Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2012-06-19

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:08 adu joined #perl6
00:11 flussence hm, awkward.
00:11 flussence this compile failure isn't caused by 32-bit *or* too new system libs
00:14 geekosaur joined #perl6
00:19 kurahaupo joined #perl6
00:59 pmichaud phenny: tell felher there was a camera in the room with my first talk; I have no idea if the video was recorded or when it might appear.  Maybe ping the YAPC::NA organizers :-)
00:59 phenny pmichaud: I'll pass that on when felher is around.
01:04 scott_ joined #perl6
01:10 am0c joined #perl6
01:11 Psyche^ joined #perl6
01:11 cj__ joined #perl6
01:17 UncleFester6 joined #perl6
01:18 UncleFester6 flussence:  it just occurred to me that your failure is on a netbook and mine is on a laptop.  Any ideas on testing if it is a limited memory problem?
01:19 UncleFester6 I think my laptop has 1 GIG ...
01:19 flussence hm, mine's got 1GB + about half a gig of swap. Never ran out of memory with it though...
01:20 sorear rakudo needs about 1-1.5 G currently I think
01:20 flussence it's 32-bit, usually about 800MB there
01:22 hypolin joined #perl6
01:22 flussence and it fails a bit too quickly to be OOM, I think.
01:39 UncleFester6 flussence: if your desktop uses bash you might try ulimit to build with limited memory ... just a thougt - I'll try to increas my swap tomorrow
01:44 mcglk joined #perl6
01:44 mcglk_ joined #perl6
01:45 flussence I might give that a try tomorrow
01:45 flussence &
01:46 UncleFester6 left #perl6
01:47 mcglk joined #perl6
01:57 tyatpi joined #perl6
01:58 mcglk joined #perl6
02:09 adu joined #perl6
02:12 colomon joined #perl6
02:13 raiph joined #perl6
02:24 adu I'm going to try being traditional here:
02:24 adu o/
02:28 quiefanatic joined #perl6
02:28 quiefanatic Good ** #perl6
02:28 adu hi quiefanatic
02:29 quietfanatic Not sure what quie is but I'm probably not it.
02:30 adu heh
02:30 adu tab completion gone wrong
02:30 * adu loves it when stuff justifies, tho
02:30 quietfanatic No, it was me typing my nick in wrong
02:31 quietfanatic but also that I guess
02:31 raiph o/
02:31 quietfanatic \o
02:32 adu o/
02:33 sorear o/ quietfanatic
02:33 raiph reading pmichaud's list slides http://pmichaud.com/2012/pres/yapcna-lists-1/slides/start.html
02:33 raiph "By default, arguments bind one-to-one without flattening"
02:34 raiph but "Slurpy parameter is both lazy and flattening"
02:35 raiph have to do ** for non-flattening slurpy
02:35 quietfanatic Still think the slurpy should be | to match with prefix:<|>
02:36 quietfanatic but who listens to me
02:36 raiph might be nicer to be consistent, and have *@sluroy mean non-flattening
02:36 raiph s/sluroy/slurpy/
02:36 raiph quietfanatic++
02:36 quietfanatic Well, in other places, ** is a multi-dimensional version of *
02:36 raiph that's what i was about to add
02:37 quietfanatic like, I think @lol[*] is different from @lol[**]
02:37 quietfanatic I could just be making this up though
02:38 raiph quietfanatic: are you saying that's a counter argument to using |*@slurpy (or *|@slurpy)?
02:38 quietfanatic raiph: No, it's a justification for *@slurpy being 1-dimensional and **@slurpy being multidimensional
02:39 quietfanatic With | I guess it would be |@slurpy and ||@slurpy
02:39 quietfanatic Hmm, maybe that's not such a good idea after all
02:39 mcglk joined #perl6
02:39 adu omg, I don't understand
02:40 adu what's the difference between my @a = 1,2,3; and my $c = [1,2,3];?
02:40 raiph oh i wasn't thinking | for indicating slurpy, just flattening, to be consistent with flattening on call
02:40 adu (slide15)
02:40 quietfanatic Hm
02:41 adu python uses * for slurpy
02:41 sorear adu: the first one declares @a, the second one declares $c
02:42 quietfanatic adu: They do the same thing in that statement, but in one case you have an @ variable and in one case you have a $ variable
02:42 sorear they are two different variables, the sigil is part of the name
02:42 quietfanatic adu: @a will generally flatten, and $a will never flatten even if it contains a list
02:42 adu is this something I would know if I knew Perl5?
02:42 quietfanatic Not exactly, no
02:43 quietfanatic In Perl5, @a must always be a list, and $a must be either a scalar or a reference to a list
02:43 leprevost joined #perl6
02:43 quietfanatic In Perl6, everything is a reference (kinda), so @a and $a are stored the same way
02:44 quietfanatic The only difference is that @a may only contain something that does the Positional[] role, and it will automatically flatten in some contexts.
02:44 adu I think I'll stick with parsing XML, that's easy
02:44 quietfanatic like in for @a { .say }
02:45 quietfanatic hrrm
02:45 adu so if I want non-flattening positional objects, then I should always use $?
02:45 quietfanatic I guess
02:45 quietfanatic yes.
02:46 adu I'm used to seeing some funky notation when you want flattening, like ',@'
02:46 raiph adu: i don't think it's that black and white
02:47 raiph adu: i think scalar containers generally don't flatten, positionals do
02:48 raiph s/positionals do/positionals generally do/
02:48 raiph but you can be explicit, to make either go the other way
02:48 adu but if I say my $y = @x; then can't I do $y[2]? which would imply it supports the positional role
02:49 quietfanatic You can.
02:49 raiph and, as this is perl, there are probably exceptions when that's the best thing overall
02:49 quietfanatic $ is the least restrictive sigil.
02:49 quietfanatic @ requires Positional, % requires Associative, and & requires Callable
02:49 quietfanatic , I believe.
02:50 raiph adu: hence the "in general" bit of "In general, non-scalars interpolate ("flatten") in lists, and scalars do not:  "
02:51 adu so @ implies Positional, % implies Associative?
02:51 quietfanatic yes
02:51 raiph goodnight all
02:51 quietfanatic at least, according to how I heard it from TimToady a few years ago :)
02:51 quietfanatic night raiph
02:52 adu ooo there is a ',@'
02:52 adu .list
02:52 adu I like that
02:52 adu raiph: good night
02:53 quietfanatic $a.list is equivalent to @($a), I believe
02:53 quietfanatic and @a.item is equivalent to $(@a)
02:53 quietfanatic and $a.lol is equivalent to @@($a)
02:55 quietfanatic Check that; TimToady says there's no @@() any more
02:55 tyatpi_ joined #perl6
03:03 adu a "few years ago"?
03:03 quietfanatic Perl 6 has been around quite a while, though it's been mutating.
03:03 awwaiid joined #perl6
03:04 adu I remember hearing about Parrot 2006 ish
03:05 quietfanatic Perl 6 first started being designed in 2000ish
03:06 [Coke] adu: parrot dates back to 2001 or so
03:07 adu I seem to remember hearing that
03:07 quietfanatic Parrot was designed primarily with p6 in mind, though it was hoping to also support p5
03:07 quietfanatic the latter project being called PONIE
03:08 sorear PONIE was officially abandoned a couple years ago
03:08 adu I thought no one understood p5 enough to reimplement it
03:08 quietfanatic Yep.  Turns out only Perl can run perl. :)
03:08 quietfanatic (5)
03:09 adu I could care less about ponie, pynie, pruby or whatever else used to run on parrot, I just need a scheme
03:09 quietfanatic Don't quote me on this, but I don't think Parrot is going to survive once Rakudo supports another platform.
03:10 adu and I've looked into what code I could find (most of what used to be in /languages is pretty hard to locate now)
03:11 quietfanatic though that probably depends on whether m0 ever comes around.
03:11 [Coke] adu: http://trac.parrot.org/parrot/wiki/Languages
03:12 adu [Coke]: yeah, I know, most of those links are dead
03:12 [Coke] (linked to from http://parrot.org/languages)
03:12 adu hence "hard to locate"
03:12 sorear Ethernet will never die.
03:13 sorear Any system developed to replace it will be called "Ethernet".
03:14 quietfanatic A rose by any other smell...
03:14 mberends joined #perl6
03:14 adu right
03:14 adu but eclectus works, but it isn't written in Parrot/PGE/Perl6Grammars, it's written in pure scheme
03:14 adu pheme is written in Perl6, and I'd like to find it so I can borrow ideas
03:15 adu but I can't find it
03:15 sorear mberends?  You return?
03:15 mberends heh
03:15 quietfanatic How is a scheme implementation written in scheme on Parrot?
03:15 [Coke] adu: I clicked on the first 18 or so. all but 2 links work.
03:15 adu [Coke]: probably the ones I just mentioned
03:16 mberends sorear: it would be rude not to
03:16 [Coke] "most are dead". please keep the hyperbole to a minimum. ;)
03:16 [Coke] ironically, those are ones originally hosted in parrot's svn server. they are probably there, just not updated to point at their new git location.
03:17 [Coke] https://github.com/parrot/pheme
03:18 [Coke] last updated: 2 years ago.
03:21 adu oOo https://github.com/parrot/parrot/tree/00edc124dc8fb07f03f256bcf81fc14262268478/languages/scheme
03:21 adu found it
03:21 [Coke] since I found it, perhaps you could update the wiki page for me to point to the new location.
03:21 adu you found pheme, i found scheme/p5
03:22 [Coke] nice.
03:22 adu I don't see an edit button
03:26 [Coke] are you logged in?
03:26 * adu registered
03:26 [Coke] FYI, pheme builds but fails immediately at runtime. I'll see if I can resurrect it.
03:26 adu I didn't get an email yet, can I still login?
03:26 adu I can do the dirty work
03:27 [Coke] eh. rather than fight it, I'd probably settle for opening a github issue at this piont.
03:29 adu I still don't see an edit button
03:30 betterworld joined #perl6
03:31 [Coke] there's a reason parrot folk don't use trac much anymore.
03:32 benabik joined #perl6
03:33 [Coke] ah. https://github.com/parrot/parrot/wiki/Languages
03:35 pmichaud I bet pheme could be quickly rewritten in nqp or perl6 :-)
03:36 pmichaud fsvo "quickly" :-)
03:37 pmichaud afk # sleep or something
03:38 pmichaud mberends: o/ !!!!
03:38 * pmichaud notices a nick he hasn't seen in a while.
03:40 mberends pmichaud: hi! thanks for email, will respond after getting enough bandwidth to watch the videos in a day or two :-)
03:41 pmichaud mberends: take your time; glad you got them, hope all is well with you
03:41 mberends yes, and ditto for you. I must also be afk for some hours &
03:42 pmichaud same here -- bbl
03:42 pmichaud afk # sleep or something, again
03:45 [Coke] msg pmichaud I have to rewrite partcl-nqp in nqp before tackling pheme that way
03:45 [Coke] ETOOMANY bots, he'll get it in backscroll.
03:46 xinming joined #perl6
03:46 neo_wulf joined #perl6
03:52 araujo joined #perl6
04:06 sisar [Coke]: http://wiki.perl6.org/phenny (re msg ...)
04:09 [Coke] sisar: yes, thank ou.
04:16 gardnan joined #perl6
04:18 quietfanatic good night #perl6
04:50 cog_ joined #perl6
04:56 geekosaur joined #perl6
04:56 birdwindupbird joined #perl6
05:45 dalek perl6.org: 742d0e5 | (John Gabriele)++ | source/ (2 files):
05:45 dalek perl6.org: Added links to categorized articles and videos pages on the wiki.
05:45 dalek perl6.org: review: https://github.com/perl6/perl6.org/commit/742d0e5b67
05:46 fhelmberger joined #perl6
05:58 cognominal joined #perl6
05:58 kaleem joined #perl6
06:07 wtw joined #perl6
06:55 sorear seen pmurias
07:06 brrt joined #perl6
07:07 kresike joined #perl6
07:07 kresike good morning all you happy perl6 people
07:13 moritz \o
07:13 sftp joined #perl6
07:15 frettled Goooood moooorning! :)
07:22 tadzik hello hello
07:22 eiro felher, no streaming at all at fpw
07:23 eiro no recording for the classrooms
07:23 eiro but i really hope we'll do screencasts after the event
07:23 eiro for the moment, i just can't install panda btw ;)
07:26 tadzik oh noes
07:30 moritz tadzik: speaking of panda, can you please push your .pod-installing commit to a branch?
07:30 tadzik moritz: oh, sure
07:31 dalek panda/pod-installing: 551cdc5 | tadzik++ | / (4 files):
07:31 dalek panda/pod-installing: First attempt at installing pod aside .pms and .pirs. Tests work, modules not really
07:31 dalek panda/pod-installing: review: https://github.com/tadzik/panda/commit/551cdc5393
07:31 tadzik there we are
07:36 sorear good morning kresike, frettled, tadzik
07:36 eiro tadzik, it's about a :3000 somewhere in nl
07:44 tadzik oh
07:44 tadzik yes
07:47 tadzik seems that the Dancer app that was serving the modules list broke :(
07:49 tadzik eiro: ok, should be fine now
08:02 sftp joined #perl6
08:03 mberends_ joined #perl6
08:05 sporous joined #perl6
08:10 cognominal joined #perl6
08:12 masak *yawn*
08:12 masak morning, meatbrains.
08:12 frettled mmmmeat
08:13 moritz good masak, morning
08:13 odoacre joined #perl6
08:13 masak They're Made Out Of Meat. http://www.youtube.com/watch?v=gaFZTAOb7IE
08:13 mberends joined #perl6
08:13 masak moritz: I'd like to chime in with the others in congratulating you. warm congratulations and best wishes.
08:13 mberends o/ masak
08:13 masak I know this was only a ceremony, and that the paperwork happened two years ago. ;)
08:14 masak mberends!!! \o/
08:14 moritz masak: thank you. And it wasn't only a ceremony, it was also a really great party :-)
08:14 masak sounds wonderful.
08:16 * masak breaks a bit of fast
08:16 mberends dunno how moritz passed the waltz test. stealth, methinks ;-)
08:18 moritz everybody was too polite to point out I failed it, t'was all :-)
08:18 sorear masak!
08:19 tadzik hello mberends, masak :)
08:19 sergot joined #perl6
08:19 sergot hi o/
08:19 sorear o/ sergot
08:19 mberends dobro jutro, ljudi :)
08:22 eiro tadzik, a dancer app ? not a bailador one ?
08:23 tadzik eiro: no, it predates Bailador :)
08:23 tadzik but it should be possible for it to run on bailador
08:23 eiro tadzik, any source somewhere ?
08:23 tadzik it's just that sometimes the response times may reach 10 seconds or so :)
08:23 eiro why ?
08:24 tadzik eiro: just this: https://gist.github.com/2953005
08:24 tadzik eiro: JSON parsing is still not so fast, for big files as the modules list
08:24 tadzik although it's well below 10 seconds these days, I think
08:24 eiro (i don't use dancer anymore btw: plack is everything i need
08:25 eiro tadzik, any NCI for that ?
08:25 tadzik eiro: nope
08:28 xinming joined #perl6
08:29 eiro well ... that's because you parse all the json at every request
08:29 eiro any way to see the json ?
08:29 eiro how is it built ?
08:29 tadzik I don't have to do that
08:30 tadzik I could keep a hash, and rebuild it only if the file is newer than it
08:30 tadzik but when I'd have to reload it, that'd take time
08:30 eiro got it
08:31 eiro tadzik, why not create a json by module ? you don't have to reload the complete database then
08:31 tadzik create json by module?
08:31 arnsholt eiro: If you're a bit familiar with C, it might not be too hard to hack together a binding to a C JSON library with Zavolaj
08:32 moritz tadzik: I think eiro wants to have many smaller JSON files instead of a big one
08:32 eiro tadzik, yep: i think it's another idea to work on atfpw
08:32 eiro moritz, right
08:32 moritz so that one parses a smaller file when talking about a single module
08:33 eiro arnsholt, any good docs about it ?
08:33 eiro tadzik, i'm pretty sure there no need of json actually
08:33 tadzik possibly :)
08:33 eiro we can do something way faster and written in perl6
08:33 moritz it was probably just the most convenient at that time
08:33 sorear eiro: at the same time?
08:33 tadzik parsing projects.json apparently takes 1340094565 1340094568 3 3
08:34 tadzik whatever that means :)
08:34 tadzik https://gist.github.com/2953027
08:34 eiro sorear, sorry ?
08:34 tadzik so about 3 seconds actually
08:34 eiro same of what ?
08:35 sorear eiro: california idiom, how to explain...
08:36 sorear eiro: 'written in perl 6' and 'way faster' - both of these are, individually, doable
08:38 eiro oh ... yes! the idea is to write perl6 stuff in perl6
08:39 eiro the json file can be split on little files with 1 info by line: nothing to parse: just aquire IOs
08:39 tadzik oh, mst is coming for fpw
08:39 eiro i asume IOs are pretty fast C bindings, so the server can run quickly :)
08:40 eiro tadzik, yep :)
08:40 tadzik nope :)
08:40 tadzik the server is pure perl 6
08:40 tadzik well, the one that I usually use for bailador
08:40 tadzik and it's the only one I know of that works with bailador
08:40 eiro tadzik, yep but perl6 uses rakudo for IOs and rakudo uses IO syscalls, right ?
08:41 eiro so how can this be slow?
08:41 tadzik well, syscalls are not the problem
08:41 tadzik but things like parsing HTTP headers in perl 6 may not be too fast on rakud'
08:41 sorear eiro: syscalls are the only way to do I/O on a Unix system as non-root
08:45 masak sorear! \o/
08:45 sorear masak: how are you?  that was a long plane trip! :D
08:47 masak heh.
08:47 masak parts of London had good wifi, but I didn't have an urgent need to check in on all y'all.
08:47 dakkar joined #perl6
08:51 arnsholt eiro: The Zavolaj docs are pretty good, yeah
08:51 arnsholt There's also the test suite, which is pretty simple
08:52 arnsholt And if all else fails, you can ask me =)
08:53 moritz one can also look at DBIish for examples
08:53 arnsholt Quite
08:59 arnsholt And this reminds me that I should try to get back to getting strings in structs working
08:59 eiro re
08:59 eiro sorear, yep! so this part is fast. storing a simple line into a scalar must be fast also (i assume=
09:00 eiro it come slow when it comes to parsing
09:00 arnsholt Maybe I can find some of those tuits in the couch...
09:00 eiro i don't like DMBS ... perhaps a redis client must be more usefull
09:01 sorear You'd be suprised what things aren't fast in current p6 impls
09:01 sorear There's a lot of hidden logic in an assignment statement
09:01 eiro arnsholt, i'll test
09:06 dakkar_ joined #perl6
09:06 moritz though scalar assignment is done in C in Rakudo, and shouldn't be too slow ;-)
09:07 eiro tadzik,
09:07 eiro https://gist.github.com/2953133
09:08 eiro ok then.. my question is: honestly: what is the current possible daily usage of perl6 ?
09:10 tadzik eiro: huh, when did that happen
09:11 tadzik I think it's also now possible to write the bootstrap.pl in pure perl 6, since we have 'use lib' now
09:13 sunflower joined #perl6
09:16 dalek niecza/non-bootstrap: 904e349 | sorear++ | lib/ (2 files):
09:16 dalek niecza/non-bootstrap: Actions pt 15
09:16 dalek niecza/non-bootstrap: review: https://github.com/sorear/niecza/commit/904e349a24
09:17 sorear sleep&
09:17 frettled sleep: missing operand
09:31 masak sleep well, sorear. dream of the conjunction of 'written in perl6' and 'way faster'.
09:36 tokuhirom joined #perl6
09:40 jerome_ joined #perl6
09:52 xinming joined #perl6
10:02 orafu joined #perl6
10:04 snearch joined #perl6
10:07 eiro tadzik, it happens when i launch boostrap.sh
10:08 eiro sorry to not investigate myself
10:08 eiro no time
10:08 eiro bootstrap.pl++
10:17 apejens joined #perl6
10:26 dalek rakudo/no-strict: a9401fb | moritz++ | src/Perl6/Grammar.pm:
10:26 dalek rakudo/no-strict: parse "no strict"
10:26 dalek rakudo/no-strict: review: https://github.com/rakudo/rakudo/commit/a9401fb47d
10:28 moritz that didn't work out the way I wanted it to. Good thing it's in a branch :-)
10:34 jaldhar joined #perl6
10:34 * masak .oO( good thing it's a DAG and none of the commit paths is special )
10:34 daxim joined #perl6
11:00 pmichaud good morning, #perl6
11:06 masak good am, pm.
11:06 mberends o/ pmichaud
11:07 tadzik \o
11:08 masak tadziku! \o/
11:09 masak is le finals hunting season over?
11:13 tadzik ba
11:13 tadzik it just starts today
11:13 tadzik the last exam is on the day I depart to FPW
11:13 tadzik I'm tired with all this
11:16 * mberends feels for tadzik: prevail, dude!
11:16 tadzik I shall :)
11:16 tadzik I have high enough grade average from the previous semesters, but this one is a disaset
11:16 tadzik * disaster
11:18 * masak empathizes
11:19 tadzik on the brighter side, it's possible that I'd be doing a Perl 6 Lightning Talk on today's Warsaw Ruby Users Group meeting :P
11:19 timotimo convert all the ruby users :]
11:20 masak if what some Ruby people say is true and Perl 6 is just Ruby, then conversion is a no-op :P
11:22 tadzik okay, gotta go to my exam &
11:22 masak good luck
11:22 tadzik thanks
11:24 GlitchMr joined #perl6
11:30 UncleFester6 joined #perl6
11:32 UncleFester6 phenny: tell flussence I bumped my swap space up by 2 gig and the build failed the same way
11:32 phenny UncleFester6: I'll pass that on when flussence is around.
11:32 UncleFester6 left #perl6
11:33 apejens joined #perl6
11:37 UncleFester6 joined #perl6
11:38 felher eiro: okay, thanks :)
11:38 phenny felher: 00:59Z <pmichaud> tell felher there was a camera in the room with my first talk; I have no idea if the video was recorded or when it might appear.  Maybe ping the YAPC::NA organizers :-)
11:38 felher pmichaud: okay, thanks :)
11:38 UncleFester6 left #perl6
11:38 timotimo i'm also interested in the recording of the first talk
11:39 JimmyZ joined #perl6
11:39 pmichaud istr someone saying there was trouble with the audio in the first talk, though... I guess we'll have to wait and see what gets published.
11:40 apejens joined #perl6
11:42 pmichaud maybe I can get my local .pm group to let me give the talks for recording :)
11:42 masak sounds like something a local .pm group could potentially be persuaded to do, yes ;)
11:43 masak jnthn, dangit, we need a local .pm group!
11:43 awwaiid where you at, masak?
11:43 pmichaud it's been a long time since our .pm group met, although a subset of our group meets monthly.  But we don't meet in a place that would be conducive to presentations or videography
11:43 masak awwaiid: http://en.wikipedia.org/wiki/Scania
11:44 awwaiid nice
11:44 masak feels like living right next door to the European continent.
11:46 Moukeddar joined #perl6
11:46 awwaiid considering that it appears to actually _be_ right next door to the European continent...
11:47 masak well, geographical location is one thing. but the bridge really helps, too.
11:47 Moukeddar Hi masak :)
11:47 pmichaud masak: is it challenging to meet up with copenhagen.pm, ooc?
11:48 masak pmichaud: not really. great idea.
11:48 masak Moukeddar! \o/
12:00 sjn \o
12:00 sjn !seen jnthn
12:01 sjn or not?
12:01 pmichaud phenny seen jnthn
12:01 pmichaud phenny: seen jnthn
12:01 pmichaud bah
12:02 pmichaud seen jnthn
12:02 pmichaud ENOALOHA
12:03 sjn pmichaud: any news on your side about the hague application?
12:03 pmichaud anyway, jnthn++ was last seen 5 days, 12 hours ago
12:03 sjn hm
12:03 pmichaud sjn: no news here, no.
12:04 sjn pmichaud: we *have to* know at absolutely latest on monday
12:04 pmichaud sjn: yes, I know.  I'll be working on trip stuff later today.
12:04 pmichaud right now I have to get $kid to $school -- bbiaw
12:04 sjn ook
12:05 pmichaud (maybe check with krunen and see where things stand on that end?)
12:05 masak pmichaud: IIRC, jnthn was with friends until Sunday and then went back to Sweden for $course.
12:05 sjn pmichaud: I've checked with krunen, thanks
12:05 sjn but I'm kinda trusting you guys to figure out this
12:09 brrt joined #perl6
12:21 dalek rakudo/no-strict: 27b55b5 | moritz++ | src/Perl6/Grammar.pm:
12:21 dalek rakudo/no-strict: remove arglist parsing for now; it does not work
12:21 dalek rakudo/no-strict: review: https://github.com/rakudo/rakudo/commit/27b55b5e42
12:21 moritz $ ./perl6 -e 'no strict; $x'
12:21 moritz Can only use get_who on a SixModelObject
12:23 att joined #perl6
12:26 dakkar joined #perl6
12:26 masak when you go MOP, error messages start sounding like philosophy.
12:26 moritz it seems it doesn't appreciate my mucking with $*UNIT_OUTER
12:27 flussence pre-emptive "argh"
12:27 phenny flussence: 11:32Z <UncleFester6> tell flussence I bumped my swap space up by 2 gig and the build failed the same way
12:27 moritz masak: you mean like "Cannot compose A, because it's not composable"?
12:28 masak moritz: mostly meant that 'get_who' sounds kinda ponderous.
12:28 moritz masak: though note that it's not get_how (which would be meta)
12:28 masak I wouldn't consider "Cannot X A, because it's not X-able" an informative error message.
12:29 masak or rather, it says the same thing twice.
12:29 moritz r: class A { }; class B does A { }
12:29 p6eval rakudo ddfb9f: OUTPUT«===SORRY!===␤B cannot compose A because it is not composable␤»
12:29 masak "Tried to X A, but A isn't X-able [so it didn't work]" is fine.
12:31 pmichaud "Once again, you've asked for the impossible, and while I can normally provide it, in this case I'm not ready to do that."
12:31 moritz pmichaud++
12:33 masak "We only provide synthetic imposibilities, not analytical ones."
12:34 am0c joined #perl6
12:36 [Coke] is p6eval setup in any special way in regards to pugs?
12:36 [Coke] extra env vars or something?
12:37 * moritz has no idea, and takes a look
12:37 moritz cmd_line    => "PUGS_SAFEMODE=true LC_ALL=en_US.ISO-8859-1 $home/Pugs.hs/Pugs/pugs %program",
12:39 masak that would explain why Pugs has encoding problems.
12:40 moritz I added that specifically because pugs had encoding problems, and at that time it solved them
12:40 masak ah.
12:41 [Coke] there's no one right LC_ALL setting, it seems, and without it, locally, I get errors from decodeUTF8' . will continue poking.
12:45 dalek rakudo/nom: d4c8e6d | moritz++ | src/core/ (2 files):
12:45 dalek rakudo/nom: more exception typos
12:45 dalek rakudo/nom:
12:45 dalek rakudo/nom: caught by Nick Glencross
12:45 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/d4c8e6d0d3
12:46 sisar joined #perl6
12:48 masak sisar! \o/
12:48 sisar masak: hey !
12:49 mikemol joined #perl6
12:49 sisar it seems we had overlapping period of absence from #perl6 (first two weeks of June)
12:50 masak no, I was actually absent.
12:50 sisar though I was not on vacation :|
12:50 masak oh. :/
12:51 sisar It was not a good time for me... all I can say is "What doesn't kill you, makes you stronger."
12:51 * moritz hugs sisar
12:51 * masak hugs sisar too
12:52 masak that saying always felt like a bit of a last resort to me.
12:52 sisar masak: so July is coming ! Another month of bloogging ?
12:52 sisar moritz, masak: thanks
12:52 masak sisar: I'll think about it. quite possibly, yes.
12:52 sisar yay !
12:53 masak it's nice to know someone would like it to happen.
12:54 timotimo oh, more perl6 blogposts? i'd like that
12:54 * moritz would love them too. But mewants to see t5 and thus p6cc2011 too
12:55 masak I promise to wrap up t5 and p6cc2011 in June.
12:55 moritz s/see/see ... finished/
12:55 sisar seems like we had very less perl6 blogs this last month (modulo moritz++'s exception reports and jnthn++'s blogs). But yes, no perl6-learning blogs.
12:55 moritz oh, and I want qast and macros!
12:55 * tadzik is back
12:55 moritz and a ponie!
12:55 masak so it's a July of introductory Perl 6 material y'all want, then?
12:56 sisar and oh, I'm not a teenager anymore. I hope my impulsiveness has somehow magically disappeared :)
12:56 masak because I'm toying with this idea of starting to write about compilers and the act of compiling. but maybe save that for next summer.
12:56 PacoAir joined #perl6
12:56 masak sisar: happy belated birthday.
12:56 masak nr: say 20.base(2)
12:56 sisar masak: thank you !
12:56 p6eval rakudo ddfb9f, niecza v18-7-ga660099: OUTPUT«10100␤»
12:56 * [Coke] wonders how to put a putStrLn into an arbitrary location in haskell code.
12:56 moritz sisar++
12:57 pmichaud wow, looks like I can get  DFW->FRA->SVG->DFW for under $1500 US
12:57 moritz [Coke]: you can't
12:57 [Coke] pugs: say 20.base(2)
12:57 p6eval pugs: OUTPUT«*** No such method in class Int: "&base"␤    at /tmp/Cku41l2LHV line 1, column 5 - line 2, column 1␤»
12:57 sisar moritz: I don't know how I forgot this, but congratulations !
12:57 moritz sisar: thanks
12:58 masak [Coke]: more LHF :)
12:59 [Coke] bah. I also cannot get ghci to run the debugger on Pugs. hurm.
12:59 moritz [Coke]: thing is, putStrLn is IO, and thus cannot be used in a pure function
12:59 [Coke] masak: I am trying to fix the decodeUTF8 issue.
12:59 masak [Coke]: is the problem with putting a putStrLn into an arbitrary location in haskell code that of monads and side effects?
12:59 [Coke] masak: blank stare.
12:59 masak sorry.
12:59 [Coke] "compilation error", mostly. ;)
12:59 masak what moritz said, instead.
13:00 moritz if a function does IO in haskell, it must be reflected in the type signature, and also in the way it's called
13:00 masak putStrLn returns IO, and so it doesn't mix well with things that don't already return IO.
13:00 moritz that's both upside and downside of a pure functional language
13:03 arnsholt There's a diagnostic function you can use anywhere to print though, isn't there?
13:03 arnsholt Can't remember what it's called
13:03 * [Coke] gives up for today.
13:03 arnsholt trace
13:03 arnsholt (Hoogle to the rescue)
13:03 masak arnsholt++
13:04 arnsholt Of course, you might get the string you wanted to print several times, but that's what you get in a non-strict language
13:05 arnsholt And I have to say I quite like the fact that you can search for a function according to the signature
13:06 masak yes, that's awesome.
13:09 Psyche^ joined #perl6
13:10 BaggioKwok joined #perl6
13:12 [Coke] arnsholt: if you speak haskell, do you have any interest in hacking on pugs?
13:12 * masak still does, but is constantly time-constrained
13:14 BaggioKwok joined #perl6
13:21 masak "Unit testing [...] is the practice of writing software to check that units of software behave in ways we expect." -- do you agree with this definition?
13:22 * masak is writing course slides
13:22 moritz yes
13:22 moritz do manual unit tests exist too?
13:23 masak I think so, but automated ones are the shit, so manual ones don't get much press.
13:23 masak manual unit tests would perhaps be necessary if the system included components outside of the reach of the software.
13:23 masak like an actual physical lever or something.
13:24 moritz then... robotics to the rescue :-)
13:24 masak heh, was thinking that too :)
13:29 moritz two years ago I had an interview with a company that does business software, and they didn't have automated tests
13:30 masak it's far to common, I'm afraid.
13:30 moritz they also had only parts of their code under version control
13:30 masak ditto.
13:30 masak sometimes I marvel at the speed some software houses manage to have despite their lack of proper tooling.
13:30 arnsholt [Coke]: I speak a tiny amount of Haskell
13:31 arnsholt I've been thinking of looking at Pugs, but my tuit-supply is low ATM, unfortunately
13:31 [Coke] arnsholt: that's more than me, and I've patched pugs. ;)
13:31 [Coke] (tuits) ah, the programmer's lament.
13:31 flussence .oO( two years ago the company I *work at* didn't have automated tests... )
13:32 moritz flussence: and you changed that?
13:32 arnsholt moritz: At Opera, the QA department had some amount of manual testing
13:32 arnsholt It turns out that testing a browser is hard in many interesting ways =)
13:32 flussence moritz: yes. on my own :)
13:32 moritz arnsholt: well, I can imagine that... exactly
13:33 arnsholt For example, you can have a bug in the rendering code, such that the screen you -want- to draw is correct, but what's shown on screen isn't
13:34 arnsholt So you need to get screenshots from the outside. Looking at the application's own framebuffer isn't enough ^_^
13:34 moritz oh, fun indeed
13:34 moritz though that sounds more like integration test than unit test
13:34 arnsholt Yeah, it's not exactly unit testing
13:35 moritz here we are really spoiled; compilers are very easy to test
13:36 masak compilers are input/output filters. filters are ideal for testing, indeed.
13:36 moritz right
13:36 masak but we still have to do some integration testing sometimes, with -e and stuff.
13:36 moritz aye
13:36 moritz masak: did you have a chance to look at my ufo branch yet?
13:36 masak ugh, no.
13:37 masak maybe tonight. at the latest on wednesday.
13:37 moritz masak: ok, great
13:37 masak um. thursday.
13:37 masak (what day is it?) :)
13:37 * masak has been on vacation
13:37 moritz r: say Date.today
13:37 p6eval rakudo d4c8e6: OUTPUT«2012-06-19␤»
13:37 moritz r: say Date.today.day-of-week
13:37 p6eval rakudo d4c8e6: OUTPUT«2␤»
13:37 masak :)
13:38 moritz r: say <Sunday Monday Tuesday Wednesday ...>[Date.today.day-of-week]
13:38 p6eval rakudo d4c8e6: OUTPUT«Tuesday␤»
13:43 masak no, I thought we went with 7 for Sunday.
13:43 * masak checks
13:43 masak aye. we did.
13:44 moritz says who?
13:44 masak S32/Temporal.
13:49 dalek joined #perl6
13:49 Util joined #perl6
13:49 masak there are arguments for both 0 and 7. we went with 7 because there is a slightly stronger case to be made for it.
13:49 masak I think feather is lagging quite badly.
13:54 PerlJam joined #perl6
14:00 dalek joined #perl6
14:02 PerlJam joined #perl6
14:03 masak train &
14:03 dalek joined #perl6
14:03 skids joined #perl6
14:09 BinGOs joined #perl6
14:15 majrmovies joined #perl6
14:24 kresike bye all
14:25 PerlJam .oO( why do I always think of West Side Story when I see "kresike"?  It's probably not even pronounced like "krupke" )
14:27 smash joined #perl6
14:28 smash hello everyone
14:28 colomon o/
14:28 TimToady hello smash you wonderful person you!
14:28 TimToady sorry, for some reason I associate that message with you...
14:29 moritz was that something that purl regularly said?
14:29 smash yeap, purl liked that
14:37 mucker joined #perl6
14:38 quietfanatic joined #perl6
14:39 sisar ETOOMANYFOO <--- what is this meme/pun ? I don't get it.
14:40 sisar does the 'e' stand for 'error' ?
14:40 moritz sisar: on UNIX there are lots of error codes which are all-caps and start with E
14:40 moritz "No such file or directory" is ENOENT
14:40 moritz (Error no entry)
14:40 huf not anymore, they've all gone treeish
14:40 PerlJam huf: heh!
14:40 sisar moritz: thanks.
14:41 sisar ETOOMANYMEMES
14:41 PerlJam sisar: ENOTENOUGHUNIX
14:41 moritz :-) you're welcome
14:41 TimToady .oO(note no ugh unix)
14:41 moritz but you usually only see that if you do some C programming
14:42 * TimToady tries to imagine caveman unix
14:42 TimToady unfortunately it's already after breakfast
14:43 huf moritz: or you're a sysadmin...
14:44 huf shitty things often just log some random error code (as the integer) and then you get to use perl to figure out what that error actually was :))
14:44 moritz huf: but that still gives you 'No such file or directory', not ENOENT
14:45 moritz $ perl -wE 'say $! = 1'
14:45 moritz Operation not permitted
14:45 huf true, true
14:46 * quietfanatic always hacked diagnostic messages into Haskell with unsafePerformIO
14:47 masak quietfanatic! \o/
14:47 quietfanatic masak!
14:47 masak hello from a train! :)
14:47 quietfanatic Hello from a smokey-smelling hotel room
14:47 * masak doesn't pretend to know how Smokey smells
14:47 quietfanatic Not very good.
14:48 masak included free of charge: second-hand smoking from the walls.
14:49 quietfanatic Well hey, can't argue with free ameni*cough cough*
14:50 masak :)
14:50 TimToady it was the last room in town, except for a flop-house several miles down the road and around the corner, and we weren't going to gamble on that
14:50 masak nod.
14:51 TimToady we were driving into town, and I saw a country club on the nagivator, so I joked "I'll bet there's a golf tournament in town, and all the hotels will be full."
14:51 TimToady we got into town, and there's a golf tournament..
14:51 TimToady and we barely got a room
14:52 TimToady I get one true prediction in my life, and I waste it on this...
14:52 masak *lol*
14:53 * masak .oO( hotel in one! )
14:53 TimToady btw, "nagivator" is not a typo, that's what I call it
14:53 [Coke] did a little birdie tell you that awful joke?
14:54 masak [Coke]: sigh. that's subpar. :P
14:54 * TimToady is teed off that all the good puns are taken
14:54 TimToady wait, there are no good puns...
14:55 * quietfanatic should learn to use a real IRC client
14:55 masak I heard a good pun this weekend.
14:55 masak can't remember it right now, though.
14:56 masak I'm quite fond of English/Mandarin hybrid puns at present.
14:57 * daxim makes a face like a horse
14:57 TimToady your ma!
14:57 masak :)
14:57 pmichaud back again
14:57 masak pmichaud, you're just in time for the horse jokes.
14:57 fgomez joined #perl6
14:58 * TimToady likes to walk into an Indian restaurant and say "Nan desu ka?"
14:58 * pmichaud takes horse jokes with a small bit of equine-imity  :-)
14:59 * quietfanatic caught a cold and is a little horse
14:59 quietfanatic best I can do folks
14:59 PerlJam quietfanatic: don't give it to the neighbors
15:00 quietfanatic That'd be a nightmare
15:00 TimToady someday some sociologist is going to figure out that its the puns that keep women away, not the loutishness...
15:00 tokuhirom joined #perl6
15:01 TimToady or perhaps those are the same thing...
15:01 quietfanatic Don't say that alout
15:01 PerlJam quietfanatic: just take some medicine and get plenty of rest and you'll be back in the saddle in no time.
15:01 TimToady keep trying...or better yet, don't...
15:02 quietfanatic I'll just relax and remain phlegmatic
15:02 TimToady I'm not sanguine about that...
15:02 jnthn Good afternoon
15:02 * jnthn is back
15:02 TimToady we'll make you regret that
15:03 jnthn uh-oh...
15:03 * jnthn quietly backs away again
15:03 flussence joined #perl6
15:03 masak we could keep keep punning for hourse.
15:03 masak jnthn! \o/
15:03 PerlJam .oO( How do you "gret" something so that you can "regret" it in the future? )
15:04 masak PerlJam: I always read that "re" as "backwards in time" (like "regress"), not "repeated action" (like "re-state")
15:04 quietfanatic and we all know 're' involves backtracking
15:05 masak oh man.
15:05 pmichaud oh, I suspect "regret" is the act of repeatedly remembering what you originally gretted.  :)
15:05 masak quietfanatic: you should hang out here more often :)
15:05 * jnthn wonders if anything much happened in his absence besides punning :)
15:05 * TimToady should trot off to the shower now
15:05 PerlJam quietfanatic: what masak said.  That last one was quite good  :)
15:05 masak jnthn: we got so low on puns without you, we had to take the reins...
15:06 quietfanatic Thanks guys, I'm still studying.
15:06 * pmichaud bridles at that last one.
15:08 masak remember us when you have more time.
15:09 arnsholt masak: The re- is more repetition than backwards
15:09 * PerlJam members again
15:10 masak arnsholt: that sounds like quite a blanket statement to me. please delve.
15:11 arnsholt Well, re- is a derivational morpheme, so it's independent meaning is fuzzy at best
15:12 arnsholt But it does indicate repetition. re-gress for example would be something like "go again" at its base. That is, "go back"
15:12 arnsholt Which has then been used metaphorically in English
15:13 PerlJam all this membering, gressing and gretting weirds me.  ;)
15:13 masak oh, ok.
15:13 masak arnsholt++ # knows stuff about etymology
15:13 arnsholt Funny enough, regret is apparently the Latin prefix with a Germanic verb
15:14 quietfanatic We're disgressing into dismembering the english langauge
15:14 quietfanatic with...um...disgrets.
15:14 PerlJam disgret, congret, ingret?
15:14 diakopter egrets
15:14 arnsholt (gråte in Norwegian, probably something similar in Swedish)
15:15 benabik diakopter: No need for the fowl language.
15:15 sisar .ety regress
15:15 phenny "late 14c. (n.), 'act of going back,' from L. regressus 'a return,' from regress-, pp. stem of regredi 'to go back,' from re- 'back' (see re-) + gradi 'to step, walk' (see grade (n.))." - etymonline.com
15:15 arnsholt Oh, and looking at my dictionary closer, re- has several more meanings, including reverse or backwards movement
15:15 masak ooh, .ety
15:15 pmichaud progress: to go forward.   congress:  to go .... backward?
15:15 arnsholt con is together
15:16 adu joined #perl6
15:16 masak as a seasoned Nomic player, I have full respect for the difficulty in getting anything legislated ;)
15:16 sisar masak: *nod*
15:16 PerlJam profit, refit, confit, outfit
15:17 * pmichaud has a contrary opinion :)
15:17 arnsholt Heh. Not sure how that one is derived
15:17 sisar .ety contrary
15:17 phenny "mid-14c., from Anglo-Fr. contrarie, from L. contrarius 'opposite, opposed,' from contra 'against' (see contra)." - etymonline.com
15:18 masak right. "contra"
15:21 adu I think there should be more guns in congress
15:22 Chillance joined #perl6
15:24 arnsholt And contra is con-ter which is the negation of con =)
15:24 adu masak: Nomic sounds fun, is it still played?
15:25 masak adu: I believe so.
15:25 masak adu: a bunch of us in Uppsala had an instance running for years.
15:26 masak I'd like to play it again some day.
15:26 adu I wonder if it's possible to write a Nomic server in Perl6
15:27 masak yes, but with the caveat that Nomic rules expressed in code has quite a different phenotype to the original legalese Nomic.
15:27 masak part of the charm of Nomic is that it's actually law running in people's brains, subject to uncertainties and interpretation.
15:28 adu well, the same might be said of MagicTG, but there's an online game for that
15:29 BinGOs joined #perl6
15:31 masak yes, we had a Magic player among us in our instance.
15:32 masak two, even.
15:32 jnthn http://irclog.perlgeek.de/perl6/2012-06-15#i_5730212 references a sha-1 in the nqp repo. But..."git show" of it here complains its an unknown revision.
15:32 masak but I venture to say that not even Magic has the metacircular property.
15:33 masak Nomic does.
15:33 PerlJam jnthn: that's a rakudo SHA1
15:33 jnthn oh...
15:33 PerlJam jnthn: (which updates the NQP revision)
15:33 jnthn ah, I see
15:34 jnthn It mentioned NQP then gave a sha-1, which logically suggests... )
15:34 PerlJam I couldn't reproduce the problem though.  But I never tried it in a pristine environment.
15:34 thou joined #perl6
15:34 Vlavv` joined #perl6
15:36 masak jnthn: I once saw a PoC search engine that took a SHA-1 as input, and gave a github project back as output ;)
15:36 jnthn There were 7 NQP commits made before the Rakudo revision bump
15:37 jnthn It'd be really handy if somebody who could reproduce the issue could figure out which one it is.
15:57 sergot joined #perl6
15:58 snearch joined #perl6
16:00 kurahaupo joined #perl6
16:05 BinGOs joined #perl6
16:08 am0c joined #perl6
16:10 sorear good * #perl6
16:11 moritz \o sorear
16:21 masak good time unit, sorear.
16:38 kaare_ joined #perl6
16:42 harmil joined #perl6
16:43 harmil left #perl6
16:43 harmil joined #perl6
16:43 harmil Good morrow perl 6
16:44 colomon \o
16:44 * colomon just ordered a Raspberry Pi
16:45 harmil Have a project in mind for it?
16:45 colomon Mostly just want to have an ARM machine in the house.
16:46 colomon make sure my $work software runs on it, etc.
16:46 harmil nice
16:46 colomon I'm guessing it won't be able to handle building Rakudo...
16:46 harmil Oh no
16:47 colomon not enough ram
16:47 colomon I would love to be wrong about that!
16:47 harmil But you *might* be able to get it to cross-compile....
16:47 harmil Building a x-compile environment for the whole p6 build chain would be … interesting
16:48 colomon but probably not ... fun
16:49 harmil Depends on what sets your $?FUN
16:50 jnthn colomon: How much RAM does it have?
16:50 sorear aren't most smartphones ARM?
16:51 sorear jnthn!  You survived too! :D
16:51 colomon 256 mb
16:51 harmil Yeah, so that's a definite cross-compile target
16:51 colomon sorear: but most smartphones are not terribly convenient for running Linux.
16:51 jnthn sorear: Well, I was taking some time afk to properly rest for some days :)
16:51 jnthn I'm back now :)
16:52 sorear colomon: oh, so your work software requires Linux.  nm then.
16:54 harmil "work software requires Linux" -> optimizer -> "work"
16:54 colomon sorear: well, Windows will work as well.  But as far as I know, no current smartphone OS is really up to it.  I'll probably be pushing it to get my stuff working with 256 mb of RAM.  (Sigh.  I'm sure it could do that back in 1998...)
16:55 vlixes joined #perl6
16:59 BinGOs joined #perl6
17:12 araujo joined #perl6
17:26 hanekomu joined #perl6
17:36 vlixes joined #perl6
18:07 [Coke] p6: say ("this" ~~ s/is/at/) # should this work eventually?
18:07 p6eval rakudo d4c8e6: OUTPUT«Cannot assign to a non-container␤  in sub infix:<=> at src/gen/CORE.setting:11207␤  in block <anon> at /tmp/TTzRu41T9w:1␤␤»
18:07 p6eval ..niecza v18-7-ga660099: OUTPUT«Unhandled exception: Writing to readonly scalar␤  at /home/p6eval/niecza/lib/CORE.setting line 589 (Cool.subst @ 88) ␤  at /tmp/xA9_CuCVLS line 1 (mainline @ 3) ␤  at /home/p6eval/niecza/lib/CORE.setting line 3917 (ANON @ 3) ␤  at /home/p6eval/niecza/lib/CO…
18:07 p6eval ..pugs: OUTPUT«Error eval perl5: "if (!$INC{'Pugs/Runtime/Match/HsBridge.pm'}) {␤    unshift @INC, '/home/p6eval/.cabal/share/Pugs-6.2.13.20120203/blib6/pugs/perl5/lib';␤    eval q[require 'Pugs/Runtime/Match/HsBridge.pm'] or die $@;␤}␤'Pugs::Runtime::Match::HsBridge'␤"␤*** Can't locate P…
18:07 hanekomu joined #perl6
18:07 [Coke] I think I just made this mistake a few weeks ago. Hurm.
18:10 fridim_ joined #perl6
18:14 [Coke] https://gist.github.com/2955693 - that segfaults the REPL. whee.
18:14 * [Coke] rebuilds.
18:16 diakopter [Coke]: there's another error in rt that segfaults after the same error.. I can't remember which one it is
18:16 diakopter hm, or maybe we just discussed it here and didn't add the segfault to the ticket
18:17 diakopter but I do remember it being the same non-container assign error
18:21 diakopter so maybe it's that particular error that triggers the segfault upon exiting the repl
18:23 moritz r: my $x = 1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1;
18:23 p6eval rakudo d4c8e6:  ( no output )
18:23 * masak feels like Alice in Wonderland
18:23 [Coke] phenny: tell masak bot to checkout 14:14 < [Coke]> https://gist.github.com/2955693 - that segfaults the REPL. whee.
18:23 phenny [Coke]: I'll pass that on when masak is around.
18:23 araujo joined #perl6
18:23 [Coke] diakopter: there we go. ;)
18:23 sorear p6: say eval '+1' x 50
18:23 p6eval rakudo d4c8e6, niecza v18-7-ga660099, pugs: OUTPUT«50␤»
18:23 sorear p6: say eval '+1' x 500
18:24 p6eval rakudo d4c8e6, niecza v18-7-ga660099, pugs: OUTPUT«500␤»
18:24 sorear p6: say eval '+1' x 5000
18:24 stephenlb joined #perl6
18:24 moritz r: my $x = 1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1
18:24 p6eval rakudo d4c8e6, pugs: OUTPUT«(timeout)»
18:24 p6eval ..niecza v18-7-ga660099: OUTPUT«5000␤»
18:24 p6eval rakudo d4c8e6:  ( no output )
18:24 moritz star: my $x = 1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1
18:25 p6eval star 2012.05: OUTPUT«(timeout)»
18:25 sorear nice, niecza's eval is actually faster than rakudo's in at least one case :D
18:25 sorear n: say eval '+1' x 50000
18:25 moritz sorear: :-)
18:25 p6eval niecza v18-7-ga660099: OUTPUT«(timeout)»
18:27 birdwindupbird joined #perl6
18:27 sorear I'm actually very suprised that pugs times out here.
18:27 sorear o/ masak in wonderland
18:27 diakopter aww; we lost the perlesque target during the move
18:30 diakopter (understandable)
18:34 * masak whistles on "One" from "A Chorus Line"
18:34 phenny masak: 18:23Z <[Coke]> tell masak bot to checkout 14:14 < [Coke]> https://gist.github.com/2955693 - that segfaults the REPL. whee.
18:34 masak [Coke]: rebuilding latest nom now.
18:36 [Coke] masak++ - diakopter says a similar bug may be pre-reported
18:36 wooden joined #perl6
18:36 masak diakopter: URL?
18:36 wooden joined #perl6
18:37 [Coke] aw, you can't perlbrew (e.g.) 5.16.0 multiple times with different config opts.
18:37 [Coke] we aren't sure.
18:45 pmichaud jnthn: btw, you might want to take a look at https://gist.github.com/2950855
18:46 pmichaud looks like the optimizer goes quadratic or something on long expressions
18:47 jnthn pmichaud: Whoa.
18:47 jnthn pmichaud: I...don't see why it should. Will have to profile to get the answer.
18:48 jnthn pmichaud: If it's not already, probably good to file rakudobug with a link to that gist.
18:48 pmichaud given th elength of the optimize and PIR steps, I'm guessing that the intermediate constants are being held in a serialization context even if they aren't needed.  but that's just a guess.
18:48 pmichaud anyway, yes, I can file a ticket
18:48 cggoebel joined #perl6
18:52 pmichaud RT #113754
18:57 adu joined #perl6
18:57 spider-mario joined #perl6
19:00 adu joined #perl6
19:01 adu joined #perl6
19:02 adu joined #perl6
19:04 [Coke] IWBNI we could get ticket stuff cc'd in here.
19:04 [Coke] new ticket, closed ticket...
19:04 * [Coke] ponders writing an RT bot. I think i still have the CLI docs lying around somewhere...
19:06 adu joined #perl6
19:08 mucker_ joined #perl6
19:09 leprevost joined #perl6
19:10 mucker joined #perl6
19:32 PerlJam btw, I won't be able to get to the rakudo release until late Thu evening my time.
19:33 [Coke] http://infiniteundo.com/post/25326999628/falsehoods-programmers-believe-about-time
19:34 PerlJam s/There are always 24 hours in a day/There are always 86400 seconds in a day/  :)
19:37 sorear most of that article can be summed up as 'never use civil time for any reason, dumbo'
19:51 mikemol_ joined #perl6
19:57 [Coke] "fine, you jerk, it's 3:56!" ?
19:57 tadzik good evening
19:58 sorear good evening tadzik
19:58 cggoebel joined #perl6
19:58 jnthn o/ tadzik
19:59 sorear o/ jnthn
19:59 jnthn o/ sro
19:59 jnthn er
19:59 jnthn o/ sorear
19:59 jnthn ...tab completion. You have to hit the right keys *before* <tab> :)
20:00 jnthn PerlJam: Late release on Thu is fine. I won't get any tuits today, but may be some tomorrow night and more likely on Thursday.
20:01 jnthn If so that just means I can tidy more things up pre-release :)
20:10 prammer joined #perl6
20:14 rjbs joined #perl6
20:14 rjbs Hey everybody!  I have a question.
20:14 moritz hi rjbs. Shoot!
20:14 rjbs p5p has had a bit of dicussion lately about valid "punctuation variable" style variable names.
20:15 rjbs I'm not so concerned about what's valid in an identifier.  That's got better guidelines.
20:15 rjbs But the question is about non-ASCII counterparts to the variables like $^X and $$ and so on.
20:15 rjbs I'm curious as to what the Perl6 solutin is.
20:15 rjbs Sorry, my keyboard has been dropping keys today.  At least, I think it's my keyboard and not my fingers..
20:16 moritz hm
20:16 flussence p6 uses $*LONGNAME for most things iirc
20:16 rjbs I wasn't even sure punctuation variables survived, since there's $*etc
20:16 moritz right; we have about 5 "real" punctuation variables
20:16 moritz $_, $!, $/ are the most important ones
20:16 flussence the worst thing we've got is "$¢", if that's even still around...
20:17 rjbs but $foo¢bar isn't a valid identifier, right?
20:17 moritz it is, but you only need that when touching the regex guts
20:17 rjbs So there's still some kind of rule about one-char var names?
20:17 moritz nr: say so '¢' ~~ /\w/
20:17 p6eval rakudo d4c8e6, niecza v18-7-ga660099: OUTPUT«False␤»
20:17 moritz not a valid identifier, no
20:17 moritz rjbs: there is a short list of valid one-char var names
20:18 rjbs Okay.  So it sounds like very sadly I can't steal from p6 here. :)
20:18 moritz $¢, $/, $! are the only ones that fall out of the pattern
20:18 rjbs Right now there's a bit of contention becaues Perl 5 is allowing some crazy stuff.
20:18 moritz the rest follow rules, like $*PROGRAM_NAME having that * sigil
20:18 moritz and the user can make $*identifer-variables too
20:19 rjbs like single-char scalars with the names \N{THIN SPACE} or \N{U+D800} or \N{COMBINING CIRCUMFLEX}
20:19 moritz s/sigil/twigil/
20:19 rjbs yeah
20:19 rjbs $* implies globality?
20:19 moritz dynamic scoping
20:20 brrt joined #perl6
20:22 masak does Niecza have enough threading to create interesting effects when the same dynamics-scoped variable is changed in two different threads?
20:25 cggoebel joined #perl6
20:26 sorear masak: yes
20:27 sorear moritz: what's the fifth?
20:28 * [Coke] is glad to see rjbs here. I follow p5p halfheartedly, and wonder occasionally if we should offer any "here's how perl 6 does it" nots.
20:28 [Coke] *nots
20:28 [Coke] *NOTES
20:28 [Coke] rjbs++ # character dropping infectious.
20:28 moritz sorear: I was thinking about @_ and %_, but then realized they don't quite fit into the category
20:29 moritz otoh I counted $_
20:29 tomaw joined #perl6
20:29 sorear TimToady talks from time to time about @!
20:30 rjbs [Coke]: I'm always interested i how Perl 6 does it, but it's always dangerous for us to think we can copy, because of the fundamental differences.
20:30 moritz sorear: aye. In some sense my exception design so far doesn#t cover the most interesting cases
20:31 sorear rjbs: how familiar are you with PL_sawampersand?
20:31 [Coke] rjbs: already had one failure in that area, neh?
20:31 [Coke] rjbs++ # keeping p5 moving.
20:32 sorear this just reminded me of a question I had - would I be correct to think that the major part of the penalty is the need to keep a copy of the orig string?
20:32 PerlJam [Coke]: I think that was just premature copying really.
20:32 rjbs sorear: a bit
20:32 rjbs [Coke]: Yeah, although I'm still a bit dubious about ~~ in p6.  Too many of the core roles' matches seem arbitrary to me.
20:33 rjbs but it's definitely saner in p6
20:33 sorear I think there's a sane subset and a "pair the spares" periphery
20:34 rjbs Yeah.  Felt like there might've been some fill-in-the-box-ism.
20:34 sorear there's nothing particularly sane for $foo ~~ @bar *to* do... imo it should be disallowed, make the user say $foo ~~ any @bar or $foo ~~ all @bar
20:34 rjbs Right, that's what I'm hoping to move to in p5.
20:35 sorear like we recently disallowed %hash{@slice}:exists
20:35 * sorear wonders if that involves adding junctions to p5
20:35 rjbs sorear: Perl6::Junction; I use it regularly
20:36 moritz rjbs: we still have some design issues with ~~ on Perl 6 too
20:36 rjbs In fact, I should make a patch for ~~ across junctions!
20:36 moritz currently  $x ~~ m:g/a/   doesn't work by spec
20:36 rjbs seen cfranks
20:37 moritz our seen bot hasn't been seen for a day or two... I don't know how long exactly :-)
20:37 sorear I do
20:37 PerlJam rjbs: so $x ~~ @a would be an error with a nice message that tells them to use Perl6::Junction and such?
20:37 PerlJam (in p5)
20:37 rjbs PerlJam: I'm not sure whether it would mention P6::J, but yes.  Let me find my proposal.
20:38 rjbs http://www.nntp.perl.org/group/perl.perl5.porters/2011/07/msg174272.html
20:39 rjbs If you read through, though, you will find that making it pluggable, as we talked about, led to some complexity.
20:39 rjbs so it hasn't happened.
20:39 sorear after the current round of niecza refactors I'm hoping to get array regexes in, maybe then we can talk about ditching the nonintuitive Array ~ Array rules
20:43 rjbs That'd be cool.  I mean, they're certainly not hurting *me* now, but they did seem to be there just to fill in the slot, rather than because they were the obvious behavior.
20:45 PerlJam rjbs:  If you keep hanging out here, we may have a "crossing the streams" moment where new features are worked out and both p6 and p5 adopt the appropriate semantics.  :-)
20:45 [Coke] someone start working on p5.5 right now. :P
20:45 rjbs This is Perl, version 5, revision 500.
20:46 [Coke] mmhehehe.
20:46 sorear rjbs: I have heard rumours of a 2-day pre-YAPC::EU event...?
20:46 [Coke] r: say (500-16)/2+2012
20:46 p6eval rakudo d4c8e6: OUTPUT«2254␤»
20:46 * [Coke] urks.
20:47 flussence .oO( perl5's versioning scheme isn't y3k-compliant... )
20:48 sorear hopefully by y3k we will all have moved to perl 7.
20:51 PerlJam we'll digitally encode our thoughts on the fabric of the universe and the natural expansion will actually execute them.
20:51 rjbs By y3k I hope to be retired and mostly writing Scheme.
21:05 masak by y3k I hope to be doing whatever has replaced programming.
21:07 masak (some higher-order design/architecture activity, most likely. or something akin to guiding programming bots along as they montecarlo and recombine on their own)
21:16 phao joined #perl6
21:16 phao left #perl6
21:21 tyatpi_ joined #perl6
21:21 bruges joined #perl6
21:22 quietfanatic joined #perl6
21:29 rjbs Yeah, I figure my programming in y3k will just be for fun.  Like to power my D&D game.
21:30 masak :)
21:30 rjbs I call for a unary operator U+232D that tests a value for its cylindricity.
21:30 sorear I try not to figure too much about y3k, there are so many unknown unknowns
21:30 sorear .u 232D
21:30 phenny U+232D CYLINDRICITY (⌭)
21:30 sorear box :|
21:33 [particle] joined #perl6
21:37 brrt left #perl6
21:40 * quietfanatic knows haskell, FTR.
21:40 quietfanatic for the folks who've been considering reviving pugs
21:42 * sorear also knows haskell.
21:43 spider-mario so do I, a little
21:43 rjbs What's the purpose of \N{SQUARE GAL}?
21:43 * geekosaur knows some haskell as well
21:43 sorear .u SQUARE GAL
21:43 phenny U+33FF SQUARE GAL (㏿)
21:44 sorear rjbs: compatibiltiy with CJK character sets and their fill-in-box-sim
21:44 rjbs Oh, I missed the ones above.
21:44 rjbs all the other SQUARE xyz ones.
21:44 sorear we have 94*94 spots in our codepage, we've already encoded a few thousand kanji, let's fill in the rest with Latin, Greek, Cyrillic, and whatever else comes to mind
22:06 sorear n: also does Bool:D;
22:06 p6eval niecza v18-7-ga660099: OUTPUT«Unhandled exception: Unable to resolve method CURSOR in type Any␤  at /home/p6eval/niecza/src/NieczaActions.pm6 line 3851 (NieczaActions.trait_mod:does @ 6) ␤  at /home/p6eval/niecza/src/STD.pm6 line 1941 (P6.trait_mod:does @ 7) ␤  at /home/p6eval/niecza/sr…
22:06 sorear I thought that looked wrong
22:08 Telgalizer joined #perl6
22:09 [Coke] if you want to hack on pugs, I can offer some suggestions on what to look at first.
22:15 rjbs "this image from comp.basilisk"
22:22 whiteknight joined #perl6
22:23 leprevost joined #perl6
22:27 skids joined #perl6
22:36 masak good night, #perl6
22:36 sorear gppd night, masak
22:52 alvis` joined #perl6
23:02 dorlamm joined #perl6
23:17 harmil Am I correct that statement-commit (single colon) in a regex has two functions: collapse all extant alternations that currently match to this alternation; prevent future backtracking. Or are those the same thing?
23:21 harmil What I want to say is "rule term { <sub-thing>: | <parenthetical> | <literal> } ; rule sub-thing { '(' ~ ')' [ <?before 'thing'> <thing> ] } ; rule parenthetical { '(' ~ ')' <expr> }" but I'm not sure if that's going to work…
23:27 TimToady no, : only prevents backtracking into teh <sub-thing>.  you want :: to commit to <sub-thing>'s LTM match, or ::> to commit to current alternative (not sure if this is implemented)
23:28 TimToady or ::: commits the current rule to this path
23:31 tyatpi joined #perl6
23:32 TimToady phở &
23:44 harmil Ok, thanks TimToady
23:45 harmil I have to say, :$x parameter passing and hash construction is so amazingly nice. I find myself doing this all the time: sub foo($a, $b, $c) { bar(:z<1>, {:$a, :$b, :$c}) }
23:46 harmil Even in Perl 5, that's much, much more cumbersome. In Python it's downright annoying.
23:46 am0c joined #perl6

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

Perl 6 | Reference Documentation | Rakudo