pugscode.org/ | nopaste: sial.org/pbot/perl6 | pugs: [~] <m oo se> (or rakudo:, kp6:, elf: etc.) (or perl6: for all) | irclog: irc.pugscode.org/
Set by Tene on 29 July 2008.
ruoso @tell pmurias, the compilation of p6opaque is failing, am I missing some command to run? 01:29
lambdabot Consider it noted.
pugs_svn r22507 | ruoso++ | [smop] preparing to implement preludescope. 01:32
ruoso @tell pmurias, I've started preludescope, it declares a m0ld block where we can define things that goes into this special scope... we then just need to use it as the ultimate outer scope when starting to parse...
lambdabot Consider it noted.
pugs_svn r22508 | lwall++ | [STD] clean up nibbler output 03:34
r22508 | lwall++ | [viv] now successfully produces transforms STD.pm to ast(ish)
TimToady s/produces //
notbenh is there a way to modify class Object from the outside? I would like to try and plug in a method, cant seem to find a way to do that in all the docs. Have I missed that? 05:57
Tene Yes. You're looking for 'is also' 05:58
class Object is also { ... } 05:59
notbenh Tene: cool thanks
pmurias ruoso: i asked this before but seemed why are some files in src prefixed with smop and others aren't? 08:23
pugs_svn r22509 | pmurias++ | [smop] a pregenerated sm0p5 shouldn't be in version control 08:38
pugs_svn r22510 | pmurias++ | [smop] fixed sm0p-- 09:43
pmurias ruoso: hi 09:44
pugs_svn r22511 | pmurias++ | [smop] reverted 22507 as it create segfaults, and after trying to get it to work i belive that plugging it into the loading sequence in smop_internal.h unnesesarly complex and frail 09:59
pmurias ruoso: what i belive is that instead of having a buch of .c/.sm0p files with a complicated loading sequence we should seperate stuff as a bunch of .ri/.m0ld files which would state their dependencies on each other 10:07
ruoso pmurias, the prefix in the files is plain inconsistency on my part :) 10:17
pmurias, hi
pmurias, that would be a great idea... the way they are now are the first thing I could think of... 10:18
we need some way to make it easier to add things to "prelude" 10:19
since the "prelude" will be mostly defined by the things that will be in src-s1p/
pmurias, about the revert... that was the part I was starting to wonder if it should be or not as part of "libsmop" 10:21
but maybe it should just be outside smop_init();
something like smop_prelude
or smop_main() 10:22
pmurias should the prefix be removed or inserted everywhere? 10:27
ruoso maybe removed, since they're already in smop/src/
then we have a usefull prefix, 10:28
which is s1p
native
slime
smop_lowlevel could be called simply lowlevel... 10:29
pmurias the lexical prelude creation could be pure m0ld 10:32
it could keep everything linked into libsmop for now, but assuming we want to target non-p6 stuff in the future too (for example python) in should be seperated at least conceptually 10:35
ruoso: should i change the paths?
pugs_svn r22512 | pmurias++ | [smop] removed the smop_ prefix from source files 11:16
ruoso pmurias, paths? you mean file names? 11:17
pmurias yes
changing the file names changes the path in a way ;) 11:18
do you think storing the ri's in a "stash" object instead of c variables would be a good idea? 11:19
ruoso pmurias, it probably would... 11:20
but some RI's need to be accessed directly
the thing is that this basic types are actually an exception...
the real types won't be available as C symbols 11:21
pmurias only the native ones, not the S1P ones should be speciall? 11:24
* special 11:26
rakudo_svn r31668 | allison++ | [pdd27mmd] Merging the pdd27mmd branch into trunk for r30235 to r31667. 11:30
masak I notice that &leave is not mentioned in S29. would it be ok if I tried to write somethine about it? 11:34
what class does &?BLOCK belong to? Block?
any other classes that have a .leave method?
(for example, what does context return? always a Block? S06 doesn't say) 11:36
also, under 'caller' in S29, it says 'See "The context and caller functions" in S06.' -- the link is broken. is it really enough to refer to the informal description in S29 instead of providing an API-perspective description in S29? or is the link just a placeholder for eventual real content? 11:38
many questions. :) if someone has the ability to mentor me through this, I believe I could do much good. 11:39
pmurias masak: i think S29 should provide an API description as that is what you expect when you lookup a function, a link is ok too if it would cause too much duplication 11:49
masak pmurias: in this case, I think S06 is more hand-wavy than a real API description 11:50
S29 could provide both the link and a minimal summary, for example
ruoso pmurias, the ones that will be written in Perl 6 are not that special... but the ones currently designated as s1p_* are special 12:00
because they are not yet using p6opaque
they won't be available in the public namespace
they won't be part of the prelude scope
ruoso later & (today elections take place in Brasil) 12:01
masak happy electing!
pmurias ruoso: how hard would it be to compile smop on windows--? 12:15
ruoso: if would like to get rid of using RI's as prototypes 12:47
s/if/i
pmurias and i can either imagine using a special prototype RI for each normal RI, or using a defined but emptyish object for the prototype such as [] for array 13:21
pmurias and there remain the issue how should the empytish one be create in the second method 13:36
* remains
rakudo_svn r31674 | pmichaud++ | [rakudo]: spectest-progress.csv update: 200 files, 4437 passing tests 13:58
ruoso @tell pmurias, about compiling on windows, i have absolutely no idea (thank god) 14:03
lambdabot Consider it noted.
ruoso @tell pmurias, the prototypes should start using p6opaque as RI as soon as we have a more high-level compiler.. the example you created in test-mildew points at this very precise direction... 14:05
lambdabot Consider it noted.
ruoso @tell pmurias, I think I know how to make the prelude loading nicer... we assume all files in src-s1p/ will be "use"d by the prelude scope, and then we add a special syntax to s1p to allow importing C symbols... 14:16
lambdabot Consider it noted.
ruoso @tell pmurias like 'our multi map (Code $expression, *@values) := ¢SMOP__S1P__map' 14:18
lambdabot Consider it noted.
pugs_svn r22513 | ruoso++ | [smop] one example on the proposal on how to automatically build the prelude using the files in src-s1p 14:40
masak what generates the L<S06/...> links on perlcabal.org/syn/S29.html ? 16:05
lambdabot Title: S29
masak I'm trying to find out what to fix to make the crosslinks work again
something on feather, perhaps? 16:09
some general POD-to-HTML thingie? but how would a general tool know what S06 was? 16:10
moritz_ re 17:30
masak: tools/smartlinks.pl
pugs_svn r22514 | moritz++ | [t] move some tests to spec/, re-worked pre-post.t, some minor corrections and 17:32
r22514 | moritz++ | rakudo fudgings
azawawi hi 17:39
azawawi @seen pmichaud 17:40
lambdabot pmichaud is in #perl6. I last heard pmichaud speak 2d 3h 31m 31s ago.
rakudo_svn r31680 | moritz++ | [rakudo] guard Str.split and Str.comb against zero width regex matches. 17:55
r31680 | moritz++ | Also refactored Str.comb to use regex(string, 'continue => pos)
pugs_svn r22515 | moritz++ | [t/spec] some test enhancements, corrections and fudging for rakudo. 17:58
rakudo_svn r31681 | moritz++ | [rakudo] five more files for spectest_regression 18:07
pugs_svn r22516 | moritz++ | [README] mention pugs-commits google group 18:53
mncharity re pugs-commits, so do we continue the "commit messages are part of the #perl6 IRC conversation" style, or switch to something more normal, but uglier on irc? lots of karma though. 19:51
moritz_ mncharity: I'm still fond of the svn bots, and have no plans on abolishing them 19:56
moritz_ (unless an overwhelming majority of #perl6 inhabitants votes differrently) 19:57
mncharity I certainly enjoy them. :) 20:05
moritz_ that's 2:0 for keeping that, I call that a majority ;) 20:08
s:2nd/that/them/
mncharity lol 20:10
moritz_ that's how applied democracy (with not-so-many participants) works 20:12
mncharity trying not to think about the election. sigh. real democracy works on "aren't they pretty? a confident? they can thus be your pilot. and brain surgeon. and president. no problem...". 20:14
*and
fullermd is willing to graciously accept your vote. 20:20
mncharity lol
pugs_svn r22517 | putter++ | [elf_h] Restructured the IRxBuild initializer.
r22517 | putter++ | Cheating for on_sbcl - because SBCL didn't like compiling a 60kB method with 100 embedded subs.
fullermd And I'm not even pretty! It's win-win! 20:21
mncharity first elf cheat for on_sbcl. :/ (ie, changing elf code solely because on_sbcl is having a problem with it). ah well. 20:24
etime. commit tomorrow. *wave* & 20:29
pmichaud I think we keep the svn bots -- I like them both. 20:38
the commit messages are useful for seeing a general idea of what's happening. the email messages provide more details.
moritz_ pmichaud: I had an idea on how to get a bit more attention from Perl 5 developers... 20:39
pmichaud: I noticed that Str.split doesn't work with zero-width assertions, and so I thought I might post it as a challenge to perlmonks.org 20:40
pmichaud: people don't like PIR, but if they give their solutions in simple perl, it should be easy to translate to PIR
and perhaps offer a CREDITS entry or something ;) 20:41
pmichaud moritz_: sure, I've invited that in the past. It'd be good to do it again. 20:43
moritz_ I'm short on tuits right now (exam preparation), but in about two weeks I should get around to it 20:44
pmichaud sounds great

Powered by Cro and the Raku® Programming Language.

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

Thank you!