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.
luqui (still talking, actually) 00:00
just a sec
pdcawley patently not reading over your shoulder.
LUQUI'S ROOMMATE => STOP DISRACTING LUKE!
putter TimToady: "they" plural, s/user/users scattered in space and time/
luqui okay, I think one semicolon is useful 00:01
because there is definitely a place for options and parameters
putter Limbic_Region: good point. though a lot of it was off line. searching for A B C and Int Num multiples might do it.
luqui but I might take that back, actually
TimToady semi is a method for yielding to another candidate you have more confidence in than yourself 00:02
pdcawley heads to bed. G'night all.
Limbic_Region wasn't volunteering
putter good night pd
luqui TimToady, how so?
TimToady night
luqui 's brain isn't thinking very clearly 00:03
I am heavily multitasking, that could do it
TimToady 's brain never thinks very clearly
drat, company meeting...will be distracted...
luqui get to work! 00:04
:-
(I will leave the final character to semantic inference)
allbery_b that, or you're speaking prolog at us
luqui hell_freezes_over :- luqui_ever_speaks_prolog_again 00:05
oh no!
there it goes...
putter re dependency management, here's a strawman proposal... 00:09
diotalevi no.
% which is just the answer to the question should you ever ask it. 00:10
putter ???
diotalevi % this is just some prolog. 00:11
TimToady everything that is make-like is bogus; everything that is non-make-like is also bogus. :)
putter ok.... continuing... ;)
diotalevi thinks an after work beer is non-make-like and isn't bogus. 00:12
TimToady pay no attention to the peanut gallery behind the curtain
allbery_b seems to have touched off a stream of consciousness
putter input is a "little" text file in cvs (using a ruby dsl I think). tasks (perhaps hierarchical) with skills required, and dependence (single flavor) on other tasks. 00:14
generate dot vis graph.
random hacker can look at pretty graph, see what needs doing that they can handle, and its connection with the whole. 00:15
00:15 lisppaste3 joined
putter rev to permits adding "stuck" entries, task "foo" is stuck because task "bar" doesn't yet do "mumble". 00:16
s/rev/rev2/
err, s/rev to/rev two/
so task "christmass" depends on tasks "test suite" and "working p6", and... 00:17
and "working p6 passes test suite", and...
thoughts? questions? 00:18
I basically want a way for anyone stopping by to see that one with p6 and minimal hs skill can be doing "move primitives from hs to p6", on which "p6 prelude" depends, on which "all implementations use a common p6 prelude" depends, etc. 00:20
TimToady you should bounce this off of obra
putter obra?
TimToady Jesse Vincent
putter it's an area of interest for him? he has done something similar? 00:21
TimToady Hiveminder
maybe
putter err, lack of simplicity. I should have just asked. "great. why?"
ahhh
hmm.. 00:25
I'm thinking of the strawman as more like the existing TASKS (or some such) file, but with dependencies, so it can be show as a graph. hmm, and that might be a better model than the dsl. 00:27
00:30 luqui joined
putter strawman proposal 2 - create a more regular version of TASKS, with shortnames/ids, and dependency info. and trivial little tool to make graph from same. 00:30
I'd like to easily see, for instance, what kp6 needs, and what it provides. how it fits in to the whole project. 00:32
thoughts? (though I remember this is the wrong time of day for this...:/ )
00:33 Psyche^ joined
putter I wonder if it could be overlayed on pod format. directory tasks/ of pods. grep for TASK and DEPENDS. 00:35
00:37 cls_bsd joined 00:39 mako132_ joined 00:50 Psyche^ is now known as Patterner 01:05 jferrero joined 01:07 spoops joined
putter anyone still around who has been watching the project? 01:41
01:44 Psyche^ joined
diotalevi thought prolog would be a nice way to write that thing. *shrug* 01:45
TimToady how do you define "watching"? 01:46
luqui diotalevi, good point 01:53
01:58 Psyche^ is now known as Patterner 02:02 SamB joined 02:06 ashelyb_ joined 02:16 I_want_job joined, I_want_job left 02:18 miyagawa joined 02:19 araujo joined, bonesss joined
putter TimToady etal: I'd like an "expert" to interview. "Is tomorrow Christmass?" "How does today principly differ from Christmass?" etc 02:26
"Is self hosting required for Ch?" "Could p6 on ghc suffice for Ch?" etc 02:28
"What are the strongest areas of the test suite?" "What are the weakest?" "What is missing entirely?" 02:29
"What are the strongest parts of the spec?" "The weakest?" "Missing entirely?"
re prolog, I thought of that. but... it increasingly seems like nothing computation is really needed. the pretty picture/whatever is primarily a driving factor for asking questions like above, secondarily a mechanism for documenting the result. 02:31
It's project analysis/characterization which is missing/unwritten, rather than a tool to compute over it. 02:34
allbery_b meant the ":-" halfsmiley, not suggesting writing anything in prolog :) 02:35
allbery_b hasn't touched prolog in some 10 years, and is kinda fuzzy on it now. and would like to stay that way :>
putter ah...
diotalevi only said prolog because you can express trees nicely there.
allbery_b bad enough they were tossing around haskell's map and foldr in prolog on #haskell the other day :) 02:36
putter DOT is quite nice. www.graphviz.org/Gallery.php
lambdabot Title: Graphviz
diotalevi regularly tosses around p5 regexps in prolog on p5p. ;-)
obra putter: start collating those tasks. no matter what format. 02:39
putter indeed. 02:41
02:41 diotalevi left 02:42 dmq joined
obra and god knows we don't have enough visibility into what needs doing in useful bite-sized chunks 02:43
03:01 nekokak_ joined
audreyt putter: I did actually give a well-received academic talk in ICFP's Dynamic Language Symposium, which is why I got invited to POPL :) 03:07
putter: however, I was unable to attend POPL this year, as it turned out that I went back to Taipei from the airport, some 20 minutes before boarding the flight, for certain private reasons 03:09
so I don't even have to write the talk now ;)
but now I need to proc certain nonfun matters instead of IRC'ing. bbiab 03:10
&
putter Oh my. I'm so sorry. Let us know if there is anything we can do to help. 03:14
audreyt yes: do try to optimize for fun even under constrained resources :) 03:16
Juerd: I've installed trac on feather a few days ago; you okay with me enabling it and link it with the svn now hosted on feather?
luqui audreyt, trac is a spam magnet. require an account to add bugs. 03:17
audreyt require an account to do any write operation.
luqui (probably already planned that)
audreyt and commitbit manages those accounts
luqui ohk
audreyt (yes, and already set up for $job)
commitbit++ # useful even for intranet projects 03:18
luqui well, enjoy nonfun matters... :-/ 03:19
audreyt goes back to the quest of contacting the travel agency for ticket refund, which is looking unlikely now at this Saturday noon
yeah, I'll try :-/
bbl
03:23 ayrnieu joined
audreyt (fwiw, it's all nonlethal and solvable, so don't worry too much, I'll eventually emerge :)) 03:24
putter ok. thanks. 03:25
03:31 szabgab joined
szabgab I still cannot compile Pugs, installed v6 and its many prereqs from CPAN is that necessary or is that what is actually causing the problems now? 03:34
03:36 [particle] joined
szabgab certainly perl Makefile.PL of Pugs did not complain on any missing modules (actually if there are CPAN prereqs for building Pugs as listed in INSTALL, would it make sense to add them to Makefile.PL as well? 03:36
I get "Unknown error" while trying to "make" pugs 03:38
03:40 mdiep joined 03:53 mdiep joined
luqui szabgab, what version of ghc? 03:53
or, where in the build process did it die?
oh, well, I have to go. I think pugs now requires ghc 6.6... 03:54
03:57 brando joined 04:02 brando joined
szabgab I have compiled and installed ghc6.6 and pugs seems to use that ghc 04:02
audreyt nopaste the build log? 04:03
szabgab hi audreyt ! 04:09
04:09 justatheory joined
audreyt szabgab: greetings 04:10
szabgab I went back to a clean PERL5LIB, so I am not sure now, why does Pusg complain about missing v6.pm ?
is that something that comes with Pugs or is that a prereq I should install from CPAN ? 04:11
audreyt that is something that comes with Pugs
szabgab so perl Makefile.PL did not seem to complain about missing it but then make stops lacking that module
audreyt can you paste the build log (or just the error lines) in
sial.org/pbot/perl6
?
lambdabot Title: sial.org Pastebot - pasteling
szabgab do you mean what I get on the screen? or is there a log file ? 04:12
audreyt what you get on the screen 04:13
make > make.log
or something like that
pasteling "szabgab" at 192.117.110.30 pasted "error while compiling pugs on a clean (?) install" (12 lines, 945B) at sial.org/pbot/22341 04:14
audreyt you have "lib" in your PERL5LIB somehow 04:15
and it picked up the perl6 File::Find
szabgab yes, is that causing problesm ?
ah I see
audreyt yes.
let me think about how to fix that
there does not seem to be a good general solution :/ 04:16
try taking out 'lib' from PERL5LIB and see if 'make' now passes? 04:17
szabgab I removed lib for PERL5LIB for now and so far it seems to be compiling ok
audreyt a general solution is to put our perl6 libs under lib6/
but for now, a warning/error in the main Makefile.PL to check for blind "lib" in PERL5LIB will do 04:18
can you add one to our main Makefile.PL? line 64ish should be a good place to put a "die" for it 04:19
also check for PERLLIB.
szabgab I'll try
audreyt szabgab++ # thanks 04:20
szabgab thank *you*
04:25 SamB joined
svnbot6 r15063 | szabgab++ | Makefile.PL - check for naked 'lib' in PERL5LIB and PERLIB 04:28
szabgab sg else: in INSTALL v6-alpha is mentioned. If I am not mistaken it is totally separate project and has nothing to do with Pugs, right ? 04:29
audreyt it is a separate implementation 04:30
part of it (the grammar engine) is used by ./pugs
otherwise no, these two are independent, 'cept they are under the same repository 04:31
szabgab so do I need to install that too ?
audreyt no
not at all :)
szabgab OK, Ill try to make that clear in INSTALL
audreyt if that's not clear enough... update INSTALL to say that
heh, preempted
szabgab++
svnbot6 r15064 | audreyt++ | * Makefile.PL - Fix szabgab++'s PERLLIB check - it was typo'ed as PERLIB 04:40
r15064 | audreyt++ | and somewhat Unix-specific; changed to use $Config{path_sep} as separator
r15064 | audreyt++ | and catch cases like ./lib ././lib .//lib etc.
r15065 | szabgab++ | INSTALL - put the v6-alpha installation notes in another file: INSTALL.v6-alpha 04:43
04:46 luqui joined
putter TimToady: ping? 04:48
04:50 awwaiid joined 04:52 spoops is now known as spoop
TimToady putter: pong 05:08
putter hmm, let's see. is there an easy way to post a png image... 05:14
TimToady poke it into a public_html somewhere? 05:19
putter www.vendian.org/mncharity/Temporary...leteme.png
which is sort of putters mind-map of p6. I need more task fodder. thoughts? 05:21
TimToady would be nice if a unified AST was in there somewhere 05:26
putter a unified ast gives you ast macros, and... what else? 05:27
TimToady event model
hopefully it gives you consistency across platforms at a lower level than syntax 05:28
I/O model
docs processing flow
putter docs processing flow? 05:29
TimToady there's a reason I'm not a project manager. :) 05:30
putter :) 05:31
05:31 kanru joined
putter what is a document processing flow? 05:31
image updated 05:32
TimToady how the documents get from where and how they're written to where and how they're read by the end user
putter ah, ok
image updated 05:34
let's see...
documentation big picture... the spec gets written, you write a book, what else happens? a reference manual? quick reference. tutorials. what else? 05:36
feel free to punt whenever. getting towards 1am here. 05:37
TimToady need to get up at 5:30 myself, sigh. 05:38
but it's a good start
putter ok. I don't have working svn at the moment, so I'll clean up and paste the DOT file. 05:39
TimToady maybe make writing the emitters to the platform a separate step... 05:40
though probably done in parallel up to a point.
Oh, the MOP!
05:41 sunnavy joined
TimToady and maybe a circularity saw if the prelude is circularly defined. 05:42
putter hmm!... 05:45
putter doesn't recall what circularly defined means vis the prelude...? 05:51
image update
szabgab ?seen nothingmuch 05:52
lambdabot nothingmuch is in #perl6, #darcs and #haskell. I don't know when nothingmuch last spoke.
TimToady things defined in terms of each other, on the assumption different platforms will nullify different parts to supply as primitives instead. 05:53
putter ah, right.
szabgab trying to run make smoke but I cannot install Test::TAP::Model from CPAN, any idea?
allbery_b "cannot install"? 05:54
it's there
szabgab I mean make test on that module fails 05:55
05:55 ofer0 joined
TimToady there is also still some work to at least move the todos to separate lines so different platforms can keep their status independently. 05:56
allbery_b yeh, force the install. there will be failures when you use it as well, but at least part of the result is valid 05:58
(but Test::TAP::Model only matters for "make smoke"; you can always fall back to "make test")
allbery_b has been getting those Test::TAP::Model errors from the start --- and his smoke test attempts alway show individual tests passing but the whole file failing
TimToady I get errors but it seems to calculate file stats right. 05:59
06:02 leed joined
putter image updated :) 06:02
szabgab ah and I was showing off with the nice HTML reports the other day but could not create them on a plain Perl 5 test suit 06:03
TimToady kinda funny to have P6 on P6 out there without anyone depending on it. :)
putter yeah. I suspect there is another path which is "some p6, not Christmas but enough to run a compiler written in p6, used to bootstrap p6 on p6 on mumble". 06:05
for some backend mumble
TimToady arguably, P6 on P6 is just part of the test suite. :) 06:06
putter :)
06:07 nipra joined
putter I guess its basically replacing a core written in mumble, with one in p6, and then bootstrapping with a mumble emitter also written in p6. 06:08
TimToady depends on what you mean by "core" 06:10
writing a VM in Perl 6 is specifically not on the agenda any time soon... 06:11
putter :)
TimToady though maybe an OS or two...
putter core + emitter + prelude = p6 06:12
"runtime" perhaps
err, core + emitter + compiler + prelude = p6
TimToady still think "core" is far too overloaded 06:13
putter agreed 06:14
TimToady it's a machine, which may or may not be virtual...
putter does everything use a compiler written in p6? even pugs?
TimToady that's the plan
putter ok
06:15 BooK_ joined 06:16 kanru joined
putter "core" -> "machine". image updated 06:20
06:21 ditto joined
TimToady I'm gonna shutup now so we can all sleep... 06:24
putter :)
DOT file up as www.vendian.org/mncharity/Temporary...leteme.dot 06:25
06:29 ditto left
putter image updated 06:31
06:47 agentzh joined, iblechbot joined 06:49 nipra joined
putter image updated 06:50
Good night all. I've no idea if/when I'll get back to the graph, so anyone/everyone should feel to take it over. 06:55
Keep up the good work. Perl 6 is *important*.
&
06:56 devogon joined
agentzh putter: sure. perl 6 is important. :) 07:11
fglock: PCR is still under development, right? 07:15
i think PCR may be a good starting point for me since it has a lot of meaningful unit tests. :) 07:16
fglock: i haven't figured out how to "run" mp6's test suite with mp6. it seems that t/*.t pick v6.pm instead. 07:18
?eval 'aa' ~~ /a**{2..2}/ 07:26
07:26 evalbot_r15026 is now known as evalbot_r15065
evalbot_r15065 *** Cannot parse regex: a**{2..2}␤*** Error: Error: quantifier not implemented: HASH(0x8fffbf0) at /home/audreyt/pugs/perl5/Pugs-Compiler-Rule/lib/Pugs/Emitter/Rule/Perl5.pm line 162.␤␤Match.new(␤ ok => Bool::False, ␤ from => 0, ␤ to => 0, ␤ str => "", ␤ sub_pos => (), ␤ sub_named => {}␤) 07:26
agentzh ?eval 'aa' ~~ /a**{2..4}/ 07:27
evalbot_r15065 *** Cannot parse regex: a**{2..4}␤*** Error: Error: quantifier not implemented: HASH(0x8fffbf0) at /home/audreyt/pugs/perl5/Pugs-Compiler-Rule/lib/Pugs/Emitter/Rule/Perl5.pm line 162.␤␤Match.new(␤ ok => Bool::False, ␤ from => 0, ␤ to => 0, ␤ str => "", ␤ sub_pos => (), ␤ sub_named => {}␤)
agentzh weird. i see there is a qunatifier implementation in Pugs::Runtime::Regex. maybe the rule parser can't handle that right now? 07:29
another question: is /a**{4}/ a valid perl 6 regex? 07:31
allbery_b suspects it should not be since ** is already a quantifier 07:33
agentzh allbery_b: ** is a meta construct according to S05. 07:34
perl 6 regex /\d ** {1..6}/ is equivalent to /\d{1,6}/ in perl 5, i think. 07:35
allbery_b had forgotten that change to regex syntax
agentzh allbery_b: ;-)
pe 07:36
allbery_b dislikes {} as a rule anyway, it's devilishly hard to handle sanely
agentzh why? 07:37
allbery_b the runtime behavior tends to be atrocious. perl and pcre note it as a pathological case; openbsd's standard regexp implementation can be easily made to take infinite time to match, and glibc's to take infinite space 07:39
07:40 mdiep joined
agentzh *nod* a lot of backtracking is involved here. 07:41
allbery_b (6-deep nested {1,6} expressions triggered the obsd and linux/glibc behavior) 07:42
agentzh ah 07:44
svnbot6 r15066 | agentz++ | PCR - added some more tests to t/01-iterator_engine.t. 07:47
r15066 | agentz++ | it's not meant to be complete; i'm just playing around
r15066 | agentz++ | PCR's internals. :)
07:55 mdiep joined 07:56 jamessan joined, devogon joined, kanru joined, BooK_ joined, SamB joined, smash_ joined, wilx joined, hcchien joined, silug joined, baest joined, sri_ joined, mugwump joined, cookys joined, audreyt joined, Gothmog_ joined, LCamel_ joined, broquaint joined, nnunley joined, takesako__ joined, pasteling joined 08:01 bernhard joined
svnbot6 r15067 | agentz++ | PCR - added 4 more tests to t/03-match.t. 08:14
08:17 Aankh|Clone joined 08:22 DHGE joined
agentzh ?eval "ab" ~~ /a?b/ 08:40
08:40 evalbot_r15065 is now known as evalbot_r15067
evalbot_r15067 Match.new(␤ ok => Bool::True, ␤ from => 0, ␤ to => 2, ␤ str => "ab", ␤ sub_pos => (), ␤ sub_named => {}␤) 08:40
agentzh ?eval "ab" ~~ /a?bg?/
evalbot_r15067 Match.new(␤ ok => Bool::True, ␤ from => 0, ␤ to => 2, ␤ str => "ab", ␤ sub_pos => (), ␤ sub_named => {}␤)
08:52 larsen_ joined 09:02 ayrnieu joined
svnbot6 r15068 | agentz++ | PCR - added some more tests to t/04-rule.t 09:08
r15069 | agentz++ | PCR - some more tests to t/05-var.t 09:17
agentzh is learning p6 regex by tests. 09:18
09:23 bernhard joined, Aankh|Clone is now known as Aankhen`` 09:25 iblechbot joined 09:54 elmex joined 10:47 Salzh joined
nothingmuch szabgab: pong 10:50
10:59 kisu joined 11:06 marmic joined 11:42 BooK joined
Juerd audreyt: Please go ahead :) 11:45
11:52 jferrero joined 11:55 bwk joined 12:07 GabrielVieira2 joined 12:14 dduncan left 12:20 larsen_ joined 12:27 elmex joined 12:28 buetow joined 13:00 ludan joined 13:08 nipra joined 13:17 chris2 joined 13:26 GabrielVieira joined 13:31 weinig joined 13:42 Limbic_Region joined 13:46 weinig is now known as weinig|bbl 14:03 nivrrex joined 14:18 Yappo_ joined 14:19 cookys joined 14:20 Yappo joined 14:32 idiotmax joined 14:46 idiotmax joined, GabrielVieira2 joined, weinig|bbl is now known as weinig 14:49 spx2 joined 14:54 weinig is now known as weinig|food 15:10 weinig|food is now known as weinig 15:37 weinig is now known as autogeneration 15:38 autogeneration is now known as weinig 15:44 polettix joined 15:57 idiotmax joined 15:58 xinming joined 16:46 sonorous joined 16:55 nwc10 joined 16:56 _bernhard joined
nwc10 Offhand does anyone know which synopsis ... ??? and !!! are in? 16:56
allbery_b S03/New operatos, at minimum 16:57
...preend I cn type 16:58
17:02 brent[] joined
nwc10 ah yes. Thanks. dev.perl.org/perl6/doc/design/syn/S03.html 17:05
17:14 nwc10 left 17:55 Limbic_Region joined 18:05 neonse joined 18:12 takanori joined 18:23 Limbic_Region joined
xinming 3 18:27
18:28 mako132_ joined
Limbic_Region hrm 18:30
18:46 sonorous joined 18:54 elmex joined 18:56 justatheory joined, bcorn joined 19:05 sonorous joined
Khisanth ?eval 1 ?? 2 !! 3 19:23
19:24 evalbot_r15067 is now known as evalbot_r15069
evalbot_r15069 2 19:24
Khisanth ?eval 0 ?? 2 !! 3
evalbot_r15069 3
19:30 foo\ joined
Khisanth ?eval 0..10:by(2) 19:33
evalbot_r15069 Error: ␤Unexpected ":by"␤expecting "_", fraction, exponent, term postfix or operator
19:44 diakopter left 20:14 rhizo joined 20:24 ill joined 20:28 ill is now known as idoublel 20:29 kanru_ joined 20:36 mako132_ joined 20:49 andara joined 21:05 beppu joined 21:11 penk joined 21:17 idoublel left 21:34 Aankhen`` joined 21:41 mako132_ joined 22:07 weinig is now known as weinig|bbl 22:11 dduncan joined 22:20 prism joined 22:21 Alchemy joined, prism joined 22:26 lambdabot joined 22:27 prism joined 22:51 mako132_ joined 23:37 justatheory joined 23:46 Vex joined 23:58 SamB joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!