Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2008-07-08

Perl 6 | Reference Documentation | Rakudo

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:10 Khisanth joined #perl6
00:35 vixey left #Perl6
00:46 Limbic_Region joined #perl6
00:55 bacek_ joined #perl6
01:06 bacek__ joined #perl6
01:06 meteorjay joined #perl6
01:07 rakudo_svn r29134 | pmichaud++ | [rakudo]: Make tools/test_summary.pl summarize TODO results also
01:09 rakudo_svn r29135 | pmichaud++ | [rakudo]: Get empty subs to properly return undef
01:09 rakudo_svn r29135 | pmichaud++ | * Refactor 'undef' to be a named_0ary
01:09 rakudo_svn r29135 | pmichaud++ | * Clean up routine_declarator a bit
01:22 cj joined #perl6
01:23 bacek_ joined #perl6
01:30 Coleoid left #perl6
01:36 PZt joined #perl6
02:14 rakudo_svn r29143 | pmichaud++ | [rakudo]:  Make 'return' with no arguments work for now. (RT #56638, masak++ )
02:14 rakudo_svn r29143 | pmichaud++ | * Patch courtesy "Carl M?\195?\164sak" <cmasak@gmail.com>
02:14 rakudo_svn r29143 | pmichaud++ | * See also note in ticket regarding proper return value.
02:27 pugs_svn r21263 | cjfields++ | add two tests for complement, squash with many-many array match
02:33 araujo joined #perl6
02:51 kanru joined #perl6
02:54 redicaps joined #perl6
03:53 PZt joined #perl6
04:10 sri_work joined #perl6
04:30 speckbot r14557 | larry++ | Clarify the role of whitespace within transliterations
04:30 speckbot r14557 | larry++ | Power up transliterations with regexes and closures
04:30 speckbot r14557 | larry++ | Formally define the implied alternation as equivalent to longest-token matching
04:39 cjfields joined #perl6
04:40 speckbot r14558 | larry++ | more clarifications and remember to thank cjfields++ this time :)
05:10 pmichaud pugs:  say \(3);
05:10 p6eval pugs: OUTPUT[CaptSub {c_feeds = [:MkFeed {f_positionals = [:IFinite 3:], f_nameds = []}:]}␤]
05:10 pmichaud pugs:  say \3;
05:10 p6eval pugs: OUTPUT[3␤]
05:11 spinclad .oO { so what kind of smiley is ]}:]} ? }
05:11 pmichaud pugs: say \(3,4)
05:11 p6eval pugs: OUTPUT[CaptSub {c_feeds = [:MkFeed {f_positionals = [:IFinite 3,IFinite 4:], f_nameds = []}:]}␤]
05:11 pmichaud pugs:  my @a = 1,2;  say \@;
05:11 p6eval pugs: OUTPUT[*** ␤    Unexpected ";"␤    expecting "@" or "::"␤    at /tmp/0UMUuTt5Vi line 1, column 21␤]
05:12 pmichaud pugs:  my @a = 1,2;  say \@a;
05:12 p6eval pugs: OUTPUT[1 2␤]
05:12 pmichaud pugs:  my @a = 1,2;  say \(@a);
05:12 p6eval pugs: OUTPUT[CaptSub {c_feeds = [:MkFeed {f_positionals = [:VList [VInt 1,VInt 2]:], f_nameds = []}:]}␤]
05:24 baest joined #perl6
05:25 Psyche^ joined #perl6
05:48 justatheory joined #perl6
05:52 rakudo_svn r29145 | pmichaud++ | [rakudo]: Refactor Perl6Capture to use Capture_PIR, implement prefix:<\\>
05:52 rakudo_svn r29145 | pmichaud++ | * This implementation is almost certainly buggy, but it's a start.
05:52 rakudo_svn r29145 | pmichaud++ | * Partially addresses RT #56684 (moritz++)
06:00 tuxdna joined #perl6
06:11 rakudo_svn r29146 | pmichaud++ | [rakudo]: More updates to empty 'return', RT #56638
06:11 rakudo_svn r29146 | pmichaud++ | * Based on comments from TimToady++
06:20 ashizawa joined #perl6
06:21 penk joined #perl6
06:28 Auzon1 joined #perl6
06:35 cls_bsd joined #perl6
06:53 PZt joined #perl6
07:11 apeiron_ joined #perl6
07:40 barney joined #perl6
08:04 iblechbot joined #perl6
08:14 wknight8111 joined #perl6
08:26 dogmatrix joined #perl6
08:30 jferrero joined #perl6
08:38 moritz_ rakudo: say \3
08:38 p6eval rakudo 29147: OUTPUT[3␤]
08:38 moritz_ my @list = 1, 2, \(3, 4); say @list.perl
08:39 moritz_ rakudo: my @list = 1, 2, \(3, 4); say @list.perl
08:39 p6eval rakudo 29147: OUTPUT[[1, 2, {}]␤]
08:42 elmex joined #perl6
08:46 bacek_ pugs: my @array = <a b c>; my $ref   = \@array; say ~$ref; say +$ref;
08:46 p6eval pugs: OUTPUT[a b c␤3␤]
08:46 bacek_ rakudo: my @array = <a b c>; my $ref   = \@array; say ~$ref; say +$ref;
08:46 p6eval rakudo 29147: OUTPUT[a b c␤0␤]
08:46 bacek_ moritz_: this is one of tests that failing...
08:58 moritz_ rakudo: eval '$x'
08:58 p6eval rakudo 29147: OUTPUT[Scope not found for PAST::Var '$x'␤current instr.: 'parrot;PCT::HLLCompiler;panic' pc 156 (src/PCT/HLLCompiler.pir:103)␤]
08:58 dogmatrix left #perl6
08:58 moritz_ rakudo: eval '$x'; say 1
08:58 p6eval rakudo 29147: OUTPUT[Scope not found for PAST::Var '$x'␤current instr.: 'parrot;PCT::HLLCompiler;panic' pc 156 (src/PCT/HLLCompiler.pir:103)␤]
09:14 cognominal_ joined #perl6
09:22 p6eval joined #perl6
09:23 chris2 joined #perl6
09:23 penk joined #perl6
09:51 meppl joined #perl6
10:04 rakudo_svn r29148 | moritz++ | [rakudo] fix RT #56692 (eval doesn't catch compilation errors).
10:04 rakudo_svn r29148 | moritz++ | Patch curtesy of bacek++
10:26 meppl joined #perl6
10:26 pugs_svn r21264 | bacek++ | [spec] unfudge more tests for rakudo in S02-builtin_data_types/flattening.t
10:56 jan_ joined #perl6
11:04 pbuetow joined #perl6
11:10 pugs_svn r21265 | moritz++ | [spec] fudge array.t to not produce passing TODO tests. Silly but convenient.
11:45 ruoso joined #perl6
11:47 wknight8111 joined #perl6
11:52 penk left #perl6
11:57 icwiener joined #perl6
12:05 rakudo_svn r29150 | jonathan++ | [rakudo] Stub in signatures. This doesn't do anything useful yet, but does lay a lot of groundwork: Signature class is in place, :(...) is now parsed and results in a Signature object, all subs and methods get a Signature object attached (though it's not populated yet) and
12:05 rakudo_svn ..some comments explaining what is going on.
12:08 pugs_svn r21266 | ruoso++ | [smop] fixing pmurias++ code. Every object that uses smop_lowlevel should support the DESTROYALL call with the object as capture. When you ask for the invocant of a capture you need to release it later.
12:10 ruoso @tell pmurias, I've fixed the mem leak you were having... basically, you had two problems in your code, the first is that you were calling SMOP__NATIVE__capture_invocant and were not SMOP_RELEASE'ing the object later... The second issue is that every object that uses smop_lowlevel needs to support an additional way of calling DESTROYALL with the invocant *AS* the capture itself (this is how I finish the loop to bootstrap the system, but that's something I sho
12:10 lambdabot Consider it noted.
12:10 ruoso uld restrict to some cases...))
12:11 ruoso btw... pmurias++ (smop hacking :)
12:13 ruoso @tell pmurias, one additional note is that when some constant identifier is assumed to be used in the core objects, you can declare it in include/smop_identifiers.h and run identifiers2C.pl that will regenerate some C files that initialize that values...
12:13 lambdabot Consider it noted.
12:15 iblechbot joined #perl6
12:27 barney joined #perl6
12:46 stevan_ joined #perl6
12:54 donaldh joined #perl6
12:57 cmarcelo joined #perl6
12:58 pmurias joined #perl6
13:02 ruoso pmurias, good morning...
13:03 pmurias ruoso: good afternoon
13:03 ruoso 10am here ;)
13:03 ruoso pmurias, have you seen the messages?
13:03 ruoso or the bot suicided?
13:03 ruoso ;)
13:04 pmurias i read the irc logs
13:04 ruoso ah... cool...
13:05 ruoso btw... I'm right now working on removing the need for the non-capture DESTROYALL call for other objects
13:05 pmurias thanks for the fix, i don't think i would have figured it easily
13:06 ruoso yeah... good old refcounting gc ;)
13:06 ruoso you always forget it somewhere ;)
13:10 pmurias i think i'll implement basic strings, when i sort-of-finish my ajaxy sms sender
13:16 drbean joined #perl6
13:16 ruoso pmurias, ok... you will only need to figure out how the API looks like ;)
13:17 ruoso which I don't really know
13:17 pmurias ruoso: do you know how to convert a string to a utf8 buf?
13:17 ruoso libiconv takes care of that...
13:20 pmurias in perl6
13:22 ruoso none of the autoboxings to native types are specified
13:22 ruoso I think TimToady is waiting to see what we come with...
13:23 pmurias you mean unboxings?
13:23 pmurias pugs: 1.int
13:23 p6eval pugs: RESULT[1]
13:24 pmurias pugs: say1.int32
13:24 p6eval pugs: OUTPUT[*** No such subroutine: "&say1"␤    at /tmp/qpRckNzb8q line 1, column 1 - line 2, column 1␤]
13:24 pmurias pugs: say.int32
13:24 p6eval pugs: OUTPUT[␤*** No such method in class Bool: "&int32"␤    at /tmp/8PE5U5Ppeg line 1, column 1 - line 2, column 1␤]
13:24 pmurias ruoso: foo.native_type_name is the smop convention?
13:25 moritz_ pmurias: $obj.OtherType coercions are in the specs
13:25 moritz_ pmurias: although it doesn
13:25 moritz_ 't specify that all conversions have to be available in this way
13:25 ruoso if that applies to native types also
13:25 ruoso that's just fine
13:26 ruoso although I still miss the bytebuf native type in the spec
13:27 pmurias buf is spec
13:27 pmurias ed
13:27 ruoso buf is a buf of ints
13:27 ruoso not a buf of bytes
13:27 ruoso that's a big difference for what you mean...
13:29 * pmurias is confused by your int/byte destinction
13:31 pmurias $foo.buf8 is speced *not* to return utf8
13:31 pmurias ruoso: what do you mean by a bytebuf?
13:32 ruoso a byte has 8 bits
13:32 ruoso a int has 32 or 64 bits
13:32 ruoso a bytebuf can be used as a parameter to strcmp
13:32 ruoso or strncmp
13:32 ruoso if buf is an array of ints
13:33 ruoso you won't be able to use it as a regular char*
13:33 ruoso which is what iconv deals with (iirc)
13:33 pmurias a buf8 is an array of bytes
13:34 ruoso oh... ok
13:34 ruoso I missed that
13:34 jhorwitz joined #perl6
13:36 ruoso pmurias, one other thing you need to be carefull is how this series of casts are done
13:36 ruoso I mean...
13:36 ruoso you need some kind of MMD to make it work
13:36 ruoso (for now you can check if it is one of your known responder interfaces
13:36 nothingmuch joined #perl6
13:36 ruoso but you probably want to cast them to Str before
13:39 pmurias ruoso: i'll check for my string class responder interface
13:39 rdice joined #perl6
13:40 ruoso cool... special casing is the ordinary way of bootstrapping the type system in smop
13:40 pmurias as i don't know what the Str->(utf8 buf8) conversion is
13:40 ruoso that's the point
13:40 ruoso if you receive something different than your known object
13:40 ruoso you need a sm0p block
13:41 ruoso that will call .Str
13:41 ruoso and then
13:41 ruoso .buf8
13:41 pmurias no
13:41 pmurias .buf8 dosn't work for utf
13:41 ruoso won't it leave the wide chars as two bytes?
13:42 ruoso (or as many as needed, actually)
13:43 ruoso pugs: "á".buf8.length
13:43 p6eval pugs: OUTPUT[*** No such method in class Str: "&buf8"␤    at /tmp/RF6XQ2Ev7G line 1, column 1 - line 2, column 1␤]
13:43 ruoso hm
13:43 ruoso rakudo: "á".buf8.length
13:43 p6eval rakudo 29151: OUTPUT[Method 'buf8' not found for invocant of class 'Perl6Str'␤current instr.: '_block11' pc 21 (EVAL_11:13)␤]
13:43 pmurias S02:851
13:44 ruoso kp6: "á".buf8.length
13:44 p6eval kp6 21266: OUTPUT[syntax error at position 9, line 1 column 9:␤"á".buf8.lengt␤         ^ HERE␤]
13:44 ruoso kp6: ("á".buf8()).length()
13:44 p6eval kp6 21266: OUTPUT[no method 'buf8' in Class 'Str'␤ at compiled/perl5-kp6-mp6/lib/Ki​ndaPerl6/Runtime/Perl5/MOP.pm line 345␤        ]
13:44 ruoso hmpf
13:44 pmurias ruoso: seen the spec link?
13:45 ruoso just read it
13:45 pmurias ruoso: why are node positions to Capturize relative
13:46 ruoso TimToady, which is the api to convert a Str into a raw array of bytes?
13:46 ruoso pmurias, because at first the frame would only point to the current node
13:46 ruoso but I changed it later to have the array and a pc
13:47 pmurias ruoso: is it ok if i change it to absolute positions (once labels are in place)
13:47 pmurias ?
13:48 ruoso I can't think of a reason against it
13:48 ruoso feel free to do it... just update all the sm0p blocks and the wiki
13:49 ruoso TimToady, more importantly... which is the api to convert a Str into a *native* array of bytes?
13:49 pmurias is smop_s1p_str the right place for my simple str implementation?
13:50 ruoso yes... that's ok
13:50 pmurias TimToady, and what is the right api to convert into utf8
13:50 moritz_ pmurias: S29, Str.bytes
13:51 moritz_ oh no, that's only the size
13:52 ruoso which is kinda weird
13:52 moritz_ I guess that Str.pack can be abused ;)
13:52 moritz_ well, I guess we'll have a Str.buf($normalization, $encoding) or some such
13:53 ruoso our buf8 multi pack( Str $template, *@items )
13:53 ruoso that seems to do it
14:01 ruoso it would be nice to have an additional signature
14:01 ruoso but actually...
14:01 ruoso why are all that "sub" instead of "method"?
14:02 ruoso our buf8 multi method pack()
14:02 ruoso a pack without template returns the raw stream of bytes
14:02 ruoso what do you think?
14:03 moritz_ in which encoding and which normalization? The default ones?
14:03 ruoso yes...
14:04 ruoso and the default is implementation-dependant
14:07 ruoso it's important to realise that this is probably only be used for low-level string comparision
14:08 moritz_ or for serialization
14:13 ruoso yeah...
14:13 ruoso serialization is expected to be implementation-dependent, right?
14:13 moritz_ depends on the kind of serialization, I guess
14:14 moritz_ I expect $anything.perl to give me valid Perl 6 that will work the same way on every implementation
14:18 ruoso yeah... but that will probably be a Str anyway
14:27 pugs_svn r21267 | ruoso++ | [smop] only a few types need to receive the non-capture DESTROYALL now... all the others only receive as captures...
14:28 ruoso pmurias, the io code can be simplified now
14:31 pmurias ruoso: ok
14:32 pugs_svn r21268 | ruoso++ | [smop] simplifying smop_s1p_io... no longer supports capture-less destroyall call...
14:32 ruoso pmurias, sorry... I went ahead and did it already ;)
14:33 pmurias no problem
14:40 pmurias had a brake from c and the segfaults are eating me... got to fetch valgrind
14:42 pmurias * break
14:48 icwiener_ joined #perl6
14:53 rakudo_svn r29152 | jonathan++ | [rakudo] Start to build signature objects. We don't put any types in yet.
15:02 rakudo_svn r29153 | pmichaud++ | [rakudo]: spectest-progress.csv update: 94 files, 1662 passing tests
15:02 rakudo_svn r29153 | pmichaud++ | * Also add a column with an estimate of the total number of spectests.
15:08 pugs_svn r21269 | pmurias++ | [smop] ok 1 is printed using io functions, i'll remove the memory leak after a walk or something
15:15 ruoso pmurias, I think you forgot to add a file
15:16 ruoso smop_s1p_str
15:18 pugs_svn r21270 | pmurias++ | [smop] removed memory leaks in str and io
15:18 pmurias ruoso: sorry
15:18 pugs_svn r21271 | pmurias++ | [smop] add missing file
15:19 ruoso noprob
15:21 ruoso pmurias, you probably want to receive a str**
15:21 ruoso char**
15:22 ruoso or better...
15:22 ruoso make a strncpy inside the create function
15:22 ruoso because you're making a free of something you didn't alloc
15:25 jfredett_ joined #perl6
15:25 ruoso pmurias, take a look at idconst.c, where I make something very similar
15:26 ruoso pmurias, specially SMOP__NATIVE__idconst_createn_nolist function
15:29 pmurias ruoso: you can change the api if you want
15:29 pmurias i'll take the much needed walk now
15:30 ruoso pmurias, it's not just about the API... it's about memory management... always try to make the allocs inside the objects... because they are the responsible for the free...
15:31 * ruoso lunch &
15:33 rakudo_svn r29154 | jonathan++ | [rakudo] STD.pm tracking: variable_decl becomes variable_declarator, add declarator and make us able to parse a signature in a declaration.
15:41 icwiener joined #perl6
15:45 meppel-san joined #perl6
16:05 pbuetow joined #perl6
16:09 justatheory joined #perl6
16:18 kanru joined #perl6
16:20 pugs_svn r21272 | lwall++ | [STD] start transition to .pmc forms so names are conserved under Perl 5
16:20 pugs_svn r21272 | lwall++ | [STD] also allow for Cat|Dog type constraints
16:21 pugs_svn r21273 | lwall++ | some renamings to go with the previous revision
16:23 pugs_svn r21274 | lwall++ | more name tweakage
16:27 pugs_svn r21275 | lwall++ | [STD5_dump_match] s/STD5/STD/ here too
16:32 pugs_svn r21276 | lwall++ | more renaming
16:32 pugs_svn r21277 | lwall++ | more renaming
16:37 pmichaud could someone take a look at S02-literals/autoref.t:51  ?   those tests look a bit odd to me.
16:37 pmichaud in particular, I'm wondering if \%hash is equivalent to \(%hash), and what that implies for the .values method
16:37 pmichaud bbl
16:59 sri_work_ joined #perl6
17:03 rakudo_svn r29158 | jonathan++ | [rakudo] Start making signatures in variable declarations work. While more work is needed to make this more general, my ($a, $b); will now declare both $a and $b.
17:04 moritz_ jonathan++ indeed
17:08 meppel-san joined #perl6
17:14 jferrero joined #perl6
17:17 pmurias joined #perl6
17:19 japhb Anyone know when this year's State of the Onion will be given?  A friend of mine will be doing OSCON for the first time and doesn't want to miss it, but says he's having trouble finding it on the OSCON website ....
17:21 bacek_ joined #perl6
17:21 pmurias ruoso: as i understand it smop's calling convention is that the called routine takes ownership of the arguments, but it might be the case of c's convention taking priority here
17:22 ruoso yes... that convention is valid in the SMOP__Object* level... not in the general malloc level
17:22 ruoso all memory allocation should be hidden inside some method call
17:22 ruoso or function call that returns low-level types
17:22 ruoso like the _create methods
17:29 pugs_svn r21278 | pmurias++ | [smop] fullfiled ruoso's request to move string allocation inside string creation method
17:29 pmurias done
17:29 ruoso pmurias++
17:34 cmarcelo joined #perl6
17:35 stevan_ joined #perl6
17:35 stevan_ left #perl6
17:35 japhb TimToady: Are you giving a State of the Onion speech this year?  If so, when and where?
17:49 pmichaud last year it was (I think) after the lightning talks, before the auction
17:53 pmichaud if that holds true again this year, it would be Thursday afternoon
17:54 japhb thanks, pmichaud
18:06 pugs_svn r21279 | ruoso++ | [smop] first sketches on the most complicated sm0p code... the code is the equivalent of: my $o = p6opaque.^!CREATE(); $o.^!methods().push($some_method); $o.^!how($some_metaclass); $o.foo(); # where $o.foo will eventually dispatch $some_method
18:07 ruoso pmurias, take a look at smop/test/14
18:07 ruoso there are two cool sm0p blocks there ;)
18:08 ruoso maybe it's time to think about a q:s1p { } macro language....
18:09 stevan_ joined #perl6
18:09 ruoso sm0p is starting to be boring because of the excessive "copy", "move_capturize" and "move_responder"
18:10 ruoso q:s1p { } could just expand some usual constructs like... my $a = something(); my $b = other(); $a.foo($b);
18:10 ruoso into the equivalent sm0p nodes
18:12 [particle] joined #perl6
18:14 pmurias food&
18:14 pmurias ruoso: i'm extending sm0p now
18:18 ruoso pmurias, if you're able to make it a little more smart (like handling variables...) that would be absolutely fantastic...
18:19 ruoso maybe change the syntax for the external variables
18:19 ruoso forcing $*foo when using a variable defined outside the block...
18:20 ruoso and using the sm0p frame operations for normal variables
18:25 pmurias ruoso: what should normal variables compile to?
18:26 dduncan joined #perl6
18:30 rakudo_svn r29164 | jonathan++ | [rakudo] Start to refactor declare_attribute a bit so it mostly doesn't dive into the parse tree, but instead takes parameters providing the information it needs. This is a step on making has ($.a, $.b) style things work.
18:36 ruoso pmurias, to the "copy", "move_responder" and "move_capturize" patterns
18:42 pmurias hmm
18:43 ruoso pmurias, take a look at the sketch I've just written
18:43 ruoso smop/test/14...
18:43 pmurias i hesitate to extend sm0p too much so that we don't end up with some pir-style monstrocity
18:44 pmurias * monstrosity
18:44 ruoso there are still two big limits in sm0p
18:44 ruoso one is that it can only represent a single block
18:44 ruoso no nested blocks...
18:44 pmurias easy one
18:44 pmurias and
18:44 ruoso and the second is that it doesn't have namespaces at any level
18:44 pmurias ?
18:45 ruoso it has SMOP__ID__* that can be guessed
18:45 ruoso variables $*outside the block
18:45 ruoso and $variables that store the result of a given node
18:45 ruoso just that
18:45 ruoso nothing more...
18:46 pmurias $* represents globals in perl6
18:47 pmurias my plan is that we extend sm0p enough to not require mixing it with c, and then start work on s1p
18:48 pmurias sounds resonable?
18:51 ruoso sm0p needs mixing with c
18:51 ruoso because of the "interpreter"
18:51 ruoso and off-the-block vars
18:51 ruoso for CPS, for instance
18:52 ruoso s1p will be something that will generate sm0p+c
18:52 ruoso (or C directly... generating the equivalent C code for the sm0p blocks that would be needed)
18:55 dduncan left #perl6
18:55 pmurias off-the-block vars?
18:56 pugs_svn r21280 | pmurias++ | [smop] changed $frame = q:sm0p {...} to fram = q:sm0p {...}
18:57 pmurias that can be fixed by having a fixed wrapper
18:58 ruoso fixed wrapper?
18:59 pmurias ...boiler plate C...;frame = q:sm0p {...};...boiler plate C...
19:03 ruoso pmurias, not all sm0p code will behave the same...
19:03 ruoso some will implement tail call
19:03 ruoso some will recurse
19:03 ruoso some will set a return value
19:03 ruoso some won't
19:04 ruoso in some cases, there won't be a goto call immediatly... but rather a set in some object for a later evaluation
19:06 pmurias if there are things that need mixing with c will do it :(
19:10 ruoso :)
19:10 ruoso pmurias, remember that we have to dislike sm0p enough to write s1p ;)
19:10 ruoso if we like sm0p too much, we'll never have s1p
19:11 pmurias i keep it in mind
19:12 pmurias but i don't like mixing in c because that makes us dependent on the c compiler
19:13 pmurias TimToady: gimme5 bug of the day [\s||#]*
19:15 pmurias TimToady: sorry, was matching my rule on a wrong string
19:17 ruoso pmurias, that's why we're pushing to s1p ASAP :)
19:18 ruoso the test/14 is already the implementation of a very important p6opaque REPR RI part
19:18 ruoso which will actually allow some translation already
19:19 ruoso we will then just need a metaclass...
19:21 pmurias the thing i'll need to be able to work on s1p is a list of short perl6 snipppets demonstrating basic features and what they compile to
19:21 ruoso basically test/14 is starting to point how that will work
19:21 ruoso it will push the methods into some p6opaque
19:22 ruoso that will have a metaclass that knows how to dispatch them
19:23 ruoso pmurias, but in fact... for that to work I now need Str
19:24 ruoso and more importantly... 'foo' eq 'bar'
19:24 ruoso or maybe not
19:24 ruoso in fact not...
19:24 ruoso because the REPR have a .^!can mehtod
19:24 ruoso and I can still special case the constant identifiers
19:25 pmurias ruoso: a initial str implementation is in
19:27 ruoso pmurias, I saw it...
19:27 ruoso pmurias, but what we need now is a dispatch code
19:28 ruoso considering I finish the REPR API, of course...
19:29 rakudo_svn r29167 | jonathan++ | [rakudo] Make has ($x, $y) work. It doesn't work with twigils just yet, but that should be a grammar change. Also remove some incorrect cuddles.
19:43 pugs_svn r21281 | pmurias++ | [sm0p] unidentified indentifiers are no longer treated as outside c vars
19:43 pmurias you mean method dispatch code
19:43 pmurias ?
19:44 pmurias sm0p now also supports comments, i forgot to mention it in the commit log
19:46 ruoso yes
19:46 ruoso cool
19:56 nothingmuch joined #perl6
19:57 cathyal joined #perl6
20:22 rakudo_svn r29173 | jonathan++ | [rakudo] Allow has ($.x, $y, $!z) to work.
20:24 xuser_ joined #perl6
20:28 pugs_svn r21282 | pmurias++ | [smop] unpredeclared identifiers work ie $foo.undecleared_anywhere
20:29 ruoso pmurias, remember to add your changes in the wiki Changelog.... ;)
20:30 pmurias ruoso: should i start a skeleton s1p implementation or are the more usefull things atm?
20:30 ruoso s1p is very much usefull
20:30 ruoso the other thing to do would be p6opaque...
20:30 ruoso but I'm already on top of that
20:31 ruoso pmurias, do you have any clue on how it will be already?
20:32 pmurias yes
20:32 ruoso cool... the example in test/14 was helpfull?
20:33 lambdabot joined #perl6
20:34 pmurias i skimed them
20:35 pmurias re any clue you meant s1p?
20:37 pmurias i plan to turn elf's ast into sm0p, likely with a simplification step between so different smop backends can be added later on
20:37 ruoso yeah... I was talking about s1p
20:39 ruoso pmurias, I need to leave now... I'll backlog later...
20:39 pmurias ok
20:44 xuser_ left #perl6
20:46 jferrero joined #perl6
21:28 armagad joined #perl6
21:34 Schwern joined #perl6
21:41 armagad joined #perl6
21:42 armagad left #perl6
21:48 gbacon joined #perl6
22:08 Schwern joined #perl6
22:10 rakudo_svn r29175 | jonathan++ | [rakudo] Stop producing blocks when we only want a signature object. Additionally, start storing the constraints in the signature object.
22:11 pugs_svn r21283 | jnthn++ | [rakudo] Some extra tests for attributes, where you declare many at once with a signature (has ($x, $y) style things).
22:18 Squern joined #perl6
22:28 Squern joined #perl6
22:38 edpratomo joined #perl6
22:38 Squern joined #perl6
22:57 japhb joined #perl6
23:20 gbacon joined #perl6
23:34 meppl good night
23:49 bacek joined #perl6

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

Perl 6 | Reference Documentation | Rakudo