pugscode.org | Beware of `make install` and Pugs, it will do bad things to your system
Set by Aankhen`` on 22 May 2007.
svnbot6 r16544 | Darren_Duncan++ | ext/QDRDBMS/ : reordered some parts of AST.pm (the file byte size is unchanged) 01:06
agentzh could anyone paste the IP address of feather? I'm having problems with my DNS server. :( 02:21
lambdabot agentzh: You have 1 new message. '/msg lambdabot @messages' to read it.
agentzh @messages
lambdabot Aankhen`` said 8d 22h 25m 54s ago: BTW, you might want to look at validator.w3.org/check?verbose=1&am...2FS02.html … I'd change the ID format to something
valid, but I'm not sure what it might break, so I'll leave that to you.
agentzh Aankhen``: thanks! I'll remove the leading underscores from the ID value :) it will break moritz++'s irclog but it's trivial to fix :) 02:24
oh, feather is out of reach due to DNS problem...
Aankhen``: feel free to update smartlinks.pl since I can't commit these days :( 02:25
Aankhen`` agentzh: Aye Cap'n. 03:21
svnbot6 r16545 | Aankhen++ | util/smartlinks.pl: 03:25
r16545 | Aankhen++ | * removed leading underscore from line anchors to ensure valid HTML.
agentzh Aankhen``: ++ 03:26
Aankhen`` Heh, thanks.
agentzh :)
Aankhen`` agentzh++ though, since all I'm doing is tweaking your work. :-) 03:27
svnbot6 r16546 | Darren_Duncan++ | ext/QDRDBMS/ : in AST.pm, added package TypeInvo and updated TypeDict to use it, made other small changes 06:14
meppl good morning 07:16
Tene hi 07:19
meppl good morning tene 07:29
xinming @tell stevan__ sorry t trouble again, Please give me the Moose repository url.
lambdabot Consider it noted.
spinclad xinming: code2.0beta.co.uk/moose/svn (or ask purl 'moose svn?' in #moose) 08:10
lambdabot Title: Revision 2378: /
xinming spinclad: thanks. 08:11
PDET hello! 08:14
PDET who wants to learn mandarin? 08:19
xinming PDET: ni gan ma? 08:20
PDET i want to find a part time job on the internet! 08:21
xinming ...
This is not the right place
PDET where should i go? 08:22
xinming PDET: bu zhi dao... 08:24
PDET thank you! 08:25
lumi TimToady++ # patch is so cool 08:41
moritz good morning ;) 09:16
svnbot6 r16547 | moritz++ | [irclog] fixed line anchors in synopsis links which were changed in r16545 09:18
IllvilJa Hm... I did a recent "make install" in my pugs build directory... Should I then count my two feet to ensure that I didn't shoot one of them off in the process? 10:52
moritz yes, you should ;) 10:53
lumi IllvilJa: moritz.faui2k3.org/irclog/out.pl?ch...-25#id_l16 -- Foot counter 11:03
lambdabot Title: IRC log for #perl6, tinyurl.com/2f47kz
lumi moritz: There's a bug in the log search (I think)
moritz lumi: tell me more 11:04
lumi moritz: I searched for that line, the link on the search page was to id_l15
moritz it has an offset of one, I know 11:05
lumi (Well, I searched for the line in *my* logs, and then searched for the nick in the search)
Okay
moritz I wanted to honor offbyone ;-)
lumi Hehe
moritz s/offbyone/offby1/
but if you think it should be fixed, I can do it
lumi How does that work for the last entry in a log?
moritz I don't see a problem with the last entry 11:06
lumi You're right, the first one
moritz the first one will get a non-existing anchor and stay at the top
lumi But it'll just get you to the top of the log, so I guess it works
Right
Well,, I'd fix it unless you want more people like me reporting it :) 11:07
moritz lumi: ok, on my TODO list
erm, @TODO ;-)
but I won't fix it until some time later, I've got a master exam today ;) 11:09
lumi But it's URGENT! :P 11:12
Good luck
Or good non-luck, as the case may be 11:13
IllvilJa moritz and lumi: if I got it right, the pugs install will do some... intresting stuff in the perl5 directories?
BTW, I did a "find /usr/lib/per5 -cmen 10080" but no files showed up. So maybe I'm lucky.
lumi IllvilJa: That's what I understand, also, it misinstalls over your perl5
I mean: That's what I understand too 11:14
IllvilJa Ok. Then I were lucky not having any misinstallations in perl5. Or perhaps I gave my find command the wrong time limit for the -cmin flag ;-).
lumi Did you spell -cmin right? 11:15
IllvilJa Hm. Maybe I should run "make install" in pugs, just to ensure that /usr/lib/perl5 get's "upgraded" :-D? (just my sick humor)
Yes. "-cmin 10080". Probably I can replace the -cmin flag with something else as 10080 is the number of minutes on a week.
lumi IllvilJa: I only asked because you mentioned it wrong a second ago, but it would've thrown an error prolly 11:16
IllvilJa Hm... when I ran that "find" command on /usr/lib/perl6 instead I also get zero files. So my use of the find command on /usr/lib/perl5 probably was a red herring... 11:17
(But when on earth does not I see any new files in /usr/lib/perl6??) 11:18
Nevermind. I'll address this (potential) problem on my machine the professional way: I'll take two aspirin and pretend it isn't there. 11:19
Tene IllvilJa: -cmin -10800 11:55
you need the '-'
svnbot6 r16548 | moritz++ | [irclog] fixed line anchors in search, lumi++ 11:58
IllvilJa Tene... thx, I'll test. 12:03
Hm... when I ran the command I just got 2 files in /usr/lib/perl5 which where touched last week. 12:05
One of them is a .pod file and the other is a .packlist file related to Perl6/Pugs, so my gut feeling is that my Perl5 install then is in pretty good shape. 12:06
Tene nods. 12:09
IllvilJa Of course, I won't let this prevent me from issuing yet another "make install" in my pugs build directory the next time there is a new release of Pugs (or parrot) ;-) 12:10
moritz presumably somebody will fix "make install" before a release 12:11
IllvilJa: in the case of parrot you don't need to be afraid, they manage their build system quite well
IllvilJa Cool! 12:12
Actually, regarding Pugs and Parrot, I'm never afraid (not even when I actually should be)...
moritz ;-)
IllvilJa The brave (or stupid?) early adopters I assume assist in getting Perl6 moving forward. 12:13
moritz the only problem with parrot's build system is that you can't "make -j 2"
IllvilJa I can live with that. And eventually, I suppose they will address that too.
moritz speaking of moving perl6 forward... do you have a commit bit?
fglock demq: ping 12:14
IllvilJa Um... no. I do unfortunately not have any time hacking ON Perl6, I'm barely looking into hacking WITH perl 6 so I suspect that a commit bit would be a bit wasted on me. 12:15
But if I get the possibility in the future to contribute with code to the Perl6 language or modules etc I'll ask if I can get that bit.
moritz IllvilJa: ok
IllvilJa But thanks for asking!
moritz IllvilJa: for example working examples/ are always welcome ;) 12:16
IllvilJa Aha!
I'll remember that.
lumi A bit wasted, haha
moritz sells commit bits on ebay ;-) 12:16
lumi I've seen them go on auction 12:17
IllvilJa Hm... commit bits and pieces... sorry, I won't go down that path. (Beware, I'm born and raised in Gothenburg, Sweden, a town whose inhabitants are notorious for their love for really bad jokes based on wording and twisting the Swedish language. So I rather stay away from "bit"-jokes... Like trying to get a "commit byte" for a workgroup of 8 ppl etc...) 12:18
Ok, got to focus on my work (which right now is coding in Perl 5, ironically). Thanks for the chat! 12:19
moritz ;-) 12:20
demq ironically coding must be hard. 12:22
youve got to steel yourself to do it. 12:23
fglock demq: you asked why Data::Bind instead of Data::Alias in v6.pm?
demq and if you arent careful you might come'a'copper. 12:24
yes
especially if you find yourself lead down the garden path.
fglock mostly because clkao wrote Data::Bind, and he was working on v6.pm at that time 12:25
demq ah
no other technical reason?
fglock I don't think so
Data::Alias looks pretty good 12:26
demq cool.
do you mind if i quote you here?
fglock sure
demq thnx
fglock but you talk to clkao too
s/you/you could/ 12:27
I'm trying to find this out in the irc logs 12:28
demq hopefully he reads p5p, and will pipe up if has something to add.
fglock I think we just ignored that Data::Alias existed - www.gossamer-threads.com/lists/perl...w_threaded 12:30
lambdabot Title: Mailing List Archive: Re: Data::Bind sematics in Perl core, tinyurl.com/yvfkt2
demq ok. 12:32
thanks. at the very least it clarifies the situation. 12:35
fglock thank you 12:36
brb &
fglock I've been thinking about how to "lexically modify the grammar" - for example, how is 'use Python' going to work? 13:14
demq die "Naughty boy" 13:15
:-)
er, die "Say no more" 13:16
fglock :)
demq the logic in the perl5 regex optimiser is twisted beyond all comprehension 13:17
fglock I see how categories are multi subs, that can be lexically redefined; but the Perl 6 grammar methods are not lexical (or are they?) 13:18
re twisted: even if you add some structure to a complex system, it is still complex 13:19
demq trust me, this is twisted. just plain old twisted. 13:22
i almost wonder if ilya was on acid when he wrote it
fglock aha, I forgot about $?PARSER
demq hmm, maybe if i took acid i could understand it.. hmmm 13:23
gaal possible janitorial Hs job, try out the new Data.Derive at www-users.cs.york.ac.uk/~ndm/derive/ as a replacement for DrIFT 13:43
lambdabot Title: Neil Mitchell - Derive
gaal though if someone has tuits and is looking for Hs stuff then moving more stuff from Pugs.Prim.* to MO methods is better! 13:44
"Hs tuitted individual" 13:45
xinming TimToady: Can we add a special variable into perl 6, which will be an alias of value returned by last executed sub? 14:08
rgs and blow up memory ? 14:09
xinming TimToady: eg, currently, in perl 5, we have, if (my $a = blabla()) { return $a }; If we have such veriable, we can use
if (blabla()) { return $last_return_variable }
rgs: No, It won't, just an alias. It won't take too much space.
lumi xinming: if blabla -> $a { return $a } # Wasn't there something like this? 14:10
rgs but you have to keep the original value around.
xinming lumi: What i want is, write something like... 14:11
blabla && return $last_excutes_variable;
rgs I'm only a perl-5'er, but in that context that doesn't sound like a good idea. Perl 6 might be different.
xinming lumi: It is still the same like 'if (my $a = blabla) { return $a }' in perl 5 14:12
demq a new keyword like returnif would do it.
demq returnif blahlah(); 14:12
xinming demq: But I don't think TimToady will add this keyword. ;-) 14:13
demq use feature returnif
:-)
gaal iirc there a similar proposal, that named this thing "it".
demq ive wanted stuff like that beofre.
xinming Because perl is different. ;-) 14:14
lumi macro returnif($a) { {{{my $t = $a; return $t if $t; }}} }
Or so...
xinming People use perl wishes to write code as less as possile.
possible*
demq its *kindof* like ||= or //= 14:15
gaal shouldn't there be an extra scope in that macro, to prevent $t from leaking?
xinming lumi: Such kind of thing is easy to add, But I think many people will like there will be built-in
lumi: It's a bit like the function 'say' in perl 5
lumi gaal: I think it's scoped to be a fresh name? Could be wrong 14:16
s/scoped/specced/
xinming sub say { print @_, "\n"; }
gaal macros get their own scope be default? doesn't sound, uh, sound.
lumi xinming: iswym, and I've been using that construct quite a bit
gaal: That's hygiene
xinming lumi: What does iswym mean? 14:17
lumi I see what you mean
gaal xinming: i see what you moose
lumi That too :)
gaal well, so long as there's a way to supress the extra scope.. 14:18
lumi if (blahblah()) { return $^it }
gaal $^ only works in ->s
lumi Oh? 14:19
I thought $^ and -> were bitter enemies
gaal let me refresh my moose..
lumi aol
xinming Perl will be a language for the next century... :-) 14:20
gaal oh, you're right; "it is illegal to use placeholder variables in a block that already has a signature"
xinming use GLOBAL '$*blabla*
use GLOBAL '$*blabla'
lumi gaal: As for the macro, :COMPILING makes it meanthe outside scope 14:22
Without it, variables default to freshness 14:23
perlcabal.org/syn/S06.html#Quasiquoting
gaal ACK, thanks
lambdabot Title: S06
lumi I guess I was using {{{}}} wrong though 14:24
gaal bbl& 14:29
lumi What's the win in Data.Derive over DrIFT? 14:45
gaal worth evaluating, but possibly: 1. faster 2. easier to use, because we can just bundle Derive instead of having anyone who wants to hack on modules that have instances need to install DrIFT themselves (which is kinda tricky with our symlink chicanery) 14:49
re-bbl&
lumi gaal: Ah, Data.Derive with its TH, I see 14:55
nothingmuch eautiful: galileo.phys.virginia.edu/classes/1...wnian.html 15:22
lambdabot Title: Einstein's Explanation of Brownian Motion, tinyurl.com/4undx
fglock nothingmuch: will you be at YAPC::EU? 15:42
nothingmuch fglock: yep 15:44
fglock great :)
nothingmuch i hope my talks get accepted 15:45
well, talk, really
the other is more of a "if you like I'm willing to give that too" kind of fller, that slightly applies to the social perl theme 15:46
moritz nothingmuch: which topic?
I mean the "real"
nothingmuch moritz: object meta programming
pasteling "nothingmuch" at 84.110.247.29 pasted "abstract" (12 lines, 1K) at sial.org/pbot/25188 15:47
moritz sounds nice ;) 15:49
nothingmuch climbing & 15:57
moritz have fun ;) 15:58
Limbic_Region salutations all 16:36
moritz hi L_R ;) 16:37
TimToady yes, "if blahblah() { return $^it }" should work 17:16
in fact, arguably "if blahblah() { .return }" should work if we take the use of $_ to be a requirement for a parameter to the block. 17:17
since "if" should respond to any block that has arity == 1 17:18
but I can also argue that the other way 17:19
since people will also expect "if .foo { .bar }" to use the same $_ for both. 17:20
renormalist When I (try to) port a P5 module to P6, what version number do you typically choose. 0.01, ie. from the beginning, or the version, that's ported? 18:31
demq when did the perl6 design phase start? 18:40
demq specifically would the perl.bootstrap be considered to be a perl6 or perl5 list? 18:42
or should it just be considered to be a perl list?
wikipedia seems to consider this important in describing "Warnocks Dilemma" 18:44
renormalist demq: just the wrong time. Probably in every timezone an important tv programm has started or such. try again in 1h. :-) 18:48
demq heh. thanks. i was starting to feel warnocked there 18:49
:-)
renormalist demq: btw, I don't understand the question. Do you ask, when the first apocalypse was written?
demq no, its silly trivia. 18:50
somebody was quibbling over the post that brought us "Warnocks Dilemma" and whether it was to a perl-5 or perl-6 list. in hindsight its a stupid debate. 18:51
as Larry has officially stated that he considers both to be just Perl. Therefore it doesnt matter
Limbic_Region actually, I thought it was on the perl6.internals list which really is parrot 18:52
but I am likely mistaken
svnbot6 r16549 | renormalist++ | - is extended|deep|also traits 18:58
demq apparently it was called perl.bootstrap 19:02
svnbot6 r16550 | lwall++ | Added note about parsing of pblocks in conditional statements. 19:16
riffraff hi 19:23
svnbot6 r16551 | renormalist++ | * cperl-mode: long time bug fixed: 20:14
r16551 | renormalist++ | distinguish between blocks and hashes now works.
renormalist Are there still the tr/// and y /// in Perl6? 20:39
TimToady tr/// only
renormalist ok. (simplifies a cperl-mode problem ...)
TimToady see S05:3260 20:40
renormalist ah, found it, thx 20:41
qmole what does the y stand for anyway? anything? 20:45
Khisanth lazY coder? :P
svnbot6 r16552 | renormalist++ | - cperl-mode: $.y no longer misinterpreted as start of y/// 20:46
moritz translyteration ;)
qmole hmm. :) 20:47
renormalist qmole: maybe they searched for a letter and someone said: Y not. And his partner answered: Why not Y? And then a third one in the room collected the most probable char statistically, which was Y in that dialogue. 20:51
renormalist should start an urban legend series about unanswerde mysteries in unix history. 20:52
gaal is this correct? ($obj but SomeRole).role_meth 21:03
or must you do something to make the invocant a clearer term?
gaal
.oO(hast is of the devil, eh?)
 21:05
TimToady gaal: I don't know how you could make it look any more like a term than the parens already do. 21:14
amnesiac <TimToady> see S05:3260 ---> that sounds like the bible! 21:16
moritz amnesiac: it kinda is - we just need a citebot ;) 21:17
amnesiac indeed, that would be pretty helpful.
moritz in the irc logs they alread link to the corresponding "verse" ;) 21:18
TimToady bbl & 21:19
moritz bbl = bible ;-)
TimToady ;)
really bbl &
gaal (($obj but SomeRole) but term).role_meth ? 21:22
I figured out why haskell drives people to distraction. In it, otherwise = True. 21:26
moritz monads drive me mad
Aankhen`` moritz: Hi, just in case you missed it, the line anchors on the synopses no longer start with an underscore. 21:27
moritz even quantum physics is easier ;)
Aankhen`` So it's just, f.e. "line_499".
moritz Aankhen``: thanks, fixed in r16547
Aankhen`` Kewl. ^_^
moritz++ for the logs BTW. 21:28
gaal moritz: in my experience, years of suffering help somewhat.
moritz thanks ;)
gaal seriouslier, pick more than one tutorial and see what they ahve to say.
moritz gaal: how very encouraging ;)
gaal: I do that already
gaal this is in opposition to most of the rest of the language, which you can learn from just say YAHT 21:29
moritz and it helped me with other topics, so I hope it will help me with monads as well
gaal oh, the YAHT presentation of monads isn't bad! it's just that (ymmv but) in my experience no single tutorial made me "get it". 21:30
we probably need a cubist tutorial writer, I mean painter, to get it all on cavas.
moritz I think the diversity is very helpfull 21:31
gaal moritz: oh, and one very specific tip: make sure you understand typeclasses.
Aankhen`` wonders whether control statements could take subs.
moritz Aankhen``: in haskell? sure ;-)
Aankhen`` Heh. 21:32
I meant in P6, of course. ;-)
moritz gaal: I decided to skimm over most of the tutorials, and then start a smal "real" project, and at the same time I read through it more carefully
the \ notation for lambda is something I can't get used to unless I'll write a whole lot of them myself ;) 21:34
gaal moritz: oh that's easy. just rewrite it in Perl 6 with -> :-)
another thing to do is to look at Parsec code 21:35
there's a lot of it in Pugs, and not all of it is crazy
moritz for me, a parsec is a unit for length/distance
gaal most of it uses do notation, but there's occasionally an explicit >>= bind
Parsec is a monadic parser combinator library 21:36
that is, it gives you tiny parsers for basic things, and you combine those to write parsers for bigger and bigger things
@google parsec daan 21:37
lambdabot www.cs.uu.nl/~daan/parsec.html
Title: Parsec
moritz so it's basically bottom up?
gaal YAHT introduces it nicely
moritz ok, I'll read that ;)
gaal it's rec descent, if i'm not mistaken, but when you write ll(1) in it it's fastest. 21:40
renormalist I have problems working in the pugs/ext/ subdir. I don't know how to use everything there autonomically outside the global Pugs makefile. I want do things like "perl Makefile.PL". "make test". Is this somewhere described?
currently I'm interested in how the tests are called. 21:41
make test doesn't do it right
gaal renormalist: in a pinch, just run make smoke instead 21:42
this is worth fixing though
you mean 'make test' doesn't descend into ext/ at all? 21:43
moritz renormalist: which module?
renormalist I want to only test my own little module. Is there an example somewhere
no in pugs yet. i struggle with basic class problems
maybe make test is ok, but another thing is worng with whateever, 21:44
gaal i think this may have been intentional, actually; once upon a time we were preflighting and the tests took too long so we didn't do ext/. (or am i imagining things?)
renormalist: to test your own stuff, sidestep the makefile and harness issues, use prove or util/prove6
could be that something isn't setting search paths correctly so it doesn't look at pugs/blib6/lib 21:45
renormalist ah, ok, porve6 works better
gaal: indeed, some path issues, but solved 21:46
gaal that would cause great slowdowns since the prelude will have to be parsed every hit, and possibly break things if someone's depending on someone else.
okay, great
renormalist btw, how can I create makefile from makefile.PL 21:47
gaal I'm glad; it's the first time I see how prove6 is actually useful for someone
"perl Makefile.PL"
renormalist hm, no. Without everything it creates a makefile that make complains about
from Pugs' make it generates other makefiles, that work 21:48
that was strange to find out, although now I know
you can test this with every module that has a :: in namespace.
in ext/
gaal then do the following: in ext/Whatever/Makefile.PL, near the top, 21:49
{ open my $e, ">", "/tmp/env.out" ; print "$_\t$ENV{$_}\n" for keys %ENV }
and run top level make
i bet it's PERL6LIB or something 21:50
renormalist mkay, I understand ,will try 21:50
gaal I suddenly noticed it's late here, moose y'all tomorrow then... 21:51
zZ&
Aankhen`` Nite.
moritz g'night 21:53
Aankhen`` gunihgt moritz
moritz is not gooing to sleep right now, just wishing gaal one ;) 21:56
Aankhen`` gunihgt moritz 22:00
Sorry, stuck in infinite loop. :-(
moritz Aankhen``: infinte loops are not too bad, unless they take up too many resources ;) 22:03
Aankhen`` Hehe. 22:04
moritz is in many semi-inifinte loops: eating, sleeping, hacking, learning... 22:05
and I don't wish them to terminate too soon ;)
Aankhen`` Haha.
bsb I've been thinking about anaphoric constructs recently too. That's ones that maintain an implicit topic like $_ 23:15
I got the name fom Paul Graham's book, he implements and "aand" that remembers the LHS for the RHS, as "it" I think 23:16
offby1 (and (odd? x) (+ 1 it))
bsb For a moment I though .{ } could mean "call this closure with the think on the left"
offby1 something like that
bsb think so, i'll check 23:17
bsb There's aand, aif, awhen, acond, eg. (aand (owner x) (address it) (town it)) 23:19
lib.store.yahoo.net/lib/paulgraham/onlisp.lisp
but .{} is a has deref for $_ right? 23:20
maybe "-> { }", without a sig?
anyway, I was thinking about it in the context of rules applied to trees, like writing an xml schema in p6rules 23:22

Powered by Cro and the Raku® Programming Language.

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

Thank you!