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.
00:07 goedel joined
TimToady yay, now passing over 11,000 again. 00:07
wolverian TimToady++ 00:32
TimToady doubtless audreyt will find a way to drop that back down to a more reasonable number over the weekend. :) 00:33
00:35 meppl joined 00:36 Quell joined
TimToady I'm wondering whether we should disallow # as quote delims. 00:48
At least one of the tests fails because rx#/# eats the # trying to traverse the whitespace to find the delimiters... 00:50
wolverian what whitespace? :)
TimToady A comment is always whitespace... 00:51
wolverian there is no comment there, is there?
TimToady there is if rx trys to traverse whiteSpace to get to the delimiter... 00:52
*tries
whiteSpace = skipMany (simpleSpace <|> comment) 00:56
00:56 Ymmv joined
wolverian what if there's only one whitespace? will skipMany ignore it? :) ah well. cough. 00:58
TimToady If you'd like to rewrite the parser in P6 rules, that'd be cool. 00:59
I'm sure a lot of people would find [ <simple_space> | <comment> ]* a lot easier to understand... 01:01
but it would still eat the #... 01:02
meppl gute nacht 01:05
01:09 oylenshpeegul joined
TimToady g'night 01:09
meppl ;) 01:10
01:14 macroron joined 01:19 justatheory joined 01:24 rbysamppi joined 01:27 flounder99 joined 01:52 oylenshpeegul left 01:59 Khisanth joined 02:02 OutThere joined 02:03 OutThere is now known as SubStack 02:11 ayrnieu joined 02:13 stclare joined 02:22 macroron1 joined 02:43 integral joined 02:49 audreyt joined
azuroth anyone feel like drawing interaction diagrams? 02:58
03:07 frederico joined 03:33 aufrank joined 03:42 FurnaceBoy joined 04:00 FurnaceBoy_ joined 04:05 FurnaceBoy joined
rbysamppi Hey...in Synopsis 6, it talks about immutable and mutable types, then value and implementation types. What's the relationship between whether or not a type is mutable and if it's a value or implementation type? Could I do things like this: @something is Array; $input is IO; &pattern is Rule? 04:41
(All those variables are being declared.)
PerlJam you need a declarator in there somewhere. 04:46
rbysamppi Yeah, I forgot them. 04:50
But assume that they're all my's. 04:51
05:12 aufrank joined
aufrank hey all 05:12
PerlJam though, I wonder if in absence of stricture, would "@foo is Array" create a package variable of the appropriate shape 05:13
I guess it must 05:17
azuroth foo bar baz quux quuux quuuux quuuuux 05:18
...
FurnaceBoy hey I know that tune. 05:32
06:09 ardder joined, Quell joined 06:24 Aankhen`` joined 06:36 neoesque joined 06:38 penk left 06:58 ghenry joined 07:11 elmex joined 07:21 szbalint joined 07:28 iblechbot joined 07:52 Southen joined
spinclad Aragone: I don't see a way to keep parallel Judy1 + JudyHS arrays, since J1 takes word-length indexes and JHS takes counted strings 07:53
but it looks like adding first, next, last, prev to JudyHS would not be too hard, as long as you don't care about the order, only that it enumerates them all. 07:55
08:08 KingDiamond joined
spinclad audreyt: see ^^ 08:40
at this point i'm ready to do four things: (1) add iterators to JudyHS and contribute them back, so we can use them for Hashes 08:42
(2) build Tries with either Array[256] or JudyL[256] arrays 08:44
(3) build the < %tokenclass | %tokenclass > tokenizer with Tries 08:45
(4) do a precedence parser based on the rest of OPTable.pir 08:47
i suppose the first thing to do is add these to -- what, not RT TODOs, that's parrot, right? just a TODO file somewhere. 08:48
re (2), i'm thinking of stepping through the Trie by octets, but with JudyL array nodes one could reasonably step through by unicode chars (codepoints, i guess). 08:52
09:03 larsen joined 09:18 pdcawley_ joined 09:29 nnunley joined 09:38 kanru joined
Juerd audreyt: /etc/sudoers updated 10:16
audreyt: /win 25
argh
s[audreyt\: /win 25][] 10:17
nnunley /me hands Juerd a new set of fingers. 10:19
Heh. Irony.
10:31 turrepurre joined, KingDiamond joined
Juerd nnunley: I could use new wrists too. Do you happen to have some spares? 10:33
nnunley I wish. 10:34
10:36 xerox joined
nnunley Hrm. So it looks 'fmap' has become deprecated. 10:37
10:54 DaGo joined
gaal nnunley: huh? 11:09
nnunley Data.FunctorM is deprecated with ghc 6.5, and with it fmap. Not sure how long the deprecation cycle is in Haskell, though. 11:19
Err.. base-1.0, that is. 11:20
Definitely doesn't hit pugs in the near future, but still. 11:21
11:30 xerox joined 11:33 iblechbot joined 11:36 elmex joined
gaal that's... interesting. got a link to a discussion? 11:39
nnunley www.haskell.org/ghc/dist/current/do...ctorM.html 11:55
gaal hmm, so so many things will now have to derive Foldable... also, it seems that the Functor class and fmap aren't deprecated, themselves, just the monadic fmapM 12:11
curious to see that fold functions are now methods 12:12
err no
just limited to class
err yes, actually, methods. 12:13
12:14 pdcawley_ joined 12:19 trampel joined 13:49 DaGo joined 13:52 kanru joined, Limbic_Region joined 13:54 neoesque joined 13:55 trampel left 14:08 lichtkind joined 14:17 miyagawa joined 14:25 justatheory joined 14:28 DaGo joined 14:35 pdcawley_ joined
audreyt spinclad: (enuming JHS) would be fantastic. 14:42
spinclad: write down any thoughts you may have w.r.t that and add it to docs/notes/? 14:44
clkao audreyt: good evening 14:45
audreyt hi
clkao got mail about yaml::syck? 14:46
audreyt yup
clkao: sorry, was afk enjoying the latest cross-talk production in the series by Dr.Lai's Performance Workshop :) 14:49
(the one after the "millenium crosstalk")
audreyt resumes procmail'ing 14:51
clkao heh 14:55
audreyt clkao: hm, YAML.pm and Y::Syck has inconsistent indent handling 15:01
clkao: what would you like me to do in the Y::Syck end? inhibit emission of double-quoted strings with indentation? imho it should be fixed at ingy's end 15:05
clkao: YAML 1.1 spec section 4.5.1.1 shows that syck's treatment is canonical 15:07
15:09 ushi joined
clkao i don't know, right now i can't switch bac kto use older svk because it uses yaml 15:10
audreyt mm 15:11
15:11 larsen joined
audreyt inhbit emission of qq+newline will fix interop 15:11
but that is potentially expensive
let ingy fix it at his end is perhaps easiest 15:12
15:13 elmex joined
clkao mmm 15:15
aufrank hi folks
audreyt aufrank: hey 15:16
what was the letter you wrote that TimToady mentioned?
15:20 FurnaceBoy joined, larsen_ joined
aufrank audreyt: I sent you two letters a couple of days ago 15:21
to audreyt at that name dot org
one was asking for comments about a developer tools SoC proposal, the other about PDL on perl6 SoC proposal
audreyt ah. right. 15:22
oh, so you cc'ed tt on the 2nd one
didn't noticed that. :)
aufrank yeah, he had been encouraging about that idea
xerox Did you find more pugs puppies to code Audrey? :-)
(Howdy)
audreyt xerox: yeah, here is one with two ideas :) 15:23
xerox Good!
audreyt I wonder if we are overflowing hs.org slots a bit, but then I don't see a "quota line" in hs.org's google page... 15:24
xerox No, it's okay.
We're thinking out how to make it fair, anyway :-)
xerox unrolls the Mentoring Monad papyrus 15:25
aufrank audreyt: I'd still love to talk things out with you/ingy/juerd if any of you has time
audreyt the PDL (unified interface to vector/array/bitmap operations) one in particular probably belongs to hs.org more than perl.org if implemented on the Hs runcore 15:26
clkao audreyt: so return value is through capture as well now?
aufrank yeah audreyt, I figured the developer tools would be a TPF proposal if anything, and the PDL would be haskell.org 15:27
audreyt clkao: yeah. return is just invoking the return-cc "function"
clkao: and it makes more sense to reuse callconv
rather than conjuring another callconv out of thin air
15:28 kanru joined
audreyt aufrank: the devtools one needs more... focus 15:28
I think "IDE support" is a clear banner 15:29
so "vector data types"->hs.org and "IDE support"->pl.org? 15:30
aufrank that makes a lot of sense. I tried to bundle in some perl6 coding, but those aspects did make it seem very scattered
yeah, that works for me 15:31
audreyt you aware of util/perl6.vim and util/cperl-mode.el ?
15:31 larsen joined
aufrank yeah, but last time I tried those they weren't up to date with the spec 15:32
that said, the existence of those is one of the reasons I tried to think of other things to bundle into the proposal
audreyt I think for IDE support, perhaps the most critical one is an underlying library that editors can call to 15:33
which all revolves around the AST/Grammar
that covers synhilight, taglookup, typeinfo, errmsgparse, and even rudimentary debugging 15:34
it'd be nice if vim and emacs and eclipse and whatnot can use perl6 to parse perl6.
(especially now that parsing perl6 without perl6 is going to be even more difficult...)
aufrank yeah, definitely 15:35
audreyt and as such is easily 3mo proj
aufrank yup
that sounds excellent
audreyt woot
aufrank I'm having trouble remembering what the name of the recent p5 pretty printing project was 15:36
parsing perl code according to a document model, IIRC
there was at least one perl.com article on it
audreyt PPI. 15:37
aufrank yeah, that's the one
audreyt Alias is occasionally here
aufrank goes to look
audreyt and always on irc.perl.org #perl
there is also TimToady's MAD
now part of bleadperl
that dumps op+doc tree as one xml tree
so, the pdl one. Hs was traditionally very much _not_ a fast numerics language 15:38
probably because people who need fortran knows where to find it
aufrank right
there's at least one student interested in trying to make haskell friendlier to scientific computing this summer 15:39
audreyt oh? url?
aufrank I saw him in #haskell, but I'm checking now
audreyt but the fun of pdl is also that it makes things easier to script 15:40
there is a lot about this that can be learned from Guy Steele et al's Fortress
it's quite p6-like (rather Scala-like) in many ways, but targetted at the scientific computing crowd
15:43 mako132_ joined
aufrank audreyt: hackage.haskell.org/trac/summer-of-.../ticket/13 and hackage.haskell.org/trac/summer-of-.../ticket/24 are the proposals that are sort of related, but I don't think either of those is from the guy I saw on #haskell 15:46
audreyt: yeah, fortress is the language I'm most anticipating aside from p6 15:47
and they're not -Ofun at this point ;)
audreyt they are rather like Arc at this point ;) 15:48
xerox But Arc isn't! :-)
audreyt (with all due respect to quux and pgraham)
15:52 jsiracusa joined
aufrank bbiab, making some food & 15:56
TimToady audreyt: did you see my earlier musings about outlawing # as a quote delimiter? 16:02
audreyt TimToady: yeah. I don't see why not 16:04
I mean, we already disallow ws
and # is arguably ws.
TimToady Just looking for simplifications...
and since that's already what Pugs does... :)
well, we're giving them most of Unicode plus [[[...]]], so I can't see that many complaints about losing #. 16:05
16:05 Ymmv joined
audreyt yay :) 16:06
TimToady also I was discussing virtualizing <%.foo|%.bar> earlier...
and wondering about the performance implications.
since we need to track the token sets along with the grammar mods... 16:07
virtualizing just makes it a little harder to keep track of when to invalidate the cache, i guess... 16:08
audreyt ...and they are lexical to boot...
TimToady lexical to the comp unit, but dynamic to the cmpiler...
audreyt hm.
TimToady but you'd kind of like to think of each %.foo entry as its own method for overriding purposes...
audreyt currently I invalidate the cache after a BEGIN that may introduct new syms 16:09
regardless of what actually happened
perf isn't great that way
TimToady which is likely suboptimal
audreyt indeed
another thought is to make cache-invalid part of sym-table-entry-add
TimToady if we keep track of all the places it's used? 16:10
or make it check a valid bit that's in a handy place...
audreyt the latter is what I was thinking
TimToady I'm wondering if a more consistent symbol policy 16:11
like we did with my declarations
would let us process macros more expeditiously.
as in "you may not use a macro earlier than it defined in a lexical scope". 16:12
theorbtwo You can use macros retroactively?
TimToady no, but you can use macros from outside the current scope
and then override them. 16:13
that's what might be made erroneous in the interests of implementation.
we basically already did that with my $foo
basically, when we start a new lexical scope, it has a single new grammar, but we just don't know it all yet. 16:14
that's the idea. not sure if it really works though.
it probably won't fly. 16:15
and stacked grammars in a single lexical scope can be useful.
theorbtwo nods.
TimToady but we then do run into the same problem as with lexical vars, which is if a child calls back into our scope as its caller, which grammar do we tell them their statement was compiled with? 16:16
In particular, if that child is an eval. 16:17
and wants to default to the same language it was called from.
we need to be able to hand that child a $?PARSER that includes the macro names that were current at that point, however we define "current". 16:19
that means that <%foo|%bar> can't be off in some other scope independent of the current parser object.
that implies %foo and %bar are attributes of that parser object 16:20
and written <%.foo|%.bar>.
audreyt uh. if it's in some scope, we can hand the child a closure. 16:21
i.e. close over %foo
and hand it to that child
even if %foo is a cache of some sort
TimToady It's a closure that has to persist from compile time to run time, is all. 16:23
audreyt but we already do that for all BEGIN-time objects
that's why we attach to the AST an "object space"
with all compile-time-result objects 16:24
and that applies to eval"" as well, yeah
TimToady yeah, I'm sure it's doable, it just needs to be efficient.
16:24 justatheory joined
TimToady mostly efficient for subsequent compile in this block, not too worried about eval. 16:24
though in a sense they're the same thing. 16:25
audreyt well, they need to be the same thing
otherwise we end up with perl5ish not-closing-over-enough-things eval"" 16:26
which would be bad.
TimToady but my earlier talk of simpler policy is basically whether %.foo has to live in the current parser object or can live in COMPILING's lexical table. 16:27
audreyt and I think it should be a lexical-table-pragma-thing 16:28
since its effect is as such
TimToady hmm, you mean use COMPILING %foo? 16:29
*<%foo>
would have to be reimported on every subrule invocation
since it could be a different parent grammar. 16:30
audreyt but a shallow lexical-pad-union thing is cheap
TimToady or do you just mean a policy to look up %foo in teh current $?PARSER?
or the current COMPILING? 16:31
audreyt the current COMPILING.
TimToady (and all of its antecedents, for overlays)
audreyt (if this is slow, then OUTER is slow)
TimToady could just write the grammar <%COMPILING::statement_control|... 16:32
audreyt that's what I mean, except perhaps %COMPILING::?GRAMMAR::statement_control
or some other nonclashing names
TimToady this is basically just inside-out grammar objects... 16:33
seems a little silly, stated like that.
audreyt well, but if they are not closed over
(which they are not usually) 16:34
TimToady they?
audreyt the %foo hashes
they only get closed over when the block has an eval
or equivalent
inside it
they are not "env", that is.
TimToady seems like it would be really easy to generate an awful lot of duplicate closures in parsing if we're not careful. 16:36
audreyt (or exported, but that amounts to closing the scope over)
and as such can be efficiently codegen'ed into a "local" register (to use parrotspeak) instead of a true lexpad entry, if we can make sure that a closure is not needed
nod.
but I thought it's the API here that you are wondering about... :)
TimToady I wonder about everything, except when I don't... 16:37
but yes, I'm mostly wondering what is the most natural API.
and I do think that people can imagine a "current parser object" more easily than they can mysterious closures. 16:38
audreyt then in that case, <%COMPILING::foo> is I think not more strange than <%.foo>
especially for people actually tweaking the p6 grammar
i.e. I wonder why we are huffmanizing this :)
TimToady except for the fact that COMPILING has a policy of one-nameset per scope, and %.foo doesn't imply that 16:39
audreyt yup.
I guess I'm biased :) 16:40
(mutable cursor properties vs. immutable closure-based updating)
TimToady well, as long as it always does everything everyone might ever want, I don't care. :) 16:41
audreyt heh :) 16:42
TimToady are you working on Capturizing pugs this weekend?
(to the extent that anyone with ADD can be said to be "working on" any particular thing... :) 16:44
audreyt lol
$job want me to port Jifty to Win32
already uncovered two XS module build failures so far
XS and win32 doesn't easily mix it seems
if I get them (and maybe more of them) resolved tomorrow, with the aid of the great CPAN community
then yes, I'd love Capturizing pugs
otherwise it'd have to wait till Monday. 16:45
TimToady k
I'm just itching to start writing bits of the Pugs compiler in P6 rules...
audreyt oh, sure, but that's quite orthogonal to capturizing... 16:46
TimToady but when we make the Capture <=> Match thing bridge across, we can start rewriting rules piecemeal.
or start thinking about auto translation. 16:47
audreyt oh. right.
nodnod. I've been thinking about autotranslation for months :)
TimToady I know. But I'd be happy with piecemeal hand translation.
if only to prototype the autotranslator 16:48
audreyt nod. how about starting with Signatures? :) (Perl6::Subs, Perl6::Bindings, etc all have incomplete specs)
and it'd equip Pugs with a good Signature parser; the current one doesn't do unpacking
(any sort of)
so is of limited use when writing tree transformers
which relies heavily on unpacking
TimToady I can start thinking about that. 16:49
have my own set of distractions though...
audreyt :) 16:52
TimToady speaking of which, I need to wander off. anything else hanging fire? 16:53
audreyt not afaics 16:57
I envy that your distractions form a set 16:58
mine is a cyclic graph with duplicate nodes
(recursively defined)
TimToady at least such a thing can be lazily defined. 16:59
not sure it's sufficiently immutable though. :)
Don't stay up past your bedtime, young lady! 17:00
But also, Have the appropriate amount of fun! :)
afk & 17:01
audreyt :) 17:05
aufrank audreyt: the match object you get back from parsing a signature should have positional args in $/<positional>[0] or just $/[0]? 17:08
audreyt I imagine $<positional>[0] would be more... descriptive 17:09
aufrank so the match object after parsing a signature should have $/<positional>, $/<named>, and $/<returns> ? 17:10
audreyt maybe... or maybe it needs to have several levels of invs 17:11
aufrank and we'd want something like $/<named><foo> to also have a <type> match for sub bar ( $foo ::Dog ) {...}
invs?
audreyt sub foo ($x : $y : $z : $w) 17:12
tiebreakers
aufrank I'm not parsing the single colon there
audreyt mm? 17:16
but it's tiebreaking
conceptually we have tiebreak, pos, nam, slurpy "zones"
aufrank I meant my human brain does not recognize what that syntax would mean in a sub =\
audreyt oh oh ok
sorry. it only makes sense in a multi 17:17
multi foo (Int $x : Str $y : Num $z) {}
aufrank oh oh, ok
so those are each potential invocants?
that's what you meant by invs
aufrank clicks
audreyt in multi there may be multiple such tiebreaking levels 17:18
aufrank ok
audreyt multi foo ($x, $y : $z, $w : $a)
currently they should all be positionals
(but it's a bit silly)
aufrank and each tie breaking level can have positionals, nameds, and slurpy zones?
audreyt no, slurpy can't be 17:19
aufrank (but not implicit paramaters)
audreyt nameds aren't required to be there in 6.0.0
aufrank ok
audreyt so essentially just multi-level of tiebreakers
then the final non-tiebreaking level
consisting of pos, named, and slurpy
aufrank got it
aufrank updates his notes 17:20
audreyt but, each positional also has a name
aufrank yeah, I saw that in the test
audreyt and in fact can have a different internal name from an external name
aufrank also, the name of the argument doesn't necessarily match the internal name of the variable
audreyt sub foo (:external($internal)) {}
aufrank right, but that is a named, not a positional, right? 17:21
audreyt right.
positionals always assume the same name
aufrank the only way to treat a positional as a named is to use its internal name as its external
audreyt (which may or may not be a bug)
aufrank right
audreyt ponders
sub foo (:x($y) is positional)
but no, it's not part of spec 17:22
according to spec, yes.
aufrank so all args matches have a slot for type constraints, external name, and internal name
the last two can be different for nameds, but are always the same for pos's 17:23
audreyt aufrank: look at src/Pugs/AST/Internals.hs
line 799
the various boolean traits can be merged into a single "traits" Set 17:24
other fields need special treatment
namely, "default", "name", and "context" 17:25
aufrank seeing it :)
audreyt and now we have external name as well.
also, that does not cover unpacking
which would need another slot per-param
to hang another Signature into a param
to allow subcasing
aufrank right, a code parameter with its own :(...) sig
audreyt as wellas 17:26
sub f (@array [$head, @*tail]) {}
or 17:27
sub register ({:$name, :$addr, *%guest_data}, $room_num) {
}
in the latter case we have an anonymous (empty named) positional
which is then unpacked-as-hash
aufrank right
audreyt so the unpacking has three forms: as-hash, as-array, and as-signature 17:28
but the unpacking param all take just the Sig
so it's just a flag on the unpacker
or even better, just reuse the same "sigil" flag: hash/array/scalar
but that's all BDUF 17:29
make something work first == more fun :)
aufrank if I manage a checkin, where should it live?
audreyt maybe src/Pugs/Parser/Signature.pg 17:30
to faciliate scw's .pg.hs translator later 17:31
aufrank ok, I'm gonna keep working on it :) 17:33
audreyt :) 17:35
g'nite
17:42 rbysamppi joined 18:02 mako132_ joined 18:06 r0nny joined 18:24 Quell joined
theorbtwo thinks length('audreyt') == length('aufrank') and substr('audreyt', 0, 2) eq substr('aufrank', 0, 2) makes me confused. 18:29
18:30 sockstat joined
aufrank theorbtwo: I was the only au.* I knew for a long time 18:33
everyone in my office signed their mails "/" ~ substr($name,0,1) 18:34
and there was already an alan who had dibs on /a
theorbtwo Yeah, Castaway and Corion joined perlmonks around the same time, and both just signed "c", only in different cases. 18:35
aufrank so I was /au and usernames and email addys followed from there
aufrank hasn't been around the monastery in a while
18:37 solarbunny joined
FurnaceBoy notices the same confusion as theorbtwo and wishes this irc client would auto-colour code people 19:00
FurnaceBoy never really investigated whether it will, just thinks it would be useful
19:06 penk joined 19:11 elmex_ joined 19:18 ruoso joined 19:19 mako132_ joined 20:01 lichtkind joined
lichtkind i got to my birthday an t-shirt with an larry wall photo and the slogan "to perl or not to perl that is here the kvetchion", fortunatly didnt had the tshirt shop chartreuse 20:27
theorbtwo fortunately the t-shirt shop didn't have chartreuse" 20:28
Er, '"'.$last
lichtkind sorry i dont understand 20:29
szbalint haha
20:29 frederico joined
theorbtwo I was trying to correct your grammar. 20:30
szbalint nice bday present.
lichtkind ok thanks thats always help
20:31 FurnaceBoy is now known as FB|afk
theorbtwo And, naturally, couldn't manage to get my own quite right. 20:31
lichtkind yes i slowly i got famous in promoting perl :)
20:41 eal left 20:47 KingDiamond joined 20:56 interferon joined 20:57 interferon left
nothingmuch what would you name the property "knows where it is in a tree" or somesuch 21:05
wolverian a type? :) 21:06
nothingmuch ?
oh
no, it's a role
i mean, what would you name the role/typeclass
that implies that you can ask a node who it's parent is
wolverian node? 21:09
nothingmuch nope, that's taken, and it doesn't necessarily know who it's parent is 21:10
wolverian child? 21:11
("its", btw)
nothingmuch yeah, sorry, it's =)
i was thinking ChildOf maybe
since that's a bit more explicit
but... meh
there must be a more compsci-savvy name
wolverian w3c uses child :)
TimToady pedigree maybe? 21:12
nothingmuch wolverian: it can be child without knowing who it's parent is
oof, phone =/
wolverian nothingmuch, adopted? :) 21:13
NotOrphan
nothingmuch oof 21:18
i hate jewish mother syndrome
wolverian heh, what?
nothingmuch a good friend of mine, who does that a lot, just phoned me up
her boyfriend dumped her 21:19
so she called herself an idiot for even trying
and yadda yadda yadda
now, if you don't comment, you're insensitive
if you agree, you're calling her an idiot
if you disagree you're violating her space because she's hurt
21:19 FB|afk is now known as FurnaceBoy
nothingmuch when people do that to me i try to be mean (== honest) 21:20
usually they realize i'm not going to fall for that
and try something else
but this time it was pretty tough
i think i'll just go with HasParent
wolverian ?eval (1..10) ~~ 2.5 21:26
ruoso making pizzas 21:27
wolverian ENOEVALBOT :( 21:28
21:32 shachaf joined
spinclad i dunno, i think you could pass on commenting on 'idiot' and go with 'yeah, that sucks' and so on 21:36
but there's lots of ways to play the game 21:38
nothingmuch spinclad: the point is to delegate bad feelings 21:55
she feels like shit
she wants me to feel like shit
i'm willing to be empathic
even sympathetic to a certain degree 21:56
but i will not feel bad just because she does
22:13 elmex joined
theorbtwo
.oO(Dropping like flies.)
 22:13
Juerd Als achter vliegen vliegen vliegen, vliegen vliegen vliegen achterna 22:16
Odin-LAP Juerd: Tongue twisters? 22:41
22:44 macroron__ joined 22:53 SamB joined
wolverian hrm, why is <> the same as $*ARGS? how is $*ARGS related to STDIN? 23:05
theorbtwo wolverian: Something like how <>, <ARGV>, and @ARGV are related in p5. 23:06
wolverian theorbtwo, I have never understood why <ARGV> reads from STDIN. 23:09
merlyn in perl5? 23:10
wolverian (I mean, it is convenient to have something that does so, but why is it called 'ARGV'?)
merlyn, yes.
merlyn only if @ARGV is empty or contains "-"
if @ARGV is empty when <> is invoked, then "-" gets shoved into it
so really, it's only when "-" is seen
wolverian ah.
merlyn++ 23:11
merlyn and then $ARGV = "-"
wolverian now it makes sense. it's invisible magic.
merlyn not entirely invisible
you can see @ARGV being nibbled, and $ARGV being set
theorbtwo (Except, of course, @ARGV doesn't appear to contain '-', so it's not quite that magical.)
merlyn well - it wouldn't at the time
theorbtwo Er, it's not quite that magic. It's even more magical.
merlyn it's an indivisible step
if <> needs to open a file, and @ARGV is empty, then @ARGV = "-" 23:12
$ARGV = shift @ARGV; open ARGV;
etc
so it's in @ARGV for a moment. :)
but you don't have access at that moment
theorbtwo Makes sense. 23:13
merlyn and you *can* put "-" in @ARGV
"this", "-", "that"
theorbtwo Somewhat more obnixiously, you can put "|foo" into @ARGV, and perl will happily run your program.
merlyn that's a *feature* :)
theorbtwo Indeed. 23:14
merlyn although most coders would be surprised by that.
23:15 Limbic_Region joined
wolverian is there a short syntax to temporize an object's attribute? something like i.e. my $foo = $*ARGS.with(irs => undef).readlines; # although I know .slurp would work too 23:25
s/i.e.// # redundant 23:26
23:38 trym joined
Juerd Odin-LAP: No, Dutch people are used to such tongue exercise. 23:41
23:47 Limbic_Region joined 23:48 kanru joined 23:49 mako132_ joined 23:58 trym joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!