»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! | Rakudo Star Released!
Set by diakopter on 6 September 2010.
dalek odel: 0d13a6c | diakopter++ | / (4 files):
[dotnet] implement coercive lt, le, gt, ge; add tests to 10-cmp.t
 00:06
ast: 91b1991 | KodiB++ | S12-construction/construction.t:
[construction.t] Added tests for an unreported bug fixed in Rakudo 037775eb.
ast: b540d54 | KodiB++ | S (3 files):
Removed some Rakudo TODOs for passing tests.
kudo: 037775e | KodiB++ | src/builtins/ (2 files):
Tweaked Parcel.item.

This fixes the regression in trans.t introduced by 5ce8fcf.
diakopter moritz_: we could add a 6model backend now 00:07
now that basic control flow & comparisons & iterations work
moritz_: for p6eval, I mean
lue y o hai o/ 01:00
sjohnson HI 01:02
ggoebel sorear: faster to compile and run sounds practical 01:23
cxreg github.com/lopnor/p6-plackdo 05:11
jasonmay whoa 05:19
cxreg any hints on making rakudo build against parrot.git? 06:05
Parrot revision r49711 required (currently r1)
cxreg short-circuits the check 06:12
sorear good * #perl6 06:40
cxreg: moritz_ is working on that
dukeleto cxreg: that is because we don't have a build system for HLL's yet that talks git fluently 07:01
cxreg: but moritz had a way to comment out some code and get it to work. obviously we need a better solution 07:02
cxreg: basically, disable that check, if you want to test Rakudo on the latest parrot
moritz_ once trac.parrot.org/parrot/ticket/1845 is resolved, I know how to fix rakudo.
dukeleto moritz_: we currently have 2 scripts in tools/dev, create_language.pl and mk_language_shell.pl that generate the code that needs to be fixed 07:04
moritz_: there is already a ticket for them, i just remembered. It is like a battlefield, with much bloodshed.
cxreg dukeleto: that's what i did
dukeleto cxreg++
moritz_: trac.parrot.org/parrot/ticket/1491 07:05
dukeleto moritz_: those tickets can almost be merged. If 1491 were fixed, 1845 wouldn't exist 07:06
dukeleto moritz_: but it doesn't matter. we just need to get it to work 07:07
moritz_ doesn't really see the relation
1491 is a pre-git ticket
dukeleto moritz_: yes. 07:08
moritz_: the problem is we are maintaining 2 almost identical things that generate the code that 1845 needs to convert to git 07:09
moritz_: what is the simplest possible thing that can work for Rakudo?
moritz_: perhaps we should just try to get Rakudo building from parrot.git, and worry about the general case of those scripts after 07:10
moritz_ dukeleto: building from parrot.git is not a problem at all. Identifying too old revisions is.
dukeleto moritz_: perhaps we implement purely checking sha1's? and then describe strings after?
moritz_: ah yes. which is the much more *interesting* issue
moritz_: but you have to patch Rakudo right now to build on parrot.git, and that isn't optimal 07:11
moritz_ dukeleto: -1 to sha1s. It needs a parrot git repo to work
it's not just not optimal. It's wrong.
dukeleto moritz_: i thought Rakudo devs usually worked against parrot git repos, to test the latest revs. Am I wrong about that?
moritz_ It's a regression.
dukeleto moritz_: can a regression exist in between developer releases? 07:12
moritz_: in the sense of "define regression"
moritz_: moving to git was more important to us that fixing those scripts, and we didn't have the resources to do everything at once. 07:13
moritz_ dukeleto: the point is that rakudo's Configure.pl checks for an existing parrot, and if the one that exists is too old, and the user asked for it, it can download parrot from git
moritz_ doesn't need those scripts fixed.
dukeleto moritz_: awesome!
moritz_ just needs a mostly linear revision number
dukeleto moritz_: so you are only experiencing the "too old rev" problem, gotcha. 07:13
moritz_ right
dukeleto moritz_: if i provided you with a "git_describe" parrot_config key, would that give you mostly what you need? 07:14
moritz_ right
dukeleto moritz_: that is very doable.
moritz_ just like I wrote in TT #1845
dukeleto moritz_: i hear ya. I was trying to make life more complicated than it already is. I am quite good at that. 07:15
zellio hrmm 07:33
moritz_ hi 07:33
zellio sorry about that, client troubles 07:35
moritz_ -> $work 07:38
zellio upon further reflection my question is better for #perl 07:39
cheers all
dalek ecza/moo: 0123456 | sorear++ | Hi:
xyz
 07:41
sorear sorear--
dalek ecza: 4045e2e | sorear++ | v6/tryfile:
[v6/STD] get heredocs working
 07:44
ecza: 9d97d25 | sorear++ | / (2 files):
Implement LTM for role parameters
ecza: 13cdb07 | sorear++ | v6/ (2 files):
Fix heredoc tweaking
sorear IT WORKS. 07:45
dukeleto sorear++ 07:57
dalek : 6574f39 | sorear++ | misc/dalek-conf.json:
Remove all perl6/ and sorear/ from poll list
 08:00
sorear everyone who can is now encouraged to migrate to push notifications 08:01
they're much faster, use less resources, don't require a commit bit, and don't require you to list branches manually 08:02
dalek announced that mu change less than 5 seconds after the push
the post-commit URL to use is {
host04.appflux.net:5000/dalek?t=freenode,perl6
if you want to hit >1 channel, use t=freenode,perl6&magnet,parrot or so 08:03
er
t=freenode,perl6&t=magnet,parrot
moritz_ sorear: could you please put those instruction in a file somewhere? 08:05
sorear Where
moritz_ mu repo if you have no better idea
sorear hmm, my web service is already getting hammered with requests for favicon.ico 08:06
moritz_ :-)
sorear perhaps this wasn't such a good idea :p
moritz_ if you have a range github IPs, you might filter by IP 08:07
have you switched the repos under perl6/ to push? 08:08
afk
JimmyZ mu: review: github.comhttps://gith[…]bae763f41050bd98c the link is wrong 08:09
sorear moritz_: yes 08:11
dalek : 89628a7 | sorear++ | misc/dalek-push.txt:
[misc] Add preliminary documentation for dalek push
dalek lek-poller: 320e165 | sorear++ | push.psgi:
Try a better multichannel way (2)
 08:19
lek-poller: ebba952 | sorear++ | push.psgi:
Try a better multichannel way
lek-poller: 51c6b9b | sorear++ | push.psgi:
Fix link generation for push mode JimmyZ++
lek-poller: a74c570 | sorear++ | push.psgi:
Try a better multichannel way (3)
 08:20
: c13a8ee | sorear++ | misc/dalek-push.txt:
[misc] Document working multichannel way
sorear github.com/perl6/mu/blob/master/mi...k-push.txt 08:34
moritz_ sorear++ 08:58
mavrc Hello. 09:09
sorear hi
dalek ecza: 4f676f7 | sorear++ | test (2 files):
Add a test for dynamic method names
 09:16
ecza: 2c4b771 | sorear++ | v6/ (2 files):
[v6] Add Stash.iterator
diakopter sorear: that's super-cool 09:51
masak oh hai, #perl6! 09:58
moritz_ \o masak 10:00
tadzik o/ 10:10
masak urgh, Gerd Pokorra suggests on parrot-dev that they simulate SVN commit numbers with a git commit hook. :( 10:39
sounds like SVN withdrawal syndrome to me.
flussence aargh 10:41
moritz_ wonders how branch merges will affect that number
flussence if they want serial numbers, use the date! 10:42
masak or that number in describe.
flussence that too, though it's (intentionally?) harder to get at
moritz_ git-describe is much more robust 10:43
if you take an old branch, and make one commit, you have a very new date 10:44
(that was also a possible problem with svn branches)
masak nom & 10:46
dd-sphere moin 11:06
moritz_ blog.mathgladiator.com/2010/11/3-pr...ed-to.html 11:09
nice idea
nero__ Hello, there :) 12:46
tadzik hello nero__ 12:47
nero__ I there something similar to qw on perl6
moritz_ sure
nero__ like @x=qw(1, 2,3)
moritz_ rakudo: say <a b c d>.join('|')
p6eval rakudo 037775: OUTPUT«a|b|c|d␤»
moritz_ it's even shorter than in p5
also qw works, but not with () (because round parenthesis nearly always mean subroutine call) 12:48
rakudo: say qw/a b c d/.join('|')
nero__ aha that's why rakudo was complain
p6eval rakudo 037775: OUTPUT«Could not find sub &qw␤ in main program body at line 22:/tmp/KqlGan_oP6␤»
moritz_ uhm, it seems that qw isn't implemented :(
nero__ I got the same error message
tadzik who cares, <> is funnier anyway :)
moritz_ alpha: say qw/a b c d/.join('|')
p6eval alpha : OUTPUT«a|b|c|d␤»
moritz_ it's a regression 12:49
moritz_ submits bug report
nero__ thanks a lot moritz, BTW the book is awesome 12:50
tadzik moritz_: is it something that can be fixed messing with grammar only?
moritz_ nero__: you're welcome, and thanks. If there's anything unclear about the book, don't hesitate to tell us/me, so that it can be improved further
nero__ moritz_, sure 12:51
moritz_ tadzik: nearly, yes
moritz_ it's probably just a call to quote_EXPR, and then an action rule that just says make $<quote_EXPR>.ast; 12:52
tadzik: see the rules for circumfix:sym<ang> and quote:sym<q> for example 12:53
tadzik there is q, qq, qx, qqx. I'll see if I can hack something out 12:56
moritz_ you'll need <quote_EXPR: ':q', ':w'>
tadzik funny, I was just looking for something to code 12:58
where does STD.pm6 live? 13:00
neroxx msg/ p6eval 13:02
tadzik qw/1 2 3/.join('|')
Method 'returns' not found for invocant of class 'Undef'
ah, I'll probably need some Actions too 13:03
moritz_ right 13:04
just make $<quote_EXPR>.ast should be enough 13:05
tadzik yeah, just looked at ang
yay, I'm fixing stuff 13:06
> qw/1 2 3/.join('|') 13:09
1|2|3
moritz_ \o/
spectest, submit patch, profit
masak tadzik++ 13:10
tadzik right, spectests, I just wanted to ask about that.
I need to uncomment stuff or something?
moritz_ S02-literals/quoting.t 13:11
175:#?rakudo skip 'qw'
tadzik that's the spectest repo?
moritz_ right
I just did
cd t/spec/
ack -w qw
tadzik I even have a commit bit there 13:12
will run the tests, send the patch and aquire glory and fame
moritz_ right; but it makes no sense to push your patch to the spectest repo before the rakudo patch is applied 13:13
tadzik yea. Can I run the spectests w/o having to pull from GH? 13:14
moritz_ yes
perl t/harness --fudge --keep-exit-code --icu=1 --jobs --tests-from-file=t/spectest.data
moritz_ or --icu=0 if you don't have ICU installed 13:15
tadzik that will take a moment 13:16
moritz_ you can also use --long=0 to skip some long running tests
(that's what 'make quicktest' does)
tadzik it starts from S02 anyway, so it's not a problem. Also, I'm not in such a hurry, I'd rather make sure I didn't break anything 13:17
moritz_ will be your 5th rakudo patch 13:18
we should start discussing giving you a commit bit
tadzik t/spec/S02-literals/quoting.rakudo ............................. ok
I can just create my fork, like KodiB did 13:19
moritz_ well, it still requires a committer to review and merge your stuff, so in the long run it's less work for us to give you access 13:20
tadzik I'd still feel better with my code being reviewed, at least for now 13:22
moritz_ that's ok 13:23
that's why I said 'start discussing', not 'give you one right away' :-)
Kodi masak: With respect to rt.perl.org/rt3//Public/Bug/Display...?id=78896: I disallowed Duration ** Int because Duration ** 2 is mathematically the same as Duration * Duration, which the spec forbids (S02:1343). Am I misunderstanding the spec?
tadzik sure :) 13:24
masak Kodi: I don't know, but personally I hope that we'll tend towards allowing things in this case.
Kodi masak: I suppose that if I do allow Duration ** Int, nobody will complain, and it's following at least one interpretation of S02. 13:25
masak aye.
flussence I'm sure someone will have a reason to want it, but relative units of measurement raised to powers don't make sense to me... 13:29
moritz_ flussence: you might want to calculate a geometric average over durations, for some reason 13:30
masak flussence: I wanted to calculate a standard deviation of a set of values that turned out to be Duration objects. standard deviation involves squaring.
flussence oh, yeah. :)
masak my general point is that it's OK to go around disallowing things and putting up idealized boundaries between things, unless you actually plan to use those things for, you know, reall stuff. 13:33
s/reall/real/
moritz_ so are there some operations that actually make sense to disallow on Duration objects? 13:34
if there are none, I'd suggest to make it a subclass of a numeric type, instead of a has-a relation
Kodi How would one intrepret sin(Duration)?
moritz_ Kodi: in terms of a power series
Kodi (It actually is supposed to be a numeric type, not a has-a.) 13:35
masak I'm not even sure we need a Duration type, so I don't think I'm the right person to ask.
flussence well, Duration * Duration is nonsense, even if the .Int values are sensible...
masak what's the horrible use case we're trying to prevent here?
people *mistakenly* multiplying Durations?
Kodi My understanding is that the Real Point of having a Duration type is to keep the spec from forcing implementations to represent durations with a particular numeric type.
masak when does that happen?
moritz_ flussence: didn't we just give good examples for no-nonsense Duration * Duration usages?
Kodi: then we can just spec it ~~ Real 13:36
Kodi moritz_: That's what we've got. I don't know the whole philosophy behind the operator restrictions.
masak I'm officially representing the rebel group who wants to remove Duration and just put a Num or Rat in its place, representing number of seconds. 13:37
I think that'd be better in all ways.
flussence ++ to that.
Kodi I think FatRat would be preferable to either.
moritz_ what about Instant?
masak instant is fine.
it feels less like a number to me in the first place. 13:38
flussence
.oO( things would be a lot simpler if someone went and measured the exact age of the universe, so we could all use absolute units... )
 13:39
Kodi Looks like TimToady wrote the restrictions on Duration arithmetic. Perhaps we should ask him what his thinking was.
moritz_ flussence: no way. You forgot about relativity
flussence then we'd also need to know how far from the centre we are, and ... Whatever :( 13:41
moritz_ flussence: we're in the centre, of course
flussence: the universe expanded, so the whole universe is what used to be the centre 13:42
masak flussence: my reflex is still to scowl when people start half-joking about Temporal. it's the only time I've kicked three people from this channel, when I was re-writing the Temporal spec, and people were goofing off about it.
tadzik . o O (a Star metamodule: just a deps.proto with Star modules)
Kodi Time is Serious Business. 13:43
masak here: irclog.perlgeek.de/perl6/2010-04-04#i_2194094 13:44
moritz_ notices that the logs don't record who was kicked 13:45
masak it was arnsholt, lue and slavik1.
moritz_ yes, I got that from the context 13:46
but it's a good sign: it's such a rare operation here that I never optimized the logs for it :-)
masak :) 13:47
moritz_ ordered a new laptop - and it actually got cheaper between my decision and ordering it 13:48
tadzik nice. Which one? 13:49
masak someone points out that Scala's startup time equals Rakudo's: twitter.com/__gfx__/status/2672047780012032
moritz_ tadzik: a HP 625 Business 13:49
tadzik this gravar looks like a Mouse author
* gravatar
moritz_ (quite low-end actually, because a new bike is also forthcoming) 13:50
tadzik All tests pass, besides config.rakudo of course
oh crap, I didn't save that test file... 13:51
moritz_ you can rerun it with running the whole spectest 13:52
tadzik that's what I'm doing 13:53
hmm, quoting.rakudo is tested, and quoting.t is not 13:54
moritz_ as soon as fudge markers are present, rakudo runs the .rakudo file instead of .t
tadzik so nothing actually got tested? I don't get it 13:55
moritz_ it just took the .t file, ran it through fudge, and ran the resulting .rakudo file 13:56
tadzik oh, ok 13:57
looks like it passes, but I'll wait for the whole spectest
tadzik a'right, tests pass with the new shiny qw//. Preparing a patch 14:18
dalek ast: bfcf5ed | (Kodi Arfer)++ | S02-builtin_data_types/instants-and-durations.t:
Allow Duration ** Int.
 14:19
flussence on the Scala subject, Rakudo's startup time is more noticeable because people start it up more often :) 14:19
tadzik :> 14:20
moritz_ why?
flussence the JVM is complete overkill for shell one-liners :D
flussence (someone will likely hit me in the face with a counterexample in the next 3 seconds...) 14:21
masak I wish Java had a default REPL. Beanshell sucks.
tadzik how do I reply to a perlbug? Replying to perl6-compiler@ won't work, will it?
masak tadzik: not in the sense of adding something to RT, no. 14:22
moritz_ just hitting reply-to in your mail client should work
afaict RT monitors perl6-compiler
masak oh?
dalek kudo: d9e2246 | KodiB++ | src/core/Duration.pm:
[core/Duration] Allow Duration ** Real.
 14:23
tadzik Mail sent. 14:24
moritz_ patch applied. 14:55
tadzik moritz_++ 14:56
moritz_ tadzik++
tadzik shall I enable the tests in roast?
moritz_ please do
even moar karma for you :-)
tadzik :) 14:57
done 14:58
dalek ast: c3e7f27 | (Tadeusz Sośnierz)++ | S02-literals/quoting.t:
Enabled qw tests (implemented in Rakudo 3aa3c)
tadzik now why isn't it tadzik?
moritz_ because you're missing in CREDITS 14:59
please submit another patch for that :-)
tadzik where do CREDITS live?
moritz_ in rakudo
dalek kudo: 9b94551 | tadzik++ | src/Perl6/ (2 files):
Implemented qw

Signed-off-by: Moritz Lenz [email@hidden.address]
tadzik duh, this prepare-send-apply cycle again :) 15:00
moritz_ I did talk about a commit bit, didn't I? :-)
tadzik yeah :)
moritz_ I can add you to CREDITS if you wish 15:01
tadzik and get the karma? No way :)
gist.github.com/672606 15:02
OTOH, it wouldn't look so... what is that word
moritz_ honest? 15:03
tadzik heh, I meant the negative term. Describing "look, I'm so awesome I'm adding myself to CREDITS"
moritz_ tadzik: fwiw, I prefer present tense in commit mesages ("Add myself" instead of "Added myself"), but that's only a minor nit
tadzik will keep that in mind 15:04
moritz_ tadzik: I know what you mean. But CREDIT entries actually do help with keeping an overview, and since you did contribute stuff, you're not an imposter :-)
dalek kudo: d859644 | tadzik++ | CREDITS:
Added myself to CREDITS

Signed-off-by: Moritz Lenz [email@hidden.address]		 15:05
moritz_ modesty is so imprinted in us that we exaggerate a bit from time to time
tadzik alright :) Btw, isn't that Impostor?
moritz_ probably
tadzik or it could be in-poster :)
alright, I'm out for today. See you! 15:08
moritz_ ciao
masak tadzik: I think the word is "boastful". and adding oneself to the CREDITS file is not that. 15:10
masak is still chuckling at moritz_++' "9 other types" punchline 17:26
colomon_ ;) 17:27
masak I have a soft spot for derivative-joke punchlines.
masak nom & 17:41
patrickas o/ 17:41
dukeleto So a two functions walk into a bar... 17:43
patrickas dang I missed masak by a few seconds ! 17:44
dukeleto The bartender says "Watch out!" There is a derivative operator on the loose! He is going around differentiating everything he sees!
The consant function ran and hid in the bathroom, because he didn't want to become the zero function 17:45
neroxx rakudo: my @x=<foo bar bazar>.join('-'); say @x;
p6eval rakudo d85964: OUTPUT«foo-bar-bazar␤»
dukeleto The derivative operator busts thru the front door and says "I differentiate you!", but the other function says "Whatever, I am e^x, i don't care"
neroxx rakudo: my @x=<foo bar bazar>; for @x -> $n {say $n.join('-');} 17:46
p6eval rakudo d85964: OUTPUT«foo␤bar␤bazar␤»
dukeleto And then the derivative operator says "Yeah, but I am df/dy!".
dukeleto stops.
neroxx is join is not working while on loop ? or I'm doing something wrong ? 17:47
patrickas neroxx the loop is taking each element
and doing a join on it
neroxx but the output don't have the (-) join 17:48
patrickas rakudo: my $n='foo'; say $n.join('-');
p6eval rakudo d85964: OUTPUT«foo␤»
neroxx the square character is my - ? 17:49
patrickas neroxx no
join will put - between two elements, you only have one element
the square charater is a newline that your font probably does not support 17:50
neroxx ah I see, so this should be like this my @x<foo bar bazar>.join('-')
rakudo: my @x= <foo bar bazar>.join('-'); say @x; 17:51
p6eval rakudo d85964: OUTPUT«foo-bar-bazar␤»
patrickas yes 17:52
neroxx where I can found documentation for p6 core modules ? 18:02
patrickas by p6 core modules do you mean modules.perl6.org/ ? If so n the github page of each module. 18:07
moritz_ next time somebody asks about core modules, point them to perlgeek.de/blog-en/perl-6/how-core-is-core.html 19:09
dukeleto maybe we need to start talking about onion modules, since onions don't have cores, and the Perl community is a fractal onion, according to TimToady 19:11
direction ... so perl6 is ready to use just not adopted yet? 19:47
lue ohayo o/
direction mornin
according to the status page, the only problem i might run into is Threading 19:48
PerlJam direction: ready to use for some purposes, not for others. 19:49
direction anything noteworthy that its not ready for besides threading? 19:50
is there a timeline / schedule for perl6 completion?
PerlJam There's no schedule for completion because that concept doesn't make enough sense 19:51
For instance, is Perl 5 "complete"? (no, it's not)
direction ok i mean widespread release
and widespread adoption 19:52
PerlJam Um ... it's available for "widespread release" right now :-)
direction yea thats why i said the latter
PerlJam Adoption we can't control as that's up to the users.
simon I like the "use v6;" convention. 19:53
PerlJam A more useful question might be "When will a Perl 6 implementation be developed to a level that's comparable to other languages like Python, PHP, Ruby, Haskell, etc.?" Unfortunately, I really don't have an answer to that question either. 19:56
PerlJam I suppose the answer is: when there are enough people using Perl 6 that we're able to flesh out all of the details of the spec and foibles of implementation. 19:59
takadonet hey all 20:10
direction hi
direction peace out (^_~) 20:11
sorear good * #perl6 20:12
takadonet sorear: yo
sorear moritz_: the CREDITS file that matters is the one in the Parrot repository 20:18
neither tadzik nor KodiB have CREDITS lines 20:20
masak ahoy! 20:20
takadonet masak: hey!
masak great to see you all. again.
frettled masak: o/ 20:21
takadonet masak: were you gone long?
masak nah, just for nom.
frettled hours!
masak potato pancakes! mmm.
masak (with lingonberry jam, sour cabbage and Turkish yoghurt) 20:29
takadonet nice 20:31
masak and super-simple to make.
lue hai masak o/ 20:42
masak lue! \o/
dalek ecs: ad8e423 | TimToady++ | S05-regex.pod:
[S05] now use :Foo and :Foo<Bar> for uniprops

The old isFoo notation was fragile and could not easily extend to Unicode properties with values. We'll use pair notation instead. The match of a pair value is via smartmatch, so both string and numeric comparisons are available, for instance.		 20:43
lue did you catch my pull request on Tardis?
sorear oooh! 20:44
PerlJam TimToady is really making use of that colon.
takadonet PerlJam: i think he likes it
masak lue: hm, haven't seen a pull request. just the diff you sent along, which I commented on. 20:49
sorear has rakudo/* been moved to push yet? 20:50
masak the pair syntax seems to work for uniprops. in some sense it's a bit arbitrary, but on the other hand it seems to map well onto what it's describing. 20:51
diakopter .beats
phenny @910
diakopter oh; whatevs
sorear masak: %hash ~~ :key<value> = %hash<key> ~~ value 20:52
masak: it's sort of like the characters are being interpreted as their property hash
masak I'll consider that a strangely consistent rationale :) 20:53
lue considers asking what uniprops are and how the isFoo notation was used, but fears the "look at the specs" answer 20:55
masak lue: that's nothing to fear. when you get a "look at the spec" answer, it's because we want to help you reach enlightenment by the quickest way possible, and the spec is the best start on that road. 20:57
huf you mean you'd like to set him on fire, right?
PerlJam huf: no, we'd like him to set himself on fire. 20:58
repeatedly.
huf oh, even better!
i guess that's the free software methodology 20:59
patrickas re o/
lue I considered asking, but then I realized the most likely answer and stopped myself (but apparently that didn't stop me from talking about what I almost did. Hmm)
masak lue: en.wikipedia.org/wiki/Apophasis 21:01
patrickas masak ready to be distracted ? 21:02
masak patrickas: bring it on! :)
lue ololiblog: rdstar.wordpress.com/2010/11/11/ya...nd-tardis/ 21:21
masak reads 21:22
dalek lek-poller: 95a85a7 | sorear++ | push.psgi:
Fix handling of defined-or with arrays

  // puts the left side into scalar context, so @{foo} // () is wrong.
Changed it to @{ foo // [] }.		 21:22
lek-poller: efec941 | bacek++ | push.psgi:
Rename IGNORE into OK and always return it.
lek-poller: 95e4263 | bacek++ | push.psgi:
Reindent code.
lek-poller: fd5ea6b | bacek++ | push.psgi:
Simplify (and speed-up) creating of @files list.
lek-poller: 92fab84 | bacek++ | push.psgi:
Use space after C<if>
lek-poller: ac797df | bacek++ | push.psgi:
Use constant to return constant response.
PerlJam reads too
masak lue: you're right in that Yapsi has a much lower threshold for starting to contribute. 21:23
dalek lek-poller: ee0aebc | sorear++ | push.psgi:
Fix backward message output

GitHub-- for wrong documentation. :/		 21:24
masak I think the main reason for that is that it sits at ~500 lines of code.
PerlJam masak: and fewer conceptual hurdles to leap 21:25
masak right. fewer everything, really. features, layers of abstraction... 21:26
PerlJam Well, I meant that to hack Rakudo, you have to know all sorts of things ... Perl 6, PIR, C, how these things interact, where the dragons in the existing code are, etc. 21:28
masak nod.
patrickas PerlJam: unless you are lucky to find things where you don't need all that .. then working on rakudo feels just like working on Yapsi (from my very minor working on rakudo experience) 21:30
PerlJam I think that's the limiting factor for Rakudo contributors actually. Even if you start off hacking on a "pure Perl" problem, you can quickly run into dragons without realizing it. 21:31
(until it's too late :)
sorear You think hacking Perl 5 might be easier? It's 100% C 21:32
masak to Rakudo's credit, it's becoming easier to hack with every rewrite :)
PerlJam masak: indeed
lue I don't think I didn't try, but Rakudo is just so complex, and there's nothing really easy to do (even the LHF, in my experience, requires you know *something* about the codebase) 21:33
Or, perhaps the real problem is not that the codebase is too complex, but that there's no Hitchhiker's Guide for it. 21:36
PerlJam lue: perhaps 21:37
masak lue: I feel your pain. I would hack more on Rakudo as well if I found it more approachable. 21:38
it's not so much that I can't do it, as that it's not a good short-term investment of my time.
PerlJam masak++ that's *exactly* how I feel.
masak PerlJam: and I'm just dressing what lue wrote in my own words :) 21:39
PerlJam Well, I like your words :)
masak \o/
sorear lue: help me to write a Hitchhiker's Guide to the niecza codebase
masak lue: yes! become a niecza contributor! 21:40
moritz_ back 21:42
lue That's a good idea. I've always thought of a Perl 6 documentation repository.
PerlJam lue: I wouldn't help sorear. That was an almost godwinian way he brought up Perl 5 earlier ;-) 21:44
Though, if there were a HHGTN, that would increase the odds that "occasional hackers" like myself participate. 21:46
masak same here.
I *want* to participate in niecza. but I don't see a good way in. some documentation would help. 21:47
sorear I think I could write one better if I knew what people were looking for
lue I should really start on my POD parser to go with this vision of a document repo. [Believe me, I've planned this whole documentation thing more than I'm letting on.] Now that I think about it, documentation is the most pleasing to me, seeing as it isn't there. 21:49
lue &afk; # nom time, afterwards I'll start fleshing out these ideas more. 21:50
sidek are there any tutorials for perl 6? I found one but it was half in German. 22:01
and, although I know basic german, I know BASIC german
PerlJam sidek: there are all sorts of resources on perl6.org 22:02
sidek thanks
moritz_ chances are the tutorial you found was from me (or lichtkind)
sidek it was on perlfoundation and looked like it was in the process of being translated 22:03
PerlJam lichtkind
sidek alright
moritz_ mine are on perl-6.de
takadonet sidek: You can ask questions here any time
moritz_ though I haven't been writing about Perl 6 in German for a year or so 22:04
sidek thanks for the offer, takadonet. I know quite a bit of perl 5, though, so hopefully I'll be fine
moritz_ the audience is just too small
dalek ecza: 2e15054 | sorear++ | docs/compiler.pod:
Add some overview of the compiler IRs
PerlJam moritz_: just the two of you? :) 22:05
moritz_ PerlJam: no, more than that, but still not enough
takadonet sidek: Looks like we are in the same boat 22:06
direction jumps on that boat 22:07
PerlJam Perl 6 does make more sense if you shift your thinking slightly away from perl-5-think and slightly towards haskell-think 22:08
takadonet PerlJam: never wrote a single line of Haskell... yet
PerlJam s/haskell/lisp-or-scheme-but-without-the-parens/ then :) 22:09
direction happens to be writing LISP as we speak
PerlJam takadonet: I didn't even really know haskell existed until pugs came along. Now it's my favorite of the FP languages. 22:10
direction: you have my condolences :)
direction eh its not so bad after you write a few thousand lines 22:11
sorear ugh, LISP 22:12
direction i write it in C-style then it gets converted to LISP style
sorear who still uses LISP?
direction Texas Instruments
PerlJam sidek: all those emacs fans
er, sorear
sidek I use lisp :( 22:13
and haskell
moritz_ lots of people use lisp. Lots of people use C. Lot's of people use Perl.
masak wants to use Lisp more
sorear Lisp is one thing
LISP is something else entirely
sidek well, I use scheme
but meh
it's a LISP 22:14
PerlJam masak: there's something deeply wrong with you
sorear elisp is closer to LISP than Lisp; I'm trying to remember where it ranks
masak PerlJam: I've learned to accept that. you should, too.
moritz_ is fascinated by people with whom something's deeply wrong 22:19
Tene I'd also like to use lisp more. I had a lot of fun writing a scheme compiler for parrot. 22:45
simon hehe 22:46
diakopter moritz_: then you're fascinated by everybody ;) 23:16
masak diakopter: that's the most dystopian I've ever seen someone add a winking smiley to. 23:17
diakopter dystopian, or theologically mainstream globally 23:19
masak that is true.
I like how in ancient China, there essentially was one philosophical school arguing that people were fundamentally good-natured, and one school arguing that people were fundamentally rotten. 23:20
and that nurture and society pulled people in the respective other direction. 23:21
diakopter too bad the respective camps' definitions of "good" and "rotten" probably didn't align very perfectly :D 23:22
let alone "nurture/society" vs "fundamentally"
masak I'm not sure I'm relaying this thousand-years-ago piece of history with total fidelity, mind. 23:23
sorear closer to 2300 23:24
masak I can't even get the dates right :P 23:25
yes, 300 BCE.
blog post! strangelyconsistent.org/blog/novemb...2010-avast
diakopter I wonder when the Uncommon Era will begin
dalek ecza: 4c09c28 | sorear++ | / (5 files):
Implement computed method names in generic roles
 23:26
masak the "two languages, same family" meme spotted in a Japanese blog: www.publickey1.jp/blog/09/perl5perl6perl2.html 23:29
dalek d: fc30a21 | TimToady++ | STD.pm6:
[STD] add colonpair as valid cclass_elem

In line with recent S05 change, <:foo['bar']> and such are now allowed as valid character class designators, along with <alpha> and <[0-9]>. The colon forms are always taken as Unicode properties.		 23:39
sorear TimToady++ 23:43
diakopter TimToady: what about variable properties ... <:foo[$scalr]> 23:44
sorear diakopter: <:foo($scalr)> 23:46
it's just a [ :lang(%*LANG<MAIN>) <colonpair> ] 23:47
sorear STD 20101111 pushed to CPAN 23:57

Powered by Cro and the Raku® Programming Language.

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

Thank you!