pugscode.org/ | nopaste: sial.org/pbot/perl6 | pugs: [~] <m oo se> (or rakudo:, kp6:, smop: etc.) || We do Haskell, too | > reverse . show $ foldl1 (*) [1..4] | irclog: irc.pugscode.org/
Set by TimToady on 25 January 2008.
meppl good night 00:22
lichtkind good night 00:23
meppl ;) 00:24
rakudo_svn r26194 | cspencer++ | Added a grep method that will parse and behave as expected when invoked in the form: 00:30
r26194 | cspencer++ |
r26194 | cspencer++ | [email@hidden.address] $_ % 2 })" or "(1, 2, 3, 4, 5).grep({ $_ % 2 })"
r26194 | cspencer++ |
r26194 | cspencer++ | Other forms (ie. "grep { $_ % 2 }, @list" and [email@hidden.address] $_ % 2 }" don't correctly parse the block and don't work as expected.
mncharity there's a lambdabot sitting on the wrong #parrot. odd. 00:58
rakudo_svn r26195 | cspencer++ | Added a first method that will work for the form: 01:00
r26195 | cspencer++ |
r26195 | cspencer++ | [email@hidden.address] $_ eq 3 })" or "(1, 2, 3, 4, 5).first({ $_ eq 3 })"
r26195 | cspencer++ |
r26195 | cspencer++ | Other forms won't parse (see previous grep commit message) until blocks are parsed correctly.
katie_d Are there any good sites that map out the inner workings of Parrot, analogous to Gisle Aas' "Perlguts Illustrated" ? I'd like to start reading the Parrot sources, but my knowledge of C isn't that great, and it would be helpful to have a "map" to help me find my way around... 06:03
PerlJam katie_d: if you download parrot there's lots of info in the docs directory. 06:05
katie_d Perljam: OK, sorry about that. I'm currently building subversion, so I haven't had a chance to browse through it yet. I'll look into that when it's done. Thanks :) 06:06
PerlJam katie_d: there's also some dead-tree info in the book "Parrot and Perl 6 Essentials" 06:07
but dead-trees are quickly outdated.
oops, it's called "PErl 6 and Parrot Essentials"
(just looked over on my book shelf)
katie_d Yes, there is an online edition now I believe. At least part of it. I'm not sure if it's the whole book though. 06:08
PerlJam katie_d: If you'd like to learn PIR and such, I've got a few articles in The Perl Review, chromatic has a few articles in The Linux Journal (I think) 06:10
oh, it's just "Linux Magazine" 06:11
katie_d OK, thanks. I'll look into those. Here is the link to the free online copy of P6/Parrot essentials btw. It is a work in progress, but they are indeed putting it up for free: dev.pugscode.org/browser/docs/tutorial 06:12
lambdabot Title: /docs/tutorial - Pugs - Trac
PerlJam well, that's more perl6y than parroty 06:14
katie_d Yes. It looks like the later parts of the book, regarding parrot's architecture, haven't been added in yet (if they are ever going to be) 06:17
PerlJam www.parrotcode.org/docs/ is what you're looking for I think 06:19
lambdabot Title: Parrot Documentation - parrotcode
PerlJam But there is some disconnect between the docs in the parrot source tree and the docs on the website so if you want the most up to date info, look at the parrot source tree 06:20
katie_d Thanks for the link. I just finished building SVN a few minutes ago and have been looking through /docs -- very interesting stuff. Hopefully in a few months, I'll be able to start helping out :) 06:22
PerlJam why a few months? Jump in now :) 06:23
katie_d PerlJam: I don't know C very well, and I've never worked on any big projects. Up to this point, I've only written 1000-3000 line scripts in Perl. I've working through K&R and "Practical C", but don't have any real world experience. So it will be a while before I'm ready to do anything as difficult as Parrot. 06:26
This is going to be my first attempt at helping out with a big open source project, so I'll probably need a few months of learning :) 06:27
PerlJam Experience is the best teacher. Find something to do and just do it so that even your learning can contribute meaningfully :) 06:28
there are lots of niches in Parrot for people of various skill levels.
spinclad katie_d: i expect you've seen by now that the parrot parts of the book are at svn.perl.org/parrot/trunk, in docs/book. 07:42
lambdabot Title: Revision 26199: /trunk
spinclad there's actually plenty to work on in parrot at PIR level or higher. 07:44
katie_d spinclad: Sorry for not responding, I was afk. I am still looking through docs/ and taking notes. But yes, I've glanced at a few of the pages in docs/book ... definitely have quite a bit of reading to do :) 07:52
sordith_ cleans the last pieces of the religion chat from the table and puts coffee and tea there for the developers 10:42
nothingmuch @tell mncharity none of the traditional AG engines work for Perl (6 or 5) because they require static typing for parent/child rels. I implemented something a bit more nasty for MO, but it works too. Either way I don't know the offerrings for HS at the moment 13:13
lambdabot Consider it noted.
lumi perl -ne '/([^=]*)=(.*$)/; $f{$1}=$2; END { print "$f{artist} - $f{title}" }' ~/.quodlibet/current 13:54
Erk. ignore. 13:55
gbacon so is anyone using the cabal build? if not, how are people building pugs? 15:09
pugs_svnbot r20057 | gbacon++ | Improvements to cabal build. Build still fails on src/Pugs/AST/Internals/Instances.hs. 15:17
diff: dev.pugscode.org/changeset/20057
lambdabot Title: Changeset 20057 - Pugs - Trac
ruoso was chatting at #io and was told that SMOP could be a natural runtime for Io... I wonder if the same could be said about erlang... 15:28
mncharity gbacon: re 'so is anyone using the cabal build?', trying to. :) re 'if not, how are people building pugs?', I suspect everyone, including the pugs evalbot, is running on old revisions. Or has lost the capability for the moment. 15:44
lambdabot mncharity: You have 1 new message. '/msg lambdabot @messages' to read it.
gbacon mncharity: I've been trying to revive it
mncharity :) :)
gbacon it's making it further through the build but still choking 15:45
see my commit message
mncharity did. was about to try. um, should that be 6.6.1 or 6.8.2? 15:46
gbacon I'm running ghc-6.8.2
and Cabal-1.3.3
mncharity nothingmuch: re 'static typing for parent/child rels', is the problem that they don't have union types?
re, ghc, ah, ok. let's see...
gbacon it ought to work with older versions of Cabal, but you know how that goes 15:48
and probably older versions of ghc too 15:49
I wonder whether 6.8.2 got stricter about wobbly types
mncharity sigh. fedora's ghc682-6.8.2-8.fc8 has Cabal-1.2.3. hmm... 15:54
www.haskell.org/ghc/docs/latest/html/libraries/ is listing things as 1.2.3. no sign of 1.3.3. typo? (he asks hopefully:) 15:56
gbacon mncharity: did you try the build? 15:57
I doubt that 1.3.3 is strictly required 15:58
1.3.3 is a devel version
mncharity in progress. my disk is being thrashed by something else, so it's taking a bit to set up a clean build.
Var/Type length mismatch in MO.Run... 16:02
and Pugs.Class
pasteling "mncharity" at 76.24.29.201 pasted "pugs build error message" (11 lines, 665B) at sial.org/pbot/30407 16:06
mncharity it's the wobbly type error in Internals 16:07
gbacon yeah, same problem I saw 16:08
mncharity at least it's replicable :)
gbacon whee!
but I didn't see an error with MO.Run 16:09
mncharity pasting...
gbacon waits 16:12
pasteling "mncharity" at 76.24.29.201 pasted "larger trace" (164 lines, 9.9K) at sial.org/pbot/30408
mncharity fyi i'm x86_64. 16:13
gbacon yeah, saw the same warning
gbacon that one appears to be long-standing: irclog.perlgeek.de/perl6/2007-03-04#i_-36356 16:16
lambdabot Title: IRC log for #perl6, 2007-03-04
mncharity looking...
ah, ok. note to future selves: we should archive copies of the build logs for releases. and perhaps monthlies. to make it easier to spot (non-)regressions. 16:18
I'm afraid I can't help much with the actual type wrangling. My record wrestling with the hs type system is... not characterized by success. 16:20
FurnaceBoy nice locution tho
TimToady mncharity: better than my record... :) 16:21
which is characterized by abject failure...
FurnaceBoy you're more of the duck type, afaik
mncharity /me googles for the wobbly type error message... 16:22
TimToady I'm just a quack...
FurnaceBoy LOL
FurnaceBoy quacks up
mncharity but takes a moment to groan first
FurnaceBoy yeah needs a GOL 16:22
gbacon mncharity: read spj's paper
TimToady Gang of Legion? 16:23
gbacon research.microsoft.com/~simonpj/pap...t-icfp.pdf
but we're making progress! :-)
[particle] groans out loud 16:24
mncharity it's never a hopeful sign when google turns up exactly zero occurrences of an error message. 16:25
FurnaceBoy nods
mncharity reading...
FurnaceBoy mncharity: perhaps it is in your uncommitted code. :)
TimToady you'll get more hits if you just search for "error"
FurnaceBoy groans
mncharity weeee. ok. so the question is why has this become wobbly. left over wobbliness seems to be "the type system shrugged here". so did something else break, or become weaker, creating the ambiguity? did this? or did the type inferencer get weaker? 16:36
svn.pugscode.org/pugs/src/Pugs/AST/...stances.hs 16:37
[particle] #ghc may have answers we don't 16:38
gbacon no, it's a wobbly type because the typechecker is conservative
there are ambiguous cases that can arrive, so it needs type annotations 16:39
mncharity (aside, I saw an experimental language once which was 2D, inside of emacs - you would write out denotational semantics, complete with lines of "---------", just as if it were a paper. I wonder if we could do something similar in p6 for typing rules:)
oh, and it would run of course. (the language) 16:40
I don't suppose the needed type annotations are... pause... obvious? 16:41
silly question. ok.
gbacon well, it already has an annotation
I wonder why the typechecker doesn't like it 16:42
cf. src/DrIFT/YAML.hs:52
mncharity and fromTypeable is from the bottom of src/Pugs/Internals/Cast.hs 16:55
mncharity /me gives up for now. was wondering if the forall might be contributing, since everything seems at least somewhat nailed down. but I'm just flinging spaghetti in the dark. 17:09
mncharity it's very quiet 20:20
gbacon mncharity: mucho progress 20:25
trying to link the pugs image now
mncharity !! 20:27
gbacon more like #%^%$&^
mncharity lol :) (sorry) 20:28
mncharity it's not polite to laugh at others' encounters with static typing. 20:29
/me svn's up, nothing new there, awaits developments with interest. :) 20:30
gbacon I can check in what I have so far if you like 20:33
TimToady καγω...moi aussi...ich auch...わしも... 20:34
mncharity and ponders parsing. () waiting on rakudo yaml ast; () STD_red is hit a STD.pm problem - could start working on STD.pm improvements (mission creep?), or worry about it later, pushing other things for now (but trying to avoid the "optimistic blocker dodging" game); () ... 20:35
mncharity re "I can check in what I have so far if you like", no need for me, just the only way I could see to potentially help was, if it had been available, to toss in another compile. but given past uniformity of error, it's not clear it'd be much help to you. 20:36
re "καγω ... わしも", I really really viscerally love having unicode. don't know why, just do. 20:38
all sorts of neat squiglies with meaning. without having to have stroke.
re parser... () redsix, sigh, could continue massaging dumped tree which is still missing much info. would really miss the concreteness of using STD.pm. It was nice to be able to say, this node's Match object should look like X - anything else is simply wrong. () kp6, ... should try pmurias's work. but don't think of kp6 as a sufficiently big dialect for me to want to do prelude work in it. but perhaps start, and get others inte 20:42
spinclad 'and get others inte'... cut off? 20:44
mncharity () plug yet_another_regex_engine into misc/winter_jig's sixrx. () use sixrx with or without (ie, Perl::Compiler::Rules) to write grammar for STD.pm. making it easier to try exploratory "pour it into bison", "into btyacc", "into PCR", etc. 20:45
re cut off,
"and get others interested? ()... outliers..." 20:46
PerlJam www.perlfoundation.org/perl5/index.cgi?gsoc2008
lambdabot Title: GSoC2008 / Perl 5 Wiki
lichtkind PerlJam: hej ho long time since i last see you
PerlJam greetings 20:47
lichtkind PerlJam: are you on board again or just visiting?
PerlJam I intended to start contributing more in Jan. but events conspired against me. 20:48
lichtkind evil conspiracy theories 20:49
PerlJam parrot/rakudo are still in the back of my mind waiting for some tuits though
and I occasionally look at STD.pm to see what I still don't understand :)
lichtkind where is rakudo lieing in the svn dint found it yesterday
haha
mncharity () write something to massage the existing kp6 ast into something which might come out of STD, and use the kp6 parser as a proxy to start working on STD-ish ast support and emitters. () start working on rx_on_re, but instead of going from yare to p6, which we don't have a way to parse, and thus to compile/run/test, instead do a ruby version first, in a p6-ish style to permit easy, perhaps automated, later translation...
lichtkind PerlJam: i make pretty good progress but only on german side of my tutorial 20:50
PerlJam mncharity: Have you signed up to be a GSoC mentor yet? 20:51
mncharity: or to participate even
mncharity PerlJam: no, and non-applicable (I suspect). :) 20:53
mugwump Is Perl signing up this year?
PerlJam mugwump: see the link above.
mugwump: btw, congrats on your perl gititude :) 20:54
mncharity hmm, I wonder if I could do a GSoC to get some educational software ideas off the back burner... but that wouldn't (yet) be p6 land. maybe OLPC... hmm... 20:55
mugwump Thanks PerlJam, it's been a lot of quite tiresome work.. 20:56
mncharity PerlJam: interesting idea, thanks!:)
"perl gititude"?
mugwump mncharity: utsl.gen.nz/gitweb/?p=perl;a=tag;h=change-33412 20:57
lambdabot Title: Use The Source, Leia! - perl/tag
PerlJam mncharity: mugwump went insane for a while
:)
mncharity ooo, shiny 20:58
PerlJam mugwump: did the last bit go quicker than the first bit? It seems like just the other day you were happy to be at 5.005 or whatever. 20:59
mugwump Well I had to write an importer for the perforce era of changes
Once I got all the conversion issues through to about Change 1647 (perl 5.005) sorted, there were very few bugs after that 21:00
PerlJam oh, I forgot about that.
mncharity Moose, "16 months ago"? what exactly is the projects/ directory?
mugwump oh, sorry, there is a lot of old stale cruft there 21:01
that was a little proof of concept conversion
mncharity np, just curious.
mugwump I took 814kB of CPAN tarballs, a 4MB SVN repository for Moose and crammed them into a 450kB packfile
mncharity re "perforce era of changes", ah, so it was a "hairy data conversion" exercise, not just a "painful broken tools system assembly" exercise. 21:02
eep
um, how? 21:03
oh, the svn wasn't compressed. even with some file commonality, the 2x compression improvement is more than i'd have guessed. 21:04
mugwump the svn is included in that
the STEVAN-trunk head has the svn history
stevan_ wait,.. what did I do? 21:06
mncharity lol
[particle] stevan_: check your trunk 21:07
mncharity we had p6 all finished, by there was a bug in one of your commits, so it's all gone. mugwump is trying to recover it, but, not much hope.
*but 21:08
PerlJam it's going to be sad christmas this year then
think of all the children!
stevan_ mncharity: yes, its all part of my plan to keep p6 as just vaporware so people will use Moose
mncharity lol :)
stevan_ [particle]: what trunk ?
pugs_svnbot r20058 | gbacon++ | Build the library, but the executable still doesn't link. Why aren't we building an Executable with cabal? 21:09
diff: dev.pugscode.org/changeset/20058
lambdabot Title: Changeset 20058 - Pugs - Trac
mncharity re sad christmas, that does restore focus, doesn't it. ok. almost end of day Mar 4. what's the right development path...
/me wonders if there is vocabulary for "right way to push through a pathless jungle".
stevan_ slash and burn? 21:10
PerlJam prefers a machete, but that's only for personal use
If we're pushing more people, we need more machetes :)
mncharity re r20058, building... 21:11
mncharity re "slash and burn?", interesting - I came across the concept of "charing"(sp?) a day or two ago. Instead of slashing down and burning the pile, you instead char the trees away. so most of the carbon hangs around, improving the soil, rather than blowing away. 21:12
mncharity but anyway... 21:14
() repackage pmurias's kp6 p5 backend as p5 class with an api, so it can be driven by other ast's... eh, the api is likely most of the weight.. :/ 21:15
hi stevan_ :) how goes Moose land? 21:16
I notice the ANTLR folks have the convention of taking picturees of themselves holding their hands up like antlers. an artistic opportunity for Moosen?
stevan_ :) 21:17
mncharity: Mooseland is sell, we need to do a big documentation push though, too many new features without docs :) 21:18
mncharity www.antlr.org/images/jean-antlr-head.jpg 21:19
mncharity ooh, "many new features"? a literature pointer or summary? 21:31
gbacon: ended with src/Main.hs:17:7: Could not find module `Pugs':, which sounds like you described. 21:32
jrockway everyone took antler pictures at YAPC::NA 2006 for some reason
gbacon mncharity: right, and if you try to help it out with -isrc, it tries to rebuild the whole stinkin' thing 21:33
mncharity but lots of happy compilation! :) progress!
:/
hmm...
gbacon ? 21:34
mncharity gbacon: nothing, just looking at our build system. 21:42
gbacon a big pile of hacks :-) 21:43
mncharity even better, interleaved strata of layered hacks :)
pugs_svnbot r20059 | cosimo++ | [t/spec] Test suite refactoring. Moved .fmt tests under S02 21:45
diff: dev.pugscode.org/changeset/20059
lambdabot Title: Changeset 20059 - Pugs - Trac
cosimo_ hi everybody 21:45
mncharity greetings 21:47
mncharity TimToady: disregarding debugability, do you have any feel for whether/how well STD.pm would work if run on a completely backtracking engine? ie, everything is interpreted as a rule or regex, and even the operator precedence parser will back out of things. ie, is the non-backtracking a requirement for correct parse, or simply efficiency? 21:55
gbacon cool, linking! 21:58
not cool: internal urp
mncharity eg, C parser have the parser->lexer feedback hack to distinguish identifiers from typenames, but if your willing to backtrack, you don't need it. I'm wondering if STD might simplify, if parts of it could be ignored, if one was willing to backtrack extensively. 21:59
mncharity wonders what an "urp" is, goes to see...
gbacon breakage
pugs_svnbot r20060 | gbacon++ | Linking pugs now, but pugs complains: Invalid grammatical category: "Bool"
diff: dev.pugscode.org/changeset/20060
lambdabot Title: Changeset 20060 - Pugs - Trac
mncharity Fortunately there's no Bool in ext/Math-Basic/lib/Math/Basic.pm, so it's "not my fault". There are some Bool's in Prelude, but the file hasn't been changed in half a year. 22:04
gbacon yeah, and it's worrisome that Bool is occurs very early in the prelude 22:07
s/is occurs/occurs/
mncharity oh, that's weird. 22:10
pasteling "gbacon" at 24.214.236.114 pasted "full list of known grammatical categories" (1 line, 936B) at sial.org/pbot/30418
mncharity I svn up'ed to 20060, and rather than make clean, first just tried make. That got me to Bool. then did a make clean; perl Makefile.PL; make, which instead failed with 22:11
ghc-6.8.2: unknown package: Pugs
mncharity but if I run make a second time, I again get Invalid grammatical category: "Bool" 22:13
which sounds like a parse failure?
or not. maybe 22:14
gbacon mncharity: in util/build_pugs.pl, you should see Pugs mentioned
gbacon mncharity: no, it's confusion deep in the internals 22:15
mncharity re 'util/build_pugs.pl, you should see Pugs mentioned', do. ?
re deep confusion, ah, ok. 22:16
gbacon do you see Pugs added to the package list in util/build_pugs.pl? 22:17
mncharity TimToady: at the other extreme, any feel for how the grammar would fare if poured into a LR(1)? Or LL(1)? Or k? I'm just trying to get a feel for the space of options. 22:19
If not, np.
looking...
gbacon mncharity: should be around line 611 22:20
mncharity yes 22:21
gbacon mncharity: sorry, 466
the build shouldn't complain about unknown Pugs module
mncharity well, err, right
(re 466)
I don't suppose it could be something silly like "Setup: /a/home/jobsearch/a_bit_of_perl6/pugsbuild/third-party/installed/share/doc/Pugs-6.2.13/LICENSE: copyFile: does not exist (No such file or directory)" preventing Pugs, just compiled, from becoming available for the next command. 22:24
I note the error does not occur the second time make is run. 22:26
Instead there is Installing, and Registering Pugs-6.2.13, and reading/saving/writing package info/config. 22:28
gbacon mncharity: there should also be a mkdir in util/pugs_build.pl 22:30
at line 619
yes, if it can't copy the license file, the install does not finish 22:31
mncharity by rm -rf third-party/installed/share/doc/Pugs-6.2.13/ I can repeat the cycle. first make fails to find Pugs, second works, fails with Bool.
gbacon maybe the whole hierarchy needs to be created :-/ 22:34
phbbbbt
mncharity however, rm -rf third-party/installed/share/doc/Pugs-6.2.13/* is _not_ sufficient to cause make to fail on Pugs.
so it's exactly the (non)existence of third-party/installed/share/doc/Pugs-6.2.13/ which is the problem
gbacon good catch
meppl good night 22:35
gbacon mncharity: update and retry 22:36
pugs_svnbot r20061 | gbacon++ | Create Pugs-$version/LICENSE so build can succeed.
diff: dev.pugscode.org/changeset/20061
lambdabot Title: Changeset 20061 - Pugs - Trac
mncharity good night meppl
meppl ;)
mncharity trying ... 22:37
mncharity looks good with the rm's. I'll try a clean build? (which takes me out of action for a bit) 22:38
gbacon let me know how it goes
mncharity will do, trying... 22:39
gbacon mncharity: sorry I didn't credit you in the commit message :-( 22:40
mncharity so very np. :) 22:41
(steel gray twilight sky reflected off of rain-wet sloped black roofs - pretty) 22:43
mncharity :) make clean; perl Makefile.PL; make, went straight to the Bool error. 22:48
gbacon [deadpan] hooray... 22:49
mncharity lol
gbacon bitrot-- 22:54
mncharity /usr/local/bin/perl util/gen_prelude.pl -v -i src/perl6/Prelude.pm -p ./pugs --output blib6/lib/Prelude.pm.yml fails with a largely empty Prelude.pm. Looking at/for Parse-YAML 22:57
./pugs -CPIL1 -e '3' is sufficient to generate the Bool error, but ./pugs --help is not. 23:00
mncharity ./pugs -c -e '3' also give s the Bool error. 23:02
pugs_svnbot r20062 | buchetc++ | [t] testcase derived from S17
diff: dev.pugscode.org/changeset/20062
lambdabot Title: Changeset 20062 - Pugs - Trac
mncharity -e "" also errors. 23:03
pugs_svnbot r20063 | buchetc++ | [Spec] start re-oraginze S17 23:07
diff: dev.pugscode.org/changeset/20063
lambdabot Title: Changeset 20063 - Pugs - Trac
perljunkie How is perl 6 coming ? 23:27
PerlJam perljunkie: nicely. 23:29
perljunkie Sweet.
cognominal_ timtoady, I think there is an error in S02, the table de precedence is wrong, according to the strangely titled sect ion feather.perl6.nl/syn/S03.html#List_...precedence comma bind tighter the assignement 23:47
lambdabot Title: S03
cognominal_ I meant error in S03
sleep time 23:50

Powered by Cro and the Raku® Programming Language.

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

Thank you!