»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.
00:11 kborer joined 00:14 TheVoiceOfEnigma joined 00:15 TheVoiceOfEnigma left 00:16 Tene joined 00:17 r0bby_ is now known as r0bby 00:25 ihrd1 left 00:28 ihrd joined, ihrd left 00:37 mrsaturn joined 00:44 kborer left
stu42j Anyone here involved with perl6.org? 00:45
s1n pmichaud: what does set_integer_keyed mean? 00:49
pmichaud: that is, i was looking at rt #68714, and i suspect i will be looking in gut.pir, but i'm not sure what the error is telling me 00:50
00:51 ihrd1 joined 00:52 Makoryu joined
carlin stu42j: most of the people here have access to modify it, what do you need to know? 00:52
stu42j I was just reading on use.perl - some people complaining about the colors... 00:53
I wanted to volunteer to help, just don't want to step on any toes.
carlin Su-Shee (who isn't here right now) did the current design 00:57
00:58 mrsaturn_ joined 00:59 mrsaturn_ left 01:09 anfedorov joined 01:14 meppl left 01:32 ihrd joined, ihrd1 left 01:34 FCO left 01:36 ihrd left 01:43 alester left 01:52 molaf_x joined 01:55 mrsaturn left 02:00 molaf left 02:01 agentzh joined 02:03 SmokeMachine joined 02:07 sharada joined 02:10 cotto joined 02:11 meppl joined 02:18 reqamst left 02:29 c9s_ joined 02:31 frew joined 02:35 colomon joined
wayland76 stu42j: Have you got a link to the use.perl article? 02:36
02:36 alester joined, PZt joined 02:41 xomas left
stu42j wayland76: use.perl.org/~sigzero/journal/39552 02:42
wayland76 stu42j: Su-shee is in Germany, IIRC, so you'll need to be on during an appropriate timezone. For me, that's evening, but I'm in Australia 02:43
Also, allow me to mention that the butterfly was personally designed by Larry Wall, and was significantly better than some of the alternatives :) 02:44
02:46 c9s left
wayland76 TimToady: If you're around, you may also be interested in commenting :) 02:49
Btw, stu42j actually has a suggested alternative: saj.thecommune.net/perl6.org/index.html 02:50
If that's the suggested alternative, then we need to speak to Su-Shee, and TimToady doesn't need to comment -- instead he can hack on the specs, or whatever it is that he does :) 02:51
phenny: Tell Su-Shee There were some complaints about he colour scheme on perl6.org. Complaints were not from p6 people, but posted at use.perl.org/~sigzero/journal/39552 02:52
phenny: tell Su-Shee There were some complaints about he colour scheme on perl6.org. Complaints were not from p6 people, but posted at use.perl.org/~sigzero/journal/39552 02:53
phenny wayland76: I'll pass that on when Su-Shee is around.
wayland76 phenny: tell Su-Shee The most constructive suggestion came from stu42j who proposed this alternative: saj.thecommune.net/perl6.org/index.html
phenny wayland76: I'll pass that on when Su-Shee is around.
wayland76 phenny: tell Su-Shee I don't mind either way myself, but I thought you might like to know. Also, stu42j is on the IRC channel when I'm writing this. 02:54
phenny wayland76: I'll pass that on when Su-Shee is around.
02:57 Makoryu left, alester left
spinclad rakudo: say 3 ne 4 03:03
p6eval rakudo 0d4fe0: OUTPUT«1␤»
spinclad (( ok then; parsed 'say (3 ne 4)' not '(say 3) ne 4' )) 03:04
03:05 tak11 joined
carlin those paler colours are a nice improvement 03:07
colomon Post-pub, Rakudo re-build, and spectest, too tired to work on Rat. In the morning before $work, maybe... 03:10
03:12 alester joined 03:20 sri_kraih_ joined 03:28 cotto left 03:35 zamolxes left, zamolxes_ joined 03:37 zamolxes_ left 03:39 molaf_x left, KatrinaTheLamia joined 03:40 sri_kraih left, sri_kraih_ is now known as sri_kraih 03:42 meppl left 03:44 xomas joined 04:03 rbaumer left, tak11 left 04:34 frew left
pmichaud use.perl.org/comments.pl?sid=43716&cid=70336 # latest (and probably last) reply to ank 04:43
04:46 anfedorov left 04:50 pyrimidine joined, bionoid left
pyrimidine pmichaud++ (for the response) 04:51
04:56 JimmyZ joined 05:12 pizzledizzle joined 05:18 pyrimidine left 05:29 JimmyZ_ joined 05:30 c9s_ is now known as c9s
mberends pmichaud++ indeed, says exactly the right things 05:33
05:33 mberends left
pmichaud woops, I ended up making one more response :) 05:38
use.perl.org/comments.pl?sid=43716&cid=70339
I think I did enough troll feeding for one day; I did try to "troll hug" a bit with the invitation for helping us figure out a threading model, but I'm not nearly as skilled at it as audreyt++ :) 05:39
05:41 nihiliad left
pmichaud oh dear, it might have worked :) 05:42
use.perl.org/comments.pl?sid=43716&cid=70340
wayland76 Are you going to tell him where he should he send the info? Or should I do it? 05:52
pmichaud Just did. 05:53
use.perl.org/comments.pl?sid=43716&cid=70342
05:57 JimmyZ left
dalek kudo: 9a37314 | pmichaud++ | docs/spectest-progress.csv:
spectest-progress.csv update: 433 files, 12465 (72.9% of 17100) pass, 0 fail
05:58
06:02 kyle_l5l joined
JimmyZ_ glad to see it. 06:11
06:11 szabgab left
spinclad (troll hugging)++ ! 06:11
carlin @karma (troll hugging) 06:12
lambdabot (troll has a karma of 0
carlin aww :( 06:13
spinclad @karma hugging)
lambdabot hugging) has a karma of 1
KatrinaTheLamia @karma hugging 06:14
lambdabot hugging has a karma of 0
KatrinaTheLamia Eris damnit! That is truly an injustice! 06:15
spinclad i think that's as well played as i've seen. turning antagonism into the prospect of synergy really changed ank's picture of the perl world, me thinks. 06:18
carlin looking at that, it seems like he changed his tune after pm explained that p6 folk are volunteers and if they weren't doing P6 they probably wouldn't be doing p5. 06:19
spinclad may this bear good, thready fruit.
KatrinaTheLamia heh... which reminds me... I need to get back into this MMO... my level 1 Perl Knave appears to have been under appreciated. 06:21
spinclad carlin: yes, i think that was an effective step 1.
KatrinaTheLamia well... time to kill some Rabbits (ie: RTFM)
moritz_ good localtime() fellow Perl 6 hackers and lurkers 06:23
sharada :) 06:24
KatrinaTheLamia hai moritz_ 06:25
moritz_ nothingmuch++ # blog.woobling.org/2009/05/your-openid-sucks.html 06:26
06:31 jauaor joined 06:42 pizzledizzle left
spinclad good verylate(), all. i'll see if i can lurk in my dreams, like audrey. 06:46
wayland76 Incidentally, I've noticed on the topic of threading that there seem to be two basic opinions floating around. 06:49
One is to do what is briefly mentioned in S17 at the moment, and have an events-underlying-threads model (which is the one I favour) 06:50
But the other model seems to be advocated by people who think POSIX threads are where it's at
My understanding is that the POSIX-threads people say that the events-threads stuff requires that the threads co-operate somehow 06:51
@karma KatrinaTheLamia
lambdabot KatrinaTheLamia has a karma of 1
moritz_ rakudo: printf "%.2f", 12465 / 17100 06:56
06:56 sharada left
p6eval rakudo 0d4fe0: OUTPUT«0.73» 06:56
06:56 sharada joined
moritz_ colomon: please use 'plan *' only sparingly in the test suite, it detoriates our statistics about how many tests we have in the test suite 06:58
07:01 iblechbot joined
dalek kudo: d2cc61c | moritz++ | tools/progress-graph.pl:
[tools] avoid overlapping labels in progress graph
07:07
07:07 moritz_ sets mode: -o dalek 07:19 szabgab joined
pmichaud we should probably find some way to be able to work with 'plan *' 07:29
or perhaps we need to get rid of the notion of computing the size of the suite (either altogether, or based on the 'plan' statements)
alester hugging++ 07:34
pmichaud: I went for brevity perlbuzz.com/2009/08/perl-6-develop...erl-5.html 07:35
and now I go to bed.
07:35 alester left
moritz_ pmichaud: maybe we can re-use the test counting logic from fudge 07:35
wayland76 I'm reminded of the Englishman who said that the verse "Greet one another with a holy kiss" should be translated "Greet one another with a hearty handshake"
07:38 Su-Shee joined
Su-Shee good morning. :) 07:38
phenny Su-Shee: 02:53Z <wayland76> tell Su-Shee There were some complaints about he colour scheme on perl6.org. Complaints were not from p6 people, but posted at use.perl.org/~sigzero/journal/39552
Su-Shee: 02:53Z <wayland76> tell Su-Shee The most constructive suggestion came from stu42j who proposed this alternative: saj.thecommune.net/perl6.org/index.html
Su-Shee: 02:54Z <wayland76> tell Su-Shee I don't mind either way myself, but I thought you might like to know. Also, stu42j is on the IRC channel when I'm writing this.
sjohnson hi Su-Shee and wayland, and others
KatrinaTheLamia huggles sjohnson 07:39
Tene japhb: Sorry for the delay. I finally added eval to nqp. It's pretty basic, and the language you're evaling needs to follow HLL interop conventions.
pmichaud Tene: is eval implemented as a keyword?
japhb Tene, Thank you!
pmichaud or as a library? 07:40
Tene pmichaud: compilers/nqp/src/builtins.pir
japhb Hopefully will be able to do more Plumage hacking next week, $day_job permitting
pmichaud oh.
KatrinaTheLamia hmm.. I could see it being used as a library as being useful.
pmichaud I'm trying to keep the number of builtin functions really small. But I might be able to accept eval for now. 07:41
(no, I'm not about to back it out... just thinking what I would like instead, if anything)
sjohnson hugs KatrinaTheLamia back :3
KatrinaTheLamia particularly is eval had more object behaviour than keyword behaviour... seriously, I could think of some interesting ideas for an eval object.
Tene pmichaud: I have no attachment to it whatsoever. Just the simplest thing that could possibly work. I'll re-do it to comply with whatever feedback you give me.
japhb Su-Shee, one thing I noticed is that the less saturated mockup of the perl6.org front page also didn't wrap weirdly like perl6.org (used?) to. 07:42
pmichaud okay.
I hadn't thought about it as builtins.pir, but I think that's likely okay for now.
Tene I don't really understand the layout of nqp very well anyway, but I didn't want to delay any longer. :)
wayland76 hugme: hug KatrinaTheLamia and sjohnson 07:43
hugme hugs KatrinaTheLamia
japhb pmichaud, seems best to minimize the keyword count in NQP, I would think.
wayland76 hugme: hug sjohnson
pmichaud Tene++ # excellent job, btw
wayland76 Hugs by proxy :)
pmichaud yes, I want to keep the keyword count low in NQP
moritz_ japhb: there were tweeks to the other (non-color) parts of the perl6.org layout recently. Maybe the decoloured version doesn't have these fixes yet
Tene japhb: you talked about json... is there a json compiler that exists as a language?
pmichaud well, more to the point, I don't want NQP to have keywords that aren't also keywords in Perl 6 :)
cognominal jnthn, are your around?
KatrinaTheLamia wayland76: you made a bot specifically for the purpose of hugging? BRAVO!
Tene I see compilers/json/, but that doesn't get installed as a language. 07:44
wayland76 No, I didn't. But it's there
But I might've just broken it :/
moritz_ KatrinaTheLamia: no. The primary purpose is of hugme is to add contributors to projects
KatrinaTheLamia bad wayland76... no cookie for you.
japhb Tene, yes, I believe that the existing JSON reader is implemented as a language. But I hadn't tried to use it yet. I don't know why it's not installed ....
moritz_ hugme: list projects
hugme moritz_: I know about these projects: json, perl6-examples, proto, svg-matchdumper, svg-plot, tufte
moritz_ hugme: help 07:45
hugme moritz_: 'hugme: (add $who to $project | list projects | hug $nickname)'
KatrinaTheLamia moritz_: ah... okay...
hugme add KatrinaTheLamia to SSBZ
hugme KatrinaTheLamia: sorry, I don't know anything about project 'SSBZ'
wayland76 hugme: hug sjohnson
hugme hugs sjohnson
Tene japhb: compilers/json certainly doesn't get installed as a language.
wayland76 ok, it works now :)
KatrinaTheLamia hmmm... well, I suppose I need to register SSBZ somewhere >.> 07:46
japhb Tene, I was agreeing with you. I was just saying I didn't happen to know *why* that was the case. :-)
Tene Ah. :)
japhb: likely because there are few people that know anything about hll interop, and json doesn't look like it's seen work in ages, so has been inherited from a time before that mattered. 07:47
japhb ah so.
KatrinaTheLamia moritz_: where do I go to add a project... I mean, I won't be listing SSBZ for a little bit (due to procrastination), but it would be nice to have people interested in it.. apart from "that crazy fan game Katrina is doing" 07:48
moritz_ KatrinaTheLamia: where do you want to list it? is a Perl 6 project?
KatrinaTheLamia Well, SSBZ was originally a Ruby project that I really got no where on (I don't think I got any actual coding done on it). Then I read up on Perl 6, and figured, that it would do nice on Perl 6. Especially since I don't expect it to be really mature any time soon. I mostly got the idea from the Iron Man blogs that Perl5 and Perl6 need more games for them. For Perl5 I am working on KREPES (Kid Radd Engine Project Existance System, I 07:51
07:52 kidd joined
moritz_ (truncated after "System, I") 07:52
KatrinaTheLamia I think the acronym ended up as..) and for Perl6 I decided to have house SSBZ. 07:53
heh... I really hate it when clients do that >.>
moritz_ anyway, if you want to add your project to the "proto" installer, just tell me or masak the URL of the project, and one of us will add it
(or anyone of the 100 other proto comitters :-)
KatrinaTheLamia proto installer? Okay... is that the base distribution of Perl6... or is that some sort of package management system for Perl 6? 07:54
moritz_ it's the prototype of a package management in Perl 6 07:55
07:55 rfordinal joined
moritz_ github.com/masak/proto/tree/master for more details 07:55
KatrinaTheLamia ah, okay ^.^ 07:56
KatrinaTheLamia still needs to start working her way around github btw >.>
moritz_ don't work your way around it - embrace it :-) 07:57
anyway, you can also use gitorious if you prefer that over github
KatrinaTheLamia well, I am still new to most of you gits ~.^
japhb Who was going to work on private attribute declaration in NQP? Was that Tene or jnthn? 08:00
Tene It wasn't me.
japhb All right, jnthn then.
Tene japhb: does that eval look sufficient for your purposes for now? 08:01
japhb Hopefully I'll cross paths with him tomorrow; in the mean time, I need to get to bed nowish. Thanks again, Tene++
Tene kk 'night
pmichaud omg itz 3am 08:02
japhb Tene, only had time to pull parrot, don't have time or brains to really look at it now, sorry. Will try tomorrow afternoon, if the planets align.
pmichaud I should get to sleep also
Tene japhb: just leave me a moose eventually.
pmichaud spent too much time working on Rats and writing use.perl comments :-|
Tene Rats? 08:03
pmichaud rakudo: say Rat.new(3,4)
Tene Ah.
p6eval rakudo 0d4fe0: OUTPUT«Could not find non-existent sub Rat␤»
pmichaud rakudo isn't updating?
Tene The gf has been talking about the rodent kind of Rat recently. >.>
KatrinaTheLamia reading through the Perl6 op chart and I could see the code `something orelse die` appear a lot >.> 08:04
pmichaud we tend to think in terms of 'fail' more than 'die' 08:05
KatrinaTheLamia I was thinking more of "GIVE ME THIS OR PERISH!" for that type of usage. 08:06
pmichaud but yes, 'orelse' fills that testing niche
sjohnson hugme: hug wayland
hugme hugs wayland
sjohnson :)
hugme: hug wayland76
hugme hugs wayland76
sjohnson that's better
Tene pmichaud: what tasks are on your list for this week?
KatrinaTheLamia well... orelse works better than what a lot of people do in other languages for that sort of thing... strikes me as cleaner.
08:07 ejs joined
Tene starts looking into db access from rakudo 08:07
08:07 jauaor left
pmichaud Tene: I'm still working out the details, but mostly I expect to be cleaning up a bit more operator stuff and then diving into nqp issues 08:07
(mainly grammar and regex related stuff)
Tene Okay, thanks. 08:08
pmichaud if you're wanting to do db access from rakudo, I'll definitely want to follow along :)
08:08 reqamst joined, jferrero joined
pmichaud I'm interested in some NCI stuff. 08:08
Tene pmichaud: You have interest in that too? I was doing it for masak.
pmichaud I'm more interested in the NCI aspects than the db ones 08:09
Tene If you have any specific interest, please express it.
Okay.
Will commits be sufficient for your curiosity, or do you want blog posts too?
pmichaud I always prefer blog posts, because (1) publicity is good and (2) others will read blogs
but that's a "do as I say not as I do" sort of thing, so I happily take whatever I get :) 08:10
because I'm notoriously bad about blog posting my own work. I even failed on IronMan last week, so I'm back to paperman status :(
moritz_ so am I
pmichaud (in my weak defense, I did have to go out of town and entertain family a lot during that week)
moritz_ masak++ is bronze
pmichaud at YAPC::EU there was a "Space Invaders" demo written in Perl 5 SDL. I've been wondering what it would take to make a Perl 6 version of it :) 08:12
Tene Someone posted a breakout clone in rakudo sdl recently 08:13
moritz_ tuits, mostly
Tene it ran very very slowly.
pmichaud yeah, I saw that it was done, and that it was really slow
that's something we could probably work on over time :)
github.com/domm/Game-PerlInvaders/
Tene He used some custom patches to parrot's SDL, iirc. I'll see about merging those back into the appropriate place. 08:16
pmichaud also, on the topic of db access from rakudo: www.asciiville.com/2009/08/topm-per...rhack.html 08:18
Tene okay, the existing mysql client library works from Parrot (but segfaults on exit?) 08:20
ah, segfaults off somewhere in libmysqlclient. 08:21
the existing mysqltest.p6 just segfaults.
08:21 ejs1 joined 08:23 payload joined 08:24 sjohnson sets mode: +oooo Tene wayland76 Su-Shee japhb
pmichaud okay, I definitely need to get to sleep. be back in a few hours 08:27
Tene Yay, segfault in get_pmc_proxy!
colomon pmichaud: is the div operator hooked up to Rat yet?
Tene I still don't really understand the pmc_proxy stuff. :)
08:28 mikehh_ left
moritz_ rakudo: say (1 div 3).WHAT 08:29
p6eval rakudo 0d4fe0: OUTPUT«infix:<div> not yet implemented in Rakudo, sorry␤in sub infix:div (src/gen_setting.pm:1592)␤called from Main (/tmp/YB7i2N0cvD:2)␤»
moritz_ that should be an easy fix.
08:29 ejs left
colomon rakudo: say new.Rat(1,3) 08:30
p6eval rakudo 0d4fe0: OUTPUT«Could not find non-existent sub new␤»
colomon rakudo: say Rat.new(1,3) 08:31
p6eval rakudo 0d4fe0: OUTPUT«Could not find non-existent sub Rat␤»
pmichaud p6eval isn't up-to-date 08:32
08:33 kyle_l5l left 08:34 kent\n left, kent\n joined, sjohnson sets mode: +v p6eval
pmichaud and afaict, commit 0d4fe0 is from 2009-08-13 08:35
08:35 hah joined
pmichaud on another topic, it might be nice for there to be some sort of link from perl6.org to 08:36
moritz_ maybe I didn't start cron in the chroot after reboot
pmichaud svn.pugscode.org/pugs/misc/camelia.txt
or some other page that explains the butterfly a bit more :)
08:36 agentzh left 08:37 agentzh joined
carlin make camelia clickable and link to there 08:37
pmichaud it might be worth writing a page that describes camelia
moritz_ aye 08:38
as soon as we have a layout that supports subpages
Su-Shee++ is working on that
Su-Shee she does. yesterday I wasted 5 hours on a broken gtk to get a recent inkscape to do shiny things.... very productive. 08:40
moritz_ well, Perl 6 development is optimized for fun, and so is the development of the accompanying websites 08:42
08:42 jaffa8 joined
jaffa8 hi 08:42
In Perl 6 regular expressions, how do I negate the effect of i 08:43
?
pmichaud :i(0)
:!i should also work, but it may be NYI in Rakudo
jaffa8 what is NYI?
pmichaud "Not Yet Implemented"
jaffa8 rakudo: /:!i do it/ 08:44
p6eval rakudo 0d4fe0: OUTPUT«perl6regex parse error: Quantifier follows nothing in regex at offset 11, found ':'␤in Main (src/gen_setting.pm:3454)␤»
pmichaud nyi
jaffa8 rakudo: /:i(0) do it/
p6eval rakudo 0d4fe0: ( no output )
jaffa8 rakudo: "DO IT"~~/:i(0) do it/; print $() 08:45
p6eval rakudo 0d4fe0: ( no output )
jaffa8 rakudo: "DO IT"~~/:i(1) do it/; print $()
p6eval rakudo 0d4fe0: ( no output )
Tene jaffa8: you want $/
pmichaud rakudo: "DO IT" ~~ /:i(0) do it/; print $/;
p6eval rakudo 0d4fe0: ( no output )
Tene rakudo: "DO IT"~~/:i(1) do it/; print $/
pmichaud rakudo: "DO IT" ~~ /:i(1) 'do it'/; print $/;
p6eval rakudo 0d4fe0: ( no output )
rakudo 0d4fe0: OUTPUT«DO IT»
pmichaud rakudo: "DO IT" ~~ /:i(0) 'do it'/; print $/;
p6eval rakudo 0d4fe0: ( no output ) 08:46
jaffa8 ok
pmichaud okay, I'm really leaving now. Computer shutting down, even :)
Tene 'night pm
jaffa8 SHould not $() work?
Tene I've never seen $(). 08:47
Is it new?
jaffa8 rakudo: "DO IT" ~~ /:i(1) 'do it'/; print ();
moritz_ no, it's old
jaffa8 rakudo: "DO IT" ~~ /:i(1) 'do it'/; print $();
p6eval rakudo 0d4fe0: ( no output )
rakudo 0d4fe0: OUTPUT«DO IT»
Tene Oh.
moritz_ what used to be $() is now spelled $/.ast
afaict
Tene ah 08:48
moritz_ but it seems to still work
don't know if it's specced that way
jaffa8 ast?
moritz_ abstract syntax tree
jaffa8 rakudo: "DO IT" ~~ /:i(1) ('do it')/; print $();
p6eval rakudo 0d4fe0: ( no output )
jaffa8 So $() should print ast. 08:49
moritz_ currently p6eval is being borked
jaffa8 rakudo: "DO IT" ~~ /:i(1) ('do it')/; print $/;
p6eval rakudo 0d4fe0: OUTPUT«DO IT»
jaffa8 rakudo: "DO IT" ~~ /:i(1) ('do it')/; print $0;
p6eval rakudo 0d4fe0: OUTPUT«DO IT»
jaffa8 borked?
moritz_ it's not reliable atm
jaffa8 What does it have to do with this? 08:50
moritz_ jaffa8: that if you write 'rakudo: $program' and you get no output, that might be the cause
jaffa8 I did not write that.
moritz_ anyway, I have a patch locally here that implements infix:<div> for ints 08:51
spectesting now
08:52 araujo left
jaffa8 $() did not print anything 08:52
That is a bug I guess
moritz_ I just explained that if p6eval doesn't print anything, it might not be rakudo's fault.
jaffa8 ok 08:53
moritz_ rakudo: say 1
p6eval rakudo 0d4fe0: OUTPUT«1␤»
jaffa8 rakudo : say 4/4
rakudo: say 4/4
p6eval rakudo 0d4fe0: OUTPUT«1␤» 08:54
jaffa8 rakudo: "DO IT" ~~ /:i (1) ('do it')/; print $0; 08:56
p6eval rakudo 0d4fe0: ( no output )
jaffa8 rakudo: "DO IT" ~~ /:i(1) ('do it')/; print $0;
p6eval rakudo 0d4fe0: ( no output )
jaffa8 rakudo: "DO IT" ~~ /:i(1) 'do it'/; print $0;
p6eval rakudo 0d4fe0: ( no output )
08:56 zamolxes joined
jaffa8 rakudo: "DO IT" ~~ /:i(1) 'do it'/; print $/; 08:56
p6eval rakudo 0d4fe0: ( no output ) 08:57
jaffa8 rakudo: "dddd" ~~ /d ** 3../; print $/; 09:00
p6eval rakudo 0d4fe0: ( no output )
jaffa8 rakudo: "dddd" ~~ /d ** 3.. /; print $/;
p6eval rakudo 0d4fe0: ( no output )
moritz_ rakudo: say "basic sanity"
p6eval rakudo 0d4fe0: ( no output ) 09:01
moritz_ p6eval's brokenness should be fixed in a few minutes
sjohnson rakudo: say 1+1 09:02
p6eval rakudo 0d4fe0: ( no output )
KatrinaTheLamia rakudo: say "I hear wayland76 loves having tea parties with his dollies and talking about all the latest gossip"
jaffa8 do you know that d ** 3.. is not implemented?
p6eval rakudo 0d4fe0: ( no output )
moritz_ jaffa8: why don't you just install rakudo locally and try it out?
carlin it's not broken, it's pining for the fjords
sjohnson wayland76, is this... *gasp*... true?
( `ー´)
jaffa8 Are you aware of the fact that of d**3 not implemented? 09:03
KatrinaTheLamia oh yes, wayland76 wears a really fetching hat. It is most becoming of him ^.^ </cutethulhu_reference>
Tene Hmm. /me is not familiar with cutethulu. 09:04
jaffa8 moritz_,I have rakudo on my machine. 09:05
moritz_,Are you aware of the fact that of d**3 not implemented?
KatrinaTheLamia Tene: I'll find the video
09:05 cotto joined
moritz_ jaffa8: so, what does your local test tell you, is it implemented? 09:06
Tene looks like google knows where it is.
wayland76 That's right, talk about me behind my back. Admittedly I have tea parties, but they usually involve me wearing a top hat and sitting next to a dormouse
moritz_ rakudo: say 'dddd' ~~ / d ** 3 / && say $/
p6eval rakudo 0d4fe0: OUTPUT«ddd␤1␤»
moritz_ rakudo: 'dddd' ~~ / d ** 3 / && say $/
p6eval rakudo 0d4fe0: OUTPUT«ddd␤»
jaffa8 moritz_,not.
rakudo: say 'dddd' ~~ / d ** 3.. / && say $/ 09:07
KatrinaTheLamia Tene: well for those too lazy: www.youtube.com/watch?v=SP-33XI4frs << it is a fairly old clip on the web.
p6eval rakudo 0d4fe0: OUTPUT«perl6regex parse error: Error in closure quantifier at offset 34, found ' '␤in Main (src/gen_setting.pm:3454)␤»
jaffa8 Can you see it now?
rakudo: say 'dddd' ~~ / d ** ..3 / && say $/ 09:08
p6eval rakudo 0d4fe0: OUTPUT«␤»
moritz_ jaffa8: you asked about **3 being implemented or not. That's not quite the same
Tene KatrinaTheLamia: I consider myself enlightened. :)
jaffa8 it was a typo
09:08 xinming left
jaffa8 if you read my earlier questions , they are right. 09:08
wayland76 So do I. Enlightenment++ window manager (e17) :) 09:09
moritz_ ok, I am aware of it
but the RT is not
jaffa8: would you care to write a ticket for that?
jaffa8 rakudo: say 'dddd' ~~ / d ** ..3 / && say $/
p6eval rakudo 0d4fe0: OUTPUT«␤»
jaffa8 there seem to be a problem with .. 3 as well.
moritz_ aye, it's probably parsed as $term ** $term 09:10
09:10 xinming joined
moritz_ rakudo: say 'dAA3d' ~~ /d ** ..3/ 09:10
p6eval rakudo 0d4fe0: OUTPUT«␤»
moritz_ rakudo: say 'dAd' ~~ /d ** ..3/ 09:11
p6eval rakudo 0d4fe0: OUTPUT«␤»
moritz_ rakudo: say 'dAd 3' ~~ /d ** ..3/
p6eval rakudo 0d4fe0: OUTPUT«dAd 3␤»
moritz_ so it's parsed as [ d ** . ] .3
jnthn morning #perl6 09:12
moritz_ oh hai jaffa8
jnthn tab completion fail
09:13 kst left
moritz_ oh right 09:13
KatrinaTheLamia hai jnthn
moritz_ jnthn hai oh
jnthn ;-)
Tene morning jnthn
09:13 kst joined
dalek kudo: 17bcd58 | moritz++ | :
Merge branch 'master' of [email@hidden.address]
09:13
kudo: a378770 | moritz++ | src/setting/ (2 files):
implement infix:<div>(Int, Int)
jnthn wonders what excitement has been happening over the last few days while he's not been paying so much attention... 09:14
jaffa8 nothing?
Tene jnthn: I'm getting segfaults from sql stuff 09:15
jnthn: I just committed 'eval' for nqp.
09:16 casiano joined
jnthn oh wow... 09:17
nice flame war...
Tene: eval for NQP - nice. :-) Did you get anywhere with library loading there yet too? :-)
jaffa8 jnthn, what flame war? 09:19
jnthn, Do you have plans to compile rakudo on regular bases as you did with pgs? 09:20
pugs
09:20 tomaw_- is now known as tomaw`
cognominal jnthn, can you help me with my blizkost patch? 09:21
whatever the kost? 09:22
:)
moritz_ jaffa8: fperrad has win32 builds of parrot and rakudo 09:23
Tene jnthn: Oh, right, I forgot that you wanted that. 09:24
although, I'd like to hear back from pm on how he wants it implemented.
He had some concerns about where I added eval.
jnthn Tene: I think japhb++ wanted it for working on the library installer stuff... 09:25
Tene pmichaud: Any concerns about 'use' in NQP?
jnthn jaffa8: No - Francois Perrad already doing a fine job of those, so no need to duplicate effort.
Tene jnthn: That's what I thought too, but then it turned out he wanted 'eval' more, or something.
I dunno. That was like a week ago.
jnthn Tene: Ah, OK. 09:26
Tene Much longer than my attention span.
jnthn I've not had time to follow things too closely of late.
cognominal: Thanks for the patch - I didn't have time to look at it just yet.
Tene it looks like I need to finally learn the pmc_proxy stuff in parrot.
I've been avoiding it for a while.
jaffa8 they are not built on daily basis as they are used to.
Tene but, now for bed.
'night
09:27 ejs2 joined
jnthn Tene: 'night 09:28
jaffa8: Ah, OK
moritz_ well, if you want pre-built versions, stick to the releases (just my opinion) 09:29
jnthn Aye, once a month ain't bad. 09:30
If somebody wants to set something more frequent up, it's probably not so bad.
09:30 maerzhase joined
moritz_ but it requires more thought 09:30
jnthn There's a makefile target in Parrot and Rakudo to make a Windows installer, IIRC.
So it may not be so much effort. I don't have a Windows box that is on every day (e.g. when I'm not at home) though. 09:31
09:32 ejs1 left
jaffa8 How did you do last time 09:32
?
jnthn jaffa8: Had some Perl script that mostly automated it 09:33
But this was way back, before Parrot had make install for example.
cognominal: Your patch - it appears to be against some earlier version of the PMCs, rather than being a complete patch to add them? 09:38
cognominal: That is, I don't see them at all in the git repo.
09:39 mikehh joined
pugs_svn r28127 | moritz++ | [t/spec] some small unfudging and fudge changes 09:40
09:41 agentzh1 joined, agentzh left
cognominal jnthn, I will check that 09:42
09:43 fridim_ joined
jnthn cognominal: ok 09:43
cognominal: would you like a commit bit, or prefer to send patches?
cognominal jnthn, anyway my problem is only with perl5.pir, I corrected one mistake, then I hit the problem 09:44
09:44 dakkar joined
cognominal A commit bit will be nice. 09:44
s/will/would/
moritz_ hugme: list projects 09:45
hugme moritz_: I know about these projects: json, perl6-examples, proto, svg-matchdumper, svg-plot, tufte
moritz_ no blizkost yet :-)
jnthn hugme?
hugme please
moritz_ jnthn: hugme is a bot that hands out commit bits to github projects
jnthn ...meh, as responsive as my admiree.
wayland76 pays attention to comments about commit bits, and commits to biting people :)
moritz_ hugme: hug jnthn
hugme hugs jnthn
wayland76 hugme: hug jnthn
hugme hugs jnthn
jnthn :-D 09:46
wayland76 hugme: help
hugme wayland76: 'hugme: (add $who to $project | list projects | hug $nickname)'
moritz_ jnthn: the only glitch is that the project own has to give me his github API key to do so
but so far it works for me, the 'perl6' user and masak++
s/own/owner/ 09:47
jnthn cognominal: Your gh ID?
moritz_ (btw "blitzkost" in german is a (IMHO bad) translation of "fast food" :-)
jnthn moritz_: lol 09:48
cognominal jnthn, cognominal
moritz_ lunch&
jnthn moritz_: At least nobody yet found a language wehre it means something really offensive. :-)
cognominal: added
cognominal: Thanks for hacking on Blizkost. :-) 09:49
cognominal I know some of the internal of Perl 5 and reading you code about roles. Seems to be a goodmatch for Blizkost
jnthn cognominal: Aye, well, we need to just get some of the basics working first. :-) 09:50
ruoso hugme, list projects 09:52
hugme ruoso: I know about these projects: json, perl6-examples, proto, svg-matchdumper, svg-plot, tufte
ruoso hugme, add ruoso to smop 09:53
hugme ruoso: sorry, I don't know anything about project 'smop'
mikehh rakudo (a378770) builds on parrot r40887, make test / make spectest (up to r28126) PASS - Ubuntu 9.04 i386 (g++)
ruoso hugme, ad project smop
hugme, add project smop
hugme, hug ruoso 09:55
hugme hugs ruoso
cognominal jnthn, I just don't get this error of not finding the pmc
jnthn cognominal: Me either, I know you're not the only person to run into it though 09:56
cognominal: Will try and spend some time on Blizkost in the next few days
09:57 kst left, kst joined
cognominal jnthn, no hurry 09:57
jnthn, you should create a google group 09:58
jnthn cognominal: Google group?
09:59 Ryan52 joined
svarg so no more haskell questions entertained in here? 10:00
cognominal jnthn, groups.google.com/ so as to have the searchable mail list 10:01
10:02 kidd left
jnthn cognominal: May be an idea, so there's at least some official place for people to post build problems / ask questions other than having to hunt down somebody on IRC. 10:03
OTOH there's already the GitHub Issues tracker.
cognominal I don't know Github that well. Anyway, that's your call :) 10:11
jnthn cognominal: Will ponder it, thanks for suggesting. :-)
OK, taking visitor out to see a castle, bbl 10:12
10:25 mikehh left 10:26 donaldh joined 10:29 pmurias joined
pugs_svn r28128 | daxim++ | [perl6.org] muted colours by stu42j 10:31
r28128 |
r28128 | use.perl.org/comments.pl?sid=43716&cid=70315
10:33 maerzhase left
pmurias ruoso: hi 10:34
moritz_ ruoso: hugme only knows about github repos
10:34 sri_kraih left, sri_kraih joined
ruoso commute & 10:35
10:35 mikehh joined, ruoso left, cotto left 10:41 masak joined
masak greetings, #perl6 10:42
masak backlogs
moritz_ oh hai masak
masak o/
moritz_ now that we have muted colors on perl6.org, can we go back to a white background? 10:44
masak we can try. 10:46
wayland76 ...and <blink> tags? :)
10:46 szabgab left
masak @slap wayland76 10:46
lambdabot puts on her slapping gloves, and slaps wayland76
moritz_ wayland76: hush
wayland76 hugme: hug everyone
hugme hugs everyone
10:47 carlin left
masak a long-time proponent of milder colours on that page, I must say I like this colour scheme much better than the previous. 10:47
10:47 carlin joined
pugs_svn r28129 | moritz++ | [perl6.org] now that we have muted colors, use white background again 10:48
r28129 |
r28129 | If you think this screams at you too much, feel free to revert this commit.
moritz_ ok, site updated 10:49
sharada hey it's got round rectangles!!
this is really Perl 6.0
masak there's something wrong with the Ironman Part of the twitter feed perlnews. it keeps repeating mst's early ironman posts. 10:50
s/P/p/ 10:51
moritz_ so does the ironman website, no?
10:52 iblechbot left
masak dunno. 10:55
yes. 10:56
maybe something wrong with mst's feed?
svarg whats with the camelia spokesbug
masak what about it? 10:57
moritz_ it's our logo
svarg logo for perl6?
moritz_ aye
svarg erm
moritz_ "erm", it's not macho enough?
svarg is that the final logo?
masak thinks it's macho enough 10:58
moritz_ svarg: nothing in Perl 6 is really final yet
masak svarg: I think no-one would complain if you tried your hand at making a logo to your liking.
svarg ok 10:59
let's see what i can come up with
the page looks nice, straight to the point
carlin an alternative logo would be nice (not to replace camelia but as an alternative) 11:00
svarg an alteernative? no, that'd be of no use
camelia is not the final piece right so why not try something else? 11:01
wayland76 svarg: No-one may complain, but there will be bikeshedding :)
moritz_ svarg: we have multiple implementations - why not multiple logos? :-)
svarg ok what do you need logos for? 11:02
carlin Camelia is nice from a fun and friendly perspective, one from a clean and professional perspective would be good
moritz_ svarg: to put it up on websites, mostly
or on books, if you write them
svarg camelia is friendly and fun but this is perl, you want to keep it a lil more reserved?
moritz_: yeah but i mean what kinda logos do you require 11:03
moritz_ svarg: we have a logo. The question is what kind of logo *you* want
for Perl 6
svarg ok
camelia is the only perl6 logo you have now righ? 11:04
i'll work on an alternative
moritz_ rakudo has a logo, but it's only for rakudo itself, not for Perl 6
rakudo.org/sites/default/files/logo.png 11:05
svarg you want an alternative for rakudo as well?
moritz_ I have no idea; if one falls out of the other efforts, it might be nice. But in the end that's pmichaud's decision
svarg alright i'll work on something, show it you in a day or two 11:06
wayland76 Well, there was an alternative to the Perl 6 logo, but it was a nightmare
masak svarg: the question 'you want an alternative for rakudo as well?' presupposes the (false) image of one single coherent group of people all wanting the same thing. :) 11:07
wayland76 IIRC, it had a couple of lion pictures stolen from the 'Net, with the slogan "Parallel lions never meet" or something :)
masak wayland76: you're just making that up. :P
11:07 flip913 joined
wayland76 masak: No, this is true -- I made it myself :) 11:07
masak ah :)
flip913 If I've got a class with (eg) an array of somethings in it, can I pass its Iterator role as the one of the class? Ie. so that iterating over the class object does iterate over the array in it. 11:08
Without re-defining all the methods needed for the Iterator role, that is.
wayland76 Yes
I don't know if it's implemented 11:09
moritz_ bugs.debian.org/cgi-bin/bugreport.cgi?bug=544399 # ITP: rakudo -- implementation of Perl 6 for Parrot
masak probably not implemented, no.
wayland76 But you do class Bar does Iterator {} class foo { has $!Bar handles ??? }
I think
But not ??? 11:10
That's me forgetting what goes there
Try S12 for more info 11:11
moritz_ or S07
but most of that really isn't implemented in Rakudo right now
svarg masak: heh, im sure there's going to be a difference in opinion but let me get started on it and show it to you
wayland76 But you do class Bar does Iterator {} class foo { has $!Bar handles Iterator; }
That's how it goes
flip913 wayland76: that's it? 'handles' ... I'll have to look at that. 11:12
masak svarg: there's always a difference in opinion. that just means people care. that's a good thing.
wayland76 It can do methods or roles or regexes that match methods :)
wayland76 loves handles
svarg masak: exactly, criticism might ask me to do more of them
ask/force 11:13
moritz_ in scruffy (ruby) I found this curious piece of code:
rendertime_renderer.instance_eval { before_render if respond_to?(:before_render) }
can I translate that to $rendertime_renderer.?before_render ? 11:14
svarg whats the caption for perl 6? 11:18
does it have one? 11:19
moritz_ I don't think there's one
svarg do you need one?
moritz_ so far I haven't found one that does Perl 6 justice
11:20 donaldh left
svarg ok you wont be needing then? :P 11:20
if you guys need it let me know
11:21 donaldh joined
carlin Perl 6: Now with more ways to do it 11:21
moritz_ well, having one wouldn't be bad
I was recently looking for something to put in perl6.org's <title> tag
and I ended up just writing 'Perl 6'
svarg ok
k will work something out 11:22
11:22 flip913 left
moritz_ hey, what about "Perl 6 - your friendly programming language"? 11:22
now let the bikeshedding begin! 11:23
svarg for the title tag i guess just perl6 will do:)
wayland76 svarg: If you want ideas, there was a long discussion about logos on the perl6-language mailing list
moritz_: Goes well with the Camelia theme :)
svarg yeah wanted to see some examples of logos previously done and rejected 11:24
moritz_ I think I have link somewhere in one of my tidings posts 11:25
let me see if I can find it
svarg yeah please
thanks
moritz_ www.nntp.perl.org/group/perl.perl6....31263.html 11:26
the rest of the thread might also be of interest to you
svarg wxactly what i wa slooking for 11:28
thx
moritz_ you're welcome
11:29 zamolxes left
svarg oh you will need to duplicate this in ascii as well? 11:32
challenging
heh
moritz_ I think the ASCII-part is not so essential 11:33
svarg from that perspective camelia looks nice
thats cool makes it more interesting, i'll do an ascii version
for rakudo whats your ascii version? just the japanese fonts? 11:34
11:34 szabgab joined
Su-Shee
.oO(still camelia not manly enough..? ;)
11:34
moritz_ s/ASCII/Unicode/ but yes :-)
svarg ok
11:38 zamolxes joined, SmokeMachine left
wayland76 Hmm. I tried making a version of camelia-in-a-suit 11:38
moritz_ slashdot.org/submission/1063859/Per...w-homepage somebody tried to /. me 11:39
now I'm truly impressed
carlin moritz_: :) I think it's been skipped over though :( 11:40
moritz_ aye, thankfully. I don't think my server could have handled that. 11:41
wayland76 I tried to slashdot myself once. It didn't work :)
Su-Shee well at least I won't be famous on slashdot for candy-pink and gay web style. ;) 11:42
11:42 SmokeMachine joined
moritz_ there are worse things to become famous for 11:44
wayland76 thinks of Michael Jackson 11:45
Su-Shee hence the ;) - in 5 days noone cares anymore anyway.
moritz_ but the 5 days of traffic are enough to throw you out of google adsense :-) 11:46
11:46 hami joined, hami left 11:47 casiano left, SmokeMachine left
moritz_ Su-Shee: he, somebody found my page by searching for "flow charts of cold process in making soap" on google images - somehow I'd expect your page to turn up instead :-) 11:49
Su-Shee moritz_: aeh, ooookeyyyy. :)) I would have thought so too. :) 11:50
I could change my meta to a better connection of "soap making" and "perl 6"... 11:51
(whenever I read certain thread I really get the impression that people see perl 6 as a personal affront...) 11:53
11:53 sharada left
carlin the same thing happened with Python 2 vs 3. PHP 4 vs 5. 11:56
It's the internet :) You could find people who see kittens as a personal affront if you tried hard enough :/
Su-Shee that's true.. but I really don't get not being able to laissez faire.. 11:57
12:03 sri_kraih left
diakopter did someone mention an ascii edition of Camelia? 12:04
masak s/ascii/Unicode/ :)
diakopter oh 12:05
masak »ö« (see topic)
Su-Shee :)
pmurias moritz_: why will traffic get you thrown out of google adsense?
diakopter at one time I was able to see such things in putty/screen/irssi... but I broke it.
Su-Shee why do I have a @ and you lot don't ...
12:06 Su-Shee sets mode: +o masak
masak gracias. 12:06
12:06 takadonet joined
takadonet morning all 12:06
12:06 ChanServ sets mode: -o Su-Shee 12:07 ChanServ sets mode: +o Su-Shee
moritz_ pmurias: extreme spikes in traffic make google suspicious, they think that fraud is involved 12:07
pmurias: I've heard multiple times that being slashdotted resulted in being excluded from adsense 12:08
12:09 jrtayloriv left
frettled OT: does anyone know where I can find Jesse Vincent's monospace Klingon font? Google is not helping. :( 12:10
moritz_ just ask him :-) obra_: feel hilighted (see above) 12:11
12:12 agentzh1 left, agentzh joined
frettled moritz_: aha! I didn't know the nick/name mapping for him. 12:12
moritz_: (thanks) 12:13
takadonet who got slashdotted? 12:14
Su-Shee noone (yet) :)
mayme moritz_ will. ;)
moritz_ but I think it's unlikely 12:15
12:17 payload left
pugs_svn r28130 | colomon++ | [t/spec] Update power.t to have real plan, un-todo stuff that now works in Rakudo, todo test that doesn't work (and may be incorrect, IMO). 12:19
colomon moritz_: How do you add test files to make spectest? I don't think I ever got power.t properly included... 12:21
moritz_ colomon: that can only be done in the rakudo repo
colomon moritz_: Where? I need to add the new rat.t i just created too, I'll make a patch with both... 12:22
moritz_ colomon: t/spectest.data 12:23
colomon moritz_++ 12:24
moritz_ colomon++ # actually doing stuff 12:25
colomon BTW, on the plan * stuff -- agree that it shouldn't be long term, but it's super handy when making a lot of changes on a test file. 12:27
I'd be willing to help try to toughen the test-counting code to deal with it, too.
moritz_ colomon: writing a simple Perl 5 script that counts the tests and add that to the test suite would be a help 12:28
I agree that 'plan *' is handy, and we shouldn't let minor toolchain problems prevent us from using it 12:29
colomon Is there a reason you can't just run the .t script and get the number of tests reported? plan * still gives that info... 12:31
moritz_ well, that works only if the test can actually be run
colomon Actually, it would have to be the make test equiv, wouldn't it? 12:32
But I would think checking tests that don't run in is a much bigger sin....
12:33 reqamst left
moritz_ most tests that aren't run are actually wrong anway 12:34
I tried it a few times, but it's nearly impossible to write more than a handful of trivial patches for NYI-features
12:35 agentzh1 joined, agentzh left 12:37 payload joined 12:38 Confield joined 12:39 kst left, kst joined
masak rakudo: say 360.0.sin 12:39
p6eval rakudo 0d4fe0: OUTPUT«0.958915723414307␤»
masak rakudo: say 360.sin
12:39 reqamst joined
p6eval rakudo 0d4fe0: OUTPUT«Method 'sin' not found for invocant of class 'Int'␤» 12:39
masak that'll confuse the "Int is a kind of Num" people. 12:40
colomon I've been pondering whether Int ought to have .sin.
moritz_ aye, I think it's a bug
rakudo: say (1/2).WHAT 12:41
12:41 d3w left
p6eval rakudo 0d4fe0: OUTPUT«Num()␤» 12:41
moritz_ rakudo: say (1 div 2).WHAT
p6eval rakudo 0d4fe0: OUTPUT«Rat()␤»
wayland76 rakudo: say 1.sin(mortal)
p6eval rakudo 0d4fe0: OUTPUT«Could not find non-existent sub mortal␤»
wayland76 rakudo: say 1.sin($mortal)
p6eval rakudo 0d4fe0: OUTPUT«Symbol '$mortal' not predeclared in <anonymous> (/tmp/DuLGNI9Eyj:2)␤in Main (src/gen_setting.pm:3454)␤»
wayland76 rakudo: say 1.sin(my $mortal)
p6eval rakudo 0d4fe0: OUTPUT«Method 'sin' not found for invocant of class 'Int'␤»
colomon moritz_, masak: Int.sin is not in the spec. But then, neither is Complex.sin, and I've already started working on that. :) 12:42
masak rakudo: say Int ~~ Num
p6eval rakudo 0d4fe0: OUTPUT«1␤»
12:42 iblechbot joined
masak colomon: thing is, it shouldn't have to be... 12:42
...in the spec.
12:43 jrtayloriv joined
colomon I can trivially add it to the test suite if that is your guys opinion. 12:44
(I mean, I have the trig.t file opened in the exact right spot and everything.)
masak that's my guys' opinion, at least.
moritz_ go for it 12:47
masak rakudo: class Foo { method bar { my @a = 1; undefine @a; } }; say Foo.bar 12:48
p6eval rakudo 0d4fe0: OUTPUT«Null PMC access in type()␤in Main (/tmp/ApxK0UzQAZ:2)␤»
masak carlin: nice catch. did you report that one? 12:49
carlin doh, no. still got the started bug report saved in my drafts 12:50
colomon moritz_: Back on modifying the list of spec tests -- now I'm getting a failure in 99problems-31-to-40.rakudo that I didn't get before?!?
carlin finishes it
masak carlin++
moritz_ colomon: that shouldn't happen :/
colomon: what kind of failure?
colomon: oh, and send the patch anyway :-)
colomon "No plan found in TAP output
"
Will try to send patches along as soon as I've finished my cereal. 12:51
diakopter ? Perl6 backend for incubator.apache.org/thrift/
moritz_ sounds like a rather neat idea 12:52
12:52 stu42j left
diakopter would certainly get attention from the Ks of fb devs 12:52
12:53 KatrinaTheLamia left
wayland76 I wonder if it could be integrated with Web.pm :) 12:54
diakopter @karma @karma 12:55
lambdabot @karma has a karma of 0
diakopter @karma--
lambdabot usage @karma(+|-) nick
pugs_svn r28131 | colomon++ | [t/spec] Introduce rat.t file to test Rat type, with a few first simple tests.
diakopter @karma- @karma
lambdabot @karma's karma lowered to -1.
masak er, not quite 100 committers to proto yet, though that would be kinda cool. somewhere between 20 and 30, though. 13:02
I'm bronze? \o/ how can you tell? 13:03
13:04 abra joined
moritz_ masak: by reading the ironman feed :-) 13:05
masak so it doesn't actually say "bronze" anywhere? you just calculated it manually? 13:06
colomon You can ask it to show your icon, and that will be bronze (if you are bronze).
masak how? where?
13:07 DanielC joined, jferrero left
colomon I saw it last night, give me a minute. 13:07
masak sits tight
moritz_ masak: www.shadowcat.co.uk/blog/matt-s-tro...munger-up/
masak moritz_: cool, thank! 13:08
colomon moritz_++
masak I'll be blogging tonight, by the way. it'll be about efficient labyrinth-making, and it'll be kinda nice.
moritz_ ironman.enlightenedperl.org/munger/.../masak.png
masak I'll be experimenting with the form of the post, and will try to make it "literal Perl 6"
moritz_ masak: I hope with cool SVG charts along the way? :-)
masak moritz_: hm, that's an idea :)
yayimbronze 13:09
PerlJam good morning 13:10
Su-Shee hi PerlJam 13:11
carlin finally sent bug report
Couldn't think of a way to word the subject properly
colomon rakudo: say 360.0.sin('degrees') 13:12
p6eval rakudo 0d4fe0: OUTPUT«-2.44921270764475e-16␤»
masak maybe I should write a 'how to word subject lines in bug reports' blog post, too... 13:15
colomon rakudo: say "360".Num;
p6eval rakudo 0d4fe0: OUTPUT«Method 'Num' not found for invocant of class 'Str'␤»
masak my algorithm is something like this: (1) start with '[BUG]' (or '[TODO]'), (2) ???, (3) end with 'in Rakudo'.
colomon Funny, I assumed the 'in Rakudo' part was assumed if you were sending the e-mail to rakudobugs. 13:16
moritz_ too
carlin didn't use "in Rakudo"
moritz_ my magic incantation is something along the lines of [TODO] implement $feature | [BUG] doing $thing gives me $bad_error_message 13:17
13:19 Vetinari` joined
colomon rakudo: say (+"360").Num 13:19
p6eval rakudo 0d4fe0: OUTPUT«Method 'Num' not found for invocant of class 'Float'␤»
moritz_ from the initial ironman announcement: "There will be a post of the month competition: a committee of good Perl writers and bloggers, who will look over the posts and pick one out to be honoured each month. The winner will get a limited edition T-shirt, and possibly other prizes from sponsors."
masak carlin: I think I'm pretty alone in doing "in Rakudo". it's almost useless, since Rakudo is the only implementation that uses RT.
carlin I should probably report the segfault too 13:20
but I'm not sure what's going on there
colomon say Num("360")
rakudo: say Num("360") 13:21
p6eval rakudo 0d4fe0: OUTPUT«invoke() not implemented in class 'Float'␤in Main (/tmp/uStjiDgG99:2)␤»
13:22 agentzh1 left
jaffa8 rakudo: m'hello' 13:24
p6eval rakudo 0d4fe0: OUTPUT«Statement not terminated properly at line 2, near "'"␤in Main (src/gen_setting.pm:3454)␤»
jaffa8 rakudo: m'hello';
p6eval rakudo 0d4fe0: OUTPUT«Statement not terminated properly at line 2, near "';"␤in Main (src/gen_setting.pm:3454)␤»
jaffa8 rakudo: m?hello?; 13:25
p6eval rakudo 0d4fe0: ( no output )
jaffa8 rakudo: "hello"~~m?hello?; print $/;
p6eval rakudo 0d4fe0: OUTPUT«hello» 13:26
jaffa8 rakudo: "hello"~~m(hello); print $/;
p6eval rakudo 0d4fe0: OUTPUT«Could not find non-existent sub hello␤»
jaffa8 rakudo: "hello"~~m&hello&; print $/;
p6eval rakudo 0d4fe0: OUTPUT«hello»
13:26 ruoso joined
pugs_svn r28132 | colomon++ | [t/spec] A number of additional sin tests, including a batch of Num.sin($base) tests which do not work for mysterious reasons, and Int.sin and Complex.sin tests which do not work because they are not yet implemented. 13:31
13:32 meppl joined 13:35 hah left, ejs1 joined 13:40 szabgab left 13:41 SmokeMachine joined 13:43 abra left
masak rakudo: module A::B; A::B::foo 13:43
p6eval rakudo 0d4fe0: OUTPUT«invoke() not implemented in class 'Undef'␤in Main (/tmp/cq9feKAC4R:2)␤»
masak I know I haven't declared my 'foo', but why the 'Undef'?
moritz_ buubot: eval: package A::B; A::B::foo 13:45
buubot moritz_: A::B::foo
moritz_ buubot: eval: package A::B; A::B::foo()
buubot moritz_: ERROR: Undefined subroutine &A::B::foo called at (eval 36) line 1.
13:45 ejs2 left
masak much more awesome error message, by me. 13:45
moritz_ masak: apparently A::B::foo autovivifies to Undef
masak but shouldn't undeclared things be not-found subroutines? 13:46
moritz_ they should, IMHO
masak submits his first rakudobug in a while
frettled yay
moritz_ masak: welcome back to the living 13:47
masak oh, I've been alive alright. :)
but thanks. :)
moritz_ :-)
pmurias ruoso: how are we going to handle the situation in smopp5 when a interpreter.goto is called from within an XS subroutine 13:48
13:48 iblechbot left 13:50 M_o_C joined
masak rakudo: sub foo() { "OH HAI" }; say foo("one argument, not zero") 13:51
p6eval rakudo 0d4fe0: OUTPUT«OH HAI␤»
masak o_O
moritz_ that's known and reported.
masak oki.
still very unsettling. :)
moritz_ rakudo: multi sub foo() { "OH HAI" }; say foo("one argument, not zero")
p6eval rakudo 0d4fe0: OUTPUT«No applicable candidates found to dispatch to for 'foo'␤in Main (/tmp/ru1RMaNdsv:2)␤»
13:57 ruoso left 13:58 _jaldhar_ joined
pugs_svn r28133 | colomon++ | [spec/t] Add a few more simple Rat tests. 14:00
14:01 rfordinal left 14:02 ruoso joined 14:11 jaldhar_ left 14:12 mberends joined 14:13 szabgab joined
mberends rakudo: "􏿽xBB􏿽xF6􏿽xAB rulez!".print 14:14
p6eval rakudo 0d4fe0: OUTPUT«��� rulez!»
masak botched latin-1 rulez? 14:15
mberends masak: hi, and yes 14:16
masak mberends: o/, and :P
I guess that should have been »ö«. and yes, I kinda like her too.
mberends masak: :-) meeting about proto-ng in a few minutes? 14:17
masak sure.
moritz_ just call it 'proton', drop the -g :-)
14:17 KyleHa joined
masak ...or pro-2... 14:18
mberends moritz_: nice one :)
masak perl6: my $a = 5; given $a { $a = 42; say $_ } 14:19
p6eval elf 28133: OUTPUT«5␤» 14:20
..pugs, rakudo 0d4fe0: OUTPUT«42␤»
masak I'm with elf on that one.
14:20 zloyrusskiy joined
moritz_ hrmpf 14:20
masak given should have for-loop semantics, and so it should copy.
moritz_ given and for both do bindings, no? 14:21
masak how could for do a binding?
(we've discussed this before)
moritz_ rakudo: my $a = 5; for $a { $a = 42; say $_ }
p6eval rakudo 0d4fe0: OUTPUT«42␤»
moritz_ masak: both 'for' and 'given' simply pass a variable to the block 14:22
masak: and during the call the parameter is bound
masak rakudo: my @a = 1,2,3; my @b = 4,5; for @a, @b { @b = 42; .say }
p6eval rakudo 0d4fe0: OUTPUT«1␤2␤3␤4␤5␤»
M_o_C mberends: I'd just like to point out that there's already an HTML editor called proton, however I don't know wether it's trademarked.
masak if we're changing the name of proto, we could call it 'neutro'. or 'electro'. :) 14:23
jaffa8 template
14:23 Psyche^ joined 14:24 REPLeffect joined
carlin just call it bikeshedo and be done with it ;) 14:24
moritz_ masak: or neutrin :-)
carlin++
masak :)
mberends I'd like to thank everyone for attending the bikeshed meeting and officially declared it closed. 14:25
masak we're not changing the name. we're just bikeshedding for the fun of it.
mberends xactly
jaffa8 bikeshedding
? 14:26
I feel on the Moon.
14:26 sharada joined
Su-Shee suggest a butterfly-like logo for proto. 14:26
masak jaffa8: it's a term for discussing details endlessly.
jaffa8: without actually producing anything of substance.
mberends jaffa8: especially pointless details such as what colour to make it. 14:27
14:29 kst left, kst joined
jaffa8 ok 14:30
carlin bikeshedding isn't a very intuitive name for it, we should change that 14:31
masak rolls eyes
mberends masak: during our last discussion we were toying with whether sudo would be required for installs, and istr you were warm to the idea of a modules directory somewhere under $HOME 14:32
masak yes.
that seems to be the proto way to me, after some pondering.
mberends is warm to that too
DanielC jaffa8: It comes from the saying that it is easier to agree on the design of a nuclear power plant than on how to make the bike shed.
jaffa8: For the power plant people will just nod their heads and agree with whatever you put down. But a bike shed is an easy project so everyone wants to have their say. 14:33
14:34 Patterner left, Psyche^ is now known as Patterner
mberends masak: re-using a previous design guideline, the module directory should survive even if proto is removed 14:34
masak dang, I could use a LAST block right now.
Juerd At Hacking At Random 2009 the workshop building was called "the bikeshed" and had four rooms: yellow, green, red, blue. 14:35
masak mberends: yes. definitely.
mberends: so maybe we shouldn't name it something-with-proto.
moritz_ $PREFIX/lib/
masak aye. 14:36
mberends some people are discussing Perl 6 modules only (eg Rakudo, Pugs) and others are thinking Parrot modules.
moritz_: yes
moritz_ so the question is what is $PREFIX?
mberends masak++ decides
something after $HOME 14:37
jaffa8 can anyone modify pugs?
moritz_ jaffa8: anyone with a commit bit can
masak $HOME/.ecosystem/lib, perhaps?
jaffa8 Does anyone understand it code?
masak both Perl 6 and Parrot use 'ecosystem'.
jaffa8: I understand a little. 14:38
moritz_ jaffa8: probably audreyt does, mostly
and others to various degrees
jaffa8 Is pugs faster than Rakudo?
moritz_ dunno 14:39
jaffa8 I mean compilation speed
masak jaffa8: try both of them and compare. :)
jaffa8 I could not compile pugs
masak in that case, Rakudo is faster :)
moritz_ :-)
jaffa8 I got stuck with ridicolous problem 14:40
mberends masak: .ecosystem resonates well with Plumage (security certificate expired) trac.parrot.org/parrot/wiki/ModuleEcosystem
jaffa8 the setup could not find a header file
masak mberends: exactly.
and it probably doesn't collide with anything else. 14:41
jaffa8 I specified the path of the header file with -I switch but it had no effect
Is anyone familiar with this?
masak jaffa8: sorry to hear that. were audreyt here, she could probably help you.
jaffa8 it is a dream. 14:42
what is the chance of that?
mberends jaffa8: only a tiny chance. the -I option is probably not being passed to the subcommand that needs it. how about copying the header file to another directory that _is_ being used for header files? 14:45
jaffa8 I wonder what directories are used.
I do not know what directories are used. 14:46
mberends dunno, it might take you an interesting while to find it in the sources.
jaffa8 pugs seems to have a better design 14:47
it can produce all kind of output.
javascript.
mberends masak: .ecosystem it is then
masak jaffa8: pugs has been in development for some time.
jaffa8 but I guess rakudo can only output PIR 14:48
masak mberends: I half-fear the tuits needed to do the shift. I suppose fear is the wrong type of anticipation, but that's how it feels. any ideas how to lessen the blow?
14:50 kidd_ joined
mberends masak: it doesn't intimidate me a much. I'll work on it first, you can repair my damage afterwards ;) 14:50
*as much
masak sounds splendid.
I had to remove the comment about config.proto deprecation, because I started receiving bug reports (!) about it. 14:51
but I still think that's a good idea, since much of the config.proto file will likely change.
so, feel free to cherry-pick that commit.
14:51 araujo joined 14:52 jaldhar joined
mberends ah, some git-learn required here :) 14:52
14:52 zloyrusskiy left
masak git-cherry-pick means 'git, I want the commit that was made over there made over here' :) 14:52
mberends bash: git-cherry-pick: command not found # this is not your Mac 14:54
masak git cherry-pick --help # brings up the man page for me 14:55
mberends reads the help now 14:56
14:57 _jaldhar_ left 14:59 nihiliad joined 15:02 ejs1 left 15:03 alester joined 15:07 Su-Shee left
jaffa8 what is the difference between c modifier in p6 and p5? 15:07
Is there any?
and what is the difference between g modifier in p6 and p5? 15:08
alester OK, everyone stop all your Perl 6 work! 15:09
Perl 5 needs threads!
masak hugs alester 15:10
alester :-)
Sometimes I just have to get on my 3000-strong soapbox and set shit straight.
[particle]1 all your complaining about perl 5 threads won't make christmas come any sooner.
15:10 [particle]1 is now known as [particle]
jaffa8 perl 5 has threads, to my knowledge 15:11
alester jaffa8: Yes, I know, we're kidding.
jaffa8: See top story on Perlbuzz.com
[particle] jaffa8 needs a link for context
perlbuzz.com/ 15:12
15:12 rfordinal joined
pmichaud Good morning, #perl6 15:13
alester pmichaud: Please exit this channel immediately and begin work on perl 5 threads.
DanielC o/ pmichaud
alester Also, let's surreptitiously talk about the things that we wnat to start PRing about.
15:14 iblechbot joined
masak pmichaud: oh hai o/ 15:15
15:15 svarg left
DanielC alester: Perhaps not everyone in this channel is interested in perl 5 threads. I know that I'm not. 15:15
masak I'm simultaneously interested and not interested in Perl 5 threads. 15:17
DanielC I'm interested in the design of cool programming languages. 15:18
jaffa8 What does that mean?
masak languages from temperate zones, perhaps? 15:19
DanielC :-)
jaffa8 Why are you interested?
alester DanielC: It was a facetious comment.
15:20 donaldh left
alester I'm actually 100% uninterested in Perl 5 threads 15:20
and prob'ly 95% uninterested in new Perl 5 features
DanielC I'm not sure I can tell you what I am interested in anything... I'm interested in math, physics, open source and space colonization, but I could not really tell you why.
15:20 sri_kraih joined 15:21 donaldh joined
jaffa8 DanielC, you cannot tell... that is interestng. 15:22
DanielC jaffa8: I could try to rationalize and guess, but that's all it would be, rationalization.
There are a lot of interesting subjects. I studied astrophysics because the subject is fascinating to me. What more reason do I need? 15:24
jaffa8 you do not want to tell..
that is my guess.
masak hugme: hug jaffa8 15:25
hugme hugs jaffa8
DanielC is bewildered that some people expect one to rationalize one's personal likes and dislikes. 15:26
alester How can I get a make target to always rerun? I can't make it a .PHONY because the file does exist. 15:27
frettled DanielC: perhaps you're one of us «ooh, SHINY!»-people :D
alester DanielC: I don't think anyone is expecting you to rationalize anythign.
masak can I get the PID of the current process from Rakudo?
alester $$? 15:28
DanielC rakudo: echo $$
pmichaud maybe from the $*OS object, if there is one
p6eval rakudo 0d4fe0: OUTPUT«Statement not terminated properly at line 2, near "$$"␤in Main (src/gen_setting.pm:3454)␤»
frettled rakudo: say $*PID
p6eval rakudo 0d4fe0: OUTPUT«Use of uninitialized value␤␤»
frettled hmm
pmichaud looks like the spec defines $*PID but Rakudo doesn't implement it
frettled rakudo: say $PROCESS::PID 15:29
p6eval rakudo 0d4fe0: OUTPUT«Use of uninitialized value␤␤»
frettled yup
It's in S28.
alester That doesn't seem like it would be too hard to implement. 15:30
frettled rakudo: say $*PROGRAM_NAME
alester assuming that Parrot will give it to us.
p6eval rakudo 0d4fe0: OUTPUT«/tmp/93niH7BmAH␤»
frettled Hmm, does rakudo implement system()?
rakudo: say system("lsof $*PROGRAM_NAME")
p6eval rakudo 0d4fe0: OUTPUT«Could not find non-existent sub system␤»
DanielC rakudo: say system('ls') 15:31
p6eval rakudo 0d4fe0: OUTPUT«Could not find non-existent sub system␤»
frettled rakudo: say exec('echo OH HAI')
p6eval rakudo 0d4fe0: OUTPUT«Could not find non-existent sub exec␤»
carlin system is run
DanielC oh... right
frettled rakudo: say run('echo OH HAI')
carlin rakudo: run('ls');
DanielC I forgot.
carlin Won't work in safe mode
p6eval rakudo 0d4fe0: OUTPUT«operation not permitted in safe mode␤in Main (lib/Safe.pm:25)␤» 15:32
frettled hee-hee
rakudo: walk('echo slowly!')
p6eval rakudo 0d4fe0: OUTPUT«Could not find non-existent sub walk␤»
frettled darnit, I thought walking was safer than running.
carlin groans
frettled Is that my coat over there?
pmichaud rakudo: say Rat.new(3,6); 15:33
p6eval rakudo 0d4fe0: OUTPUT«3/6␤»
pmichaud rakudo: say Rat.new(3,6)+Rat.new(1,5);
p6eval rakudo 0d4fe0: OUTPUT«21/30␤»
frettled neat.
pmichaud _o/
rakudo: say 3/6 15:34
p6eval rakudo 0d4fe0: OUTPUT«0.5␤»
pmichaud that part needs fixing.
jaffa8 DanielC: All it means to me you do not know yourself.
frettled rakudo: say Num.new(Rat.new(3,6)+Rat.new(1,5));
p6eval rakudo 0d4fe0: OUTPUT«0␤»
DanielC @ignore jaffa8 15:35
lambdabot Not enough privileges
pmichaud rakudo: say (Rat.new(3,6)+Rat.new(1,5)).Num
p6eval rakudo 0d4fe0: OUTPUT«21␤»
frettled que?
pmichaud wt...?
rakudo: say (Rat.new(3,6)+Rat.new(1,5)).numerator
p6eval rakudo 0d4fe0: OUTPUT«21␤»
frettled It interpreted it as a string and stopped at the slash?
pmichaud rakudo: say (Rat.new(3,6)+Rat.new(1,5)).denominator
p6eval rakudo 0d4fe0: OUTPUT«30␤»
pmichaud something else must've happened there
checking
rakudo: say Rat.new(3,6).Num 15:36
masak in the interests of release-early: github.com/masak/perl6-literate/
p6eval rakudo 0d4fe0: OUTPUT«3␤»
15:36 M_o_C left
pmichaud rakudo: my $a = Rat.new(3,6); say $a; 15:36
p6eval rakudo 0d4fe0: OUTPUT«3/6␤»
pmichaud rakudo: my $a = Rat.new(3,6); say $a.numerator + 0.0 15:37
p6eval rakudo 0d4fe0: OUTPUT«3␤»
mikehh rakudo (a378770) builds on parrot r40892 - make test / make spectest (up to r28133) PASS - Ubuntu 9.04 amd64 (gcc)
pmichaud rakudo: my $a = Rat.new(3,6); say $a.numerator + 0.0 / $a.denominator
ohhhhh
p6eval rakudo 0d4fe0: OUTPUT«3␤»
pmichaud oops
my fault
rakudo: say 3.Num
p6eval rakudo 0d4fe0: OUTPUT«3␤»
pmichaud fixing. 15:38
frettled \o/
masak: I'm not sure whether that code is neat, scary, or both.
masak jaffa8: do you have an overwhelming need to inform people that they do not know themselves?
frettled: thank you. it's just a state machine. :) 15:39
I pondered whether to make it a grammar instead. it should be possible.
especially with variables thrown into the mix somehow.
jaffa8 masak: I was suprised that this had scared him so much.
15:39 abra joined
jaffa8 masak: to go as far as to ignore me. 15:39
masak jaffa8: I think it could be taken as an insult. 15:40
jaffa8 masak: Are you not surprised yourself?
15:40 baest left
masak jaffa8: no. though you do not indend it, you sometimes come off as slightly trollish. 15:40
15:40 baest joined
jaffa8 masak: it is hard to know where the limit of the other person are. 15:40
masak jaffa8: now, I know you have good intentions, but not everyone knows that.
frettled masak: Making it a grammar seems like a neat challenge. 15:41
masak frettled: sir, I challenge thee.
15:41 awwaiid joined
frettled masak: Oh, look over there, isn't that a diversion? 15:41
masak where? where?
frettled masak: right behind that B+ tree!
masak turns head 15:42
I don't see anything behind that B+ tree.
frettled sneaks away stealthily.
Run today, and live to run another day!
masak shouts "btw, you have commit access to perl6-literate now!" after frettled
frettled mutters an expletive that would make mst proud. 15:43
pmichaud frettled: he'd probably curse you for it.
frettled haha :D
masak my advice would be to run the test suite while refactoring. then it should be a piece of cake.
frettled mm
frettled places the idea on the ToDo list of learning :D
diakopter can p6eval pull/use a .pm from a url? 15:44
frettled that looks like something that shouldn't be permitted, and definitely not in safe mode 15:45
DanielC masak: I very much doubt that jaffa8 is anything but a troll. It is absurd to take one comment on IRC and declare that someone doesn't know himself. In my case the claim is particularly absurd, but you'd have to know me well to know that (I spend a lot of time in introspection).
masak: I *can* talk at great length about myself, my motivations and my inner mind, but I will not waste my time with a random troll on IRC. 15:46
15:48 justatheory joined
masak DanielC: you'll have to excuse me for intruding into the situation. events lately have made me interested in social IRC/blog dynamics. I personally don't believe that jaffa8 is a troll, but I'm still open to new impressions. 15:48
dalek kudo: dd7587a | pmichaud++ | src/setting/Rat.pm:
Fix Rat.Num.
masak DanielC: I do, however, believe that he managed (unwittingly) to insult you, and I do understand why.
sjohnson morning all 15:49
jaffa8 DanielC: My statement were inaccurate. there seem to be a lack of knowledge about yourself in certain area. Sometimes I forget about distances that separate people. I could also noticed that I stopped talking to you before this.
DanielC masak: It's not so much about insult. I just have better things to do with my time than justify my motivations to a random person on IRC.
masak DanielC: aye, that too. 15:50
jaffa8 DanielC: why did not you just say so?
masak jaffa8: I think it's all about protocol and context. there are some things you can say to strangers, and some things you can't.
frettled jaffa8: it should not be necessary to say so. This is an IRC channel about Perl 6, and when people don't want to discuss themselves here, they certainly shouldn't be pressured into doing so. 15:51
masak jaffa8: "I think you do not know yourself" would be one of the things you can't.
frettled Can we get back to being cozy and cuddly regarding Perl 6, Rakudo, Parrot etc. now? :D
masak oki! \o/
masak cozes, cuddles
pmichaud No! We need psychedelic attack butterflies!
moritz_ hugme: cuddle pmichaud
hugme cuddles pmichaud
jaffa8 frettled: who is pressured? noone is forced to answer a question. 15:52
DanielC jaffa8: Rather, you have seem to have a very naive and narrow-minded view. It takes a certain amount of maturity and knowledge to know the limits of understanding. When I was young I was naive like you, and I thought I knew myself better than I do. With time I learned to be more humble about what I think I know.
15:52 kst left 15:53 kst joined, cotto joined
DanielC jaffa8: You don't know many of the things you think you know. For example, your memory is a lot more fragile and fallible than you think. Likewise it is with your understanding of the roots for your likes and dislikes. 15:53
jaffa8: It is then really absurd that you would then come and tell me that I don't know myself.
masak DanielC: are you sure this fellow did not insult you? 15:55
seems to me he hit a nerve.
DanielC masak: Oh, he did. What I meant earlier is that the insult is not the main point.
masak ah.
point taken.
DanielC masak: But he did hit a bit of a nerve.
masak DanielC: I know the feeling. 15:56
DanielC I hope that what I wrote above didn't come out too angry. It was meant to be an explanation.
moritz_ alester: hm, you can use .PHONY even though the file exists, no? 15:57
DanielC But human as I am, it's hard to be dispassionate in my reply.
alester but not if I don't know which
jaffa8 DanielC: I would be eager to answer something, I was told not to. if you want pm me.
moritz_ alester: ok
pmichaud rakudo: say Rat.new(3,4).Num; 15:58
p6eval rakudo 0d4fe0: OUTPUT«0.75␤»
DanielC jaffa8: Not particularly. My wife and I are going to the gym now.
pmichaud \o/
15:58 cdarroch joined
pmichaud rakudo: say (Rat.new(3,6)+Rat.new(1,5)).Num 15:58
p6eval rakudo 0d4fe0: OUTPUT«0.7␤»
pmichaud \o/
jaffa8 Danielc: then that is it.. you know I do not want to be kicked out. 15:59
frettled pmichaud++
moritz_ pmichaud: but I think the revision number at the start of p6eval's output is wrong, I have to investigate which paths are wrong on the server
pmichaud moritz_: it undoubtedly is wrong. 0d4fe0 was an August 13 commit
masak pmichaud++ 16:00
frettled Where does p6eval find its revision number from?
moritz_ frettled: from an outdated file
frettled moritz_: aha 16:01
I thought the version number might be compiled in.
pmichaud I haven't quite figured out how to do that yet.
frettled Hmm. 16:02
carlin rakudo: say 10i;
p6eval rakudo 0d4fe0: OUTPUT«0+10i␤» 16:03
carlin rakudo: say 10i.perl;
p6eval rakudo 0d4fe0: OUTPUT«Statement not terminated properly at line 2, near ".perl;"␤in Main (src/gen_setting.pm:3454)␤»
carlin rakudo: say 10i.WHAT; 16:04
p6eval rakudo 0d4fe0: OUTPUT«Statement not terminated properly at line 2, near ".WHAT;"␤in Main (src/gen_setting.pm:3454)␤»
carlin rakudo: say WHAT 10i;
p6eval rakudo 0d4fe0: OUTPUT«Complex()␤»
frettled pmichaud: where's the code for the p6eval thingy? 16:05
pmichaud frettled: I have no idea. :-|
frettled pmichaud: ack :(
carlin the pugs repo
frettled hmm! 16:06
carlin misc/evalbot
16:06 szabgab left
moritz_ aye 16:06
I should add the build script too 16:07
16:08 bionoid joined
frettled Hmm, I see. 16:09
It tries to fetch from /home/evalenv/rakudo/rakudo_revision. 16:10
I also see the difficulty in changing the behaviour. 16:11
16:12 sevvie left
frettled An ugly hack could be to have a special command line syntax for rakudo that printed the revision at the beginning of each line, but, urgh. 16:12
Also, that will mess up everything. Hrms. 16:13
frettled envisions: rakudo OUTPUT«0d4fe0 OH␤0d4fe0 HAI␤0d4fe0 IDJOT␤» 16:14
16:15 Karthz joined
carlin there is .git/ORIG_HEAD 16:15
masak how do I die() in Rakudo without generating a stack trace? 16:16
frettled masak: with honor?
:)
carlin: that's a bit long, though, isn't it? 16:17
carlin: or is the revision printed by p6eval just the first six chars?
mberends masak: is perl6-literate the same as inside out pod?
masak mberends: I think so.
mberends: www.haskell.org/haskellwiki/Literate_programming 16:18
dalek kudo: 87f1a27 | pmichaud++ | docs/spectest-progress.csv:
spectest-progress.csv update: 433 files, 12478 (72.9% of 17110) pass, 0 fail
16:19
masak 72.9%!
carlin frettled: it could be made to only do that 16:20
I guess
takadonet come on 73%!
carlin almost 3/4 of the way there 16:21
frettled carlin: should be a simple substring extraction, yes
pugs_svn r28134 | moritz++ | [evalbot] add build script for rakudo and parrot
masak oh, and I just pushed a script that runs a literate Perl 6 script.
frettled scaryneat. 16:22
masak & # going out for a run
16:22 masak left
pmichaud the percentages being reported are actually off -- we need to fix that 16:23
16:23 Su-Shee joined
pmichaud it's not that Rakudo is passing a larger percentage of the test suite, it's that our methods for computing the size of the test suite are starting to produce smaller numbers 16:23
carlin evalbot must already do it, looking at the build script
pmichaud i.e., we can no longer look at "plan" statements for a rough estimate of the number of tests in the suite. 16:24
pugs_svn r28135 | moritz++ | [evalbot] hopefully fix the location of the rakudo_revision file 16:25
frettled \o/
16:26 cognominal_ joined, p6eval left, p6eval joined
moritz_ rakudo: say "OH HAI" 16:26
p6eval rakudo : OUTPUT«OH HAI␤»
frettled hm.
moritz_ Can't open file '/home/evalenv/p/rakudo_revision': No such file or directory at evalbot.pl line 179.
pugs_svn r28136 | moritz++ | [evalbot] not only the location, also the name of the file has changed. moritz-- 16:28
16:28 p6eval left, p6eval joined
moritz_ rakudo: say "OH HAI" 16:28
p6eval rakudo : OUTPUT«OH HAI␤»
pugs_svn r28137 | pmichaud++ | [t/spec]: Unfudge some more Rat-related tests.
moritz_ goes home and will fix that later 16:29
pmichaud "Perl 6 -- Rats, Cats, and butterflies."
Su-Shee pmichaud: you forgot the parrot. 16:30
pmichaud You're correct, I did. 16:31
carlin the cat got there first
16:32 cognominal left
carlin Night all 16:33
16:34 carlin left 16:36 Karthz left 16:38 ejs joined
jnthn rehi (for 30 mins or so) 16:41
pugs_svn r28138 | moritz++ | [evalbot] typo in file name + shorten revision number 16:43
jnthn pmichaud++: Looks like you've been doing some nice work on Rats. :-) 16:44
pugs_svn r28139 | moritz++ | [evalbot] remove double pipe
moritz_ moritz--
pmichaud jnthn: Thanks. Mostly I decided it was low-hanging fruit and if I just got something started, others could quickly pick it up and run with it for a while. :) 16:45
(came out of our local hackathon here on Saturday)
btw, from the hackathon we started tagging some rt tickets as "[LHF]" ("low hanging fruit") 16:46
these are things that someone might be able to resolve quickly with just a bit of effort
jnthn pmichaud: Ah, nice idea. :-)
16:46 zamolxes left
jnthn I'll keep that in mind when going through RT in the future. 16:46
(e.g. to tag those that I think would be LHF) 16:47
pmichaud we also decided that tickets assigned to [email@hidden.address] aren't likely to be LHF :)
jnthn :-P
pmichaud (except maybe for [email@hidden.address] :)
jnthn So long as you didn't decide assigning to [email@hidden.address] is the way to tag High Hanging Fruit. ;-)
I'll probably be around some more during the next few days. 16:49
Will try and get a Rakudo day in.
pmichaud excellent
jnthn Flying to Tokyo on Fri.
pmichaud me too -- my schedule is looking quite nice now that kids are back in school
jnthn Cool. :-) 16:50
Su-Shee jnthn: ooooh, how nice.
jnthn Su-Shee: Aye, I hope so. :-)
Su-Shee: Well, being in Tokyo will probably be the nice part rather than the 12 hours of flying. ;-) 16:51
Su-Shee jnthn: I see delicious noodle soup in your future.. :)
jnthn Mmmmm. :-)
moritz_ rakudo: say "hi" 16:52
p6eval rakudo 87f1a2: OUTPUT«hi␤»
moritz_ looks if 87f1a2 makes any sense 16:53
lo' and behold, it does!
16:55 rjh joined
moritz_ after only 5 comitts to fix two stupid thinkos/typos... 16:57
pmichaud afk # lunch 16:59
17:00 RonOreck left 17:04 Ryan52 left 17:05 masak joined 17:06 hercynium joined 17:07 szabgab joined
jnthn moritz++ # persistence :-) 17:11
erm
perseverance
...damm, my English sucks.
moritz_ self.write_to('moritz-state.txt')
jnthn ;-)
masak jnthn: oh hai! 17:13
jnthn lolitsmasak! :-D 17:14
masak: How's things?
masak jnthn: wonderful. I'm back from a very pleasant weekend in Härnösand.
jnthn: how's your things?
17:15 abra left
moritz_ closely escaped capsizing on Friday ;-) 17:15
the other boat was not so fortunate
jnthn masak: Had a friend visiting for a long weekend, which has given me excuses to wander around nice places and go to some nice food places. :-) 17:16
moritz_ (but that's not so tragic with small dinghies)
masak \o/
jnthn Ah, and for a little beer too. ;-)
moritz_ beer. Why did we expect that? ;-)
jnthn looks up where Härnösand is 17:17
17:17 jhelwig left
jnthn masak: ooooh...that's north. :-) 17:17
masak looking back at my weekend, I think I mostly consumed hugs. I have a whole month's worth with me back home. :)
jnthn: yes. it's the northernmost I've ever been. 17:18
thrice, now.
jnthn hugs++
masak we have a yearly get-together there.
jnthn was happy his source of those is back in town
Even though I'm about to leave here for a month. 17:19
I changed train at Sundsvald - that's the closest I got to there. But I was in Ostersund for a bit.
I mostly remember it raining my entire time there. ;-)
masak we had fairly nice weather, I'd say.
some rain, but quite a bit of sun too.
jnthn Aye. I'd enjoyed good weather up at Abisko the couple of days before I went there.
And I had quite nice weather in Stockholm the days after Ostersund. 17:20
So on balance, I did well...I just didn't really get to see the city at its nicest.
The regional history museum there was thankfully good though.
jnthn needs to give Blizkost some attention on this week's Rakudo day 17:21
masak jnthn: I'm still slightly puzzled that there's a place called Abisko in my country. one where you've been and I haven't, at that. :)
jnthn masak: If there's wifi, it'd be an epic place for a Perl 6 hackathon. ;-)
masak wow. 17:22
jnthn If we do it in summer, then the sun never sets, so we'd never have to sleep.
So it'd be like, really productive.
;-)
masak :D
17:22 DanielC left
colomon Doing a C++ build in one window and running make spectest in another (with a virtual Windows box running in the background) really slows down my Mac. 17:23
jnthn I'm surprised and very happy to have already given out three commit bits to Blizkost, and that @other have already way out-stripped my patch count. :-) 17:24
moritz_ TEST_JOBS=1 make spectest # just one parallel test job
jnthn On $rakudo_day I hope I can at least get some basic support for being able to use a return value from eval'ing Perl 5 code back in Rakudo. 17:27
masak that would be extremely nice. 17:34
jnthn I think it's probably the simplest thing that pushes it from "just a little toy" to "maybe useful for something" 17:35
17:35 rfordinal left 17:36 ejs left
dalek kudo: 5c1bc83 | (Solomon Foster)++ | t/spectest.data:
Add S32-num/power.t and S32-num/rat.t.
17:38
masak jnthn: nod.
17:39 nbrown left
jnthn masak: Of course, that's dangerous, 'cus that's the stage where you'll also start submitting bug reports on it. :-) 17:39
masak very likely, yes. 17:40
jnthn \o/ 17:41
masak++
OK, dinner time :-)
back later perhaps :-) 17:42
colomon Any advice for adding a sub in class namespace? I've coded a simple GCD calculation function for Rat, but I'd rather not pollute the global namespace with it. 17:44
moritz_ my sub gcd ... 17:47
although I had troubles with lexical subs in the setting
colomon: you can also put it in Rat.new
colomon local subs can be defined in methods? 17:48
moritz_ that's not what I meant
I meant that you just can put the logic into the new() method
colomon moritz_: sure, but that strikes me as inelegant. :) 17:49
let me try my sub first and see how that works.
moritz_ colomon: why? Rat is defined to be immutable
so whenever you change something, you have to create a new Rat 17:50
colomon Good point.
moritz_ so you can put the simplification logic in there anyway
colomon Still, I think I just like simple functions. :)
Think I just discovered that Rat - doesn't work yet.
(Adding tests while I wait for my C++ code to compile.) 17:51
moritz_ colomon: anyway, your next patch should add yourself to CREDITS
colomon I haven't done that anywhere yet. ;)
moritz_ (rakudo patch, I mean)
colomon I think pmichaud++ got a little carried away with the cutting and pasting in Rat -- subtraction operator is just the addition operator again with a different name. 17:52
:)
17:53 kst left
moritz_ patches + tests welcome! 17:53
17:53 kst joined 17:54 ejs joined
colomon They will be coming soon. 17:54
Huh. So actually, the easiest way to get this right might be to commit the rakudo-change in git, then roll it back locally, then get the spectest working without it, then commit that to svn? 17:55
moritz_ colomon: or just skip any new tests brutally 17:56
17:56 japhb left 17:57 sevvie joined
masak rakudo: say Rat.new(1,2) 17:57
p6eval rakudo 5c1bc8: OUTPUT«1/2␤»
masak rakudo: say Rat.new(1,2) - Rat.new(1,2)
p6eval rakudo 5c1bc8: OUTPUT«4/4␤»
masak o_O
pmichaud yeah, I might've missed a sign change 17:58
colomon I've got it fixed locally here, will push patch in a few minutes.
masak \o/
then I'll stand down the bug alert.
17:58 payload left
moritz_ pmichaud++ # laying the foundations so that the rest of the Rat work ist mostly LHF 17:58
TimToady I think the page now looks drab; maybe we should have two top-level pages: the Kansas version, and the Oz version, and people can bookmark whichever one they want
.5 * :-) 17:59
colomon moritz_: Amen.
Okay, got my sub gcd working. Give me a minute or two to sort things out and then check them in. 18:00
masak TimToady: that might be a wise idea. to me, it doesn't look drab, it looks like someone who knows which colors can go together picked the colors.
TimToady I know, we'll tweek the colors depending on whether they say perl6.org, perl6.com, perl.info, etc :)
masak :) 18:01
moritz_ perl6.biz is black and white, then
frettled perl6.name is what? :D
masak perl6.info will have a fixed-width green font on black background, and no graphics.
frettled perl6.com will have Flash?
masak no please no
TimToady well, Camelia can change the colors of her wings at will 18:02
Su-Shee I take the Oz-version though I doubt that those colors are possible on computers. ;)
TimToady also, one other quibble I have is that people keep using the "patches welcome" meme, but this is true only to a certain extent 18:03
Camelia is fairly non-negotiable by now
moritz_ TimToady: patches welcome to fix that ;-)
SCNR
pmichaud we do need a link to a page that describes Camelia a bit better
moritz_ aye
pmichaud just a sec
(fixing)
TimToady Editorial policy can't easily be decided by committee 18:04
pmichaud ...but it can be published, which makes discussions on the topic a bit easier
TimToady troo
frettled Groo's clever twin?
pmichaud right now there's no published policy, so it's easy to say "anything goes" even though it really doesn't 18:05
TimToady any twin of Groo's would be clever by comparison
frettled pmichaud: Yep. A published policy would be nice to point towards. 18:06
TimToady pugs/misc/camelia.txt was a start on that
pmichaud hmm. I tried making "Camelia" (in the header text) a link to camelia.txt, but it doesn't visually look right. 18:07
maybe a link from "butterflies" instead
moritz_ or or just the first Camelia 18:08
18:08 Woody2143 left
TimToady another thing that should be pointed out is that this is just the Perl 6 logo, not necessarily the logo of any distribution of Perl 6. 18:08
Red Hat and Ubuntu don't use Tux. 18:09
colomon Rat patch sent.
pmichaud I'll push what I just did, others can patch
pugs_svn r28140 | pmichaud++ | [perl6.org] Create some links to camelia.txt .
TimToady also, the complaint that it looks juvenile is quite accurate, because it's *intended* to look that way as a sign of renewed youth
moritz_ is sorry for brining such pulitic discussion to #perl6 18:10
pmichaud svn.pugscode.org/pugs/docs/feather/...index.html
moritz_ pmichaud: looks good
s/good/fine/
pmichaud oh, here's another fix
(working) 18:11
frettled :)
[particle] the download link should explicitly mention rakudo, because right now it looks like rakudo *is* perl 6 18:13
TimToady someone needs to start Perl6 University so we can get perl6.edu. :)
pugs_svn r28141 | pmichaud++ | [perl6.org]: Revise link for Camelia in "Specification"
cognominal_ jnthn, I added blizkost to ohloh : www.ohloh.net/p/blizkost/
TimToady [particle]: I was thinking that would likely be part of Su-Shee's redesign
pmichaud svn.pugscode.org/pugs/docs/feather/...index.html # reload
I changed the entry for "Camelia" in the Specification
moritz_ what about perl6.mmil? :/
pmichaud it used to be "camelia.pdf"
now it talks more directly about Camelia herself, with a link to the PDF 18:14
frettled pmichaud++
moritz_ colomon: spectesting your patch
TimToady the text still says PDF
moritz_ (and I removed some trailing spaces)
pmichaud ...text still says PDF?
frettled I think the change is only in svn right now. 18:15
pmichaud it hasn't pushed to perl6.org yet -- there's a delay
frettled It will happily migrate to the front page when the winds are right, some butterfly has to flap its wings a bit first.
moritz_ TimToady: there are two links - one says PDF and goes to a PDF
TimToady ah, I see
frettled Let's overexplain it once more, just for -Ofun ;) 18:16
ruoso gives up trying to make the perl6.org site work for people with larger fonts
it's broken again
TimToady hopefully that can also be dealt with in Su-Shee's redesign
[particle] maybe we need some ruby folks.
18:16 Woody2143 joined
pugs_svn r28142 | colomon++ | [t/spec] More simple Rat tests. 18:16
moritz_ ruoso: since the design will be replaced soonish there's not really a point 18:17
pmichaud I also made a link from the logo itself to the camelia.txt page 18:18
Su-Shee I log everything.
frettled panics
moritz_ Su-Shee: so do; I see /topic :-)
masak on my screen, the design always looked wonky due to my browser window not being wide enough. instead of three boxes per line, I have two. :/
sri_kraih is thinking about drawing a "badass" version of camelia :)
ruoso masak, that's my case as well, because of the larger fonts 18:19
TimToady for all their problems, tables do a better job of keeping the text in the boxes
frettled strangely enough, the width is no problem on my netbook
it's the height :)
masak ruoso: we're both suffering from widebrowserism, then.
ruoso TimToady, tables break accessibility 18:20
frettled mobile browsers (earlier known as "phones") and netbooks are "forcing" web designers to rethink their web pages
moritz_ ruoso: not true
TimToady then make perl6.info accessible :)
moritz_ ruoso: I know blind people, and in general they don't have problems with tabled layouts
masak giveupandusetables.com/ 18:21
frettled Su-Shee: One wish for the redesign: please place a download link near the top of the page, where people will see it quickly :)
pmichaud within the PmWiki community we went through very long and heated discussions about the merits of table versus non-table layouts
frettled pmichaud: and ended up doing plaintext? :D
TimToady let's throw 361 RFCs at Su-Shee and see if she takes six months to recover like I did :)
pmichaud the final answer was that tables weren't ideal, but they were often less evil than the alternatives (more)
frettled TimToady: point taken :D 18:22
pmichaud of course, it's quite different in a wiki context because the content isn't static
most approaches to making layouts table-less also assume that the relative sizes of the components are known and/or fixed
frettled I like floating elements, but there are still differences in rendering that are less than trivial to fix.
pmichaud but the conclusion that we came to as a group was that making good-quality layouts using tables was far easier than doing table-less layouts 18:23
moritz_ they don't have to be rendered identical; they just have to be rendered all OK
frettled I did some work on this as part of my master studies in 2000-ish, and things haven't improved much regarding floating elements since then.
(but they have improved)
pmichaud in particular, table-based layouts are much easier to maintain over time
frettled yes, except for vertical placement
pmichaud anyway, that's my experience on the topic 18:24
moritz_ anyway, I'll leave that decision to th(os)e who actually implement it
TimToady it would be nice to have context-sensitive floating elements that know how they got floated and change things that don't influence how the float was calculated, such as color
frettled mm
18:24 kst left
pmichaud TimToady: maybe we should throw some cups and work on redesigning CSS :) 18:24
frettled Recent CSS and AJAX can help there. 18:25
18:25 kst joined
TimToady I wouldn't mind the 3-blocks to 2-blocks problem nearly so much if the blocks could change their color 18:25
Su-Shee frettled: the download link was placed right in the middle of the page, but since more links were added, the layout started to scroll.
TimToady frettled: I remind myself that CSS also could stand for Cross Site Scripting
frettled TimToady: we use XSS for that these days :) 18:26
[particle] blocks can change their color, if js in enabled
pmichaud I've alw.... what frettled++ said
[particle] we can determine browser width, and load the proper css file
pugs_svn r28143 | moritz++ | [t/spec] unfudge Rat test that colomon++ has fixed in Rakudo
TimToady well, js is okay by me, but flash is too far
moritz_ votes against JS 18:27
frettled I couldn't agree more regarding Flash.
moritz_: JS is nice if the site works without it.
pmichaud (couldn't agree more) perhaps you're just not trying hard enough
:)
frettled That is, menus etc. should continue to work if JS is disabled.
[particle] right
dalek kudo: 4fc254d | (Solomon Foster)++ | src/setting/Rat.pm:
Add gcd function and use it to reduce Rats to simplest form. Also fix - operator.
frettled pmichaud: :)
[particle] and the site should work if css is disabled, too 18:28
pmichaud 0
+ $numerator /= $gcd;
0
+ $denominator /= $gcd;
won't work.
moritz_ uhm, why?
pmichaud because infix:</> of two Ints produces a Rat
TimToady you need div
moritz_ oh. 18:29
colomon Errr... I thought infix / produced a Num, and infix div produced a Rat?
that certainly seems to be how things work today?
TimToady see most recent change to S03
colomon rakudo: say (1 / 3).WHAT
p6eval rakudo 5c1bc8: OUTPUT«Num()␤»
TimToady I decided Rats pretending to be Nums was probably a bad idea 18:30
colomon rakudo: say (1 div 3).WHAT
p6eval rakudo 5c1bc8: OUTPUT«Rat()␤»
TimToady just last week
pmichaud S03:698
C<< infix:</> >>
If both operands are of integer type, the operator returns the
corresponding C<Rat> value.
TimToady is more consistent with the use of 3/7 as a Rat literal too. 18:31
18:31 japhb joined
ruoso still unconfortable about the idea that Num doesn't include all types of numbers 18:31
TimToady and we didn't have a way to get at the integer division primitive
"include" is ill defined 18:32
colomon Errr... does that imply that everything we've done with Rat has div and / backwards?
pmichaud colomon: it might.
ruoso means the "Complex ~~ Num" sense of "include"
pmichaud I hadn't been looking at infix:<div> much
colomon Boy, it sure does.
TimToady we need to come up with a set of roles that describe behaviors, but I think those are a higher level than Int, Num, and Rat
colomon Any objections to me taking ten minutes to swap those usages, then? 18:33
(It sounds much more appealing than my $work at the moment.)
pmichaud no objections
ruoso why can't we use the math definition?
moritz_ colomon: no, you even get 15 ;-)
pmichaud TimToady: possible fossil at S02:636 18:34
moritz_ pmichaud++ # reviewing commits
TimToady fixing 18:35
pmichaud my interpretation "Complex ~~ Num" being true would be that everywhere that one could use a Num argument, a Complex argument would be handled as well. I think this would tend to confuse a lot of end-programmers. 18:37
18:37 hercynium left
masak nod. 18:37
pmichaud if I write sub foo(Num $x) { ... } I wouldn't expect $x to be able to be a Complex
or, if Num "includes Complex", then I want a type that excludes them.
moritz_ like, Real 18:38
pmichaud (and it would probably have to be called "Float")
I wouldn't want "Real" because floating-point representations aren't really Real :)
ruoso Real is the math name
TimToady Flt
ruoso and it does make a lot of sense
moritz_ the problem with "Float" is that people will ask where the Doubles are
pmichaud I always read "Flt" as "flight", though.
ruoso pmichaud, that's the point where I think we need to split conceptual types from storage types
Float is a storage types 18:39
*type
TimToady Int, Num, and Rat are Storage Types (abstract)
ruoso Real is a conceptual type
TimToady I repeat, define the math roles at a higher level than that, please
ruoso TimToady, I'd replace Num by Float in that case
TimToady Imaginary ~~ Numeric is fine
Float is ugly
that's why it's Num 18:40
pmichaud Numeric++
ruoso I'm fine if the Math types have long names
colomon Okay, I've just gotten div for Rat out of the spectests, as far as I can tell. Of course, this will break the world until Rakudo catches up (which I will attempt next). Should I check it in now? Seems unwise...
TimToady and Real can be Rat|Num
ruoso Numeric, Complex, Imaginary, Real, Integer, Natural
TimToady that is, both Rat and Num does Real
pmichaud colomon: you can always temporarily fudge the tests 18:41
ruoso TimToady, that's what I mean, yes
colomon pmichaud: True. Suppose I should go ahead, do that, check things in, and then look at fixing the operators themselves?
ruoso in fact we have both Rational and Irrational as subsets of Real 18:42
and Integer is a subset of Rational
pmichaud colomon: the other approach would be to submit the test changes along with the rakudo patch :)
TimToady Int, Num, and Rat are like the intermediate layer of a neural net that crossbars the inputs to the outputs
pmichaud colomon: still another approach would be to get you a Rakudo commitbit :)
colomon pmichaud: You mean at the same time, via the two different methods of patching? Or generating a patch file for pugs as well.
pmichaud colomon: generate a patch for pugs as well 18:43
it's just "svn diff"
colomon Sure, but I have a commit bit there! ;)
ruoso www.mathsisfun.com/sets/images/number-sets.gif
TimToady that is, pragmatic abstractions like Rat are mediating between the mathematical abstractions and the machine types
pmichaud yes, but by submitting them together you can make sure the tests get updated the same time as rakudo
TimToady so the third layer is appropriate 18:44
ruoso TimToady, agreed... so we have three sets of types
pmichaud anyway, I'd personally go with the fudge approach if it were me
ruoso Numeric, Complex, Imaginary, Real, Irrational, Rational, Integer, Natural
pmichaud maybe we should just call the types N, Z, Q, R, I, and C :-P 18:45
ruoso pmichaud, we could use the proper unicode symbosl
pmichaud even .... exactly :)
jeekobu Transcendental, Algebraic
colomon ruoso++ for a good kind of insanity
pmichaud rakudo: say "\c[DOUBLE-STRUCK CAPITAL R]" 18:46
p6eval rakudo 5c1bc8: OUTPUT«ℝ␤»
TimToady just keep your mits off my TLTs
jeekobu Computable
pmurias please don't make perl 6 use silly letters 18:47
TimToady define "silly letter"
ruoso TimToady, TLT?
TimToady Three Letter Types
pmichaud afk for a bit
ruoso pmurias, the math roles are going to have specific uses, I think it's pretty much ok to have them with the unicode symbols 18:48
I even think the mathematicians will thank us
there isn't a symbol for Numeric, I guess 18:49
moritz_ is there any numeric type more general than Complex? 18:50
18:50 molaf joined
TimToady the roles should be spelled out, but we can make it easy to get the symbols, as with Sets 18:51
18:51 Zloyrusskiy joined
TimToady I think this should probably be general policy 18:51
pmurias molaf: surreal numbers
sorry moritz_
18:51 masak left
colomon moritz_: Yes, but it's sort of orthogonal to the divisions you guys have been looking at. 18:51
TimToady maybe a pragma saying "whenever you have the choice of importing Unicode symbols, do so"
pmurias moritz_: i was wrong sorry
colomon Basically you define really abstract concepts and see what fits. 18:52
pmurias use i-have-a-fancy-keyboard;
18:52 payload joined
colomon Like "groups" and "rings". 18:52
moritz_ you mean like groups, rings, vector spaces, hilbert spaces
allright
ruoso ℂ, 𝕀, ℝ, ℚ, ℤ, ℕ
18:52 Zloyrusskiy left
pmurias are those distinct letters or just a traditional way of writing them 18:53
18:53 kst left
ruoso pmurias, they are specific to the math use 18:53
colomon I guess I'd like to see a more formal proposal for this. The geek part of me loves getting the unicode symbols in there, but it's not at all clear to me how you can usefully program using this information.
sharada what about H
18:53 kst joined
ruoso TimToady, what about Complex? are you going to define a TLT for it? 18:54
TimToady it's been considered in the past
sharada
TimToady but none of them look very good
pmichaud colomon: for most programs I suspect one wouldn't need to use these higher-level types 18:55
ruoso TimToady, because there isn't another name for the conceptual set\
18:55 masak joined
TimToady R_I :) 18:56
ruoso RnI 18:58
alester OK, who has the serious psychological issues/ Fess up!
colomon pmichaud: I guess what I'm trying to say is that for most things, it seems like the knowledge that this numeric type forms a ring is much more useful than the knowledge it is a complex number. 18:59
and it seems like a lot of work "just" to give yourself the ability to have, say, complex numbers defined using Rats rather than Nums. 19:00
ruoso N2D
Number with 2 Dimensions
colomon Actually, maybe that's not the greatest example, because I can see room for a "complex" role no matter what.
TimToady ruoso: no, the dimensions are wonky :) 19:01
from a vector-space point of view
literal alester: I'm working on automated testing for vim-perl, will commit it to my fork soon, just so you know
alester AWESOME
you saw my ticket?
literal yeah
alester literal: That is SO SO SO SO SO KICKASS THANK YOU 19:02
I mean, assuming you do it. :-)
colomon Hey, I've got the rat.t spectest running again.
alester Otherwise it's just SO KICKASS for trying.
colomon Now let's see how badly I blew up the rest of the world.
literal I've only got a few small kinks to work out, but I have to go play some pool, so I thought I'd let you know I've got something
alester But if you can pull it off, you get four extra SO
TimToady I have enough trouble with one 19:03
literal :)
19:04 ezra joined
ruoso TimToady, C99 does define a complex type, but I'm not sure there is platform specific native support for math with complex numbers... 19:04
maybe there isn't a point in a non-high-level Complex type 19:05
colomon Hmmm... maybe the right way to state my objection is, what functions make sense to define on Int that one should not define on Rat? In other words, why would the abstract roles be different?
19:05 |Ryan52 joined
ruoso hmm... no... I'm wrong... 19:05
every number ~~ Complex
but not every number ~~ R_I
because R_I would mean it actually has two slots (one for the real part, other for the imaginary) 19:06
TimToady yes, the pragmatic types imply something about slots, without overspecifying the slots 19:07
that seems like a good way to look at it
but I do think the math folk would rather write Complex and have it pun into a class with a default implementation of R_I, as it were 19:09
ruoso hmm
colomon sharada: Ah, yes, you were pointing out the obvious set of numbers that contains Complex. I forgot that H was Hamilton numbers (quaternions).
ruoso it does make a lot of sense...
PerlJam wonders if he's a "math folk" then
ruoso because Rat is simply a number with two slots as well
and so is Float 19:10
19:10 M_o_C joined
TimToady but also a relationship between the slots 19:10
ruoso sure... the type defines how many slots and what they mean
then you have the actual implementation types (native?) 19:11
TimToady some of them are considered native 19:12
but there are potentially many implementation types that look more like objects
gets fuzzy
native more or less means the hardware will get upset if you try to add slots to this 19:13
19:13 zamolxes joined
ruoso TimToady, native to Perl 6, I mean, not to the hardware 19:14
TimToady derivation is constrained not to change representation
19:14 RonOreck joined
ruoso yeah... that's the point... the vm needs to be able to do calculations with them 19:14
TimToady augment int32 { has $.foo } # kaboom 19:15
ruoso I think it's quite sane to constrain the changes in representation of number implementation types
so we leave some room for optimization 19:16
you can customize the other two layers already
ruoso .oO( thinking about slots, it makes even less sense to call Float as Num, because it's actually two numbers.. ) 19:17
colomon Errr... if Float is two, then the standard Complex implementation is four, no? 19:19
ruoso colomon, 2 slots that *might* hold 2 slots each 19:20
19:20 donaldh left
ruoso colomon, it's not that Complex itself holds 4 slots 19:20
after all, you can have a Complex built with two Integers
1 + 1i
19:20 donaldh joined
ruoso but Complex as the conceptual type doesn't imply the number of slots 19:22
since 1 ~~ Complex
R_I does
(I think that name will stick) 19:23
moritz_ still doesn't know what R_I is supposed to mean
ruoso Real and Imaginaryh 19:24
19:24 cotto left 19:28 masak left
colomon moritz_: When I said ten minutes, I wasn't factoring in the fact that the spectests take longer than that to run by themselves on my machine. Sigh.... 19:36
moritz_ colomon: no hurry. 19:37
colomon Well, I should be doing $work at some point here... 19:38
ruoso TimToady, do you think we have a spec-candidate definition already? 19:39
moritz_ rakudo: 3 % * 19:40
p6eval rakudo 4fc254: OUTPUT«Ambiguous dispatch to multi 'infix:%'. Ambiguous candidates had signatures:␤:(Any $a, Whatever $b)␤:(Any $a, Whatever $b)␤in Main (/tmp/zLrp2K91Qt:2)␤»
moritz_ OH HAI; SOMEBODY BROKE MY SVG::PLOT MODULE
or something ;-)
19:41 jaffa8 left 19:43 |Ryan52 is now known as Ryan52
moritz_ dammit, I should include every feature I use in the test suite somewhere 19:44
colomon Wow, running make spectest has generated two rakudo's both of which are currently using 1+ GB of real memory (2+ GB of virtual memory) and almost all my CPU cycles. 19:45
And hey, core dump!
Parrot VM: PANIC: out of mem! 19:46
Looks like that was arith.t. 19:47
moritz_ that's usually an indication for an infinite loop
ruoso decommute &
19:47 ruoso left
pugs_svn r28144 | moritz++ | [t/spec] test for RT #68894, $number % * 19:49
pmichaud looks like there's lots of copy-paste errors in Whatever.pm 19:50
19:50 Vetinari` left
pmichaud fixing 19:50
colomon I'm guessing the entire section of Int / in arith.t is probably wrong with the current spec?
moritz_ colomon: that could very well be the case 19:51
19:54 masak joined
pmichaud my @a = <a b>; 19:54
my $t = join '', map { @a[$_ % *] }, 1..5;
is $t, 'ababa', '$_ % * works';
19:54 sevvie left
pmichaud ...shouldn't that be 'babab' ? 19:54
moritz_ ah, Whatever.pm line 65 should be s/Whatever/WhateverCode/
pmichaud: or 0..4, right
pmichaud anyway, I have Whatever.pm fixed locally, spectesting now 19:55
pugs_svn r28145 | moritz++ | [t/spec] fix $_ % * test, pmichaud++
moritz_ pmichaud: did you try if my starry obfu works after the operator changes? 19:56
pugs_svn r28146 | lwall++ | [S02] remove fossil noticed by pmichaud++ 19:57
pmichaud moritz_: I meant to do so, but didn't get a chance to do it.
I hope it still works :|
moritz_ pmichaud: I just tried, still works
pmichaud \o/
moritz_ it's a nice stress test
pmichaud and "whew!"
bbiab 19:58
colomon Wow, that's a lot of failing tests. 20:00
20:00 pmurias left
colomon Is Inf an Int? 20:02
rakudo: say Inf.WHAT 20:03
p6eval rakudo 4fc254: OUTPUT«Num()␤»
20:03 Hashwolf joined
colomon rakudo: say NaN.WHAT 20:04
p6eval rakudo 4fc254: OUTPUT«Num()␤» 20:05
20:09 SmokeMachine left
pugs_svn r28147 | moritz++ | [perl6.org] (permissive) robots.txt 20:11
20:12 cotto joined
moritz_ oh wow, on Saturday we had 2.4k visitors on perl6.org 20:13
20:13 sharada left 20:16 krunen_ left
pmichaud is there a google analytics tag on that site? 20:17
moritz_ no
and I won't add one (privacy)
pugs_svn r28148 | pmichaud++ | [t/spec]: Unfudge now-passing RT #68894. 20:18
pmichaud Ticket resolved.
moritz_ thanks
dalek kudo: e005879 | pmichaud++ | src/setting/Whatever.pm:
Fix copy-pasto errors in Whatever.pm .
pmichaud afk again for a while (run to store) 20:20
20:23 Zloyrusskiy joined, Su-Shee left 20:25 takadonet left 20:28 dakkar left, SmokeMachine joined
colomon About 1/3 of the way through the swap(div, /) spectest errors... 20:31
Errr... anyone out there know why »div« might work when >>div<< fails with a "ResizablePMCArray: Can't pop from an empty array!" error? 20:35
frettled just finished glancing at the Wikipedia article for quaternions. Shame on you, colomon and ruoso, now you got me thinking about how I could write a sensible blog post about it. ;) 20:39
moritz_ colomon: might be a mis-parse 20:41
colomon Actually, it's weirder than that, I misread the error. 20:42
moritz_ I know this error from unterminated ternary operators
rakudo: 1 ?? 3
p6eval rakudo 4fc254: OUTPUT«ResizablePMCArray: Can't pop from an empty array!␤in Main (src/gen_setting.pm:3454)␤»
colomon I switched »/« to »div«, and that seems to have broken >>xx<< twenty lines later.
20:42 iblechbot left
frettled Is this what we call involuntary refactoring? 20:43
colomon No, it's >>div<< after all?
frettled: yes.
moritz_ colomon: »div« and >>div<< should be identical 20:44
frettled IIRC, all unicodish variants have ASCII equivalents.
colomon I think >>div<< isn't parsing correctly.
moritz_ note that it's missing from build/gen_metaop_pir.pl 20:45
so the >>div<< hyper op is never generated 20:46
colomon Any hints on how to fix that?
lisppaste3 moritz_ pasted "generate >>div<< etc. for colomon++" at paste.lisp.org/display/86326 20:47
moritz_ (note that this change might require a reconfigure) 20:48
frettled hmm, it's really that easy to make a new one?
moritz_ dearly hopes so 20:49
colomon Okay, cross your fingers... 20:50
20:50 mikehh_ joined
frettled Then again, it's not _really_ easy, because it makes it slightly awkward for user-defined hyper operators. 20:50
moritz_ frettled: does it?
rakudo: sub infix:<foo>($a, $b) { "$a foo $b }; say <a b c> »foo« (1, 2, 3) 20:51
p6eval rakudo 4fc254: OUTPUT«Unable to parse block; couldn't find final '}' at line 2, near "\"$a foo $b"␤in Main (src/gen_setting.pm:3454)␤»
moritz_ rakudo: sub infix:<foo>($a, $b) { "$a foo $b" }; say <a b c> »foo« (1, 2, 3)
p6eval rakudo e00587: OUTPUT«a foo 1b foo 2c foo 3␤»
frettled hmm
colomon rakudo: sub infix:<foo>($a, $b) { "$a foo $b" }; say <a b c> >>foo<< (1, 2, 3) 20:52
p6eval rakudo e00587: OUTPUT«a foo 1b foo 2c foo 3␤»
20:52 alester left
frettled Okay, then that's not «slightly awkward», it's just awesome. 20:52
moritz_ ;-)
frettled hmm. 20:53
colomon moritz_++ # >>div<< now works.
20:54 kst left 20:55 kst joined
frettled rakudo: sub infix:<fnord>(%a, %b) { %a.keys Z %b.keys }; say <a b c> Z (1, 2, 3) >>fnord<< <x y z> Z (7, 8, 9) 20:58
p6eval rakudo e00587: OUTPUT«Parameter type check failed; expected Associative, but got Int for %a in call to infix:fnord␤in sub infix:fnord (/tmp/rixx0Nfi7A:2)␤called from Main (/tmp/rixx0Nfi7A:2)␤»
frettled Grok error in frettled line 956 234 522.
moritz_ rakudo: sub infix:<fnord>(%a, %b) { %a.keys Z %b.keys }; say %(<a b c> Z (1, 2, 3)) >>fnord<< %(<x y z> Z (7, 8, 9)) 20:59
p6eval rakudo e00587: OUTPUT«␤»
moritz_ rakudo: sub infix:<fnord>(%a, %b) { %a.keys Z %b.keys }; say (%(<a b c> Z (1, 2, 3)) >>fnord<< %(<x y z> Z (7, 8, 9)) 21:00
p6eval rakudo e00587: ( no output )
moritz_ rakudo: sub infix:<fnord>(%a, %b) { %a.keys Z %b.keys }; say (%(<a b c> Z (1, 2, 3)) fnord %(<x y z> Z (7, 8, 9))
p6eval rakudo e00587: OUTPUT«say requires an argument at line 2, near " (%(<a b c"␤in Main (src/gen_setting.pm:2550)␤»
21:01 ruoso joined
moritz_ frettled: your >>fnord<< expects a list of hashes on both sides - you don't supply them 21:01
frettled List of hashes, even. Hrm! Good point.
21:02 ceci left
moritz_ rakudo: sub infix:<fnord>(%a, %b) { %a.keys Z %b.keys }; say ({ a => 1}, { b => 2 }) >>fnord<< ({ c => 2}, { d => 4 }) 21:03
p6eval rakudo e00587: OUTPUT«a cb d␤»
moritz_ better? ;-)
frettled absolutely :)
21:03 mikehh left
ruoso back 21:04
frettled o/~
rakudo: sub infix:<fnord>(%a, %b) { %a.keys Z %b.keys }; %h=({ a => 1}, { b => 2 }) >>fnord<< ({ c => 2}, { d => 4 }); say %h.perl; 21:05
p6eval rakudo e00587: OUTPUT«Symbol '%h' not predeclared in <anonymous> (/tmp/msOQqWBy94:2)␤in Main (src/gen_setting.pm:3454)␤»
frettled whooops. 21:06
moritz_ my %h = ... will be just fine
frettled rakudo: sub infix:<fnord>(%a, %b) { %a.keys Z %b.keys }; my %h=({ a => 1}, { b => 2 }) >>fnord<< ({ c => 2}, { d => 4 }); say %h.perl;
p6eval rakudo e00587: OUTPUT«{"a c" => ["b", "d"]}␤»
frettled moritz_: yes, I keep forgetting that, stupidly.
pugs_svn r28149 | moritz++ | [irclog] enable default_escape in search template, jrtayloriv++
frettled I'm too used to writing Perl 5 non-strict one-liners. 21:07
21:10 Zloyrusskiy left 21:13 mikehh_ is now known as mikehh
ruoso considering patching the spec for the three-layered number system... but maybe I should wait for TimToady's approval 21:14
colomon Only one .t file left to patch... 21:15
21:15 bioSlayer joined
bioSlayer well, hello everybody 21:16
moritz_ hi
bioSlayer Hey, actually, well, I got to here through the link you put in at PerlMonks.org
masak bioSlayer: o/ 21:17
ruoso colomon, oh... are you doing it already?
colomon I'm working on the / switch with div. 21:18
bioSlayer I have this question, If Perl 6 is the next big thing happening in Perl, how do you perceive a transition experience from Perl 5 to Perl 6
I hope this can be in context with the ongoing conversation and pard me for the interruption
moritz_ ruoso: colomon is patching the test suite and rakudo to conform to the current spec
colomon Looks like I may have to dig into the Temporal code...
moritz_ bioSlayer: there are several options; one is embedding Perl 5 code in Perl 6 programs
bioSlayer: jnthn is working on a prototype that enables that, see github.com/jnthn/blizkost/tree/master 21:19
masak bioSlayer: you're not at all interrupting anything, by the way. we love that kind of question.
ruoso bioSlayer, additionally, there's a plan on modifying the p5 interpreter so it will be able to run Perl 6
masak colomon: oh hai. I have an ongoing fork where I work with the Temporal code. wanna collaborate? 21:20
colomon masak: It's just going through and changing / to div repeatedly...
bioSlayer and how about the syntax structure and the commands, do we really like have to forget about Perl 5 to learn this one, thing is I have just started learning Perl these two months
masak colomon: ah. never mind, then. :)
moritz_ bioSlayer: some things are different, but all in all it feels rather similar to Perl 5 21:21
masak bioSlayer: Perl 5 knowledge will keep being useful for many years to come.
bioSlayer: with that said, I know and use both Perl 5 and Perl 6 daily, and I greatly prefer the syntax and semantics of Perl 6.
bioSlayer ummm, so if I learnt the major things out there in Perl 5, I can still learn or rather "Transit" to Perl 6 on the fly? 21:23
frettled Yes, sortof. 21:24
From my point of view, it helps to have some experience with other programming languages as well.
21:24 [particle] left, [particle] joined
bioSlayer this is good news then , I have nothing to worry about :) thanks everybody.... 21:25
masak bioSlayer: have the appropriate amount of fun. :)
bioSlayer for sure I would, and you too :), I will help you all with Perl 9 probably, wait for me everyone :P 21:26
frettled bioSlayer: what masak said. There is also an increasing amount of documentation about how to learn basic Perl 6 out there, plus quite a few blog entries if that floats your boat.
bioSlayer: If you start with Perl 6 now, I'm sure you'll be helping out with Perl 6, not perl 9! 21:27
masak I'd recommend checking out Planet Perl 6.
bioSlayer <frettlet>thanks for the uplift
masak, I would check planet Perl 6 21:28
masak excellent.
moritz_ planetsix.perl.org/
bioSlayer @morits thanks :)
lambdabot Unknown command, try @list
masak we're here if you have any questions or suggestions.
lambdabot: hush, humans talking.
21:28 Whiteknight joined
frettled bioSlayer: oh, as you can see, there's a stupid bot here thinking that sentences starting with an @ are meant for it :D 21:29
bioSlayer hahaha
I am sure you guys are having fun, I woul catch up and see what I can participate with... 21:30
ruoso hugme, hug bioSlayer
hugme hugs bioSlayer
frettled bioSlayer++
colomon Dang it, I'm down to the very last failing test, and I have no idea what is up with it.
masak bioSlayer++
moritz_ colomon: what's the offending test? 21:31
bioSlayer :)
moritz_ @karma bioSlayer
lambdabot bioSlayer has a karma of 2
lisppaste3 colomon pasted "Last failing test" at paste.lisp.org/display/86327
colomon It appears to go into an infinite loop. 21:32
21:33 tak11 joined
moritz_ colomon: maybe turn your current rakudo + spectest changes into a big patch each, and put them in RT. It's much easier to look at that when knowing what you changed, and how 21:33
21:34 bioSlayer left
frettled rakudo: my %AB = {A => { a => 1, c => 3, b => 2 }, B => { y => 8, x => 7, z => 9 }}; %AB = %AB{A}.keys.sort Z %AB{B}.keys.sort; say %AB.perl; 21:36
p6eval rakudo e00587: OUTPUT«Could not find non-existent sub A␤»
frettled uh, fnord. 21:37
moritz_ %AB<A>
frettled rakudo: my %AB = {A => { a => 1, c => 3, b => 2 }, B => { y => 8, x => 7, z => 9 }}; %AB = %AB<A>.keys.sort Z %AB<B>.keys.sort; say %AB.perl;
moritz_ instead of %AB{A}
p6eval rakudo e00587: OUTPUT«{"a" => "x", "b" => "y", "c" => "z"}␤»
frettled moritz_: yep
masak frettled: and you don't really need those outermost curlies.
frettled rakudo: my %AB = A => { a => 1, c => 3, b => 2 }, B => { y => 8, x => 7, z => 9 }; %AB = %AB<A>.keys.sort Z %AB<B>.keys.sort; say %AB.perl;
p6eval rakudo e00587: OUTPUT«{"b" => "y", "c" => "z", "a" => "x"}␤»
frettled neatski.
masak++
masak bows 21:38
frettled Now for the better trick: let's say that I want to join the keys of %AB<A> with the values of %AB<B>, by sorted keys.
masak with happy salutations from the secret society of bracket removal.
frettled \o/
I have trouble seeing how I can do that easily. Any suggestions?
masak frettled: use a map on the right side of the 'Z'. 21:39
21:40 kst left
frettled masak: hms? 21:40
masak let me whip up an example.
21:40 kst joined
frettled thanks :D 21:41
masak: as a measure of my gratitude, please listen to this youtube clip of a great classic afterwards. :) www.youtube.com/watch?v=5FSHS3DLElE
(also entertaining for other Scandiwegian-comprehending listeners) 21:42
masak thanks, I will :)
oh, Fingal. :)
isn't that him sitting over there?
frettled isn't he dead?
masak no, he's moving about. 21:43
frettled :)
BTW, I'm using this hash juggling in a new blog entry, because I just realized that I might be able to save quite a bit of work in a refactoring job somewhere. :) 21:44
masak rakudo: my %AB = A => { a => 1, c => 3, b => 2 }, B => { y => 8, x => 7, z => 9 }; %AB = %AB<A>.keys.sort Z map { %AB<B>{$_} }, %AB<B>.keys.sort; say %AB.perl; 21:45
p6eval rakudo e00587: ( no output )
moritz_ rakudo: say 1
p6eval rakudo e00587: OUTPUT«1␤»
masak well, it workedforme.
frettled heh :)
moritz_ oh, at this time of the hour the server recompiles a rakudo
masak rakudo: my %AB = A => { a => 1, c => 3, b => 2 }, B => { y => 8, x => 7, z => 9 }; %AB = %AB<A>.keys.sort Z map { %AB<B>{$_} }, %AB<B>.keys.sort; say %AB.perl; 21:46
frettled ah!
p6eval rakudo e00587: OUTPUT«{"a" => 7, "b" => 8, "c" => 9}␤»
masak there you go.
frettled thanks!
moritz_ which means that the load is high, and chances for a timeout are even higher
masak :)
the intricacies of providing a self-refreshing evalbot. 21:47
frettled yep
moritz_ aye. *sigh*
why doesn't somebody else take care of it? I'm not that good with sysadmin stuff
github.com/moritz/hugme/ # source of hugme, for your inspection :-) 21:48
masak \o/
frettled moritz_: if someone else took care of it, then we couldn't complain to you :) 21:49
moritz_ frettled: aye; and I'd spent more time on the test suite and cool IRC bots like hugme 21:50
21:50 KyleHa left, FCO joined
frettled moritz_: that's a pretty good point. 21:53
I'm not volunteering right now because of personal issues, but perhaps someone else has a chance of helping out.
(that is, I'm not _reliable_ now) 21:54
moritz_ if somebody wants to step up, let it be known that it's no magic; I'd make the current server available to that person, and give an introduction, answer questions etc. 21:57
21:58 PacoLinux left 21:59 M_o_C left
masak use.perl.org/~kentnl/journal/39561 # more butterflies 22:01
I'm getting conflicting signals from that post.
the point seems to be "hey I tried to re-do the logo, but it still pretty much sucks". 22:02
22:02 jrtayloriv left
moritz_ aye 22:02
frettled rakudo: my %A = { a => 1, b => 2 }; my %B = { z => 9, y => 8 }; %AB = %A.keys.sort Z map { %B<$_> }, %B.keys.sort; say %AB.perl;
masak not the most convincing argument, I'd say. 22:03
p6eval rakudo e00587: OUTPUT«Symbol '%AB' not predeclared in <anonymous> (/tmp/kzhQmY8CST:2)␤in Main (src/gen_setting.pm:3454)␤»
frettled argh! sigh.
rakudo: my %A = { a => 1, b => 2 }; my %B = { z => 9, y => 8 }; my %AB = %A.keys.sort Z map { %B<$_> }, %B.keys.sort; say %AB.perl;
p6eval rakudo e00587: OUTPUT«{"b" => undef, "a" => undef}␤»
masak frettled: not <$_>
frettled: {$_}
frettled rakudo: my %A = { a => 1, b => 2 }; my %B = { z => 9, y => 8 }; my %AB = %A.keys.sort Z map { %B{$_} }, %B.keys.sort; say %AB.perl;
p6eval rakudo e00587: OUTPUT«{"a" => 8, "b" => 9}␤»
frettled masak: yes, I forgot that <> only works with literals
masak yeah. that's what it does. 22:04
frettled masak: huh, what's the point of that blog post of kentnl's, anyway? sigh. 22:05
moritz_ hates it that JSON doesn't allow trailing commas after lists
masak kent\n: oh hai. I like the style of the logo. it looks mature. 22:06
frettled Now I just wonder if I've gotten my Perl 5.10.1 code right, haha.
masak kent\n: it's kinda the antithesis of Camelia. 22:07
kent\n masak: I had hoped at very least it might give somebody a few alternative ideas :)
masak I think it might.
pmichaud good evening, #perl6 22:10
kent\n masak: the vibe I was trying to do was "programmers and designers dont do well being the same person, but its plain to see the existing idea is a bit nasty and dated, even for a programmer, heres proof, now go hire somebody who knows what they're doing plz" 22:11
22:11 SmokeMachine left
masak kent\n: I half-got that this was what you wanted to say. 22:11
kent\n the existing logo is fine... if your target audience is about 6. 22:12
years old.
masak yes.
your point being? :)
kent\n my point is made earlier :)
masak oh, ok.
PerlJam cultural wonkiness is our forte ;) 22:13
kent\n PerlJam: with that logic, we should have stuck with goatse for a logo =P
PerlJam kent\n: except that aesthetics matter and goatse is unappealing to almost everybody. 22:14
kent\n yes. quite.
cultural wonkiness is probably fine if you only want to attract existing perl programmers, but like churches, you can't cater explicitly to the inhouse, you have to be prepared to meet the outside world or you'll crust up and die 22:15
frettled kent\n: I think you may be placing a bit too much importance in a logo.
kent\n possibly.
but I'd rather not be giving out low-hanging lambast fruit 22:16
PerlJam kent\n: If it 's fine for 6 year olds, it's fine for you too :)
frettled kent\n: PHP's huge userbase doesn't seem to be overly bothered by their completely lackluster logo :)
PerlJam frettled: PHP has a logo? 22:17
frettled exactly :)
kent\n that was my next question.
frettled static.php.net/www.php.net/images/php.gif
and what about Python, or Ruby?
kent\n yeah, but now imagine if they'd slapped barney the purple dinosaur on their logo
it would be amusingly apt 22:18
22:18 ruoso left
frettled That might get them some attention. ;) 22:18
kent\n python: snakes
ruby: rubies
frettled Just one ruby, currently.
And Python's use of stylized snakes looks like something a 7-year old may have come up with while bored in class. Yet it doesn't seem important. 22:19
kent\n I'm more familiar with the snake on the front of the books than on their logos
frettled as we're all probably more familiar with the Camel on the Camel Book than … uh, I don't even _know_. 22:20
22:20 KatrinaTheLamia joined
moritz_ the onion? 22:20
PerlJam garlic may be more apt in some cases :)
frettled Isn't that the Perl Foundation's logo?
kent\n the onion seems to be only known heavily by in-house
although, tis the only offical one we can use thats not copyright to Oriely 22:21
frettled perl.org confusingly uses both.
sjohnson perl => pearls :)
PerlJam sjohnson: that's the wrong kind of confusion there.
kent\n even with pearls, big balls of white opaqueness are really hard to make a decent logo out of with only ~6 colours to play with 22:22
pmichaud Perl Foundation == Onion. Perl 5 == Camel (trademarked by O'Reilly)
22:23 kst left, kst joined
kent\n decides Perl's logo should be a fecking great block of cheese 22:23
why? no reason. 22:24
pmichaud I think masak++ has the cheese logo rights :)
Juerd Is Camelia trademarked?
PerlJam carves the cheese into something resembling a troll just for kentn :)
frettled Juerd: yes.
Juerd It has a small "TM". That's ugly :(
pmichaud Juerd: trademarked, yes, not registered yet (afaik)
Juerd I hope it will be like tux, in the sense that you can use it for whatever purpose without having to ask anyone. 22:25
pmichaud Juerd: svn.pugscode.org/pugs/misc/camelia.txt
masak I hereby cede all cheese logo rights. :)
kent\n alternatively, a chilli-bean ( if it is not already used, but I bet it is ) would be a good compliment to an onion
pmichaud The Camelia image is a trademark of Larry Wall, and permission is
granted for non-exclusive use to label anything related to Perl 6,
provided the image is labeled as a trademark when used as a main logo
on a page.
Juerd pmichaud: That's good enough, as everything in existence is related to Perl 6 somehow :D 22:26
PerlJam kent\n: you should read that page pmichaud just mentioned too
kent\n we could just revamp the onion logo
I've on occasion glossed over it
Juerd kent\n: Have you seen Camelia? She's *beautiful*
kent\n I'm sure somebody could retain the beauty, and approach a better degree of reality 22:27
PerlJam "reality"?
Juerd Reality!
Hah! :)
Ever since junctions I haven't cared... ;)
kent\n If you could get the proper metalic sheen of butterfly wings, that'd be ossum
PerlJam kent\n: have you seen padre's splash screen? 22:28
kent\n thinks he has
pmichaud I think TimToady++ wants Camelia to be about as real as Tux is.
Juerd "It is not necessary to so label icons"... Yay!
I have the onion logo *illegally* on juerd.nl, because it doesn't have that exclusion to the requirement. 22:29
(Haven't asked permission either, though.)
kent\n would like to see an onion, complete with glossy brown skin tones and green shoot emerging
PerlJam kent\n: I'm sure you could find one like that at your local market ;) 22:30
kent\n but not a photo, sort of cgi-ish
Juerd en.wikipedia.org/wiki/Camelia # hmmm
Not bad either in terms of prettiness. 22:31
frettled A disambiguation page is clearly needed. 22:32
PerlJam camelia is also some type of flower IIRC
frettled «Camellia»
PerlJam ah, two els 22:33
However ... camelia.sourceforge.net/
We just have to win the meme war. Given the controversy already, we're sure to do so ;) 22:34
frettled hee-hee
Okay, let's fix the Wikipedia article first, and then move the current Camelia article to a new place, and … :D
kent\n www.hallagulla.com/urdu/photo/data/..._Onion.jpg # is nice, but whiter and slightly more toony would be awesome 22:35
frettled …or maybe I should just go get my beauty sleep, I have to pick up my car in 6.5-ish hours. Bleh. :(
Juerd kent\n: There's no use in replacing the onion logo. It's already a well known and recognised logo around the world.
(in the community)
frettled g'night!
moritz_ 'night 22:36
PerlJam good night frettled
Juerd Good night
masak 'night, frettled
pmichaud A bit of historical perspective on Tux (and that might be relevant to Camelia): gist.github.com/178756
kent\n yeah. I just figgured, perl6 is inspired by some of the old with a lot of the new, so I can't see why an revamps for it could be an option. but meh.
TimToady cultural wonkiness is one of the invariants 22:37
that makes Perl 6 the same as Perl 5
and the fact is, we *are* aiming to impress 6-year-olds 22:38
and these days, toons are big business :) 22:39
PerlJam Pokemon Camelia should go over well. 22:40
TimToady and so far, I haven't seen anyone jolted by Camelia that couldn't use a little jolting
PerlJam Anime Camelia for the older crowd
kent\n man, I can see a rule 34 coming a mile off :( 22:41
TimToady to the pure, all things are pure :) 22:42
kent\n I can smell circular logic the same way 22:43
TimToady it's all circular at some level or other
masak or axiomatic. 22:44
PerlJam masak++ 22:45
TimToady axioms circle themselves
masak :)
well, if you're using _that_ definition of circular, then sure...
TimToady almost all cultural fights involve someone assuming something is axiomatic that someone else wants to take a circularity saw to 22:46
I personally want to take a circularity saw to corporate logo culture
kent\n asserts the world really does revolve around himself, and nobody can prove concretely the contrary 22:47
TimToady 'coure not, that's what time is for 22:48
*course
22:48 FCO left
moritz_ now if you find a good, non-circular explanation of what time is... the Nobel Price might be yours! 22:49
TimToady well, maybe next time around
kent\n easy.
time isn't.
pmichaud I'd come up with a good non-circulation explanation of time but it's likely to take me too long to do it. 22:51
TimToady one could probably figure out a sequence of logic from that premise that indicates the world doesn't revolve around you
pmichaud *non-circular
kent\n there is no time, I'm just imagining it. 22:52
TimToady since, obviously, revolution cannot occur without time :)
kent\n not true.
I could just be imagining things are revolving
they might not exist at all.
masak that's not a very interesting hypothesis, that's all.
kent\n say what you will, you've just been unthunk 22:53
this line of thought is boring
NEXT
22:54 sevvie joined
masak I've discovered what I think is a very strange bug having to do with Str methods. I think I'm too tired to track it down, however. 22:54
pmichaud it's very likely
TimToady what color is it?
pmichaud we're still getting many of the fundamentals in place
masak my blog post is blocking on it, so it'll have to wait 'til tomorrow.
I can already report that it's a lot of fun to be writing literate Perl 6. 22:55
TimToady I think I've figured out the refactor of 'is context'
masak if you want, I can gist you a preview of the post.
pmichaud oh, good timing :)
masak: if you gist me a preview, I might be able to work on the bug a bit tonight
masak pmichaud: if you want. :) I can say it's quite involved.
pmichaud eagerly awaits 'is context' improvements, since that's his next major area of focus 22:56
TimToady in a nutshell, any var declared as $*foo is contextual
pmichaud so, my $*foo = 5; # contextual
TimToady so my $*answer = 42
pmichaud I like it
TimToady yes
pmichaud I like it _very much_
TimToady then there's the readonly question
pmichaud yes, I was about to ask that :)
TimToady I think there's a new declarator
for which I've looked at what seems like several hundred names 22:57
but for now, call it 'def'
pmichaud so, what was the several hundred plus one'th name?
'def'
hmmm.
TimToady which is a declaration of a readonly of any sort
def $x = 42; $x = 43; # boom
pmichaud reminds me a bit too much of python
TimToady too bad :)
pmichaud you decided against 'constant'? 22:58
TimToady constant evaluates = at compile tim
*time
pmichaud ah.
TimToady this evaluates at run time
masak pmichaud: gist.github.com/178765
TimToady so combining, you get def $*answer = 42
pmichaud hmmm
masak pmichaud: line 223 causes the Str method to return nothing.
TimToady and then any subroutine gets $*answer as readonly 42
Juerd Doesn't "def" want to be the opposite of "undef"? 22:59
TimToady but other threads might be running with different answers
Juerd Can you undef a def?
Can you def something to be undef?
pmichaud readonly-ness still feels like a trait to me, but that's probably because I still think of "is rw" and "is copy" and "is ref" having similar effects on variables
truly, I'd almost expect read-only to be the default behavior of := 23:00
my $x := 42; $x = 43; # boom
TimToady it represents the same semantics that parameters get by default
my $x := 42; $x := 43; # non-boom
def $x := 42; $x := 43; # boom
23:00 kst left
masak TimToady: then maybe call it 'bind'? 23:00
TimToady that was one of many words I considered 23:01
23:01 kst joined, rhr_ left
masak is not surprised 23:01
TimToady it might or might not even be the best one
pmichaud I agree with Juerd that def <-> undef is a bit disconcerting
I don't have any better suggestions at the moment
actually 'readonly' comes to mind :)
TimToady too ong
*long
pmichaud I suspect it's uncommon enough that length isn't an issue, though. 23:02
Juerd ensure, force, keep, uphold
pmichaud I agree it's lengthy
sometimes clarity trumps brevity, I think.
TimToady yes, the thesaurus has many words in those categories
Juerd uphold $foo = 42;
TimToady I would really like both
pmichaud agreed, me too.
masak 'night, y'all. 23:03
23:03 masak left
TimToady since this is likely to replace 'my' in single assignment contexts 23:03
Juerd TimToady: That's why I'm providing a filtered view :)
pmichaud I truly doubt I'll come up with any words that aren't in the hundreds you've considered, but I'll see if any spring to mind
Juerd TimToady: The whole thesaurus would flood the channel
TimToady I rather liked 'posit', but even that is a bit long
pmichaud 'fix' or 'fixed' 23:04
TimToady <-> fixed point
Juerd pmichaud: Brilliant :)
pmichaud 'fix' <-> fixed point seems less jarring to me than def <-> undef
TimToady so $x = 42; # :-) 23:05
Juerd pmichaud: I'd *love* to say "fix $this for $me;" Is assignment mandatory? :)
pmichaud Juerd: I suspect not :)
TimToady there are also lots of possible variants that change the = instead of the declarator
pmichaud ooooh
that would be useful to
and there's always 'but' 23:06
my $x = 42 but readonly
(too long, yes, I know)
Juerd Can a 'but' avoid future ':='?
TimToady one could even force 'my $x == 42;' to work
pmichaud I like changing the '=' because one could have something that is rw for a while and *then* becomes readonly afterwards
TimToady then == is out :) 23:07
so what operator means 'posit'?
my $x <- 42
Juerd What does the word posit mean? 23:08
TimToady my $x ≣ 42;
it means 'suppose for now' 23:09
pmichaud isn't that "let" ?
TimToady yes, but we've used that
unless we kill both temp and let
pmichaud we still have "let"?
ah.
Juerd suggests my $x <-> 42 and finds a place to hide
23:09 justatheory left
pmichaud we already have <-> as a lambda :) 23:09
it does mean "rw", however :)
TimToady that's why Juerd is hiding
pmichaud and -> means "ro"
Juerd I'm hiding for multiple reasons
TimToady so <- should ro the $x :) 23:10
Juerd I remember that <- used to be a problem, but that's fixed by the whitespace rules now.
pmichaud <:= 23:11
TimToady and, of course, there's all sorts of arrows in unicode
Juerd Still, though, it reads as "write only" because of the way arrows are used with for.
TimToady of course, CS folks who use arrow to mean assignment will be confused :) 23:12
Juerd Won't CS folks be confused, anyway, by Perl 6?
pmichaud right along with the math folks :) 23:13
Juerd I know I am :D
pmichaud: They have been confused by every programming language so far :)
TimToady if we made declarations at the end of feed operators readonly, then my $x <== 42 would work
pmichaud Yes, but I still somewhat like the notion of being able to mark a container as "fixed" at some point other than its declaration. 23:14
Juerd Visually, "is readonly" is lighter than "<==". It's also easier to type...
pmichaud I don't know why I like that notion, I just do.
colomon moritz_, pmichaud, masak: RT #68898 is the complete Int / Int swap with div patch, both rakudo and spectest bits. It's mostly there, and hopefully the e-mail and commit note successfully explain its flaws.
Juerd pmichaud: That would be wonderful for when you're paranoid about AaaD
pmichaud Juerd: it also reminds me of the way we do things in other contexts (more)
TimToady $x <- 42; assign and set readonly 23:15
pmichaud For example, often a process will start out with superuser privileges, and then it drops those privileges after doing some bit of setup. This is the same sort of thing for individual containers
Juerd <- has the same problem that def has. It looks like the opposite of something that's already there but has very little to do with it.
23:15 nihiliad left
TimToady it means bind to this with readonly semantics in both cases 23:16
23:16 rhr joined
TimToady and the ends of <-> just cancel :) 23:16
pmichaud I don't know that "set readonly" has to be an operator; I'd be fine with it being de-huffmanized and there being a short declarator form.
Juerd pmichaud: I imagine such a fix operater used on configuration. my %foo = eval $config-file's-contents; fix %foo;
pmichaud Juerd: exactly.
but in that sense it almost sounds like 'final' 23:17
(which might be an argument against it)
Juerd Well, at least "fix" has the antonym "break" ;) 23:18
TimToady why wouldn't you just say fix %foo = eval ...
Juerd TimToady: Because you want to do something in between too.
TimToady or my %foo <-- eval...
that's what do blocks are for :)
pmichaud :)
Juerd TimToady: my %foo = eval ...; %foo.values >>~~ s/.../.../; fix %foo;
TimToady FP is your friend (sometimes)
pmichaud and yes, one could always do the building in another variable, then assign to the "fixed" one.
Juerd TimToady: I can't really answer any "why would you want that" question, though, because I already feel that way about making variables immutable anyway. 23:19
TimToady I have a soft spot in my heart for imperative programming, but I'd rather encourage FP where not inconvenient
23:20 donaldh left
pmichaud The only counter-case I can come up with is that I might want to have a sub or trait that "fixes" one of its parameter arguments. 23:20
but it's a weak case.
TimToady it really is pretty much identical to the semantics we want parameters to default to
23:20 donaldh joined
TimToady I suppose we could steal := for this, at some expense 23:20
Juerd I suspect that that expense would not be that large in practice. 23:21
TimToady it would be more of a unification with sig binding 23:22
Juerd my $foo := ...; $foo := ...; feels weird anyway.
TimToady rebinding would still be possble with an appropriate declarator 23:23
this is all very Hmm-y
Juerd TimToady: :dwis
pmichaud that doesn't feel so weird to me -- it just means I'm re-binding an existing symbol
certainly I'd want &foo := ... to work
Juerd Convincing example.
TimToady basically, we want a light container bind, and a heavier container/contents bind and make readonly 23:24
Juerd forgot what ::= meant. Which synopsis is that?
pmichaud ::= is a compile-time bind
TimToady could also be stolen 23:25
pmichaud yes.
KatrinaTheLamia Juerd: ::= is compiler time bind
Juerd No synopsis where I can read the details? :)
pmichaud Synopsis 2, I think.
maybe 3.
Juerd (Thanks for the swift answers though)
KatrinaTheLamia I dunno, I don't mind doing := over and over again... would be good, actually.. make it so there are less syntax errors of "=" in boolean statements.
Juerd used to know all the synopsis numbers. Now there are 42 times as many, and I forgot even the ones I knew.
wayland76 S03 operators 23:26
pmichaud wayland76 ftw
wayland76 And S11 modules
..according to fgrep -c '::=' * | grep -v ':0'
pmichaud anyway, just going through this discussion I naturally tend to use "fix" as the verb (much more so than I'd want to use "def") 23:27
Juerd Under what circumstances would one prefer compile time binding?
pmichaud anyway, I'm being called to dinner
Juerd Bon apetit
TimToady when one gets tired of writing BEGIN { ... := ... } :)
Juerd s/p/pp/
pmichaud TimToady: Big +2 on the my $*foo declarations -- it's perfect. 23:28
TimToady my STD is already written that way, and works :)
thinking hard about stealing ::= to mean bind and lock
pmichaud I bet I can add trivially add it to Rakudo as well.
I'd be find with ::= meaning "bind and lock"
*fine
I haven't had a lot of use for :=
er, ::=
(I've had a *lot* of use for :=) 23:29
(not to mention that it makes excellent smileys when used at the end of a parenthetical thought :=)
okay, gone for now
TimToady chow
23:31 jrtayloriv joined
Juerd In Rakudo's source, I've acked for ::=, and it appears that it isn't actually used, whereas := is. 23:32
But maybe this just means that ::= isn't implemented yet. Haven't checked the list. Don't even know if there is one.
23:33 nbrown joined 23:38 Hashwolf left
pugs_svn r28150 | lwall++ | [S03] steal ::= for readonly binding ala sigs 23:42
23:45 justatheory joined
TimToady I think we still have the context trait, but $*foo variables automatically get it set, along with $_, $/, and $! 23:47
Juerd What is "virtually identical" in practice? 23:48
wayland76 "quite different"? :) 23:50
23:57 sevvie left 23:58 pyrimidine joined, pyrimidine left