pugscode.org/ | nopaste: sial.org/pbot/perl6 | ?eval [~] <m oo se> | We do Haskell, too | > reverse . show $ foldl1 (*) [1..4] | irclog: irc.pugscode.org/
Set by diakopter on 11 July 2007.
avar pugs.blogs.com/pugs/2007/10/kindaperl6-now-.html <= what's topaz? 01:03
lambdabot Title: Pugs: KindaPerl6 now runs on sbcl, clisp and ecl
moritz_ good localtime() 06:47
agentzh hmm, my svk pugs repos is broken again :/ 07:24
"Retrieving log information from 18539 to 18539
Modification of non-creatable array value attempted, subscript -1 at /usr/local/share/perl/5.8.8/SVN/Mirror/Ra.pm line 784."
does anyone have some idea on what is going on here? 07:25
moritz_ svk bug is my guess :) 07:29
agentzh moritz_: heh 07:32
haven't touched my pugs copy 2 or 3 months...
today when i finally have the chance to get back to it...it breaks... 07:33
sigh
moritz_ agentzh: use git and git-svn ;)
pmurias or plain old subversion ;) 07:34
agentzh moritz_: git also put something like .svn in the directories, which is not grep-friendly, no?
*pugs
*puts
moritz_ agentzh: just a .git in the top level dir
agentzh hmm, sounds good :) 07:35
moritz_ agentzh: and it's compressed, so you won't accidently find files by grepping with rgrep $pattern .
agentzh great
moritz_
.oO( or just use rgrep $pattern * )
agentzh goes to try git-svn.
apparently svk is not "pugs-friendly" 07:36
moritz_ anyway, I'm off to university &
integral avar: topaz is a failed rewrite of perl5 in C++, by chip.
It did funky stuff like changing the class of a C++ object at run time. Crazy.
pmurias integral: is changing the vtable at runtime allowed in standard C++? 07:37
Aankhen`` JUST USE MERCURIAL YOU HIPPIES
integral pmurias: I'm not sure the standard actually requires you to use a vtable
Aankhen`` Mmm, everything sounds better when you make it uppercase and add the phrase "YOU HIPPIES" at the end.
pmurias integral: you have to use an equivalent 07:38
cj Aankhen``: is that the fanciest these days? What about git? 07:39
Aankhen`` cj: Everyone seems to love git, I prefer Mercurial. Mercurial shares *many* features with git, as it originated through the same chain of events; it just has a lot more polish than git.
integral pmurias: but not an equivalent that allows mutation :-) 07:40
pmurias integral: name a (semi-)sensible nomutable alternative 07:42
integral instead of using pointers to pairs of (vtable,data); use pairs of (vtable,pointer to data) 07:43
so a function receiving one of these can't change its caller's idea of the object's type
integral & 07:45
agentzh oh...svnbot is not around? 08:59
moritz_ agentzh: seems not ;( 09:00
r18540 | agentz | 2007-10-22 08:56:47 +0000 (Mon, 22 Oct 2007) | 2 lines
agentzh okay...anyway, git-svn works perfectly for me :)
moritz_ agentzh++ test commit in git-svn
agentzh :)
thanks :) 09:01
moritz_ I tried to mimic svnbot, but I'm too lazy to do it properly ;)
agentzh: did you clone the complete repo with history?
agentzh no 09:02
just fetched the HEAD 09:03
moritz_ ok 09:03
I plan to make a deep clone with an up-to-date git-svn some time
agentzh wow
moritz_ and sync it regularly
agentzh that'll be great ;)
agentzh maybe i can do it on my server as well ;) 09:04
moritz_ but first I have to learn a bit more about git and git-svn and the like
agentzh but that server is not ubuntu...
git is a lovely creature :) 09:05
pmurias svk's false conflict suck 09:52
agentzh pmurias: have i reverted your previous commit? 09:53
pmurias: i'm not sure...being a new git-svn user...
agentzh " [kp6] ShortCircuit passes thunks to ||,// and &&" 09:54
agentzh goes to look at the trac
agentzh r18542 r18541 09:55
pmurias agentzh: no, svk reports possible conflict bla,bla,bla and aborts the commit for me 09:56
agentzh ah...okay
pmurias: maybe switch to git-svn as well ;) 09:57
pmurias fglock: hi 09:59
fglock pmurias: hi 10:00
pmurias agentzh: it only reports the possible conflict warning when there are none, and handles real ones correctly
agentzh k
fglock how about change Visitor::Emit* to Visitor::Emit::* 10:01
fglock also add Visitor::Optimize::FoldConstant and such 10:04
pmurias fglock: makes sense
pmurias fglock: i'm trying to move the shortcircuiting hack into a visitor 10:18
fglock ok 10:19
brb & 10:22
agentzh "PCR - Pugs::Runtime::Tracer now works; also added t/17-tracer.t which is passing now :)" 10:27
pmurias fglock: do you have any idea why the test i checked in dosn't work
? 10:29
agentzh: how do i use the tracer? 10:30
agentzh pmurias: not for kp6 ;) 10:31
pmurias i know this
agentzh pmurias: the plan is to write a Firefox addon for the tracer UI
pmurias: Pugs::Runtime::Tracer only prints out all the necessary info as plain text 10:32
pmurias: in a format designed by myself...
pmurias: such outputs will finally be read by that FF addon that i previously mentioned and generate graphical outputs
pmurias: that's the plan 10:33
pmurias: a bigger plan is to add this to kp6 as well ;)
pmurias: but i'd like to make PCR work first :) 10:34
pmurias: because most of the work on the PCR side is almost done :)
fglock pmurias: looking
meppl good morning 10:44
pmurias good morning meppl 10:46
meppl good morning pmurias 10:48
pmurias fglock: sorry, the test i checked in was completly screwed up 10:56
fglock is svnbot down? 12:14
masak fglock: yes 12:17
don't know why
moritz_ svnbot runs on top of pugs, right? 12:20
pmurias pugs is pretty stable those days 12:21
;)
fglock lunch &
fglock b 12:51
allbery_b bcpl
pmurias fglock: do you have a plan for onion? 13:25
fglock pmurias: working on it right now, 13:26
pmurias fglock: good ;) 13:27
fglock I and ruoso are making it a $work project 13:29
fglock we already have a company product that uses Perl 6, 13:32
so it makes sense for the company to put resources in Perl 6 development
[particle] considers working there 13:34
fglock assigns desk space for [particle] 13:41
pmurias & 13:44
fglock signatures now include types 14:45
moritz_ kp6: 'test' 15:10
exp_evalbot r18556: OUTPUT[Can't open --secure: No such file or directory at script/kp6 line 274.␤]
moritz_ what is the currently recommended way to run kp6 with the --secure option? 15:11
fglock moritz_: looking 15:16
moritz_ r18555 | fglock++ | [kp6] initial impl of types in signatures 15:19
r18556 | fglock++ | [kp6] improved bare block parsing
r18557 | fglock++ | [kp6] proper hash x pair x bare-block parsing
sorry, duplicates :)
fglock svnbot overshoot 15:19
moritz_ I'm currently hacking a pure perl5 svnbot 15:20
integral secure may just need the line in the GetOptions block uncommented 15:21
pugs_svn r18557 | fglock++ | [kp6] proper hash x pair x bare-block parsing 15:39
r18558 | fglock++ | [kp6] Grammar - allow an extra comma
[particle] moritz_: why switch to perl5 svnbot? 15:41
moritz_ [particle]: because it works 15:44
avar removed --secure support in kp6 overhaul, oops 15:45
moritz_ [particle]: in general I'm in favor of making everything in perl6, but since pugs is unmaintained atm and kp6 isn't there yet...
avar he probably means why not use one of the 10e6 current rss->irc bots 15:46
or are you just going to use plagger?
moritz_ [particle]: I think many evalbot-problems where pugs problems in fact
avar: what's plagger?
avar search.cpan.org/dist/Plagger
I've used it with traq to run a svn ircbot
you just set up a yml file and tell it to start running 15:47
moritz_ avar: might be a good idea, but now it's a bit too late ;)
avar well, the universe always needs more ircbots
moritz_ avar: unless you just bump in your solution here, in which case I'll withdraw my bot 15:48
diversity++
avar I just told you what you can use, I can't be bothered to set it up:)
moritz_ avar: just give me the code, then I'll take a look at it ;) 15:49
[particle] rolls his eyes 15:49
avar cpan -i Plagger
buubot avar: Sorry, I couldn't find a module that matched: -i
[particle] kicks buubot 15:50
moritz_ ah, I didn't see it has an IRC plugin
avar plagger.org/trac/browser/trunk/plag...s/irc.yaml 15:51
lambdabot Title: /trunk/plagger/examples/irc.yaml - Plagger - Trac
avar just change the feed name and bot name
moritz_ damned, that's too easy ;) 15:52
[particle] bonus: it's purl perl 5 ;) 15:54
pugs_svn r18559 | fglock++ | [kp6] bare blocks compile 16:16
avar [particle]: that's a bonus? 16:26
:)
moritz_ no, it's a requirement ;) 16:35
dlocaus [particle]: ping 16:41
[particle] pong
dlocaus Friday, was last my day of work. 16:42
[particle] congratulations? :)
dlocaus last Friday. So, I have some time now to start on kp6. My fedora SATA raid died this morning, but I still have my trust Windows XP
[particle] fglock: ping
dlocaus (yes: Congrautlations, I wanted to have a party)
so, what do you guys need help on? 16:43
fglock pong
moritz_ dlocaus: what would you like to do?
moritz_ dlocaus: for example masak++ thought about starting a parrot backen 16:43
[particle] we have a willing and able contributor for kp6
fglock dlocaus: hi 16:44
dlocaus moritz_: I want perl6 out the door, I'm tried of seeing it next "bash" and other shell scripting languages on job offers.
I'm a decent perl5 programmer (at least I think so), I don't have a *nix box at the momment... died this morning.
but, I'm willing to do what it takes to get perl6 out the door.
moritz_ ok 16:44
dlocaus well, almost willing to do anything.
[particle] here... drink this. 16:45
dlocaus no blue pill?
fglock dlocaus: can yo get a perl5 and svn installed? maybe Strawberry perl and TurtoiseSVN
you
dlocaus I have tortise SVN already installed, and I have ActiveState perl installed.
checking on straberry perl. 16:46
fglock also, ask for a commit bit :)
dlocaus I have a commit bit on kp6.
moritz_ dlocaus: read pugs.blogs.com/pugs/2007/06/kindaperl6_proj.html and the roadmap it links to
lambdabot Title: Pugs: "KindaPerl6" project roadmap
fglock ActiveState might just work
dlocaus [particle] got me one.
moritz_ [particle]++ 16:47
[particle] activestate works fine for me
it's the makefile unportability that doesn't work (although that shouldn't prevent development)
dlocaus moritz.faui2k3.org/pugs_pod/kp6-roadmap isn't responding for me. 16:48
moritz_ damned
pugs_svn r18560 | fglock++ | [kp6] 'for' now compiles to a method call, so that it can work on lazy lists
Khisanth grab the pod then :)
[particle] dlocaus: fglock started the kp6 effort, and is shepherding the contributors
dlocaus aye, I've been reading up the documentation 16:49
fglock dlocaus: can you hack the Makefile.PL for windows compat?
dlocaus I probably could with a hammer, if you want I'll give it a shot. 16:50
cj are you folks done yet?
avar yay .for :)
dlocaus waves a hammer at Makefile.PL
moritz_ I'v just disabled the irc logs for the moment 16:51
faxathisia hi avar
moritz_ they were slowing down the server too much
now I'm checking the logs
avar can Class For be removed then?
faxathisia: hi
fglock avar: I think so
dlocaus fglock: I'll start on the Makefile.PL
fglock dlocaus++
[particle] cj: not yet. check back tomorrow :)
fglock avar: now need to implement next/last/return as exceptions 16:53
moritz_ one IP blocked and irc logs reenabled 16:54
fglock dlocaus: you'll need an utf8-capable editor too
try out adding your name to AUTHORS
[particle] gvim++ 16:55
dlocaus fglock: I'm downloading ghc ATM.
moritz_ dlocaus: you don't need that for kp6 hacking 16:56
dlocaus moritz_: the Makefile.PL required it.
[particle] urk
maybe you're looking at the wrong makefile.pl?
moritz_ dlocaus: you don't need the top level makefile for kp6
dlocaus '--version' is not recognized as an internal or external command,
operable program or batch file.
moritz_ dlocaus: go to v6/v6-KindaPerl6 and run the Makefile.PL there 16:57
dlocaus moritz_: ok.
I need to get some modules... 16:58
[particle] hands dlocaus the keys to cpan 16:59
dlocaus hurm... sorry, I'm used to unix. How do I run make on winxp?
moritz_ 'nmake' perhaps? 17:00
dlocaus ah, ok.
moritz_ if you have Visual Studio blah installed
dlocaus fatal error U1000: syntax error: ')' missing in macro invocation 17:01
Stop.
I'll guess, I start waving my hammer now. :)
pugs_svn r18561 | fglock++ | [kp6] rm 'For' ast 17:02
fglock dlocaus: it should run without make
TimToady sneaks out the back door 17:04
pugs_svn r18562 | dlo++ | Added myself (CPAN: DLO) to the AUTHORS list. 17:05
moritz_ dlocaus++ # first commit
dlocaus fglock: successfully added myself.
fglock: I don't understand. " perl Makefile.pl; nmake " correct? 17:06
[particle] dlocaus: yep, there's some make- or shell- specific code in that makefile. likely needs to be converted to use ExtUtils:: code in perl
eg $(patsubst ...) 17:07
TimToady I always used dmake on Windows; I suppose that's strawberry Perl though 17:09
[particle] i have nmake, make, and dmake. usually use nmake, cause that's what most winfolk are used to
i guess we forgot to lock the back door, cause TimToady's back 17:10
TimToady came back around the front. It's not that far...
but I'm pretending to be a customer now
moritz_ TimToady: so you want to contribute to perl6 development now? ;-) 17:11
TimToady DARN IT, WHY IS THIS TAKING SO LONG???
[particle] ha!
moritz_ that's what "customers" here want
pugs_svn r18563 | fglock++ | [kp6] more s/For//g 17:13
TimToady well, my skill set is very limited, I really am not up to much other than a few documentation cleanups here and there...
[particle] i think there's a synopsis or two that needs some fleshing out.... 17:14
TimToady those synopses are scary 17:16
cj who wrote those anyway? 17:23
masak TimToady: I can see the slashdot headlines: "Larry Wall intimidated by Perl 6 synopses: 'those synopses are scary'"
[particle] wait till he looks at the implementations....
masak [particle]: they're already a bit scary. pugs scares me sometimes. 17:24
...in a good way
[particle] seems like pugs is waiting for the competition to catch up
...and waiting... 17:25
masak :(
[particle] everyone else has a new object model now... what about pugs? 17:26
masak I think it has been on its way in since june 17:27
it's residing somewhere in the computer and/or mind of audreyt++
meanwhile, strange bugs are being found, like t/builtins/arrays/init-in-exported-sub.t 17:28
I followed that one pretty far into the Pugs jungle, but in the end the scary noises of Haskell beasts got the better of me 17:29
integral There is some stuff in the pugs svn that needs hooked up for the MO stuff. But, either someone needs to just go ahead and Do it; or they need to figure out how to find out the status from audreyt 17:30
masak integral: I think both sound like a good idea 17:31
who knows which one will pay off?
maybe one, maybe both
last I heard, audreyt was finding a sustainable lifestyle 17:32
what that means for her possible reappearance here, I don't knoe
s/knoe/know/ 17:33
is there someplace one can read up about the MO stuff?
I don't think I can do anything about it, but it would be nice to be able to comprehend what needs changing and why 17:34
cj TimToady: makes me think of something you wrote once 17:36
www.testearly.com/2007/08/17/fire-y...lazy-ones/
lambdabot Title: Test Early 􏿽xBB Fire your best people&#8230;reward the lazy ones, tinyurl.com/3b6uwc 17:37
cj botsnack
pugs_svn r18564 | fglock++ | [kp6-perl5] .for with multiple parameters 17:45
TimToady masak: r15828 was audreyt's checkin of MO refactoring, at which point she estimated two audreydays remaining to get it to work. (she reverted it as the next patch and then just installed the pad refactoring) 17:52
the person to ask for MO documentation is probably nothingmuch
[particle] ...who's trekking in peru atm.
masak oh, so r15828 contains a lot of future goodness? like a branch? 17:53
worth checking out.
literally
TimToady what two precrashaudreydays corresponds to in mere mortal time is anyone's guess...
[particle] remember to minimize reality to the system tray
masak [particle]: I'm on a Mac, I use Cmd-H 17:54
TimToady masak: gaal also probably knows a lot about it 18:00
TimToady nothingmuch has some docs at or near nothingmuch.woobling.org/MO/lib/MO/Overview.pod 18:02
masak TimToady: great, thx! 18:05
TimToady also stevan_ probably knows more about the theoretical end of MOP than anyone
[particle] agreed. his mind is a theoretical kitchen floor. 18:06
TimToady but he's not into Haskell as much, I suspect
masak why not?
stevan_ cause he likes Ocaml bettter :P
TimToady [particle]: MO is a dessert wax AND a floor topping! 18:07
stevan_ nothingmuch is currently hiking the moutains of South America,.. so he is uhm,... inaccessible :)
masak: were you thinking of re-starting the MO Haskell effort?
masak: you can also look at nothingmuch's MO talk slides here - www.iinteractive.com/moose/object_m...start.html 18:09
lambdabot Title: meta meta meta meta, tinyurl.com/26cv6l
stevan_ it might help some
TimToady stevan_: so just rewrite pugs in Ocaml. ^_^
masak stevan_: thx 18:10
stevan_: no, nothing so grand as re-starting an effort
right now I'm just curious
stevan_ TimToady: k,.. I will start now, and should be done by thanksgiving
masak: the focus of MO was to break the MOP into 2 seperate peices
TimToady er which...nevermind...
stevan_ TimToady: the thanksgiving shortly before the christmas release of perl 6 :) 18:11
masak: MO builds meta objects in a purely functional way,.. so that they can be easily manipulated and transformed without breaking things too much
TimToady but then you'd be able to write pugs in prerelease Perl 6 instead of Ocaml...
stevan_ true 18:12
I was thinking the Ocaml dialect of Perl 6,..just grammer hacking yah know :)
TimToady *nodnod* :)
stevan_ masak: MO does all this at "compile" time...
masak: then when its ready, it "emits" a class 18:13
[particle] i think that kind of statement deserves texas quotes.
stevan_ there is then a whole framework for the "runtime" of MO
which handles the bits that couldnt be optimized away in compile time
[particle]: what are texas quotes??
[particle] <<texas quotes>> 18:14
stevan_ ah
masak: MO is much eaiser to understand if you view at two seperate frameworks :) 18:15
that said there are only 2 people who actually understand MO,.. audreyt and nothingmuch
[particle] you don't understand it, you just put it on cpan? 18:16
stevan_ no no,.. this is nuffin's baby
his $pet_project for work
awwaiid You don't have to understand The Matrix to employ it 18:17
stevan_ [particle]: he was annoyed by how I did the class compilation in Moose... and wanted to fix it
he couldnt (blame perl5) and so decided to go out on the pure research limb and write MO instead
[particle] ah, that's right. mo is moose without the ose
stevan_ yup 18:18
it will probably serve as the basis for Moose v2
TimToady can we have that by halloween?
[particle] TimToady: wrong channel. try #ponie :P 18:19
masak stevan_: cool! I already understand it a little better 18:20
[particle] i wonder when we'll have evalbots for ponie and topaz here....
masak will assault nothingmuch++ with questions when I've read his stuff
stevan_ TimToady: if you can get chromatic to pay me to write it,.. yes,.. you can have it by halloween :)
masak: nothingmuch wont be back in civilization until mid-November 18:21
masak: so feel free to ask me until he gets back,.. and I will try and help
masak stevan_: goodie, I'll be back with questions eventually 18:24
pugs_svn r18565 | ruoso++ | [yap6] new scalar dispatcher. we now have a scalar container available in the runtime... 18:35
r18566 | ruoso++ | [yap6] Makefile.cvs now has two helper targets "build-world" and "clean-world" 18:42
pugs_svn r18567 | ruoso++ | [yap6] Clarifications in the roadmap inside the README file 18:46
pugs_svn r18568 | ruoso++ | [yap6] Clarifications in the roadmap inside the README file 18:54
r18569 | ruoso++ | [yap6] fixing ChangeLog and NEWS to point to README 19:03
masak 'night 19:18
dlocaus How does one do 19:43
my_var = `perl -e "blah blah blah"` in a make file?
TimToady depends on the make program. in the worst case, you filter the makefile yourself 19:46
moritz_ my_var=$(shell perl -e "foobar") iirc
at least in GNU make 19:47
dlocaus I have to use "nmake" :)
moritz_ but somehow I don't think that's protable
dlocaus windows xp
[particle] i think $(shell ...) works 19:47
moritz_ I haven't used windows for programming since 2002 or 2003 19:48
rindolf [particle]: in GNU make?
dlocaus nmake (microsoft's make)
[particle] reads the fine manual 19:51
dlocaus [particle]: I'm not sure nmake supports any method of my_var = $(shell perl -e "code") 19:55
[particle] no, i can't find it 19:56
dlocaus I think that Makefile.PL will have to do the expansion itself.
or code will have to be added to Makefile.PL to do the expansion.
[particle] no, you can generate a file, and read it in 19:57
dlocaus perl Makefile.PL; perl script/expansion_WinXP.pl; nmake?
[particle] inside makefile, $PERL some_script_to_modify_env.pl >some_file && import some_file 20:00
dlocaus ok 20:01
[particle] dlocaus: maybe it's just easier to do the expansion at Makefile.PL time and regenerate the Makefile more often 20:07
dlocaus [particle]: I agree.
probably more portable too. [reading microsofts infile documentation, which doesn't seem to be portable] 20:08
[particle] (make portability)--
dlocaus so, do 20:09
if ( $OS eq 'Win32' ) {
do expansion;
warn "note about needing to regenerate Makefile"
}
[particle] that seems good start 20:11
...to be a...
moritz_ [particle]: do you know a good URL that explains TGE? 20:21
not for me, I wan to link it on perlmonks
Aankhen`` moritz_: Tree Grammar Engine.
Now just link to the line in the IRC logs.
Aankhen`` ducks. 20:22
[particle] moritz_: sadly, there's no documentation outside the source at the moment 20:23
i could point you to the source dir, where TGE.pir et al live
moritz_ [particle]: if that contains documentation that's fine
[particle] svn.perl.org/parrot/trunk/compilers/tge/
lambdabot Title: Revision 22402: /trunk/compilers/tge
moritz_ [particle]: thanks 20:24
pugs_svn r18570 | fglock++ | [kp6] Grammar - fixed default value in signatures 21:08
r18570 | fglock++ | [kp6] Grammar - pointy subs allow a parameter list, without parenthesis
r18570 | fglock++ | [kp6-perl5] - default values in parameters work
fglock kp6: for [1,2,3,4,5] -> Int $a, Int $b => 42 {say " ($a,$b)"} 21:19
exp_evalbot r18570: OUTPUT[Can't open --secure: No such file or directory at script/kp6 line 274.␤]
pugs_svn r18571 | moritz++ | [kp6] try to fix --secure option in script/kp6 21:21
moritz_ kp6: for [1,2,3,4,5] -> Int $a, Int $b => 42 {say " ($a,$b)"} 21:22
exp_evalbot r18571: OUTPUT[syntax error at - line 2, near ")␤ ("␤Execution of - aborted due to compilation errors.␤]
moritz_ my $list = [1, 2, 3, 4, 5]; for @( $list ) -> Int $a, Int $b => { say "$a|$b"; }; 21:23
kp6: my $list = [1, 2, 3, 4, 5]; for @( $list ) -> Int $a, Int $b => { say "$a|$b"; };
exp_evalbot r18571: OUTPUT[syntax error at position 43, line 1 column 43:␤my $list = [1, 2, 3, 4, 5]; for @( $list ) -> Int $a, Int $b => { say "$a|$b"; }␤ ^ HERE␤]
moritz_ kp6: my $list = [1, 2, 3, 4, 5]; for $list -> Int $a, Int $b => { say "$a|$b"; };
exp_evalbot r18571: OUTPUT[syntax error at position 38, line 1 column 38:␤my $list = [1, 2, 3, 4, 5]; for $list -> Int $a, Int $b => { say "$a|$b"; }␤ ^ HERE␤]
moritz_ kp6: my $list = [1, 2, 3, 4, 5]; for $list -> Int $a, Int $b => 42 { say "$a|$b"; }; 21:24
fglock hmm - it works here 21:24
moritz_ kp6: say "test";
fglock: I recommend to checkout a clean copy of kp6 from time to time
then you'll soon notice when you forgot to commit a recompile, 'svn add' a file etc. 21:25
fglock i'll try a make forcerecompile
moritz_ the bot died with "Segmentation fault" 21:26
kp6: say "test";
exp_evalbot r18571: OUTPUT[Can't locate object method "test" via package "True" (perhaps you forgot to load "True"?) at - line 1.␤]
moritz_ kp6: "test".say;
exp_evalbot r18571: OUTPUT[DISPATCH: calling perl on invalid object:$VAR1 = 1;␤␤ at compiled/perl5-kp6-mp6/lib/KindaPerl6/Runtime/Perl5/MOP.pm line 33␤ main::DISPATCH(1, 'perl') called at - line 16␤]
moritz_ that's an evalbot breakage 21:27
dlocaus does kp6 pass all tests w/o errors?
moritz_ dlocaus: no
dlocaus ok
pasteling "moritz_" at 89.13.155.241 pasted "kp6 test failures" (16 lines, 1K) at sial.org/pbot/28174 21:28
moritz_ perl script/kp6 --secure -e 'say "test"' 21:29
works for me
pugs_svn r18572 | moritz++ | [evalbot] try to fix kp6: evaler 21:31
moritz_ kp6: say "test" 21:32
pugs_svn r18573 | fglock++ | [kp6] make forcerecompile
exp_evalbot r18572: OUTPUT[test␤]
moritz_ kp6: my $list = [1, 2, 3, 4, 5]; for $list -> Int $a, Int $b => 42 { say "$a|$b"; };
exp_evalbot r18572: OUTPUT[1|2␤3|4␤5|42␤]
moritz_ fglock++ # many grammar fixes
Tene kp6 have threading yet? 21:33
moritz_ I always thought the syntax for default params was an assignment 21:33
Tene: no
.oO(but lazy gather/take ;)
TimToady moritz_: yes, it's just =, not =>
fglock oops 21:34
fixed 21:35
I even read the specs sometimes
pugs_svn r18574 | fglock++ | [kp6] default param is = instead of => moritz++
fglock Tene: I'm told threading in Perl 6 over Perl 5 is not worth it, but it would work in other backends 21:37
fglock kp6: for [1,2,3,4,5] -> Int $a, Int $b = 42 {say " ($a,$b)"} 21:38
exp_evalbot r18574: OUTPUT[ (1,2)␤ (3,4)␤ (5,42)␤]
pugs_svn r18575 | dlo++ | Added an $^O check for MSWin32. If MSWin32 is present certain make gnuisms $(patsubst) $(wildcard) are removed and replaced with filename listing.
r18575 | dlo++ |
r18575 | dlo++ | See warning at bottom of Makefile.PL. This code does not re-generate the kp6_mp6_OBJ kp6_mp6_modules_OBJ list when you execute nmake. You have to re-execute perl Makefile.PL to rebuild the list. You need to do this when you add a new file to directories listed in "sub
..constants".
fglock sleep & 21:39
moritz_ dlocaus: does it compile now for you? 21:40
dlocaus Could someone please download kp6's Makefile.PL on a Windows System, with nmake and let me know if it works?
yes.
well, "perl Makefile.PL" appears to work fine, the tests have errors.
moritz_ dlocaus: see my paste from some time agoe
dlocaus: sial.org/pbot/28174 21:41
lambdabot Title: Paste #28174 from "moritz_" at 89.13.155.241
moritz_ that's what I see (and what I think is normal)
dlocaus moritz_: running tests. 21:42
dlocaus moritz_: t/kp6/grammer/17-complex-return took up 1gig of ram, and then crashed. 21:56
moritz_ not good :/ 21:57
dlocaus moritz_: I had more tests pass... I svn updated about 3 hours ago.
cj can someone tell me why register-based VMs are better for implementing dynamic languages than stack-based VMs? 23:08
SamB are they?
Patterner define 'better' 23:12
TimToady if you want a longer answer than you want rather than a shorter answer than you want, you should ask on #parrot instead 23:15
Tene #parrot on irc.perl.org 23:16
meppl good night 23:23
dlocaus does anyone know a good article explaining p6's grammar and tokens? 23:46

Powered by Cro and the Raku® Programming Language.

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

Thank you!