Perl 6 - the future is here, just unevenly distributed

IRC log for #plparrot, 2010-08-03

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

All times shown according to UTC.

Time Nick Message
01:34 dukeleto cxreg: pointy blocks seem to work, as least using them in a basic way
02:02 dukeleto http://news.perlfoundation.org/2010/08/2010q3-grant-proposal.html
05:34 justatheory joined #plparrot
15:11 davidfetter dukeleto, what remains in order to merge setof into master?
16:04 justatheory joined #plparrot
16:59 dukeleto davidfetter: making it actually work, there is just a test
16:59 davidfetter ah
16:59 davidfetter is there a return_next ?
18:00 * davidfetter nudges dukeleto
18:43 dukeleto davidfetter: what do you mean?
18:44 davidfetter so as i understand it, there's just one way to get SETOF: put it all in a ResizeablePMCArray and then return that
18:45 davidfetter other languages have a "batch" method as above, and a "continuous" method, usually called something like RETURN NEXT
18:45 davidfetter the "batch" method has sharp upper limits as to the size of the result set
18:46 * davidfetter using metaphors from chemistry here, and hopes they're not too opaque
18:53 dukeleto davidfetter: when you said "is there", i meant to ask, what do you mean? Does PL/Parrot support return_next ? No.
18:53 davidfetter ok
18:54 * davidfetter frantically assembling something for tonight's sf.pm meeting
18:54 davidfetter i want to describe the metes and bounds of the current system as accurately as possible
19:46 justatheory joined #plparrot
21:19 dukeleto davidfetter: for a list of all the featues we don't have, but plan to implement, the github issues is a good list
21:20 dukeleto davidfetter: you can say that there are branches for SPI and SETOF support currently being hacked on
21:20 dukeleto davidfetter: is there a public link describing your talk tonight?
21:20 davidfetter dukeleto, will do :)
21:21 davidfetter meetup.com has something about the meeting. i'll put any slides i make in the web site
21:46 dukeleto davidfetter: sounds good
21:46 davidfetter i'm planning a live demo, if i can get this up on my mac :)
21:47 dukeleto davidfetter: oh boy!
21:48 davidfetter cxreg, what've you been up to lately?
21:49 davidfetter cxreg, oh, and do you have a photo somewhere i can use?
21:49 * davidfetter likes to give credit where it's due :)
21:49 dukeleto davidfetter: i have a nice PL/Perl6 example for you to show off
21:50 davidfetter sweet!
21:52 * dukeleto getting it ready
21:54 davidfetter w00t!
21:57 dukeleto davidfetter: i just pushed some new tests, and updates to the PL/Perl6 page, they should be on pl.parrot.org in a few minutes
21:58 dukeleto davidfetter: it is the fibonacci example that uses named arguments via a pointy block
21:58 davidfetter :)
21:58 davidfetter so you didn't want to use @_.shift() ?
21:59 dukeleto davidfetter: that doesn't give anybody named arguments
22:00 davidfetter ah
22:00 davidfetter ok :)
22:00 dukeleto -> $a, $b { ... } is better than { my $a = @_.shift; my $b = @_.shift; ... } or { my ($a,$b) = @_; ... }
22:01 dukeleto it makes people feel like they are actually using Perl 6
22:01 davidfetter good poing
22:01 davidfetter point*
22:01 davidfetter hrm. if pl/perl6 supported it, what would the SETOF version look like?
22:02 davidfetter .oO(what does marsellus wallace look like?!?)
22:03 dukeleto davidfetter: SETOF is for returning aggregates
22:03 davidfetter um, no
22:03 davidfetter it's for returning >1 row
22:03 dukeleto davidfetter: that is what I meant, i guess
22:03 davidfetter so "the first $1 fibonacci numbers" would be one example of a SETOF
22:04 dukeleto davidfetter: sure. The PL/Perl 6 code would return a list, and then PL/Parrot would marshall that list into the appropriate PG datatype
22:05 dukeleto davidfetter: if you attempt to return a list from a PL/Perl 6 function now, I am pretty sure PL/Parrot blows up
22:05 davidfetter um, right. so what would the perl6 code look like?
22:05 dukeleto it might just be changing the [+] to a [\+], maybe
22:07 dukeleto actually, no
22:07 dukeleto just remove the [+]
22:08 dukeleto > my $limit = 10; say ~(1, 1, *+* ... $limit)
22:08 dukeleto 1 1 2 3 5 8
22:08 dukeleto the leading ~ is the stringification operator, which puts the spaces between the elements of a list
22:08 davidfetter perl6 -e 'say ~(1, 1, *+* ... 5)'
22:08 davidfetter 1 1 2 3 5
22:08 davidfetter hrm. interesting
22:09 dukeleto pl.parrot.org is updated now
22:10 davidfetter oh, sweet :)
22:10 dukeleto SETOF is hard to test. PL/Perl just sees if the operation succeeded, but it doesn't actually test that the SETOF was created correctly
22:10 dukeleto i don't think pgTAP has SETOF-aware functions
22:10 * dukeleto could be wrong
22:10 justatheory dukeleto is wrong
22:10 davidfetter ask justatheory. i'm pretty sure he knows ;)
22:10 * dukeleto is happy to be wrong
22:10 dukeleto justatheory: sweet! i can test functions that do SETOF stuff?
22:11 justatheory dukeleto: See results_eq(), set_eq() and bag_eq(). http://pgtap.org/documentation.html#Pursuing+Your+Query
22:11 justatheory yes
22:11 dukeleto justatheory++
22:11 justatheory dukeleto: ’Course, you might have removed them from your copy.
22:12 * davidfetter <3 bag_eq
22:12 davidfetter makes it way easier to have equality without those pesky ORDER BYs
22:19 dukeleto the pgTAP in PL/Parrot has set_eq and bag_eq, sweet.
22:19 dukeleto justatheory: i only removed schema-verifying stuff
22:22 justatheory dukeleto: Cool.
22:30 davidfetter hrmf
22:31 davidfetter During configuration the following steps failed:
22:31 davidfetter 53:  auto::icu
22:31 davidfetter You should diagnose and fix these errors before calling 'make'
22:31 davidfetter that's on osx 10.6.latest
22:31 davidfetter trying to make parrot
22:31 davidfetter 55fab15af95b9f76f095cae90f147f53f5604136
22:32 dukeleto hmmm.
22:32 dukeleto does stuff still work?
22:33 * davidfetter waits while make churns
22:43 davidfetter hrm. xcode update
22:43 * davidfetter tries again
22:57 davidfetter what's the easiest way to get parrot/rakudo* for osx?
23:01 dukeleto doesn't R* have a downloadable thingy?
23:01 * davidfetter checks
23:01 dukeleto has anybody actually tried PL/Parrot with Rakudo* ?
23:02 davidfetter i'm about to ;)
23:02 dukeleto davidfetter++
23:16 davidfetter um, so, how badly will i break things if i don't include ICU?
23:21 davidfetter oh, and is there something special i need to do, other than adjust $PATH, in order to use a parrot at some wack location?
23:32 justatheory joined #plparrot
23:49 cxreg davidfetter: here's me in front of a null pointer: http://cxreg.com/cxreg.jpg
23:49 davidfetter *envy*
23:49 davidfetter you've got one of the shirts with a 9 on it
23:50 cxreg yeah, I didnt actually make it to oscon, but an orker brought me one
23:50 * davidfetter didn't make it to oscon either :(

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