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.
00:09 ashleyb joined 00:17 Ziggy6 left 00:50 lisppaste3 joined 00:53 putter joined, nipotaway is now known as nipotan 01:01 drupek12167 joined
putter Hmm. en.wikipedia.org/wiki/Pugs says "Due to licensing and concerns about the lack of copyright assignments from contributors, it is unlikely that The Perl Foundation will bless Pugs as the official implementation of Perl 6." 01:08
It was added in early Nov by anon. 01:09
Does this strike anyone else as bogus? Ie, if they meant the haskell implementation, it seems to underestimate how much of the codebase will be churned before xmas, and overestimate the number of people who have made substantial contributions. Either of which would mitigate the concern expressed. 01:12
TimToady that's funny, considering S01 specifies that we not consider any implementation "the official Perl 6 implementation". An implementation can be "an official implementation" if it passes the test suite. 01:45
dduncan TimToady, regarding r13553, I suggest you reformat that cascaded conditional example so that the !! is on the leading end of lines rather than the trailing end 01:47
and they line up vertically with the assignment
I refer to the example under: +A tabular form is also popular for cascaded conditionals: 01:49
TimToady gonna sic Perl::Critic on me? :)
dduncan I noticed it myself, without the help of a program
and I agree with the rationale of the change
TimToady I don't always follow PBC myself... :)
dduncan but was your choice here arbitrary, or did you have the trailing !! on purpose? 01:50
01:50 drupek12167 joined, TimToady joined
dduncan but was your choice here arbitrary, or did you have the trailing !! on purpose? 01:50
TimToady I just threw the pieces of gold into the fire, and lo! out came this Golden Calf... 01:51
putter re "official"/"that's funny": correcting... 01:52
TimToady I probably shouldn't be putting stylistic notes into the S's in any event. 01:53
dduncan but the examples could still be consistent with PBP if you don't oppose them 01:54
leading by example isn't leading by doctrine
SamB TimToady: that is the lamest excuse evar! 01:56
it was lame the first time, and it is really really lame this time! 01:57
putter Hmm, S01 was probably not written in the month of "Nob". No, I cant say that. It does seem likely to be a typo however. 01:59
TimToady SamB: so break the stone tablets again. :) 02:01
SamB do I look like my face glows? 02:02
no!
SamB hasn't got any stone tablets
anyway, if I had one, I'd want to run programs on it
allbery_b thinks decent laptops are heavy enough as is, thanks :) 02:03
02:09 Schwern joined
putter "flintstones, meet the..." 02:20
in p6 rules, can <{ }> interpolated regexs be modifiers? eg /<{/:i/}> GoOd ThInG ThAt InTeRpOlAtEd/ ? 02:37
TimToady sure, but they're scoped to the subrule, which doesn't do you much good. 02:38
putter ah, ok
TimToady making pretty good progress on the Official Grammar, btw 02:39
putter :)
TimToady to keep "actions" from cluttering things, I'm writing the rules such that they can be preprocessed
putter good choice :) 02:40
TimToady so we have rules like:
when => rule { <block> {*} }, #ACT sc when
the {*} is a placeholder where you might want an action 02:41
and #ACT sc when is the identifier of that action
in this case "statement_control when"
putter ah. I can live with that. hopefully nothing with have two actions? 02:42
TimToady the hope is that the grammar can be used as-is for mere syntax validation, but preprocessed into various derivative grammars for other purposes
putter yes
TimToady nothing says the preprocess couldn't s/<'{*}'>/{...} {...}/ 02:43
looked at various ways of sneaking the identifier into the rule, but it really cluttered things up.
putter <action 42> ? 02:44
TimToady prefer symbolic
02:44 dmq joined
putter <action-42> .... rule action-42 { <null> } ? 02:45
TimToady also it doesn't look like a longest-token stopper
and {*} looks like a closure because it is one
just doesn't do much.
so you really only have to translate the ones that you want to add an action to. 02:46
putter hmm, do we have a way to "inline" a subrule? eg, gram Acts does Base { rule when { <inline Base::when> {random cruft} } 02:48
TimToady well, there's always eval... 02:49
putter I suppose inline'ing is merely bonus. could just rule when { <Base::when> {somewhat different code} }
pasteling "TimToady" at 71.139.17.198 pasted "Top of the grammar to ya!" (181 lines, 6.4K) at sial.org/pbot/22748
putter oooo, looking... 02:50
TimToady there's probably some form of textual macroing, but it could *really* screw up the visual semantics of which field is $3
now all I need is define all my terms. 02:51
putter chuckle 02:53
TimToady oh, the ' ' key in a magic hash tells it what to check for between the key and the value.
putter hmm, one thing I've never understood about token... actually a couple of things.... 02:54
TimToady so it's easy to force word boundaries there, for instance
putter is the non-backtracking here needed for a correct parse, or merely efficiency?
TimToady not sure. there may be other things that token implies that regex wouldn't. 02:56
for a while was considering that token/rule implied longest token on | while regex didn't, but decided consistency was better
putter so is there any reason all those tokens could be rules instead? 02:57
TimToady but mostly it's "token" to keep reminding you that it doesn't do whitespace like "rule".
putter ah 02:58
TimToady tokens don't do :sigspace
though "token { \{ <expression> \} }" automatically gets whitespace inside because <expression> is a rule 02:59
presumably with auto-ws fore and aft
02:59 justatheory joined
putter so when users locally modify the p6 grammar, they have a fighting chance against backtracking rules, but if they run afoul of a token, they lose? 02:59
TimToady we try to push as many tokens into hashes as possible to keep them officially independent. 03:00
and deletable as well as addable
03:02 Ziggy6 joined
TimToady dinner, so entering attention deficit mode. & 03:02
putter in redsix each grammar just collected all the locally visible rules named token_foo and put them in the token hash. what's the real way one does gram B does A { ...let me fiddle with the token set... }
ah, bye. cheers. 03:03
03:09 ayrnieu joined 03:46 Ziggy6 left 03:59 bcorn joined 04:13 mako132_ joined 04:21 finchely joined 04:28 ProperNoun joined
svnbot6 r15162 | putter++ | misc/pX/Common/yet_another_rules_engine - progress. re_tests.t is 60% okay. See README. 04:54
04:55 Lunchy_ joined 04:59 leed joined 05:00 amnesiac joined 05:03 Lunchy__ joined 06:13 BooK joined
TimToady @tell putter backtracking vs not doesn't have anything much to do with mutability, nor does token vs rule. rules don't backtrack by default either. 06:20
lambdabot Consider it noted.
TimToady @tell putter we're not changing the grammar out from under ourselves in any case. we're only changing the grammar to parse the rest of the block. 06:21
lambdabot Consider it noted.
TimToady @tell putter it's sort of a continuation idea. But yes, how one actually does the overrides is the real question. The problem with the hash approach is it tempts us to reinvent inheritance badly. 06:22
lambdabot Consider it noted.
TimToady @tell putter and the old <%foo|%bar> notation is really a form of MI. even the new /%foo|%bar/ is still kinda MIish. anyway, will think on this derivation mechanism s'more. 06:25
lambdabot Consider it noted.
06:52 bsb left
tene ?eval say "foo" 07:05
07:06 evalbot_r15161 is now known as evalbot_r15162
evalbot_r15162 OUTPUT[fooā¤] Bool::True 07:06
07:29 Lorn_ joined, idiotmax_ joined 07:39 drrho joined 07:44 ofer0 joined 07:45 Southen joined 07:47 marmic joined 07:56 rgs joined 08:00 Yappo joined 08:13 devogon joined 08:16 iblechbot joined 08:56 kanru joined 08:59 andara joined 09:02 elmex joined 09:10 nipra joined 09:18 |Lupin| joined, rfordinal_ joined 09:27 rfordinal_ joined 09:29 andara left 09:55 SCalimlim joined 09:56 SCalimlim joined 10:10 ruoso joined 10:18 howl joined 10:20 howl is now known as cornelius 10:21 drupek12167 joined 10:29 buetow joined, shayx joined 11:07 shayx_ joined 11:13 jferrero joined 11:25 dduncan left 11:45 iblechbot joined, nipotan is now known as nipotaway 11:54 drupek12167 joined 12:03 chris2 joined 12:13 pbuetow joined 12:23 TimToady joined 12:47 nipra joined 12:51 Cornelius joined 12:58 Cornelius is now known as z0889 13:04 sunnavy joined 13:08 kisu joined 13:19 z0889 is now known as z090 13:20 Limbic_Region joined 13:57 shayx joined 14:02 seano joined 14:03 VanilleBert joined 14:23 [particle] joined 14:29 ayrnieu joined 14:32 iblechbot joined 14:36 jferrero joined 14:43 vel joined 14:55 bonesss joined
|Lupin| 4/part 15:13
oops, sorry
15:13 |Lupin| left 15:18 ProperNoun joined 15:22 nipra joined 15:34 hexmode joined 15:45 amnesiac joined 15:54 VanilleBert left 16:03 justatheory joined 16:13 ProperNoun joined 16:29 thepler joined 16:53 nipra joined 17:03 elmex joined 17:20 kanru joined 17:22 ofer1 joined
beppu @babel en es perl is funny 17:44
lambdabot el Perl es divertido
allbery_b just off the top of my head, that looks slightly odd :) 17:45
beppu @babel en es pythonistas are not to be underestimated. they are devious and sly. 17:47
lambdabot los pythonistas no deben ser subestimados. son desviados y sly.
TimToady @babel jp en 恙恔恄 17:49
lambdabot Plugin `babel' failed with: IRCRaised Error: Language jp not supported
TimToady 恔ļ¼ 17:50
xinming éƒé—·ć€‚ć€‚ć€‚ 17:51
ē”觉äŗ†ļ¼Œ å¤§å®¶ę™šå®‰ć€‚ :-D
TimToady ćŠä¼‘ćæćŖ恕恄怂 17:54
17:56 xinming_ joined
Juerd I see funny characters. 17:58
18:06 Teratogen joined
Teratogen will Perl 6 have winged comments? 18:06
[particle] yes
code; # comment 18:07
TimToady never heard them called that...
[particle] it's a c thing... /* comment */
Teratogen winged comments are like /* comment */
TimToady end-of-line comments is what I call them
Teratogen err, what [particle] said
TimToady do you mean #{...} comments then? 18:08
Teratogen probably
18:08 nipra joined
Teratogen code /* comment */ code 18:08
TimToady code #[[[ comment ]]] code
where any set of brackets works
Teratogen ah
nice
TimToady is disallowed if # is in column 1 though. 18:09
that's always line-end comment even if followed by bracket
because people do 'a,'zs/^/#/ and expect it to work 18:10
also, we encourage people to use pod for block comments
18:15 BooK_ joined 18:16 theorbtwo joined 18:17 nipra joined
Juerd #(((((((((( hug comment )))))))))) 18:20
Hm, no 18:21
#(((((((((( hug comment ))))))))))
:)
18:33 gene9_ joined, nipra joined 18:35 CardinalNumber joined 18:47 theorbtwo joined 18:53 theorbtwo joined 18:54 bonesss joined 18:56 hexmode` joined 19:12 hexmode`` joined, ezraburgh joined 19:15 GabrielVieira joined 19:30 justatheory joined 19:43 theorb joined 19:49 drupek12167 joined 19:50 bernhard joined 20:00 REPLeffect joined
masak perl6 cleans up the following things: regexes, OO and functional declarations 20:07
with "cleans up", I mean something like "wholly replaces old with new" 20:08
those three areas alone are reason enough to upgrade 20:09
20:09 ProperNoun joined
masak but then perl6 makes a whole lot of other improvements too 20:09
20:12 nipra joined
PerlJam masak: you make a good argument for upgrading but it has one important flaw ... no perl6 compiler. :) 20:12
20:19 elmex joined 20:23 hexmode`` joined
TreyHarr1s i thought there was a perl 5 backport of given/when, but I can't find it on cpan. am i hallucinating? if not, what's the model name? 20:40
s/model/module/ 20:41
20:41 hexmode`` joined
wolverian TreyHarr1s, 5.10 will have it; right now you can use Switch.pm 20:41
20:42 Daveman joined
wolverian (note that Switch is a source filter) 20:42
TreyHarr1s wolverian: ahh, i missed the 'Perl6' import when i looked at Switch.pm a moment ago.
wolverian :)
20:43 VanilleBert joined 20:44 rashakil__ joined 20:48 VanilleBert left 21:03 miyagawa joined 21:28 buetow joined
masak PerlJam: true. But pugs is enough of a taste of one for me to feel excited. I want to show its goodness to people 21:28
realistically, how far away is a real p6 compiler written in p6? 21:29
TimToady I'm working on the official grammar right now.
masak sounds good
TimToady but a compiler is a lot more than just a grammar... 21:30
SamB yes, but a grammar is a good starting point for a compiler
TimToady there are lots of good starting points, and we've started at most of 'em already. :) 21:31
SamB heh
has everyone had a chance to start at each of them yet?
TimToady a chance? sure. 21:32
SamB is referring to the saying "everything has been said, but not everyone has had a chance to say it yet"
TimToady you can say that again 21:33
that goes without saying
SamB why would I want to say it again? not everyone has had a chance to say it yet!
TimToady "Say it again, Sam!"
[particle] has joined #puns 21:34
21:34 hexmode joined
TimToady has joined #pun6 21:35
21:37 mdiep joined 22:01 mdiep joined 22:07 nipra joined 22:13 CardinalNumber joined 22:14 luqui joined, CardinalNumber joined 22:31 dduncan joined 22:36 sahadev left 22:38 justatheory joined 22:44 REPLeffect_ joined 22:46 weinig|away is now known as weinig, fridim joined
fridim hi. 22:46
22:55 polettix joined 23:06 Aankhen`` joined, Psyche^ joined 23:08 weinig is now known as weinig|away 23:15 Psyche^ is now known as Patterner 23:27 ProperNoun joined 23:35 Odin- joined 23:39 bonesss is now known as bonesss`away 23:50 davidfetter joined
davidfetter Lorn_, oi 23:50
23:52 CardinalNumber joined

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!