»ö« | perl6.org/ | nopaste: paste.lisp.org/new/perl6 | evalbot usage: 'perl6: say 3;' or rakudo: / pugs: / std: | irclog: irc.pugscode.org/ | UTF-8 is our friend!
Set by Juerd on 28 August 2009.
00:05 jrtayloriv left, tak11 joined
lisppaste3 diakopter annotated #87203 "yamlized 2" at paste.lisp.org/display/87203#1 00:09
diakopter that's what it looks like without the ::Dumps 00:10
down in the ...
s1n pmichaud: i replied about the scheduling (second sat. after dallas.p6m is this weekend), let me know what you want to do 00:11
diakopter I guess if it's a Double_Double, it should always descend into .nibble 00:12
(afk )
00:13 sri_kraih left
pugs_svn r28265 | lwall++ | [viv] properly hoist EXPR fields, diakopter++ 00:16
00:16 rhr_ joined
dalek kudo: fed7b0e | pmichaud++ | build/Makefile.in:
Add Mapping.pm to setting (missing from b29506b commit).
00:18
00:19 ihrd joined
TimToady diakopter: say(1+2) was just a stupid bug--hoist didn't return it's %r 00:20
so you should have more annotations in the outer node now
00:20 ihrd left
TimToady *its 00:22
00:23 justatheory joined, sri_kraih joined 00:28 rhr left
TimToady diakopter: working on a decent sub nibbler to bypass the AUTOLOAD botch 00:43
00:44 dukeleto joined
diakopter oo cool 00:49
dalek kudo: c52c780 | pmichaud++ | perl6.pir:
Work around inferior loop problem causing segfault on exit.
00:54
00:54 SmokeMachine left 00:56 nihiliad joined 01:01 maja joined, jrtayloriv joined 01:02 felipe left 01:03 sm^^^ left, sm` joined
TimToady dinner & 01:06
01:12 maja_ left 01:22 wolverian left 01:27 wolverian joined
diakopter STD/viv seems ... faster ... with those last commits 01:34
01:36 snarkyboojum_ joined, rhr_ left 01:38 snarkyboojum__ joined 01:40 rhr joined
diakopter vijs: say(say(3)) 01:41
p6eval vijs 28265: OUTPUT«3␤1␤␤ time in interpreter: 0.012802 s␤»
s1n diakopter: what's vijs? 01:52
01:54 snarkyboojum left, snarkyboojum__ is now known as snarkyboojum
s1n diakopter: emits JS from p6? 01:54
diakopter uses Larry's STD parser in Perl 5; emits its AST to JS; interpreter walks the AST
could be a "definitional" interpreter eventually... slow of course without optimization phases.
s1n interesting 01:55
02:01 snarkyboojum_ left 02:03 tann left 02:09 hercynium left 02:14 jauaor left 02:17 jiing left 02:20 jaldhar joined
colomon pmichaud, moritz_: The post office bounced my CLA. Is the Hillsboro, OR 97123 address still correct for the Perl Foundation? 02:22
02:29 am0c joined 02:30 dukeleto left
jrtayloriv I'm trying to understand this rule --> pastebin.com/d61f15c2 ... The first line of the rule says to look for zero or more <command_or_definition>s, but I don't understand what [ $ || <panic: 'Syntax error'> ] saying. It seems from the context that it is somehow saying "if you didn't find any <commmand_or_definition>s, then something is wrong -- exception", but I don't see how it works. 02:39
pugs_svn r28266 | diakopter++ | [vijs] better not-implemented error msg 02:41
arnsholt jrtayloriv: $ means end-of-string
So if it doesn't find the end-of-string marker after the program, it invokes the panic routine (|| is the same as | in p5 regex) 02:42
jrtayloriv arnsholt, What is "end-of-string" marker? How would it not be present? 02:45
02:47 felipe joined
diakopter it means if <command_or_definition>* failed to consume all of the input (and then it looked for the eof but didn't find it), there must be a syntax error 02:47
jrtayloriv diakopter, Oh, I see. So all that [ $ || <panic: 'Syntax error'> ] is getting is leftovers from the <command_or_definition>* match ... and since there shouldn't be leftovers, we die horribly. 02:48
Thanks :)
diakopter yw 02:49
02:53 dukeleto joined
jrtayloriv diakopter, So the point where [ $ || <panic: 'Syntax error'> ] would be matched (and panic() called) is when it reaches the end of the first line of whatever is left of the input after <command_or_definition>* , correct? 02:56
diakopter probably not linewise..?
02:57 stephenlb left
diakopter it would panic at the beginning of whatever input wasn't consumed 02:57
jrtayloriv Oops -- I see. That was just me, being stupid ... nevermind. 03:00
colomon rakudo: say atanh(1); 03:10
p6eval rakudo c52c78: OUTPUT«Divide by zero␤»
03:17 snarkyboojum left
pmichaud colomon: as far as I know that address is the correct one. 03:17
dalek kudo: 2eca8b3 | pmichaud++ | docs/announce/2009-09:
More announcement updates for the release.
03:20
kudo: ea6448f | pmichaud++ | docs/ChangeLog:
More ChangeLog updates.
03:25 meppl joined 03:29 gabiruh_ is now known as gabiruh 03:33 jjore left 03:37 jjore joined 04:10 frew_ left
dukeleto seen jnthn 04:11
04:20 nihiliad left 04:25 tak11 left
dukeleto jnthn: ping! 04:33
04:35 ms111 joined
pugs_svn r28267 | lwall++ | [my.t] missing ; 04:39
r28268 | lwall++ | [viv] fix up nibbler's AST node a bit 04:41
04:48 ms11 left 04:51 justatheory left, rfordinal joined 04:59 cdarroch left
diakopter reprecompiles STD 05:00
05:02 jrtayloriv left
diakopter that'll be perfect for topping the day's irclog. 05:03
05:03 zloyrusskiy joined
pugs_svn r28269 | diakopter++ | [viv] extrude extraneous Dumps 05:03
diakopter er, or not. silly timezones. 05:04
05:12 dalek left, dalek joined
dukeleto diakopter: good localtime() :) 05:13
05:14 awwaiid left 05:16 awwaiid joined, jrtayloriv joined 05:17 jrtayloriv left, jrtayloriv joined 05:21 finanalyst joined
dukeleto cognominal: ping! 05:22
05:27 zamolxes left 05:41 justatheory joined 05:45 zpmorgan joined 05:53 justatheory left 05:59 synth left 06:08 sm^^^ joined 06:09 Aisling left 06:18 sm` left 06:30 blaze-x left 06:32 yves left 06:33 rfordinal left 06:40 orafu left, orafu joined 06:47 am0c left
moritz_ good morning 06:55
dalek kudo: 73a3952 | pmichaud++ | docs/spectest-progress.csv:
spectest-progress.csv update: 436 files, 15497 (71.5% of 21671) pass, 0 fail
06:56
kudo: 8effbb7 | pmichaud++ | docs/announce/2009-09:
Update announcement with spectest results.
06:59 iblechbot joined 07:09 rfordinal joined 07:13 blaze-x joined 07:15 yves joined 07:16 viklund joined
dukeleto moritz_: howdy 07:18
a good morning (for a release) it is. at least in some timezone
moritz_ dukeleto: oh hai
at least for some people ;-) 07:19
(not a very good morning health-wise for me)
dukeleto moritz_: blizkost now passes lots of tests (including most of t/base from the perl 5 test suite) : leto.net/dukeleto.pl/2009/09/blizko...tests.html
moritz_: sorry to hear that. 07:20
moritz_ cool
never mind, will be fine tomorrow again
jnthn dukeleto: omg wow
dukeleto jnthn: there you are :)
jnthn dukeleto: Just invoking the blizkost exe? 07:21
dukeleto jnthn: i can't seem to do "make blizkost", but parrot ./perl5.pir t/foo.t works
jnthn: see github.com/jnthn/blizkost/issues#issue/10
jnthn: as well as "make test" 07:22
jnthn I think the makefile is probably wonky.
moritz_ it is
yesterday when a PMC changed, a 'make' didn't update anything
jnthn aye, I noticed such oddness. 07:23
Patches welcome...
I'll try and fix it in a bit if nobody else does though.
dukeleto moritz_: oh noes. if I can't trust "make", who *will* I trust ?
jnthn It's an annoying bug.
dukeleto moritz_: gmake is also required on FreeBSD but the Configure.pl doesn't know about that 07:24
07:25 donaldh joined
moritz_ the blizkost@exe@ does not depend on any PMCs 07:26
dukeleto that seems not-so-good 07:27
moritz_ only "build" (the default target) depends on $(BLIZKOST_GROUP) 07:28
jnthn ah
That'd perhaps be why.
moritz_ maybe it would be saner to delegate build to blizkost@exe@
and make the latter depend on $(BLIZKOST_GROUP)
dukeleto jnthn: do you have any comments re: github.com/jnthn/blizkost/commit/31...1b041bb938 ? 07:29
jnthn: should we use Test::More in the test suite? Test.pm works
jnthn dukeleto: Hmm, not sure off hand. 07:30
Basically though, things to keep in mind are:
1) Some tests in whatever are better than no tests at all.
2) While tests for integration with other specific HLLs are OK, we probably want to be doing a bunch of tests from PIR, since this is aimed at being a compat layer for Perl 5 / Parrot, not just Perl 5 / Rakudo. 07:31
dukeleto jnthn: understood. Next up, PIR tests. 07:32
jnthn The real hard debugging task ATM is fixing the segfault I mentioned in github.com/jnthn/blizkost/commit/38...6bff439e83 07:33
moritz_ another problem with the Makefile 07:34
clean:
jnthn If that can be fixed, then we'll be starting to be useful.
moritz_ $(MAKE) $(PMC_DIR) clean
shouldn't that be $(MAKE) -C ($PMC_DIR) clean?
is the -C option mostly understood by the different make utilities?
jnthn Not sure - that bit for sure came from the make language shell tool. 07:35
moritz_ what it does currently is nonesense 07:36
pmichaud ugh, make_language shell is evil. :-) 07:37
I much prefer rakudo's approach :)
(which avoids the recursive makefiles)
moritz_ oh, now I understand 07:42
MAKE := make -C
pmichaud that seems evil also :)
moritz_ that explains a lot
pmichaud (which is why I didn't use it for rakudo)
also, the make language shell tool depends on Parrot's configure scripts 07:43
jnthn popping to Osaka for a bit - back later
pmichaud popping to bed for a bit... back later 07:45
07:46 mberends joined
pugs_svn r28270 | lwall++ | [numeric.t] syntax error 07:48
07:52 hanekomu joined
mberends Good morning! Aww - I just missed a useful discussion about make. My 􏿽x800.02 worth is that building a Windows Rakudo with the Strawberry toolchain works if mingw32-make is renamed to make. I tried patching build/gen_parrot.pl to accept alternative names for make, as per [particle]++'s suggestion, but so far those builds still fail. The name 'make' is probably hardcoded somewhere down the line. :-( 07:53
moritz_ mberends: it seems to me that build/gen_parrot.pl uses whatever parrot uses for $make 07:55
mberends: at which point does the build fail? 07:56
07:57 zamolxes joined 07:58 rbaumer joined
moritz_ mberends: what's the entry for make in parrot/config_lib.pasm 08:00
s/\\/?/
mberends moritz_: I'm away from that machine atm, will look again soon. I was trying to work around the parrot/config_lib.pasm:242, which is hardcoded "make"
moritz_ oh.
that's a parrot configure bug then 08:01
mberends it seems we need to consider gmake, nmake and mingw32-make as well
08:01 cmv joined
moritz_ afaict config_lib.pasm is generated by parrot's Configure 08:02
08:02 cmv left
moritz_ the correct solution is to have it set a proper entry for P0["make"] 08:02
mberends: shall we take that to #parrot? 08:03
mberends moritz_: sure, and I'll move to the Windows machine with the problem
08:05 cmv joined 08:11 jauaor joined 08:29 klaus_ joined
klaus_ good morning 08:30
moritz_ hello 08:31
klaus_ the perl6 wiki has problems?
some broken links
moritz_ which wiki? there are several :/ 08:32
klaus_ like: www.perlfoundation.org/static/2.14....x.cgi?smop
moritz_: perlfoundation
08:32 TopoMorto joined
moritz_ klaus_: ok. Where did you find that link? 08:32
klaus_ several wikis, thats bad ;-)
moritz_ aye 08:33
oh, I see
klaus_ www.perlfoundation.org/perl6/index....mentations
moritz_ that's *very* bad
so all internal links are broken
I'll see if I can find somebody to report to 08:34
klaus_ moritz_: thanks
sjohnson hi all 08:35
moritz_ hi 08:36
sjohnson hi moritz_
been away in India for the last week 08:37
missed programming in Perl :]
Matt-W Morning
sjohnson hi matt! 08:39
08:47 TopoMorto left 08:59 zamolxes left
klaus_ moritz_: you may ask Martin Berends 09:01
www.perlfoundation.org/perl6/index....0720163017
moritz_ mberends: feel hilighted 09:03
mberends: did you replace those links manually, or was that some wiki software hiccup?
09:22 zamolxes joined, masak joined
viklund o/ masak 09:22
masak viklund: \o 09:23
viklund: thanks to Tene++, I have pretty good db support in Web.pm now. you could check out Squerl if you want. 09:24
viklund nice ;) 09:25
09:25 ejs joined
viklund Squerl? 09:25
as in squirrel?
masak viklund: it's already sufficiently implemented to start thinking about trying to move November over to it.
viklund: yes, I learned "Squirrel" is one of the pronunciations of "SQL" :)
viklund masak: sounds good, november really needs some love (features)
masak and bugfixes... 09:26
to be honest, I've sort of been holding out for November (the month).
viklund yep, but it's still up, I think this is the longest time w/o crash ever!
masak \o/
stability, here we come!
viklund otoh I don't know if it has been that much used...
masak maybe not. 09:27
viklund last change was 88 days ago by lichtkind...
masak lichtkind++ 09:32
viklund btw masak, have you seen lichtkind++ around lately? 09:34
(or anyone else for that matter)
hmm,
masak not really.
moritz_ 5 days ago in #perlde
viklund thx moritz++
masak FBI finally got to him, huh?
viklund masak, that was mean 09:35
masak sorry :/
viklund (but funny)
masak :)
09:38 dakkar joined
flip913 quit 09:41
sorry
ESC is not / in IRC
masak flip913: you can log out anytime, but you can never leave :)
flip913 :wq doesn't work either
moritz_ masak: lol, I just thought exactly the same 09:42
flip913 masak: yes, thanks ;-)
09:42 flip913 left
moritz_ then try :x ;-) 09:42
mberends asks forgiveness for breaking www.perlfoundation.org/perl6/index.cgi 09:43
moritz_ mberends: just fix it ;-) did you really change all these links manually? 09:44
09:44 rbaumer left
mberends moritz_: that was a page edit via FireFox. I'd better just undo it. 09:44
moritz_ mberends: ok, I thought it was wiki software fail (not quite impossible with socialtext--) 09:45
reqamst: I see you pushed things to your fork of tufte - just wanted to tell you that you have commit privs to my copy, so you can push directly (IMHO it makes things easier, because I don't have to watch the fork queue) 09:47
mberends moritz_: www.perlfoundation.org/perl6/index.cgi reverted 09:48
moritz_ mberends++
09:49 azawawi joined
azawawi hi 09:49
masak o/
azawawi masak: I finally got my November membership card :) 09:50
masak azawawi: great! viklund++
azawawi indeed viklund++
viklund viklund--, it took me 8 days...
moritz_ reqamst: ah, I see you just tested, so never mind 09:51
azawawi viklund: better late than never, viklund++ :)
masak indeed viklund-- :P
lunch &
moritz_ viklund: if you're happy to hand out your github API key I can make hugme add committers to november
hugme: list projects 09:52
hugme moritz_: I know about these projects: json, perl6-examples, proto, svg-matchdumper, svg-plot, temporal-flux-perl6syn, tufte
viklund moritz_: sounds like a good idea
once I had lunch
09:53 jauaor left
azawawi moritz_: could you please add #sdl to your wonderful irc logger? :) 09:53
moritz_ azawawi: which network? perl or freenode?
azawawi moritz_: on perl
moritz_ i think let me double check
moritz_: yup irc.perl.org 09:54
09:54 araujo left
moritz_ viklund: ok, when you're back from lunch you need to /msg me the API token from github.com/account and tell me who should have the powers to add new committers 09:54
09:55 gfldex left
azawawi moritz_: many thanks :) 09:55
moritz_ azawawi: irclog.perlgeek.de/sdl/today at your service - please announce the logs in the channel
09:56 donaldh left
moritz_ (it'll take a day or so to update the index pages, they are heavily cached) 09:56
azawawi moritz_: cool
moritz_++
moritz_: done. 10:00
moritz_: Let the Perl sdl games begin :)
moritz_ ++
10:02 azawawi left 10:11 zamolxes left 10:14 KatrinaTheLamia left 10:15 hanekomu left 10:16 M_o_C joined
moritz_ bah, github seems down 10:18
10:22 rbaumer joined 10:41 riffraff joined, zamolxes joined 10:43 payload left
masak github had DDoS issues yesterday. 10:46
10:49 rfordinal left 10:52 payload joined 10:53 hugme left, hugme joined
moritz_ hugme: show november 10:53
hugme moritz_: the following people have power over 'november': TimToady, [particle], masak, moritz_, pmichaud, viklund. URL: github.com/viklund/november/
masak moritz_++ hugme++ 10:54
moritz_ and viklund++
mikehh rakudo (8effbb7) builds on parrot r41317 - make test PASS / make spectest (up to 28270) FAIL - Ubuntu 9.04 amd64 10:56
rakudo - t/spec/S05-capture/caps.t - Non-zero wait status: 11 - Parse errors: Bad plan. You planned 21 tests but ran 19
rakudo - it passes ./perl6 t/spec/S05-capture/caps.t however - also
rakudo - make t/spec/S05-capture/caps.t fails with the same error (ran both at least 3 times) and note
rakudo (8effbb7) builds on parrot r41312 - make test / make spectest (up to 28270) PASS - Ubuntu 9.04 amd64
moritz_ ah, that's a post-1.6 parrot 10:57
mikehh moritz_: yes
masak another candidate for porting to Perl 6: www.yuml.me/diagram/class/samples 11:01
I'd like both an offline version and an URL-based web service.
moritz_ once you have an offline version, the URL version is not too hard 11:02
I've never been much of an UML fan myself 11:05
but I'd like to have a nice tool for plotting graphs 11:06
'dia' (the gnome project) is quite nice, but somehow the graphs it produces are always ugly
11:11 iblechbot left
masak we talked about over lunch today how there are no sweet-spot FOSS UML makers out there. 11:12
there one we've tried are either ugly or unstable.
moritz_ not surprised 11:13
11:17 donaldh joined
cognominal say eval( 'sub A::t { $_[1]->{a} }; bless {}, A', :lang<perl5> ).t( { :a<ok> } ) 11:19
rakudo: say eval( 'sub A::t { $_[1]->{a} }; bless {}, A', :lang<perl5> ).t( { :a<ok> } )
p6eval rakudo 8effbb: OUTPUT«Method 't' not found for invocant of class 'Nil'␤»
11:21 cdarroch joined 11:29 Mimir joined
Mimir Hello, I don't know much Perl 6 but I'd like to contribute to the development process. How can I help? 11:31
moritz_ Mimir: hello and welcome to Perl 6 11:32
Mimir: what are your skills? what would you like to do?
we need people who write applications and libraries, people who write tests, websites, compilers, documentation, ....
art work, good ideas - there's nothing we can't need ;-) 11:33
Mimir Well, I know Perl 5, Haskell, and some C. How about tests? What's the right way to go about that?
moritz_ perlmonks.org/?node_id=780001 here KyleHa describes his experience with adding tests 11:34
11:34 mikehh left
Mimir Okay, thanks. 11:35
moritz_ if you want to write tests, you need a commit bit for the repository where the tests live
if you tell (or /msg) me your email address, I can give you access
Mimir [email@hidden.address] 11:36
moritz_ Mimir: you should get mail within the next few minutes
Mimir Got it! Very good, very good. 11:37
moritz_ you can obtain the test suite with 'svn checkout svn.pugscode.org/pugs'
11:37 synth joined
moritz_ it's a custom to add your name to the AUTHORS file to test your commit bit 11:37
Mimir Guess I'll have to learn how to use SVN first, heh. 11:39
moritz_ aye, but it's not much that you need most of the time
Mimir Works for me.
11:46 Mimir left 11:50 riffraff left 11:53 hugme left 11:54 hugme joined 11:58 charsbar joined
dalek kudo: 4de8215 | (Martin Berends)++ | src/builtins/globals.pir:
[builtins/globals.pir] for ~/.perl6/lib use $HOMEDRIVE$HOMEPATH if $HOME does
11:58
11:58 charsbar left, charsbar joined 12:00 charsbar_ left 12:01 charsbar left, charsbar joined 12:05 abra joined 12:15 zamolxes_ joined
viklund oops, p6ism in p5: for (<>) 12:20
(technically a hybrid)
moritz_ <> is forbidden in Perl 6 12:21
std: <>
p6eval std 28270: OUTPUT«===SORRY!===␤Obsolete use of <>; in Perl 6 please use lines() or () instead at /tmp/63lP0ucfnu line 1:␤------> <⏏>␤FAILED 00:02 36m␤»
12:22 zamolxes_ left
viklund aye 12:30
eject?
moritz_ eject? what do you mean by that?
viklund or what is that yellow symbol? 12:31
in the std output
moritz_ yes
$ echo ⏏ | utf8-dump
\N{EJECT SYMBOL}
viklund ha
moritz_ very handy ;-)
viklund utf8-dump, where can I get that? 12:32
moritz_ perlgeek.de/blog-en/perl-tips/utf8-dump.html
no rocket science really 12:33
viklund thx
well, you need to know it's there 12:34
usually that is the hard part
moritz_ aye
moritz_ doesn't feel sufficiently productive for coding, but maybe it's enough for blogging 12:35
or maybe I should work on perl6.org instead 12:36
12:38 payload left
masak is blogging 12:39
Matt-W is trying to focus his mind on his afternoon's task
moritz_ I found blogging much easier when you actually did something you can blog about ;-) 12:40
12:41 hanekomu joined 12:44 zamolxes left
Matt-W moritz_: yes, it is :) 12:45
moritz_ ok, I'm working on perl6.org now 12:56
the idea is to make the design of the sub pages so horrible that others can't stand it and chime in with a better design
arnsholt Sounds like an excellent plan =) 12:58
Matt-W lol
I'm trying to make a web page piled high with JavaScript work in IE6... 12:59
moritz_ make a single XMLHttpRequest which returns a rendered version of that page as .jpg ;-)
13:02 iblechbot joined
Matt-W lol 13:04
thing is all the JavaScript is jQuery stuff
so it's kind of supposed to work
13:04 __ash__ joined
Matt-W but I think the google earth plugin is causing mischief with its interfaces 13:04
masak man, the use.perl.org main page is ugly. :(
moritz_ oh yes 13:05
Matt-W all of use.perl.org is ugly
moritz_ we should have a use.perl6.org with camelia design and cheerful colors
13:07 saran joined
Matt-W Hmm 13:07
To be honest, I don't like Camelia
saran Saran is back after studying some perl basics
*perl6 13:08
Matt-W So I'm just going to wince quietly in a corner
saran: welcome back!
13:09 zloyrusskiy left
saran yesterday was discussing with mask and mberends abt working on ".perl" function which simulates Data::Dumper in Perl 5 13:09
diakopter Camelia might like darker chroma
Matt-W Yes I saw that
And it currently doesn't handle circular references very well 13:10
saran matt-w . yes . i need to work to find why it hangs while doing circular references 13:11
Matt-W If it's not handling them at all at the moment, I expect it hangs because it keeps following them
diakopter saran: if you want to see an algorithm that handles reference cycles (and sibling/cousin multiple references) while serializing, see the ToJS.pm I committed to pugs/src/perl6 yesterday 13:12
masak loliblogged: use.perl.org/~masak/journal/39639
13:12 hercynium joined
diakopter it could be optimized a bit by eliminating the stashing of references for items that *aren't* referenced more than once (internally or otherwise) 13:13
optimized for reading, I mean.
masak Matt-W: I started out mildly disliking Camelia.
13:14 saran left
Matt-W masak: I don't mind the concept, I'd just like something that looks a bit more... professional 13:14
Awful though that word is
masak Matt-W: what sold me was watching the hilarious reactions ("childish", "gay", "girly" etc) on Twitter. people clearly need to loosen up.
Matt-W: yes, that was my first reaction, too. not professional enough.
Matt-W Maybe they do, but we're talking about first impressions
diakopter it's professionally intended to be counter-culture, I assume.
masak aye. 13:15
Matt-W And if you're looking at corporate adoption and you say at your workplace let's do this in Perl 65
s/65/6/
diakopter or 65
Matt-W And they look at perl6.org and see something that's decorated like a kindergarten, what are they going to say?
Go away and weep because their grey suits aren't colourful enough?
Maybe we can do colourful, I'm not a designer, I just think it needs *something* 13:16
diakopter "our company can't afford to be associated with such counter-culture; our available hiring field would diminish significantly" 13:17
Tene points out the django pony.
masak: i love the blog post.
masak: have you seen keep? conway.rutgers.edu/~ccshan/wiki/blo...shell/keep 13:19
13:19 saran joined
saran Saran : am sorry..my network hangs out often 13:20
Matt-W diakopter: this is unfortunately how companies think :(
moritz_ well, maybe perl6.com should show larry in a suit rather than camelia ;-)
Matt-W likes Strawberry Perl, maybe we just need a really nice big Camelia
diakopter django pony: ok, the distinction there in manager-think is between a "technology/product" and a "language/platform". Generally, commitments to the latter are seen as much more binding than commitments to instances of the former, regardless of whether the perceived classificiation is "correct" or whether the distinction is even valid to any granularity.
Matt-W I'd say a designer, but preferably not the ones who did my employer's new look 13:21
diakopter syntaxes and compilers are still seen as much darker magic than apis and runtimes.
masak Tene: thank you. no, haven't seen keep. 13:22
masak checks it out
moritz_ diakopter: well, in a sense they are ;-)
Tene masak: the description is on this page: conway.rutgers.edu/~ccshan/wiki/blo...der_shell/
13:22 snearch joined
masak looks like a good starting point for tote. :) 13:23
diakopter moritz_: yeah, usually, I know. :) but sometimes I let myself deludedly forget :P
hercynium is trying to figure out if his memory of Python formerly having a cartoonish logo is correct
diakopter green viper?
Tene masak: the gf has a couple of sketches that you asked for, and is going to scan them and finish them sometime soon, just is busy with school.
13:24 SmokeMachine joined
hercynium I think it was this: www.vanille-media.de/images/python.png 13:24
diakopter --outline-drawings++ 13:25
saran masak : i am back today after my office work
hercynium I'm sure O'Reilly will come up with something suitably "professional" when the time is right
masak saran: cool.
diakopter (--outline_drawings++ : appear to decrement, but actually stay neutral :) 13:26
hercynium should have put "suitably" in quotes as well :)
moritz_ hercynium: and hold the copyright on it, just like the camel - not what we want
masak hercynium: ISTR Guido resisted any snake-like mascots for a long time.
hercynium moritz_: indeed
Matt-W It's a shame, because the camel is good
Matt-W also likes the onion
pugs_svn r28271 | moritz++ | [perl6.org] intial layout for menus 13:27
hercynium however, they probably *will* have their own icon, and people in corporate life will associate it
moritz_ there's nothing wrong with that
just like rakudo has its own logo
hercynium agreed :) 13:28
saran mosak : trying to understand this RT ticket rt.perl.org/rt3//Public/Bug/Display...l?id=58922 13:29
moritz_ moritz.faui2k3.org/tmp/perl6.org.png 13:30
13:32 snearch left
masak saran: are you experimenting with names of misspelling my nick? :) 13:32
I'll see you faster if you write 'masak'. :P
moritz_ ma<tab key> 13:33
saran masak : am sorry abt the typo :)
masak no prob. :)
Matt-W moritz_: Does Blizkost really come under Perl 6 Compilers?
masak saran: can I help you in some way?
Matt-W (game of table tennis) & 13:34
moritz_ Matt-W: have fun
don't know about blizkost
diakopter saran: shouldn't that remain stalled until the condition in pmichaud's comment is satisified? (or help that condition be satisfied?) 13:35
saran masak : give me a minute..lemme understand the RT ticket
masak diakopter: probably, yes.
diakopter: we're just seeing if we can un-stall it at this point. 13:36
diakopter there are only a couple of options: create a new syntax (like yaml's numbered object IDs) that lets object literals reference other parts of the same object literal expression, or make it statement/phase-oriented 13:37
saran masak : in this oneliner <code> perl6 -e 'my @a = (1); @a[0] := @a; @a.perl' </code> . am more curious abt @a[0] :=@a; # is this equal to @a[0] = @a[0]:@a ? what it does ? 13:38
masak saran: the ':=' is a binding.
saran: it makes @a[0] point to the same memory location as @a.
Matt-W moritz_: No fun, somebody's having a meeting next to the table tennis table 13:39
masak or, I should say, the same container.
moritz_ diakopter: well, .perl is specced to return perl code. 'do { my $tmp = [1, 2]; $tmp[2] := $tmp } ' would be avalid return value
diakopter right, I wasn't (at that point) arguing for one or the other option... but the more I think about it, the more I suspect a new self-referencing object literal syntax would be extremely handy
moritz_ Matt-W: :( my next match is on Monday, and I haven't played for months
13:40 cmv left
saran masak : Ok.. got it. in my words..one of the element in the array holding the parent array memory value ..correct me if i am wrong ? 13:40
masak saran: sounds correct to me. 13:41
saran: and thence the circularity.
diakopter it might be something like relative directory references (. or .., except not dots, or something), or alternatively absolute references with in-expression IDs
moritz_: oooohhh! 13:42
what about this
a syntax that allows just a shorthand way of typing what you just wrote... a twigil on a variable that makes it "expression-scoped"... so the do{} isn't necessary 13:44
saran masak : thanks.as u said yesterday am going to check out the function call on src/setting/List.pm, line 42 13:45
diakopter moritz_: my @b = do { my $tmp = [1, 2]; $tmp[2] := $tmp } becomes my @b = [1, 2, @=b] ... or something 13:46
masak saran: yes. it might be useful to know that this is the code that runs when you do .perl on a list.
13:46 punter joined
diakopter moritz_: or using the relative "level" idea, (so "b" isn't necessary): [1, 2, @..] 13:48
(yes, I know I got the sigils incorrect in those examples, but you see my point I hope)
moritz_ well, one could think about it 13:49
masak diakopter: sounds like a cool idea for a syntax-modifying module.
diakopter masak: but not a cool idea for the STD language? 13:50
masak in my humble opinion, no.
too specialised.
13:51 saran left 13:52 saran joined, guest1235123 joined 13:53 ruoso joined
diakopter masak: I'm trying not to sound overly defensive, but when you denigrate the idea so summarily without giving any explanation, how else am I going to sound when I ask, "how is it too specialised?" 13:53
masak diakopter: sorry if I sounded too dismissive. I'll try to motivate my feeling, though in the end it's only a feeling. 13:55
in my day-to-day programming, I don't encounter situations where I need to introduce loops in my data structures. 13:56
when one does, the feature is perfect, which is why I think it would make a nice module.
diakopter My motivation is this: self-referencing data structures are extremely common; being able to serialize them declaratively (in-line, non-statement-oriented) without acrobatics would be extremely helpfull.
helpful.
well there's the difference.
moritz_ Perl 6 programmers try to solve every problem with syntax ;-)
diakopter you don't deal with those structures
masak but putting it in core would complicate the language for everyone, with fairly little average gain.
13:57 payload joined
diakopter honestly, I think it *lessens* the complexity of the language, by making it much more declarative. There are many occasions where statement-oriented code can be "inlined" more by introducing in-expression relative references. 13:58
13:58 punter left
saran masak : As i am a newbie lemme understand --> how do i iterate over the loops --> www.programmersheaven.com/2/Perl6-FAQ-Basics#for 13:58
masak rakudo: for 1,2,3 -> $value { say $value } 13:59
p6eval rakudo 4de821: OUTPUT«1␤2␤3␤»
masak saran: like that, basically.
diakopter masak: (I'm not merely referring to serialization/.perl; I'm talking about source code in general). Many a programmer has to "refactor" an expression out into several statements just because they have to "predeclare" some reference. 14:00
masak saran: you can also use @array.map, like the code in List.pm does.
diakopter: I hear ya. do you have a very convincing use case? 14:01
diakopter ok. the serialization of STD.pm's AST. 14:02
14:02 zamolxes joined
masak diakopter: but that can be solved with the do blocks we've been talking about. no need to introduce wholly new syntax. 14:02
diakopter: in fact, that's how Data::Dumper does it. 14:03
I'd go so far as to say that $root is a great bit more readable than %........
diakopter that can't solve mutually recursive cycles (indirect cycles). Also, it doesn't solve multiple references across the tree (that aren't a parent of each other)
masak why wouldn't it solve mutually recursive cycles? 14:04
saran masak : i cant understand what do u say " @array.map " in List.pm 14:05
colomon rakudo: say atanh(1); 14:06
p6eval rakudo 4de821: OUTPUT«Divide by zero␤»
colomon rakudo: say atanh(.9);
p6eval rakudo 4de821: OUTPUT«1.47221948958322␤»
saran masak : you mean this line " return '[' ~ self.map({ .perl }).join(", ") ~ ']'; "
colomon rakudo: say atanh(1.1);
masak saran: yes.
p6eval rakudo 4de821: OUTPUT«NaN␤»
masak saran: it says 'self.map' but 'self' here is a List.
colomon rakudo say atanh(2); 14:07
saran masak : just a little confusion..it is better to clear all small small doubts before getting into the problem
masak saran: my point is that it's another way to loop over lists, and sometimes it's preferable to 'for'.
saran: oh, definitely.
colomon rakudo: say atanh(2);
diakopter Let me revise that negative case to be: a combination of a mutually recursive cycle and cross-tree references. In the most pathological case, it's just assigning a temporary "named" (or "keyed") variable to each reference in the tree, and then binding all the links afterwards.
p6eval rakudo 4de821: OUTPUT«NaN␤»
14:07 ssm left
colomon rakudo: say atanh(3); 14:07
p6eval rakudo 4de821: OUTPUT«NaN␤» 14:08
masak saran: I'd recommend that you approach this by actually writing a test, and then try to make Rakudo pass it. that will probably help in giving the effort a short-term goal.
diakopter masak: My (perhaps) larger point is that the ability to serialize a data structure as a single expression makes it [have the potential to be] perfectly round-trippable. 14:10
14:10 Zloyrusskiy joined
saran masak : wat do u mean by "writing a test" . you mean simulating the circular reference error 14:10
14:11 Zloyrusskiy left
diakopter saran: type out what you think the output of that test case should be 14:11
masak saran: yes. having a short program that causes the error will help you greatly.
saran: the program in the RT ticket will do.
saran: as I see it, your first task is to make Rakudo not hang. your second task is to make .perl output something sensible that creates such a structure. 14:12
saran masak : yes i am trying to understand the " :=" operator..
masak saran: have you found S03?
Tene masak: if we get a good version of fairy-camelia, and get a version of django running on parrot, then we'd have an excuse to make a logo for "Perl 6 on django", the form of which should be obvious. ;) 14:13
saran masak : wat is S03 ?
masak diakopter: yes, I understand. it's just that I find a some advantages in not intriducing any new syntax at all. Data::Dumper seems not to suffer too much from its way of doing things. 14:14
saran: perlcabal.org/syn/S03.html
saran masak : am studying this which may help to understand the ":=" --> www.programmersheaven.com/2/Perl6-F...es#binding
masak saran: that might be good. the synopses are the source of all knowledge, though. 14:15
Tene: :) 14:16
guest1235123 does anyone think a guide to the channel's notation would be helpful? and if so, link in chan topic? 14:17
Tene guest1235123: if it seems like we have significant notation, sure. It doesn't seem that way to me, but I'm not in a position to effectively judge. :) 14:18
14:18 smash joined
smash hello everyone 14:18
Tene guest1235123: you do want to remember, though, that nearly nobody ever reads the topic.
HI SMASH
14:19 Zloyrusskiy joined
colomon Tene: sad to say, it took me about two months to figure out what "afk" meant. ;) 14:19
Tene colomon: that's very much not specific to this channel... but I don't have any idea what proportion of #perl6 members are new to IRC. 14:20
14:20 KyleHa joined
colomon There probably is a useful generic guide to IRC out there somewhere... 14:21
masak I think the conventions of #perl6 would be a bit difficult to pin down.
people here tend to like to bend notation, adapting it on the fly.
which, come to think about it, is pretty appropriate. 14:22
colomon :)
saran masak : got understood from the ":=" aka it is alias what we use symbolic references in perl5 by this link www.programmersheaven.com/2/Perl6-F...es#binding 14:23
masak saran: that's close enough, I think.
saran: in Perl 6, it's called 'binding'. 14:24
saran: you can do $a := $b, and then when you change $a, $b will also change.
14:25 justatheory joined
saran masak : yes..got it clearly. Thanks for the link 14:25
14:26 Zloyrusskiy left 14:31 Zloyrusskiy joined
pmichaud good morning, #perl6 14:34
Tene hi pm
masak goooooooood morning, pmichaud!
ruoso bom dia 14:36
14:36 nihiliad joined
hercynium #perl6 is pmichaud's vietnam? 14:36
;-)
Matt-W oh hai pmichaud 14:37
hercynium imagines robin williams
14:37 saran left
pmichaud yes, #perl6 is my personal quagmire :) 14:43
diakopter moritz_: ok, here's a non-syntactic alternative
fix { [1,2,$_] }
14:44 Intensity left
diakopter suggestion courtesy of someone off-channel who can remain anonymous (or not) if they wish :P 14:44
masak likes fix
diakopter now, deriving those fix expressions for complex structures might be tricky. probably there's some algorithm somewhere for it. 14:45
for .perl, I mean
masak I'm having an easier time picturing that fix might blow up for some cases of loopy references, though. 14:46
14:47 Psyche^ joined
diakopter otoh, the syntax I suggested can be considered shorthand for "recursive lets" 14:47
colomon Is there a clean way to combine "when" with a flag? 14:49
Something like when $in_setup && /blue/ { do something }
where $in_setup is a Bool? 14:50
pmichaud colomon: what you just wrote should work.
colomon It won't try to $_ ~~ $in_setup or something like that? 14:51
pmichaud it will actually $_ ~~ ($in_setup && /blue/)
14:53 Intensity joined
jnthn o/ 14:53
colomon and ~~ Bool is just Bool, and then it does $_ ~~ /blue/ ? 14:54
That's awesome.
moritz_ at surely works for & instead of &&
rakudo: given 'foo' { when True && /oo/ { say "bar" } } 14:55
p6eval rakudo 4de821: OUTPUT«bar␤»
moritz_ rakudo: given 'foo' { when True && /oof/ { say "bar" } }
p6eval rakudo 4de821: ( no output )
moritz_ rakudo: given 'foo' { when False && /oo/ { say "bar" } }
p6eval rakudo 4de821: ( no output )
colomon woo-hoo!
moritz_ seems to work.
pmichaud it doesn't thread over &&, if that's what you're asking 14:56
it does a normal ($in_setup && /blue/)
moritz_ aye
pmichaud if $in_setup is false, then the ~~ expression becomes ~~ False
moritz_ but /blue/ binds to the correct $_ 14:57
pmichaud if $in_setup is true, then we match /blue/ to $_
14:57 klaus_ is now known as muixirt
pmichaud (which is the default behavior for any regex in boolean context) 14:57
moritz_ and ~~ False is always false
so everything is fine ;-)
masak Perl6++
14:58 Patterner left, Psyche^ is now known as Patterner
colomon thanks! 14:59
jnthn ooh...a new masakpost! :-D 15:01
masak :)
15:02 alester joined
masak this iron man thing actually works. nowadays, I have a little mental variable that tells me when I need to blog. mst++ 15:02
moritz_ aye 15:03
TimToady I blog on twitter once a week or so now :) 15:04
jnthn masak++ # got f**king blogging ;-)
moritz_ actually I think I still have one spare post that's going to be published on time
jnthn didn't quite get dragged into Twitter yet. 15:05
pmichaud I find twirssi to be very helpful for twitter stuff :)
tweets I'm interested in are just another window in my irssi client :) 15:06
jnthn that's a nie way to manage it 15:07
*nice
I guess the real issue is that I didn't convince myself of he value of it yet.
pmichaud well, since I generally have irssi open anyway, it's easy to just notice activity in the twitter window
in some ways, twitter is like an aggregator for blog comments 15:08
masak no, that's RSS. 15:09
pmichaud I've seen RSS aggregate blog postings, but not the comments :)
viklund most blogposts have a special rss-feed just for the comments 15:10
15:10 agentzh joined
viklund (or maybe I should say, most bloggingsoftware) 15:10
pmichaud does use.perl have one?
jnthn I guess maybe I should give twitter a go at some point. 15:11
finanalyst rakudo: my %h= <a b c> Z (1,[2,3,4],5); my @a=2,3,4; say ?(all(%h<b>) == any(@a));my %hs= <a b c> Z (1,[<a b c>],5); my @as=<a b c>; say ?(all(%hs<b>) eq any(@as))
pmichaud it bugs me that I don't get notices when someone comments on an article I wrote
p6eval rakudo 4de821: OUTPUT«1␤0␤»
jnthn I'm afraid of getting too hooked by it.
viklund dunno
jnthn Then, I didn't get hooked on Facebook, it's just a useful tool for keeping in touch with various people who prefer to communicate that way...
finanalyst pmichaud: junctions with eq seems to have stopped working
__ash__ my twitter account is um... dead, it gets update from an iphone app that records how long my jogs are and thats all thats on my twitter account right now 15:12
viklund doesn't look like it
pmichaud rakudo: say (any(<3 4 5>) eq 4).perl
p6eval rakudo 4de821: OUTPUT«any(Bool::False, Bool::True)␤»
finanalyst rakudo: my %h= <a b c> Z (1,[2,3,4],5); my @a=2,3,4; say ?(all(%h<b>) == any(@a));my %hs= <a b c> Z (1,[<a b c>],5); my @as=<a b c>; say ?(all(%hs<b>) eq any(@as))
p6eval rakudo 4de821: OUTPUT«1␤0␤» 15:13
finanalyst should both be 1
it used to work
pmichaud rakudo: my @as = <a b c>; say [<a b c>] eq @as; 15:14
p6eval rakudo 4de821: OUTPUT«1␤»
pmichaud rakudo: my %hs = ('b', [<a b c>]); say %hs.perl;
p6eval rakudo 4de821: OUTPUT«{"b" => ["a", "b", "c"]}␤»
masak no parens needed. :)
pmichaud rakudo: my %hs = ('b', [<a b c>]); my @as = <a b c>; say %hs<b> eq @as;
p6eval rakudo 4de821: OUTPUT«1␤»
pmichaud oh, yes, of course 15:15
Rakudo is giving the correct output.
oh, wait, maybe not.
rechecking.
colomon ooo, just got burned by using $1 instead of $0. 15:16
finanalyst rakudo: my @x=<a b c>; say ?(all(@x) eq any(<a b c>))
p6eval rakudo 4de821: OUTPUT«1␤»
pmichaud rakudo: my %hs = ('b', [<a b c>]); my @as = <a b c>; say all(%hs<b>) eq any(@as);
p6eval rakudo 4de821: OUTPUT«all(any(Bool::False))␤»
pmichaud rakudo: my %hs = ('b', [<a b c>]); my @as = <a b c>; say all(%hs<b>.list) eq any(@as);
p6eval rakudo 4de821: OUTPUT«all(any(Bool::True, Bool::False), any(Bool::False, Bool::True), any(Bool::False, Bool::True))␤»
pmichaud all(%hs<b>) consists of a single element 15:17
cognominal jnthn, probably, I don't know HLLCompiler well enough, but I don't know yet how to use load_library in blizkost?
finanalyst but when the elements are integers not strings, the junction flattens
pmichaud no, it becomes 3
here. 15:18
rakudo: my %hs = ('b', [<a b c>]); say all(%hs<b>).perl
p6eval rakudo 4de821: OUTPUT«all(["a", "b", "c"])␤»
finanalyst rakudo: my %h= <a b c> Z (1,[2,3,4],5); my @a=2,3,4; say ?(all(%h<b>) == any(@a))
p6eval rakudo 4de821: OUTPUT«1␤»
pmichaud note that it's all(["a", "b", "c"]) and not all("a", "b", "c")
so the request of using eq becomes a test for "a b c" eq "a" || "a b c" eq "b" || "a b c" eq "c" 15:19
here, watch this way
jnthn cognominal: See what use impl does in Rakudo
pmichaud rakudo: my %h= <a b c> Z (1,[7,8,9],5); my @a=7,8,9; say ?(all(%h<b>) == any(@a)) 15:20
p6eval rakudo 4de821: OUTPUT«0␤»
pmichaud correct.
jnthn cognominal: But basically, it takes some params and hands back a hash.
cognominal impl?
jnthn implementation
KyleHa rakudo: class A { has @.a; method empty { @.a = () }; A.new.empty;
p6eval rakudo 4de821: OUTPUT«Unable to parse block; couldn't find final '}' at line 2, near ""␤in Main (src/gen_setting.pm:3469)␤»
pmichaud finanalyst: figured it out yet?
KyleHa rakudo: class A { has @.a; method empty { @.a = () }; }; A.new.empty; 15:21
p6eval rakudo 4de821: OUTPUT«Cannot assign to readonly variable.␤in method A::empty (/tmp/ymny7wpRiy:2)␤called from Main (/tmp/ymny7wpRiy:2)␤»
15:21 rfordinal joined
finanalyst pmichaud: nope. i had a similar line in code that was working. I have returned to it and it stopped working 15:22
15:22 mspaulding joined
muixirt where is the separation of statements with semicolons specced? 15:23
pmichaud finanalyst: well, perhaps Rakudo was broken before. But all of the examples above look correct to me.
finanalyst \action thinking
pmichaud muixirt: S04, I think.
I'm not sure it's "specced" beyond the information about statement-ending blocks in S04. 15:24
masak it doesn't seem to be.
it's just taken to be the same as in Perl 5, I guess.
finanalyst rakudo: my %h= <a b c> Z (1,[7,8,9],5); my @a=7,8,9; say ?(all(%h<b>) == any(@a)) 15:25
p6eval rakudo 4de821: OUTPUT«0␤»
finanalyst rakudo: my %h= <a b c> Z (1,[2,3,4],5); my @a=2,3,4; say ?(all(%h<b>) == any(@a))
p6eval rakudo 4de821: OUTPUT«1␤»
15:25 cmv joined
pmichaud rakudo: my %h= <a b c> Z (1,[7,8,9],5); my @a=2,3,4; say ?(all(%h<b>) == any(@a)) 15:25
p6eval rakudo 4de821: OUTPUT«1␤» 15:26
TimToady muixirt: and S03, and STD
pmichaud because all([7,8,9]) == any(2,3,4)
(because +[7,8,9] == 3) 15:27
muixirt TimToady: a semicolon is an operator?
TimToady kinda
finanalyst pmichaud: I am really getting lost here
TimToady "sequencer" precedence 15:28
pmichaud finanalyst: in numeric context, all([7,8,9]) is the same as all(3)
TimToady same as feed operators
15:28 ms111 left
finanalyst ah now I see. Not intuitive! 15:28
pmichaud so the correct way to get what you want is all(@%h<b>) 15:29
finanalyst how then to flatten [7,8,9] to 7,8,9?
pmichaud however, Rakudo doesn't have the prefix @ yet, so the workaround is all(%h<b>.list)
finanalyst crossed msgs
TimToady std: all(@%h<b>)
p6eval std 28271: OUTPUT«===SORRY!===␤Anonymous variable requires declarator at /tmp/PzW11pYAfA line 1:␤------> all(@⏏%h<b>)␤ expecting twigil␤FAILED 00:02 37m␤»
TimToady std: all(@(%h<b>)) 15:30
pmichaud although I think that @(%h<b>) works
p6eval std 28271: OUTPUT«Potential difficulties:␤ Variable %h is not predeclared at /tmp/J9ysZUvmTW line 1:␤------> all(@(%h⏏<b>))␤ok 00:03 37m␤»
pmichaud rakudo: my %h = 'b', <7 8 9>; say all(@(%h<b>)).perl
p6eval rakudo 4de821: OUTPUT«all("7")␤»
15:30 Zloyrusskiy left
pmichaud oops 15:30
rakudo: my %h = 'b', [<7 8 9>]; say all(@(%h<b>)).perl
p6eval rakudo 4de821: OUTPUT«all("7", "8", "9")␤»
pmichaud there.
finanalyst thank you!!!
TimToady rakudo: my %h = 'b', [<7 8 9>]; say all(%h<b>.values).perl 15:31
p6eval rakudo 4de821: OUTPUT«all("7", "8", "9")␤»
15:32 drbean_ joined, drbean_ left
finanalyst .values makes it more understandable 15:33
phenny AttributeError: 'NoneType' object has no attribute 'startswith' (file "/var/www/inamidst.com/htdocs/phenny/modules/validate.py", line 15, in val)
jnthn oh noes phenny is has a fail
finanalyst .oh noes 15:34
pmichaud ".py" ??
pugs_svn r28272 | kyle++ | [t/spec] Test for RT 69192
finanalyst blasphemy!
TimToady nah, just crudity
pmichaud mail [email@hidden.address] -s "[BUG] phenny appears to be written in Python" 15:35
mail [email@hidden.address] -s "[TODO] Implement 'use Python'"
TimToady
.oO(pmichaud is a stand-up camelian)
15:36
pmichaud I'm often accused of changing my colors.
15:36 drbean_ joined 15:37 mspaulding left
Tene pmichaud: I can bump HLL interop support for pynie higher on my task list if you'd like. 15:37
pmichaud Tene: that's okay, I won't be needing it anytime soon. 15:38
15:38 iblechbot left
pmichaud I have to get some grammar engine work done first 15:38
15:39 mspaulding joined
jnthn yay! 15:39
pmichaud: Does PGE have contextual var decls yet?
masak \o/
pmichaud jnthn: not yet
masak is there a Rakudo release today? 15:40
pmichaud particle++ is release manager for today
masak [particle]++
15:40 rfordinal left
pmichaud jnthn: I'm likely to do protoregexes this next week, then contextuals and lexicals after that 15:41
today/tomorrow I have to do $otherjob stuff :-(
jnthn oh, re-ordering?
It worked out better that way?
pmichaud well, I'd like NQP to be parsing regexes, using something similar to the way STD.pm parses regexes 15:42
so in order for that to happen I really need protoregexes
jnthn aha, OK.
Sounds sane.
jnthn looks forward to it
masak also 15:43
TimToady jnthn: www.programmersheaven.com/2/Perl6-F...es#binding is now wrong on the subject of ::=
colomon pmichaud: Still looking at the Perl Foundation address after my bounce from their Oregon address -- their contact page has a "Grand Ledge, MI" address.
Tene pmichaud: you expect to need python support for something? for what?
pmichaud then when NQP is parsing regexes, I can add the contextual declarations to NQP's parser much more straightforwardly than trying to add it to PGE's current regex parser
jnthn pmichaud: OK, makes sense from an efficiency POV.
pmichaud Tene: I was making a funny earlier
Tene pmichaud: Ah, right. ><
pmichaud jnthn: also, protoregexes are the outstanding item on my Hague grant 15:44
Tene still waking up... slowly.
jnthn TimToady: I need to give that FAQ a good look over generally for spec movements...it's not really kept up.
pmichaud I think I might fix := and ::= for Rakudo day this next week
15:44 drbean left 15:45 hercynium left, jferrero joined
masak great times ahead, it seems. 15:45
jnthn Yeah! 15:46
I'm liking this month.
:-)
Tene It looks like my life might be starting to settle down soon, hopefully... 15:48
15:48 ejs left
Tene masak: I heard you got squirrel working well... just with sqlite, or does it have mysql support too 15:49
15:49 am0c joined
Tene ? 15:49
masak Tene: it had a name change for starters. it's now known as Squerl. :)
Tene: it uses your SQLite3 module.
no MySQL support yet. would be nice to have that too. 15:50
will blog about it tomorrow.
right now I'm mostly porting and passing tests from Sequel. I would blog today, but I want a few more tests before I do. 15:51
jnthn Awww...squirrel was a cute name.
15:51 mspaulding left
masak it still sounds the same. :) 15:52
TimToady rhymes with Perl
Tene masak: i can download from your github repo, yes? 15:54
masak aye.
I'm currently filling up t/squerl/02-dataset.t... slowly. 15:55
Tene oh, it's in web
masak yes.
it would be nice if tote were already written at this point, then I would have used it for writing and passing tests. :) 15:56
Tene Yay, github crashing again!
masak they're being DDoSed. :/
Tene orly?
dukeleto says "screw you" to the lamers DDoS'ing github
masak Tene: twitter.com/github/status/4056522781 15:57
yes, it seems really meaningless. why would anyone want to DDoS github? 15:58
guest1235123 Squerl, pronounced like an Amerkin would? I like.
15:58 guest1235123 left
dukeleto masak: perhaps they hate freedom and/or free software 15:59
masak dukeleto: :)
sjohnson hi TimToady ! 16:01
back from India
masak o/ sjohnson
sjohnson hi masak too
:)
masak how was India?
sjohnson surprisngly a lot of fun 16:02
didnt thin i'd have so much fun in a 3rd world country
think 16:03
masak nice to hear.
India does seem an interesting country.
Tene masak: "Apparently the attackers are targeting a specific project on GitHub" -- twitter.com/github/status/4034616135 16:05
16:05 araujo joined
dukeleto Tene: yeah, that is really lame 16:05
16:06 mspaulding joined
masak still -- I wonder which project, and why. 16:06
Tene Me too.
araujo hello
dukeleto araujo: good localtime() 16:07
masak araujo: \o
araujo hi dukeleto masak
:)
16:08 Zloyrusskiy joined, Zloyrusskiy left
colomon rakudo: acosh(-1) 16:10
p6eval rakudo 4de821: ( no output )
colomon rakudo: say acosh(-1)
p6eval rakudo 4de821: OUTPUT«NaN␤»
dukeleto rakudo: say acosh(NaN) 16:13
p6eval rakudo 4de821: OUTPUT«NaN␤»
masak food &
16:13 masak left 16:14 rfordinal joined
colomon cosh on reals has to be positive, so acosh of a negative is NaN. Makes perfect sense. :) 16:16
16:17 jferrero left
dukeleto colomon: indeed 16:19
colomon I'm still feeling through these things while I work on the trig tests.
dukeleto colomon: i wrote most of the tests dealing with NaN in t/spec, if you have questions, let me know
colomon Ooo, hey, I do have one. 16:20
dukeleto rakudo: say acosh(Inf)
p6eval rakudo 4de821: OUTPUT«Inf␤»
16:20 abra left
colomon Okay, two, actually. I don't think you wrote any tests for Inf and inverse functions? 16:20
16:21 ejs joined
dukeleto that may be wrong. since lim( cosh(x) ) = Inf as x -> +- Inf, it is not a one-to-one function there and acosh(Inf) should probably be NaN 16:21
i need more coffee to make be sure
colomon: I just did :) 16:22
colomon: yes, it looks like the inverse functions don't know that Inf is special
rakudo: say acosh(-Inf)
p6eval rakudo 4de821: OUTPUT«NaN␤»
dukeleto ok, that is good
jnthn sleeep 16:24
dukeleto colomon: perl 5's Math::Complex says that acosh(Inf) = inf (not sure I buy that) and I just found a bug when you try to do acosh(-Inf): Math::Complex::make: Cannot take real part of '-Inf' 16:25
colomon Are you saying you turned up a Perl 5 bug? :) 16:26
dukeleto colomon: in general, any function where multiple input values map to the same output value, the inverse function at that point should return NaN
colomon: looks like it. fun fun
16:27 cognomore joined
colomon dukeleto: that can't be right, that would blow the heck out of the normal inverse trig functions 16:27
dukeleto colomon: i am not talking about periodic functions
that i should have clarified
colomon so any non-periodic functions where multiple input values map to the same output should be NaN.
that makes sense, I suppose. 16:28
My first question, which got sidetracked:
dukeleto the inverse function at that point should be NaN, that is.
colomon: so basically, this means the hyperbolics, or to say in another way, the exponentials and logs 16:29
colomon only an issue with cosh, isn't it? (he said, looking at the graphs of the hyperbolic functions)
dukeleto colomon: sech(x) is an even function as well 16:30
colomon: this happens with even functions on the real line
colomon ah, yes, nearly invisible on the graph I was looking at.
dang it, I can't find the number I was looking at in the log. 16:31
rakudo: say atanh(1)
p6eval rakudo 4de821: OUTPUT«Divide by zero␤»
colomon That's it.
rakudo: say atanh(1.1)
p6eval rakudo 4de821: OUTPUT«NaN␤»
colomon atanh is NaN if x > 1, divide by zero if x == 1
I'm think it should be NaN if x >= 1 16:32
16:33 lionWrk joined
dukeleto colomon: they both are forms of "NaN", just that one is loud and one is quiet. Dividing directly by zero usually seems to trigger an exception, where doing other non-sensical things gives you a NaN value 16:33
colomon: you may want to ask TimToady what he wants
16:33 lionWrk is now known as LionMadeOfLions
pmichaud yes, I'm wondering if returning a failure is actually more correct here 16:33
just like divide by zero should return a failure 16:34
dukeleto pmichaud: you mean a Failure exception object thingy ?
pmichaud yes
__ash__ ruby makes a distinction between Nan and runtime errors like division by zero, btw
dukeleto pmichaud: see, you *are* interested in ieee754-2008 ;)
16:35 rfordinal left
pmichaud in particular, my $a = 3 / 0; should not throw an exception 16:35
colomon I'd be okay with that (though there'd be tests to be rewritten :) ) as long as the failure is consistent for atanh(1) and atanh(1.1).
dukeleto __ash__: the perl 6 spec says that you should be able to lexically define how NaN's are propagated (quiet or signalling)
colomon pmichaud: Actually, 3 / 0 returns a valid Rat.
pmichaud okay, 3.0 / 0 then 16:36
colomon rakudo: my $a = 3 / 0; say $a.perl;
dukeleto rakudo: say 3.0 / 0
p6eval rakudo 4de821: OUTPUT«1/0␤»
rakudo 4de821: OUTPUT«Divide by zero␤»
dukeleto rakudo: say 3/0
p6eval rakudo 4de821: ( no output )
__ash__ dukeleto: ah, got ya, makes sense
pmichaud rakudo: my $a = 3.0 / 0;
p6eval rakudo 4de821: OUTPUT«Divide by zero␤»
pmichaud rakudo is wrong there.
(I tried fixing it once but ran into a few issues so saved it for another day) 16:37
note that "say 3.0 / 0" *does* throw an exception
because we're then trying to take the value of a Failure, which throws its associated exception
dukeleto pmichaud: so you mean, if you attempt to stringify a Failure, it croaks ? 16:38
pmichaud if the Failure has been unhandled, yes.
er, has not been handled
dukeleto thinks about being unhandled 16:39
16:39 zamolxes left
pmichaud S04: "If you test a C<Failure> for C<.defined> or C<.true>, it causes C<$!> 16:40
to mark the main exception as I<handled>; the exception acts as a
relatively harmless undefined value thereafter. Any other use of the
C<Failure> object to extract a normal value will throw its associated
exception immediately.
16:41 jauaor joined
jauaor :) 16:41
16:41 iblechbot joined
pmichaud afk, errands 16:41
colomon rakudo: say acosh(0) 16:44
p6eval rakudo 4de821: OUTPUT«NaN␤»
colomon rakudo: say acosh(1)
p6eval rakudo 4de821: OUTPUT«0␤»
16:49 finanalyst left
colomon masak: "tote" == "Totally Obsessive TDD Environment" ? 16:54
16:55 agentzh left, zaphar_p2 joined 16:57 kidd_ joined 16:58 jrtayloriv left 17:00 saran joined 17:01 zostay left 17:02 zostay joined 17:03 desertm4x joined
moritz_ [particle]: how's the release going? 17:04
(or when?) ;-)
17:07 zaphar_ps left 17:09 sevvie left
colomon dukeleto: re multiple input values mapping to one output value -- cosh graphs a parabola-style shape, with cosh(x) == cosh(-x). by your argument, acosh should only be defined for 1. Presumably it is actually defined to take the positive branch of the function, and therefore acosh(Inf) = Inf is perfectly reasonable. 17:20
17:21 __ash__ left 17:24 jrtayloriv joined 17:28 saran left 17:39 abra joined 17:48 pmurias joined
pmurias diakopter: what i think would be a very cool thing would be to have a debugger/tracer for vivjs which would highlight the exact bit of code that is executed 17:49
as the AST you are executing has a direct mapping to places in source code 17:50
diakopter pmurias: that would be cool... 17:51
like a dynamic edition of that std-highlight webpage form thing
pmurias yes, something like that
diakopter moritz_: is /msg p6eval anything_not_matching_prefixes supposed to try [and fail] to call STD_syntax_highlight 17:54
pmurias diakopter: it's not supposed to fail 17:57
the path in the evalbot is incorrect 17:58
diakopter o
pmurias it moved to STD_highlighy 17:59
s/y/t/
std_highlite actually
18:00 reqamst left, smash left
TimToady std: say "howdy $world" 18:00
p6eval std 28272: ( no output )
diakopter std: "$world"
p6eval std 28272: ( no output ) 18:01
18:01 sm` joined
TimToady std: say "howdy $world" 18:02
p6eval std 28272: ( no output )
TimToady grrr
sjohnson uh oh
TimToady std: 42
p6eval std 28272: OUTPUT«ok 00:02 37m␤» 18:03
TimToady std: say "howdy $world"
p6eval std 28272: ( no output )
TimToady probably not giving it enough time to recalculate the "" grammar, but that ought to be cached already, y'd think
sjohnson it's because it knows that $world will be "world" anyway 18:04
DOES NOT COMPUTE
18:04 abra left
TimToady std: say "howdy $IO-monad" 18:04
p6eval std 28272: OUTPUT«Potential difficulties:␤ Variable $IO-monad is not predeclared at /tmp/SlySsoyROO line 1:␤------> say "howdy $IO-monad⏏"␤ok 00:04 38m␤» 18:05
TimToady that was the problem :)(
TimToady tries to figure out what his typo smiley means...
18:05 rbaumer left
TimToady a grimace, perhaps 18:05
18:06 sm^^^ left
TimToady diakopter: see ^^ 18:06
18:07 sjohnson sets mode: +v p6eval, ejs left 18:13 cmv left
diakopter std: "HOWDY $WORD STOP" 18:19
p6eval std 28272: ( no output )
18:20 he_ left
diakopter cornfusion 18:20
18:20 he_ joined 18:21 stephenlb joined
diakopter std: "howdy $IO-monad" 18:21
p6eval std 28272: ( no output )
diakopter std: say "howdy $IO-monad" 18:22
p6eval std 28272: ( no output )
diakopter std: say "","howdy $IO-monad" 18:24
p6eval std 28272: OUTPUT«Potential difficulties:␤ Variable $IO-monad is not predeclared at /tmp/IiQXrPPXE4 line 1:␤------> say "","howdy $IO-monad⏏"␤ok 00:04 39m␤»
18:26 bloonix_ left, bloonix joined, he_ left 18:27 he_ joined 18:30 _jaldhar joined 18:31 jaldhar left 18:34 bloonix left 18:38 sjohnson left
pugs_svn r28273 | fglock++ | [mp6] Javascript emitter and runtime updates - grammar works 18:38
18:40 sjohnson joined
diakopter rakudo: say("Meesa »ö«") 18:41
p6eval rakudo 4de821: OUTPUT«Meesa »ö«␤»
diakopter pugs: say("Meesa »ö«")
p6eval pugs: OUTPUT«Meesa »ö«␤»
pugs_svn r28274 | lwall++ | [STD] don't record null nibbles unless whole string is '' 18:46
r28274 | [viv] handle bracketing quotes correctlier
diakopter two for the price of 1
diakopter reluctantly types `make`
I mean, excitedly. 18:47
jrtayloriv Is it OK if I move the <ws> rule of a grammar to the bottom, following all of the places where it is used? Or is using a rule/token similar to C functions, in that I must declare it somehow before I use it? 18:53
18:53 azawawi joined
TimToady it's like methods, so it doesn't matter what order you declare them 18:54
jrtayloriv And it's location in the source make any semantic difference at all?
TimToady, thanks
TimToady in fact, in STD it's not just *like* methods, it *is* mehtods
*methods
rakudo too, these days 18:55
colomon rakudo: say cotanh(0)
p6eval rakudo 4de821: OUTPUT«Divide by zero␤»
TimToady though perhaps not Perl 6 methods...
colomon rakudo: say cotanh(.5)
p6eval rakudo 4de821: OUTPUT«2.16395341373865␤»
colomon rakudo: say cotanh(100) 18:56
p6eval rakudo 4de821: OUTPUT«1␤»
colomon rakudo: say acotanh(1) 18:57
p6eval rakudo 4de821: OUTPUT«Divide by zero␤»
colomon TimToady: That reminds me -- we meant to ask you if trig functions should be allowed to return divide by zero. 18:58
rakudo: say acosh(1.1)
p6eval rakudo 4de821: OUTPUT«0.443568254385115␤»
colomon wrong one 18:59
rakudo: say acosh(1)
p6eval rakudo 4de821: OUTPUT«0␤»
colomon rakudo: say atanh(1)
p6eval rakudo 4de821: OUTPUT«Divide by zero␤»
colomon rakudo: say atanh(1.00001)
p6eval rakudo 4de821: OUTPUT«NaN␤»
TimToady if you get divide by zero when the limit is going to Inf, why not return Inf?
colomon rakudo: say atanh(.99999)
p6eval rakudo 4de821: OUTPUT«6.10303382276111␤»
colomon rakudo: say atanh(.9999999999) 19:00
p6eval rakudo 4de821: OUTPUT«11.859499013855␤»
colomon TimToady: Good question.
TimToady seems like we should use Inf/NaN where it makes sense, and reserve exceptions for, er, exceptions that are more exceptional... 19:01
colomon I was thinking atanh(1) ought to be NaN to match all the atanh values greater than 1, but Inf does seem to make sense there.
TimToady++
TimToady after all, we may be passing these values off to some other language that hasn't a clue about unthrown exceptions
19:02 pmurias left, pmurias joined
colomon I will try to work that into my tests (though I don't know if I can figure out the PIR to actually implement it). 19:02
19:03 donaldh left, azawawi left 19:04 donaldh joined, am0c left
Tene I really need to update some of the web stuff to use context vars where appropriate. 19:10
pugs_svn r28275 | colomon++ | [t/spec] Split the trig function by trig function info off into a separate file, trig_functions. Add tests for inverse functions on rational numbers. 19:11
19:16 sri_kraih left 19:17 sri_kraih joined, sri_kraih left 19:21 sjohnson left, sjohnson joined, sjohnson left 19:23 sjohnson joined, meppl left
sjohnson sorry about all the reconnects 19:25
19:25 sri_kraih joined 19:35 kent\n joined 19:36 ruoso left 19:37 sevvie joined 19:40 Aisling joined
pugs_svn r28276 | colomon++ | [t/spec] Add inverse Int tests as well. 19:50
colomon The trig test generator generates 6238 tests at the moment. (Not included in the official spectests yet...) 19:51
All pass on my copy of rakudo, planning on checking them in after the release today. 19:52
pmichaud (after release) +1 20:02
20:03 _jaldhar left 20:07 chturne joined
chturne #join #pugs 20:07
20:17 zerhash joined 20:25 desertm4x left 20:33 Khisanth left
pugs_svn r28277 | mimir++ | Added self to AUTHORS. 20:40
20:45 Khisanth joined 21:04 iblechbot left 21:12 zamolxes joined 21:14 cmv joined 21:23 M_o_C left
dalek kudo: 05e7604 | (parrot user)++ | :
fixed bad math
21:25
kudo: 41f7ba7 | (parrot user)++ | :
Merge branch 'master' of github.com/rakudo/rakudo
kudo: 2953ac7 | (parrot user)++ | docs/release_guide.pod:
fix typo
21:29 athomason left 21:31 SmokeMachine left, zerhash left 21:35 mberends left 21:36 kidd_ left 21:44 ihrd joined 21:47 kent\n left, ihrd left 21:48 KyleHa left 21:49 dakkar left 21:51 jauaor left
pmichaud particle++ # Rakudo #21 21:59
[particle] (thundering hoarde)++ for making the release easy this month 22:00
somehow i think of rakudo devs more as pirates than pack animals 22:01
22:03 kolibrie left, kolibrie joined 22:05 masak joined
masak o/ 22:05
sjohnson hi
masak I didn't plan to come back tonight. but I discovered a number of bugs, so I thought I'd report them before going to bed. 22:06
nothing beats writing app code for finding bugs.
sjohnson pats masak on the back 22:07
masak rakudo: sub foo(Bool :$x) {}; foo(:x)
p6eval rakudo 2953ac: OUTPUT«Parameter type check failed; expected Bool, but got Int for $x in call to foo␤in sub foo (/tmp/M4rN5Eiet4:2)␤called from Main (/tmp/M4rN5Eiet4:2)␤»
masak submits rakudobug
rakudo: class A { has $.a; method x(Str $:x) {}; method all() { $!a } }
p6eval rakudo 2953ac: OUTPUT«Invalid twigil used in signature parameter. at line 2, near ") {}; meth"␤in Main (src/gen_setting.pm:1262)␤» 22:08
masak submits rakudobug
oops, no that's not it.
rakudo: class A { has $.a; method x(Str :$x) {}; method all() { $!a } }
p6eval rakudo 2953ac: OUTPUT«No such attribute '$!a'␤in Main (/tmp/gwwL4ES70M:2)␤»
masak there you go.
rakudo: say "oh hai" .= uc
p6eval rakudo 2953ac: OUTPUT«OH HAI␤»
masak rakudo: say "oh hai" .= "uc"
p6eval rakudo 2953ac: OUTPUT«.= must have a call on the right hand side at line 2, near " .= \"uc\""␤in Main (src/gen_setting.pm:2722)␤»
masak submits rakudobug
22:10 tak11 joined 22:12 lucs_ joined
diakopter re: Cobra & Ioke thread... ships passing in the night. 22:13
22:15 Kodi joined
Kodi Could someone change the Pugs SVN username "mimir" to "Kodi", or just give "Kodi" a commit bit (email [email@hidden.address] I'd like to be to referred to by my first name in SVN logs. 22:15
masak rakudo: class A { has $.a; method x(Str $x) {}; method all() { $!a } }
p6eval rakudo 2953ac: OUTPUT«No such attribute '$!a'␤in Main (/tmp/sHzJVfEqwb:2)␤» 22:16
masak rakudo: say "oh hai"."uc" 22:18
p6eval rakudo 2953ac: OUTPUT«OH HAI␤»
masak rakudo: my $a = "oh hai"; say $a.="uc" 22:19
p6eval rakudo 2953ac: OUTPUT«OH HAI␤»
masak rakudo: my $a = "oh hai"; say $a .= "uc"
sjohnson heh
p6eval rakudo 2953ac: OUTPUT«.= must have a call on the right hand side at line 2, near " .= \"uc\""␤in Main (src/gen_setting.pm:2722)␤»
pmichaud std: my $a; $a .= "uc";
p6eval std 28277: ( no output ) 22:20
pmichaud hmmmmm
sjohnson std's been busted for a while now
pmichaud std: my $a; $a .= uc;
sjohnson std: 3;
p6eval std 28277: OUTPUT«ok 00:03 38m␤»
std 28277: OUTPUT«ok 00:03 36m␤»
sjohnson nevermind
pmichaud std: my $a; $a .= 'uc';
p6eval std 28277: OUTPUT«Potential difficulties:␤ Possible obsolete use of .= as append operator; in Perl 6 please use ~= instead at /tmp/kNhxZT6Qhy line 1:␤------> my $a; $a .= ⏏'uc';␤ Useless use of quotes at /tmp/kNhxZT6Qhy line 1:␤------> my $a; $a .= 22:21
..'uc'⏏;␤ok…
pmichaud yes, that's what I expected.
masak Matt-W: thanks for the kind words on p6l. 22:23
22:23 Kodi left
masak colomon: (re TOTE eq 'Totally Obsessive TDD Environment') sure, why not? :) 22:25
diakopter for posterity (I mean backscroll), I sent a pugs commitbit invite to Kodi as requested 22:26
22:26 alester left
Tene Me too! 22:27
22:28 patspam joined 22:29 nothingmuch left, nothingmuch joined 22:30 KatrinaTheLamia joined
Tene hi katrina 22:30
22:34 pmurias left
pugs_svn r28278 | Kodi++ | Insignificantly changed lead of AUTHORS to test new commit bit (I'm Kodi Arfer). 22:35
masak KatrinaTheLamia: \o 22:37
yay, I'm on reddit! www.reddit.com/r/programming/commen..._on_crack/
in summary, the commenters say it's been done already, it's trivial, they misread TDD twice, and they wish there was a Perl 5/.Net version. :) 22:38
diakopter 77 22:39
hm
(oops)
KatrinaTheLamia hai masak 22:41
hi Tene
22:48 nihiliad left, patspam left
masak use.perl.org does not degrade gracefully when js is turned off... editing a post turns up a blank page. 22:54
pugs_svn r28279 | diakopter++ | [vijs] non-interpolated Str can be said; TimToady++ for fixing/optimizing lots in STD/viv 22:55
diakopter vijs: say("Meesa »ö«") 22:58
p6eval vijs 28279: OUTPUT«Meesa »ö«␤␤ time in interpreter: 0.015203 s␤»
diakopter purty
22:58 lucs_ left
masak hm... even with js, use.perl doesn't permit me to edit my post. :/ 23:02
avuserow masak: I think you can blame slashcode for that 23:03
masak oh? 23:04
avuserow yeah, I don't think they have editing. slashdot doesn't at least, and they both run slashcode 23:05
masak up until today, I've been able to edit my posts just fine. 23:06
and I think I've seen updates to Slashdot posts as well.
not comments, mind. posts.
diakopter vijs: say("Meesa »ö«");say(say(' hi hi ', say 445))
p6eval vijs 28279: OUTPUT«Meesa »ö«␤445␤ hi hi 1␤1␤␤ time in interpreter: 0.045923 s␤»
avuserow masak: my mistake, thought you meant comments
masak rakudo: class A { method foo { say "A" } }; class B is A { method foo { say "B"; nextsame } }; B.new.+foo 23:09
p6eval rakudo 2953ac: OUTPUT«B␤A␤A␤»
pmichaud that looks correct to me 23:14
masak same here.
I had an old sticky note where it Null PMC'd. it doesn't any more.
rakudo: class A { method foo(::T $ :) { say T.perl } }; A.new.foo 23:15
p6eval rakudo 2953ac: OUTPUT«A␤»
masak std: class A { method foo(::T $:) {} } 23:16
p6eval std 28279: OUTPUT«ok 00:03 40m␤»
masak rakudo: class A { method foo(::T $:) {} }
p6eval rakudo 2953ac: OUTPUT«Invalid twigil used in signature parameter. at line 2, near ") {} }"␤in Main (src/gen_setting.pm:1262)␤»
masak cackles
pmichaud: is it reasonable to expect Rakudo to parse the above '$:' as '$ :'? 23:17
pmichaud is std parsing it that way? 23:18
masak good question.
it's not throwing an error, at least.
pmichaud I suspect std might be parsing it as '$:' 23:19
pmichaud looks
masak why then doesn't it complain?
pmichaud why should it? $: is valid 23:20
it's the same as method foo(::T $)
masak blinks
yes, that's my point.
er, unless it isn't. 23:21
TimToady it has a param sep of ':'
so I think it's making it the invocant
pmichaud okay
that's fine -- that's what masak++ is asking for
let's see what Rakudo is doing
masak pmichaud: why would 'method foo(::T $:)' be the same as 'method foo(::T $)'? the first one has an invocant, the other one has that plus a param. 23:22
pmichaud masak: just as $ is an anonymous placeholder variable, $: could be considered that as well depending on the parsing
TimToady $: requires a name at STD:1783
pmichaud (it's not that way, but it's possible for a grammar to treat it as such)
yes, I see that
TimToady I don't know why param_var isn't slurping it as a twigil though 23:24
pmichaud right, that's my point :)
masak pmichaud: do you think rt.perl.org/rt3/Ticket/Display.html?id=67846 would be hard to fix? I'm considering trying it myself. it's getting to be enough of an itch after tonight's hacking session. 23:25
pmichaud looks like param_var wants a \w after the <twigil>
[<?before <.twigil>\w> <twigil>]?
TimToady line 3062, yes
that would be it 23:26
pmichaud Rakudo doesn't have that check.
We can add it.
masak \o/
pmichaud: should I submit this to RT?
pmichaud sure.
masak makes it so
TimToady why do I get the feeling that colon is overloaded to within an inch of its life already?
pmichaud ...inch?
that much?
TimToady nanomicro? 23:27
pmichaud better.
masak attoparsec.
TimToady probably with a large negative imaginary component
diakopter non-breaking space
pmichaud zero-width space :-)
diakopter I mean. that.
TimToady a BOM maybe
that has a large negative imaginary component, I suspect
pmichaud yes, it's aBOMinable
masak no argument there. 23:28
diakopter pmichaud++ # often translating my thoughts to typing better than I do.
TimToady sec...
pmichaud rakudo: class A { method foo(::T $ :) { } } # just checking 23:30
p6eval rakudo 2953ac: ( no output )
pmichaud okay
let me fire up my other box and see if I can make the trivial change and spectest it while still working on $otherjob tasks :)
TimToady back 23:31
vijs: say 'hello ' ~ 'world' 23:32
pmichaud TimToady: would [ <twigil> <.before \w> ]? be a better construction?
p6eval vijs 28279: OUTPUT«execute(): execute error: undefined at viv line
..142.␤[{"T":"noun__S_value","BEG":4,"WS":1,"_specific":1,"SYM":"value","value":{"T":"value__S_quote","BEG":4,"WS":1,"_specific":1,"SYM":"quote","quote":{"T":"quote__S_Single_Single","BEG":4,"WS":1,"_specific":1,"SYM":["'","'"],"nibble"…
masak 'night, #perl6. 23:34
TimToady o\
masak that looks like someone scratching his head. :) 23:35
TimToady pmichaud: possibly, though it may be that I had that and was working around some difficulty
masak: that's because we don't know what to think when you're gone
pmichaud I know what to think: "Gee, I wonder when masak++ will be back."
TimToady well, if I said /o\ you'd feel guilty about leaving 23:36
masak :)
diakopter TimToady: lol.
masak I'll try to sleep quickly.
diakopter _o_
masak flight attendant.
23:36 masak left
diakopter |_o_| 23:37
23:37 jrtayloriv left
pmichaud |-o-| # tie fighter 23:37
TimToady sending semaphores?
»ö« <-- Darth Camelia
23:38 frew_ joined
sjohnson tie fighter heh 23:38
i remember my brother one that old DOS game in a raffle
we never played it
one => won 23:39
diakopter what a great game. I still play the theme jingles on a piano whenever I see one. kindof annoying of me, I know. 23:40
23:42 jrtayloriv joined 23:43 sevvie left
colomon I never played TIE Fighter, but I played a good bit of X-Wing vs TIE Fighter, and it was great fun. 23:54