svn switch --relocate svn.openfoundry.org/pugs svn.pugscode.org/pugs/ | run.pugscode.org | spec.pugscode.org | paste: sial.org/pbot/perl6 | pugs.blogs.com
Set by avar on 16 November 2006.
svnbot6 r15172 | lwall++ | First snapshot of proposed standard Perl 6 grammar in standard Perl 6. 00:07
TimToady (it is by no means complete) 00:08
mostly a translation of pmichaud++'s current grammar into something more like its final form. 00:09
still missing a lot of terms, and a lot of {*} commentary 00:11
just checked it in so you could see the direction I'm headed with it... 00:12
svnbot6 r15173 | lwall++ | add label token 00:46
svnbot6 r15174 | lwall++ | more grammar cleanup 01:59
TimToady n 02:08
y, even 02:09
meppl i always read "TimToday" 03:45
SamB dyslexic a bit? 04:10
allbery_b no, just the usual tendency to map something to what's familiar 04:16
I went through that too, I use "today" rather more often than "toady"
so the latter always looks like a typo
tene It took me a while to figure out what meppl had typed differently. 04:23
I kept reading it as "Toady" 04:24
dduncan I imagine that some people read it as TIMTOWTDI 05:01
allbery_b I gather that's how it's *supposed* to be read 05:31
svnbot6 r15175 | lwall++ | much grammar tweaking 07:20
jrockway heh, good to see a lwall++ commit :) 07:25
TimToady yeah, he does so little work around here and gets so much credit for it... 07:26
good to see him pulling his weight a little more. :) 07:27
jisom eh, someone has to tell all the programs how to behave
jrockway so is YAPC::Asia going to have a TimToady presence this year? 07:29
TimToady nope, have to be on the east coast that week. 07:30
alas
wouldn't mind going to YAPC::Asia every year...
jrockway for me it's a good excuse to get $WORK to pay for (part of) a trip to japan :) 07:31
shay|p6 guys, is there a place where all the made test are publicly available? 08:49
(I'm doing some tests, and I won't like to re-invent the wheel ;)
tene There's pugs svn... 08:50
shay|p6 oh I see, that "t" folder?
tene Yep. :) 08:51
shay|p6 great, working on it, thanks
tene svn.pugscode.org/pugs/t/README
is a file you want to read.
shay|p6 Indeed, I'm on it now
dmq yay, rsc said hed let us use his code if we want. 10:22
putter Do we have an operator precedence parser written in p5 yet? 19:05
Doesn't look like kp6 has one. Drat. (v6/v6-KindaPerl6/lib/KindaPerl6/Grammar.pm) 19:09
There is a misc/pX/fglock/Precedence.pm, but it looks like an unfinished draft. 19:11
tewk started a transliteration of PGE (written in parrot) to P5. misc/pX/tewk/Rules-P5/ 19:16
Just about a year ago. :/
audreyt putter: perl5/./Pugs-Compiler-Rule/lib/Pugs/Grammar/Precedence.pm 19:17
based on Parse::Yapp. pretty usable.
perl5/Pugs-Compiler-Perl6//lib/Pugs/Grammar/Expression.pm uses it.
putter thanks! looking... 19:19
putter sigh. choices, choices. 19:26
rodi well, after a long absence, I was making some modifications to ext/Automata/Cellular/*, and have discovered I never renewed my commit bit after the repository switchover... anyone around who could help with that? 19:45
Limbic_Region rodi - I think so 19:49
rodi - I have only done it once on the new system
rodi - give me a few
rodi Limbic_Region: thanks, I appreciate it. 19:51
Limbic_Region rodi - you are listed as "pending"
do you want me to send a new one or tell you the address the last one went to? 19:52
rodi Limbic_Region: I'm sure it went to my yahoo address, but I'm afraid I probably deleted it :-\ mind sending another? 19:53
Limbic_Region oh, and btw rodi - it is a @yahoo.com address which ate mine 19:54
I ended up having audreyt resend to my gmail account
sure
will do
if you don't get this one we can try another
rodi coolio, I'll check shortly.
Limbic_Region sent 19:55
used same info as the existing one
putter Limbic_Region: perhaps document the procedure somewhere? unless we three are unusual, I imagine it's going to come up again... 20:00
gaal evil: clemens.endorphin.org/liskell 20:01
lambdabot gaal: You have 1 new message. '/msg lambdabot @messages' to read it.
Limbic_Region putter - what?
lambdabot Title: Liskell - clemens.endorphin.org
Limbic_Region I have no idea what you are talking about
putter person had old repository account. lost email. wonders what to do now. 20:02
Limbic_Region putter - that's not the case here
rodi Limbic_Region: email came through, password updated. Doesn't seem to have updated in the repository yet, but that my require some sync time. Thanks.
Limbic_Region putter - person had old repository account and new repository account but lost email
putter , puzzled, punts 20:03
Limbic_Region in any case, I just did an "add" with the exact same details as were "pending" 20:04
PartclCoke So i'm looking into implementing tcl's [trace] on parrot. requires hooks for pre-HLL subs, post, and then pre and post on each *line* of the sub. I'm looking for the perl6 equivalents. S06 makes a blink-and-you'll-miss-it reference to PRE/POST/ENTER/LEAVE/KEEP/UNDO. Are these explained in any more detail anywhere? 20:08
PartclCoke (and, is there any interest tcl's apparently slightly more fine grained settings) 20:13
... finally, should I be asking this on the mailing list. =-) 20:15
(I see the test for this section in pugs is TODO...) 20:16
lumi Closure traits? 20:20
S04 maybe?
PartclCoke ah. a cross reference from s06 would have been most helpful there, danke. =-) 20:21
PartclCoke so it looks like tcl's [trace add execution (enter|leave) ...] are like perl6's ENTER and LEAVE; I have no desire to implement any of the other stuff that perl6 needs, and it looks like tcl has a few more related features that perl6 doesn't. Hurm. 20:24
though it looks like many of those traits can be implemented *in terms of* ENTER & LEAVE, so I don't feel too guilty.
any sample code for those? 20:25
lumi Other than the tests? 20:25
PartclCoke Didn't see any tests in pugs. 20:26
saw, "TODO"
lumi t/closure_traits/* ?
PartclCoke (in svn.pugscode.org/pugs/t)
lambdabot Title: Revision 15175: /t
PartclCoke ah. hadn't looked for enough different kinds of traits. thanks again. 20:27
lumi np
PartclCoke ... I am continually reminded of how ugly I think this is. :|
lumi How is it ugly? 20:28
lumi Looking at the smoke, it looks like everything in closure_traits passes 20:29
PartclCoke perl6? Most of the samples I see leave me thinking "powerful, but ugly". If I had anything more articulate, I'd be on the language list with it. =-)
lumi There's always Lisp 20:30
PartclCoke eh. perl5 is fine. =-)
Hurm. no way to add ENTER/LEAVE at run time?
lumi You find Perl 6 uglier than 5?
PartclCoke typically, yes. 20:31
perhaps I'll have a change of heart if I ever actually use it. who knows.
lumi Just wait til Christmas 20:32
putter Could someone remind me of the relative performance of p5 sub's vs methods? Steady state, not first call. 20:37
dmq i think its like 30% slower 20:53
putter dmq: my thanks 20:55
dmq i dont know that you should quote it, i just recall that from somewhere.
putter understood
if it were a small integer factor, I'd be inclined to preemptively optimize. but for 30% on a subsystem, nah. 20:56
putter is transliterating the redsix precedence parser into p5. happy to discuss choice if anyone cares. 20:57
putter Anyone know what the current mechanism of choice is in p5 for throw-catch aka exceptions aka longjump? Error? 22:10
audreyt Exception::Class 22:12
(if you want structured exception objects)
moritz hi, I just wantet to hint that the "Binary builds" link on pugscode.org is a 403 ;) 22:13
and the "illustrations" link as well 22:14
putter there's only a single exception flavor, and I don't really need a payload. so simple is fine.
E::C is slower than Error?
moritz: thanks for the report! :) 22:15
moritz putter: np ;)
putter audreyt: btw, my thanks for the pointer 22:16
svnbot6 r15176 | audreyt++ | * Fix broken links on pugscode.org as pointed by moritz++.
audreyt putter: not heavily, I don't think.
moritz: fixed, thanks
putter danke
audreyt moritz: next time around you can commit to docs/feather/pugscode.org/ too :)
moritz audreyt: without password? cool ;) 22:17
putter (he asked for it...:) 22:17
moritz audreyt: I still get a 403, but it could be a transparent proxy that's causing the problem 22:18
putter moritz: if you /msg me an email address, a repository account invite will land in your inbox... 22:19
audreyt moritz: it'll refresh in but 5mins
moritz: a commit bit is in your inbox now :) 22:20
putter lol
audreyt customary, add yourself to AUTHORS as the first commit. Welcome aboard! search.cpan.org/dist/Perl6-Pugs/doc...c/Hack.pod has some more ptrs.
lambdabot Title: Pugs::Doc::Hack - How to hack on Pugs - search.cpan.org
moritz ok, I'll read that ;) 22:21
audreyt and if you're not actually mlenz at physik dot uni-wuerzburg dot de, yell now and I'll have the commit bit re-sent :)
moritz audreyt: I am ;)
putter wonders if project_planning/TASKS is mentioned in Hack.pod 22:22
audreyt putter: no; please do
putter: also, have you used/adminned a trac before? 22:23
now that I finally get some homeostasis back, setting up trac may make xref code/issue easier
masak audreyt: happy to hear about the homeostasis 22:24
putter re before, no
audreyt masak: somewhat ironically, $job is competing fun with pugs :) as jifty dodges many perl5 pains, and we import more and more of perl6 into production
masak audreyt: that's great 22:25
putter re "xref code/issue", ??
audreyt (as of this writing I'm doing this RDBMS-synchronized-with-SVK backend for jifty that turns any hosted app instantly into synchronizable offline app)
putter: code tree cross-referenced with issue tracking as well asdocs
moritz speaking of production... is there a perl6 mod_perl module for apache?
putter ooooooooo...
masak wow 22:26
audreyt moritz: there was an early port of pugs on mod_parrot, but currently plain CGI is the only option.
moritz: otoh, I guess you can
use perl5:FCGI;
though I havn't tried it myself.
moritz audreyt: ok, I was just curious ;) 22:27
masak & 22:28
audreyt hm, GHC sure has improved in the month I was away. 22:31
full "make profiled" build of Pugs: 7min (was 30min) 22:32
(that's from realclean.)
see if it translates into smoketime as well...
putter any chance of hs-prime moving the prelude off monads onto arrows? ;) 22:49
putter TimToady: Any thoughts on when the bottom half of the sandwich might have a draft? 23:02
I'm pondering transliterating the "generate all the operators" section of redsix, but it's known to be so very non-spec. 23:03
(and if you use a script to generate the p6, the script would in itself be useful (to skip having to parse the p6)) 23:04
anyone: you could of course also play. the spec is in dev.perl.org/perl6/doc/design/syn/S03.html . (the redsix kludge in svn.pugscode.org/pugs/misc/pX/Commo...napshot.rb , section ###*** Operator declarations) 23:09
lambdabot Title: Synopsis 3: Perl 6 Operators - perl6:
putter wanted - p5 code which enumerates the operators, each with all its associated info. 23:10
(when my operator precedence parser hatches, it will be hungry...) 23:12
moritz *grml* pugs (svn == 15176) doesn't build, error message is ``Setup: error reading ./.setup-config; run "setup configure" command?'' 23:31
svnbot6 r15177 | putter++ | project_planning/TASKS - new task: Mention TASKS file in Hack.pod.
moritz any idea how to fix this?
and yes, I installed the modules mentioned in INSTALL.v6-alpha ;) 23:32
putter curious
one sec
wolverian hey, ruby2 stole our lambda!
those bastards :) 23:33
moritz shoot 'em ;)
wolverian I suppose it's a good sign ..
tene wolverian: explain more?
allbery_b lambdabot hangs out in ruby channel? yeh. seems the Rails folks are vbecoming addicted to haskell 23:34
putter TimToady: I wasn't thinking. Would need terms too. A rather more involved task. So... whenever. It's not pressing just yet.
allbery_b maybe you can corrupt them to perl6 :)
wolverian tene, ruby2 uses -> (a, b) { ... } to denote a lambda, apparently.
or so I gathered from redsix :)
putter yes
lol 23:35
allbery_b oh, that
wolverian what does the comment that older rubies didn't have a real lambda mean? are they broken like in python?
putter python lambda's are limited to a single expression (no statements) 23:36
pasteling "evalbot_r15176" at 194.145.200.126 pasted "Pugs build failure" (375 lines, 20.9K) at sial.org/pbot/22811
wolverian and they're read only
wolverian stares at evalbot_r15175
putter ruby 1.* lambdas... a = 2; lambda (a) {...} the a arg is bound to the same cell as the a = 2. 23:37
wolverian oh. it doesn't grab its own context? that's no lambda.
putter lambda(b) would work as expected. as long as b isnt mentioned in surrounding context. :/ 23:38
wolverian well, ruby's scoping is broken in any case.
putter rb scoping has many nice characteristics. 23:40
moritz: ghc --version ? 23:43
moritz putter: 6.6
(from debian etch)
putter (so much for that hypothesis...) 23:44
moritz putter: moritz.faui2k3.org/tmp/pugs-build-log <-- full build log
(`script` messed up a few characters, though) 23:45
tene Anyone have anything interesting to say about the current state of Perl6 Rules in pugs? 23:45
putter moritz: so its not finding network-2.0 for network-any 23:48
moritz putter: ok, so what do I do about it? 23:49
putter related to these? packages.debian.org/unstable/source...ll-network packages.debian.org/changelogs/pool.../changelog ? 23:50
lambdabot Title: Debian -- haskell-network
putter do you have haskell-network installed? 23:51
moritz no, it's not in testing
so I'll just have to get the unstable version...
putter curious that we haven't hit it before. 23:52
tene: I'm too out of date to have anything interesting to say. You might look at a recent smoke to get an idea of current status. 23:53
moritz it's not even listed at packages.debian.org/unstable/devel/
lambdabot Title: Debian -- Software Packages in "unstable", devel section
putter I didn't have to install it separately. Could it be (supposed to be?) part of the normal ghc package? 23:54
moritz putter: then it wouldn't be listed separately als pool/main/h/haskell-network 23:55
moritz putter: (at least I think so ) 23:55
I'm currently playing with my pinning to get a haskell package from unstable 23:56
putter tene: if you explore, and it turns out pugs is using a p5-side implementation of p6 regexs, I would be interested in hearing that. 23:58

Powered by Cro and the Raku® Programming Language.

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

Thank you!