»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend!
Set by sorear on 4 February 2011.
00:05 tokuhiro_ joined 00:09 atrodo left 00:11 leprevost joined 00:16 supernovus left 00:28 jerome left 00:29 Circlepuller joined 00:30 Circlepuller_ left 00:42 jerome joined 00:43 Targen left
dalek rl6-roast-data: 64a98b7 | coke++ | p (2 files):
today (automated commit)
 00:59
perl6-roast-data: f50d826 | coke++ | / (4 files):
perl6-roast-data: today (automated commit)
00:59 dalek left, dalek joined, ChanServ sets mode: +v dalek 01:02 tokuhiro_ left 01:07 aramisf left 01:11 japhb_ left 01:14 aramisf joined 01:15 aramisf is now known as Guest20650 01:20 FROGGS_ joined 01:22 flightrecorder left 01:24 FROGGS left, flightrecorder joined
[Coke] huh. there were 9 of those, and dalek only showed 2. 01:25
diakopter poor guy got booted 01:27
01:28 mjreed joined 01:33 macae left
[Coke] ugh. parrot-nqp used to autoconvert based on HLL mappings. looks like I lost that using nqp 01:36
01:36 Targen joined 01:37 mjreed left 02:02 isBEKaml joined 02:04 buubot_backup left
diakopter is this valid C code: if (1); 02:06
sorear yes 02:07
diakopter cool 02:08
I have an opportunity to use it
sorear that is a valid C statement. it does nothing.
you can shorten it to just ;
diakopter uhm.
02:08 buubot_backup joined
diakopter surely you don't think I was asking about the case where 1 is there in particular 02:08
sorear actually I did 02:09
diakopter I would have thought it was pretty clear I was asking about the lack of statement after the condition
sorear there is a statement after the condition
; is legal anywhere C expects a statement
diakopter I... .... am *surprised* ... you thought I was asking whether 1 was a valid condition. what kind of moron do you take me for? 02:10
geekosaur huh?\ 02:11
sorear I'm just trying to answer the questions without regard to who's asking
shachaf took the question the same way. 02:12
diakopter ok. thanks for the info that ; is always a valid statement. I knew it was ok after while () but I wasn't sure about if.
02:13 buubot_backup left
sorear actually I thought you were asking whether "if (1);", as a whole, was valid 02:13
diakopter yes, but that would be an insane question because it's obviously a no-op 02:14
I mean, who would ask how to write a no-op?
sorear I think it's best not to second-guess people's questions.
diakopter I don't see any "second-guessing" there at all. 02:15
I see deduction from assumption of sanity and pragmatism 02:16
shachaf wonders what context makes "if (...);" useful. 02:17
Anyway, "sanity" is a pretty vague thing, and figuring out edge cases of C can be tricky enough when you assume that people say what they mean. :-) 02:18
diakopter if (a || b || c || d) { return; } at the end of a void function
shachaf Does (a || b || c || d) have side effects?
diakopter nope
shachaf Then what's the point?
diakopter I was just curious if I could shorten the { return; } to ; 02:19
02:19 PacoAir left, PacoAir_ joined
colomon www.johndcook.com/blog/2012/10/01/d...leverness/ # perl 6 disparagement in the comments... 02:19
shachaf You can, but you can also shorten the whole thing to either nothing or "(a || b || c || d)".
diakopter ah 02:21
sorry I'm wrong; yes they do have side effects
they are each CAS operations
shachaf OK, then you still don't need the if. 02:22
diakopter if another thread is competing to do the exact same process, the one that fails first can fail fast
s/fails first/fails to CAS first/ 02:23
really, only the first one needs checked actually
02:35 macae joined 02:37 benabik left
diakopter colomon: do you disagree with the comment? 02:38
colomon diakopter: very definitely 02:39
diakopter can you support your opinion?
02:41 benabik joined 02:46 orafu left 02:49 orafu joined 02:59 Pleiades` joined 03:18 adu joined 03:44 sivoais left 03:46 Tanktalus left 03:48 sivoais joined 03:49 isBEKaml left 03:56 jerome left 03:58 macae left 03:59 Tanktalus joined 04:10 isBEKaml joined, jerome joined 04:25 telex left 04:27 bro_grammer left 04:30 leprevost left 04:31 telex joined 04:36 fgomez left 04:42 skids left 04:44 fgomez joined 04:49 cognominal left 04:55 cognominal joined 04:56 _jaldhar_ left 05:00 PacoAir joined, PacoAir_ left 05:02 adj13 joined
adj13 perl6: say 3 05:03
p6eval rakudo 3d31af, niecza v22-6-g9e5350d: OUTPUT«3␤»
adj13 rakudo
05:06 thou left 05:07 adu left 05:11 salala joined, adj13 left 05:13 salala left 05:15 aharoni joined 05:20 tomaw left 05:21 birdwindupbird joined 05:25 cognominal left
moritz good morning 05:30
sorear good morning moritz.
05:31 Chillance left
isBEKaml moritz: good morning 05:33
moritz \o sorear, isBEKaml 05:35
05:39 sivoais left 05:40 sivoais joined 05:44 cognominal joined 05:46 Kim_Breitwieser joined 05:51 isBEKaml left 05:53 aloha left 06:08 aloha joined 06:20 flightrecorder left
mathw o/ 06:20
06:20 FROGGS_ left
mathw was up late launching a bargeful of errors waiting to happen and would now very much like another job for two days of the week 06:21
sjohnson o/ 06:24
sorear "another job"?
mathw well I'd like to swap what I currently do for two days a week for something else
sorear do you currently have multiple jobs? would you, after the swap? 06:25
mathw I currently have two part-time jobs which fill the week between them. I very much like one of them (which used to be full-time, and I wish it still was), but the other one causes me continual grief 06:26
sorear ah, I see 06:28
06:29 isBEKaml joined
mathw ever done a schema migration on a database designed by someone who doesn't think referential integrity is important? 06:30
sjohnson mathw.C++ 06:31
06:31 isBEKaml|2 joined
mathw Wouldn't mind doing some C++ 06:31
:)
06:32 isBEKaml|2 left, isBEKaml|2 joined 06:33 isBEKaml|2 left, isBEKaml|2 joined, isBEKaml|2 left 06:34 isBEKaml left
sjohnson mathw: what's the main source of grief? lack of job satisfaction? boring tasks? annoying coworkers? 06:34
mathw it's a very small company, so it basically boils down to one coworker who just doesn't think the same way I do 06:35
almost always takes the shortest path to a solution
sjohnson i hear ya 06:36
mathw and is very fond of giving me a task then implying it's my fault when I don't manage to include this obscure bit of it that he never told me about
he's never had to produce proper specs before, because he's never had to communicate them to another programmer
not that he wants to be a programmer anyway
the best bit about this job is that I got to learn Dancer and write some serious Perl 5 06:37
sjohnson well, you're not alone
mathw yeah it's just hard to take because I had an idyllic job with a terrifyingly intelligent colleague who thinks a lot like me (but different enough for us to reinforce each other's weak points) but the company's been having trouble and they couldn't keep me on full time anymore
sjohnson :[ 06:40
mathw so that job's more annoying now because we can never deliver in the time frames they got used to
becaues we don't have as much time as we did!
even my productivity-per-day has gone down, because there's a context-switching penalty involved in this 06:41
plus more time (proportionally) spent dealing with user support
sjohnson sounds like a recipe for woe 06:44
mathw yeppers
am currently entertaining approaches from a recruiter who contacted me
sjohnson it is kind of amazing why something so simple would be so hard for managers to understand
mathw but he hasn't got back to me after passing my CV on to a company yet
I don't want to leave the good job, but if I have to... well, I have to 06:45
moritz hard choices are hard
mathw yes 06:46
I've considered freelancing, but I have no idea how to get work or what to really do.
06:50 wtw joined
mathw and I'm not sure I ahve the patience for the admin, so I'd rather just get a job :) 06:52
06:52 buubot_backup joined 06:56 buubot_backup left 07:00 atrodo joined 07:02 grondilu joined
grondilu I've recompiled rakudo and now when I try to import a module (previously imported with Panda), I get: 07:03
===SORRY!===
Missing or wrong version of dependency 'src/gen/CORE.setting'
07:03 nebuchadnezzar joined
sorear afaik that's expected 07:04
grondilu do I have to reinstall panda or something? 07:05
07:06 atrodo left
sorear you have to tell rakudo to recompile the modules. 07:06
I'm not sure how best to do that, but just using panda to reinstall the modules should be enough... 07:07
alternately, you can delete the precompiled modules
which should be .pir files in ~/.perl6 or something like that
but IANA rakudo expert.
grondilu tries removing *.pir files 07:08
sorear IANA rakudo user, for that matter, so take everything I say carefully 07:10
grondilu well, now I can run panda at least. It takes a long time though, since nothing is precompile anymore. 07:11
07:12 cognominal left
grondilu runs "panda install panda" (which is quite odd) 07:14
mathw :) 07:15
Programmers often deal with absurdities. It's part of the fun.
grondilu it failed. It wants an already precompiled JSON::Tiny 07:16
07:20 cognominal joined
moritz panda has a 'rebootstrap.pl' command 07:21
which you have to run (from its source directory) whenever you install a new rakudo version
and you want to keep your old modules 07:22
grondilu moritz++: thanks 07:23
07:23 sqirrel joined
sorear o/ sqirrel 07:24
aharoni Hallo. How can I check that I have the right version of ICU for building Rakudo? (I'm running Ubuntu.) 07:27
sorear aharoni: I guess you could start the "Configure.pl" script, stop it when it starts building Parrot, then check to see what Parrot's configure thinks about your ICU 07:28
aharoni sorear, I already ran `perl Configure.pl --gen-parrot --gen-nqp' once, and it cloned everything, so if I run it again, it finishes very quickly and doesn't say anything about what it did or didn't find. Even after I did `make realclean'. 07:30
Or did you mean that I should run `make' and see what it says about ICU?
sorear aharoni: maybe try grep has_icu parrot/config_lib.pir 07:31
aharoni set $P0["has_icu"], "0" 07:32
does "0" mean that I don't have it?
moritz correct
--gen-parrot doesn't rebuild if there's a parrot already there
what you can do is 07:33
rm -rf install/
and then run perl Configure.pl --gen-parrot again
aharoni aha, "Is ICU installed....................................yes" 07:35
maybe it's because I installed libicu_dev after running Configure for the first time. 07:36
But I didn't know that I should delete install/ before re-building. I assumed that `make realclean' would take care of it.
07:37 Patterner left
sorear these days the only "clean" I trust is "git clean" 07:37
07:39 Psyche^ joined, Psyche^ is now known as Patterner
moritz aharoni: well, 'make realclean' only cleans rakudo's files 07:40
07:53 cognominal left 07:59 grondilu left, cognominal joined
aharoni When I didn't have ICU installed properly, S02-types/version.t tests with Greek letters failed. After installing libicu_dev and re-building, they pass. 08:02
TimToady back to Tokyo from Nikko (along with jnthn & ingy)
aharoni Should I maybe add a comment about this to S02-types/version.t?
sorear aharoni: you should modify t/spectest.data and add the NEEDS_ICU marker or however that's spelled to the S02-types/version.t line 08:03
there are a number of test files that Rakudo autoskips if ICU is not available
aharoni wonders what other tests it auto-skips. 08:04
moritz grep icu t/spectest.data
aharoni I would actually love it not to skip Unicode-related tests, given that I care about foreign alphabets a lot. 08:05
08:06 fhelmberger joined
jnthn afty, #pelr6 08:06
oops, forgot to type...
aharoni sorear, moritz, i huess that the marker is `# icu'
a comment in the end of the line.
sorear aharoni: why test something that is known a priori to fail?
jnthn, aharoni needs to know the history of ICU in Parrot 08:07
also, goof afternood
aharoni sorear, If you don't care about Unicode and don't want to install ICU, then it's fine to skip it. But without that failure that I had I wouldn't even notice that I don't have ICU installed properly. 08:08
(Although I would probably notice it after some time.)
Is there maybe some make target than runs everything without skipping? 08:09
jnthn History? Parrot used to bundle ICU within it and I spent ages getting the Windows build of it to work and if you had a Parrot you had an ICU, iirc. They they tossed it out of the repo and now some people have ICU and some done. 08:10
I've no idea how to do it these days.
s/done/don't
aharoni On Ubuntu it's just a matter of installing libicu_dev.
jnthn Yeah, I'm on Windows.
aharoni Oh. 08:11
jnthn I don't see ICU as our long-term or even medium-term solution anyways.
aharoni Anyway, how can I submit a patch that marks these tests as ICU-dependent? rakudo.org/how-to-help/ says to ask on IRC :) 08:12
jnthn But yeah, it's needed for doing Unicode stuff on Rakudo today
aharoni: Add the # icu markers. 08:13
sorear aharoni: make a patch, then either post it on irc using a pastebin, or via a github pull request, or a github ticket with patch, or whatever else suits you
jnthn aharoni: Then create a patch and send it to [email@hidden.address]
aharoni Oh, GitHub, of course. I just cloned and didn't even look at github.com :) 08:14
OK.
sorear note that rakudo has an unusualy (by #perl6 standards) restrictive commitbit policy
aharoni No problem. 08:15
08:15 buubot_backup joined
sorear for most other repos we'd give you commit access right now and tell you to fix it yourself 08:15
08:15 kresike joined
sorear but rakudo does not give commit bits without a signed committer license agreement 08:15
kresike good morning all you happy perl6 people
sorear hmm, I should do that someday soon
o/ kresike
kresike hello sorear o/
tadzik hello 08:19
jnthn Wow, no Rakudo or NQP commits while I was gone in Nikko
o/ tadzik
sorear o/ tadzik 08:21
tadzik jnthn: do you plan to look at the threads issue? 08:22
08:22 Kim_Breitwieser left
tadzik nqp vs parrot-nqp inconsistency 08:22
jnthn tadzik: What threads issue? 08:23
aharoni Mmm... `make spectest' clones from github.com/perl6/roast . Can I tell it to clone from my fork in some nice way?
moritz aharoni: you can simply clone yourself into t/spec/ 08:25
aharoni: then rakudo will just 'git pull'
aharoni: but of course I can also just give you commit access to perl6/roast, and then you can use the official one :-)
aharoni: what's your github ID?
jnthn tadzik: I thought the threads thing in Parrot got backed out?
tadzik jnthn: gist.github.com/3817314
aharoni moritz, amire80
tadzik jnthn: backed out by whom 08:26
aharoni moritz, it's very nice of you :)
jnthn tadzik: I thought I saw a post saying that the threads merge wasn't meant to have happened.
tadzik huh, don't recall that
08:26 tomaw_ joined
tadzik otoh, I'm not on parrot-dev anymore, I think 08:26
aharoni maybe I can just pull from perl6 and push to my fork.
moritz aharoni: welcome. Have the appropriate amount of fun
tadzik dukeleto was asking me about blockers recently
aharoni moritz, thank you.
jnthn tadzik: Was the lexpad issue fixed?
tadzik: nqp, unlike parrot-nqp, does sub lookups lexically, consistently. 08:27
tadzik jnthn: yes, see gist
it works in parrot-nqp
jnthn Even for a custom lexpad? 08:28
tadzik well, I don't know :)
it fails in perl6/nqp
jnthn Hm 08:29
I'm really not sure from the backtrace, aside from it seems to be trying to construct an exception for some reason
I'm not sure why, but the root issue may be that something causes an exception
08:32 kurahaupo left 08:33 dakkar joined
sorear tadzik: there was a force push reverting the threads merge. there was a general message to parrot-dev telling people to recheckout. 08:34
tadzik I'll check out the ML 08:35
jnthn did that at uni... :) 08:37
08:41 FROGGS joined 08:42 cognominal left
sorear o/ FROGGS 08:42
aharoni Hehe, apparently there is a pull request about the Greek letters in version.t in GitHub already - github.com/rakudo/rakudo/pulls . 08:44
08:44 cognominal joined 08:46 flightrecorder joined 08:48 skids joined 08:49 buubot_backup left
FROGGS hi sorear 08:49
lestrrat jnthn: your talk video is up now yapcasia.org/2012/talk/show/645d66c...4e6aeab6a4 08:52
jnthn: do you have your slides up anywhere?
jnthn lestrrat: No, will fix that now
Thanks for the reminder
lestrrat :)
jnthn lestrrat: jnthn.net/papers/2012-yapcasia-modules.pdf 09:03
diakopter o.o
lestrrat jnthn: registered! thanks :) 09:04
09:06 MikeFair___ left 09:07 buubot_backup joined 09:08 tomaw_ is now known as tomaw
Su-Shee lestrrat: I _just_ read your yapc::asia posting. you made it look like an amazing event i'm very envious about. :) 09:09
hello everyone.
09:09 buubot_backup left
jnthn lestrrat: Awesome that the video is up so fast, BTW! 09:10
lestrrat++
lestrrat If you felt it that way, then I did my job right ;)
jnthn: please thank this guy -> twitter.com/941 09:11
I'm like the foreign relations / tech / finance guy. He does the rest
09:12 buubot_backup joined
lestrrat if you haven't seen me talk about him, please peruse slides 40-42 www.slideshare.net/lestrrat/running...na-2011/40 :) 09:13
09:17 kurahaupo joined
jnthn lestrrat: have done so :) 09:19
sorear it's funny. from what I've seen in the past, I thought lestrrat was generally opposed to the concept of videoing talks, yet yapc::asia has the promptest videos I've seen yet
lestrrat sorear: huh? what made you think that way?
TimToady sorear: you sure you aren't confusing lestrrat++ with someone else? 09:20
lestrrat I AM against ustream'ing, if you ask :)
but I don't think I've gone and talked about it much :) 09:21
sorear blogs.perl.org/users/yapceurope_201...ent-122259
lestrrat Ah, I must have been thinking ustreaming but typed otherwise, or something. 09:22
sorear what's ustreaming?
lestrrat people around here likes to www.ustream.tv/ their conferences (I don't know about europe or na) 09:23
i.e. stream it live
also, we lower the quality of our videos.
on purpose.
that's what I wanted to say in that comment, I think.
since I was asked "I think - without wanting to say anything bad or picky - that video could be even cooler, perhaps with respect to the readability of the slides. " 09:24
to which I was like, "well, yeah, I can give you some video, but not the real thing"
sorear now what I would *really* like in the conferences is some way to make the days longer so that you don't have to schedule interesting talks against each other 09:25
24 hours is never enough
lestrrat I'd hate that. the organizers would have to work longer hours ;) 09:26
sorear cannot speak for yapc::asia specifically but had many painful decisions to make at ::europe
lestrrat yeah, I purposely put interesting talks back-to-back. otherwise you get one room full, and the other room empty. that's bad for the non-rock-star speakers. demotivates them a lot. 09:27
TimToady at ::asia it largely comes down to whether you understand very rapid　日本語. :)
gets easier to choose if you don't :)
lestrrat the language barrier is one of the reasons why we are making the videos and slides available so quickly
most JP people probably didn't understand what (exactly) larry was saying at the venue, 09:28
sorear probably for the best I didn't go, then. my 日本語 listening comprehension skills are basically nonexistant
lestrrat but they can figure it out if they watch the videos carefully.
TimToady which is why I picked something that can largely be understood by reading the code on the screen 09:29
though if they can't see that from the video, it'd be a problem
sorear lestrrat: what goes into making the videos presentable? why do na and europe take months?
lestrrat sorear: Right, but you don't need Japanese or English to understand stuff like this: www.youtube.com/watch?v=3C7Ngq6bM4M :)
sorear: beats me. we just have an HD handycam on each room, and just download, cut, and upload. done. 09:30
I have no clue why it would take more than a week or two to do the above work.
sorear lestrrat: is that the reduced quality version? 09:31
lestrrat yep.
sorear I guess I need new glasses, because I can't see any unusal blurring :D 09:32
or rather, "it seems in focus to me :D"
lestrrat well let me just put it this way: if we used the original quality video, you would be able to see the scratches on the projector screen. it's full HD. 09:33
TimToady: I meant to ask you - since you did live coding, you have no slides, right?
just wanted to make sure.
sorear well I had better head off to sleep. 09:34
lestrrat++ 941++ # even if I didn't go I consider this a valuable community service
&
lestrrat thanks. g'night
mhasch lestrrat: "live coding" might still have been prepared in advance... :-) 09:35
TimToady lestrrat: in a sense I had 1377 slides
lestrrat meh :)
TimToady since it was a vim keystroke file
lestrrat oh, so you can actually repro the entire step ? 09:36
TimToady but one can see teh before and after on rosettacode.org :)
in the Perl 5 and Perl 6 entries for strand sort
but yes, I can repro the whole thing
lestrrat I think a lot of the people there were wondering how you were typing with one hand :) let me know if you feel like posting that keystroke file somewhere 09:37
mathw I like this ocncept
(sorry, too tired to type properly this morning) 09:38
vim keystroke file as presentation tool. Very cool.
TimToady the technology has a few rough edges yet 09:40
you have to be very careful not to use screen-relative commands like H
mathw I imagine setting it up is quite tedious 09:41
TimToady well, depends on how safe you want it to be
every command saves and restores the current position in my file, so if I accidentally or purposefully move the cursor, it starts up again that the correct spot 09:42
but that clobbers your position if you do j or k across a short line, so I had to compensate for that too
and my popups didn't work quite right
not surprising since I hacked that capability in within the hour before the talk :) 09:43
Su-Shee can't you just make it a "filetype" and just disable the keystrokes for this foobar.mysuperpresentation filetype?
09:44 arnsholt_ is now known as arnsholt
TimToady also, inserts go in all as a single transaction, when you'd want to simulate typing one char at a time sometimes 09:44
I'm sure there are lots of capabilities within vim that I haven't explored; I was trying to do the simplest thing that would work, and that was a little more complicated than what I used in ::eu, which didn't quite work 09:45
mathw Something tells me that one day there'll be a Perl script which produces files vim can digest from a friendly specification
Su-Shee I love the idea though. :) I never warmed up with anything "real presentation" like, I usally made everything html.. 09:46
mhasch you'd also have to pay a price for more flexibility (shortcuts and elaborations)
Su-Shee mathw: vim has a perl api, you can just enable it or use vimscript.
TimToady what the popups did was start a gnome-term with zoom=3 or so, fullscreen
I'd hit a return to exit that, and then I was back in vim 09:47
Su-Shee I made a note "look into vim as presentation tool" for me. :)
TimToady wasn't trying to use embedded Perl
mathw Su-Shee: then I'd have to learn vimscript :)
Which I don't really object to, but it would be more fun to write something in Perl 6 or Haskell
Su-Shee mathw: www.ibm.com/developerworks/linux/li...index.html and there's a part 2, 3, 4 and 5 too. 09:48
TimToady my vim script is all of three lines
map <F12> :let feed = readfile(input("vs: ", "x", "file") . '.vs')
map <Insert> :call feedkeys((len(feed)?remove(feed,0):''),'t')
mhasch TheDamian mentioned a little project "vimpoint" back in '08 or so; dont' know whether he pursued it further. 09:49
TimToady map <Insert>:
that's supposed to be a ^V-space
oh, I had to write a program to take a keystroke file and split it into separate lines per command 09:50
I couldn't find any way to get vim to update the screen if I didn't come back to a user prompt on each command 09:51
mathw I will admit to an influence from my colleague, who needed to generate an invoice so wrote an invoice-generating program.
09:52 FROGGS left, flightrecorder left
mhasch TimToady: did ^L not work? 09:53
TimToady didn't seem to
I tried lotsa things
mathw huh 09:54
TimToady in the end I decided doing each command when I type space as a macro gave me more flexibility anyway
it certainly gave me the capability of recovering in ::eu when my cursor got off from where it belonged
mathw which is definitely worthwhile 09:56
aharoni I am very much a newbie, but I suspect that S32-io/IO-Socket-INET.pl creates the file server-ready-flag for testing and doesn't delete it. 10:00
So `git status' in the spec repo shows it as untracked.
10:02 skids left 10:03 wk left
moritz patches to change that would be appreciated 10:05
10:06 daxim joined 10:12 grondilu joined
grondilu Check out this: perl6 --target=pir -e 'sub infix:<m+>($x, $y) { $x + $y }' 10:13
===SORRY!===
Serialization Error: could not locate static code ref for closure 'infix:sym<m+>'
? 10:14
Am I missing something? 10:15
"perl6 --target=pir -e 'sub m($x, $y) { $x + $y }'" works just fine. 10:16
Please do not all anwser in the same time :/ 10:22
In case you wonder, I wanted to precompile one of my library to make it faster. But I got the above error. 10:24
10:27 cognominal left 10:28 cognominal joined
TimToady most of the experts are asleep or having Indian food :) 10:29
grondilu ok, I'll wait, then.
10:31 flightrecorder joined, FROGGS joined 10:37 _jaldhar_ joined
grondilu r: sub postfix:<%> { $^x / 100 }; say 14% + 1; say 13 % 10; 10:46
p6eval rakudo 3d31af: OUTPUT«1.14␤3␤»
grondilu cool :)
FROGGS ya, it is :o) 10:53
10:56 mucker joined 11:05 Su-Shee_ joined 11:07 Su-Shee left 11:08 Su-Shee_ is now known as Su-Shee
flussence re: irclog.perlgeek.de/perl6/2012-10-02#i_6031394 -- I have! :D 11:12
mathw flussence: fun, isn't it! 11:13
flussence worse than that, I had to *fix* it into a proper structure 11:14
mathw I *want* to 11:15
but I can't
flussence (which is real fun when the entire system you've been given is undocumented...)
mathw oooh
at least I've got the author of the system in question here to beat up
sorry, ask questions of
11:17 sisar joined 11:20 leont joined, SamuraiJack joined
aharoni moritz, I tried to fix that remaining test temp file issue: github.com/perl6/roast/pull/25 11:23
11:24 sisar left 11:30 Tanktalus left
moritz aharoni: nice. I'll probably won't get around to look at it today, but tomorrow your chances are very good 11:36
aharoni Thank you, moritz.
moritz (tomorrow is national holliday in .de)
ingy jnthn: hi 11:40
jnthn: can you pass me my beer please?
11:41 cognominal left
jnthn does so 11:41
11:41 cognominal joined
ingy drinkz 11:41
jnthn grondilu: Problem is that pre-compilation of things with custom operators doesn't work. 11:42
grondilu: It's already on my "stuff that really needs fixing" list.
11:43 gootle left 11:44 Tanktalus joined, gootle joined 11:48 simcop2387 left 11:50 gootle left 11:56 leprevost joined, simcop2387 joined 11:57 gootle joined
flussence I *really* *really* like .?() now. There's a thing I'm trying to do in $dayjob that'd be much easier if there was a short-circuiting -> op 12:00
12:01 mikemol left
grondilu jnthn: ok. 12:02
12:04 ireekennational joined
leont Ah, a maybe operator, nice 12:05
12:06 xinming joined
FROGGS flussence: what does it do? 12:10
.?() I mean
12:10 SamuraiJack left
Timbus r: Any.?this-fails(); say 'alive'; 12:13
p6eval rakudo 3d31af: OUTPUT«alive␤»
daxim doc.perl6.org/language/operators#postfix+.%3F
the point is that you can safely chain method calls 12:14
Timbus kinda
r: class A { method foo(Int $a){"got int"} }; say A.?foo('bar'); 12:15
p6eval rakudo 3d31af: OUTPUT«Nominal type check failed for parameter '$a'; expected Int but got Str instead␤ in method foo at /tmp/7WDloz6g6k:1␤ in block at /tmp/7WDloz6g6k:1␤␤»
Timbus r: class A { multi method foo(Int $a){"got int"} }; say A.?foo('bar');
p6eval rakudo 3d31af: OUTPUT«Cannot call 'foo'; none of these signatures match:␤:(A : Int $a, Mu *%_)␤␤ in method foo at src/gen/CORE.setting:367␤ in block at /tmp/DoVzMTjshe:1␤␤»
12:19 benabik left
jnthn .? is not about bindability, just "is there a method with the right name" 12:25
12:30 Tanktalus left
FROGGS so I can ask the object and it tells me (Bool) that the method exists or does it call the method? 12:32
jnthn Calls it if it exists
FROGGS daxims answer speaks for the latter
k
what does it do if the method doesnt exist? 12:33
is there a $! or something?
12:33 tokuhiro_ joined
moritz it returns Nil 12:34
FROGGS thanks 12:35
moritz if you want $! set, you do to try $obj.?method() instead
12:37 _jaldhar_ left 12:41 hoelzro|away is now known as hoelzro 12:44 Tanktalus joined 12:53 atrodo joined 12:56 flightrecorder left, FROGGS left
[Coke] wonders what diakopter's problem is. 12:59
12:59 hash_table joined 13:02 Guest20650 is now known as aramisf 13:03 PacoAir left
[Coke] (referring to the discussion that occurred after sorear answered his C question) 13:03
13:03 PacoAir joined 13:04 am0c joined, leprevost left 13:06 cognominal left
[Coke] to clarify: sorear was trying to be helpful, and IMO, diakopter reacted very (too?) defensively. 13:07
nqp precedence question: +@args[0].getList() 13:08
is the + modifying the result of the getList method? 13:09
13:09 flightrecorder joined
[Coke] jnthn: does nqp-latest support parrot HLL mappings for pmcs? 13:09
moritz [Coke]: yes, the + binds less tight than method calls 13:10
jnthn [Coke]: CompUnit lets you specify a hll
[Coke] jnthn: yes, but I'm not seeing args autoconverted.
jnthn Odd...you have custom PMCs? 13:11
[Coke] quite possible I lost something in the upgrade process so far, just wondering if I should expect that to work (especially for slurpy args getting autoboxed into my list type)
they are classes now.
jnthn How do you set up the mapping?
13:12 FROGGS joined, ireekennational left
[Coke] I'm not sure that code survived. I'm just trying to find out if it's expected to work before I try to chase it down. 13:12
moritz++ #thanks. 13:13
jnthn: I don't see any declarations to try to make that mapping work, no.
13:14 snearch joined
[Coke] I suspect I'm going to be better off not relying on the autoconversion working (and therefore having methods available on every object that I put there), and instead move those methods to subs that can work with various types. 13:15
(that .getList() for example, is also available on strings to convert them to a list - but not on builtin strings, and not on slurpy @args) 13:16
sorry, on TclStrings and TclLists 13:17
13:18 leprevost joined 13:22 kurahaupo left, leprevost left 13:26 grondilu left 13:30 cognominal joined 13:32 GlitchMr joined 13:34 tokuhiro_ left, hash_table left 13:58 bluescreen10 joined 14:01 flightrecorder left 14:03 FROGGS left 14:04 flightrecorder joined 14:11 snearch left
pmichaud good morning, #perl6 14:11
moritz good am, pm 14:12
14:12 stopbit joined
dalek kudo/nom: c1ddea8 | pmichaud++ | .gitignore:
Remove '*~' from .gitignore .
 14:12
14:12 sctt joined 14:16 FROGGS joined
jnthn hi, pmichaud 14:20
14:24 jaldhar joined, wtw left 14:30 kaare__ joined 14:32 hoelzro is now known as hoelzro|away
[Coke] misread that commit hash as my last name. 14:34
O_o
14:36 jaldhar left 14:37 sctt left, hash_table joined 14:38 buubot_backup left
[Coke] adds more caffeine to be safe. 14:38
14:40 Kim_Breitwieser joined
mathw \o/ 14:40
14:41 jaldhar joined
mathw contacts FedEx pls send caffeine stop urgent stop 14:41
mathw is in need of tea
14:41 kst left
jnthn [Coke]: It would be far more worrying if you misread a sha-1 hash as your first name... :) 14:42
14:43 cognominal left
jnthn sleep & 14:44
14:49 sirrobert joined 14:50 sctt joined
sirrobert r: "a b c 4 e" ~~ /$<num>=(\d+)/; say $<num>; say $num; 14:50
p6eval rakudo 3d31af: OUTPUT«===SORRY!===␤Variable $num is not declared␤at /tmp/OZlTDMIt_b:1␤»
sirrobert how do I refer to the var after capture?
mathw wonders what locale jnthn is in right now
14:50 cognominal joined
sirrobert r: say "a b c 4 e" ~~ /$<num>=(\d+)/; 14:51
mathw r: "a b c 4 e" ~~ /$<num>=(\d+)/; say $/<num>;
p6eval rakudo 3d31af: OUTPUT«｢4｣␤ num => ｢4｣␤␤»
rakudo 3d31af: OUTPUT«｢4｣␤␤»
FROGGS r: "a b c 4 e" ~~ /$<num>=(\d+)/; say $<num>;
sirrobert mathw: ah, thanks
p6eval rakudo 3d31af: OUTPUT«｢4｣␤␤»
FROGGS ya
mathw $/ is the default match object
sirrobert thanks, both
14:51 tokuhiro_ joined
mathw I was kind of surprised I remember that correctly actually :) 14:52
I'm not very current on the finer points of matches
I should write some more Perl 6 :)
sirrobert ahhh... I see my issue
the repl doesn't keep $/ from one command to the next.
FROGGS I had no chance to write something for the last weeks 􏿽xF3.􏿽xF2 14:53
mathw aaah
flussence r: my $m = "a b c 4 e".match(/$<num>=(\d+)/); say $m<num>; 14:54
p6eval rakudo 3d31af: OUTPUT«｢4｣␤␤»
sirrobert flussence: hm! I didn't know that 14:55
FROGGS feels a little bit like javascript
sirrobert I guess it's returning a Regex which uses a postcircumfix operator '<>' to access named captures?
mathw oh yes you can save match objects
r: my $m = "a b c 4 e".match(/$<num>=(\d+)/); say $m.WHAT;
p6eval rakudo 3d31af: OUTPUT«Match()␤»
mathw it's a Match object, which contains the results of a regex match operation, and it implements the necessary roles to let you access named captures in a hash-like way and numbered captures in a list-like way 14:57
sirrobert cool
14:57 thou joined
mathw so you can use postfix [] for numbered captures 14:57
sirrobert ahh, that's cool 14:58
mathw and {} and <> and «» for named ones
sirrobert and the special var $/ stores "last created Regex" ?
mathw that's the Match object from the last match operation
the Regex object just has the pattern in it, Match is where the action is once you start doing things 15:00
15:03 sctt left 15:04 atrodo left, atrodo joined, MayDaniel joined 15:05 birdwindupbird left 15:13 cognominal left 15:14 macae joined 15:20 benabik joined 15:31 skids joined 15:33 am0c left 15:35 benabik_ joined 15:37 benabik left, benabik_ is now known as benabik, flightrecorder left, FROGGS left 15:38 tokuhiro_ left 15:41 benabik_ joined
dalek rlito: 5488b30 | (Flavio S. Glock)++ | / (4 files):
Perlito5 - js - migrate array and hash accessors from "js3" to "js"
 15:43
15:45 benabik left, benabik_ is now known as benabik 15:48 Targen left
dalek rlito: 414f594 | (Flavio S. Glock)++ | README-perlito5-js:
Perlito5 - js3 - README - add milestones
 15:51
15:54 wk joined 16:01 cognominal joined
kresike bye folks 16:03
16:03 kresike left 16:13 wk left 16:18 daxim left 16:21 SamuraiJack joined 16:29 wk joined 16:31 macae2 joined 16:33 macae left 16:35 macae2 left 16:38 bluescreen10 left 16:40 kurahaupo joined 16:41 Chillance joined 16:44 fgomez left, macae joined, kurahaupo left 16:46 rummik- joined, _jaldhar joined 16:48 aloha left, rummik left, skids left, jaldhar left, skids joined 16:49 cognominal left 16:50 bluescreen10 joined 16:53 aloha joined 16:54 dakkar left 16:57 kurahaupo joined 17:01 leont left 17:15 kurahaupo left, tomaw_ joined
diakopter [Coke]: you're right; I do have a problem. I'm trying to get help. <- please don't take as an excuse 17:16
17:16 fhelmberger left
[Coke] hugme: hug diakopter 17:21
hugme hugs diakopter
[Coke] diakopter: 17:24
whoops. ENOMESSAGE, sorry about that. 17:40
17:41 mjreed joined
mjreed hello. What do we have in the way of introspection? How do I see what methods are defined on an object, what variables are defined, etc? 17:42
[Coke] r: say Int.^methods; 17:45
p6eval rakudo c1ddea: OUTPUT«Int Num Rat FatRat abs Bridge chr sqrt base expmod is-prime floor round ceiling sign conj rand sin asin cos acos tan atan atan2 sec asec cosec acosec cotan acotan sinh asinh cosh acosh tanh atanh sech asech cosech acosech cotanh acotanh unpolar cis Complex log exp …
[Coke] I don't know about variable introspections. 17:48
tadzik r: say Int.^attributes 17:51
p6eval rakudo c1ddea: OUTPUT«No such method 'gist' for invocant of type 'BOOTSTRAPATTR'␤ in method gist at src/gen/CORE.setting:4825␤ in sub say at src/gen/CORE.setting:7250␤ in block at /tmp/c76MFh_Rn8:1␤␤»
tadzik hehe, to clever
r: say DateTime.^attributes
p6eval rakudo c1ddea: OUTPUT«$!year $!month $!day $!hour $!minute $!second $!timezone &!formatter $!saved-offset␤»
17:52 buubot_backup joined
GlitchMr r: say Str.^attributes 17:53
p6eval rakudo c1ddea: OUTPUT«No such method 'gist' for invocant of type 'BOOTSTRAPATTR'␤ in method gist at src/gen/CORE.setting:4825␤ in sub say at src/gen/CORE.setting:7250␤ in block at /tmp/UA2KNGzukW:1␤␤»
GlitchMr r: say Set.^attributes
p6eval rakudo c1ddea: OUTPUT«%!elems␤»
mjreed Thanks! 17:57
17:57 mucker left 18:00 benabik left
[Coke] should attributes be telling us the priva attributes? 18:00
tadzik there is only one type of attributes :) 18:04
they can only have public accessors
18:06 leont joined 18:13 erkan left 18:14 erkan joined, erkan left, erkan joined, aharoni left 18:19 tokuhiro_ joined 18:21 _jaldhar left, kaleem joined, _jaldhar joined 18:24 Targen joined 18:25 kaleem left 18:29 fgomez joined 18:34 tokuhiro_ left
mjreed ok, another dumb question. Given declaration 'my Foo @foo', what's the best loopless way to initialize it to contain $n instances of Foo.new()? 18:35
GlitchMr mjreed: my Foo @foo = Foo.new xx 8 18:36
mjreed that won't get you 8 refs to the same Foo? Cool. Thanks.
GlitchMr Where 8 is your number of instances
...
oh, perhaps it will
oh, it doesn'
t
It doesn't
18:37 Targen left
GlitchMr mjreed: 18:38
> rand xx 8
0.26045819402589 0.758167870898859 0.671191730109182 0.588028023807915 0.819969979708805 0.181961273527108 0.60098118721211 0.477937755681602
It's feature :-)
rn: print join ' ', join xx 8
p6eval niecza v22-6-g9e5350d: OUTPUT«===SORRY!===␤␤Undeclared routine:␤ 'xx' used at line 1␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1437 (die @ 5) ␤ at /home/p6eval/niecza/src/STD.pm6 line 1147 (P6.comp_unit @ 37) ␤ at /home/p…
..rakudo c1ddea: OUTPUT«===SORRY!===␤CHECK FAILED:␤Undefined routine '&xx' called (line 1)␤»
GlitchMr rn: print join ' ', rand xx 8
p6eval rakudo c1ddea: OUTPUT«0.599440810809941 0.501246903880627 0.043752178355259 0.387303263970303 0.751670024253219 0.833972400039769 0.432674116824582 0.104679755674702» 18:39
..niecza v22-6-g9e5350d: OUTPUT«0.57264244257129837 0.13834876713266073 0.27021880693278222 0.91654168996798879 0.640887339897867 0.84707201963619894 0.87091507942924051 0.020385839985863231»
18:41 mjreed left
colomon rn: class Foo { my $counter = 0; has $.index = $counter++; }; my @foo = Foo.new xx 8; say @foo>>.index; 18:42
p6eval rakudo c1ddea, niecza v22-6-g9e5350d: OUTPUT«0 1 2 3 4 5 6 7␤»
GlitchMr infix:<xx>, expression repetition operator 18:43
I guess it explains the operator
18:43 Exodist left
GlitchMr Foo.new xx 2 is like map { Foo.new }, ^2 18:44
colomon yes 18:51
18:53 XphX joined 19:00 XphX left 19:01 Targen joined 19:05 aharoni joined
nyuszika7h ohai 19:14
tadzik hello nyuszika7h
19:14 leont left
nyuszika7h rn: print 'hi' 19:14
p6eval rakudo c1ddea, niecza v22-6-g9e5350d: OUTPUT«hi»
19:18 kst joined 19:26 benabik joined 19:30 FROGGS joined 19:31 flightrecorder joined
GlitchMr hi 19:33
rn: say 'hi' 19:34
p6eval rakudo c1ddea, niecza v22-6-g9e5350d: OUTPUT«hi␤»
sirrobert how can I insert breakpoints in my code for the perl6-debug to catch (rakudo)
?
I mean, not during runtime 19:35
GlitchMr I guess you cannot... yet 19:38
sirrobert ok, thanks =) 19:40
GlitchMr rn: Exception.new(:message<Hello, world!>).throw 19:45
p6eval niecza v22-6-g9e5350d: OUTPUT«===SORRY!===␤␤Undeclared name:␤ 'Exception' used at line 1␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1437 (die @ 5) ␤ at /home/p6eval/niecza/src/STD.pm6 line 1147 (P6.comp_unit @ 37) ␤ at /ho…
..rakudo c1ddea: OUTPUT«␤ in block at /tmp/kho4NU2HNG:1␤␤»
GlitchMr Uhmmm... how can I throw exceptions properly?
benabik rn: throw "oops"
p6eval niecza v22-6-g9e5350d: OUTPUT«===SORRY!===␤␤Undeclared routine:␤ 'throw' used at line 1␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1437 (die @ 5) ␤ at /home/p6eval/niecza/src/STD.pm6 line 1147 (P6.comp_unit @ 37) ␤ at /hom…
..rakudo c1ddea: OUTPUT«===SORRY!===␤CHECK FAILED:␤Undefined routine '&throw' called (line 1)␤»
benabik oops.
GlitchMr I know I can use die 19:46
But those exception types are to be used, aren't they?
Oh, I have to extend them. 19:47
ok
sirrobert dunno =)
19:55 rindolf joined 19:57 rindolf left, rindolf joined
flussence r: X::AdHoc.new(:payload<hai>).throw 19:57
p6eval rakudo c1ddea: OUTPUT«hai␤ in block at /tmp/oHnBmLfdHd:1␤␤»
GlitchMr Perl 6 debugger is awesome: gist.github.com/3822919 19:58
sirrobert So, I have a hash. When I try to "say %.hash.perl" the program pins the CPU at 100% (even in the debugger). How can I inspect it?
GlitchMr But it looks better in action, with colors
sirrobert: it's probably recursive hash then 19:59
Currently, Perl 6 freezes on recursive hashes
sirrobert What's a recursive hash?
a hash with itself as one of the elements?
one of the values?
GlitchMr > my $a = {}
().hash
> $a<a> = $a
yes
sirrobert huh, ok. I don't *think* I did that, but maybe so =)
thanks, GM 20:00
GlitchMr This hash contains $a<a><a><a><a><a><a><a><a>...
20:00 antigremlin joined, GlitchMr left 20:07 sqirrel left
aharoni In the documentation of Types at doc.perl6.org, there are many specifications that say something like "multi sub chr(Int:D ) returns Str:D" 20:27
What does the <:D> stand for? 20:28
moritz defined
ie Int:D allows 1, 2 or 3 but not the Int type object
aharoni I'm not sure that I understand the difference. 20:29
moritz r: sub f(Int $x) { say $x.perl }; f(3)
p6eval rakudo c1ddea: OUTPUT«3␤»
moritz r: sub f(Int $x) { say $x.perl }; f(Int)
p6eval rakudo c1ddea: OUTPUT«Int␤»
moritz r: sub f(Int:D $x) { say $x.perl }; f(Int) 20:30
p6eval rakudo c1ddea: OUTPUT«Parameter '$x' requires an instance, but a type object was passed␤ in sub f at /tmp/FbDX8HwLgJ:1␤ in block at /tmp/FbDX8HwLgJ:1␤␤»
moritz r: sub f(Int:D $x) { say $x.perl }; f(3)
p6eval rakudo c1ddea: OUTPUT«3␤»
sirrobert or, if it helps:
r: say defined(Int); say defined(4);
p6eval rakudo c1ddea: OUTPUT«False␤True␤»
sirrobert (moritz's was more thorough =) 20:31
aharoni OK, I understand the difference now, but I'm not sure what is f(Int) good for. Can I do anything with that object?
20:31 birdwindupbird joined
moritz sure 20:31
you can use it for type checking, for example
or for creating new instances
or for introspection 20:32
sirrobert typechecking is the big deal (imho):
moritz r: say grep Int, 1, 2, 'string', ['arr', 'ay']
p6eval rakudo c1ddea: OUTPUT«1 2␤»
moritz there I pass the Int type object to the grep() function
sorear or for type-friendly semipredicates, Maybe-style 20:33
sirrobert my Int $foo; #do something else#; some_func($foo);
You don't know if $foo is defined or not at runtime, say.
But it does have to be an Int.
20:33 kurahaupo joined
aharoni Ah, finally found it in perlcabal.org/syn/S12.html . 20:34
I would expect it to appear in S03 - Operators, but I don't see anything like it there.
moritz well, actually S06 is about subroutines, and the signatures belong to subroutines 20:35
so I'd expect it in S06, or maybe S02, beause everything is in S02 :-)
aharoni I couldn't find it in S02 either. Of course, it's possible that I don't understand what to look for. 20:36
moritz indeed, it's only in S12 20:37
sirrobert What's the difference that makes the second one of these hang: 20:40
r: my %foo = {}; %foo ,= {a=>1}; say %foo;
p6eval rakudo c1ddea: OUTPUT«("a" => 1).hash␤»
sirrobert r: my $foo = {}; $foo ,= {a=>1}; say $foo;
p6eval rakudo c1ddea: OUTPUT«(timeout)»
sirrobert r: my $foo = {}; say $foo; $foo ,= {a=>1}; say $foo; 20:41
p6eval rakudo c1ddea: OUTPUT«(timeout)().hash␤»
sirrobert it's definitely a hash before the freeze
benabik $foo = $foo, {a => 1} ? 20:43
sirrobert except $foo is initialized as a hash
r: my $foo = {}; $foo ,= $foo, {a=>1};
p6eval rakudo c1ddea: ( no output )
benabik r: say {}, {a =>1} 20:44
p6eval rakudo c1ddea: OUTPUT«().hash("a" => 1).hash␤»
sirrobert no freeze
say {} ,= {a => 1};
benabik r: say {{}, {a =>1}}, {b => 2}
p6eval rakudo c1ddea: OUTPUT«Block.new()("b" => 2).hash␤»
sirrobert r: say {} ,= {a => 1};
p6eval rakudo c1ddea: OUTPUT«Cannot assign to a readonly variable or a value␤ in block at src/gen/CORE.setting:12001␤ in block at /tmp/JCV68ujC4X:1␤␤»
sirrobert that makes sense =)
benabik ().hash("a"=>1).hash ?
sirrobert: I don't think ,= does what you think it does. 20:45
sorear sirrobert: my $foo = {}; $foo ,= {a => 1} # $foo is now self referential
sirrobert benabik it does in the first example...
benabik I think it's creating something recursive.
sirrobert hmm
why does the first one work?
r: my %foo = {}; %foo ,= {a=>1}; say %foo;
p6eval rakudo c1ddea: OUTPUT«("a" => 1).hash␤»
sorear because you're using the % sigil
benabik r: my $foo = {}; $foo = %{$foo, a => 1}; say $foo; 20:46
p6eval rakudo c1ddea: OUTPUT«===SORRY!===␤Non-declarative sigil is missing its name␤at /tmp/vIWEU0UONO:1␤»
sirrobert so I can't *quite* use a hash stored in $ transparently...
benabik Bah.
sirrobert r: my Hash $foo = {}; $foo ,= $foo, {a=>1};
p6eval rakudo c1ddea: OUTPUT«Type check failed in assignment to '$foo'; expected 'Hash' but got 'Parcel'␤ in block at src/gen/CORE.setting:12001␤ in block at /tmp/Ch4GzkH6ZL:1␤␤»
20:46 benabik left
sirrobert r: my Hash $foo; $foo ,= $foo, {a=>1}; 20:47
p6eval rakudo c1ddea: OUTPUT«Type check failed in assignment to '$foo'; expected 'Hash' but got 'Parcel'␤ in block at src/gen/CORE.setting:12001␤ in block at /tmp/YuRcfXoaRw:1␤␤»
sirrobert huh
sorear sirrobert: don't use ,= ever, it's very wasteful
sirrobert I can *see* that the difference is $ vs %, but I don't understand why.
sorear: what's a better way to concatenate two hashes?
,= was suggested
or "merge" two hashes 20:48
sorear can't you use .push for that?
anyway 20:49
sirrobert I dunno
sorear $foo ,= {} expands to $foo = $foo, {}
or rather $foo = ($foo, {})
$foo may be a hash, but the variable itself doesn't know that
so after that statement, you are left with $foo being a Parcel pair ($foo, {})
sirrobert the reason that's a bummer is that means I can't do things like: %hash<x> = {...};
[Coke] r: my $foo = {a=>1}; my $bar = {bar =>2}; $foo.push($bar); say $foo.perl;
p6eval rakudo c1ddea: OUTPUT«Trailing item in Hash.push in block at /tmp/qiJedzNkOu:1␤␤{"a" => 1}␤»
sirrobert then %hash<x> ,= %hash-y 20:50
[Coke] r: my $foo = {a=>1}; my $bar = {bar =>2}; $foo.push($bar.kv); say $foo.perl;
p6eval rakudo c1ddea: OUTPUT«{"a" => 1, "bar" => 2}␤»
[Coke] sirrobert: there's a version with push
sirrobert [Coke]: thanks =)
20:50 SamuraiJack left
sirrobert my other version was a for loop. 20:50
[Coke] de nada
yah, I would ahve used a for loop, too. sorear++ 20:51
20:53 Targen left 20:55 hash_table left
sirrobert r: my %foo = {a=>1}; my %bar = {b=>2}; say %foo.push: %bar; 20:57
p6eval rakudo c1ddea: OUTPUT«("a" => 1, "b" => 2).hash␤»
20:57 adam7504 joined
sirrobert [Coke]: fwiw, the .kv isn't needed. 20:57
[Coke] it was in my example. 21:03
look at the scrollback. 21:04
21:06 Kim_Breitwieser left
sirrobert yeah, I was saying I re-tried your example... it works without the .kv 21:07
in case it was helpful
colomon sirrobert++ 21:11
21:15 sirrobert left 21:16 cognominal joined 21:18 adam7504 left, birdwindupbird left 21:19 antigremlin left 21:30 rindolf left
[Coke] I'm saying I just got an error without the kv. 21:31
can you show an example here that doesn't use the colon syntax, uses the $ var, and doesn't get the error?
(you changed both of those from my example: i'm betting it's the %-sigils that let it work.
21:40 flightrecorder left, FROGGS left 21:45 hash_table joined 21:50 rindolf joined 21:51 kurahaupo left 22:00 benabik joined 22:02 kaare__ left 22:04 stopbit left
sorear OT: SHA3 contest closes! www.nist.gov/itl/csd/sha-100212.cfm 22:04
rindolf sorear: interesting. 22:06
sorear: someone once interested me in optimising one of the entries for this competition, but they ended up not wanting my help.
22:10 bluescreen10 left 22:19 MayDaniel left 22:31 PacoAir left 22:36 whiteknight joined 23:31 Chillance left 23:45 macae left 23:50 rindolf left 23:51 benabik left 23:57 jerome left, benabik joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!