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.
Auzon rakudo: say 1.true 01:46
p6eval rakudo 29623: OUTPUT[1␤]
Auzon rakudo: say 0.true
p6eval rakudo 29623: OUTPUT[0␤]
pugs_svn r21400 | Auzon++ | [gsoc_spectest] moved given.t to spec/, cleaned up syntax, fudged for rakudo. 01:57
Auzon rakudo committers: t/spec/S04-statements/given.t is ready for spectest_regression:
pugs_svn r21401 | lwall++ | [STD] successfully parses heredocs and trims leading whitespace. 03:25
pugs_svn r21402 | lwall++ | [STD] list all expectations, not just ones matching first letter 05:09
TimToady STD successfuly parses about 2/30 of t/spec 05:48
*2/3
pugs_svn r21403 | lwall++ | [kv.t] must have space before -> and { 05:53
Auzon TimToady++: Very impressive! 05:54
Auzon How do you convince STD.pm to check it? 05:55
pugs_svn r21404 | lwall++ | [teststd] program to run STD over t/spec 07:02
r21405 | lwall++ | [Cursor] short circuit some () returns 07:17
r21405 | lwall++ | [next.t] fix some malformed for loops
r21406 | lwall++ | [ord_and_chr.t] some malformed double-quoted strings 07:22
r21407 | lwall++ | three more malformed tests spotted by STD 07:35
r21407 | lwall++ | (remember that if() and for() are function calls!)
pugs_svn r21408 | fglock++ | [v6.pm] implemented string interpolations \x[99] and \c[FOUR TEARDROP-SPOKED ASTERISK] 10:27
pmurias v6: say 4 10:31
moritz_: we don't have v6.pm in the evalbot?
pmurias (lex poisoning)-- 11:05
pugs_svn r21409 | pmurias++ | [pixie] code vanishing protection 11:12
r21410 | pmurias++ | [STD5_dump_match] added an option to mark arrays with a circumfix [] 11:13
pugs_svn r21411 | pmurias++ | [smop] 11:19
r21411 | pmurias++ | s1p handles integers and multiple parameters now
r21411 | pmurias++ | + SMOP__S1P__Array_create()
r21411 | pmurias++ | + Str.perl
pmurias is adding a use v6; to switch to perl6 in C a sane idea?
pmurias is there a p6 C grammar somewhere? 11:28
pugs_svn r21412 | pmurias++ | [elf] converted the elf AST description into perl6 11:30
r21413 | pmurias++ | fixed teststd to work on systems without . in $PATCH 11:37
pugs_svn r21414 | pmurias++ | [teststd] added use strict 13:48
pugs_svn r21415 | rhr++ | [ord_and_chr.t] typo - '{' is 123 13:52
pmurias TimToady: i'm trying to inline tryfile into teststd (wraping it in eval), so far i have found problems with $::HIGHWATER and $::HIGHEXPECT, but when even when resetting them manualy the subsequent parses don't work 14:19
are you aware of things which might be causing that? 14:20
TimToady pmurias: well, anything declared our that isn't re-initialized in either Cursor.pm or STD.pmc 14:45
also, anything declared my at file scope 14:46
(that isn't functioning as constant)
in theory %lexers could get poisoned, but it shouldn't happen 14:47
it's most likely one ours at the front of STD.pmc
s/on/one of the/
gah, can't tipe this morning 14:48
(will be afk for a couple days starting in an hour or so)
pmurias TimToady: the lex/ directory gets poisoned often 15:07
pmurias i'll check all globals (not only directly in main::) and lexicals than 15:20
pmurias * then 15:30
TimToady lex/ poisoning would affect separate tests as well 15:54
TimToady it's only semantic changes to STD and Cursor that can affect lex/; otherwise it's immutable data 15:56
'k, heading out for OSCON now...
bbl & 16:00
pugs_svn r21416 | fglock++ | fixed char class with spaces 16:50
pugs_svn r21417 | Auzon++ | [gsoc_spectest] testing that block eval is not implemented. also cleaned up eval.t. (added 1 test) 18:33
pugs_svn r21418 | fglock++ | [PCR] unicode fixes 18:43
pmurias fgl 18:44
sorry
Auzon pmurias: Are you a rakudo committer?
pmurias Auzon: no 18:46
i'm not even a rakuo harcker
* rakudo hacker
Auzon ah, ok 18:47
rindolf Hi all. 18:52
Auzon Hi. 18:53
pugs_svn r21419 | fglock++ | [v6.pm] version 0.022 # 827 tests pass 18:54
pmurias rindolf: hi 18:58
rindolf Hi Auzon , pmurias
pmurias rindolf: want to do some perl6 hacking or looking for casual discussion? 19:02
rindolf pmurias: just discussion.
pmurias: still didn't really start with Tucan or Spark. 19:03
pmurias: email to @perl.org works now.
There was an email I wanted to write.
Should I blog about the OSDClub "puts ruby" meeting on use.perl.org?
OSDClub Tel Aviv == Tel Aviv Linux Club + Perl-IL etc. 19:04
pmurias "puts ruby" = ?
rindolf is watching the Gilmore Girls. 19:09
4th Season.
pmurias: a presentation about Ruby. 19:10
By ik_5
pmurias dosn't watch tv
* doesn't
rindolf: why is it called "puts ruby"? 19:13
rindolf pmurias: it's from torrent.
pmurias: because puts is like say in Ruby.
I think.
pugs_svn r21420 | fglock++ | [PCR] unicode fix
r21421 | fglock++ | [v6.pm] require new PCR 19:14
pmurias a "say perl" would be a better name
(for a perl presentation) 19:15
pmurias real time html rendering while editing code in firebug is cool 19:18
pmurias what is the recommended way to ask for an AST review? 20:08
should i mail p6-compiler, it seems to contain mostly rakudo patches and bug reports 20:11
pmurias nothingmuch: should the metainformation for an object be stored using a special API provided by the object representation or is it dependend on the metaclass and should be stored using the API for normal attributes? 20:20
nothingmuch what kind of meta information? 20:23
pmurias nothingmuch: the methods an object has 20:24
nothingmuch that's altering the metaclass
it has it's own API
its
or if you're editing the object's proto
then it's still the same metaclass api
pmurias hm 20:25
nothingmuch it isn't the representation though
pmurias and in prototype based oo?
nothingmuch the best answer I have for that is "it depends" 20:26
but technically I would say since the meta and the repr and the object in prototype oo *can* be different in the pure sense
then still not the same
by different I mean it might be the same data, the same object, but a different responder
pmurias ruoso sketch the api like this www.perlfoundation.org/perl6/index....mop_oo_api 20:27
lambdabot Title: SMOP OO API / Perl 6
pmurias * sketched
nothingmuch yeah
note the '^'
this denotes it is a meta call
also, this is specific to perl 6 prototype OO 20:28
it's useful as a base API
pmurias the ^! is an object representation call
nothingmuch but for some models it doesn't make sense
in SMOP land repr is the meta
if your question was about some hypothetical object viewable from perl 6 then the answer is that they may be different 20:29
if you're asking about SMOP, check with ruoso but in general meta calls are made on the repr
pmurias i was asking because i want to implement this
nothingmuch implement where?
pmurias in SMOP ;) 20:30
nothingmuch oh
definitely ask ruoso
i'm not your guy =
=)
pmurias i just want to know what you think 20:32
nothingmuch IIRC i agreed with ruoso
but I'm not fully up to date on the model
i don't know if it diverged since the PTP 20:33
W
pmurias what do you mean by the meta in "in SMOP land repr is the meta"
the metaobject instance is different from the object representation 20:34
nothingmuch i meant that all meta calls are goverened by the repr IIRC 20:35
the confusing bit is that we use 'meta' to mean one of two things
well, one of like 432 things
but mainly it's either the instance of the metaclass, or the concept of what governs the object's interface 20:36
i meant the second in this case
i hope that clarifies instead of confuses ;-)
pmurias it mostly generalised ;) 20:37
nothingmuch at any rate, conceptually the representation type and the object interface stuff don't need to be the same
pmurias there are 3 things which govern the interface the RI,the repr and the metaclass instance
nothingmuch typically the representation is a product of the meta stuf
define govern ;-)
as I see it there is only one: 20:38
pmurias influence
nothingmuch at runtime it's the RI
at compile time it's the meta class instance, for creating the repr and the RI
and the repr kinda ties the knot
pmurias different the way the ruoso's previous attempt in smop tried to do it 20:39
nothingmuch well, again, i'm not 100% up to date
pmurias the metaclass instance is also used in the runtime
nothingmuch that's the RIs business though
nobody except the RI knows that the RI is delegating to the metaclass
so on the outside looking in only the RI is responsible 20:40
pmurias aren't the object representations supposed to be indistinguishable from the outside? 20:41
nothingmuch at some point the RI has to be consulted
the repr is kind of a meta meta
since it contains enough info to locate the RI 20:42
or maybe I'm being confused
anyway, i'm sorry but I have to get back to $real_work, and my brain is not fully functional today
so I can't multitask
hunt down ruoso, he'll be much more helpful
pmurias i would have to track him down ;) 20:43
pmurias i anybody else responsible for the meta stuff? 20:43
nothingmuch i dunnop
i only ever discussed this stuff with ruoso/audreyt/TimToady to my recollection 20:44
i'm pretty out of the loop
pmurias i'll implement it the way i would like it to be than ;) 20:45
nothingmuch at any rate run it by ruoso 20:46
pmurias ok
zamanfou is now away: In the rain my tears are forever lost. 21:40

Powered by Cro and the Raku® Programming Language.

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

Thank you!