»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg camelia perl6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by sorear on 25 June 2013.
diakopter lue: what step is next? :) 00:02
lue build & testing. Takes a while :) 00:03
00:08 dmol left, ajr_ left 00:09 raiph left 00:18 Entonian joined 00:19 Entonian left 00:22 kaare_ joined 00:28 raiph joined 00:31 jnap left 00:36 andrers52 left, Ben_Goldberg joined, andrers52 joined 00:37 BenGoldberg left 00:39 BenGoldberg joined, kaare__ joined, zakharyas left 00:42 Ben_Goldberg left 00:44 kaare_ left 00:46 treehug88 joined 00:52 treehug88 left, logie left 00:58 andrers52 left 01:02 jnap joined 01:06 jnap left
lue can anyone look this over and tell me if any of these errors need to be dealt with? gist.github.com/lue/422c00aa994f82dd22b7 01:07
01:07 jnap joined
lue (the filestat.t failure is because I have a noatime mount) 01:07
Looks fine to me, at least. 01:11
O.o perl6-lwp-simple fails magnificently 01:14
01:17 BenGoldberg left
diakopter lue: are you sure you had libicu installed? 01:18
lue jsonrpc fails too. It's thanks to a "missing or wrong version of dependency" of MIME/Base64/PIR.pm6. Not sure why.
diakopter: updated gist with modules-test failures 01:19
✘ ⑩ modules-test failed 01:20
01:21 BenGoldberg joined
lue & 01:22
01:48 raiph left 02:02 beastd left 02:03 census joined, pera left 02:12 thou left 02:49 kivutar left
lue hello again o/ 03:04
So, no-one yet can point me in a direction to this failure? :/ 03:06
oh, huh. Looks like `make install` fails on precompiling the offending modules. I just didn't catch that earlier it seems. 03:10
(LWP::Simple has the same "missing or wrong version" error as the tests, while JSON::RPC needs LWP precompiled to work.) 03:11
TimToady timotimo: in the normal Perl 6 RC page, the link to the unimplemented tasks is on the right in the Your Help Needed box 03:14
[Coke] lue: did make test work for the updated lwp-simple in teh R* release?
I mentioned that a few days ago in here, see you tripped over it too. 03:15
lue [Coke]: *updated* lwp-simple...? /me checks the logs for LWP 03:16
dalek kudo-star-daily: 08b25ec | coke++ | log/ (5 files):
today (automated commit)
rl6-roast-data: 609f1cf | coke++ | / (6 files):
today (automated commit)
[Coke] lue: i just mean latest.
lue last commit from 23 Nov
yeah, seems to be latest 03:17
[Coke]: I'll remember to check the R* runs before complaining next time, so I at least know more when I do :) 03:19
do you recall how you fixed it? 03:20
[Coke] I never fixed anything. I just reported the failure... and I see it's gone in today's daily run 03:21
03:21 kurahaupo_mobile joined
[Coke] last failed on Sat Dec 21 15:27:47 2013 03:21
had at least 2 days where it was working.
(including today) 03:22
lue *headdesk* I just _knew_ this was why updating the MIME::Base64 module was troublesome earlier. I just _knew_ this error was the reason behind that. And now I have proof :)
irclog.perlgeek.de/perl6/2013-11-23#i_7905580 <-- evidence, for the curious
[Coke] (when I asked if you fixed it, I didn't realize my daily run had cleared up.) 03:23
lue
.oO(the insidious part of the error is that it's the things *using* ::Base64, not the module itself, that fail. Which would make git blame and related strategies a bit more involved.)
03:24
...I'm gonna need to make 2013.12.1 tag for R*, aren't I? :)
oh, nvm, tagging doesn't happen yet. phew. 03:28
[Coke] you only need to make a .1 tag if you made a non .1 tag already 03:31
03:32 xenoterracide left
lue [Coke]: yeah, like I said, my brain got muddled for a second :) 03:35
03:43 xenoterracide joined 03:45 gfldex left 03:47 gfldex joined
lue is busy trying to find a command that will leave MIME::Base64 alone in the next release cycle 04:05
Hrm. The only solution I've seen so far is a linux-specific [ test ] for the foreach command :/ 04:09
04:12 colomon left 04:20 colomon joined, mtk joined, odoacre joined
dalek ar: 2a5ed8f | lue++ | modules/Perl6-MIME-Base64:
Revert Perl6::MIME::Base64 to working commit.

  moritz had fixed this last month, I just didn't know about it when I
updated this module.
04:26
ar: 2b96edf | lue++ | tools/star/release-guide.pod:
Update release guide to deal with Perl6::MIME::Base64

There's no easy way to modify the existing submodule command to skip Perl6::MIME::Base64, in a cross-platform way. So instead presented here is a way to fix Perl6::MIME::Base64 getting a merge commit.
The sooner we don't need this addition to the Release Guide, the better.
04:27 xinming_ joined 04:28 BenGoldberg left 04:29 xinming left, BenGoldberg joined
lue ✔ ⑨ generate tarball (again) 04:36
04:45 kaare__ left 04:46 jnap left 04:49 kurahaupo_mobile left 04:51 skids joined 04:54 preflex joined, ChanServ sets mode: +v preflex 05:07 logie joined 05:09 SamuraiJack joined 05:14 logie left 05:17 jnap joined, kaleem joined 05:21 jnap left 05:46 BenGoldberg left 05:48 rurban1 joined, rurban1 left 05:55 fridim_ left 05:57 rurban1 joined 06:05 raiph joined
moritz wishes everybody a good and merry UGT Christmas Even morning 06:08
06:20 logie joined 06:26 logie left 06:51 kaleem left, kaleem joined, kaleem left, kaleem joined 06:56 raiph left 06:57 kaleem left, kaleem joined, rurban1 left 07:28 xinming_ left 07:33 darutoko joined 07:37 rindolf joined 07:58 mberends joined 07:59 odoacre left 08:14 odoacre joined 08:45 xinming joined 08:47 xinming left 08:50 xinming joined 08:52 xenoterracide left 08:53 xenoterracide joined 08:54 dmol joined 08:55 xinming left 08:57 xinming joined 09:01 xinming left 09:03 spider-mario joined, xinming joined 09:08 xenoterracide left 09:10 brrt joined 09:11 Alina-malina joined, raiph joined 09:12 xinming left 09:27 mberends left 09:31 xinming joined 09:35 xinming left 09:41 mberends joined 09:51 xinming joined 10:12 raiph left 10:16 kivutar joined 10:31 xinming_ joined, xinming left 10:50 Alina-malina left 10:51 Alina-malina joined 10:53 denisboyun joined 11:02 finanalyst joined
finanalyst r: my @s=(0,1,0.5,0.4)[^3]; my @t=((0,1,0.5,0.4) xx 1)[^3]; @s.say;@t.say 11:04
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«0 1 0.5␤0 1 0.5 0.4␤»
finanalyst should both be the same I think
11:04 brrt left
finanalyst r: my @s=(0,1,0.5,0.4)[^3]; my @t=((0,1,0.5,0.4) xx 1)[^3]; say @s == @t 11:06
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«False␤»
11:15 odoacre left 11:19 odoacre joined 11:31 ivanshmakov joined 11:46 GlitchMr is now known as xfix 11:58 denis_boyun__ joined 11:59 denisboyun left 12:05 wooden left 12:07 brrt joined 12:08 kivutar left, wooden joined, wooden left, wooden joined 12:25 kurahaupo_mobile joined 12:26 kaleem left 12:30 kurahaupo_mobile left 12:31 wooden left 12:32 wooden joined 12:39 colomon left 12:40 colomon joined 12:54 xinming_ left, xinming joined 13:02 xinming left 13:03 xinming joined 13:07 tgt joined 13:08 xinming left 13:09 brrt left, xinming joined 13:12 atroxaper joined 13:21 xinming left, xinming joined 13:22 kivutar joined 13:23 dmol left 13:25 atroxaper left 13:26 atroxaper joined 13:27 dmol joined 13:30 atroxape_ joined, atroxaper left 13:31 araujo_ joined 13:34 colomon left 13:35 ajr joined, araujo_ left, ajr is now known as Guest70190, araujo joined, Guest70190 is now known as ajr_, colomon joined 13:38 gfldex left 13:44 cognominal left 13:48 sftp joined, sftp_ left 13:50 retupmoc1 joined, dbrock joined 13:55 sorear joined 13:56 __rnddim__ joined, orevdiabl joined 13:57 FOAD_ joined 13:58 ponbiki_ joined 14:02 ggoebel117 joined, pmichaud_ joined, Bucciarati_ joined 14:05 jnap joined, frettled_ joined 14:06 lue left, gfldex joined, FOAD_ is now known as FOAD 14:11 genehack joined, Celelaptop joined, zamolxes joined 14:12 ingy joined, Bucciarati_ left 14:14 lumimies joined 14:18 Bucciarati joined 14:19 kivutar left 14:25 kivutar joined 14:34 wooden left 14:36 rurban1 joined 14:39 dayangkun left 14:41 wooden joined, wooden left, wooden joined 14:42 kaleem joined 14:51 wooden left 14:52 wooden joined 15:25 rurban1 left 15:26 kaleem left 15:30 tgt left 15:35 ajr_ left, rurban1 joined 15:36 ajr joined, ajr is now known as Guest18972 15:37 Guest18972 is now known as ajr_, [Sno] left 15:42 colomon left 15:43 colomon joined 15:46 cognominal joined
raydiak good morning, #perl6 15:47
moritz good morning, raydiak 15:48
diakopter good morning, self
15:54 samebchase joined, filly joined 15:59 Psyche^ joined
raydiak cyberuniverses.com/pray/#pray-image-9 - hopefully the last pic before code release (which is what I said last time, but I couldn't help myself) 16:00
timotimo csg \o/ 16:03
16:03 dayangkun joined
moritz raydiak++ 16:04
timotimo is your nickname raydiak because you do raytracing?
moritz raydiak: if you release the code tomorrow, you can post the announcement as day 25 of the advent calendar :-)
raydiak timotimo: sorta, but it's more based on my real name..."Ray" is literally my middle name :) 16:05
diakopter diak is your last name?
raydiak moritz: I thought all the days were gone now...???...
diakopter opter is timotimo's middle name
moritz raydiak: well, advent calendars typically only go to 24, which is why only the days upto 24 are taken 16:06
raydiak: but there wouldn't be any harm in adding another day :-)
raydiak diakopter: ha ha :) it's too early for me to come up with anything clever to answer that with
moritz: oh right...well, that's a thought; I do have half a day of peace to myself 16:07
diakopter early in the human eon
raydiak that too 16:09
16:11 colomon left, zakalwe joined 16:18 dmol left, colomon joined 16:31 colomon left, atroxape_ left, atroxaper joined 16:32 colomon joined
timotimo diakopter: my first and last name are both timo 16:33
16:44 colomon left 16:50 finanalyst left 17:00 jnap left 17:06 spider-mario left 17:10 atroxaper left 17:20 colomon joined
colomon raydiak++ # !!!!!!! 17:21
raydiak thanks :) that's the impression I was going for 17:23
17:27 spider-mario joined 17:28 spider-mario left
colomon raydiak: I look forward to having the time to check it out more completely. busy desperately cleaning the house at the moment. :) 17:28
17:29 spider-mario joined 17:30 jnap joined
raydiak colomon: it looks better from a distance anyway...then you don't notice as much that it doesn't do internal reflection, subsurface scattering, or caustics yet 17:32
17:35 jnap left 17:36 ajr joined, ajr_ left 17:37 ajr is now known as Guest64029, Guest64029 is now known as ajr_
colomon raydiak: I'm sure those features can wait till next week. ;) 17:46
raydiak: what's your output format? 17:47
raydiak colomon: just ppm right now 17:48
timotimo are the times terrible? :) 17:56
17:58 Mouq joined, kivutar left
raydiak quite :) but I haven't made any attempt to profile or optimize yet, other than to keep it possible to work on in reasonable amounts of time 17:58
timotimo right 17:59
if you find something that should be much faster, feel free to ping me; i'm always interested in doing little optimizations
something that obviously should be faster ;)
as in: easy to prove at compile time :P 18:00
Mouq finalanyst: re: irclog.perlgeek.de/perl6/2013-12-24#i_8036731
r: my @s=(0,1,0.5,0.4)[^3]; my @t=((0,1,0.5,0.4) xx 1)[0][^3]; @s.say;@t.say
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«0 1 0.5␤0 1 0.5␤»
Mouq r: my @t=((0,1,0.5,0.4) xx 1); @t.perl.say 18:01
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«Array.new(0, 1, 0.5, 0.4)␤»
Mouq hm :?
r: my @t=(0,1,0.5,0.4); @t.perl.say
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«Array.new(0, 1, 0.5, 0.4)␤»
Mouq hrm 18:02
timotimo so the (0,1,0.5,0.4) xx 1 generates a LoL?
r: ((0, 1, 0.5, 0.4) xx 1).perl.say
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«((0, 1, 0.5, 0.4),).list␤»
timotimo yeah
r: ((0, 1, 0.5, 0.4) xx 2).perl.say
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«((0, 1, 0.5, 0.4), (0, 1, 0.5, 0.4)).list␤»
timotimo and you expect the [^3] to flatten the LoL before accessing? 18:03
kind of makes sense; it already says .list at the end
and the .list of that should be flat
raydiak timotimo: I'll be sure to let you know if I notice anything obvious when I start profiling (if japhb doesn't beat me to it...he mentioned a "barely contained obsession with optimizing 3D graphics code") 18:04
Mouq timotimo: I feel the way it is probably good for consistancy though, e.g., (0, 1, 0.5, 0.4) xx $i
18:04 kurahaupo_mobile joined
timotimo :D 18:04
Mouq: well, the thing is that the postcircumfix seems to disregard the list context it seems to be in 18:06
r: for (0, 1, 0.5, 0.4) xx 2 { say $_ }
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«0␤1␤0.5␤0.4␤0␤1␤0.5␤0.4␤»
timotimo in a for loop you have to .tree it in order to get two iterations out of that
r: for (0, 1, 0.5, 0.4) xx 1 { say $_ }
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«0␤1␤0.5␤0.4␤»
timotimo same here 18:07
18:07 Rotwang joined
timotimo it could be argued that it's a bug that postcircumfix:<[ ]> doesn't evaluate it in the list context like a for loop would 18:08
Mouq r: for (0, 1, 0.5, 0.4), (0, 1, 0.5, 0.4) { say $_ }
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«0␤1␤0.5␤0.4␤0␤1␤0.5␤0.4␤»
Mouq r: ((0, 1, 0.5, 0.4), (0, 1, 0.5, 0.4)).perl.say 18:09
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«((0, 1, 0.5, 0.4), (0, 1, 0.5, 0.4))␤»
timotimo interestingly, it doesn't output .list there
i'm not 100% sure how all that really works inside; and also pmichaud said he'd do a full revision of the list synopsis 18:10
colomon raydiak: you haven't posted the source yet? 18:11
18:15 kurahaupo_mobile left 18:21 SamuraiJack left 18:28 kurahaupo joined 18:31 jnap joined 18:36 rurban1 left, jnap left 18:41 darutoko left 18:44 SamuraiJack joined
raydiak colomon: no, but any day now 18:45
18:50 SamuraiJack left 18:51 __rnddim__ is now known as lue
lue hello world o/ 18:51
raydiak \o
18:53 census left
lue ✔ ⑩ Test tarball 18:53
✔ ⑪ tag 2013.12 release 18:57
18:57 tgt joined
lue could someone with the proper permission scp the tarball for R* to the rakudo.org location? The MD5 of the tarball on my end is c0b4f1b780182530c6714f01cf36d87b . 18:59
(posting the release announcement to rakudo.org is also something I can't do.) 19:00
Rotwang I know that I've asked it already, but have you considered making this process more automatic? 19:04
TimToady It's hard to automate the WTF bits. 19:05
lue Rotwang: Perl6::MIME::Base64 would've been *very* fun this month on automation :) 19:06
19:09 rurban1 joined 19:13 rurban1 left
Ulti if I wish to map a list to triplets from that list what is a perl6ism for this that returns the new list using something like for @list -> $a,$b,$c {} works but is there something like map that consumes more thna one value? 19:14
timotimo isn't for the same as map, basically? :P 19:15
i'm confused; why isn't eqat available in the stage0s? we recently rebuilt the stage0s :( 19:17
Ulti I dunno I just tried map {"$^a$^b$^c"}, @list and it didn't work out 19:18
timotimo i just mean: why not use that exact for?
Ulti because its a bit unwieldy 19:19
timotimo well, use the for just to make triplets then?
Mouq Ulti: the map you wrote works fine for me, latest perl 19:22
r: map {"$^a$^b$^c"}, (0, 1, 0.5, 0.4) xx 3
camelia ( no output )
Mouq r: say map {"$^a$^b$^c"}, (0, 1, 0.5, 0.4) xx 3
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«010.5 0.401 0.50.40 10.50.4␤»
Mouq r: say map ->$a, $b?, $c? {"$a$b$c"}, 0, 1, 0.5, 0.4
camelia rakudo-parrot d72797: OUTPUT«use of uninitialized value of type Mu in string context in block at /tmp/kbpOETfyA1:1␤␤use of uninitialized value of type Mu in string context in block at /tmp/kbpOETfyA1:1␤␤010.5 0.4␤»
..rakudo-jvm d72797: OUTPUT«use of uninitialized value of type Mu in string context␤use of uninitialized value of type Mu in string context␤010.5 0.4␤»
TimToady r: .say for map -> $a, $b, $c { "$a$b$c" }, 1..9
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«123␤456␤789␤»
TimToady r: .say for map *~*~*, 1..9 19:23
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«123␤456␤789␤»
Ulti weird
ahh ok its that I used @list X @list X @list without surrounding () 19:25
TimToady yes, that includes the {} in the first list :) 19:26
Ulti makes sense 19:27
TimToady r: .say for map *~*~*, list <a b c> X <a b c> X <a b c> 19:28
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«aaa␤aab␤aac␤aba␤abb␤abc␤aca␤acb␤acc␤baa␤bab␤bac␤bba␤bbb␤bbc␤bca␤bcb␤bcc␤caa␤cab␤cac␤cba␤cbb␤cbc␤cca␤ccb␤ccc␤»
TimToady anything that drops precedence will work
19:32 jnap joined
Ulti neat 19:32
TimToady r: .say for map *~*~* <== <a b c> X <a b c> X <a b c>
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«aaa␤aab␤aac␤aba␤abb␤abc␤aca␤acb␤acc␤baa␤bab␤bac␤bba␤bbb␤bbc␤bca␤bcb␤bcc␤caa␤cab␤cac␤cba␤cbb␤cbc␤cca␤ccb␤ccc␤»
TimToady though that might not continue to work if rakudo actually treats it as a statement sep 19:33
timotimo i'm glad this isn't a bug in perl6 or rakudo, and instead just works
TimToady r: <a b c> X <a b c> X <a b c> ==> map *~*~* ==> map *.say 19:34
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«aaa␤aab␤aac␤aba␤abb␤abc␤aca␤acb␤acc␤baa␤bab␤bac␤bba␤bbb␤bbc␤bca␤bcb␤bcc␤caa␤cab␤cac␤cba␤cbb␤cbc␤cca␤ccb␤ccc␤»
TimToady that should continue to work, I think
unless we require .map someday, which istr jnthn++ suggesting at one point 19:35
errands & 19:36
19:36 jnap left
Ulti btw this is what I'm working on gist.github.com/MattOates/8117240 19:54
not sure if that is gross or nice... I get what is going on and it's elegant but perhaps not simple 19:55
timotimo Ulti: [X] (@!bases xx 3) :D 19:58
19:58 ajr_ left
timotimo Ulti: i like your use of the whatever star in the aminos array, as the sequences that would generate them don't exist (is my guess anyway) 20:00
Ulti timotimo I dont think that does what I want..?
maybe it does
timotimo although it's not *actually* a whatever star :)
Ulti oh thats the symbol for a 'stop' codon
timotimo oh 20:01
ok
also, a list of one-digit strings should .Str the same way it would .join(""), no?
r: <foo bar baz>.Str.say
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«foo bar baz␤»
timotimo r: <foo bar baz>.gist.say
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«foo bar baz␤»
timotimo no, that's not it
r: my %foo = <a b c d e f g> Z=> (1, 2, 3, 4, 5, 6, 7); my $a = "c"; say %foo<<$a>> 20:04
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«3␤»
colomon raydiak: if you want someone to look at it before you show it to the general public…. ;)
timotimo Ulti: you can replace {" ... "} with << ... >>
Ulti k k
timotimo to completely confuse anyone ever :) 20:05
Ulti the [X] (@list xx 3) is a no go though since its not three lists but one list made from three appended by the looks of it
timotimo also i think your Str @!bases would look better with < ... > syntax as well
Ulti timotimo yeah I realise I could make some epic single line :) 20:06
timotimo no, i think that really would look nicer :)
Ulti I meant more the Z=> sorry
timotimo oh? no, i just needed a hash to test if <<>> will interoplate and turn into postcircumfix:<{ }> 20:07
and i thought Z=> was the shortest way
Ulti ohhh
timotimo in the same vein, i could have written Z=> 1..*, or even used <a b c d e f g>.kv
hm. i wonder if there's a reason to have .vk and .sriap, too 20:08
Ulti <<>> is blowing up 20:09
type Any in string context :S
timotimo oh?
gimme a sec.
20:10 rurban1 joined
timotimo r: my %foo; (given {%foo<<$^a$^b>>} { .WHAT, .arity }).say 20:11
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«(Block) 2␤»
Ulti yeah thats what I have... hmm 20:12
timotimo weird. can you update the gist with your exact code?
Ulti really it would be nice to get rid of the {} to have something as expressive as the previous map *~*~*
timotimo grondilu also suggested the [X~] $@!bases xx 3 20:13
Ulti timotimo yup one second
timotimo that's wicked concise
the $ prefix to turn the list into being item context is pretty slick
i should remember todo that more often
Ulti done 20:14
ahh so that gets you a reference to the list and then 3 of those
yeah thats nice though the plan wasnt to golf lol
I like it though
timotimo "reference to" isn't a good way to say it, IMO 20:15
oh hey
your join, map, comb looks just like a .trans 20:16
Ulti hmmm $@!bases throws up too 20:17
"Unsupported use of $@ variable as eval error; in Perl 6 please use $!" 20:18
timotimo whoops, that's a bug.
timotimo has a look at the grammar
Ulti $(@!bases) does work 20:19
20:19 grondilu joined
Ulti which I got via the usefule message: Unsupported use of ${@!bases}; in Perl 6 please use $(@!bases) 20:19
timotimo m) 20:20
Ulti so in Perl5 that would be scalar context... scalar context means the object representing a list type?
grondilu weird. $@ works fine here (rakudo star november)
Ulti This is perl6 version 2013.09 built on parrot 5.5.0 revision 0 20:21
grondilu r: my @a = rand xx 2; say $@a
camelia rakudo-parrot d72797: OUTPUT«0.249971135842742 0.298091508719278␤»
..rakudo-jvm d72797: OUTPUT«0.1677135985885968 0.2767544824499587␤»
timotimo interestingly the rule to find $@ looks for \W, '(' or a sigil
wow, ulti, that's old!
Ulti grondilu: might be the class variableness
timotimo: heh yeah :P
I am on my work laptop rather than my personal one (xmas)
timotimo it b0rks on my local version, too 20:22
Ulti so what is $(@list) doing?
because it feels different from what you might imagine from Perl5
timotimo it's the same as @list.item 20:23
afterwards, @list will not interpolate
r: say (1, 2, (3, 4, 5), 6, 7).list.perl; say (1, 2, $(3, 4, 5), 6, 7).list.perl
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«(1, 2, (3, 4, 5), 6, 7).list␤(1, 2, $(3, 4, 5), 6, 7).list␤»
grondilu also, you can do: return %codon_table{map *~*~*, $!seq.uc.comb}.join
timotimo oh, excuse me, .perl will still give you its structure
r: .say for (1, 2, (3, 4, 5), 6, 7).list; .say for (1, 2, $(3, 4, 5), 6, 7).list 20:24
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«1␤2␤3␤4␤5␤6␤7␤1␤2␤3 4 5␤6␤7␤»
timotimo you see how there's no newlines between the second 3, 4, 5?
that's the effect $... and ....item have on lists 20:25
Ulti okedoke but what $($string)
is it just meaningless
timotimo no change, $string is already in item context, because it is in a $-sigiled variable
grondilu r: my $x = rand; say $$$$$$$$$$$$x 20:26
camelia rakudo-jvm d72797: OUTPUT«0.4587225161057409␤»
..rakudo-parrot d72797: OUTPUT«0.276195594432892␤»
grondilu ;-)
timotimo that's also why we have \ sigil-less variables; these don't have an item/list context or anything like that
grondilu: oh shit! it behaves differently on our two backends!
time to throw one away!
Ulti hmmm so scalars feel refrency
timotimo i have no past with perl5 or earlier.
no history* 20:27
Ulti so is something like $list = (1,2,3) a list thats always in item context?
grondilu should remember to avoid using rand for value examples when calling r:
timotimo yeah
grondilu r: my $a = (1, 2, 3); .say for $a, 4, 5, 6 20:28
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«1 2 3␤4␤5␤6␤»
grondilu r: my $a = (1, 2, 3); .say for @$a, 4, 5, 6
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«1␤2␤3␤4␤5␤6␤»
timotimo anyone else think the evalbot should write ␤ instead of just ␤?
20:28 tgt left
Ulti doc.perl6.org/type/Str doesn't have .trans :P 20:29
Ulti reads the real docs
sergot Hi! o/ 20:32
Ulti hello 20:33
20:33 jnap joined
Ulti timotimo: yeah that would be a little easier to read 20:33
sergot Christmas gift from poznan.pm: act.yapc.eu/plpw2014/ ! :)
timotimo Ulti: it may be in Cool instead of Str
Ulti yeah I looked in Cool too 20:34
its in Str in the core shizzle in Rakudo
timotimo it's mentioned there, but there's nothing about it here
Ulti reading the source rather than the docs
timotimo mhm, yeah
it's in the synopses in any case. 20:35
Ulti I think maybe the docs.perl6.org hasn't been rebuilt in a while 20:36
or is it not coming from the source code?
also I cant believe I forgot about [X~] considering I used it to make some decks of cards only a few months ago 20:37
timotimo i do think it comes from the source code
20:38 jnap left
timotimo er, doesn't it come from the docs/ folder? 20:38
Ulti yeah I doubt its the actual source because there aren't any pod comments
or comments...
retupmoc1 re: MIME::Base64 and R* - the problem is that precompilation breaks somehow on the conditional require 20:39
timotimo that's right
the core setting is already painfully slow to compile
retupmoc1 without precompilation, everything works correctly
timotimo if we also parsed Pod6 inside it ... ouch
20:39 retupmoc1 is now known as retupmoca, SoemGyu joined
timotimo i know for a fact that there's still lots to be optimized in the pod6 parsing we do. 20:40
Ulti timotimo use a perl5 script to strip it all and mangle variable names to be short like JS "compilers"
timotimo well, to be honest ... i tried to optimize it and didn't get any wins
they are actually called "uglifiers", at least i heard that name used often enough.
lue retupmoc1: yeah, I read the commit on the precompile-fix branch :)
Ulti heh
retupmoca lue: I thought you probably had, but I wanted to make sure 20:41
Ulti or obfuscators
retupmoca since iirc moritz spent a lot of time figuring out what was going on last time
lue I can imagine figuring it out wasn't very fun for moritz++ 20:42
pmichaud, jnthn, masak, PerlJam, tadzik, moritz: if one of you could scp the 2013.12 R* tarball (md5sum c0b4f1b780182530c6714f01cf36d87b) and post the release announcment on rakudo.org, that'd be great :) 20:44
20:48 [Sno] joined, Mouq left 20:58 V_S_C joined
timotimo oh 21:00
$@ wins out over $(@!foobar) even though the latter is longer (this is LTM), because it matches earlier. 21:01
21:02 dayangkun left
timotimo i'll add a lookahead that makes sure we're not actually looking at $@!... before we throw out the error. 21:02
Ulti timotimo++
timotimo std: class foobar { has @!bases; method frob { say $@!bases } }; 21:03
Ulti also I am failing to "get" inheritance in P6 ;_;
camelia std 4731beb: OUTPUT«ok 00:01 125m␤»
timotimo looks at STD before changing rakudo
do you want to open a rakudobug so i can tag the test for it? 21:04
Ulti thought it would be easy to make a DNA or RNA sequence subclass where just the @!bases list was different :S
timotimo can do if you step me through opening one which I imagine starts with me getting an account to submit a ticket
timotimo @! is private
Ulti what so a subclass can't define it? 21:05
timotimo if you want to override the way @!bases behaves from a subclass, you have to go through an accessor that gets overrided
you can't define it?
Ulti yeah but the methods from the super class use the undefined one
timotimo you should just be unable to access the superclass' one, no?
from the subclass, that is 21:06
Ulti I also tried making it @. but it didnt change anything
timotimo it should; but you have to access them with @. as well
Ulti its more the methods are in the super class and I wanted the sublcasses to just define a different parameter
timotimo if you access them with @!, you'll end up accessing the private one of either the subclass or the superclass
Ulti yeah no thats what I did... I'll keep playing
timotimo can you update the gist so i have something concrete to look at?
Ulti oh I must have missed one before 21:07
just tried again ;]
its working now
timotimo \o/ 21:08
Ulti win
timotimo is always glad to help
oh, the last thing grondilu suggested is also nice 21:09
using a list indexing on the codon table
21:11 rurban1 left
timotimo hah 21:11
rakudo has <?before \W | '(' | <sigil> > and STD has <!before \w | '(' | <sigil> > 21:12
21:16 V_S_C left
timotimo Ulti: are you going to file the bug about $@! for me? 21:17
21:17 dayangkun joined
timotimo r: use Test; eval_lives_ok '$omg'; 21:19
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«not ok 1 - ␤# Error: Variable '$omg' is not declared␤»
moritz lue: where's the tarball?
timotimo fröhliche weihnachten, moritz :) 21:20
moritz timotimo: ebenso :-)
timotimo Ulti: i'll file the bug if you don't want to. 21:21
21:21 kaare__ joined
lue moritz: I gave you an md5sum so I didn't have to find a place to put it :P . (If I hadn't lost access to my feather account...) 21:21
moritz lue: since tarballs contain timestamps, I can't easily recreate a tarball with the same md5sum 21:22
lue oh. :/ 21:23
moritz: would you mind terribly a 12MiB email? 21:26
moritz lue: yes, I would
lue does not posses an off-hand way to store files, except perhaps Google Drive...
timotimo moritz: you could simply use the md5sum to bruteforce all timestamps 21:27
moritz lue: I'll rather recreate the file, and upload it for your inspection
lue moritz: sounds good.
21:27 mberends left
moritz lue: moritz.faui2k3.org/tmp/rakudo-star-....12.tar.gz 21:27
lue was almost tempted to post a Base64 text file gist :P 21:28
21:29 kurahaupo left
diakopter hee 21:29
timotimo hah
Ulti: i'm filing it now 21:31
21:33 raiph joined, jnap joined 21:35 kivutar joined
timotimo actually. i can just put a test into the spectests without further documentation 21:37
21:38 jnap left 21:39 kurahaupo_mobile joined
dalek ast: 0a9e8fb | (Timo Paulssen)++ | S32-exceptions/misc.t:
ensure $@!foobar doesn't get parsed as $@ by accident.
21:40
kudo/dont_carp_dollar_at_bang: 6d54dc8 | (Timo Paulssen)++ | src/Perl6/Grammar.nqp:
parse special_variable:<$@> just like STD does.
21:42
timotimo spectests that commit 21:43
Ulti head asplodes from trying to do a super super class that checks the validity of input strings on construction 21:45
timotimo sounds like you just want to compose in a role, honestly 21:46
lue moritz: I should be able to tell you if the tarball is good in a few minutes :) 21:47
timotimo (that's the result i got from shaking my magic 6-ball :)) )
Ulti the problem I have is that once I have my own constructor the magic that overrides goes all to hell because in BUILD I do a bind to $!something rather than $.something 21:49
if a subclass has some defined values that were undefined in the super class do these just get mapped as if they were variable => value passed in the call to new? 21:50
if I have a BUILD in the super class 21:51
moritz -> sleep, R* upload will have to wait for tomorrow
timotimo need to get my power brick plugged into this laptop now 21:52
21:54 denis_boyun__ left, BenGoldberg joined
timotimo i don't understand why your BioInfo::Sequence just invariably dies 21:55
diakopter Death does what Death wants
timotimo Ulti: i think you want something like method residues { <T C A G> } instead of a member 21:57
alternatively, your DNA/RNA sequences could do a role Sequence[Str @residues] 21:58
Ulti timotimo: testing is why it invariably dies 21:59
timotimo anyway, i think you want something different from a member variable 22:04
it's a property of the class rather than of any one instance
so either do the method or the role?
22:05 aindilis joined
dalek kudo/nom: 6d54dc8 | (Timo Paulssen)++ | src/Perl6/Grammar.nqp:
parse special_variable:<$@> just like STD does.
22:09
timotimo Ulti: commit passes all spectests including the new one i made for the bug you found 22:10
22:11 raiph left
timotimo Ulti: do i understand correctly that you want the sequence to be checked for wrong characters? like an X for example? 22:11
Ulti timotimo not sure I do want the role because some sequences wont be DNA ;) 22:12
or RNA and use residues which have a tonne of different operations on
timotimo hmm.
Ulti but the concept of a constrained alphabet common to them all is still true
timotimo in that case i'd go for a RestrictedToAlphabet role that has a @.alphabet or a method like that 22:13
now i'm thinking about what the smartest way to reject strings that have invalid characters in them is.
iirc, if you interpolate an array into a regex, it'll work like an LTM alteration
let me check that
r: my @alts = <foo bar baz>; "bar bar baz" ~~ m:s/ <@alts> +/ 22:14
camelia ( no output )
timotimo r: my @alts = <foo bar baz>; say "bar bar baz" ~~ m:s/ <@alts> +/
camelia rakudo-parrot d72797, rakudo-jvm d72797: OUTPUT«「bar bar baz」␤␤»
timotimo cool.
so you can just match for ^ <@.alphabet> $
Ulti yeah I already knew you can expand that list to be a character class I did it previously in a little grammar that parsed these sequences working out if it was DNA or RNA 22:16
well guessing
timotimo :)
for that i would have bag $foo.comb, tbh 22:17
Ulti well this was a grammar that parsed some of the file formats and it just had some rules that let it choose between the three depending on what matched 22:19
timotimo oh
well, in that case ... :)
Ulti so I'm trying to make some classes to build actual objects that are useful from the sequences parsed :) 22:20
lue preflex: tell moritz the only file to differ between the two tarballs is <tarball>/MANIFEST, and *I think* it's just that the sorting of files in them is different. I also don't spot any files that only appear in one tarball. 22:25
preflex Consider it noted.
22:25 gfldex left
timotimo lue: send the file through sort on both ends and diff the result 22:25
22:25 grondilu left
lue timotimo: I was just thinking of doing something like that. 22:25
timotimo unix, baby! :) 22:26
diakopter eunuchs, baby
timotimo a eunuch's baby?
that would appear, at first glance, to be unlikely
22:28 colomon left, xenoterracide joined
lue I'm relieved, diff <(sort FILE1) <(sort FILE2) gave me no differences :) 22:29
timotimo \o/
you should preflex moritz that right away :) 22:30
lue preflex: tell moritz: Just confirmed it's a simple case of the files being listed in a different order in the MANIFESTs. Note that this was *just* the top-level MANIFEST; the others were precisely the same between tarballs.
preflex no
lue preflex: tell moritz Just confirmed it's a simple case of the files being listed in a different order in the MANIFESTs. Note that this was *just* the top-level MANIFEST; the others were precisely the same between tarballs.
preflex Consider it noted.
lue preflex: tell preflex to not be so rude next time.
preflex what
timotimo :D 22:31
lue
.oO(My tarball-identicalness check falls apart if one file between two versions changes in such a way to not change the md5sum. Allow me to play with SHA1, just for fun.)
22:32
22:32 colomon joined
BenGoldberg lue, how often does one file change without changing it's md5sum? 22:33
timotimo especially "on accident" :) 22:34
lue BenGoldberg: rarely I would imagine. I just feel like playing with the script I made some more :D
BenGoldberg Well, that's ok then ;)
22:34 jnap joined
lue should possibly get around to writing that Perl 6 SHA-3 thing sometime. After all, there's no sha3sum on his computer... 22:35
timotimo :D
i remember seeing a presentation on how keccak works
pretty nice 22:36
lue timotimo: yeah, I really liked trying to figure out it works a while ago. Of course, I've since forgotten everything :)
22:39 jnap left
timotimo still taking suggestions for a topic i could hold a lightning talk on 22:45
lue timotimo: first thought that popped into my head was the time-bending properties of phasers. Would that fit in a lightning talk format? 22:47
22:47 cognominal left
timotimo only if i can hold the talk in the same time-bent format *and* show people it makes sense to do so 22:48
22:49 cognominal joined
lue :) 22:49
timotimo perl5 had phasers, too, right?
geekosaur sort of 22:50
lue idk
timotimo only very limited phasers?
lue
.oO("The introductory material is stored in the penultimate slide *click* *click* *click* *click*")
geekosaur phasers are a more principled version of what perl5 has
timotimo well, one gag i could do is have a slide open that says BEGIN { say "it is now {now}" }; say "it is now { BEGIN now }"; say "it is now {now}" 22:51
and then say only two of those strings during the presentation
22:52 kurahaupo joined 23:03 Ben_Goldberg joined 23:05 BenGoldberg left 23:06 tgt joined 23:07 xenoterracide left, BenGoldberg joined 23:10 Ben_Goldberg left 23:14 xenoterracide joined
timotimo i like the idea with the phasers more and more, but i think it'd give a huge "OMFG WHY WOULD ANYONE EVER WANT THIS IN A LANGUAGE" reaction 23:15
23:18 jnap joined
lue timotimo: now - BEGIN now 23:19
:)
timotimo i know of that idiom
but still ...
the idea would probably be having slides that have phasers and reading/acting them out in the way they would be run if it were code 23:20
lue oh! Now I remember what trick I was trying to think of (which lead me to suggest phasers): proto sub will-be-multi(|) { pre-stuff; {*} post-stuff; }
23:21 araujo left, araujo joined 23:22 kaare__ left
timotimo that should work, right? 23:22
lue timotimo: sounds like it could work 23:24
The KEEP/UNDO phasers seem like really interesting ones, by the way.
timotimo yes, they do
do you know python's "for ... else" construct?
and try ... except ... else?
23:24 tgt left
lue timotimo: It's been a looong time since I've used Py3 23:25
timotimo that's in py2, too
lue
.oO(Also good: CATCH is actually a phaser)
timotimo has been for a rather long time, i believe
23:38 finanalyst joined 23:53 colomon left 23:54 colomon joined 23:57 rindolf left