»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.
dalek href="https://modules.perl6.org:">modules.perl6.org: 5ddfccd | raydiak++ | web/lib/P6Project.pm:
Revert "Allow HTTP redirection" - causes other problems; discussion needed

This reverts commit e5af4fc7246755cbb3c0e055778c84834002867a.		 00:12
raydiak gotta run, back in a while
01:16 FROGGS_ joined 01:20 FROGGS left 01:30 raiph joined
raiph .tell FROGGS Just want to let you know I am working on a CPAN MIRRORED_BY grammar. My start is at gist.github.com/raiph/63508a5f4ade000acc7f 01:32
yoleaux raiph: I'll pass your message to FROGGS.
01:48 ilbot3 joined 01:49 adu joined
raydiak okay, so the problem with modules.perl6.org is that raw.github.com now redirects to raw.githubusercontent.com 01:53
there are 2 ways to fix it 01:54
either change the script that generates the module list for the website so it handles the redirects properly
or update all urls to point to the new domain 01:55
from a technical efficiency standpoint, it'd be better to use the new direct urls 01:57
01:57 klapperl_ joined
BenGoldberg There's also solution 1B, which is to cache the redirects ;) 01:57
raydiak but idk if there would be an ramifications (eg for tools like panda) would be if we changed *every* url in META.list all at once 01:58
hm
basically it's a question of whether we want to know when something has been redirected (by having it break), or just not have it break and have it work as long as possible, even when buried under umpteen layers of years-old redirects 01:59
01:59 klapperl left
raydiak if we cache then it eliminates the performance problem if we don't care about having updated urls 02:00
BenGoldberg Cache and give a warning ;)
And, perhaps, deprecate the old urls, eventually. 02:01
raydiak why are the best compromises always the most complex to code? :)
02:01 plobsing joined
BenGoldberg The best compromises are the ones least uncomfortable for the largest number of humans. Since humans are both complex and uncomfortable creatures, it's inevitable that the compromises will be too :) 02:02
raydiak hah...well put 02:03
lue raydiak: the clear solution is to abstract away the URLs :) 02:10
02:10 hoverboard joined, ivan`` left 02:15 ivan`` joined, raiph left, ivan`` left 02:19 ivan`` joined, adu left 02:20 ivan`` left 02:26 ivan`` joined
colomon tadzik: it appears my rakudobrew / panda problems came from having other installs of rakudo floating around. everything seems to be working now. (crosses fingers) 02:28
02:33 xragnar_ joined, xragnar left, xragnar_ is now known as xragnar
colomon tadzik: though when I do rakudobrew build-panda on my Mac, for some reason it doesn't create a link to the panda script in the .rakudobrew/bin directory? 02:36
02:37 pecastro joined
colomon doing rakudobrew switch moar does install it there. 02:37
(well, a link, I presume)
panda install NativeCall
resolve stage failed for NativeCall: Project NativeCall not found in the ecosystem
colomon is having one of those days when every single step fails. :\
ecosystem is broken at the moment? 02:39
on my Linux box, doing panda list produced the expected list of files. Then I did panda update and panda list again, and this time it is only the installed modules plus HTTP::Headers, HTTP::Message, and HTTP::Request 02:40
02:41 skids left
raydiak colomon: those are the only 3 modules not pointed at raw.github.com, which as of today, is a redirect...so its breaking stuff 02:42
colomon raydiak: tell me more, please?
I see they're pointed at raw2.github.com? 02:43
raydiak colomon: right, that one isn't a redirect yet
so those 3 still work
suppose we could just update all the urls in META.list
colomon to raw2 ? 02:44
raydiak no, to where they're redirecting to
raw.githubusercontent.com
I didn't know anything besides modules.perl6.org was broken 02:47
colomon are you sure that's the right URL? my naive attempts to access it have all just failed. 02:48
raydiak go to any of the urls in META.list and look at where your browser takes you
colomon WTF
I would swear that didn't work the first time I tried it. 02:49
and did the second time.
raw2.github.com seems to work as well
raydiak yeah but I'm assuming that will redirect eventually too 02:50
lue raydiak: I think the best solution might be to take away the place.github.com/ from the list, leaving things like moritz/json/master/META.info
(maybe make the first entry of the list the common part of the URL?)
raydiak but then we as module authors are inextricably tied to github with no choice to host elsewhere 02:51
iow it adds the artificial limitation that all modules must reside on the same domain
lue raydiak: maybe instead something like GITHUB_RAW:jrandom/... ?
dalek osystem: 48b2383 | (Solomon Foster)++ | META.list:
Switch all raw.github.com to raw.githubusercontent.com.
 02:52
raydiak that's not a bad idea
colomon wonder how long it will take it to get from there to feather?
lue I don't know how often modules.perl6.org is updated.
raydiak have to fix the META.list url in the script that does that too 02:53
I got it
dalek href="https://modules.perl6.org:">modules.perl6.org: 62151e8 | raydiak++ | web/build-project-list.pl:
Update META.list URL
 02:54
02:55 btyler joined 03:03 arturo_ left
dalek osystem: 4da5508 | raydiak++ | META.list:
Update URLs for Acme::Addslashes and Text::Abbrev
 03:04
03:13 benabik left 03:28 benabik joined 03:31 benabik left 03:44 sivoais joined 03:54 sivoais_ joined 03:55 sivoais left 03:56 hoverboard left 03:57 hoverboard joined, hoverboard is now known as moistcherry 04:07 btyler left 04:19 sivoais_ is now known as sivoais, sivoais left, sivoais joined, moistcherry left 04:20 BenGoldberg left 04:26 SamuraiJack joined 04:30 Psyche^ joined 04:34 Psyche^_ left
avuserow raydiak++, colomon++ # fixing the ecosystem 04:55
04:58 xenoterracide_ left
timotimo retupmoca: can you explain why you used read_int and read_varint instead of just read_int in the serialization and deserialization of CStruct? 05:49
05:52 vaskozl left 05:58 vaskozl joined 06:09 jlaire left 06:10 jlaire joined
dalek href="https://modules.perl6.org:">modules.perl6.org: 051a448 | raydiak++ | web/lib/P6Project/Hosts/Github.pm:
Catch GitHub API errors
 06:14
href="https://modules.perl6.org:">modules.perl6.org: 5569415 | raydiak++ | web/lib/P6Project/Hosts/Github.pm:
Fix github raw url in another place
href="https://modules.perl6.org:">modules.perl6.org: c3e1be6 | raydiak++ | web/build-project-list.pl:
Add missing line break after error list
raydiak modules.perl6.org should fix itself at the next cron run, now; temporary working mirror at cyberuniverses.com/perl6-modules/ 06:17
note panda and modules.perl6.org will still break when redirects are introduced 06:29
will touch base tomorrow when more people are online. good night, #perl6 \o
timotimo thank you! 06:30
gnite!
06:31 kaare_ joined 06:48 logie left, xinming_ joined 06:51 xinming__ left 06:55 darutoko joined 06:57 vaskozl left 06:59 ClarusCogitatio joined 07:03 vaskozl joined 07:04 Rotwang joined 07:05 darutoko left, darutoko joined 07:08 FROGGS_ is now known as FROGGS
FROGGS raiph++ 07:08
moritz \o
FROGGS morning moritz 07:09
07:10 darutoko left 07:11 darutoko joined
timotimo o/ 07:11
moritz somehow I've managed to roally screw up my laptop
after a few small updates, and booting to windows (and updating windows from 8.0 to 8.1), diretories under /var/lib were missing
like, /var/lib/{dpkg,postgresql} 07:12
it was bad enough that network wouldn't start anymore, and neither could I run apt/dpkg
07:12 kaare_ left
moritz so I tried to reinstall 07:12
timotimo what o_O
moritz which needed several attepts
and ended up killing both windows and linux 07:13
and didn't finish, 'cause it claimed there was an IO error while reading from the CD/DVD (which really is a USB stick)
so tried a different USB stick, same result
summary: linux broken, windows broken, all fucked up 07:14
(I did back up /etc/ and /home to an external drive before)
07:14 ponbiki_ joined
moritz I guess I should have checked sha1sums of the image before I started to download it 07:15
07:16 kaare_ joined, Rotwang left
FROGGS ohh dear :/ 07:16
moritz oh dear indeed :( 07:19
timotimo poor you ;( 07:20
moritz very strange; downloading that ubuntu image via firefox or wget consistently produces wrong sha1sums 07:28
downloading it via wget on a totally different machine in a totally different network works, though
now trying the wget -> scp cascade
moritz suspicious 07:29
masak mornin', #perl6 07:30
07:30 darutoko left
moritz sha1sum broken even after scp transfer. WTF? 07:34
\o masak
moritz tries again, and feels very stupid
"trying the same thing twice and expecting a different resut is a sign of stupidity" -- paraphrasing EY 07:35
timotimo this code is sooooo slooooow :| 07:36
moritz and it worked the second time. Now I fell *really* stupid 07:38
07:47 arturo_ joined 07:54 dmol joined
timotimo you can watch the screen build up >_< 07:54
FROGGS moritz: I'd suggest you do a memtest 07:56
timotimo is the hard drive failing? 07:59
08:02 pdcawley left 08:04 vaskozl left 08:05 pdcawley joined 08:06 fhelmberger joined
FROGGS it can also be the sata controller... there are many things that could cause this 08:09
08:10 vaskozl joined
masak moritz: apparently the "definition of insanity" quote comes from Narcotics Anonymous: en.wikiquote.org/wiki/Narcotics_Anonymous 08:10
(at least that's very likely the origin)
wikiquote++ 08:11
08:14 fhelmberger left 08:22 darutoko joined 08:32 arturo_ left 08:45 denis_boyun_ joined
tadzik colomon: I think I fixed it 08:57
08:57 sqirrel joined 09:02 denis_boyun__ joined 09:03 kaare_ left 09:04 denis_boyun_ left
vendethiel is there a "rule vs token vs regex" in the FAQ or something ? I recall something similar, but nothing in S05 09:05
arnsholt It's in S05 09:07
09:07 vaskozl left
arnsholt token is regex :ratchet, and rule is regex :ratchet :sigspace 09:09
vendethiel Ok, in "Regexes constitute a first-class language, rather than just being strings" found it
I think that could be moved to a FAQ Q/A ?
arnsholt Could be, yeah 09:10
09:12 pdcawley left 09:15 vaskozl joined
vendethiel m: my 'foo bar' ~~ my anon rule { 'foo' 'bar' } 09:22
camelia rakudo-moar 440739: OUTPUT«===SORRY!=== Error while compiling /tmp/M_7gE4LikT␤Malformed my␤at /tmp/M_7gE4LikT:1␤------> my ⏏'foo bar' ~~ my anon rule { 'foo' 'bar' ␤ expecting any of:␤ scoped declarator␤»
vendethiel m: my 'foo bar' ~~ (my anon rule { 'foo' 'bar' })
camelia rakudo-moar 440739: OUTPUT«===SORRY!=== Error while compiling /tmp/Zdi4yn4WJn␤Malformed my␤at /tmp/Zdi4yn4WJn:1␤------> my ⏏'foo bar' ~~ (my anon rule { 'foo' 'bar'␤ expecting any of:␤ scoped declarator␤»
vendethiel m: say 'foo bar' ~~ (my rule foo { 'foo' 'bar' })
camelia rakudo-moar 440739: OUTPUT«｢foo bar｣␤␤»
vendethiel m: say 'foo bar' ~~ (my regex foo { 'foo' 'bar' })
camelia rakudo-moar 440739: OUTPUT«Nil␤»
vendethiel Like `regex : Backtracks, no significant space. token : Doesn't backtrack, doesn't have significant space. rule : Doesn't backtrack, has sigspace` 09:26
09:27 Rotwang joined, Rotwang left
FROGGS is 'no significant space' clear enough for a beginner? 09:30
vaskozl FROGGS: what about him?
FROGGS maybe 'ignores whitespace' might be easier to understand
or something like that 09:31
vaskozl perl6 introduces significantly more whitespace doesn't it?
FROGGS errrm, what? 09:33
vaskozl like loop ($j=0; $j < 4; $j++) 09:34
compared to: for($j=0;$j<4;$j++)
FROGGS ahh
well, yes
it uses whitespace to distinguish between subcalls and things that are not 09:35
vaskozl FROGGS: yes I know now
what about the $j<4 becomming $j < 4 09:36
I've always liked perl because I don't have to remember how much whitespace there was where
FROGGS you don't have to put space around an infix if it is obvious and there is only one waay to parse it 09:37
if you would introduce an < post- or prefix you would have to put space around the infix
vendethiel FROGGS: maybe that could point to a relevant place in the syn or something else ?
FROGGS vendethiel: perhaps, yes (not that I know the relevant places offhand :o) 09:38
vaskozl: I find the rules in P6 about whitespace even easier to understand, and therefore the code is easier to understand for me
vendethiel m: sub postfix:<< < >>($n) { $n ** $n }; say 5< 09:39
camelia rakudo-moar 440739: OUTPUT«3125␤»
FROGGS m: say 1+2; sub postfix:<+>(|) { }; say 1+2 # vaskozl: see 09:40
camelia rakudo-moar 440739: OUTPUT«===SORRY!=== Error while compiling /tmp/SF5kDmizJJ␤Two terms in a row␤at /tmp/SF5kDmizJJ:1␤------> say 1+2; sub postfix:<+>(|) { }; say 1+⏏2 # vaskozl: see␤ expecting any of:␤ postfix␤ inf…»
vaskozl m: my $i;loop ($i;$i<10;$++){ say "The numbers is $i";}
camelia rakudo-moar 440739: OUTPUT«===SORRY!=== Error while compiling /tmp/eI4DcCr9PA␤Whitespace required before < operator␤at /tmp/eI4DcCr9PA:1␤------> $i;$i<10;$++){ say "The numbers is $i";}⏏<EOL>␤ expecting any of:␤ postfix␤»
FROGGS vaskozl: I like that way, because you can always be sure to parse it right
vendethiel m: class A { has $.n; method postcircumfix:<( )>($n) is export { 'hey you !'; } }; sub postfix:<< < >>($n) { new A(:$n); }; say 5< 3; 09:41
camelia rakudo-moar 440739: OUTPUT«===SORRY!=== Error while compiling /tmp/83ZPvhYy8S␤Unsupported use of C++ constructor syntax; in Perl 6 please use method call syntax␤at /tmp/83ZPvhYy8S:1␤------> !'; } }; sub postfix:<< < >>($n) { new A⏏(:$n); }; s…»
vendethiel m: class A { has $.n; method postcircumfix:<( )>($n) is export { 'hey you !'; } }; sub postfix:<< < >>($n) { A.new(:$n); }; say 5< 3;
camelia rakudo-moar 440739: OUTPUT«===SORRY!=== Error while compiling /tmp/iw12IGSbfh␤Two terms in a row␤at /tmp/iw12IGSbfh:1␤------> fix:<< < >>($n) { A.new(:$n); }; say 5< ⏏3;␤ expecting any of:␤ postfix␤ infix stopper␤…»
vendethiel m: class A { has $.n; method postcircumfix:<( )>($n) is export { 'hey you !'; } }; sub postfix:<< < >>($n) { A.new(:$n); }; say 5<(3); 09:42
camelia rakudo-moar 440739: OUTPUT«hey you !␤»
vendethiel :D
09:42 sqirrel left
vaskozl m: loop (my $i;$i <'10';$i++){ say "The numbers is $i";} 09:42
camelia rakudo-moar 440739: OUTPUT«use of uninitialized value of type Any in numeric context␤use of uninitialized value of type Any in string context␤The numbers is ␤The numbers is 1␤The numbers is 2␤The numbers is 3␤The numbers is 4␤The numbers is 5␤The numbers is 6␤The n…»
vendethiel $i = 0; $i < 10 ?
vaskozl what's the uninitailised value?
vendethiel: oh thx 09:43
vendethiel m: loop (my Int $i;$i <10;$i++){ say "The numbers is $i";}
vaskozl m: loop (my $i=0;$i <'10';$i++){ say "The numbers is $i";}
camelia rakudo-moar 440739: OUTPUT«Invocant requires an instance, but a type object was passed␤ in method Bridge at src/gen/m-CORE.setting:4491␤ in sub infix:<<> at src/gen/m-CORE.setting:4404␤ in block at /tmp/VRoB8Ssf2K:1␤␤»
rakudo-moar 440739: OUTPUT«The numbers is 0␤The numbers is 1␤The numbers is 2␤The numbers is 3␤The numbers is 4␤The numbers is 5␤The numbers is 6␤The numbers is 7␤The numbers is 8␤The numbers is 9␤»
vendethiel m: loop (my int $i;$i <10;$i++){ say "The numbers is $i";}
camelia rakudo-moar 440739: OUTPUT«The numbers is 0␤No such method 'STORE' for invocant of type 'Int'␤ in sub postfix:<++> at src/gen/m-CORE.setting:4554␤ in block at /tmp/ekw87rbgor:1␤␤»
vendethiel ah yeah, there's no postfix:<++> 09:44
vaskozl vendethiel: huh?
vendethiel "int" is a native type
FROGGS there is no such postfix for that type YET 09:51
it should get one though :o)
m: loop (my int $i;$i <10;$i=$i+1){ say "The numbers is $i";}
camelia rakudo-moar 440739: OUTPUT«The numbers is 0␤The numbers is 1␤The numbers is 2␤The numbers is 3␤The numbers is 4␤The numbers is 5␤The numbers is 6␤The numbers is 7␤The numbers is 8␤The numbers is 9␤»
FROGGS m: loop my int $i;$i <10;$i=$i+1 { say "The numbers is $i";}
camelia rakudo-moar 440739: OUTPUT«===SORRY!=== Error while compiling /tmp/hHN2_hDkvn␤Missing block␤at /tmp/hHN2_hDkvn:1␤------> loop ⏏my int $i;$i <10;$i=$i+1 { say "The numb␤ expecting any of:␤ statement list␤ scoped bloc…»
FROGGS ohh, sad
but yeah, makes sense 09:52
(because of the semicolon)
10:15 IllvilJa left 10:16 vaskozl left, rindolf joined 10:20 yoleaux joined, ChanServ sets mode: +v yoleaux
tadzik wow, new ubuntu actually ships with perl 5.18 10:21
10:22 vaskozl joined
nwc10 is it newer than the Rakudo they package? :-) 10:22
Ulti tadzik: also rakudo from December 2012
nwc10 so, yes! :-)
Ulti sorry 2013!
nwc10 oh. :-(
Ulti keep forgetting its 2014
because I'm olde 10:23
moritz faq.perl6.org/#token # already in the FAQ 10:24
because it's, like, a frequently asked question :-)
anybody wants to hack up perl6/doc htmlify.pl? 10:26
the index page could list the title of the "Language Documentation" docs next to the link 10:27
vendethiel moritz: I double-checked ôO 10:35
How in the name of camelia did I miss that 10:36
10:45 dmol left
colomon tadzik: fixed what? 10:48
tadzik colomon: 'panda' not available after build-panda 10:51
colomon tadzik++
tadzik colomon++ # #1 bugfinder
colomon now if the ecosystem was just functional.... 10:53
colomon does not think the problem is tadzik++'s fault 11:01
tadzik that depends. What's dysfunctional? :) 11:03
colomon Pretty much everything.
right now, from my systems the ecosystem only appears to have three modules in it.
HTTP::Headers
HTTP::Message
HTTP::Request
all the modules located at raw.github.com were not working. 11:04
last night I changed ALL the raw.github.com modules to raw.githubusercontent.com
which seems to work better if I call it up directly.
but this has had no apparent effect on the ecosystem. 11:05
might just be that feather hasn't updated the list?
tadzik let me see 11:06
11:06 SamuraiJack left
tadzik :o 11:06
crazy
ah, the new ones are SSL'd 11:07
and lwp-request doesn't like that
colomon tadzik: irclog.perlgeek.de/perl6/2014-04-26 (discussion starts, pauses briefly, then continues)
tadzik colomon: the modules that work use raw2.github.com, maybe we should switch to that 11:08
colomon but you may have just figured out more than we did already
can do for the ecosystem
tadzik yes, that fixes it 11:09
I'll do that in the ecosystem repo as well 11:10
dalek osystem: f4865e5 | (Solomon Foster)++ | META.list:
Try raw2.github instead of raw.githubusercontent.
 11:12
tadzik ah, you did that alreayd :)
colomon :)
tadzik but you did .github.com.com :P
11:12 spider-mario joined
tadzik I fixed the list on feather for the time being 11:13
dalek osystem: 3c61acf | (Solomon Foster)++ | META.list:
.com, not .com.com.

Sigh.
tadzik looks good now :)
colomon raydiak found a spot in modules.perl6.org that needed to be fixed as well
github.com/perl6/modules.perl6.org...62151e893a 11:14
tadzik also, we may just switch that list to user/repo
instead of the full url
colomon tadzik++ # fixing feather, panda works again!
tadzik there are two places that'll need to be upated after that
but I think it's for the best
colomon how would we indicate to use something other than github? 11:15
tadzik ah :) 11:16
dalek href="https://modules.perl6.org:">modules.perl6.org: 5f7a6a1 | (Solomon Foster)++ | web/build-project-list.pl:
Switch $list_url to use raw2.github instead of raw.githubusercontent.
 11:19
11:24 vaskozl left 11:26 darutoko left 11:27 darutoko joined 11:28 vaskozl joined
masak heh. is Burnside's lemma too advanced for an introductory Python course? :P 11:29
ooh, let's make that today's mini-challenge, actually.
a factory is making necklaces with 16 beads each; 13 black and 3 white. they become ridiculously popular, with people collecting unique ones.
how many unique necklaces are there? 11:30
the necklaces have reflection symmetries, and rotational symmetries. there's no buckle. 11:31
this problem is simple enough that you can solve it with a one-liner in Perl 6, but scripts are fine too.
colomon m: (2..*).map(*.is-prime).first-index(73).say 11:39
camelia rakudo-moar 440739: OUTPUT«(timeout)»
colomon m: (2...*).map(*.is-prime).first-index(73).say
moritz colomon: try .grep
camelia rakudo-moar 440739: OUTPUT«(timeout)»
colomon moritz++
colomon--
m: (2...*).grep(*.is-prime).first-index(73).say
camelia rakudo-moar 440739: OUTPUT«20␤»
colomon m: my @primes := (2...*).grep(*.is-prime); my @reversible := @primes.grep(*.flip.is-prime); say @reversible[^10] 11:41
camelia rakudo-moar 440739: OUTPUT«No such method 'is-prime' for invocant of type 'Str'␤ in block at /tmp/w5fQ7g5pu6:1␤␤»
colomon m: my @primes := (2...*).grep(*.is-prime); my @reversible := @primes.grep(*.flip.Int.is-prime); say @reversible[^10]
camelia rakudo-moar 440739: OUTPUT«2 3 5 7 11 13 17 31 37 71␤»
colomon m: my @primes := (2...*).grep(*.is-prime); my @reversible := @primes.grep(*.flip.Int.is-prime); say @reversible[^20] 11:44
camelia rakudo-moar 440739: OUTPUT«2 3 5 7 11 13 17 31 37 71 73 79 97 101 107 113 131 149 151 157␤»
11:46 xfix joined
masak m: say "31".is-prime 11:51
camelia rakudo-moar 440739: OUTPUT«No such method 'is-prime' for invocant of type 'Str'␤ in block at /tmp/pEqc9xCT4j:1␤␤»
masak fair enuff.
m: say 3.5.is-prime
camelia rakudo-moar 440739: OUTPUT«No such method 'is-prime' for invocant of type 'Rat'␤ in block at /tmp/yggiVl0KUt:1␤␤»
masak why is .flip on Cool, but .is-prime on Int?
colomon I vaguely recall a discussion on why .is-prime should be on Int instead of Cool. 11:53
it was a while ago. :)
moritz because primality testing is very integer specific
11:53 zakharyas joined
moritz and not everything can be coerced sensibly to Int 11:53
whereas everything Cool has a clear and well-defined coercion to Str 11:54
masak ok, makes sense.
colomon Are GTK+ and GTK the same thing? 11:55
moritz seems like they aren't 11:57
The name GTK+ originates from GTK; the plus was added to distinguish an enhanced version # from en.wikipedia.org/wiki/GTK+ 11:58
colomon but it also seems like GTK+ is what people mean when they say GTK? Or is that wrong? 12:00
moritz it would seem to me, but I'm not really sure
12:01 SamuraiJack joined
_sri wonders what a perl6-ish rewrite of Mojo::EventEmitter with supplies and promises would look like mojolicio.us/perldoc/Mojo/EventEmitter 12:02
perhaps something like $foo.on('bar').tap({...}) 12:08
_sri also wonders if there's a way to get a promise only for the next value of a supply 12:10
$foo.on('bar').next.then({...}) 12:11
colomon is scared by the OS X build instructions for GTK+ 12:12
"If you have MacPorts or Fink installed, you must remove all traces of them from your environment before you try to run jhbuild. " 12:14
moritz "excise all traces with fire, or preferably with nukes and lasers from outer space" 12:15
colomon seriously, these are some of the most ridiculous instructions i've ever seen for building software. 12:24
12:27 vaskozl left 12:34 vaskozl joined 12:54 skids joined 12:56 xenoterracide_ joined 13:26 SamuraiJack left 13:32 vaskozl left 13:35 vaskozl joined 13:39 dmol joined
geekosaur os x is rather a challenge; if you start digging into *why* they say that, the road leads ultimately to decisions made by apple 13:51
however, if you have macports/fink/homebrew installed, you can install gtk+ via that and ignore those build instructions 13:52
FROGGS "lovely" 13:53
yoleaux 01:32Z <raiph> FROGGS: Just want to let you know I am working on a CPAN MIRRORED_BY grammar. My start is at gist.github.com/raiph/63508a5f4ade000acc7f
FROGGS ohh, wb yoleaux
colomon geekosaur: alas, I just have broken, leftover bits of a macports install around
geekosaur those instructions are specifically for building against a pure-apple ecosystem, and that means living with out of date and sometimes incompatible versions of libraries shipped by apple and mixing in the modern stuff provided by third party package managers is a recipe for weird link errors or core dumps
colomon I suppose trying to fix my macports might be the most straightforward thing to do. 13:54
geekosaur iconv in particular is an infamous point of collision because apple ships an ancient one and if you get any parts of a newer one tangled up in it your programs WILL dump core on startup 13:55
14:00 pippo joined
pippo Hello #perl6! 14:00
FROGGS hi pippo 14:01
pippo I am trying to install DBIish with panda but if fails tests. Any ideas? gist.github.com/anonymous/11320743
moritz tries it too 14:04
dalek osystem: 441275d | (Andrew Egeler)++ | META.list:
Add Net::SOCKS
FROGGS pippo: what is your `perl6-m --version` ? 14:06
moritz oh, it's actually a NativeCall test failure 14:07
pippo FROGGS: This is perl6 version 2014.04-98-g440739c built on MoarVM version 2014.04-29-ga109e8d
moritz: yes!
FROGGS pippo: that should work out...
must be a regression, since it was working for 2014.04 14:08
moritz works on HEAD for me too
pippo FROGGS: unfortunately does not on my box. Anything relevant to 32/64 bits? 14:09
FROGGS pippo: most of us are on x86_64, so I can only tell about that
that is, x64 should work
pippo I have a 32 bits box. 14:10
FROGGS :/
pippo: can you make an RT ticket with some system information?
pippo I can put everything in a gist. RT I do not know how to do it. Sorry :-( 14:12
14:13 skids left
FROGGS pippo: you can email to [email@hidden.address] 14:18
but a gist is fine also, I'll create the ticket then 14:19
pippo: would be enough if you pasted `perl -V` to the gist you posted already 14:21
pippo FROGGS: gist.github.com/anonymous/11321347
FROGGS: let me know if I need to add some other infos.
FROGGS pippo: please add the out of `perl -V` 14:22
pippo FROGGS: sorry I did not see your request. I'll do...
FROGGS np :o) 14:23
pippo FROGGS: here it is: gist.github.com/anonymous/11321460 14:25
FROGGS pippo++ # thank you 14:26
14:26 Rotwang joined
pippo FROGGS: thanks to you. o/ 14:27
14:27 pippo left
FROGGS I won't have time for it today but hopefully tomorrow 14:29
but loop labels for nqp-j are kinda burning under my fingernails, though I tried that already and failed :( 14:30
masak FROGGS++ # courage 14:32
grondilu mentioned failure to install NativeCall on 32bit yesterday. Pippo's issues seem very much related. I could install NativeCall locally by replacing 'int' with 'int32' in relevant places in the *.t files. 14:39
14:40 pippo joined, xenoterracide_ left 14:41 vaskozl left
pippo grondilu: I'll try and let you know ... 14:41
14:47 vaskozl joined
pippo grondilu: I confirm I have same results here i.e. tests work if int32 iso int. 14:55
o/
14:55 pippo left 15:00 xenoterracide joined 15:24 Rotwang left 15:40 guru joined 15:41 guru is now known as ajr_ 15:42 darutoko- joined, darutoko left 15:46 vaskozl left, logie joined 15:52 vaskozl joined 15:57 zakharyas left 16:03 benabik joined, Maddingue joined 16:04 logie left 16:11 guest10 joined, guest10 left 16:15 hoverboard joined 16:16 lizmat joined 16:21 Shozan is now known as SHODAN 16:23 xenoterracide left, xenoterracide joined 16:26 ssutch joined
masak misses sorear 16:27
FROGGS yeah
16:31 ssutch left
raydiak morning #perl6 16:34
moritz good UGT, raydiak 16:35
raydiak \o 16:37
16:37 hoverboard left
raydiak I recommend installing ssl on feather and pointing the links to their new location. wouldn't count on raw2 being not a redirect forever. 16:39
also, was wondering if we want redirection to work with the modules.perl6.org script, or if we want to know when they change (by them breaking) 16:41
timotimo perhaps we want reports from the cron job to be received by many members of the community 16:43
and output things like that at the top if possible
or at least easy to see at a glance 16:44
16:49 denis_boyun joined, denis_boyun__ left
raydiak apologies, apparently my morning has other plans for me. :P back in a while. 16:50
17:11 zakharyas joined 17:21 hoverboard joined 17:23 anaeem1 joined
timotimo :) 17:25
that's fair
17:47 anaeem1 left 17:48 xenoterracide left 17:49 xenoterracide joined 17:50 anaeem1_ joined 17:52 anaeem1_ left 17:53 anaeem1_ joined 17:55 denis_boyun left 17:57 rurban joined 18:10 jtpalmer left 18:11 telex left, jtpalmer joined 18:12 telex joined 18:15 xenoterracide left 18:18 anaeem___ joined 18:21 anaeem1_ left 18:26 woosley left 18:28 woosley joined 18:29 zakharyas left
dalek rl6-most-wanted: e1cc7bc | (Andrew Egeler)++ | most-wanted/modules.md:
Mention Net::IMAP
 18:31
18:40 anaeem___ left 18:45 anaeem1_ joined 18:47 hoverboard left 18:48 kurahaupo joined 18:49 raiph joined 18:50 darutoko- left 18:53 xinming_ left 18:59 xinming_ joined 19:07 rindolf left 19:08 raiph left 19:09 raiph joined 19:20 anaeem1_ left 19:35 anaeem1_ joined
vaskozl Hey I'm trying to learn perl6 and I'm confused as to why you have all these "methods" that you put after a scalar or so with a dot instead of just having functions 19:35
I don't really like these methods for now :(
masak I'm willing to explain, but I need to understand your concern/confusion first. 19:36
vaskozl yah I've never really understood object orientation
19:36 Su-Shee joined
masak methods are like functions, but they are "tied" to a certain class, so that all objects of that class can be called with that method. 19:36
Su-Shee good evening.
masak Su-Shee! \o/
vaskozl hey
masak vaskozl: here, let me put together an example, so we can talk about it.
vaskozl masak: examples would probably help 19:37
raiph By default: when regex partially match, then fail, they backtrack to try another way; tokens don't backtrack; rules don't backtrack and add <ws>+ between elements. #6now #grammar #regex #rule #token #faq irclog.perlgeek.de/perl6/2014-04-26#i_8640947
vaskozl I don't really understandthe part: that all objects of that class can be called with that method
raiph: please take small steps.. 19:38
masak m: class Computer { has $.name; method greet() { say "Bleep blorp, I am $.name" } }; my $c1 = Computer.new(:name<COMPUTRON>); my $c2 = Computer.new(:name<FORTRANATOR>); $c1.greet; $c2.greet
camelia rakudo-moar 440739: OUTPUT«Bleep blorp, I am COMPUTRON␤Bleep blorp, I am FORTRANATOR␤»
masak vaskozl: hoping that was a small enough step. please study the code above.
vaskozl tries to proccess what he sees
masak there's a class declaration, telling you what a "Computer" is.
raiph vaskozl: what I wrote has nothing to do with your dialog with masak, please ignore it 19:39
masak it contains an attribute $.name, which all Computers have.
and a method, greet, which all Computers do.
next, we create ("instantiate") two new Computer objects, and put them in $c1 and $c2.
then we call their methods.
note *especially* that each Computer object knows what it's called. that's the crucial bit.
objects are bundles of data (like $.name) traveling along with shared methods (like greet). 19:40
hth
19:41 hoverboard joined
vaskozl masak: what do you call an object? 19:41
masak the things in $c1 and $c2. 19:42
vaskozl aren't those scalars?
masak the variables are scalars, yes.
the values inside of them are objects.
we can ask values to identify themselves. hold on. 19:43
m: say 42.WHAT
camelia rakudo-moar 440739: OUTPUT«(Int)␤»
masak m: class Computer { has $.name; method greet() { say "Bleep blorp, I am $.name" } }; my $c1 = Computer.new(:name<COMPUTRON>); say $c1.WHAT
camelia rakudo-moar 440739: OUTPUT«(Computer)␤»
masak see? one is a normal Int, the other is a Computer object.
(the cool thing being that "Computer" is a type that we just added to the programming language ourselves) 19:44
m: say [].WHAT; say {}.WHAT; say (2/3).WHAT 19:45
camelia rakudo-moar 440739: OUTPUT«(Array)␤(Hash)␤(Rat)␤»
vaskozl so why would we just do: sub computer { say "Bleep blorp, I am @_[0]"; }; computer('COMPUTRON'); computer('FORTRANTOR');
*wouldn't
masak you can, in a lot of cases. 19:46
but objects *persist* after you've called them.
vaskozl persist?
masak you can see that from the order in which I did things: create $c1, create $c2, call $c1.greet, call $c2.greet
with your solution, you call them once, and then they're gone. 19:47
that's, like, the opposite of persisting :)
yes, "persist" as in "stay around for as long as you need them, holding the data you've put into them".
vaskozl masak: call $c1.greet, when did we do that?
masak: OOH, it holds the original input! 19:48
masak yes!
*lightbulb on*!
:D
vaskozl but why don't we just have $inputcomputer1 and $outputcomputer1
masak not sure what you mean.
vaskozl $output1 = sub($input1); 19:49
masak the cool thing is that we can "put away" the names of the computers into those objects, and then they are not really our concern any more. they "belong" to the objects, and the objects manage that data.
19:49 anaeem1_ left
masak ah, well what you're suggesting works in small cases, but it quickly gets messy for bigger things. 19:49
vaskozl what was the $c1.greet you mentioned? 19:50
masak the big win here is that objects "hide" their data and expose everything through the methods instead, so in some sense we don't have to care as much about how the objects do things.
$c1.greet is a method call I did above.
the crucial thing is that I defined greet as a method on the class ("Computer"), but then I called it on an object ($c1).
vaskozl did we have $c1.greet in our code? 19:51
oh yah I see
masak yes, we did :) 19:52
vaskozl wow now I think I understand it
masak \o/
vaskozl: let's take a bigger example. say you have a window manager.
vaskozl so we can have a bunch of different methods
and each method may use different attributes
masak vaskozl: each individual thing on the screen that you care about can be an object: windows, text boxes, buttons, labels, etc.
vaskozl I'm taking a leap of fait here 19:53
19:53 rurban left
masak vaskozl: but they often have things in common, so their classes often share many attributes and methods between each other (through various mechanisms such as inheritance) 19:53
vaskozl: for example, you can ask almost any object for its width, or tell it to place itself somewhere on the screen.
or maybe there's a "click" method that gets called when the user clicks on one of the objects on the screen. 19:54
now, all this *could* be done with just subroutines as you propose, but it's usually nicer to separate things into their own classes and objects.
think of it as a way of organizing things.
vaskozl alright
masak object oriented programming is not a God-given. it's just one of many possible ways to manage complexity. 19:55
vaskozl it seems to me that's it makes managing a bunch of similar things more easily 19:56
like a lot of buttons
masak right.
you can have hundreds of button objects on-screen, but they're all managed by the same Button class.
so each method in that class only needs to exist in one place in memory.
and each Button object holds its individual state, such as position, button text, and whether it's being pressed down or not. 19:57
vaskozl masak: I'm stunned by how quickly and efficiently you are explaing this to me
masak vaskozl: I'm glad.
vaskozl: at $dayjob, I teach.
vaskozl masak: and there are a bunch of default methods in perl6? 19:58
like .get and .words
masak yeah, there are hundreds.
m: for "do you feel lucky punk".words -> $word { say "$word.uc()." } 19:59
camelia rakudo-moar 440739: OUTPUT«DO.␤YOU.␤FEEL.␤LUCKY.␤PUNK.␤»
masak .words takes a string and divides into chunks of non-whitespace.
vaskozl what's the .uc method do?
masak .uc takes a string and Upper-Cases it.
there's also .lc and .tc :) 20:00
feel free to try.
vaskozl oh..
that's pretty cool
masak those particular methods are "string methods".
I don't know whether they're actually on the Str class in this case, but for simplicity, let's assume they are. 20:01
m: say ?Str.has("uc")
camelia rakudo-moar 440739: OUTPUT«No such method 'has' for invocant of type 'Str'␤ in block at /tmp/5mOhsLa1jq:1␤␤»
masak m: say ?Str.^has("uc")
camelia rakudo-moar 440739: OUTPUT«No such method 'has' for invocant of type 'Perl6::Metamodel::ClassHOW'␤ in block at /tmp/aeXUv_Tv9H:1␤␤»
masak hrm.
m: say ?Str.^does("uc")
camelia rakudo-moar 440739: OUTPUT«False␤»
masak there we go.
m: say ?Str.does("uc")
camelia rakudo-moar 440739: OUTPUT«True␤»
masak m: say ?Str.does("scary-freakin-dinosaurs") 20:02
camelia rakudo-moar 440739: OUTPUT«True␤»
masak o.O
m: say Str.does("scary-freakin-dinosaurs")
camelia rakudo-moar 440739: OUTPUT«True␤»
masak oh, I'm using the wrong method :P
retupmoca m: say ?Str.can("scary-freakin-dinosaurs")
camelia rakudo-moar 440739: OUTPUT«False␤»
masak m: say ?Str.can("uc")
camelia rakudo-moar 440739: OUTPUT«True␤» 20:03
masak m: say ?Str.can("scary-freakin-dinosaurs")
camelia rakudo-moar 440739: OUTPUT«False␤»
masak that's better.
ignore my fumbling with "has" and "does" -- apparently, I'm tired.
vaskozl I got kinda lost with all the ?Str.^ and what's true or false
masak yeah, just look at the last two evals.
objects "know" what you can call on them and what you can't.
20:04 Mouq joined
masak let's get back to basics. 20:04
vaskozl so are you checking if scary-freakin-dinosaurs is a method?
masak right.
(it isn't)
vaskozl surprise surprise!
masak Perl 6 can surprise you :)
for example, we have a Mu class.
someone also added this lately: 20:05
m: say 42.WHY
camelia rakudo-moar 440739: OUTPUT«Life, the Universe and Everything␤»
vaskozl m: say 42.why
camelia rakudo-moar 440739: OUTPUT«No such method 'why' for invocant of type 'Int'␤ in block at /tmp/BDAywA9HeU:1␤␤»
masak case-sensitive.
needs to be .WHY
vaskozl why is it in all caps?
masak some methods are special.
.WHAT and .WHY are kind of "debugging methods".
not meant for a healthy program.
vaskozl masak: ok that makes sense 20:06
masak so they shout to mark that something is going on.
oh, speaking of all-caps things...
gonna show you something. :)
Mouq m: say "1+20*2+e**(2*pi*i).re".EVAL 20:07
camelia rakudo-moar 440739: OUTPUT«42␤»
Mouq m: say "1+20*2+e**(2*pi*i).re".EVAL.WHY
camelia rakudo-moar 440739: OUTPUT«(Any)␤»
masak m: for reverse 1 .. 10 -> $count { LAST { say "liftoff! lots of smoke and fire!" }; say "$count..." }
camelia rakudo-moar 440739: OUTPUT«10...␤9...␤8...␤7...␤6...␤5...␤4...␤3...␤2...␤1...␤liftoff! lots of smoke and fire!␤»
Mouq m: say "1+20*2+e**(2*pi*i).re".EVAL.Int.WHY
camelia rakudo-moar 440739: OUTPUT«Life, the Universe and Everything␤»
masak vaskozl: see the countdown above :)
Mouq also o/
masak vaskozl: and the LAST block, which gets executed after the last iteration.
Mouq: \o
20:08 LLamaRider joined
masak vaskozl: in this case, the "LAST" is uppercase because the special thing going on is "out-of-the-flow" timing of things. the LAST doesn't get executed where it is, but afterwards. 20:08
in the same vein: 20:09
m: say "HAI"; BEGIN { say "OH" }
camelia rakudo-moar 440739: OUTPUT«OH␤HAI␤»
masak (BEGIN blocks run during compilation; the rest of the program runs during runtime.)
Mouq m: END {say now - BEGIN now}; my $expensive-computation = [+] 0..100000 20:10
camelia rakudo-moar 440739: OUTPUT«0.60834820␤»
20:10 vaskozl left
Mouq m: END {say now - BEGIN now}; my $expensive-computation = [+] 0..100000; say $expensive-computation 20:11
camelia rakudo-moar 440739: OUTPUT«5000050000␤0.6682453␤»
20:11 vaskozl joined
vaskozl s/between to/between two/;s/keywoard/keyword/ 20:11
oh.. did I lose connection?
masak for a little while.
see /topic for a backlog :) 20:12
vaskozl masak: uhm what in the topic?
20:12 woolfy joined
vaskozl what was the last thing I said? 20:12
masak vaskozl: irclog.perlgeek.de/perl6/today
woolfy: \o 20:13
woolfy masak \o/
lizmat and jnthn are tired from NLPW and King's Day festivities and bicycle trip
masak I can imagine :)
vaskozl I don't really like too much all caps :/, which I think perl handles quite well 20:15
masak most of the keywords are lowercase :)
the all-caps things are reserved for special stuff.
Mouq really likes that he can do $var-name rather than $var_name or $varName. so much nicer and easier to type 20:16
vaskozl I'm gonna paste my question I asked when I DC'd (sorry):
is there an object that removes newlines from a string allowing us to regex the entire file instead of line by line?
I often want to extract something between two <plaintext>extract me!</plaintext> tags in a file
that I know to be after some keyword 20:17
and I never know how to do it :/
currently I'm stupidly running trough sed to remove newlines and passing the input as one line
s/object/method/
masak vaskozl: I think what you want is the IO.slurp method. 20:18
vaskozl: it reads in a file as a single string, which you can then manipulate as you see fit.
FROGGS "myfile.txt".IO.slurp.subst(/.../, '', :g)
Mouq m: say "ab\ncd\n<plaintext>\nextract me!\n</plaintext>" ~~ ms/'<plaintext>' .*? '</plaintext>'/
camelia rakudo-moar 440739: OUTPUT«｢<plaintext>␤extract me!␤</plaintext>｣␤␤»
Mouq m: say "ab\ncd\n<plaintext>\nextract me!\n</plaintext>" ~~ ms/'<plaintext>' <( .*? )> '</plaintext>'/ 20:19
camelia rakudo-moar 440739: OUTPUT«｢extract me!␤｣␤␤»
vaskozl alright that's cool! 20:21
how'd we do it with STDIN?
masak m: say get
camelia rakudo-moar 440739: OUTPUT«There were three men came out of the West␤»
masak m: say get.subst(/men/, 'really cool dudes')
camelia rakudo-moar 440739: OUTPUT«There were three really cool dudes came out of the West␤»
vaskozl masak: what did you just do? 20:22
lue m: say $*IN.slurp.subst(/men/, 'really cool dudes', :g); # :P
camelia rakudo-moar 440739: OUTPUT«There were three really cool dudes came out of the West␤Their fortunes for to try␤And these three really cool dudes made a solemn vow␤John Barleycorn must die␤␤They've ploughed, they've sewn, they've harrowed him in␤Threw clouds upon his head…»
masak lue: :)
vaskozl: 'get' reads a line from STDIN.
vaskozl: .subst does the substitution.
FROGGS vaskozl: there is a textfile attached to camelia's stdin 20:23
vaskozl oh...
masak vaskozl: lue++ just showed how to do the substitution "globally" (:g), many times.
oh, yes. textfile. forgot to mention.
I have no idea what that poem is.
but I realize I haven't had enough irreverent fun with it since it changed from the Austrian national anthem. :)
20:24 raiph left
Mouq vaskozl: $*IN.slurp or lines.join ($*IN is set to STDIN by default) 20:24
vaskozl: We have the evalbot's STDIN set to github.com/perl6/evalbot/blob/master/stdin
Timbus slurp works on $*IN on nits own doesnt it?
vaskozl now I know where to contribute...
Timbus just like lines and get
masak m: for lines() -> $line { say $line.subst(/\w+$/, -> $/ { $/.uc }) }
camelia rakudo-moar 440739: OUTPUT«There were three men came out of the WEST␤Their fortunes for to TRY␤And these three men made a solemn VOW␤John Barleycorn must DIE␤␤They've ploughed, they've sewn, they've harrowed him IN␤Threw clouds upon his HEAD␤And these three men made a …»
masak Timbus: aye. 20:25
Timbus cool
Mouq Timbus: lines() is just $*IN.lines AFAIK
masak vaskozl: what Timbus++ just said. some methods (like .slurp) have a subroutine form (like slurp) with some defaul behavior.
Mouq Oh 20:26
masak Mouq: more like $*ARGFILES.lines, but yes.
Mouq Oh
20:26 Rotwang joined
masak default* 20:26
Mouq didn't realize there was a slurp sub 20:27
FROGGS most ppl expect that for basic file IO me thinks
masak it's a nice convenience. 20:28
it's like STDIN and STDOUT are "there", in a kind of ambient global presence in the program, even though they're really objects.
vaskozl get gives only the first line for me :/
Mouq Does anyone know it jnthn++'s golf-helper script works on OS X? 20:29
masak vaskozl: yes, get does one line at a time.
vaskozl: if you want the whole file, use slurp.
vaskozl: if you want a list of lines, use lines.
vaskozl masak: can we get the whole input at once?
lets say we have a file with names on each line 20:30
and we want to print all names between the two specific names John and Billy 20:31
Mouq vaskozl: "< masak> vaskozl: if you want the whole file, use slurp." 20:33
vaskozl oh right
yah I just realised
Mouq :)
masak vaskozl: < masak> vaskozl: I think what you want is the IO.slurp method.
vaskozl yah, you might have to tell me 5 times before I get it O_O
vaskozl goes on to fix all his stupidly done scripts up to now 20:34
Mouq m: for lines() { .say if m/^They/ ff m/^And/ }
camelia rakudo-moar 440739: OUTPUT«They've ploughed, they've sewn, they've harrowed him in␤Threw clouds upon his head␤And these three men made a solemn vow␤They've let him lie for a very long time␤Till the rains from heaven did fall␤And little Sir John sprung up his head␤They've…»
lue forgot what ff is supposed to do 20:35
vaskozl is wondering the same
Mouq "flipflop"
It's false until the first thing is true
vaskozl and why do we have a 'Threw'?
Mouq Then it's true until the second thing is true 20:36
masak because it picks all the lines in between.
Mouq And then it's false again
vaskozl ah..
masak *after* the line that matches.
Mouq (it's not the most commonly used operator)
masak m: for lines() { .say if m/^They/ ff^ m/^And/ }
camelia rakudo-moar 440739: OUTPUT«They've ploughed, they've sewn, they've harrowed him in␤Threw clouds upon his head␤They've let him lie for a very long time␤Till the rains from heaven did fall␤They've let him stand till midsummer's day␤Till he looked both pale and wan␤They've …»
vaskozl I thought it was some sort or "or" in the beggining
lue Mouq: seeing ff reminded me somebody needs to create a new table of operators by now :) 20:37
masak ff is not very commonly used.
vaskozl I saw youtu.be/x4Ep0VKa_ZI with Larry Wall the other day. 20:38
was stunned by them vim skills
20:39 benabik left
vaskozl You guys are awesome btw, thx for all the help! 20:41
masak vaskozl: we're happy you stopped by.
vaskozl: keep coming back :> 20:42
vaskozl in #perl I'd get !perlbot rtfm, tutorial, stupid
masak: I never leave :D
lue no problem, we're glad you have a better understanding of object oriented programming now :)
masak hehe, #perl keeps being lovable as always... :)
vaskozl: I'm happy someone asked me to explain OO tonight. put me in a good mood ;)
vaskozl masak: yay!
vaskozl is going to have some weird dreams tonight 20:43
20:43 pippo_ joined
lue masak: IIUC, the kindness here is all thanks to au++, no? 20:43
20:43 pippo_ left 20:44 pippo joined
vaskozl I also love how quickly you solved the -p flag and added multiple flag support like -pe and -ne 20:44
masak lue: dunno about "all", but she gave it a really good start.
wow, -pe has been added now? great! 20:45
vaskozl++
shows how we need newcomers with the priorities in all the right places :P
vaskozl it was literally added 5 minutes after I asked about it 20:46
masak that's a nice amount of minutes.
vaskozl sometimes I think it'd be better if we were using base 12 20:47
btw, does masak++ increase rating or something? 20:48
or is it just a way of saying kudos? 20:49
lue
.oO(great, another karma-check-fest incoming...)
karma lue
vaskozl T_T
lue .karma lue
masak not sure we have a karma bot at the mo.
lue vaskozl: don't worry about it, not a bad thing :) 20:50
Mouq m: rand(42) # www.youtube.com/watch?v=x4Ep0VKa_ZI 20:51
camelia rakudo-moar 440739: OUTPUT«===SORRY!=== Error while compiling /tmp/DUu5GJMo4b␤Unsupported use of rand(N); in Perl 6 please use N.rand or (1..N).pick␤at /tmp/DUu5GJMo4b:1␤------> rand⏏(42) # www.youtube.com/watch?v=x␤»
Mouq \o/
20:53 rurban joined
vendethiel <awesome people>++ 20:53
20:55 ajr_ left 20:57 rurban left 20:59 kurahaupo left
vaskozl where can I read more about ms/'<plaintext>' <( .*? )> '</plaintext>'/ syntaxness? 21:00
vendethiel S05, probably :)
lue vaskozl: there's also doc.perl6.org/language/regexes, which might be easier to read than S05 21:01
avuserow colomon: I just updated Audio::Taglib::Simple to use proxies. I agree that it seems like a more appropriate way to handle editing in this case. Let me know if you have any other comments or feedback on it. I'd also love to hear about anything built with it
vaskozl S05 being? 21:02
vendethiel lue: definitely !
vaskozl: the Synopse 05 21:03
Mouq vaskozl: perlcabal.org/syn/S05.html
vaskozl I'm trying something like ms/'keyword .*? <plaintext>' <( .*? )> '</plaintext>'/
but I just get false 21:04
Mouq: thx!
Mouq vaskozl: You have the .*? in the quotes
21:04 rurban joined
Mouq The quotes actually make the .*? a literal string that the regex matches againts 21:04
8st
bluh
s/nts/nst/
vaskozl argh ok!
so qutoes need to be escaped now! 21:05
Mouq vaskozl: Yup!
vaskozl: "Alphanumeric characters and the underscore _ are literal matches. All other characters must either be escaped with a backslash (for example \: to match a colon), or included in quotes"
vaskozl: doc.perl6.org/language/regexes#Literals 21:06
vaskozl Mouq: thx again!
why am I getting weird squares before and after the output?
Mouq vaskozl: Perl 6 regexes take some adjusting from PCRE, but it's definitely worth it 21:07
lue m: say uniname($_) for <｢ ｣>
camelia rakudo-moar 440739: OUTPUT«HALFWIDTH LEFT CORNER BRACKET␤HALFWIDTH RIGHT CORNER BRACKET␤»
vaskozl what are they caused by? 21:08
lue vaskozl: Rakudo uses those corner brackets to surround the output, because they're the shortcut for Q[...] 21:09
(like how '' is the shortcut for q[...], and "" the shortcut for qq[...])
Q is basically a raw string: no backlashes or other fancy tricks work. 21:11
vaskozl won't even pretend he understand
21:11 bjz joined
vaskozl I don't get these with a simple perl -e 'say "hello"' 21:11
why are these annoying things there now?
FROGGS vaskozl: say in P6 does an output for human eyes 21:12
lue r: my $foo; say "Double quotes: \\ $foo \""; say 'Single quotes: \\ $foo \''; say ｢\\ $foo \｣ # note I didn't bother with a second ｣, to avoid a compiler error
camelia rakudo-{jvm,moar} 440739: OUTPUT«use of uninitialized value of type Any in string context␤Double quotes: \ "␤Single quotes: \ $foo '␤\\ $foo \␤»
..rakudo-parrot 440739: OUTPUT«use of uninitialized value of type Any in string context in block at /tmp/tmpfile:1␤␤Double quotes: \ "␤Single quotes: \ $foo '␤\\ $foo \␤»
lue r: my $foo = 42; say "Double quotes: \\ $foo \""; say 'Single quotes: \\ $foo \''; say ｢\\ $foo \｣ # note I didn't bother with a second ｣, to avoid a compiler error
FROGGS if you what an output that machines understand user either print, or say ~$x
camelia rakudo-{parrot,jvm,moar} 440739: OUTPUT«Double quotes: \ 42 "␤Single quotes: \ $foo '␤\\ $foo \␤»
Mouq vaskozl: You don't get them with perl6 -e 'say "hello"' either 21:13
vaskozl I imagine the squares I'm seeing aren't supposed to be squares are they?
Mouq vaskozl: We output them so it's clear that it's a match
vaskozl: No, they're not :)
FROGGS true, corner brackets do only appear around match objets
objects*
lue vaskozl: they should look like corners, top left and bottom right respectively. 21:14
Mouq vaskozl: www.fileformat.info/info/unicode/ch.../index.htm
vaskozl: www.fileformat.info/info/unicode/ch.../index.htm
Mouq is very distractable today
21:14 Rotwang left
Mouq ran the steeple chase at a meet yesterday 21:15
it was a lot of fun and one the best physical challenges I've had in a while
21:17 vaskozl left
lue Heh, I notice doc.perl6.org hasn't a page on quotes, perhaps I should write one in a bit :) 21:17
Mouq lue: doc.perl6.org/language/terms 21:18
Oh, I never gen'd a new doc.perl6.org
But please add to what we've started on it :)
There, I updated it somewhat… 21:19
lue I almost think quoting construct could/should be a separate page (we got a separate page (at least kinda) for that Regex section, and I should think a sublang like Q deserves its own page :P) 21:20
Mouq lue: Then do it :)
lue Perhaps later, I'm busy doing something else at the moment :) 21:24
Mouq lue: Ok :)
21:24 vaskozl joined 21:25 benabik joined 21:37 raiph joined, benabik left
dalek c: fe92708 | Mouq++ | / (2 files):
Use the TITLE for "Language Documentation", if possible

  moritz++ ( irclog.perlgeek.de/perl6/2014-04-26#i_8641084 )		 21:38
Mouq Oh crap 21:39
The links now point to the wrong place -_-
21:39 kl joined 21:40 bjz left 21:42 grondilu left, hoverboard left
raiph .ask jnthn why do you have P5 code in your "reactive programming in perl 6" presentation ("while (!eof($fh)) { my $line = <$fh>; next if $line =~ /^\#/; # … }")? 21:43
yoleaux raiph: I'll pass your message to jnthn.
21:49 dmol left
FROGGS raiph: sometimes it is good to start with something what the audience knows, and then take them bit by bit to the thing you want to show 21:57
21:59 kl left
avuserow is DESTROY still a thing? I can only find two passing mentions of it in the specs, and one test that isn't in spectest.data 22:00
arnsholt Destructors are tricky 22:01
avuserow I don't need prompt destruction, just for it to happen eventually...
arnsholt With refcounting you get timely destruction and, more crucially, a known order of destruction 22:02
With many (most?) other GC algorithms, you can't even make guarantees for the order in which objects are destructed
masak I don't see the enthusiasm around DESTROY that would ever make it a useful feature, even if it were implemented. 22:03
arnsholt Which makes DESTROY harder to implement (IIRC), and also a bit dangerous
Cf. the thing in Java where a finalizer can revive objects if you screw up
FROGGS .tell daxim I have information about ß.uc... if you are interested please either read your privmsg or ping me 22:04
yoleaux FROGGS: I'll pass your message to daxim.
timotimo Mouq: what is a steeple chase? 22:05
avuserow yeah, I understand that destructors are difficult in this way. I guess I have to bubble up the free requirement from the native library to module users
I noticed that after reading ~1100 files, this library stops working if I don't free its pointers 22:06
FROGGS timotimo: hürdenlauf 22:07
timotimo ah
vendethiel :o)
FROGGS yay, $dayjob done, now I am tired :/ 22:08
Mouq timotimo, FROGGS: not quite hürdenlauf, more like en.wikipedia.org/wiki/3000_metres_steeplechase (though this was only a 2000m steeplechase) 22:09
FROGGS Mouq: that is hürdenlaufig enough for me :o) 22:10
timotimo aye
FROGGS well, for somebody sitting on a couch anyway 22:11
timotimo :D
you do your steeple chase while you're sitting on a couch?
that's a pretty amazing feat
Mouq FROGGS: Haha. The difference (other than distance) is that the barriers are shaped so that they can be stepped on and can't be knocked over, and there's a big ol' pit of water to jump off a barrier into 22:13
(The pit is deeper further in, so you try to jump as far out as possible. Jumping over completely is possible, but not really advised) 22:14
22:14 molaf joined, zakharyas joined
timotimo so combined running and swimming? :) 22:15
Mouq timotimo: Thankfully not :) The pit is maybe 1.5+ m long 22:17
Oh 22:18
FROGGS 1.5m+ is enough for my wife to swim in there :o)
Mouq "The water jump consists of a barrier followed by a pit of water with a landing area 3.66 metres (12.0 ft) wide × 0.70 metres (2.3 ft). It then slopes upward from 700 millimetres (28 in) deep to level with the surface of the track."
Mouq is only off by a factor of 2
grfx.cstv.com/schools/niu/graphics/...050909.jpg 22:20
FROGGS yeah, I don't even know how that is called here... 22:22
gnight all
Mouq 'gnight FROGGS :) FROGGS++ for the loop labels stuff, btw, I was super excited to see work on that 22:23
timotimo aye!
22:23 vaskozl left
raiph .tell jnthn I guess what I mean is that the PDF should at least mention that that first fragment is P5 code (though I'd prefer that the whole PDF was P6 code) so it works for folk who don't know or care about P5 22:26
yoleaux raiph: I'll pass your message to jnthn.
masak 'night, #perl6
raiph night
Mouq night masak!
dalek c: ea7d720 | Mouq++ | / (2 files):
Fix url creation for language documentation
22:28 zakharyas left 22:29 vaskozl joined
dalek c: 4cadbce | Mouq++ | htmlify.pl:
*Actually* fix url creation for language documentation
 22:39
Mouq needs to make htmlify do caching so that it doesn't take 13 minutes to verify a simple commit 22:40
s/Mouq/Someone/
:P
22:42 colomon left 22:58 colomon joined
dalek kudo/nom: 46b38f7 | Mouq++ | src/core/Routine.pm:
Make sure Routine.perl makes an EVALable string

E.g. (sub {1+1}).perl would produce something like
  'sub() { #`(Sub|4345502344) ... }', which will parse sub() as a
function. This fixes Routine.perl to be more like
  'sub () { #`(Sub|4345502344) ... }'.		 23:02
avuserow I think I'm going to do a non-NativeCall module for my next project 23:03
timotimo have an idea yet? :)
lue imagines an output for Routine.perl that reconstructs code from the AST, and unintentionally becomes a "this is how a REAL computer writes this sub"
* a "this is ..." method 23:04
avuserow I have several
many of them are from the most-wanted list
I also would find it useful to have some sort of simple single-file cache for stuff like cronjobs 23:05
in one of my P5 projects, I wanted to be able to keep track of state from the last 5 or so days 23:06
so something that makes that easy is high on my list
either that, or bindings to Last.FM's API :) 23:07
timotimo www.youtube.com/watch?v=HOfll06X16c - oh fun!
23:11 rom1504 joined
Mouq ++avuserow 23:13
timotimo that's how bugs get discovered :)
i'm fond
not necessarily fond of anything in particular. just fond. 23:14
avuserow oh or I could patch this:
m: say Buf.new eq ""
camelia rakudo-moar 440739: OUTPUT«(timeout)»
timotimo oh! i recall i wanted to have a discussion with people with more knowledge about this topic
avuserow well I guess the other impls have a "better" infinite recursion message 23:15
p: say Buf.new eq ""
timotimo whether to introduce a candidate for eq for the Buf types on the lhs or rhs or whether to drop the does Stringy from class Buf
camelia rakudo-parrot 440739: OUTPUT«maximum recursion depth exceeded␤current instr.: 'print_exception' pc 146395 (src/gen/p-CORE.setting.pir:61588) (gen/parrot/CORE.setting:11337)␤called from Sub 'Stringy' pc 224364 (src/gen/p-CORE.setting.pir:92270) (gen/parrot/CORE.setting:1044)␤ca…»
timotimo yeah, moar doesn't have a finite stack size ;)
23:15 spider-mario left
timotimo er 23:15
it doesn't have a limit to the exact number of stack frames
avuserow whoa, infinite stack? that's impressive :D
:P
timotimo like a turing machine, but different 23:16
Mouq timotimo: That video is really cool 23:23
lue timotimo: if you ask me, Buf shouldn't do Stringy, but others may disagree :) 23:25
23:26 hoverboard joined, btyler joined
timotimo lue: you're definitely more "in the know" when it comes to all the kinds of stringy things 23:26
but i do recall there was quite an argument last time :)
lue Heh :) My opinion boils down to the fact that Buf can handle more than Stringy (namely, non-textual data), and Perl 6 doesn't tend to define subset relations of that nature 23:28
r: say Int ~~ Rat # math would say True 23:29
timotimo that seems like a good point of reference
camelia rakudo-{parrot,jvm,moar} 440739: OUTPUT«False␤»
23:29 BenGoldberg joined
timotimo avuserow: how about you prepare code and tests in a pull request (well, either a pull request to rakudo and a pull request to roast, or a pull request to rakudo and a gist that shows the problem/solution) 23:30
23:30 vaskozl left
timotimo since pull requests are a good way to get a discussion going 23:30
avuserow sure
lue timotimo: What doesn't help the discussion is that, if you look at Rakudo, Stringy is literally nothing at the moment.
timotimo and the discussion is closer to the code (at least compared to irc)
lue So 'Buf does Stringy' serves only to cause things like that infinite recursion issue :) 23:31
(The biggest issue I see with detaching Buf and Stringy is that ~& etc. make sense for buffers, and not strings; yet the ~ impies stringiness, and I can't think of a good "this is a buffer" symbol in its place right now) 23:33
timotimo m: say "\c[SNOWMAN]" 23:34
camelia rakudo-moar 46b38f: OUTPUT«☃␤»
timotimo always the correct answer 23:35
BenGoldberg If you do $bufone ^ $buftwo ... what would be expected to happen?
23:35 mtj_ left
lue BenGoldberg: a Junction appears :P 23:35
in seriousness, S03 talks about it. 23:36
BenGoldberg (or |, or &)
23:36 vaskozl joined
BenGoldberg bonks self on the head 23:36
Ok, how about +|, +&, +^, then?
lue (And always say that ~^ etc. on C<Str>s is a likely design error, one that wouldn't exist if Buf didn't do Stringy *wink*)
BenGoldberg ponders %|, %&, %^ 23:37
Naah
timotimo that seems weird 23:38
timotimo looks at his keyboard
lue BenGoldberg: for ~&, for example: " does numeric bitwise AND on corresponding integers of the two buffers, logically padding the shorter buffer with 0 values." ... "This operator is distinguished from numeric bitwise AND in order to provide bit vectors that extend on the right rather than the left (and always do unsigned extension)."
timotimo how about Γ?
BenGoldberg .u Γ
yoleaux U+0393 GREEK CAPITAL LETTER GAMMA [Lu] (Γ)
lue timotimo: the way I see it, the acceptable buffer symbol would either be ` or something outside the ASCII range, with some sort of Texas equivalent then needed.
timotimo oh, ` i don't like 23:39
lue (looking at my plain ol' US keyboard at least :P)
timotimo it tends to hide
Ξ is a pretty letter :P
lue timotimo: yeah, but AFAICT ` is the only thing Perl 6 doesn't do anything with. Never said it was a good pick :)
timotimo hehe.
BenGoldberg How about we allow the user to use some sort of compile-time pragma, which says that, until the end of the scope, ^|& are all Buf ops, math ops, or junction ops. 23:40
timotimo yikes
i'm going to bed for today
lue should write a blog post on his opinions before too long, so the inefficiencies of IRC for long explanations don't cause problems again
avuserow m: say [1, 3, 5] +| [2, 4, 6]
camelia rakudo-moar 46b38f: OUTPUT«3␤»
timotimo gnite and good luck :) o/
avuserow oh right
lue ♞ timotimo o/
avuserow o/ timotimo 23:41
BenGoldberg Good night :)
vendethiel m: sub prefix:<Γ>(Bool $n) { return !$n }; sub infix:<Ξ>($a, $b){ $a == $b }; say Γ5 Ξ False; 23:42
camelia rakudo-moar 46b38f: OUTPUT«===SORRY!=== Error while compiling /tmp/DZjK1OZU2A␤Undeclared routines:␤ Γ5 used at line 1␤ Ξ used at line 1␤␤»
avuserow m: say [1, 3, 5] <<+|>> [2, 4, 6]
camelia rakudo-moar 46b38f: OUTPUT«3 7 7␤»
vendethiel m: sub prefix:<Γ>(Bool $n) { return !$n }; sub infix:<Ξ> is looser(&infix:<Γ>)($a, $b){ $a == $b }; say Γ5 Ξ False;
camelia rakudo-moar 46b38f: OUTPUT«===SORRY!=== Error while compiling /tmp/27JAzgYrXN␤Missing block␤at /tmp/27JAzgYrXN:1␤------> n }; sub infix:<Ξ> is looser(&infix:<Γ>)⏏($a, $b){ $a == $b }; say Γ5 Ξ False;␤ expecting any of:␤ …»
vendethiel meh
lue I don't think Γ is a good choice, one might confuse it with the function so named (not likely, but still) 23:43
(besides, I'd prefer a symbol over a letter if at all possible)
vendethiel doesn't Γ usually mean "not"?
or is it the middle one ?
bcode that'd be ¬ :)
lue vendethiel: Γ is the greek capital letter Gamma, no relation to that negation operator :)
avuserow m: say Buf.new(1, 3, 5) <<+|>> Buf.new(2, 4, 6)
camelia rakudo-moar 46b38f: OUTPUT«3␤»
vendethiel bcode: thanks !
.u ¬ 23:44
yoleaux U+00AC NOT SIGN [Sm] (¬)
lue m: say $_ for Buf.new(1,3,5)
camelia rakudo-moar 46b38f: OUTPUT«Buf:0x<01 03 05>␤»
lue m: say $_ for Buf.new(1,3,5)[0..*]
camelia rakudo-moar 46b38f: OUTPUT«1␤3␤5␤»
avuserow I have a few wishlist items for bufs (but who doesn't have some area they'd like to change)
23:45 pippo left
lue heh, typing "buffer" into codepoints.net gave me this: ⎌ :) 23:46
vendethiel .u ⎌
yoleaux U+238C UNDO SYMBOL [So] (⎌)
23:46 btyler left
vendethiel 's bad eyesight makes him a unable to see most of these unicode shinies 23:46
bcode for me it's the font size, in this case 23:47
lue same here, a lot of Unicode characters aren't always easily discernible at my usual font sizes.
23:48 btyler joined 23:51 raiph left

Powered by Cro and the Raku® Programming Language.

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

Thank you!