»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.
hoelzro any compiler wizards around? 01:18
I'm working on the S26 stuff, one of my tokens is being parsed twice, and I can't figure out why 01:19
dalek rl6-bench: bbca251 | (Geoffrey Broadwell)++ | analyze:
Show missing scores in text history
 01:58
sergot o/ 06:19
dalek rl6-bench: 73cf7ef | (Geoffrey Broadwell)++ | bench:
Minor cleanup at top of bench
 06:41
rl6-bench: d80285c | (Geoffrey Broadwell)++ | bench:
Factor known-tags() out of `bench list-tags`
rl6-bench: 207e457 | (Geoffrey Broadwell)++ | bench:
Add `bench quickstart` command

Performs all the steps necessary to do a default benchmark of all recent perl5, nqp-moar, and rakudo-moar tags (plus nqp-moar/master and rakudo-moar/nom), showing the resulting performance history summary.
Not fully tested yet.
japhb And with that, sleep& 06:42
Ven .lines doesn't take windows's \r into account, and I've had to remove them myself several times. Should it take care of that ? 09:06
moritz yes
though the exact mode how that should happen isn't clear to me
maybe \cr\lf should become one grapheme that matches \n 09:07
Ven that does raise a ETOOMAGICAL for me 09:08
Ven Type check failed for return value; expected 'Bool' but got 'Junction' <- looks like new rakudo broke my game of life :( 09:31
moritz seems like you shouldn't return a Junction when you declared that you return a bool :-) 09:32
maybe 'return so $expression_with_junction'
Ven I don't return a Junction, though. 09:33
0 <= $y <= $.dim && 0 <= $x <= $.dim && @.grid[$y][$x] // False; <- how can this return a function ? Except if @.grid ... contains one.
moritz Ven: just assign it to a variable, and say it's .^name
then you know.
Ven m: my $a = do given 3 { when 3 { 1 == 1 | 2 } }; say $a.perl
camelia rakudo-moar 736027: OUTPUT«any(Bool::True, Bool::False)␤»
Ven WAT.
m: my $a = so do given 3 { when 3 { 1 == 1 | 2 } }; say $a.perl 09:34
camelia rakudo-moar 736027: OUTPUT«Bool::True␤»
moritz doesn't find it WAT
you write a Junction, it returns a Junction.
if you don't want a junction, use || instead of |
or so, yes 09:35
Ven no, I write `1 == 1 | 2`, not `1 | 2`. But eh.
jnthn Operations involving a Junction produce a Junction, which flattens away in a boolean context. 09:37
betterworld m: my $a = do given 3 { when 3 { 1 == 1 | 2 } }; say ?$a.perl
camelia rakudo-moar 736027: OUTPUT«True␤»
Ven Yeah, but apparently it used to automatically flatten or something.
Ven if I do `my Bool @array`, should it auto flatten ? 09:40
jnthn No. 09:41
moritz that's a type check declaration, not a type coercion declaration 09:43
hoelzro morning #perl6 11:39
I'm having an issue with my work on S26; my #= comment token seems to be matched *twice* by Grammar.pm, and I can't figure out why 11:44
here's a patch, script, and sample output file: gist.github.com/hoelzro/6f356615290a55802f0e
zengargoyle junctions are starting to remind me of Monads, once you put something in one, you stuck with it forever... 11:52
cognominal zengargoyle, unlike monads junctions have a very specific purpose. And for monads, idris addresses the issue ias explained in eb.host.cs.st-andrews.ac.uk/drafts/...torial.pdf Not sure it can tricklle down to haskell though. 12:00
hoelzro here's my crappy workaround for that: github.com/hoelzro/rakudo/commit/6...de24ef27c3 12:15
dalek Heuristic branch merge: pushed 21 commits to roast/S26-WHY by hoelzro 12:18
carlin ohh didn't know dalek could do that 12:19
hoelzro pats dalek on the dome
FROGGS_ hoelzro: I guess there is some <?comment> somewhere? 12:31
FROGGS_ hoelzro: so that the look ahead is done and the match itself 12:31
(could be a <!before <comment> > also)
arnsholt The joys of producing your thesis in LaTeX: I've just spent my whole day to write ~130 lines of TikZ code to produce a page with line drawings on it 12:32
arnsholt Yay, productivity! \o/ 12:32
FROGGS_ \o/ 12:32
*g*
[Coke] someone who uses latex! have you checked out simon cozen's SILE? 12:36
[Coke] RT now has 784 Rakudo tickets. 12:39
nwc10 Perl 5 only has 1400 - you're more than halfway to maturity :-) 12:40
[Coke] m: say eval "3" 12:41
camelia rakudo-moar 736027: OUTPUT«===SORRY!=== Error while compiling /tmp/ROwfXHV9Fy␤Undeclared routine:␤ eval used at line 1␤␤»
lizmat m: say EVAL "3"
camelia rakudo-moar 736027: OUTPUT«3␤»
lizmat m: say EVAL 3
camelia rakudo-moar 736027: OUTPUT«3␤»
lizmat m: say EVAL "'3'" 12:42
camelia rakudo-moar 736027: OUTPUT«3␤»
lizmat m: say (EVAL "'3'").WHAT
camelia rakudo-moar 736027: OUTPUT«(Str)␤»
[Coke] lizmat++: there was a ticket about that, just double checking before I closed it.
lizmat m: say (EVAL "3").WHAT
camelia rakudo-moar 736027: OUTPUT«(Int)␤»
arnsholt [Coke]: Linky? 12:45
[Coke] arnsholt: to what? 12:46
arnsholt Oh sorry, SILE 12:47
Always interested in new LaTeX stuff =)
[Coke] github.com/simoncozens/sile
"SILE is a text formatting engine. It's sort of like TeX, except it isn't 35 years old, doesn't use idiosyncratic font and output formats, and it isn't written in Pascal." 12:48
arnsholt Sounds interesting, but that'd also have the drawback of not having all of my fun TeX stuff =) 12:49
One of the reasons TeX is so darn resilient is that it's actually a complete programming language
carlin
.oO( Perl 6 is a programmig language. It's sort of like Perl 5, except it isn't 20 years old... )
 12:51
[Coke] given that simon had the first commit in the parrot repo, that should not surprise you. :)
I looked at it briefly, but it seems almost like html2pdf to me, based on how I'd use it. 12:52
lizmat Question: should we ever allow loading of a pre-compiled compunit if the associated source file is not there (anymore) ? 13:03
if we consider the pre-comp file to be a cache, to be refreshed when needed, a pre-comp binary file only makes that impossible 13:04
on the other hand, in the case of a .jar: there may not be a source file available for rakudo to use to create the pre-comp binary 13:05
jnthn Note that pre-comp binaries are tied to a particular Rakudo build also. 13:06
FROGGS_ not having the ability to rebuild a .moarvm file does not imply disallowing to use that .moarvm file when it is still valid
jnthn Folks certainly may wish to distribute a binary build, but I think they use case is going to be shipping a complete package including a matching Rakudo. 13:06
*the
FROGGS_ true 13:07
jnthn Which is fine, really; disk is cheap.
lizmat so binary only is ok? 13:13
in the context of a complete package and source obfuscation, I guess :-)
FROGGS_ yes 13:14
timotimo source obfuscation is a brilliant idea!
FROGGS_ also, the source could be somewhere else at runtime... so as long the cache is okay we should use it
timotimo we should ship a tool that strips annotations like line numbers out of moarvm bytecode %)
jnthn lizmat: Perhaps "binary is OK" depends on the compunitrepo in question.
lizmat: But to a first approximation, I'd say allow it. 13:15
lizmat okidoki.. just checking :-)
hoelzro ah ha 13:39
I think you're right, FROGGS_
FROGGS_++
FROGGS_: extra points if you can help me think of a better solution than my existing one =) 13:40
hoelzro isn't sure if you can tell if a token was matched via lookahead or otherwise
hmmm...do action methods get called for lookaheads?
FROGGS_ no 13:41
action methods are called for every matching token
hoelzro ah ha
timotimo is there a way to capture the result of a lookahead anyway?
hoelzro so maybe I should be doing my work there...
FROGGS_ timotimo: no
timotimo: maybe in an ugly way... 13:42
hoelzro: yeah, that is what actions are for, basically :o)
timotimo heh. put a { $external_var = $/.clone } in there or something
(eew)
hoelzro =)
FROGGS_ eww, yeah :o)
lizmat NRpm meeting& 13:55
hoelzro masak: ping 14:33
dalek kudo-star-daily: 2cec0c6 | coke++ | log/ (14 files):
today (automated commit)
 14:37
atroxaper m: say { return False; }() 14:46
camelia ( no output )
atroxaper m: say { False; }()
camelia rakudo-moar 736027: OUTPUT«False␤»
masak hoelzro: pong 14:57
hoelzro masak: is ufo obsolete? it's broken with the latest perl6-m 14:58
masak I don't think it's obsolete as in "there's no concievable use for it". 14:59
I haven't used it myself much.
masak sorry to hear it's broken on m. 14:59
maybe I can help?
masak clones it and tries to run it
hoelzro masak: I actually have an outstanding PR to fix it =) 15:00
masak hm, getting a deprecation warning...
hoelzro I haven't been keeping up with the module ecosystem stuff in P6 land much, so I'm wondering if there's something newer, flashier, and better available =)
dalek o: c188859 | (Rob Hoelz)++ | bin/ufo:
Remove deprecated usage of $*VM
 15:01
o: 37479e4 | (Rob Hoelz)++ | bin/ufo:
Coerce %*CUSTOM_LIB paths to Strs
o: f950bac | (Carl Mäsak)++ | bin/ufo:
Merge pull request #14 from hoelzro/master

Fix ufo to work with latest Mokudo (2014-06-11)
masak hoelzro: merged.
hoelzro masak: thanks!
masak thank *you*.
dalek o: f88b3bd | Kamil++ | README:
Removed repos that changed into alien panda
 15:02
o: 9653e94 | (Carl Mäsak)++ | README:
Merge pull request #15 from teodozjan/readme

Removed repos that changed into alien panda
o: 5ab0f4b | (Kamil Kułaga)++ | README:
Removed repos that simply disintegrated
 15:09
masak handled all other pull requests for ufo while I was on it. 15:10
colomon hoelzro: I think most people are using panda instead of ufo these days... 15:12
hoelzro ah, I see 15:13
I haven't looked at panda in a while
teodozjan colomon: Compilin big project multiple times with panda hurts :) 15:15
colomon teodozjan: not sure what you mean by that, but I'm sure that panda can be improved...
teodozjan colomon: when using ufo there is incremental compilation, afaik panda always starts over 15:17
colomon has to admit he hardly ever compiles at all during module development…. then uses panda when things are pretty stable. 15:18
teodozjan was abducted by ufo from world when people have habit of compiling projects 15:20
s/when/where :) 15:21
carlin panda is optimized to increase slacking off time
ufo's alien tech was just too efficient
teodozjan panda understands corporation needs ;) 15:22
hoelzro so I just discovered that perl6advent.wordpress.com/2011/12/1...ng-perl-6/ uses #= as leading comments 16:03
should that be corrected?
are advent posts edited after the fact if the language changes in ways that invalidate them?
flussence I think it's okay, as long as you leave a footnote saying it was changed 16:05
masak yes, sometimes they are edited.
depends on the original author, I think.
japhb hoelzro: What's the cycle for #= deprecation? I end up using it a lot ... 16:10
hoelzro japhb: I have no idea =P
I've just been working to make rakudo follow the spec
the fix is pretty simple; just replace #= with #|
japhb In nom, or another branch? 16:11
hoelzro it's in S26-WHY under github.com/hoelzro/rakudo
hoelzro it *could* be in this month's release 16:11
japhb Yeah, just a lot of stuff to update, a fair amount of it shared -- so I have to make sure I update the code and users' perl6 at the same time.
hoelzro but I might hold off having it merged until 2014.08 16:12
if you look at roast, there's a file called S26-TODO under the S26-WHY branch
japhb If you decide to hold off, I would put a deprecation warning in 2014.07's changelog
hoelzro which illustrates how much is left to be done =)
sounds like a good idea to me
my program that makes use of this stuff won't be ready until after 2014.08 is out anyway 16:13
who's the release manager this month, anyway?
japhb Hmmm, I don't see one in docs/release_guide.pod 16:15
PerlJam "#= deprecation"? What's that about? 16:19
masak it's an old spec change from October 2011. 16:20
PerlJam So, #= isn't going away, just its meaning changes? 16:21
masak github.com/perl6/specs/commit/7d19...cc39206bd6
it's being split in two, favoring predictability rather than DWIM.
I bet there is quite a bit of interesting discussion in the IRC logs around that time, too. 16:22
I remember Haddock (a similar Haskell doc tool/syntax) being discussed as an influence.
PerlJam Aye, I've been thinking of it as that commit says, but when I saw "#= deprecation", it sounded like it was going away. 16:23
grondilu Hello, maybe the following warning while compiling MoarVM is worth signaling:
src/6model/reprs/CPointer.c: In function ‘set_int’:
src/6model/reprs/CPointer.c:33:17: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast] body->ptr = (void *)value;
(it's the only warning I get so I thought maybe it matters= 16:24
hoelzro PerlJam: I guess it's more of an "existing behavior of #=" deprecation =)
jnthn grondilu: On 32-bit, ooc?
grondilu linux 32 bit indeed. 16:25
jnthn Yeah, it warrants a look. 16:25
grondilu there's a similar warning in exceptions.c: 16:26
src/core/exceptions.c: In function ‘search_frame_handlers’:
src/core/exceptions.c:81:61: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast] MVMuint64 block_label = block_has_label ? (MVMuint64)(f->work[eh.label_reg].o) : 0; 16:27
grondilu wonders if that could be related to the NativeCall issue in tests for 32 bit machines 16:28
hoelzro lunch & 16:33
dalek rl6-roast-data: 12c6a3c | coke++ | / (5 files):
today (automated commit)
 16:58
ast: 9eb26a8 | coke++ | integration/99problems-51-to-60.t:
fudge for rakudo.moar
 16:59
dalek osystem: b74dfb3 | (Michal Jurosz)++ | META.list:
Add IO::Socket::SSL (required by HTTP::UserAgent)
 17:07
rindolf TimToady: hi, are you feeling better? 17:26
mj41 sergot: Hi github.com/sergot/io-socket-ssl/pull/1/files pls 17:30
sergot Oh, I didn't notice that, mj41++ thank you!
mj41: anyway, it's not in the ecosystem yet 17:31
mj41 sergot: it is github.com/perl6/ecosystem/commit/...7c93862490 :-)
trying to install HTTP::UserAgent 17:32
sergot ohh, I wanted to put it there a bit later :)
right :)
mj41++ thanks
sergot mj41: it still has a lot of bugs though 17:36
mj41 sergot: np, you should fix them (it is in the ecosystem already) :-) 17:37
sergot mj41: yes, it's my gsoc project, I will :) 17:38
retupmoca sergot: re IO::Socket::SSL - is it possible to "upgrade" an existing socket connection to ssl/tls ? 17:41
(such as required by the STARTTLS option in smtp, for example)
sergot retupmoca: it is, by using .set-fd, just need to pass the file description of your connection 17:43
or it'll be 17:44
I mean, it is , but I have never tested it, so I cant say how it works :)
retupmoca ah, ok
I assume that takes the C-level fd?
retupmoca I have to figure out how to get that then 17:45
sergot retupmoca: yes 17:46
sergot retupmoca: what are you using to connect? 17:47
retupmoca IO::Socket::INET.new
sergot Yeah, my goal is to get the fd from it, could you let me know if you figure out how to do this? 17:48
It's seems to be easy when we talk about files, but not so easy in case of sockets.
it seems*
FROGGS_ can't we establish an unsecured connection via openssl, and upgrade it afterwards? 17:50
this way we would have an fd for sure
sergot hmm, I'm not aware of this 17:56
I'll find out
mj41 modules.perl6.org is updated, but panda uses feather.perl6.nl:3000/projects.json (not updated yet) 18:03
hmm, feather is up to date now 18:06
raiph hoelzro: could we pretty please have #< and #> (and if at all possible, #<...> too)?
hoelzro raiph: I'm just following the spec =)
what would #<...> do? 18:07
raiph change the spec :)
#<...> would be same as or instead of #`< ... >
hoelzro would it also be a decl comment? 18:08
or just a regular one?
I'm all for #< and #>, but I think it merits discussion
jnthn The reason for the ` is so you end up with a line commenting out the opening of some bracketing construct ending up commenting out the entire construct by accident. 18:08
jnthn uh, so you don't end up with 18:09
sergot mj41: exec: panda update
raiph m: say 3 #`< embedded comment > + <apple orange> 18:11
camelia rakudo-moar 736027: OUTPUT«5␤»
mj41 sergot: I did, but modules.perl6.org is not updated at the same time as the feather.perl6.nl (used by "panda update")
sergot: Successfully installed HTTP::UserAgent 18:12
raiph m: say 3 + <apple #`< embedded comment > orange>
camelia rakudo-moar 736027: OUTPUT«7␤»
mj41 sergot: after "yum install openssl-devel"
sergot mj41++: \o/
raiph sergot++ mj41++ # writing/exercising important modules 18:14
PerlJam m: say (<apple #`< embedded comment > orange>).perl 18:17
camelia rakudo-moar 736027: OUTPUT«("apple", "#`<embedded", "comment>", "orange").list␤»
PerlJam huh. That's weird
sergot raiph: thanks :)
japhb PerlJam: Why weird?
japhb that the spaces inside the inner brackets got squeezed? 18:18
PerlJam yeah
japhb Yeah, that is kinda odd.
PerlJam m: say (<<apple #`< embedded comment > orange>>).perl
camelia rakudo-moar 736027: OUTPUT«("apple", "#`<", "embedded", "comment", ">", "orange")␤»
jnthn m: say Q{{ 1 { 2 } 3 }} 18:19
camelia rakudo-moar 736027: OUTPUT« 1 { 2 } 3 ␤»
jnthn Following same rules as things like that, I believe. 18:20
m: say Q{ 1 { 2 } 3 }
camelia rakudo-moar 736027: OUTPUT« 1 { 2 } 3 ␤»
jnthn That's more clear.
It keeps a count of extra openers it saw
Note the whitespace splitting is a postprocessing step.
sergot mj41: any feedback will be appreciated 18:20
raiph jnthn: I'd appreciate a demo of your earlier point (commenting out by accident) using camelia 18:21
japhb jnthn: I think his point was that with < (as opposed to <<), the inner spaces *disappeared*. 18:23
If the spaces were still there but the splitting hadn't happened, that would be a different issue.
jnthn japhb: That's what I meant by it being a post-processing step with <...>. We parse it like any other string, and then split it afterwards. It's <<...>> that works at a smarter, AST level. 18:33
raiph: It was more a concern with {...} really: if you were commenting out lines of code using #, and just happen to comment out a line starting with a {, it's preferable that it doesn't suddenly change meaning from being a single-line to a bracketed comment. 18:35
` isn't used in code in Perl 6, so it's not a risk with that.
masak circumfix:<` `> has been unofficially marketed as a nice target for slangers. 18:42
[Coke] jnthn: (like any string and split afterwards) - where is that, I have a bugfix I need to put in. :) 18:44
(something about non-break space)
cognominal "unofficially marketed". Perl shines in stealth marketing... except when #perl6 was created in freenode. 18:46
jnthn [Coke]: Probably in Perl6::Actions::Q as some kind of postprocess method.
vendethiel cognominal: what happened then, 18:49
s/,/?
japhb jnthn: One of us is confused. I might have expected ("apple", "#`<", "embedded", "comment", ">", "orange").list and could understand the bug if it returned ("apple", "#`< embedded", "comment >", "orange").list or ("apple", "#`< embedded comment >", "orange").list. But why would the spaces disappear, instead of *either* appearing in the output strings or being used to split on? 18:51
jnthn japhb: Oh...now I see 18:53
OK, that's is weird.
japhb yeah
PerlJam japhb++ for saying what I was thinking 18:56
cognominal vendethiel, people do stats about activities on well know places and conclude that Perl is dead. But most perl activity is secluded on perl.org irc servers.
masak Rakudo compiler release is in two days, and we don't have a release manager signed up :(
PerlJam The only problem with using IRC as a communication medium for work is that sometimes ... you actually have to do some work ;) 18:57
jnthn wonders why it's become so hard to find release managers :( 18:59
masak I could sign up for it, but I'd prefer to delegate it. 19:00
cognominal vendethiel surely, people who relies on whatever metrics fails to see interesting things because Kuhnian paradigm changes make old metrics obsolete.
PerlJam jnthn: people are busy? 19:01
jnthn PerlJam: Maybe the average busy-ness of channel inhabitants has increased in recent months... :) 19:02
masak I agree -- it used to be easier.
I have released Rakudo twice this year. timotimo, tadzik, FROGGS, and sergot have all released it once. 19:03
masak looking back at 2013, we see isBEKaml, Coke, and lizmat as release managers, apart from the ones already mentioned. 19:04
2012: japhb, pmichaud. 19:05
FROGGS_ I think I can do it on Thursday, but I#d prefer August
masak 2011: jdhore, mberends, arnsholt.
FROGGS_ I'd*
masak FROGGS_: [Coke] has August. is September OK? :) 19:06
FROGGS_ September is fine, aye
dalek kudo/nom: 84289aa | (Tobias Leich)++ | docs/release_guide.pod:
sign up for September release
 19:07
masak FROGGS_: dang, you beat me to it ;) 19:07
FROGGS_ :P
masak had to do a `git pull --rebase`, which slowed me down...
FROGGS_ masak: I just edited on the website
masak anyway, FROGGS++
masak FROGGS_: I'm pretty sure I'm faster on the console than on the website, though. less network latency. 19:08
also, vim.
FROGGS_ hehe
all in all it just took about four seconds or so the way I did it
FROGGS_ (the release guide is in my browser history) 19:09
cognominal vendethiel, "We have had some terrible news. Perl is secluded in the west wing". /me paraphrases Big Lebowsky. 19:10
dalek kudo/nom: fe26e31 | duff++ | docs/release_guide.pod:
Sign up for Oct release
 19:11
masak PerlJam++ 19:13
jnthn Yay.
masak ok, so it isn't that people won't do releases. nice. 19:13
it's that they won't do the Thursday release :P
dalek kudo/nom: 38ec697 | (Tobias Leich)++ | docs/release_guide.pod:
sign up for July, who needs spare time anyway
 19:14
masak note: I will do it if no-one else does. but I'd prefer it if someone else did.
ooh!
FROGGS++
FROGGS_: can I grab September, then? 19:15
FROGGS_ in exchange jnthn++ needs to fix the jvm stuff :P
masak: sure, if you like
in fact, I don't mind doing both
dalek kudo/nom: 4a3e4d9 | masak++ | docs/release_guide.pod:
[docs/release_guide] oust FROGGS++ from Septembe
 19:15
lizmat masak, FROGGS_: July is impossible for me because I'll be on a plane to Portland and zonked out on the far end
masak r
masak lizmat: "zonked out on the far end" sounds like a good name for an ambient rock band. 19:16
[Coke] masak, FROGGS_: happy to take sept instead of august, no problem. 19:17
masak either works for me. 19:17
[Coke] (or later, if needs be. it was just teh first one I knew I'd be able to do when I signed up)
masak I'm just happy we now have releasers for all the way to October :> 19:18
jnthn lizmat: Feel free to pick Nov or Dec. ;)
jnthn At this rate we can be done for the rest of the year. :) 19:18
masak \o/
#perl6++
FROGGS_ well, there is also star
jnthn Yes...
FROGGS_ btw, it might make sense to do both star and the compiler release in the same run 19:19
PerlJam FROGGS_: have you done them both like that before?
jnthn I'm not sure that'll help...given Star can involve chasing modules up...
FROGGS_ ewww, I've got a 48cm thingeny on my pillow that stinks >.<
masak "thingeny"? 19:20
FROGGS_ PerlJam: I always do a last star smoke test when doing a compiler release, so I see that there is no point release needed to recover from a regression
[Coke] PerlJam: the 2014-10 release is on the 23, not the 16th.
mind if I fix it?
FROGGS_ masak: yeah, I mean it is cute when it sleeps, but it does not smell nice atm
PerlJam sure, I just left the date that was there. 19:21
[Coke] it's "first thursday after 3rd tuesday", not "third thursday", which has bit me at least once in the past.
PerlJam FROGGS_: I haven't looked, but if you have any pointers about compiler release + star release, it might be useful to add that to the release guide. 19:22
FROGGS_ PerlJam: the star release has its own guide
dalek kudo/nom: fdf0a22 | (Elizabeth Mattijsen)++ | docs/release_guide.pod:
Pick up November
 19:22
FROGGS_ and of course it takes time to do both 19:23
masak m: for 1..12 -> $month { my $d = Date.new(Date.today.year, $month, 1); ++$d until $d.day-of-week == 2; $d += 14; $d += 2; say $d }
camelia rakudo-moar 736027: OUTPUT«2014-01-23␤2014-02-20␤2014-03-20␤2014-04-17␤2014-05-22␤2014-06-19␤2014-07-17␤2014-08-21␤2014-09-18␤2014-10-23␤2014-11-20␤2014-12-18␤»
FROGGS_ and perhaps it is not that easy to do the star with a not yet done compiler release at all...
I'll think about it this Thursday
dalek kudo/nom: c9d3ca9 | (Will Coleda)++ | docs/release_guide.pod:
fix date, add more releases, steal one.
 19:24
[Coke] dammit, the live edit dropped lizmat's change in favor of mine. fixing... 19:25
dalek kudo/nom: b46af89 | (Will Coleda)++ | docs/release_guide.pod:
give lizmat++ her release back

github-- for the live edit not merging in changes to commits that happened since I started editing!		 19:26
masak m: sub first(Date $d, &p) { $d++ until &p($d); $d }; for 1..12 -> $month { say Date.new(Date.today.year, $month, 1).&first(*.day-of-week == 2).later(2, 'weeks').later(2, 'days') }
PerlJam [Coke]++ for illustrating why my suspicion of the live edit thingy is justified :)
camelia rakudo-moar 736027: OUTPUT«Cannot modify an immutable Date␤ in sub postfix:<++> at src/gen/m-CORE.setting:1834␤ in sub first at /tmp/GAPB2Awrr4:1␤ in block at /tmp/GAPB2Awrr4:1␤␤»
masak m: sub first(Date $d is copy, &p) { $d++ until &p($d); $d }; for 1..12 -> $month { say Date.new(Date.today.year, $month, 1).&first(*.day-of-week == 2).later(2, 'weeks').later(2, 'days') }
[Coke] I will volunteer for a star release perhaps when I get mynew laptop with enough memory. :)
camelia rakudo-moar 736027: OUTPUT«Too many positional parameters passed; got 3 but expected 1␤ in method later at src/gen/m-CORE.setting:15692␤ in block at /tmp/DCLLC01ufY:1␤␤»
jnthn masak: I guess it's weeks => 2, or :2weeks 19:29
masak m: sub first(Date $d is copy, &p) { $d++ until &p($d); $d }; for 1..12 -> $month { say Date.new(Date.today.year, $month, 1).&first(*.day-of-week == 2).later(:2weeks).later(:2days) }
camelia rakudo-moar 736027: OUTPUT«2014-01-23␤2014-02-20␤2014-03-20␤2014-04-17␤2014-05-22␤2014-06-19␤2014-07-17␤2014-08-21␤2014-09-18␤2014-10-23␤2014-11-20␤2014-12-18␤»
masak \o/
jnthn
.oO( Did I just tell masak how to use an API he designed? :) )
masak re-re-designed, but yes.
that last shuffle-around was *good*, mind, and left us in a better place than before, with .delta 19:30
...but I'm not yet used to it ;)
[Coke] have people actually be doing step 10 so that spectest runs clean? 19:32
FROGGS_ [Coke]: yes, at least for moar and parrot
[Coke] (given the daily failures I'm seeing in the spec test suite, I imagine not) 19:33
huh. ok.
FROGGS_ moar is (was?) hard because of S17
[Coke] there are dozens of failures in jvm & parrot at the moment.
FROGGS_ :/
itz_ l 19:44
cognominal lizmat++, should'nt there be mention of the MIME RFCs in S22, and a way to declare media type according to tools.ietf.org/html/rfc2046 19:50
Ven can I say that the whatever star will "wrap around the closest expression and turn it into an anonymous function ?" 20:14
colomon "closest"? 20:15
colomon as far as I know, it would be more accurate to say it will turn an operator into an anonymous function most of them time…. 20:16
Ven m: (say *)(3) 20:17
camelia rakudo-moar 84289a: OUTPUT«*␤Cannot find method 'postcircumfix:<( )>'␤ in block at /tmp/FD3GWtDLhN:1␤␤»
Ven mmh, yah 20:18
colomon and the effect runs up through a series of operators (again, most of the time)
Ven m: (3*(5+*))(5) # p sure this works ... 20:50
camelia ( no output )
Ven m: m: 3 # this segfaults on my machine 20:51
camelia ( no output )
cognominal m: multi sub say(Whatever) { sub ($) { say "Whatever! What did you expect?" } }; (say *)(3)
camelia rakudo-moar b46af8: OUTPUT«Whatever! What did you expect?␤»
cognominal Ven: I suppose you could do interesting things with functions that do not support Whatever by catching the exception and recalling the function appropriately to have it your way. But I don't know enough about exceptions to get the original arguments. 20:59
Ven that seems like magic a tad too dark for me 21:00
(and for my tutorial :).)
cognominal magic? that sounds more like pure evil! just one step of a pure EVAL. 21:02
jnthn EVAL *is* pure if the code does no side-effects :P 21:04
cognominal indeed, Perl 6 is strict but usually impure 21:07
carlin hm, `anything:` causes a segfault 21:10
m: ItsAlive: say 'alive'
camelia rakudo-moar b46af8: OUTPUT«alive␤»
carlin aww :(
oh, only on the REPL 21:11
avuserow how hard is making a star release? I may have some tuits and interest in that this month. If not to do it entirely, to lend a hand somehow if that's possible? 21:35
[Coke] IME, it's more time intensive than hard. 21:36
avuserow I've been thinking recently when lurking that I ought to take a release sometime but my timezone does not align with the channel's peak activity very well
[Coke] Check out the release docs in the star repo.
github.com/rakudo/star/blob/master...-guide.pod 21:37
masak ++avuserow
avuserow I don't want to sign up formally since I'm afraid I'll forget or get busy :\ 21:38
(for a main rakudo release I mean)
masak nodnod
all in due time :)
masak 'night, #perl6 22:06
japhb Good night, masak. 22:07

Powered by Cro and the Raku® Programming Language.

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

Thank you!