Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2006-05-15

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:30 avar joined perl6
00:38 weinig|away is now known as weinig
00:42 mjk joined perl6
00:43 bsb joined perl6
01:05 meppl gute nacht
01:09 FurnaceBoy n8
01:17 svnbot6 r10346 | fglock++ | PCR t/09-ratchet - 2 todo, 1 skip; added warnings for all unimplemented switches
01:28 FurnaceBoy_ joined perl6
01:30 * audreyt goes to run a series of interlocking errands just to get into U.S...
01:30 audreyt (which is a notoriously time-consuming process around here)
01:31 clkao believe me, it's much better than the french embassy
01:31 davidfetter getting into the u.s. is a giant PITA :(
01:31 audreyt clkao: I thought french is part of schengen
01:31 clkao good thing is you get 5-year visa, rather than stupid 2-month one after spending 5 hours
01:31 davidfetter this is why PostgreSQL's 10th anniversary thing in July will be in Tronno
01:31 clkao yes, schengen visa is horrible in london. but alright in taiwan
01:31 audreyt clkao: I'm not sure US still issue multientry 5year anymore for tourists
01:32 davidfetter :(
01:32 clkao audreyt: they do. i just got one last week
01:32 audreyt will find out tomorrow
01:32 audreyt oh. that's good
01:32 davidfetter where in the u.s. will you be?
01:32 clkao audreyt: sent a mail to you 'bout ref
01:32 audreyt davidfetter: redmond->chicago->boston->portland
01:32 audreyt clkao: got it. will proc tonight
01:32 clkao thank you
01:33 obra audreyt: that reminds me i should get bos->pdx tix
01:33 davidfetter hrm. portland. when?
01:33 audreyt davidfetter: oscon
01:33 audreyt obra: cool. I plan to take the same plane as you
01:34 * davidfetter is still unsure whether he can get to oscon, as the pg anniversary is 8-9 july
01:34 obra  audreyt: oh. good to know :)
01:34 davidfetter audreyt, will you be @ conisli?
01:34 audreyt that's yapc::sa? probbly yes
01:35 clkao mmmmmmm
01:35 davidfetter cool. i'm hoping i can have enough of pl/parrot done to talk about it there :)
01:35 audreyt ooh nice :)
01:36 audreyt I admit I've been not catching up with any Pg/Mysql progress for months now
01:36 obra pl/parrot?
01:36 audreyt <- abducted by the infinitely hackable sqlite
01:36 davidfetter sqlite is only finitely hackable compared to postgresql ;)
01:36 davidfetter obra, are you familiar with PostgreSQL?
01:36 audreyt well, speaking as someone who doesn't know C...
01:36 obra davidfetter: some
01:36 audreyt ...sqlite is easier to hack :)
01:37 davidfetter i don't know C either
01:37 audreyt wow
01:37 audreyt and you hack PL/Parrot? :)
01:37 davidfetter obra, well, one of the things you can do in PostgreSQL is write functions in lots of different languages
01:37 audreyt obra: stored procedure
01:37 davidfetter i see my role as organizer :)
01:37 audreyt or rather, injected embedded functions
01:38 davidfetter right now, there are separate projects for PL/Perl, PL/PHP, PL/Ruby...
01:38 davidfetter i'd like to get it so there is just one: PL/Parrot, with small (ideally 0) changes for each of the front-end languages
01:39 audreyt so essentially the mod_parrot vision
01:39 audreyt with a different setting
01:39 davidfetter yes :)
01:39 audreyt laudable :)
01:39 davidfetter and besides, i think it would help larger teams
01:40 davidfetter ppl who write python wouldn't have to fight w/ppl who write perl
01:40 davidfetter because they'd both have each other's stuff available :)
01:40 sselva joined perl6
01:40 obra *nod*
01:41 davidfetter that's the "vision" part
01:41 davidfetter so i am now trolling for ppl who will help move it forward in a more practical way
01:41 Quell joined perl6
01:42 sselva left perl6
01:42 audreyt TimToady: does these two forms differ?
01:42 audreyt f(1, ***) <== @x
01:42 audreyt f(1) <== @x
01:43 audreyt or is the latter equiv to this:
01:43 audreyt f(1; ***) <== @x
01:43 audreyt also, how does multipe pipe batches fit?
01:43 audreyt f(1, ***, 2) <== @x <== @y
01:43 audreyt does that *** only interpolate @x
01:43 audreyt and @y goes into another separate pipe batch ?
01:44 audreyt f(1, @x, 2 ; @y)
01:44 audreyt because
01:44 audreyt f(1, @x, @y, 2)
01:44 audreyt would mean that *** does a flatten of batches
01:44 audreyt the spec is a bit unclear about this.
01:51 SamB joined perl6
01:54 dduncan joined perl6
01:58 dduncan as a follow-up to my comments of a few days ago about OSCON 2006 plans ...
01:59 dduncan for various reasons, I'm now about 90% sure that I will NOT go to OSCON this year ...
01:59 dduncan but rather I will try to attend at least one YAPC
01:59 dduncan that occurs later on
02:00 davidfetter dduncan, may i PM ?
02:00 dduncan the question then becomes, with what seems like a half-dozen-plus YAPC events happening each year, which are the best to focus on if I have to be selective, where the most happens and/or most people in attendance ...
02:00 dduncan davidfetter, IRC pm is fine
02:00 dduncan if its stuff that shouldn't be said in public
02:02 davidfetter fisl 7.0 had > 5500 attendees. dunno how many of those were at YAPC::Brazil, which was part of that
02:04 audreyt dduncan: YAPCNA :)
02:04 dduncan that's possible, but YAPC NA is even earlier than OSCON, so I may not have enough prep time
02:05 dduncan I suppose the biggest issue is getting cheap flights at the last minute
02:05 dduncan you're a lot more experienced with regular world travel, so probably know things ...
02:06 dduncan will there be a YAPC NA hackathon?
02:06 audreyt yes.
02:06 audreyt will be great if yo ucan be there
02:07 audreyt lots of #perl6 people and #parrot people will be there
02:07 audreyt http://yapcchicago.org/wiki/?HackAThon
02:07 dduncan in your experience, what's the latest one can book a round trip flight reliably?
02:07 audreyt reliably, maybe two weeks before
02:08 audreyt if you are wiling to pay for business class... pretty much anytime.
02:08 dduncan I just need a ride and don't care for amenities, so whatever's easy to get and inexpensive
02:09 dduncan also, it may be on the website, but I also have to figure out where to stay ...
02:10 dduncan looking at website ...
02:14 dduncan okay, I'll try to get to YAPC NA 2006 plus hackathon ... should know more definitely in the next 2 weeks or so
02:20 shachaf joined perl6
02:45 diotalevi joined perl6
02:46 weinig is now known as weinig|sleep
03:15 svnbot6 r10347 | fglock++ | PCR t/09-ratchet.t - char classes like <alpha> are inlined
03:36 hikozaemon joined perl6
03:37 kanru joined perl6
03:49 coumbes joined perl6
03:58 mako132_ joined perl6
04:00 nirgle joined perl6
04:40 shachaf joined perl6
04:50 shachaf joined perl6
05:04 Khisanth joined perl6
05:25 Quell joined perl6
05:56 bsb left perl6
06:01 svnbot6 r10348 | Darren_Duncan++ | ext/Relation/ : updated Relation.pm to refactor code between semijoin() and join(), then add some algorithmic optimization to join(), then add new method semidifference() with aliases semiminus() and not_matching()
06:03 dduncan left perl6
06:09 justatheory joined perl6
06:12 justatheory joined perl6
06:33 justatheory joined perl6
06:37 jserv-- joined perl6
06:40 domm joined perl6
06:40 iblechbot joined perl6
06:50 marmic joined perl6
06:54 domm grr, freenode won't let me /msg Juerd cause I'm not a registered user
06:54 domm Juerd: I sent you a mail wrt train travel to vienna, please reply soon...
06:55 rgs domm: or you can /msg him on irc.perl.org too
06:57 Juerd Hi domm
06:57 domm rgs: oh well, yes, thanks
06:58 rgs hi, Juerd & domm !
07:02 Southen joined perl6
07:03 oozy joined perl6
07:30 drrho joined perl6
07:30 Ymmv joined perl6
07:48 kanru joined perl6
08:38 Aankhen`` joined perl6
09:00 nnunley joined perl6
09:03 cognominal joined perl6
09:08 Aragone is now known as Arathorn
09:11 Arathorn morning everybody
09:36 justatheory joined perl6
09:51 justatheory joined perl6
10:01 justatheory joined perl6
10:12 justatheory joined perl6
10:17 Aankhen`` joined perl6
10:18 justatheory joined perl6
10:28 ludan joined perl6
10:37 justatheory joined perl6
10:47 ruz joined perl6
10:53 justatheory joined perl6
11:00 justatheory joined perl6
11:04 ruz joined perl6
11:09 chris2 joined perl6
11:09 justatheory joined perl6
11:22 justatheory joined perl6
11:24 jsiracusa joined perl6
11:29 neoesque joined perl6
11:30 kane_ joined perl6
11:35 mago joined perl6
11:38 justatheory joined perl6
11:52 justatheory joined perl6
11:55 mako132_ joined perl6
12:10 justatheory joined perl6
12:11 iblechbot joined perl6
12:17 Limbic_Region joined perl6
12:18 kolibrie joined perl6
12:21 ruoso joined perl6
12:21 justatheory joined perl6
12:23 Juerd What's the difference between $foo and $.foo again?
12:23 Juerd Perl6/Overview/Object.pod doesn't explain this; should I be looking elsewhere?
12:24 justatheory joined perl6
12:27 azuroth maybe $foo is just a 'my' and $.foo is a has?
12:27 Juerd S12 of course. I've been doing other stuff too long.
12:28 Juerd Public attributes have a secondary sigil of "dot", indicating the automatic generation of an accessor method of the same name
12:28 azuroth like ruby's @foo I guess?
12:29 Juerd I don't know.
12:29 Juerd I've not used Ruby in ages
12:30 azuroth oh, @foo might not declare an accessor. just makes it an instance var
12:30 justatheory joined perl6
12:50 jiing_ joined perl6
12:56 justatheory joined perl6
13:11 justatheory joined perl6
13:14 justatheory joined perl6
13:19 justatheory joined perl6
13:21 cdpruden joined perl6
13:26 justatheory joined perl6
13:27 weinig|sleep is now known as weinig
13:30 justatheory joined perl6
13:36 justatheory joined perl6
13:37 svnbot6 r10349 | putter++ | sprintf_and_as.t - corrected typo.  sprintf("%d",2.4) is "2", not "3".
13:38 justatheory joined perl6
13:43 justatheory joined perl6
13:45 sri_ joined perl6
13:50 justatheory joined perl6
13:52 justatheory joined perl6
13:54 gaal putter!
13:56 kane_ what Test::* module would i use to test commandline tools for desired output and exit codes?
13:57 obra kane_: Test::Expect
13:57 obra er oops. that's a p5 module
13:57 justatheory joined perl6
13:57 kane_ p5? that's what i want...
13:59 obra Cool
13:59 justatheory joined perl6
13:59 kane_ and it's by acme... must be good code ;)
13:59 kane_ thanks obra
14:00 obra :)
14:00 obra and acme is responsive to patches
14:00 FurnaceBoy joined perl6
14:00 obra rt 3.6.0rc2 has a set of cli tests based on it
14:01 justatheory joined perl6
14:04 justatheory joined perl6
14:08 justatheory joined perl6
14:10 justatheory joined perl6
14:13 svnbot6 r10350 | fglock++ | PCR - subrule-call rewrite under way
14:16 Qiang joined perl6
14:46 cratylus joined perl6
14:46 svnbot6 r10351 | putter++ | Common/Regexp-ReplaceEngine.pm wrap-up.  Added some notes, and moved Aside/.
15:01 elmex joined perl6
15:03 cratylus joined perl6
15:19 mako132 joined perl6
15:20 cratylus_ joined perl6
15:29 cratylus joined perl6
15:34 svnbot6 r10352 | putter++ | re-override/NOTES - an AUTHORS section attributed this new file to me. I removed it as undesired and misleading.
15:34 svnbot6 r10352 | putter++ | Yes, re-override was a fork of some exploratory coding I did.  But neither the code, nor its embedded __END__ notes, were more than development transients.  Given pugs' openness, anyone is of course free to distribute them as they judge fit.  But the quality gap between these development scribbles, and what I consider releasable code and docs, is too large for me to wish my name on such distribution.  Which is why I keep removing it.
15:34 svnbot6 r10352 | putter++ | Thanks for the intent. ;)
15:35 FurnaceBoy nicely structured essay: introduction, body, conclusion.
15:36 FurnaceBoy makes me wonder if commit messages should include HTML, javaScript and inline images.
15:36 FurnaceBoy and GPG signatures
15:38 sahadev joined perl6
15:43 mako132 joined perl6
15:48 nnunley_ joined perl6
15:50 neoesque joined perl6
15:50 nnunley__ joined perl6
15:56 nnunley_ is now known as nnunley
15:58 pdcawley_ joined perl6
16:01 scw_ joined perl6
16:28 Maddingue joined perl6
16:30 penk joined perl6
16:30 penk joined perl6
16:31 kane_ joined perl6
16:35 sri_ joined perl6
16:38 svnbot6 r10353 | putter++ | builtins/my.t - Added comment: test $d ($d interpreted as $OUTER::d prior to mid-block my $d declaration) is inconsistent with the current spirit of S04.
16:38 svnbot6 r10353 | putter++ | It is not however strictly inconsistent.  While defined as an error, compilers are allowed, but not required, to complain.  But a fully conformant compiler will fail the test.
16:38 svnbot6 r10353 | putter++ | It might be worth noting in the spec why this ambiguity is desirable.
16:41 hexmode joined perl6
16:48 fglock joined perl6
16:50 sri__ joined perl6
16:54 fglock how about making 'token' match at pos by default - regex :ratchet :p { ... }
17:00 fglock does "atbtok" ~~ token{ tok } fails? since the first 't' is not followed by an 'o' and backtracking is disabled
17:03 TimToady fglock: already does assume :p if called as a subrule.
17:03 TimToady but a case might be made for doing it always.
17:04 fglock I think :p would make it easier to learn/compile, as it it is mostly supposed to be used inside subrules anyway
17:04 TimToady but if you want to use a token directly, why wouldn't you just use //?
17:05 fglock isn't // a regex?
17:06 TimToady yes, so you'd have to anchor it explicitly, but that's like to be shorter than token {...} inlined still.
17:06 TimToady *likely
17:06 TimToady I'm not really arguing against you--it's something that just hasn't been decided.
17:07 fglock I'm asking because I have to decide whether to fix or to skip a test :)
17:07 TimToady when used as a top-level matcher, there's something to be said for anchoring it on *both* ends.
17:07 FurnaceBoy is now known as CurmudgeonBoy
17:08 fglock sure
17:08 aufrank joined perl6
17:08 TimToady that would in some way increase the "meaning" of token/rule vs regex.
17:10 TimToady and would probably increase the utility of $input ~~ $pattern.
17:10 TimToady for validation of numbers and such.
17:11 gaal TimToady: is it okay to rename the 'reduce' builtin to 'fold' and add one for 'scan'? My understanding is that 'reduce' is the general term for them both.
17:12 TimToady I don't like that usage of "scan" at all.
17:12 TimToady It's far too overloaded to go with that one specific meaning.
17:12 gaal the names ending with 'morphism' are worse
17:12 TimToady which has very little to do with normal usage of the word.
17:13 TimToady besides, there's no "reduction" taking place with it.
17:13 TimToady you get the same number of values out.
17:14 TimToady so I think this is just one of those places where Haskell is cork-brained from a linguistic point of view.
17:14 gaal Okay. I'll try to think of a better name, but I really do think the function should be builtin, since it's easy on the runcore and relatively hard on to express in userland
17:14 wolverian .reduce :keep? kind of fights itself semantically
17:15 gaal .reduce :but_keep_adding_liquid
17:15 TimToady you want something more like the idea of taking more and more terms in mathematics...
17:15 TimToady refine maybe
17:15 gaal .omgitsalive
17:15 wolverian .triangulate? :)
17:16 TimToady [\X]
17:16 wolverian that's where I got it from..
17:16 TimToady triduce.
17:16 gaal [\X] is fine but doesn't accept lambdas, as we said..
17:16 wolverian ha! tridunce
17:16 gaal well, as you said :)
17:16 TimToady traduce
17:16 wolverian [\x -> ...] @list
17:17 gaal wolverian: syntax error: operators don't permit whitespace
17:17 gaal hmm!
17:17 wolverian it's probably a bad idea to emulate the lambda syntax in one specific place...
17:17 gaal so maybe this could be made to work with pointy subs after all? :)
17:17 TimToady if some initial sequence was unlikely to be used in [...] composers, we could distinguish based on longest token even without the ]
17:18 TimToady [\X] was not intended to be a lambda, H notwithstanding.
17:18 fglock deltas
17:19 TimToady we could possibly go with [->...] and [\->...] though.
17:19 gdickie Perhaps [\ ... ] is an operator?
17:19 TimToady at the moment, it's a metaoperaotr.
17:19 gaal gdickie: it's already a metaoperator
17:20 svnbot6 r10354 | fglock++ | PCR - t/09-ratchet.t - tokens are only tested for :p cases
17:20 TimToady maybe induce
17:21 particle_ is [/.] an operator yet? i think it'd fill your screen and logs with useless garbage :)
17:22 gdickie What's the real difference to the language-user between a metaoperator, and a circumfix operator on Code? The latter might allow [\ -> { ... } ] or [\ myfun ]
17:23 TimToady that latter is too much like a [...] composer
17:24 TimToady but the "real" difference?  hmm.
17:24 TimToady a metaoperator is making stricter promises about what is known at compile time, i think.
17:25 TimToady it has to be based on something that's already completely declared.
17:25 TimToady and more to the point, its interface completely known.
17:26 TimToady if we had something like [-> { $^a mumble
17:26 CurmudgeonBoy is now known as FurnaceBoy
17:26 TimToady [-> $a, $b {...}]
17:26 TimToady then we know the interface at that point, which might be good enough.
17:27 TimToady In the more general case we might just have to assume that the function mentioned takes two arguments, in the absence of type inferencing.
17:28 TimToady but even [-> $a, $b {...}] is ambiguous with a list consisting of one pointy sub.
17:28 Juerd TimToady: Do you have an overview that indicates when to use a role, when to use a class?
17:28 TimToady so we'd have to special case it.
17:28 Juerd TimToady: Something end users can understand :)
17:28 TimToady How the heck would I have one of those when I don't understand it myself?  :)
17:29 gaal how about .radish for a name. It gets asymptotocally closer to a point
17:29 TimToady I'm just a cargo-cult designer.  I leave it to others to understand it all...
17:29 Juerd TimToady: Err...
17:29 TimToady .radish and .reduce are too similar in pronunciation when drunk.
17:29 Juerd I kind of hoped you'd understand the matter...
17:30 Juerd What other languages have roles?
17:30 TimToady none.  Smalltalk has traits though.
17:30 Juerd Then it's not cargo culting :)
17:31 TimToady As with any cargo-culter, I add my own interpretation when I borrow the symbology.  :)
17:31 svnbot6 r10355 | fglock++ | PCR - updated benchmark
17:32 TimToady maybe .reduce and .preduce
17:33 gaal .produce even?
17:33 gaal because it produces the result (by stages)
17:33 TimToady has possibilities.
17:33 gaal or was that what the pun was working on?
17:34 TimToady produce the next rabbit out of the hat...
17:34 TimToady I was thinking the pun, but also the "pre"ness of it, as in "what reduce has internally before it gives you the final result"
17:35 gaal .proginate
17:35 TimToady but "pro" can mean "in front", I think, so maybe it works anyway.
17:35 Juerd TimToady: But you have no view of what should be a role and what should be a class?
17:35 gaal s/i/e/
17:36 Juerd That'll be an incredible mess then, the first few years.
17:36 TimToady I like the *duce relationship
17:36 Juerd And I won't have anything useful to say in my Perl 6 OO talk :)
17:36 DaGo joined perl6
17:37 TimToady roles are for abstractions that don't make a lot of sense to instantiate, such as generics
17:38 TimToady that is, to instantiate without more information.
17:38 gaal well, "re-" is back and "pro-" is forward in the ducere etymology
17:39 TimToady I think it can be viewed as working in either Latin or English.
17:40 gaal I meant, it's consistent that way too. I'm for (heh) it.
17:41 bernhard joined perl6
17:43 TimToady Juerd: it's already been an incredible mess for the first few years, which is why you should probably be asking Stevan about this and not me.
17:43 gdickie Somebody mentioned mathematical series a while ago; a mathematician would call the result of [\+] the "partial sums" of the operand. So if [ ... ] is reduce, [\ ... ] is partial-reduce or partial-reductions.
17:44 gaal fwiw, I thought context dependence was a beautiful solution for [\]. If only say [+] weren't in list context.
17:44 gaal .parduce :)
17:44 TimToady I wasn't gonna say it...
17:45 aufrank we're golfing now?  a par two hole?
17:45 gaal aufrank: K does fold AND scan metaoperators in one character each
17:46 TimToady It's no accident that (real) golfers are also called "hackers".
17:46 aufrank gaal: par-deuce
17:46 TimToady *cough* APL *cough*
17:47 PerlJam good $localtime #perl6
17:47 gaal APL, K and J are all related, aren't they?
17:47 TimToady I know J is derived from APL.  Don't know about K.
17:47 Juerd stevan: Do you have an overview that explains to end users when to use classes, and when to use roles?
17:48 TimToady Juerd: or we could always Ask Slashdot.  :)
17:48 Juerd God, please, no.
17:49 PerlJam are there any languages that support the concept of roles out in the wild?
17:49 aufrank stevan borrowed heavily from scala's traits, as I understand it
17:50 gaal K was developed, quoth Wikipedia, by "Arthur Whitney, an influential APL programmer".
17:50 merlyn influential in general, or just about APL?
17:50 merlyn that's an odd use of "influential"
17:51 merlyn maybe he was influential *because* he used APL?
17:51 PerlJam Juerd: perhaps there's some "best practices" wrt traits (roles) in the scala docs
17:51 gaal maybe he was influential because he developed K :)
17:52 PerlJam gaal: so clearly Larry is an influential C programmer :)
17:52 TimToady that's BASIC to you.
17:54 gaal Surely he's influenced the way people think about C programming, and he did it with C.
17:54 TimToady so maybe use P; puts you into the APL subset of P6...
17:54 TimToady though I suppose L should be the next one...
17:54 TimToady use PL;
17:54 TimToady swipe 'em both.
17:55 gaal L for ladder, that which is discarded etc.
17:56 gaal where's audreyt, she gets my oblique Wittgenstein jokes
17:56 TimToady well, I gotta pretend to go and work.  Not being much use right now, I'm afraid, due to incipient grandfatherhood...
17:57 gaal I gotta .produce
17:57 gaal happy incipience
17:58 aufrank wow, that is the first time I actually parsed pro-duce as the verb "produce"
17:58 aufrank in this conversation, that is
17:58 * PerlJam is the anti-duce
17:58 TimToady well, as long as you've invoked her...  audreyt: I'll think about *** and pipes in general as I'm driving up to Seattle.  There's something declarative I want there, but I'm not sure we've nailed it yet.  Trying to figure out how it all desugars.
17:58 particle_ driving to seattle?
17:59 FurnaceBoy incipient? impending?
17:59 TimToady incipient grandfatherhood, as I said...
17:59 TimToady or perhapsh excipient...
17:59 FurnaceBoy "beginning to happen or develop" ... it's been incipient for a while, then. :)
17:59 TimToady perhapsh I should not drink so early in the day.
17:59 gaal bon nascence, TimToady
18:00 TimToady official due date was yesterday, but nuttin' happened.
18:01 kolibrie babies are "never" born on the due date
18:01 * particle_ was two weeks late, and has been trying to catch up since
18:01 * kolibrie was late too
18:02 TimToady I'm pretty sure I was conceived on Christmas...
18:03 particle_ i guess that's why perl6 will be born on christmas
18:03 TimToady I come by it honestly...
18:03 TimToady commutating... &
18:04 solarbunny left perl6
18:05 ruz joined perl6
18:06 stevan Juerd: ping
18:14 penk joined perl6
18:21 marmic joined perl6
18:26 Maddingue joined perl6
18:29 spinclad [\...]: how about .adduce: lead another item to the heap, and leave it there
18:30 spinclad ..., and pile it on
18:30 wolverian I like .produce
18:32 wolverian btw, why was [\] invented? [] in list context being a scan too weird?
18:33 TimToady yeah, basically, [X] is deemed to be too useful in list context, and the whole point of calling it "reduce" is that you end up with one thing.
18:34 TimToady (or at least something less)
18:35 TimToady And while I've often wanted a reduce operator, I can't recall ever wanting a produce operator.  Probably just lack of imagination on my part, but still, it's a Data Point.
18:35 particle_ .greengrocer
18:36 PerlJam does [\X] need the huffmanization?
18:36 gaal well, it was a nice unification
18:37 gaal @triangulars ::= [\...] ...
18:37 gaal @factorials := [*] 1 ..*
18:37 gaal @pidigits := [+] ...
18:37 gaal [\] that is
18:38 TimToady but it was falling into the Haskell trap of huffmanizing the abstruse
18:38 TimToady not to mention the APL trap
18:39 ludan joined perl6
18:39 TimToady PerlJam: we could certainly live without [\X] but it's awfully "cute", to quote at least two of @Larry...
18:39 aufrank has [\...] discussion been pretty much contained in this channel?  no bleed out to p6l yet?
18:39 * aufrank tries to figure out what he missed
18:39 ludan hi
18:39 gaal I can't argue against that. I just loved the elegance
18:39 gaal aufrank: it was in a thread called "Scans"
18:40 PerlJam aufrank: well, a synopsis has been quite publicly updated on p6l  :)
18:40 TimToady arguably, the [\X] should have been marked as conjectural, but it wasn't, so it isn't.  :)
18:41 TimToady Just like *** and ==> *
18:41 aufrank I remember [+] as reduce
18:42 aufrank maybe this is the problem:  what is \X a standin for?
18:42 PerlJam [\+] for instance,  (X is just some operator)
18:42 PerlJam er, binary operator even :)
18:43 PerlJam [\X] does have the nice allusion to Captures that bridges the mental gap to "save intermediate results"
18:43 aufrank ?eval ([+] [1, 2, 3])
18:44 TimToady 3, I hope.
18:44 TimToady actually, (1,2,3)
18:45 TimToady hmm.  maybe need something about forcing numeric context on a single argument there...
18:46 TimToady well, ([+] [1, 2, 3]) + 1 is 4, so maybe not an issue.
18:47 loud- joined perl6
18:48 TimToady on another subject entirely, I was thinking about unifying {...} and ; parsing a bit by saying that ; also may be continued with a , or followed by an immediate postfix.  Then you could say
18:48 TimToady sort 3,1,2;.say
18:49 TimToady or sort 3,1,2;\   .say
18:49 TimToady not that you'd generally want to write it that way, but it seems like {...} and ; parsing could be more alike than different.
18:50 aufrank gaal: I figured it out.  I missed triangulate in r9216.  I got too distracted by tr/\*/\[,\]/
18:50 TimToady though it seems by that analogy that say 1,2,3;,4 is the same as 1,2,3,4
18:52 TimToady and arguably the sort-say should be written sort 3,1,2 ==> say
18:52 TimToady or sort 3,1,2 ==> ***.say in the current (non)conjecture.
18:52 wolverian an arrow to a verb is weird.
18:53 PerlJam having so many trigraphs in perl seems weird to me, but maybe only because I'm not used to it.
18:53 TimToady how so?
18:54 wolverian the arrow is a verb.
18:54 TimToady how so re "arrow to verb", not trigraphs
18:55 wolverian I read it as "go to" or "put into"
18:55 TimToady "put into the input queue of"
18:55 PerlJam wolverian: it's only weird because the noun is hidden.
18:56 ruz joined perl6
18:56 wolverian yeah, it's not so weird that I won't get used to it.
18:56 TimToady I think it's clearer than foo | bar in the shell...
18:57 wolverian it is that..
18:57 PerlJam TimToady: but we have had decades to get used to shell pipes and only a short time to get used to perl pipes :)
18:58 wolverian the shell | would make more sense if it still had a hole in it.
18:58 TimToady that's at right angles to how I think of the pipe
18:58 TimToady I think it's a very skinny pipe.
18:59 wolverian well, a leaking pipe.. it randomly picks parameters and sends them to the neighbouring statements
18:59 PerlJam So is there a unicode version of ==> that's thicker so that I can increase my throughput?  ;)
19:00 wolverian there is, but I can't type it with this keymap, only → . and I don't know if it's official.
19:00 TimToady ⇛       21DB            RIGHTWARDS TRIPLE ARROW
19:00 wolverian don't you want the double arrow?
19:00 TimToady ⇉       21C9            RIGHTWARDS PAIRED ARROWS
19:01 TimToady the funny thing about
19:01 TimToady ⇒       21D2            RIGHTWARDS DOUBLE ARROW
19:01 gaal Beh. I added "produce" and [\] but am getting SVK errors in pushing it. Don't work on hacking this, folks
19:01 TimToady is that
19:01 TimToady ⇐       21D0            LEFTWARDS DOUBLE ARROW
19:01 TimToady is shorter in my font than the rightwards one.
19:01 TimToady you'd think they'd be symmetrical.
19:02 TimToady course, there's also dingbats:
19:02 wolverian they're the same length here.
19:02 TimToady ➔       2794            HEAVY WIDE-HEADED RIGHTWARDS ARROW
19:02 TimToady ➙       2799            HEAVY RIGHTWARDS ARROW
19:02 TimToady ➛       279B            DRAFTING POINT RIGHTWARDS ARROW
19:02 TimToady ➝       279D            TRIANGLE-HEADED RIGHTWARDS ARROW
19:02 TimToady ➞       279E            HEAVY TRIANGLE-HEADED RIGHTWARDS ARROW
19:02 TimToady ➡       27A1            BLACK RIGHTWARDS ARROW
19:03 wolverian oh, and the double arrow is probably =>, so that the unicode variants are shifted by one, because there's -> and => (both the same length)
19:03 TimToady well, it's not the only thing screwy about my current font...
19:04 TimToady I was actually thinking pipes should maybe be:
19:04 TimToady ⇠       21E0            LEFTWARDS DASHED ARROW
19:04 TimToady ⇢       21E2            RIGHTWARDS DASHED ARROW
19:05 aufrank invisibly different from → here
19:05 TimToady or maybe
19:05 TimToady ⇦       21E6            LEFTWARDS WHITE ARROW
19:05 TimToady ⇨       21E8            RIGHTWARDS WHITE ARROW
19:06 TimToady there do exist:
19:06 TimToady ⟸       27F8            LONG LEFTWARDS DOUBLE ARROW
19:06 TimToady ⟹       27F9            LONG RIGHTWARDS DOUBLE ARROW
19:06 TimToady but those are even shorter in my font...
19:07 aufrank  do those show up with boxes around them for everyone else?
19:07 aufrank or is that my font or my dark background?
19:07 TimToady not me.
19:07 TimToady they just look like a shorter version of the "short" ones in my font.
19:07 wolverian my font doesn't have them, so they show up as ? in boxes
19:08 TimToady I think they don't show up in Firefox either, last I checked.
19:08 aufrank wolverian wins, I think
19:09 TimToady is this the circular scale where it wraps around in back so that wins and loses is the same thing? :)
19:09 TimToady sort of like communist and fascist meet around back at totalitarian?
19:10 TimToady and blue and red meet around back at purple?
19:10 * FurnaceBoy (h) ASCII
19:10 TimToady I am not familar with the (h) operator...
19:11 TimToady Is that ASCII for ❤ ?
19:12 clkao (y)
19:12 clkao (Y) actually
19:12 FurnaceBoy <3
19:12 PerlJam (h) is the hate operator.  FurnaceBoy was expressing his hatred for ASCII
19:12 TimToady not to be confused with [Y]
19:12 FurnaceBoy (heart)
19:12 TimToady or [\Y]
19:12 FurnaceBoy PerlJam, !hate
19:12 clkao (\/)
19:12 clkao (v)
19:12 FurnaceBoy I see <3 for it, but I always think icecream, or flowers...
19:13 TimToady [nor]
19:13 FurnaceBoy or file descriptor 3...
19:13 wolverian (v) is just hot ..
19:13 * FurnaceBoy wolfwhistles
19:13 * FurnaceBoy ... in EBCDIC
19:13 gdickie [Y] transposes a list of lists. Nice!
19:13 FurnaceBoy (v) conflates a crate of plates.
19:17 PerlJam so ... earlier I said that the operator in [X] must be binary.  Is that true?  Could it be N-ary as long as it prodoces a scalar result?
19:18 TimToady for syntactic purposes it must currently be defined as infix:<X>
19:18 TimToady otherwise it's too ambiguous with [...] as an array composer.
19:19 PerlJam okay, good. :
19:19 TimToady It's probably sufficient to say &infix:<X> ::= &X or some such.
19:20 TimToady presuming the signature of X is $x/$y compatible.
19:27 svnbot6 r10356 | gaal++ |  r10423@sike:  roo | 2006-05-15 21:52:25 +0300
19:27 svnbot6 r10356 | gaal++ |  * parse and honor [\] metaoperator
19:27 svnbot6 r10356 | gaal++ |  * produce builtin
19:27 svnbot6 r10356 | gaal++ |  * update tests
19:27 gaal clkao++
19:28 TimToady http://wall.org/~larry/umaybe has a list of potential Unicode characters for Perl.  I'm particular fond of
19:28 gaal afk&
19:28 TimToady ⁇       2047            DOUBLE QUESTION MARK
19:28 TimToady ‼       203C            DOUBLE EXCLAMATION MARK
19:29 TimToady mostly just thinking ahead to specifying which characters *might* have mappings someday, and if so, what they'd be.
19:33 TimToady all systems are go for lunch &
19:41 cratylus left perl6
19:52 clkao gaal: :)
19:56 aufrank joined perl6
20:03 aufrank joined perl6
20:04 gaal TimToady: docs/quickref/unicode has a few suggestions
20:06 gaal √ ≡ etc.
20:27 TimToady I'm taking my list down since docs/quickref/unicode is much better organized.
20:28 nothingmuch joined perl6
20:29 justatheory joined perl6
20:44 loud- left perl6
20:50 larsen joined perl6
20:52 shachaf joined perl6
21:12 aufrank joined perl6
21:23 penk joined perl6
21:25 ruoso joined perl6
21:26 jsiracusa joined perl6
21:37 azuroth I can mount my mp3 player and delete one directory on it, but after that it says "this is a read only filesystem!"
21:42 szbalint I suppose linux + usb + storage?
21:43 szbalint If yes you might enable usb debug messages in the kernel and check dmesg
21:44 azuroth hmm, okay. I might just try reformatting the thingy as fat first, because it's got a whole bunch of crazy files on it
21:47 FurnaceBoy did you check system log?
21:47 FurnaceBoy a media error can force it to switch to read only
21:47 azuroth no, I didn't ;-p where is the system log?
21:48 FurnaceBoy /var/log/messages, typically
21:48 FurnaceBoy or you can also just hit 'dmesg' for recent ones
21:49 azuroth ah, cool. I'll check later; hopefully this copying is working now after the format. I think I'm going to miss my train
21:50 FurnaceBoy good luck
21:55 azuroth at least two songs works :D
22:04 froh-doh joined perl6
22:18 weinig joined perl6
22:30 penk joined perl6
22:31 [BlueT_at_Mars] joined perl6
22:32 [BlueT_at_Mars] penk: :p
22:37 svnbot6 r10357 | lwall++ | reverse examples are conflicting with built-in, renamed to rev
22:47 TimToady ?eval say"foo"
22:49 svnbot6 r10358 | Darren_Duncan++ | ext/Relation/ : TODO-ize the test suite since module prerequisites not ready yet
22:50 TimToady madgolfer.p6 is failing on the say"foo" problem.
22:52 TimToady that should parse.  just because say.bar requires the absence of whitespace doesn't mean that say"foo" requires whitespace...
23:01 rashakil_ left perl6
23:11 penk left perl6
23:22 ruoso joined perl6
23:26 Juerd stevan: It's unlikely that we will be online and active at the same time. Try the asynchronous aspect of IRC, and skip the ping :)
23:26 Juerd afk
23:26 Juerd (z)
23:45 bsb joined perl6

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

Perl 6 | Reference Documentation | Rakudo