing performance before correctness.
09:20 pdcawley Oops, hit return.
09:20 autrijus spinclad: yes, it means we can restart targetting.
09:20 broquaint Hrm, then it might be some time ...
09:20 autrijus before then it was impossible -- the function calling convention keeps changing in a hourly basis
09:25 nothingmuch TSa is silly
09:25 nothingmuch "i would call intentional exceptions terrorism"
09:25 pdcawley No!
09:25 pdcawley WTF!
09:25 nothingmuch me doesn't understand why he insists on calling errors exceptions
09:25 pdcawley I'm not sure 'silly' begins to cover that.
09:26 nothingmuch when the two, taken most literally, mean what they mean and not wha the wants them to mean
09:26 pdcawley Thomas "Humpty Dumpty" Sandlass
09:26 pdcawley Words mean precisely what he wants them to mean.
09:26 Juerd has joined #perl6
09:26 nothingmuch heh
09:26 nothingmuch so, have you forgiven me yet?
09:27 pdcawley Nah. You're unforgiveable.
09:27 pdcawley ; grins.
09:27 nothingmuch is that the name of a pop song or something like that?
09:27 pdcawley Oops, irc is not a mud. Muds are not irc.
09:27 pdcawley Unbelievable.
09:28 nothingmuch ttp://dictionary.reference.com/search?q=exception
09:28 spinclad unforgetable
09:28 nothingmuch One that is excepted, especially a case that does not conform to a rule or generalization.
09:28 pdcawley Someone should do Unforgiveable as well.
09:28 nothingmuch http://dictionary.reference.com/search?q=error
09:29 nothingmuch The condition of having incorrect or false knowledge.
09:29 nothingmuch when an error happens, an exception in the normal control flow is used to notify code which can make the decision of what to do next
09:30 scook0 so an erroneous situation causes an exception to normal control flow...
09:33 nothingmuch yep
09:36 wolverian has quit IRC ("leaving")
09:39 nothingmuch there
09:39 * nothingmuch hopes he's formalized it
09:39 nothingmuch scook0: i've expanded a lot on that
09:40 wolverian has joined #perl6
09:40 nothingmuch http://www.nntp.perl.org/gro​up/perl.perl6.language/23276
10:07 iblechbot has quit IRC (Read error: 110 (Connection timed out))
10:18 scook0 has quit IRC (Read error: 104 (Connection reset by peer))
10:19 wolverian has quit IRC ("leaving")
10:20 wolverian has joined #perl6
10:25 scook0 has joined #perl6
10:28 wolv has joined #perl6
10:28 wolv has quit IRC (Client Quit)
10:29 SamB has quit IRC (Connection reset by peer)
10:32 wolverian has quit IRC ("leaving")
10:34 wolverian has joined #perl6
10:50 gaal has quit IRC (Read error: 110 (Connection timed out))
11:01 SamB has joined #perl6
11:03 * spinclad happily backjournals amid the backlogging
11:08 nothingmuch bah
11:08 scook0 huh?
11:08 nothingmuch i wonder if TSa has actually used exceptions in code or not
11:09 nothingmuch he seems to be very confused not about how they work, but how they save programmer time
11:11 scook0 to be honest, I rarely even try to understand what he's talking about any more
11:11 hangring has joined #perl6
11:12 nothingmuch in the use fatal err fail thread, he
11:12 nothingmuch 's making some point
11:12 nothingmuch which i think is wrong
11:12 nothingmuch but come to think of it, i have no clue why he reached that point
11:12 wolverian I agree that culturally, exceptions are more like errors than events
11:13 wolverian I also agree that that is wrong. :)
11:13 hangring has quit IRC ("Download Gaim: http://gaim.sourceforge.net/")
11:13 wolverian (so I agree with both of you, I think)
11:14 nothingmuch wolverian: i'm not contradicting that
11:15 nothingmuch all I'm saying is that till now exceptions have only been dealing with errors
11:15 nothingmuch but exceptions are just a way of leaping up to known sanity, as a reaction to an event
11:15 wolverian right.
11:16 nothingmuch and the only events we're used to reacting to, with an exception, are errors
11:16 hangring has joined #perl6
11:17 scook0 I still don't understand what point TSa is trying to make
11:17 nothingmuch wrt 'return is two orthogonal things?'?
11:17 nothingmuch err, take one ? out of there
11:19 scook0 not so much that bit
11:19 scook0 more the 'least thing that *defines* an exception'
11:20 nothingmuch i think he's really confusing return value things
11:21 nothingmuch perhaps I should explain the Either monad
11:21 nothingmuch and fail
11:21 nothingmuch vs return
11:21 gaal has joined #perl6
11:21 nothingmuch hi gaal
11:28 iblechbot has joined #perl6
11:33 cognominal hum, what is this Blondie thing? What the difference of purpose with PIL?
11:34 cognominal is this a PIL new generation?
11:34 wolverian I think autrijus talked about it in a recent journal
11:34 scook0 I believe Blondie is its own little independent experiment
11:35 scook0 ask nothingmuch
11:35 wolverian it is not PIL2
11:35 wolverian right; it's nothingmuch's project
11:35 scook0 it's a sandbox for testing certain compilation techniques
11:36 cognominal thunking, among other thing I have gathered...
11:36 wolverian I just love how ./configure just has to check for Fortran support when I'm compiling a C IRC client
11:36 wolverian we.hates-software.com, I guess, although it's an old hate.
11:37 cognominal yea, the world of pure ideas is so confortable
11:37 cognominal and code is so disgusting
11:37 cognominal I hope Perl6 will mitigate that
11:42 gaal hi nothingmuch
11:43 gaal i want my brain back
11:43 nothingmuch what happenned to it?
11:43 gaal $work has been sucking it away, i feel depleted
11:53 spinclad geoffb++  # good article, expresses much of my own happiness with how things are moving lately
11:59 * nothingmuch actually has a comment on that, now that i'm thinking about it
11:59 nothingmuch perl isn't taking all those features and assimilating the,
11:59 nothingmuch it's doing something deeper
11:59 nothingmuch the features are evaluated for usefulness
11:59 nothingmuch and distilled
11:59 nothingmuch such that perl is a very simple language
12:00 nothingmuch with which we can implement a vast, useful core library
12:00 nothingmuch with a cool language based on all the langauges listed in the article
12:00 nothingmuch so that the result feels like the next perl, WRT perl 5
12:02 spinclad i'm beginning to see perl6 and haskell sharing much core mechanism...
12:04 spinclad these language cores are becoming community property
12:04 nothingmuch oh, and i forgot: the result makes sure that it not only assimilates perl 5, but that it assimilates the practical features of the other languages too
12:04 spinclad or especially the ideas and structures that make them up
12:04 nothingmuch aye
12:14 kolibrie has joined #perl6
12:27 kgftr|konobi theorbtwo: oioi
12:28 scook0 has quit IRC ("leaving")
12:28 xinming_Beijing has quit IRC (Read error: 113 (No route to host))
12:28 xinming_Beijing has joined #perl6
12:35 cognominal spinclad: "I'm beginning to see perl6 and haskell sharing much core mechanism..." I don't think there is a expliicit pattern matching system a la haskell in Perl6. Too bad, the more I learn haskell the more I like pattern matching
12:39 draget has joined #perl6
12:39 spinclad well, not yet anyway, at least in the surface language.  i was actually thinking of the implementations;
12:40 draget $new =~ s/\$number\n//;
12:40 draget i've done something wrong, but can't work out what
12:40 draget anyone know why that doesn't work?
12:40 spinclad draget: perl5 or perl6?
12:41 draget perl5, is their a channel for 5, I didn't see it in the list
12:42 cognominal anyway, I find there is a similarity between regex and pattern matching.  regex deals with chars, amd pattern matching with types. I am note sure what to think about that parallelism
12:42 dec has quit IRC (Read error: 110 (Connection timed out))
12:43 draget basically i need a variables contents to be searched for and replaced with nothing
12:43 spinclad draget: several, among them #perl, #perlhelp, #perlcafe, ##perl...
12:44 szabgab has joined #perl6
12:44 draget hmmm Gaim didn't list them
12:45 QtPlatypus cognominal: It is possable to make strings of things other then chars.  I wonder if its possable to make strings of types and use the regex engion to do pattern matching.
12:46 spinclad draget: #perl6 is about perl6 development, which bears only somewhat more than a historical resemblance to perl5 these days
12:46 draget oh, sorry, i'll find a different channel
12:47 cognominal QtPlatypus: indeed
12:47 dec has joined #perl6
12:48 cognominal perl6 regexing is very much the explicit building of a tree
12:48 spinclad cognominal: i expect pattern matching, type inference and such to be of great interest in p6l and implementation work
12:48 cognominal and this potentially branches (no pu intended) on attribute grammars and so many things
12:48 spinclad (and considering we have a goal of migrating the implementation to perl6, that calls for them in the language...)
12:49 cognominal the apocalypse on regexen convinces me that Perl6 would be great.
12:49 draget has left
12:49 spinclad yeah, them too
12:49 spinclad (AGs)
12:50 cognominal but with these new insighst, may be we will see it only scratches the surface
12:51 spinclad in any case it must pass @Larry, and seeing it at work will make a significant difference with that
13:01 nothingmuch luqui: ping
13:04 kakos has joined #perl6
13:10 hangring has left
13:13 fglock has joined #perl6
13:14 Qiang has joined #perl6
13:18 Limbic_Region has joined #perl6
13:20 Limbic_Region I am sure everyone has already seen and read autrijus and luke's journal entries at use.perl - but if you haven't, do so - exciting stuff
13:24 cognominal thx, Limbic_Region
13:25 nothingmuch has quit IRC (Read error: 110 (Connection timed out))
13:38 GoCooL has joined #perl6
13:51 autrijus :)
13:51 * autrijus is soaking in mad techniques from GHC
13:51 szabgab has quit IRC (Remote closed the connection)
13:53 dudley_ has joined #perl6
13:58 frodo72 has joined #perl6
14:00 PerlJam nothingmuch++ (nice analogy :)
14:00 SM_ax has quit IRC ("Download Gaim: http://gaim.sourceforge.net/")
14:21 integral hmm, is luqui's blog on one of the planets?  doesn't seem to be on planetsix
14:26 _justin has joined #perl6
14:28 wolverian has quit IRC (Remote closed the connection)
14:28 _justin_ has joined #perl6
14:28 _justin has left "Konversation terminated!"
14:29 _justin_ has quit IRC (Remote closed the connection)
14:29 _justin has joined #perl6
14:30 ods15_ has joined #perl6
14:30 svnbot6 r7200 | fglock++ | * pil-run - Junction constructor accepts either Array or ARRAY
14:30 fglock integral: http://use.perl.org/~luqui/journal
14:32 evalbot_7199 has quit IRC (Remote closed the connection)
14:32 evalbot_7200 has joined #perl6
14:36 autrijus integral: I just suggested to obra to add it
14:36 dudley_ has quit IRC (Read error: 110 (Connection timed out))
14:37 justatheory has joined #perl6
14:41 ods15_ has quit IRC (Read error: 104 (Connection reset by peer))
14:42 ods15 has quit IRC (Read error: 110 (Connection timed out))
14:47 integral thanks autrijus !
14:48 Limbic_Region integral - planetsix?
14:50 Limbic_Region if there is a planetsix, where is it and why isn't it listed in the planetarium at planetperl?
14:51 _justin has quit IRC (Remote closed the connection)
14:52 integral Limbic_Region: there's: planet.perl.org, planetsix.perl.org and planet.parrotcode.org
14:52 Limbic_Region I knew about 2 of the 3
14:52 autrijus it should be more prominiently displayed.
14:53 Limbic_Region and the middle one doesn't make sense - it should be planet.perl6.org
14:53 Limbic_Region but I couldn't find it in the planetarium
14:54 autrijus www.perl6.org is owned by pair networks, Daniel Wright
14:54 Limbic_Region so planet.perlsix.org
14:55 Limbic_Region *shrug* - It doesn't look like it aggragates from any sources I am not already reading so
14:56 autrijus er no, I mean writing daniel wright and/or pair to re-dns it
14:56 autrijus *nod*
14:56 gaal hi, need some more help with ast hacking
14:56 autrijus gaal: yo
14:56 gaal hey :)
14:56 gaal re mergeStmts
14:57 autrijus aye
14:57 gaal remember the Label thingie? i don't see how it should go in
14:58 autrijus mm perlsix.(org|net|com) is entirely unregistered
14:58 autrijus okay...?
14:58 autrijus we had two options; annotate Stmts with another field
14:58 autrijus or invent a Label node to annotate stuff
14:59 gaal we have ruleBlockBody, wher wanted to apply the Labels
14:59 gaal yeah
14:59 gaal first option:
14:59 gaal so we have Stmts !Label !Exp !Exp
15:00 gaal but then mergeStmts has to carry the current Label over to the next Exp
15:00 gaal mergeStmts (Stmts x1 x2) y = mergeStmts x1 (mergeStmts x2 y)
15:01 gaal that would lose Labe information, no?
15:01 gaal unless something like
15:02 autrijus mergeStmts (Stmts l1 x1 x2) y = mergeStmts x1 (mergeStmts l1 x2 y)
15:02 autrijus is the form
15:02 autrijus assuming y is not itself a Stmts
15:02 autrijus which will be a separate case nayways
15:02 gaal right, but ont the rhs, where does the label go?
15:02 gaal for x1 that is
15:04 autrijus I see.
15:04 autrijus that suggests mergeStmts wants to be threading a Label with it
15:04 autrijus :: Exp -> Exp -> Reader Label Exp
15:04 autrijus or simply
15:04 autrijus :: Label -> Exp -> Exp -> Exp
15:05 autrijus the label argument would be the enclosing environment
15:05 autrijus the pragmas in effect, so to speak
15:05 autrijus and then some primitives to merge them.
15:05 autrijus it's a bit hairy.
15:06 autrijus I'll go back to my hotel now; be back in 30 mins; I've been havking all day and I'll hack some more after dinner :)
15:07 gaal enjoy dinner :)
15:07 autrijus :)
15:07 gaal (funny being in the same timezone as you :)
15:15 gaal does some thing like mergeStmts (Stmts l1 x1 x2) y = do { env <- getRuleEnv ; local (const env{ Label = l1 }) (mergeStmts l1 x1 (mergeStmts l1 x2 y))
15:15 gaal make any sense?
15:17 gaal .oO(it will when this is a Reader?)
15:18 ods15 has joined #perl6
15:27 cognominal has quit IRC (Remote closed the connection)
15:28 cognominal has joined #perl6
15:29 cognominal has quit IRC (Client Quit)
15:29 cognominal has joined #perl6
15:43 vkm_ has joined #perl6
15:47 khisanth_ has joined #perl6
15:53 SamB has quit IRC (Read error: 110 (Connection timed out))
15:53 joepurl has joined #perl6
15:58 r0nny has quit IRC ("Verlassend")
15:59 khisanth_ has quit IRC ("Leaving")
16:15 joepurl has quit IRC (Read error: 113 (No route to host))
16:16 frodo72 has quit IRC ("Ciao a tutti / bye to all")
16:20 vkm_ has quit IRC ("leaving")
16:21 saorge has joined #perl6
16:26 ods15_ has joined #perl6
16:31 erg has joined #perl6
16:36 elmex has joined #perl6
16:39 ods15 has quit IRC (Read error: 110 (Connection timed out))
16:44 Amnesiac has joined #perl6
16:47 elmex has quit IRC ("leaving")
16:50 ods15_ is now known as ods15
17:11 putter has joined #perl6
17:11 putter fglock: ping/pong?
17:17 ods15_ has joined #perl6
17:23 autarch has joined #perl6
17:23 autarch seen clkao
17:23 jabbot autarch: clkao was seen 6 minutes 25 seconds ago
17:23 autarch has left
17:27 wolverian has joined #perl6
17:28 wolverian btw, is anyone planning a haskell implementation on perl6? :)
17:30 ods15 has quit IRC (Read error: 110 (Connection timed out))
17:33 wolverian another thought: do grammars have default parse rules, or so? grammar Foo { ... } Foo.parse(...);
17:34 ods15_ is now known as ods15
17:35 joepurl has joined #perl6
17:40 fglock putter: pong
17:42 elmex has joined #perl6
17:43 elmex_ has joined #perl6
17:43 elmex_ has quit IRC (Client Quit)
17:45 SamB has joined #perl6
17:45 Jooon has left
17:48 putter hi fglock.   hmm, deep recursion in chaos.  how poetic. :)  running smoke...
17:50 putter how goes?
17:55 mitrow has joined #perl6
17:57 gaal has quit IRC (Read error: 110 (Connection timed out))
17:58 mitrow has quit IRC (Client Quit)
17:58 gaal has joined #perl6
18:00 fglock_ has joined #perl6
18:01 fglock_ putter: sorry for the delay
18:01 putter np
18:01 putter :)
18:01 ruz has quit IRC (Read error: 104 (Connection reset by peer))
18:02 fglock_ I implemented Junctions yesterday - I don't  know if this is related to the smoke problem
18:02 fglock_ I only ran small tests
18:02 putter fglock++ # Junctions!! :)
18:04 putter test/smoke had lots an lots of increment on undef errors.  they went away when I if(0)'ed the new versions of ++ in PrimP5.
18:05 putter sorry :/
18:05 fglock_ I messed a bit with ++ and -- (may be my fault then)
18:05 fglock_ I was trying to fix String++
18:06 dada has quit IRC ("I have a bike you can ride it if you like")
18:07 putter did it break with mm2?
18:08 fglock_ it should not break - the problem may be elsewhere. OTOH - array slices is broken
18:09 putter map too.  and... lots.
18:09 fglock_ it is nice that the Prim subs didn't have to reimplemented for Junction to work
18:10 putter anything which explicity does a ref($x) eq 'Foo' needs  $x->isa('Foo')
18:10 putter :)
18:11 fglock_ ok - I've already fixed some of those isa()
18:12 nothingmuch has joined #perl6
18:13 putter hi nm
18:13 putter fglock_++
18:14 putter I love smokes which are already out of date by the time they finish. :)
18:15 putter sense of rapid progress
18:18 putter is misc_junctions.t supposed to run a long time (4+ minutes)?  lots of tests are currently non-terminating, so I'm wondering whether to kill it.  I don't know whether the non-terminations are due to the transition, or changes made since then.
18:19 putter ah, it finished.
18:20 putter perhaps I'm being too quick on the kill button.  perhaps things are just slower.  ??
18:20 ruz has joined #perl6
18:22 fglock_ Junctions are slow - they need to call each sub multiple times
18:22 putter ok
18:22 putter np
18:26 putter stevan: ping?
18:27 fglock_ it is possible to use threads, but it could be even slower, because the threads would be very short-lived
18:27 nothingmuch evening
18:27 pdcawley has quit IRC ()
18:27 putter evening nm
18:27 Limbic_Region I haven't been paying attention, but are the junctions slow because of a recursive implementation?
18:28 gaal has quit IRC (Read error: 110 (Connection timed out))
18:28 Limbic_Region Q::S changed to an iterative implementation some time ago fwiw
18:28 Amnesiac has quit IRC (Read error: 104 (Connection reset by peer))
18:28 Limbic_Region though Q::S is quite limited in functionality compared to junctions
18:28 putter fglock_: threads in p5, eeep!  p5 threads are still the "clone the universe" monsters, yes?
18:29 * nothingmuch wonders why his smokes suck so much
18:29 Limbic_Region putter - COW is still in the works last I heard
18:31 putter ;)
18:32 fglock_ Limbic_Region: I'm not sure if it qualifies as iterative or recursive
18:32 Limbic_Region putter - see http://yapc.mongueurs.net/yapc/talk/?tid=85 for relavent information
18:33 Limbic_Region fglock_ - well, unless it is in p5 I doubt I would be much help.  Was just noting that Q::S has an iterative implementation if necessary/interested
18:33 putter Limbic_Region: tnx for link
18:34 hlen has quit IRC (Read error: 110 (Connection timed out))
18:34 Limbic_Region it isn't the one I was looking for btw
18:34 Limbic_Region the one I was thinking of was more recent and focused specifically on the threading model
18:34 Limbic_Region *shrug*
18:35 putter fglock, LR: so what is a short characterization of junctions (whatever their current state is) vs Quantum::Superpositions?
18:36 fglock_ Limbic_Region: yes, we are talking about the p5 implementation - perl5/Perl6-Value/lib/Perl6/Code.pm
18:36 Limbic_Region dunno cause I don't know that the current state is
18:37 Limbic_Region I have spent a fair amount of time under the hood of Q::S though
18:37 Limbic_Region too bad binding isn't overloadable in p5
18:37 Limbic_Region you can't do if any( @foo ) =~ /bar/ in Q::S
18:38 fglock_ the problem is, (1|2)==3 expands to (1==3)||(2==3) and takes twice the execution time - bigger Junctions take even longer, and if you have Junctions in both sides, the times multiply
18:39 Limbic_Region fglock - I doubt seriously my understanding of Q::S is going to be much of a help to you without knowing more about the p6 junction specifications
18:39 Limbic_Region and I stopped following that when I found out that
18:40 Limbic_Region all( any(@foo) eq any(@bar)) wouldn't do what I wanted (out of the box)
18:40 larsen has quit IRC ("later")
18:41 Limbic_Region and I know even less about MM1/2 then I do about p6 junctions
18:42 Limbic_Region unless a lot more DWIMery has been added to the p6 junction specifications - I find them less then useful
18:42 r0nny has joined #perl6
18:42 Limbic_Region s/then//
18:42 Limbic_Region s/then//
18:43 fglock_ just in case - Code.pm line 223 - the implementation takes just 20 lines (called from lines 252 and 367)
18:44 Limbic_Region looking now
18:44 fglock_ it wraps the normal function call (the Code->do() function)
18:44 fglock_ s/function/method/
18:45 Limbic_Region it isn't recursive but that doesn't mean it is necessarily optimized either
18:46 Limbic_Region oh wait
18:46 Limbic_Region it is recursive
18:47 Limbic_Region I didn't see that exand_junctions calls do which calls expand_junctions
18:47 fglock_ yes - that's when you have 2 parameters that are junctions, for example
18:48 fglock_ it needs to call the subroutine (n * m) times
18:48 stevan putter:pong
18:49 Limbic_Region fglock_ - then Q::S should be able to help
18:49 Limbic_Region it uses Class::Multimethods btw
18:51 Limbic_Region hmmm - or maybe not
18:51 putter stevan: are there any mm construction methods which can't be chained?
18:51 stevan putter: not that I know of,.... but I have never tried it so hard to say
18:51 stevan can you given an example?
18:51 putter ie, ->new()->superclasses()->role​s()->add_foo()->add_bar()...
18:52 stevan k
18:52 stevan lemme see whats up
18:52 putter oh, no
18:52 putter no known problem.  just checking on intent as I design something
18:52 stevan ah
18:52 stevan well lets see
18:53 stevan new() will return a new thing of what you want
18:53 stevan if it is Class, then one of these methods will work
18:53 kgftr|konobi theorbtwo: moo?
18:53 stevan but they do not allow further chaining becuase they accessor/mutators
18:54 stevan superclasses returns the array of superclasses, etc
18:54 putter ah.  ok.  tnx.
18:55 stevan putter: BTW - endless recursion in chaos usually means that your method is recursing
18:55 stevan it would say where it really is, except all method calls must pass through there
18:56 fglock_ Limbic_Region: there is also Perl6::Junction
18:57 pdcawley has joined #perl6
18:58 nothingmuch hola pdcawley
18:59 putter stevan: re recursion.  ok, tnx.  I have to run a smoke before the recent junction changes to see if it's the mm2 transition or a new problem.
18:59 pdcawley hola.
19:00 Limbic_Region fglock_ oh yeah, that is likely to be closer to the p6 junction specification then Q::S anyday
19:00 gaal has joined #perl6
19:01 nothingmuch hola gaal
19:04 hlen has joined #perl6
19:11 fglock__ has joined #perl6
19:13 fglock__ putter: did the Junctions cause the recursion error?
19:14 r0nny has quit IRC ("sleep")
19:15 constant has joined #perl6
19:17 putter fglock__: no idea.  the smoke of current state is still running, and I haven't done any specific tests.  the test which recursed was...
19:17 chip has quit IRC (Read error: 113 (No route to host))
19:18 integral has quit IRC ("leaving")
19:18 fglock_ has quit IRC (Read error: 113 (No route to host))
19:18 Limbic_Region has quit IRC (Read error: 113 (No route to host))
19:18 constant is now known as integral
19:18 autrijus has quit IRC (Read error: 113 (No route to host))
19:20 broquaint has quit IRC (Read error: 113 (No route to host))
19:20 fglock has quit IRC (Read error: 113 (No route to host))
19:20 fglock__ has left "Fui embora"
19:21 evalbot_7200 has quit IRC (Read error: 113 (No route to host))
19:21 fglock__ has joined #perl6
19:22 putter well, the first one was t/builtins/arrays/splice.t  .  but several failed, blowing memory or time limits.
19:27 fglock__ splice isn't working in pil-run - that's MM2 related, I think (it was working before that) - I'll take a look in the code
19:31 fglock__ putter: Array.pm line 356 (slice method) - 'if ( ref($list) ne 'Array' )' - this may be a problem
19:31 dduncan has joined #perl6
19:32 putter yep.  !($list->isa('Array'))
19:33 putter would you like to grep for all instances of ref(  or shall I? :)
19:33 fglock__ please do
19:33 putter k
19:35 fglock__ !( ref($list) && $list->isa('Array') ) -- because $list may be a p5 thing
19:35 fglock__ bbiab
19:47 fglock__ back
19:59 wilx` has quit IRC (Read error: 104 (Connection reset by peer))
20:06 wilx has joined #perl6
20:06 wilx has quit IRC (Remote closed the connection)
20:08 putter ok, let's see, ref's...
20:09 wilx has joined #perl6
20:13 wilx has quit IRC (Remote closed the connection)
20:15 autrijus_tw is now known as autrijus
20:20 pdcawley has quit IRC ()
20:20 fglock__ putter: UNIVERSAL::isa() doesn't work for MM2 too
20:21 wilx has joined #perl6
20:35 rantanplan_ has joined #perl6
20:35 putter ok, that's next
20:35 nothingmuch stevan: ping
20:36 nothingmuch those germans don't seem to understand control flow at all ;-)
20:37 * fglock__ is trying to find out how Quantum::Superpositions is different from Junctions
20:38 nothingmuch i forgot how, but the answer is "subtly"
20:38 nothingmuch please note: http://search.cpan.org/~cfranks/​Perl6-Junction-1.10/Junction.pm
20:38 fglock__ Limbic_Regions said there are optimizations in Q::S that make it non-recursive
20:39 fglock__ s/Regions/Region/
20:41 svnbot6 r7201 | putter++ | Perl6::Value, Perl6::Container, PIL-Run - fix uses of ref() for mm2.  P6::Value and ApiX use separate p6v_isa and p6_isa because their requirements are different, and P6::V does not yet depend on ApiX.
20:46 fglock__ Perl6::Junction and Q::S implementations are much more complex than pil-run, because they have to overload p5 primitives. I think it is not very hard to apply Q::S optimizations to pil-run.
20:48 fglock__ I wonder what's the meaning of 'qbop' and 'qblop' method names
20:50 fglock__ 'quantize binary operation' and 'quantize binary logical operation' ?
20:51 stevan nothingmuch: poing
20:53 fglock__ $self->poing() is a reflexive op
20:53 svnbot6 r7202 | putter++ | Perl6::Value, Perl6::Container, PIL-Run - fix uses of isa() for mm2. fglock++
20:59 svnbot6 r7203 | putter++ | Perl6::Value, Perl6::Container, PIL-Run - p6v_isa() bugfix. 'p6v_isa' -> 'Perl6::Value::p6v_isa'.
21:04 fglock__ putter: slices are working! putter++
21:05 fglock__ is @a[3|4] supposed to work?
21:05 svnbot6 r7204 | putter++ | PIL-Run/lib6/Test.pm: revert use of p6_isa.  Test.pm is p6, not p5.
21:07 putter fglock_: sanity looks sane for the first time since mm2.  do you know of any unresolved mm2 problems, or shall I smoke to look for them,
21:07 fglock__ it looks fine to me so far
21:07 putter ok.  smoking...
21:08 fglock__ pugs too doesn't accept @a[3|4]
21:10 fglock__ putter: do the tests expect true and false to stringify as 'bool::false' and 'bool::true'? I think the correct is '' and '1', but I don't remember where I saw this
21:11 Qiang has quit IRC ("home")
21:11 putter ?
21:11 fglock__ just in case you saw some failure related to this
21:12 putter pugs behaves as you describe ('' '1')
21:13 fglock__ hey, that's right - if I type ~(1==1) it works
21:13 putter and so does pugs -Bperl5.  :)
21:13 fglock__ but (1==1) prints bool::true
21:14 fglock__ I guess that's (1==1).perl
21:14 fglock__ :)
21:14 * geoffb fades in
21:14 putter the pugs repl shows .perl.  the pugs -Bperl5 repl is still showing .str.   which will change, but not until after mm2 transition.  hopefully today?
21:14 geoffb spinclad, thanks for the kind words; nothingmuch, feel free to comment in the blog comments.  :-)
21:15 fglock__ geoffb: nice article
21:15 geoffb fglock__, thanks
21:18 putter fglock: re ->increment, is that speced somewhere?  there seem to be a lot of method names ->foo which seem like they should be ->FOO or ->_foo_ or something.  then again, ->id() is spec :(.
21:20 putter anyone want to email p6l suggesting ->id() should be ->object_id() (ala ruby), as "id" is too useful to users to waste on such a little written method?
21:20 fglock__ these methods can be uppercased; increment is not specced
21:20 putter k
21:20 fglock__ increment is used by $x++
21:22 putter increment only came to mind because I had seen it recently.  which triggered a general "namespace clutter?" thought.
21:25 fglock__ yes, many things could actually be multisubs in prelude now
21:25 putter hmm... $x=3; undefine($x);  isn't working.  complains undefine() is ... undefined.
21:25 fglock__ checking
21:26 nothingmuch oh my, what a long post
21:26 nothingmuch putter: yes, I agree
21:27 nothingmuch i like $foo.id for e.g. primary key in a object relational mapper, or a domain specific attribute in my domain specific use for IDs
21:28 fglock__ $a.undefine works - I think it doesn't undertand that undefine is a method
21:28 * geoffb joined p6l to be a Good List Citizen when posting, but already regretting it
21:29 wolverian has quit IRC (Remote closed the connection)
21:29 nothingmuch geoffb: it's addicting ;-)
21:29 nothingmuch the first week i was like "fun! perl 6!"
21:29 nothingmuch then i said to myself "uggh.. this is too much"
21:29 nothingmuch now it is the only list that I really read almost fully
21:29 geoffb wow, good for you I guess.
21:30 geoffb I skipped #1, and now sit square at #2.
21:30 geoffb I no longer read lists much, except for very low traffic (like 1-2/day), and even then only a couple of those
21:30 putter fglock: tnx... that suggests a bigger problem (in ApiX)...
21:30 nothingmuch i am subscribed to several lists
21:30 nothingmuch but i mark-read-thread almost all
21:31 nothingmuch the reason I read is that I like to keep watch for things that interest me
21:31 nothingmuch for example the catalyst mailing list
21:31 nothingmuch p6i
21:31 nothingmuch darcs-users and darcs-devel
21:31 geoffb I am beginning to make peace with the fact (albeit slowly) that too much interests me
21:31 nothingmuch i don't post or even read those, and I am by no means a developer, but I still like to watch the subjects
21:31 nothingmuch i think everyone here has that problem
21:32 * nothingmuch is completely unable to cope with it
21:32 nothingmuch look at harrorth
21:32 nothingmuch i have so much I want to do with it
21:32 nothingmuch but I feel that it's just ever so slightly lost the trend at which i learned from it
21:32 nothingmuch so I naturally moved on, even though I really want to work on i
21:32 nothingmuch t
21:33 geoffb nodnod
21:33 putter hmm.  (\@a).isa('Array')  is false?
21:33 geoffb at least you had an open repo
21:33 geoffb s/had/have/
21:33 nothingmuch yes
21:33 nothingmuch i'm all up for that now
21:33 nothingmuch if i'm going to have these 60% projects I might as well share the unfinished work
21:34 nothingmuch and not hide it waiting for the never reached release date ;-)
21:34 putter fglock__: pugs says  (\@a).isa('Array') is true.  pilrun says not.  ?
21:34 kolibrie has quit IRC ("leaving")
21:34 * geoffb finally just made a /dev/urandom section of his site to hold that stuff
21:34 nothingmuch pdcawley_: do you have a paypal account?
21:35 nothingmuch ah yes, I have a project that is supposed to make /dev/random into a tree automatically for my website
21:35 fglock__ putter: that's a problem in Ref.isa - Code.pm 325
21:35 nothingmuch i know how it'll work, but it's too much to implement it right now
21:35 * nothingmuch needs to quit his job, and be a student
21:35 fglock__ it should "auto-deref" Array and Hash
21:35 fglock__ just like it does with other methods
21:36 putter can you fix?
21:36 fglock__ sure
21:36 nothingmuch pdcawley_: anyway, at the rate I'm posting I guess I should support you & matt with $2 a month subscription payments... That way you can pay for the coffee it should take you to read through my mails ;-)
21:37 nothingmuch shit, rindolf has a new project
21:37 putter tnx :)
21:38 nothingmuch http://perlmeme.org/
21:40 nothingmuch "creating a simple cgi script" has the dillemas: "object oriented or not", makes you create a CGI object without knowing what it represents, talks about output without telling you what the CGI environment actually looks like, and encourages using the HTML generation features of CGI
21:40 fglock__ putter: done
21:40 nothingmuch what an aweful example
21:40 fglock__ putter: there is a bug - tied Array and Hashes should not auto-deref - I'll fix this later
21:40 nothingmuch there is absolutely no value, it's only merit is scaring the n00bs away so that they will go and do PHP
21:41 sili has quit IRC (Nick collision from services.)
21:41 sili has joined #perl6
21:42 sili has quit IRC (Nick collision from services.)
21:42 fglock__ bbiab
21:43 putter ok.  thanks fglock.
21:43 * nothingmuch goes to read and sleep
21:46 sili__ has joined #perl6
21:46 svnbot6 r7205 | fglock++ | * perl5/Ref - Ref to Array isa('Array')
21:47 fglock__ has left "Fui embora"
21:48 putter hmm.  so if an object has AUTOMETH, does that mean one can  userandomstringhere($o) to call it?
21:49 flgr_ has joined #perl6
21:49 flgr_ is now known as flgr
21:49 wolverian has joined #perl6
21:54 iblechbot has quit IRC (Connection timed out)
22:36 wolverian has quit IRC (Remote closed the connection)
22:40 wolverian has joined #perl6
23:08 rantanplan_ has quit IRC ("Leaving")
23:20 Juerd Does anyone happen to know why feather is down, or since when it has been down?
23:21 rafl The last time I was logged in was yesterday, IIRC.
23:22 dduncan has quit IRC ()
23:23 SamB has quit IRC (Connection reset by peer)
23:28 rafl Juerd: AFAIK evalbot ran on it. Maybe its part time can help
23:28 Juerd Oh, heh, indeed it can
23:28 Juerd And not just its
23:28 Juerd 21:19 -!- chip [n=chip@194.145.200.126] has quit [Read error: 113 (No route to host)]
23:28 Juerd 21:19 -!- Limbic_Region [n=81217719@p3m/member/Limbic-Region] has quit [Read error: 113 (No route to host)]
23:28 Juerd 21:19 -!- fglock_ [n=c8115922@feather.perl6.nl] has quit [Read error: 113 (No route to host)]
23:28 Juerd 21:19 -!- autrijus [n=autrijus@feather.perl6.nl] has quit [Read error: 113 (No route to host)]
23:28 Juerd 21:21 -!- broquaint [n=broquain@ts3-126.twistspace.com] has quit [Read error: 113 (No route to host)]
23:28 Juerd 21:21 -!- fglock [n=c8b690e6@feather.perl6.nl] has quit [Read error: 113 (No route to host)]
23:28 Juerd 21:22 -!- evalbot_7200 [n=evalbot_@feather.perl6.nl] has quit [Read error: 113 (No route to host)]
23:29 Juerd That's all feather, in different stages of reverse (wow, some people have been online for a long time!)
23:29 Juerd Or rather, had been.
23:29 rafl Yes. Any Idea what's wrong with feather?
23:29 Juerd No
23:29 Juerd All I know is that it's unreachable.
23:30 Juerd I got an email from chip
23:30 Juerd I can't stand though that I just got home from a place very nearby feather
23:30 Juerd I briefly considered checking my email there
23:30 Juerd And didn't
23:30 Juerd Fuck.
23:30 Juerd But - it'll have to wait until I've slept
23:30 Juerd It's 0132 am here
23:30 rafl Here as well.
23:31 Juerd And as this box isn't making money, it can wait ;)
23:32 SamB has joined #perl6
23:34 saorge has quit IRC (Connection timed out)
23:34 meppl gute nacht
23:48 joepurl has quit IRC (Read error: 110 (Connection timed out))

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

Perl 6 | Reference Documentation | Rakudo