pugs.blogs.com/dist/Perl6-Pugs-6.2.13.tar.gz released! | run.pugscode.org | spec.pugscode.org | paste: sial.org/pbot/perl6 | pugs.blogs.com
Set by audreyt on 17 October 2006.
00:02 dduncan joined, Odin- joined, beppu joined, penk joined, nothingmuch joined, jiing joined, Patterner joined, nelhage joined, stevan_ joined, araujo joined, Narcisse joined, meppl joined, takesako joined, autark joined, osfameron joined, masak joined, Shabble joined, spinclad joined 00:10 justatheory joined
TimToady @tell aufrank 'my $foo is context;' 00:20
lambdabot Consider it noted.
00:29 mako132_ joined 00:42 justatheory joined 00:54 vel joined 00:58 mauke_ joined 00:59 frankg joined, markstos joined 01:03 mauke_ is now known as mauke
markstos I just made the mistake of using 'given' when I meant 'when'. It would be nice if there was a warning emitted if given {}is used, without doing anything useful with it. 01:04
?eval for <a b c> -> $rm { given Str { say "$rm is a Str" } default { say "$rm was not a Str, it was: "~$rm.WHAT } }
evalbot_r14504 OUTPUT[a is a Str␤a was not a Str, it was: Str␤b is a Str␤b was not a Str, it was: Str␤c is a Str␤c was not a Str, it was: Str␤] undef
01:08 bluejay joined 01:09 bluejay left 01:14 weinig|food is now known as weinig
revdiablo ?eval given Str { say $_ } # How can the usefulness be determined by a compiler? 01:32
evalbot_r14504 OUTPUT[Str␤] Bool::True
markstos I'm not sure if it's even possible or practical. I'm just reporting a "wish" from a user's perspective. 01:33
01:33 nipotaway is now known as nipotan
eviltwin_b it might be possible to warn if the block has neither whenclauses nor uses of $_, I suppose 01:34
(more precisely, references $_ neither explicitly nor implicitly)
markstos eviltwin_b: That's the kind of thing I had in mind. 01:35
01:50 Aankh|Clone joined 02:05 lambdabot joined 02:07 MacVince joined
markstos I found a Perl 5 embedding case we might want to address. If I have a Perl 5 string and pass it to a Perl 6 routine that checks to see if the arg is a string, it won't work. 02:20
the type appears to be "{obj-perl5}" instead. 02:21
02:29 mac93 joined
markstos Nevermind. There isn't a bug there like I thought. 02:30
bcorn best book for getting started with haskell? 02:58
araujo Introduction to Functional Programming - Bird 03:01
bcorn araujo: thanks 03:07
03:13 nicodaemos joined 03:20 weinig is now known as weinig|zZz 03:22 zMott joined 03:54 Yappo is now known as a 03:55 a is now known as Yappo, zMott left 04:03 nicodaemos joined 04:27 scw joined 04:30 SubStack joined 04:51 elmex joined 05:00 root4o joined, mjk joined 05:09 root4o left 05:15 BooK joined 05:29 DHGE joined 05:36 mdiep joined 05:43 chaoslawful joined 05:47 marmic joined 05:50 mdiep joined 05:55 araujo joined 06:09 seano joined 06:13 nipotan is now known as nipotaway 06:14 iblechbot joined 06:24 ozan joined 06:25 ozan is now known as ozayam 06:27 nipotaway is now known as nipotan 06:32 kanru joined 06:37 Pomin joined 07:26 ditto joined 07:58 nipotan is now known as nipotaway 08:02 qmole joined 08:15 nipotaway is now known as nipotan 08:49 kensanata joined 08:58 ruoso joined 09:07 bcorn joined 09:22 andara joined 09:44 Avochelm joined 09:46 iblechbot joined 09:55 dduncan left 09:57 chris2 joined 10:04 buetow joined 10:18 Pominx joined 10:38 kanru joined 10:53 andara left, andara joined 11:10 fglock joined 11:14 xinming joined 11:21 norageek joined 11:31 Psyche^ joined 11:37 vytautas joined 11:47 weinig joined, Psyche^ is now known as Patterner 12:12 iblechbot joined
svnbot6 r14505 | fglock++ | v6/emitter(2) - updated HACKING 12:12
12:17 Limbic_Region joined 12:24 nipotan is now known as nipotaway
svnbot6 r14506 | fglock++ | v6/emitter(2) - updated HACKING 12:36
fglock implementing a fully optimized v6/emitter(2) is probably not the way to go - this will take too long 12:40
Limbic_Region yes fglock, and the rest of the p6 project is moving so fast - you don't want to be left behind 12:44
</tounge_in_cheek>
fglock Limbic_Region: :) 12:45
Limbic_Region: I have to use my human resources sparingly
Limbic_Region that I understand - there isn't enough hours in the day to learn all that I want, spend time with my family, earn a living at $work, and be productive too 12:48
I have come to the conclusion however that my hacking put in perspective of the mathematical world is definately "pure" and not "applied" 12:49
12:49 nicodaemos_ joined
fglock Limbic_Region: an API layer can fix that 12:50
Limbic_Region fglock - sure, but that's someone else's problem 12:55
fglock Limbic_Region: right :) 12:56
Limbic_Region fglock - are technical books expensive where you are? 12:57
fglock Limbic_Region: reasonably expensive, yes 12:58
Limbic_Region and what are the custom tax on gifts from US? 12:59
fglock Limbic_Region: there is no taxes for books, afaik
Limbic_Region and you are in .br right? 13:00
fglock Limbic_Region: yes
Limbic_Region hrmm - I was completely unaware of this problem but you are not alone
fglock brb - meeting & 13:01
Limbic_Region just the other day I was chatting with a fellow from argentina who indicated the cost prohibitive nature of books their
13:13 weinig is now known as weinig|away 13:21 [particle] joined 13:24 thepler joined 13:42 c6rbon_ joined 13:49 vel joined 13:51 chris2 joined
nothingmuch @seen cmarcelo 13:51
lambdabot Last time I saw cmarcelo was when I left #darcs, #gentoo-haskell, #ghc, #haskell, #haskell-blah, #haskell-overflow, #haskell.es, #haskell.it, #haskell.se, #haskell_ru, #oasis, #perl6 and #scannedinavi
an 5d 11h 56m 38s ago, and .
13:55 rlb3 joined 14:00 c6rbon_ is now known as c6rbon 14:12 hexmode joined 14:22 aufrank joined 14:23 cjeris joined 14:30 iblechbot joined 14:49 nusgnaf joined 14:52 rlb3 joined 14:55 nusgnaf left
svnbot6 r14507 | audreyt++ | * Pugs.Parser: Parse for (but do nothing with) "of" types in return type: 15:04
r14507 | audreyt++ | sub f (Int $x --> Hash of Int) {...}
masak I get "user error (Incompatible version number for compilation unit" all the time
even after following the advice in "Consider removing /usr/lib/perl6/site_perl/Prelude.pm.yml and make it again"
audreyt rm blib6/lib/*.yml? 15:05
lambdabot audreyt: You have 1 new message. '/msg lambdabot @messages' to read it.
audreyt @messages
lambdabot TimToady said 5d 21h 54m 7s ago: MOH is just a rotated HOW. :)
audreyt weird, lambdabot has a confused memory
[particle] sure does
nothingmuch heh 15:06
svnbot6 r14508 | audreyt++ | * ext/MessageBus: New extension for a Pub/Sub message bus with 15:07
r14508 | audreyt++ | multiple channels and pluggable cache backends.
gaal audreyt: in your autopugs, please add the following after make 15:09
rm /tmp/refresh-prelude
chmod a+w blib6/lib/Prelude.yml 15:10
and add some moose to rebuild if /tmp/refresh-prelude exists :)
audreyt gaal: the autobox is in evalbot-loop 15:11
gaal that way no need to impersonate you
audreyt please add it yourself :)
autopugs even
(not being lazy; the hotspring net here is not exactly fast)
gaal okies :)
sure
be more lazy, it's good for you :) 15:12
masak hotspring net? :)
gaal uh, someone remind me, how do you set up a unix dir so that every file created in it is a+w? 15:13
or belongs to a certain g, also good
nothingmuch @tell cmarcelo i'd like to discuss the purity of the meta object in the haskell versions, please ping
lambdabot Consider it noted.
wilx Change umask?
nothingmuch gaal: my flight control friend has access to very up to date weather data
he'll give us an update tomorrow
though i'm skeptical =( 15:14
gaal nothingmuch: I saw those metaobjects debauching last night, I think it's high time cmarcelo and you take care of their purity
eviltwin_b certain group: give dir that group, flag as setgid
gaal if not a little late
eviltwin_b (doesn't work on some network filesystems) 15:15
gaal wilx: I want everybody who writes there to moose those perms, even if their umask is unfun
eviltwin_b: yeah, that's good. what's the incantation? chmod gs+w ? 15:16
eviltwin_b chmod g+s 15:17
gaal and then g+w?
eviltwin_b change the group first as some kernels will clear the setgid
yes
gaal great, thanks
eviltwin_b keep in mind that the group write will still depend on the umask 15:18
masak so, I want a something equivalent to an inheritable private method in a perl6 class... is there a way to do this?
eviltwin_b if you need finer control you need a kernel+fs that speaks posix acls
...or a network filesystem with its own permissions foo, but that has its own issues :)
gaal eviltwin_b: the only purpose here is to allow all ferather users to ask autopugs to regen its prelude when it's broken (happens occasionally) 15:19
even if they don't have sudo privs
but also allowing autopugs' user to remove the flag file
masak er, or even an inheritable method at all... is this implemented? I can't get pugs to find my base class method 15:21
masak studies the tests 15:22
ah, needed to prefix the method call with "self.". now works satisfactorily 15:27
audreyt or $. 15:28
$.meth(1,2,3)
15:30 bonsai joined
masak audreyt: yes. just found that out myself and changed the syntax to that :) 15:31
svnbot6 r14509 | audreyt++ | * MessageBus: Method name cleanup from obra++ 15:34
fglock re MO: is there a paper or reference about it? I need to find out how this interacts with v6 emitter 15:40
svnbot6 r14510 | gaal++ | * allow users to request an evalbot refresh by touching
r14510 | gaal++ | /tmp/autopugs/refresh
r14510 | gaal++ | (not enabled on feather until we set up everyone to
r14510 | gaal++ | be members of a common group, to prevent the file from
r14510 | gaal++ | accidentally not being removable by the evalbot-loop;
r14510 | gaal++ | that would trigger endless refreshes.)
gaal Juerd: could you please add a new group on feather, and make everyone (including new users) members of it? see 14510 15:41
masak ?eval class A { has @.x = <a b c>; method y( $i ) { return @.x[ $i ]; } } class B is A { method z( $i ) { return $.y( $i ); } } B.new.z( 1 )
evalbot_r14509 \undef
masak evalbot_r14509: I would have liked a 'b' from you
seems that when I try to access a base class' member variables through a method called from a derived class, I get nothing back 15:42
Juerd gaal: I don't understand what the purpose would be of a group that everyone is in. Unix has owner, group, and world. Doesn't world suffice?
masak now, I'm open for the possibility that I've misunderstood something, but barring that, I'd say this is an error or a missing feature 15:43
gaal Juerd: is there a way to ensure a file created in some dir is o+w? 15:44
Juerd gaal: A directory can be g+s, which automatically sets the group. This group can then be that of the evalbot's user.
gaal: I don't think /tmp is a good place for this, by the way. It's cleaned on bootup. 15:45
gaal: Also, aren't we better off with a command that sends a signal to the process?
audreyt masak: initializaer did not inherit properly; please writ a test or locate one
masak audreyt: oki 15:46
Juerd In any case, I find adding everyone to the same group a lot of maintenance for such a small feature.
There are so many ways to do this. You could even build a web interface, or let people request this from IRC.
gaal Juerd: not via evalbot, fwiw. this is for use when evalbot breaks :) 15:47
the signal approach makes the most sense
Juerd Why not refresh automatically on breakage?
Breakage tends to be detectable
lumi gaal: Why not via evalbot? The breakage is just pugs not working
fglock <lazy> how is MO different from Moose? </lazy> 15:48
gaal hmm.
15:49 BooK joined
gaal okay, I'm going away for ~3hr, will do that if somebody doesn't beat me to it 15:50
15:51 justatheory joined
svnbot6 r14511 | masak++ | [t/oo/inheritance.t] 15:51
r14511 | masak++ | * added test for inherited initializer
audreyt fglock: MO is compiled. 15:53
fglock audreyt: in the sense that it does not depend on an intermediate language? 15:56
like Moose is implemented in perl5 (the VM language) 15:57
does compiled mean it belongs to the VM level? 16:00
like Parrot should do 16:01
svnbot6 r14512 | andara++ | [runpugs]
r14512 | andara++ | -now runs latest release by default, latest devel as option
r14512 | andara++ | -runpugs2: devel version, now with non-interactive sessions (not live yet)
r14511 | masak++ | [t/oo/inheritance.t]
r14511 | masak++ | * added test for inherited initializer
16:01 GeJ joined
masak seems svnbot6 is confused today, too 16:05
I'm getting undeserved karma here :)
audreyt fglock: compiled means that it can take a set of declarations and emit native code in (say) perl5 or haskell code for them 16:08
leo » Joined channel "#parrot". 16:13
» Checking for new commits (ignore error messages)... *** Cannot cast from VHandle {handle: temp-svnbot-20777-532} to Object at /home/audreyt/pugs/examples/network/svnbot.pl line 152, column 13-22
svnbot troubles
hi btw 16:14
fglock leo: hi 16:20
audreyt: so it could even be written in Perl6 16:21
no, I don't think this is what you mean 16:22
you mean it can choose to either inline code or to make calls (?) 16:23
16:26 buetow joined 16:27 quix joined 16:28 quix left 16:35 cj joined
cj audreyt: whee! You're coming to talk to us on Friday! 16:36
16:38 DHGE joined 16:40 cmarcelo joined
cmarcelo nothingmuch: ping 16:40
16:42 the_dormant joined
svnbot6 r14513 | fglock++ | v6/emitter(2) - added multisub-example.pl 16:43
fglock cmarcelo: would you help me with MO? 16:44
the_dormant Hi all, I'm learning perl6 (w/ very little background on perl) and I'm looking for the equivalent of -C, -A on files 16:45
anyone help please
xinming the_dormant: perl 6 uses utf-8 by default 16:46
cmarcelo fglock: sure. but I know very little bit about it :o)
fglock cmarcelo: np - let's learn together 16:47
PerlJam the_dormant: perl6 still has those operators I believe.
fglock cmarcelo: the input data for MO comes from an AST transformation, right? 16:48
cmarcelo I think so. At least the tests I read suggest that.. 16:49
fglock cmarcelo: the output data is native source code?
xinming PerlJam: I think It is depend on the compiler. ;-) 16:50
PerlJam: If we use perl 5 as vm, then, We don't need to worry about this issue. 16:51
cmarcelo i don't know exactly. it builds a bunch of objects in perl5 (or hs in case of moh), that could be serializable. but there are method definitions which are in MO right now just refs to p5 code, but it could be ref to a more abstract representation of code (part of a p6 ast, for example), which can be serializable too.. 16:54
makes sense?
fglock cmarcelo: the output code makes no use of OO, since MO desugars it into basic operations? 16:56
cmarcelo basic operations == using MO vocabulary to call a method instead of doing $obj.meth() ? 16:57
fglock yes 16:58
cmarcelo in the code inside method definitions I'm not sure if this actually happens (looks like it should)...
fglock: what do you have in mind? 17:00
fglock cmarcelo: I'm writing a code generator for v6.pm 17:01
cmarcelo: gaal said you should take care of metaobjects purity - what is he referring to? 17:02
cmarcelo: re have in mind - I'm looking for alternate architectures 17:04
cmarcelo i'm (trying to) translate MO code into Haskell. i guess he said that regarding the debate: whether the objects in moh should be datatypes full of IORefs/TVars (mutable things) or should be just pure data types. this has consequences on how I should deal with attributes inside objects (how do the accessors) and how do inherithance too.. 17:06
fglock: I thought that much of this MO-thing would be (in a sense) part of pugs AST... this make sense? 17:09
fglock cmarcelo: it depends how much preprocessing you are doing during parsing 17:10
cmarcelo fglock: (i used tried to install v6 from CPAN and my machine freezed [froze?] doing lots and lots of IO during 01-tap.t test in v6-alpha) 17:11
s/used//
fglock in order to execute BEGIN blocks you have to go very far
cmarcelo: there are some problems regarding Cache::Cache module and access rights - if you make tests as non-root 17:12
but this should only make it 50% slower 17:13
cmarcelo i'll try again today. with non-root it should be slower but work fine? 17:14
fglock yes
and it takes about 80MB of free ram, otherwise it will page fault a lot 17:15
cmarcelo: how is code that uses MO faster than current code? 17:17
17:20 weinig|away is now known as weinig
cmarcelo fglock: don't know about performance, but the idea is that MO provides more "meta-candy" than what we have now.. 17:26
17:26 mdiep joined
cmarcelo fglock: nothingmuch and audreyt can enlighten you more about this... 17:36
fglock cmarcelo: thanks - this is starting to make sense
17:42 andara left 17:49 nicodaemos joined
leo any one here to fix the mentioned pugs and/or svnbot err? 17:51
18:01 weinig is now known as weinig|bbl 18:02 fglock left 18:12 [particle] joined, buubot joined 18:19 ozayam joined 18:40 dmq joined 18:41 stevan__ joined 18:59 the_dormant_ joined 19:03 the_dormant_ left, the_dormant_ joined 19:04 DHGE joined
the_dormant_ Hi all, anyone knows how to get the attributes of a file in pugs (eqv of -C, -A in perl 5) 19:05
kolibrie 12:47 < PerlJam> the_dormant: perl6 still has those operators I believe. 19:08
audreyt we don't currently implement -A -M etc 19:09
please add the ones you want to t/operators/filetest.t
and they'll get hacked in
you do have a commit bit?
if not, what's your email? :) 19:10
19:10 svnbot6 joined
audreyt leo: hm, that means you are running an outof date pugs 19:11
trunk pugs does not exhibit that error
(I just restarted svnbot6 to make sure)
the_dormant_ [email@hidden.address]
leo audreyt: I'm running it on feather, whatever that is :)
it dies after 6 mins, when querying again the rev 19:12
PerlJam feather is still running an old ghc isn't it? 19:13
integral PerlJam: it reports 6.6
leo Version: 6.2.13 (r14513) 19:14
PerlJam ah. For some reason I thought feather wasn't updated.
svnbot6 r14514 | audreyt++ | * svnbot.pl: Trying to fix bug without knowing what's going on... 19:20
audreyt leo: try again? 19:21
19:21 HiTech69 joined
leo started ... waiting 5 mins ... 19:21
TimToady the_dormant_: commitbit sent, please add yourself to AUTHORS as 1st commit 19:23
leo audreyt: seems to have survived a ?check request - lookin good 19:24
the_dormant_ TimToady: ok, thx 19:28
19:28 ruoso joined
leo audreyt: that seems to have fixed it - thanks a lot 19:31
audreyt np :) 19:33
leo svnbotl successfully did report r15014 - thanks 19:35
audreyt woot 19:36
cj audreyt: ! 19:49
audreyt: got lunch plans on friday?
audreyt cj: I think I'll do lunch in amazon no? 19:53
19:54 justatheory joined
cj audreyt: sure. The cafeteria doesn't serve the best food, but it's convenient for those in PacMed. You can get there from US1 (where you'll be speaking) on a shuttle. There is a lot of "international" food in the very near vicinity to US1, if you want to eat asian food. 19:56
[particle] thinks it's worth a trip to salumi if you can get there for lunch (and you eat meat)
ooh, better, audrey's at us1
19:58 the_dormant_ joined 20:02 nicodaemos joined
audreyt cj: I'll let you decide :) 20:06
lunch &
20:13 justatheory joined 20:16 larsen_ joined
cj audreyt: ha! You must think I'll be paying, then. I'd have to run that by the wife. 20:16
audreyt: how many folks do you want to see at lunch? 20:17
20:22 weinig joined
audreyt cj: I wouldn't mind... 20:36
any number of folks, that is 20:38
20:39 the_dormant_ joined
cj alrighty. I'll see if I can gather some people. We'll make a decision about where to eat based on how many people show up. Do you want to do lunch before, between or after your talks? 20:40
20:41 fglock joined
svnbot6 r14515 | kudra++ | Placeholder 20:41
20:45 the_dormant_ left 20:54 jferrero joined 20:55 doepiedoe joined
doepiedoe Hey ;) 20:55
i'm trying to translate a perl script to php, but can someone tell me what this piece of code does? 20:56
buubot correct syntax is: translate to/from language string
wilx Huh... 20:57
You are doing it in the wrong direction :)
The translation. 20:58
integral Pugs doesn't have a PHP backend yet? 21:01
21:05 SubStack joined
fglock integral: misc/Parser-Mini/pil1_json_emit_php.pl - but this was almost a joke 21:07
home & 21:08
21:08 ruoso joined, fglock left 21:09 the_dormant_ joined 21:19 Limbic_Region joined 21:46 mauke joined
qmole 2 21:47
22:01 justatheory_ joined 22:14 GeJ joined 22:16 Aankhen`` joined 22:20 BooK joined 22:41 thepler joined 22:44 bcorn joined 22:55 mako132_ joined 22:56 weinig is now known as weinig|bbl 22:58 justatheory joined
svnbot6 r14516 | the_dormant++ | added -M,-C,-A operators check 23:03
23:29 weinig|bbl is now known as weinig 23:41 dmq joined 23:47 thepler joined
cmarcelo @tell nothingmuch just commited more bits on moh, take a look to see where I'm having trouble understanding MO. I'll be back in about 2hr. 23:55
lambdabot Consider it noted.

Powered by Cro and the Raku® Programming Language.

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

Thank you!