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.
pugs_svn r18806 | avar++ | * [kp6] deleted some redundant pod 00:03
pugs_svn r18807 | fglock++ | [kp6-roadmap] added Milestone 5.2 - Working bootstrapped grammar 11:21
ruoso kp6: my $a = gather for (1 .. Inf) { say kp6++; take 1; }; print $a[10]; 13:02
exp_evalbot r18807: OUTPUT[error in Block at compiled/perl5-kp6-mp6/lib/KindaPerl6/Grammar/Sub.pm line 753, <> line 1.␤*** Syntax Error in Block: missing closing curly bracket } ␤] 13:03
ruoso kp6: my $a = gather { for (1 .. Inf) { say kp6++; take 1; }}; print $a[10];
exp_evalbot r18807: OUTPUT[error in Block at compiled/perl5-kp6-mp6/lib/KindaPerl6/Grammar/Sub.pm line 753, <> line 1.␤*** Syntax Error in Block: missing closing curly bracket } ␤]
ruoso kp6: my $a = gather { for ( 1 .. Inf ) { say "kp6++"; take 1; }}; print $a[10];
exp_evalbot r18807: OUTPUT[error in Block at compiled/perl5-kp6-mp6/lib/KindaPerl6/Grammar/Sub.pm line 753, <> line 1.␤*** Syntax Error in Block: missing closing curly bracket } ␤]
ruoso kp6: my $a; $a = gather { for ( 1 .. Inf ) { say "kp6++"; take 1; }}; print $a[10];
exp_evalbot r18807: OUTPUT[error in Block at compiled/perl5-kp6-mp6/lib/KindaPerl6/Grammar/Sub.pm line 753, <> line 1.␤*** Syntax Error in Block: missing closing curly bracket } ␤]
ruoso kp6: my $a; $a = gather { for ( 1 .. Inf ) -> $x { say "kp6++"; take $x; }}; print $a[10]; 13:04
exp_evalbot r18807: OUTPUT[kp6++␤kp6++␤kp6++␤kp6++␤kp6++␤kp6++␤kp6++␤kp6++␤kp6++␤kp6++␤kp6++␤11]
moritz_ @karma kp6 13:09
lambdabot kp6 has a karma of 1
pugs_svn r18808 | fglock++ | [kp6] fixed typo in script/kp6 14:39
pugs_svn r18809 | fglock++ | [kp6-roadmap] minor updates 15:21
moritz_ updates the html version 15:25
pugs_svn r18810 | fglock++ | [kp6-roadmap] possible next milestones 15:28
fglock moritz_: sorry :P
moritz_ fglock: no problem, it's really just two lines on the command line
pugs_svn r18811 | dlo++ | Added note about how to forcerecompile everything. 15:42
pmurias is make forcerecompile;make different than make -B? 15:44
moritz_ uhm, dunno 15:45
pmurias: yes, make -B always tries to rebuilt Makefile 15:48
and dies with
==> Your Makefile has been rebuilt. <==
==> Please rerun the make command. <==
`make -j' for kp6 produces a load of nearly 15 ;) 15:49
moritz_ script/run_tests_kp6_mp6_perl5rx.pl is broken for me 16:13
it just prints "FAILED--no tests were run for some reason." twice, and then "some tests failed" - that's all
pmurias perl script/run_tests.pl --backend=perl5rx works for me 16:16
pmurias i get the same error with the former 16:16
moritz_ maybe we should just make the former a wrapper to the latter 16:16
moritz_ no need to maintain the same code twice 16:16
pmurias we could remove the former 16:17
moritz_ even better ;)
pugs_svn r18812 | moritz++ | [kp6] 16:20
r18812 | moritz++ | * make forcerecompile
r18812 | moritz++ | * delete broken script/run_tests_kp6_mp6_perl5rx.pl, the same can be achieved
r18812 | moritz++ | with script/run_tests.pl --backend=perl5rx
r18812 | moritz++ | * updated link in docs/pk6-parrot-boundry.txt
fglock I'm working out how to split kp6 devel into tasks 16:41
I've got a list of tasks, but it doesn't correspond exactly to the roadmap
maybe splitting the roadmap into tracks, and allow people to work in parallel 16:44
moritz_ fglock: which tasks, for example? 16:45
fglock moritz_: things that belong to - longest token; namespaces; scopes; threads; regex emitters 16:46
stm, autovivification, lazy lists, junctions 16:50
the roadmap is the "main track"
moritz_ sounds like a plan
dlocaus sorry, I didn't realize the perl6 irc window was closed out 17:11
dlocaus Would anyone mind if I attempted to make Makefile.PL a bit more documented? 17:12
pmurias i don't think anyone would mind documentation
dlocaus ok
pmurias unless it's deliberately misleading :) 17:17
dlocaus I'll try and avoid that :)
pugs_svn r18813 | dlo++ | [KP6] Corrected documentation (minor)
fglock kp6: rmdir "x" 17:36
exp_evalbot r18813: OUTPUT[no method 'say' in Class 'Str'␤ at compiled/perl5-kp6-mp6/lib/KindaPerl6/Runtime/Perl5/MOP.pm line 334␤ KindaPerl6::Runtime::Perl5::MOP::__ANON__('DISPATCH=HASH(0x874ed10)', 'say') called at compiled/perl5-kp6-mp6/lib/KindaPerl6/Runtime/Perl5/MOP.pm line 162␤
..main::DISPATCH('DISPATCH=HASH(0x874ed10)', 'say') called at - line 16␤]
fglock kp6: rmdir "x"; say 1;
exp_evalbot r18813: OUTPUT[1␤]
moritz_ fglock: --secure is broken :(
pasteling "dlocaus" at 63.192.210.11 pasted "Can't locate v6.pm in @INC?" (12 lines, 1K) at sial.org/pbot/28484 17:54
dlocaus I have a v6 in "pugs/perl5/Pugs-Compiler-Perl6/lib/v6.pm", is that the one I'm supposed to use?
moritz_ dlocaus: you forgot a redirection symbol, I guess 17:55
perl script/kp6 -Cperl5 src/KindaPerl6/Runtime/Perl6/Array.pm compiled/perl5-kp6-mp6/lib/KindaPerl6/Runtime/Pe
that should > compiled/perl5-k... 17:56
s/should/should be/
dlocaus sorry, even with out the redirection symbol it doesn't work
pasteling "dlocaus" at 63.192.210.11 pasted "opps, cut-n-paste booboo" (10 lines, 975B) at sial.org/pbot/28486 17:57
fglock dlocaus: it works here
moritz_ same here 17:58
dlocaus is there a path that I am missing from PERL5LIB?
PERL5LIB=./inc/v6-MiniPerl6-lib5:src
moritz_ dlocaus: perhaps you have a miscompiled/svn conflicting/whatever repo 17:58
fglock script/kp6 should set it's own lib
try removing PERL5LIB ? 17:59
moritz_ dlocaus: just to be sure, check out the kp6 part of the pugs repo into a clean dir, and try it there
dlocaus removing PERL5LIB didn't work, I'm checking out a clean copy 18:01
fglock I usually do: svk revert -R .
then: svk up -s 18:02
fglock this is almost like a clean copy 18:02
moritz_ should 'make test' default to --backend=perl5rx 18:03
?
fglock moritz_: not sure - perl5rx is less portable 18:04
it is being used in kp6-kp6 for performance reasons 18:05
dlocaus I'm trying again, I think it has something to do with "make forcerecompile"
fglock trying forcerecompile here 18:07
dlocaus I'm always thinking I've gotten a faster computer, but everything stays the same. 18:09
moritz_ ;)
I noticed that you have to run 'make' twice after forcerecompile
dlocaus wonders if a directory could be jinxed. 18:11
moritz_ dlocaus: after a recompile it works fine for me 18:12
fglock MiniPerl6 requires 2 compilation cycles
dlocaus I'm wondering if that is what is doing it, I'm not compiling it twice. 18:13
fglock the first compilation compiles the compiler, and the second compiles the runtime using the new compiler
fglock hmm - this sounds confusing 18:14
dlocaus fglock: is there some way of fixing the Makefile.PL so it does it twice automatically for us?
fglock not sure how to do it, mp6 uses a separate script for bootstrapping 18:16
but kp6 should not require 2 cycles yet, because it still uses mp6 18:17
dlocaus Ah ha!
I think I've got it
dlocaus - start with a clean respoitory 18:20
perl Makefile.PL
make forcerecompile
make -j
fglock make -j might be a problem
dlocaus I think the -j forces the code to "jump" ahead and attempt to compile files that haven't been created yet.
pugs_svn r18814 | fglock++ | [kp6-perl5] make forcerecompile 18:22
dlocaus blames moritz_--
irclog.perlgeek.de/perl6/2007-11-01#i_135836
lambdabot Title: IRC log for #perl6, 2007-11-01
fglock it would be better to create a whole new compiled/, and later move it to the right place 18:23
dlocaus ;)
fglock trying out -j 18:24
dlocaus it works, I've done it without -j several times.
fglock yes - it doesn't work
I mean, -j doesn't work here
dlocaus which, was why I was thinking that the directory was jinxed. I was doing it differently
moritz_ takes the blame, but points out that if the dependencies in Makefile were specified complete it wouldn't break 18:25
s/complete/completely/
fglock I've got a bunch of broken files like: compiled/perl5-kp6-mp6/lib/KindaPerl6/Runtime/Perl6/Array.pm.temp
fglock I think it is trying to overwrite files that are in use 18:26
it can't compile to Array.pm because there is another compiler using Array.pm
so we really need a temp directory 18:27
pmurias kp6 uses Array.pm
fglock instead of temp files
pugs_svn r18815 | dlo++ | [kp6] Updated forcerecompile documentation to note that you cannot use "make -j"
moritz_ dlocaus++ # I was about to suggest that ;)
fglock you can use -j, but you need to change the Makefile a bit 18:31
dlocaus is still trying to figure out how everything works :)
which, is why I'm attempting to documenting the Makefile.PL => Makefile. 18:32
fglock doesn't have Makefile-fu
pugs_svn r18816 | fglock++ | [mp6] v6-MiniPerl6-JVM moved away 18:35
pugs_svn r18817 | dlo++ | [kp6] I have updated the Makefile.PL to issue a warning not to use "make -j" after doing "make forcerecompile" as 18:39
r18817 | dlo++ | this will generate odd errors.
r18817 | dlo++ |
r18817 | dlo++ | irclog.perlgeek.de/perl6/2007-11-07#i_139627
r18817 | dlo++ |
lambdabot Title: IRC log for #perl6, 2007-11-07
pugs_svn r18817 | dlo++ | I have also updated the Makefile.PL to tell us "what COMPILE" means when it is compiling.
fglock dlocaus: I'm trying to split kp6 into small tasks, as people have been suggesting 18:47
dlocaus fglock: thank you! :)
fglock: I've been trying to figure out how kp6 is put together. I've embarked on documenting Makefile.PL ( and some cosmetic clean up ) to get better understanding 18:48
fglock have you taken a look at the drawings at v6/docs/ ?
it would be nice if you could comment on which tasks you have identified so far 18:49
dlocaus I have
fglock I'm working out a tasks x roadmap diagram
dlocaus The problem is understand what is doing what. Given that one language is compiled to parse another language and then perl6 code is translated/compiled into perl5 code to provide runtime for perl6 code hasn't been making things easy ;) 18:50
dlocaus snickers
Right now, I'm seeing if I can factor perltidy out of the Makefile and into script/mp6 script/kp6 when it is used in Makefile 18:54
pugs_svn r18818 | dlo++ | [kp6] Makefile will now removing temporary (dollar)@.temp files that it creates when a compilation failes 18:56
dlocaus fglock: moritz_: would either of you mind if I updated script/mp6 to use perltidy by default, like script/kp6 does? 18:57
fglock dlocaus: no problem
pmurias fglock: why is (??{ eval '$_rule_something' }) used instead of (??{$_rule_something})? 19:12
fglock pmurias: because it must be interpolated as late as possible (i think) 19:12
pmurias aren't both forms the same thing? 19:13
fglock I think there are some tests - looking 19:14
under examples/token* - but I'm not sure which one is testing this 19:15
there is some docs in v6/docs/p6regex-on-p5regex.txt 19:16
cathyal hi 19:17
ask away?
pmurias yes
or rethink the question
whatever you prefare
prefere 19:18
* prefer
pmurias lacks spelling-fu
fglock pmurias: the prototype scripts used eval as a way to implement inheritance - v6-KindaPerl6/temp/backtracking-recursive-subrule.pl 19:24
pugs_svn r18819 | fglock++ | [kp6] moved temp/ to misc/ 19:28
r18820 | fglock++ | [kp6] moved token experiments from examples/ to misc/ 19:30
fglock several things in examples/ and docs/ could be moved to misc/ 19:33
pmurias cathyal: what about you question? 19:38
pugs_svn r18821 | fglock++ | [kp6] move docs around 19:39
r18822 | fglock++ | [kp6] set html docs to "text/html" 19:43
pmurias fglock: the eval should be removed as it is puzzling and possibly slowe 19:44
dlocaus yea! 19:45
pugs_svn r18823 | dlo++ | [kp6] 19:46
r18823 | dlo++ |
r18823 | dlo++ | script/mp6.pl now will use Perl::Tidy if available to clean up produced files.
r18823 | dlo++ |
r18823 | dlo++ | script/mp6.pl now accepts a --noperltidy option
dlocaus that was an odd pain in the butt.
pugs_svn r18823 | dlo++ |
r18823 | dlo++ | mp6.pl produces the exact same output with or without perltidy as the old mp6.pl file does.
fglock pmurias: I'll investigate that - probably needs more tests to be sure
pmurias you could remove it and test if everthing works 19:47
fglock dlocaus: cool
pmurias: right now perl5rx is broken by the delimiter bug; it needs to be fixed first
so we can actually run big tests 19:48
pmurias: is it ok to move tokeninline into scripts/ ? 19:52
and kp6-token-debug 19:53
pmurias fglock: yes
pugs_svn r18824 | fglock++ | [kp6] base dir cleanup 19:56
fglock home & 20:01
pugs_svn r18825 | dlo++ | [KP6] I updated Makefile.PL to take advantage of script/mp6.pl perltidy change (last commit?). These files are 20:15
r18825 | dlo++ | changed as a result of perltidy being run on them. At least for the files I checked, the changes appear to be
r18825 | dlo++ | cosmetic.
gbacon small bug in script/kp6 20:19
dlocaus gbacon: what is it?
gbacon print on line 385 has a spurious comma 20:20
moritz_ gbacon: do you have a commit bit?
gbacon prevents clisp tests from running
no :-(
moritz_ gbacon: /msg me your email and desired nick
and you'll get one
dlocaus I will fix that, one second 20:21
pugs_svn r18826 | dlo++ | [kp6] removed misplaced comma, (gbacon++) 20:21
dlocaus commited
nwc10 In Perl 6, if you do the Perl 6 equivalent of the Perl 5 my $foo = \&bar; 20:22
does it have a side effect equivalent to the Perl 6 equivalent of the Perl 5 sub bar;
ie, does taking a reference to a function implicitly declare a stub that affects method resolution?
pmurias nwc10: no 20:23
moritz_ no, it doesn't
nwc10 OK.
pmurias it's my $foo = &bar;
nwc10 what happends in Perl 6 if you try to invoke the subroutine via the reference and it does not exist?
moritz_ gbacon: it's a custom to add your name to AUTHORS as a first commit (and commit test at the same time)
nwc10: it dies 20:24
pmurias nwc10: you can't take a reference to subroutine which does not exist
moritz_ nwc10: at the time of referecing it, not at calling time
nwc10 um. AUTODLOAD/
you can in Perl 5.
moritz_ aye
gbacon roger 20:25
moritz_ perl 6 breaks compatibility deliberately
nwc10 so you get a reference that will die always? even if you later define sub bar {...}
moritz_ nwc10: no, you get an compile time error in the line of the assignment
nwc10 Ah.
that's what you meant by "it dies" :-)
moritz_ my $foo = &nonexisting_sub; files
s/files/fails/
English is hard ;) 20:26
nwc10 ah. interesting.
Um. typing is hard too. I'm a native speaker and my fingers still do that sort of thing for me. :-)
moritz_ ;-)
that's because you're not a native writer ;)
pmurias native typer 20:28
* typist
moritz_ typo!
pugs_svn r18827 | dlo++ | [kp6] updated the Makefile.PL documentation to include note about forcerecompile and make -j
nwc10 thanks 20:30
pugs_svn r18828 | dlo++ | [kp6] changed _patsubst_wildcard subroutine documetnation from POD to comments, so the documentation will not mess up existing Makefile.PL POD documentation 20:31
moritz_ dlocaus++ # documentation fixes 20:33
dlocaus :)
I'm going to work on another project for a while, /me waves 20:35
moritz_ 'later 20:35
pugs_svn r18829 | gbacon++ | First commit. 20:43
gbacon whee!
moritz_ gbacon: welcome to Perl 6 development!
gbacon feels great 20:44
moritz_ I never planned joining the development.. then I wanted to test pugs, and reported a dead link here 20:45
somehow I never left that channel ;)
gbacon I was active on p5p a few years ago
I miss it
is most kp6 development done on sbcl? 20:46
clisp builds take a little tweaking
moritz_ most of kp6 dev is done with the perl5 backend ;) 20:47
pmurias clisp was used at least a bit by avar 20:48
pmurias but i think sbcl is the most used lisp implementation 20:49
moritz_ but any clisp fixes are welcome as well 20:50
pmurias of course
gbacon sounds good.. looking forward to it 20:51
fridim I wish I could help writting perl5 code. Someone would tell me : we need a function that takes this type of argument, and it returns this, or do this work, but I guess it is not that simple :) 20:53
moritz_ fridim: sadly that's true most of the time
fridim I am very bad at reading other people code
moritz_ fridim: but you could find simple builtin functions to implement
fridim: for example List.shuffle might be a good starting point 20:54
you could implement that in pure perl 6 in the Runtime::Perl6::Array in kp6
moritz_ or a pure perl random number generator 20:55
fridim oh
moritz_ or any method from S29 that doesn't look too complicated
perlcabal.org/syn/S29.html
lambdabot Title: S29
fridim random number generator, using modulo :] 20:55
moritz_ fridim: that's a starting point, yes 20:56
it's not very elaborate, but it works
fridim what do you mean?
It is not a physical generator, that should work pretty well
moritz_ r(n) = (a * r(n-1) + b) mod limit 20:57
fridim I am not sure to understand the gaol
moritz_ the goal is to a) advance perl6 implementation and b) make you familiar with the code 20:57
fridim it sounds good 20:58
ggoebel do the mp6, kp6, onion projects have a collective name? 20:59
lambdabot ggoebel: You have 1 new message. '/msg lambdabot @messages' to read it.
moritz_ they're all under v6/, so you could kinda use that as a name
usually we just say kp6/mp6 21:00
fridim I saw "clisp". Is perl6 code transformed into lisp code and then compiled? 21:01
moritz_ that's one of the options of kp6, yes 21:02
fridim s/Is perl6 code/Can perl6 code be/
ok 21:03
moritz_ but the backend is not yet as complete as the perl5 one
fridim efforts seem to be relatively divided :)
moritz_ yes :/
but in a volunteer's project everybod hacks what he/she likes 21:04
i.e -Ofun
fridim -OnotReadyForChristmas :]
syle2 -OMistressForChristmas 21:07
fridim That may be some part of "divide and conquer" principle... 21:09
ggoebel is onion abbreviated as o6 or on6?
or op6 21:10
pmurias ggoebel: i don't think onion is abbreviated 21:12
ggoebel what was mp6 originally written in? 21:32
pmurias mp6 was bootstraped by lrep 21:34
which was written in perl5
ggoebel so mp6 was a one off from lrep 21:40
and kp6 was a one off from mp6
and onion is a one off from kp6?
moritz_ right 21:41
pugs_svn r18830 | ggoebel++ | add self 22:00
ggoebel I've created a v6 page on the developer wiki at dev.pugscode.org/wiki/v6 and replaced the broken links to mp6 and kp6 on dev.pugscode.org/wiki/ to reference it 22:02
moritz_ good job, ggoebel++ 22:23

Powered by Cro and the Raku® Programming Language.

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

Thank you!