»ö« | perl6.org/ | nopaste: paste.lisp.org/new/perl6 | evalbot usage: 'perl6: say 3;' or rakudo: / pugs: / std: , or /msg p6eval perl6: ... | irclog: irc.pugscode.org/ | UTF-8 is our friend!
Set by wolfe.freenode.net on 30 October 2009.
00:00 xabbu42 left 00:03 TiMBuS joined, nihiliad left
jnthn pmichaud: I've got a start on the packages stuff. I'll continue on it tomorrow. 00:03
pmichaud: I think I'm past the initial conceptually tricky bit (figuring out how it should look). 00:04
00:04 pnate joined
jnthn So now it's "just code" for a while. Expect to commit something tomorrow. 00:04
00:04 pnate2 left 00:10 kidd` left, kidd` joined, Gruber joined 00:13 icwiener left 00:15 orafu left 00:16 pnate2 joined, orafu joined 00:17 nbrown left, tak11 joined
bpetering hai everyone :) 00:18
jnthn: (backlogging a lil): what's the issue with Parrot/Rakudo/MSVC++ ? 00:21
00:21 alester joined
bpetering reason being, i'm stuck on a vista machine for the next 2 weeks, and if masak wants his tests i need to get Rakudo running :) 00:22
just wondering what is the best way to do so.
jnthn bpetering: ohhai 00:23
bpetering: As far as I know, it's something to do with manifest.
Juerd bpetering: feather? :) 00:24
jnthn It gets right to the end and then fails when making perl6.exe.
Juerd: shush! We might get somebody to fix Rakudo on Vista! :-P
bpetering haha :P
jnthn :-)
Juerd jnthn: You think anyone else would notice? ;)
Anyway, I'm off to bed 00:25
bpetering I think vista needs to die, and if anyone has Rakudo running on it, it's slightly more reason to let it live :)
Juerd Good localtime.
bpetering night Juerd :)
jnthn Juerd: Well, I guess the other people who are stuck on Vista. ;-)
night Juerd o/
bpetering jnthn: i'm a little impatient, and judging from masak++'s latest blog post, he is too. 00:26
so i'm thinking cygwin.
also, how hopeful are you about Rakudo on Win 7? thinking about going that route in 2 weeks or so. :) 00:27
00:27 reid04 left 00:28 reid04 joined
bpetering err, "masak++++'s SECOND latest". My goodness that man is productive. 00:29
00:31 pnate left, pnate joined
bpetering jnthn: oops, reread what i wrote, don't interpret that as demanding anything from you. (waking up) 00:36
jnthn bpetering: I'm going to be doing Rakudo dev on Win7 in a month or so.
bpetering: I think it'll share the same issue as Vista. 00:37
00:38 orafu left
bpetering jnthn: ok, thank you :) 00:38
00:38 orafu joined
s1n yay, i can haz references! 00:38
s1n in reference (pun) to larry's recent commits 00:39
00:39 pnate2 left 00:44 Wolfman2000 joined
Wolfman2000 Evening. Does anyone know if there are any web frameworks already working on Perl 6? 00:44
00:44 pnate2 joined
bpetering Wolfman2000: check out github.com/masak/web 00:45
Wolfman2000 bpetering: thanks. *goes to read*
bpetering Wolfman2000: np 00:46
jnthn: and keep up the good work... you and pmichaud have been fearsomely productive of late :) 00:47
Wolfman2000 ...this seems like it can be useful. ...when it's time for me to upgrade to Perl 6 that is. 00:49
bpetering Wolfman2000: what's stopping you, out of curiosity?
Wolfman2000 bpetering: still used to 5.10.0
waiting for Perl 6 to become more...mainstream I guess 00:50
bpetering Wolfman2000: we expect to take over the world sometime around April next year :)
Wolfman2000 Then I'll keep an eye out
bpetering seriously, read about "rakudo star" - it'll make it possible for Perl 6 to start becoming mainstream :) 00:51
Wolfman2000 So that's why April 2010 is the goal. perl 6 is being released...not feature complete 00:54
bpetering Wolfman2000: Perl 6 is a specification developed at the same time as the implementations of it 00:55
Wolfman2000 I don't know if that's a good thing or not
bpetering so by necessity, the implementations have releases which implement and incomplete spec
00:55 pnate2 left
bpetering *an incomplete 00:56
00:56 Chillance left
bpetering we tend to think so, because then Perl 6 can avoid the problem of "oops, the language design was frozen way before anyone tried to do anything useful with it" 00:56
Wolfman2000 ...the reason I asked about the web development stuff...I right now use Catalyst for one of my websites. If you require Perl 6 testing, any sort of tutorial or guide to convert my page to the Perl 6 web framework could be useful. 00:58
bpetering Web.pm (URL i linked you to) isn't really ready for production use 00:59
01:00 alester left
Wolfman2000 I'm guessing...not much is ready 01:00
01:00 pnate left
bpetering we have nothing near Ruby on Rails yet, true 01:01
01:01 quietfanatic joined
bpetering but we're also aiming way higher than Ruby 01:01
Wolfman2000 Well, this is a Perl room
bpetering that's not a negative comment about Ruby. "It's a nice language, I use it occasionally" 01:03
Wolfman2000 Hard to tell: sarcasm on the web isn't my strong point 01:04
bpetering Wolfman2000: neither mine. (that wasn't sarcastic) :)
TimToady well, you can tell how much we like other languages by how much we borrow back from them. :) 01:05
and certainly, we like a lot of Ruby because some large percentage of it was borrowed from Perl 5 :)
Wolfman2000 TimToady: I was not aware of the language borrowing 01:06
TimToady Perl 6's twigil semantics are essentially Ruby's sigil semantics, for one example
and we're very much more Everything Is An Object than we were
on the other hand, Ruby's made some mistakes too, and we've tried not to borrow those :) 01:07
Wolfman2000 I think all languages have made mistakes at least once 01:08
PHP is looking to make one themselves with the namespace separator
...why \?
TimToady well, at least they'll *have* namespaces...finally... 01:09
and hey, Perl used ' for a while, which was a bit of a problem
bpetering Wolfman2000: the consensus around here is that "PHP is a demon child which should have been strangled in the crib"... (or worse) 01:10
Wolfman2000 bpetering: ...hope you don't mind that I actually have some websites that use PHP then. 01:11
bpetering Wolfman2000: not at all, it succeeded for a reason - namely, it got some things right
TimToady wel, one of our main rakudo developers has a wiki in php :)
what they got right was making it dropdead easy to get started 01:12
Wolfman2000 TimToady: most wikis are in PHP
TimToady most disruptive technologies succeed by doing something better and something else worse 01:13
but I suspect the php world is quickly reaching its natural limits, much as Perl 5 has. 01:14
both will see reduced growth in the future
Wolfman2000 ...do you require more testing with Web.pm? 01:16
bpetering Wolfman2000: we'd love any sort of contribution. Testing Web.pm would be great :)
TimToady which "you" are you referring to? "you" in this case is probably some folks who are asleep right now...
Wolfman2000 "you" in general
I have...4.7 GB of space on my virtual server. I shouldn't require that much for Perl 6, right? 01:17
TimToady I suspect that's probably a bit more than enough 01:19
Wolfman2000 ...stupid aptitude. told it to install git and it didn't even put it in the path 01:20
My virtual private server is Ubuntu, TW
BTW*
01:21 quietfanatic left 01:23 quietfanatic joined
bpetering Wolfman2000: like TimToady said, the author of Web.pm lives in Sweden, and so is probably asleep right now. 01:24
Wolfman2000 bpetering: Sweden...what's the time difference...5, 6 hours from EST?
bpetering GMT+2, I think 01:25
bpetering looks for DateTime.pm or somesuch
jnthn It's 2:34 right now in Sweden. 01:34
jnthn lives in the same timezone
2:34am that is ;-)
01:36 agentzh joined
sjohnson Perl 6 borrows a lot of good things from Perl 5 too :) 01:37
Wolfman2000 sjohnson: I figured as such 01:38
well, I have enough time to get git, get perl 6, and hopefully figure out enough of Web.pm to put one of my websites on it
Of course...I would like to be able to keep as much of my URL names the same as possible. 01:39
01:39 alester joined
bpetering Wolfman2000: all the best :) 01:40
01:40 nbrown joined
Wolfman2000 ...I just heard my dad laugh at one of the more recent Get a Mac ads 01:46
Sounded like PC was claiming that each version of Windows would have none of the problems the previous one did. "Trust Me". 01:47
Based on my own experiences with windows...I'll say no
pugs_svn r28978 | lwall++ | [STD] $_ is ref 01:48
01:50 ihrd joined, ihrd left
carlin zaslon: still alive? 01:52
zaslon Sorry, I don't understand that command
carlin :)
I think that false alarm about pm's blog was caused by a temporary connection problem ... need error handling there. Successfully alerted to masak's post though :-) 01:55
zaslon: add carlin theintersect.org/category/perl6/feed/ 02:04
zaslon I am now following carlin's blog
carlin zaslon: link carlin
zaslon carlin's blog is at theintersect.org
carlin zaslon: remove carlin 02:05
zaslon I am no longer following carlin's blog
02:10 xp_prg joined 02:15 JimmyZ joined
Wolfman2000 okay, installing Perl 6 now...and I forgot there are different...builds, I guess, of Perl6. There's Parrot, Pugs, Rakudo (which seems to use Parrot)...any others? 02:22
revdiablo Wolfman2000: Parrot is a VM, not a Perl 6 implementation. Rakudo is a Perl 6 impl that uses Parrot. Pugs is a Perl 6 impl that doesn't use parrot. 02:23
Wolfman2000 ah
revdiablo Wolfman2000: There are other Perl 6 implementations at various stages of donenness, SMOP and STD.pm (the latter of which is really just a grammar, as far as I know) 02:24
Wolfman2000: Rakudo is probably the most done, or at least the one with the most promising future as far as I can tell
bpetering Wolfman2000: Pugs hasn't been developed in a while... even so it's nearly as advanced as Rakudo is today, though jnthn and pmichaud are catching up fast :)
pmichaud I'm not sure I'd agree with that. 02:25
bpetering pmichaud: oh?
Wolfman2000 I'll just stick with rakudo for now since that's the one installing
pmichaud in some ways we've surpassed pugs a fair bit.
jnthn I think in some areas we're waaay past. :-)
revdiablo I thought Rakudo passed Pugs, but I could be wrong
jnthn Especially by the time ng lands.
:-) 02:26
anyway, sleep for me - night!
bpetering night jnthn, sleep well, get up late :)
diakopter 'night
revdiablo Interesting. Googling for Rakudo, I see perlbuzz.org as the top result and perlfoundation.org as the 2nd. Rakudo.org is the 4th 02:27
Wolfman2000 is still compiling Perl6...or is it Parrot right now? Gah, not used to compiling during a compile I guess.
02:27 xp_prg left
Wolfman2000 goes back to reading the synopsis 02:27
revdiablo Wolfman2000: Most sufficiently advanced computer systems have multiple layers that can or need to be compiled separately
Wolfman2000: Think of parrot as a required library =) 02:28
Wolfman2000 ...alright, I'll buy that
02:28 mjk joined
alester perlbuz.com 02:29
perlbuzz.com
and maybe that's because of perlbuzz.com linking to Rakudo.org everywhere
revdiablo Oh yeah, .com
Wouldn't that increase rakudo.org's pagerank? 02:30
Maybe the googlebots are intoxicated
alester who knows?
02:32 mjk left
bpetering Wolfman2000: which parts of the synopsis are you reading? 02:34
Wolfman2000 read a bit of S09, looking at S06 now 02:35
02:35 patspam joined
bpetering it can get topsy-turvy at times :) 02:35
Wolfman2000 agreed 02:36
also, right now redoing the compilation of Perl6: seemed like it hung for too long to my liking
revdiablo Wolfman2000: If you know perl 5, you might want to take a look at this series of posts: perlgeek.de/en/article/5-to-6 02:37
I find them much more easily digestible than the synopses
Wolfman2000 a refresher course wouldn't hurt. good idea
02:38 xp_prg joined
bpetering Wolfman2000: do you have a lot of programming experience? 02:39
Wolfman2000 bpetering: bacherlors in Computer Science. specialized more in the web development end, though.
bpetering Wolfman2000: oh, me too :) (still working on the degree though) 02:40
Wolfman2000 already have the degree from NCState
...okay, on the same line it hung...allegedly
02:40 agentzh left
lisppaste3 wolfman2000 pasted "Compiling rakudo hangs here." at paste.lisp.org/display/89705 02:41
bpetering Wolfman2000: that came up a day or two ago, let me backlog and maybe i can find it for you :) 02:42
Wolfman2000 ...okay, it's advancing now
that file just took awhile
doesn't exactly help that I'm SSHing into basically a linux box in the middle of nowhere 02:43
revdiablo Wolfman2000: Running the build in screen, I hope 02:44
Wolfman2000 actually...no
never did figure out how to get screen to work on Terminal
Besides: interesting to see the build process
src/thread.c:1386: warning: suggest braces around empty body in ‘do’ statement Don't know how much you guys care about keeping your code warning free, but this looks like an interesting one. 02:45
revdiablo It's usually pretty easy. You type "screen" and then it seems to do nothing, you are simply given another shell. The nice part is if your connection drops, you can ssh back in and reconnect to the terminal without losing anything.
bpetering Wolfman2000: that looks very interesting!
Wolfman2000 why would the compiler make suggestions? I thought it was supposed to just parse what it could or choke. 02:46
I never understood that.
02:49 araujo left
revdiablo Wolfman2000: Some things are legal but may be problematic. Warnings are often added after the feature's been around long enough to require backwards compatibility 02:50
Wolfman2000: The warning doesn't break back compat, but does dissuade use of problematic features
Wolfman2000 I believe someone mentioned recently that Perl 6 is being designed to avoid mistakes. Wouldn't keeping the warnings in...be a mistake? 02:52
also, new slow file: src/ops/core_ops_cg.c
bpetering hugs Wolfman2000 02:55
Wolfman2000 ...I thank you, but no thanks
don't know you that well
carlin bpetering: we have bots for that :-) 02:56
hugme: hug Wolfman2000
hugme hugs Wolfman2000
carlin Wolfman2000: how much RAM does the box you're compiling on have?
Wolfman2000 256 MB
bpetering Wolfman2000: well, if you hang around long enough maybe you will :) 02:57
Wolfman2000 alright...guess I may as well learn about screen
I've disconnected from my box now 02:58
what's the quick and dirty that I need to know for using screen to connect to my box?
03:00 xp_prg left 03:01 Whiteknight left
bpetering Wolfman2000: not sure i follow :| 03:01
Wolfman2000 ...nevermind. I think I figured out some of the basics 03:02
www.kuro5hin.org/story/2004/3/9/16838/14935
I've got my box screen detached...I assume it's running, doing it's thing...I just don't know if it's using much of my cable modem though or if it's going faster 03:03
bpetering Wolfman2000: don't think it's possible to go faster than that 03:04
Wolfman2000 than what it was going while I wasn't screened you mean?
bpetering pretty much, not sure i have the words to express exactly what i mean 03:05
revdiablo Wolfman2000: If you ssh back in, running "screen -x" will reconnect to your prior screen session 03:07
Perhaps I replied a little too late =)
Wolfman2000 revdiablo: I've actually been using screen -r to reconnect to my detached screen
revdiablo That works too. The difference being -x will let you connect to a session even if it's still connected elsewhere.
03:08 patspam left
Wolfman2000 ...even with screen, seems like some files just take awhile to compile 03:10
bpetering sometimes things don't happen as quickly as you want 03:11
i've been wanting to upgrade my laptop for a long time, give it some more RAM... make compiles quicker :) 03:14
Wolfman2000 my laptop (primary computer) already has 2 GB, but it's over 2.5 years old 03:15
bpetering Wolfman2000: time for an upgrade? :) 03:16
Wolfman2000 bpetering: no money
revdiablo Wolfman2000: Screen won't usually speed you up much, but it'll sure be nice if you get disconnected halfway through an hour-long compile
Wolfman2000 ...does perl6 normally take an hour to compile? 03:21
TimToady no, you're probably thrashing badly with only 256m 03:22
Wolfman2000 Let's see...4 websites to my name on that box. One on Catalyst::Perl, one on Pylons, one on CodeIgniter, and one with plain PHP. Using nginx for all four of them. 03:23
well, lighttpd + nginx for the PHP solutions
ooh...Perl 6 offers multi subs. Nice. :) 03:24
carlin Thought so ... there's a memory leak that wasn't there a week ago 03:29
03:29 lidden left
Wolfman2000 carlin: ...you mean some of the files I mentioned...are causing leaks? 03:29
carlin Wolfman2000: no you should be fine 03:30
Wolfman2000 ...was fine
and...my attempt at sending hte process to the background failed so I could check memory
...ooh boy. swap space is near full 03:31
03:35 cotto left
Wolfman2000 still...if it's not one file, it's another that likes to have trouble compiling 03:35
And I don't exactly have a lot of websites I can just...take off
bpetering Wolfman2000: which one? 03:36
Wolfman2000 src/ops/core_ops_cg.c <-- that file still hates me 03:37
brb: need a drink 03:38
03:42 patspam joined, patspam left 03:44 envi^office joined 03:45 zloyrusskiy joined 03:46 zloyrusskiy left
Wolfman2000 sorry about that. 03:46
Wolfman2000 restarts the compiling process again, then just C-a d's it 03:47
revdiablo Wolfman2000: yeah, that's the ticket 03:49
03:50 zloyrusskiy joined
Wolfman2000 ...okay, this is rich. I can assign custom operators to get things like 5! to mean factorial of 5. 03:50
revdiablo Wolfman2000: It seems like a lovely feature. I wonder how it will play out in practice. 03:51
Wolfman2000 I have to wonder if utf-8 characters can be used for these custom operators
bpetering Wolfman2000: (x)emacs?
Wolfman2000 bpetering: if you're asking what type of text editor I use on my VPS box... (x)emacs > vi(m), but nano > (x)emacs
japhb Wolfman2000, yes, you can define operators with the full Unicode character set. 03:52
Wolfman2000 japhb: Just confirmed. € is used for a doubling operation
japhb Makes defining math modules quite nice. Like actually using the native symbols for set operations, vector calculations, calculus ... 03:53
03:53 zloyrusskiy left
lisppaste3 wolfman2000 pasted "Square Root Prefix Function?" at paste.lisp.org/display/89706 03:54
Wolfman2000 Hopefully someone who already has Perl 6 can give that a shot 03:55
quietfanatic rakudo: sub prefix:<√> ($x) {sqrt $x}; say √2 03:59
p6eval rakudo 74f561: OUTPUT«1.4142135623731␤»
quietfanatic Can't change precedence yet though
I think
rakudo: sub prefix:<√> ($x) is tighter(&infix:<*>) {sqrt $x}; say √2*4 04:00
p6eval rakudo 74f561: OUTPUT«No applicable candidates found to dispatch to for 'trait_mod:is'␤in Main (file <unknown>, line <unknown>)␤»
quietfanatic rakudo: sub prefix:<√> ($x) {sqrt $x}; say √2*4
p6eval rakudo 74f561: OUTPUT«5.65685424949238␤»
04:00 JimmyZ__ joined
quietfanatic rakudo: sub prefix:<√> ($x) {sqrt $x}; say √8 04:01
p6eval rakudo 74f561: OUTPUT«2.82842712474619␤»
quietfanatic huh
oh
rakudo: sub prefix:<√> ($x) {sqrt $x}; say √16
p6eval rakudo 74f561: OUTPUT«4␤»
quietfanatic hum
Wolfman2000 ...what is the NL character?
quietfanatic That's a newline 04:02
Wolfman2000 fanciest way I've ever seen a newline
quietfanatic Unicode FTW
Wolfman2000 what character is it?
err, point I guess
quietfanatic U+2424: ␤ 04:03
Wolfman2000 I wonder if buu can put that into buubot
ENUMS! 04:04
True enums in Perl 6...oh man
so far, the main enums I've liked the best were Java's enums. They provided...more than the traditional enum
quietfanatic They're not entirely perfect yet, but they're close
Perl's enums will act like an int or a type 04:05
For instance, Bool is an enum
Wolfman2000 True and False
...wait, Perl now has True and False? 04:06
quietfanatic enum Bool <False True>
Wolfman2000 Good grief, you're going to make me squee so much I may hurt myself.
quietfanatic They're still 1 and 0, just enum'd
IIRC
Wolfman2000 ...better than 0 and -1 at least 04:07
bpetering i agree completely
quietfanatic Yes; in Perl 5 I attach a ! to the beginning of every system() call
Wolfman2000 ...I meant -1 and 0 04:08
where -1 is true
quietfanatic Well -1 is true...
rakudo: if -1 {say "-1 is true"}
p6eval rakudo 74f561: OUTPUT«-1 is true␤»
Wolfman2000 quietfanatic: I'm not questioning that
I just mean...1 and 0 are more universally known as true and false 04:09
versus -1 and 0
quietfanatic of course
pmichaud plumage's Configure.nqp should use pir::load_bytecode(...) instead of load_bytecode(...)
oops, wrong chan
Wolfman2000 ...can anyone perhaps tell me what a good revision of rakudo is for me to compile? I think the latest git/svn version is...not working for me. 04:10
...wait. what version OF GCC do I have/ 04:11
?*
lambdabot Maybe you meant: . ? @ v
Wolfman2000 ...doesn't seem like I have that bug 04:12
Wolfman2000 puts in the optimize fix...maybe that will do it
revdiablo Wolfman2000: I would probably start with one of the monthly releases 04:13
04:13 stephenlb left
Wolfman2000 revdiablo: If this doesn't work, I'll look into it 04:13
TimToady if it's the thrashing problem, the monthly probably won't be any better 04:14
Wolfman2000 No clue if it's thrashing
that's the problem
revdiablo Wolfman2000: Is it bailing with an error or simply taking longer than you expect?
Wolfman2000 latter 04:15
I force cancel it with C-c
TimToady the core ops file is probably a huge switch statement
revdiablo Wolfman2000: Wait it out and see if it finishes 04:16
Wolfman2000 this time, I intend to
revdiablo Wow, the rakudo spectest progress graph looks pretty amazing these days 04:17
Wolfman2000 revdiablo: got a pic or link?
04:17 envi^home joined
revdiablo Wolfman2000: rakudo.org/status 04:17
I wonder if any of the tests I wrote for pugs are still in there. =) 04:18
Wolfman2000 alright...looks like I'm waiting 40 minutes before checking 04:19
Hmm...is there no color for tests that are expected to fail, but pass anyway? 04:20
revdiablo Wolfman2000: btw, you can create another window in your screen session with C-a C, then bounce between them with C-a C-a
Wolfman2000 Next thing I need to learn about screen: how to keep my colors.
revdiablo I mean C-a c
04:21 JimmyZ left
Wolfman2000 revdiablo: C-a C-a seemed to work 04:21
...oh, I see what you meant
04:22 mepplock joined
Wolfman2000 TimToady: is it possible to make the core ops file...well, many smaller files? 04:29
or at least, is there some way to identify if my box is being thrashed? 04:30
I can take down my websites for a few hours if need be.
TimToady you'd have to ask one of our other experts, like pmichaud++ 04:32
pmichaud Wolfman2000: it's taking forever to compile parrot ?
Wolfman2000 pmichaud: rakudo...which uses parrot, so yes
seriously, the two of you have good timing 04:33
TimToady I paged him
I think he's asking whether it's during the Configure --gen-parrot or during the make?
pmichaud it's during the make 04:34
if it's core_ops
it's during parrot's make, sorry
you're right, it's during rakudo's Configure --gen-parrot
Wolfman2000 so it's both
pmichaud what os, wolfman2000?
(hard for me to find details in the backscroll)
Wolfman2000 Ubuntu 8.04 I believe
could be 8.10
but not any higher than that
pmichaud what processor?
64-bit or 32-bit? 04:35
Wolfman2000 don't recall that one offhand
pmichaud maybe "uname -a" will help
if it says x86_64, then 64bit
Wolfman2000 long as C-z works
pmichaud if it says i386 or i686, then 32-bit
Wolfman2000 64 bit 04:36
Wolfman2000 just made a new connection to his box
pmichaud how much memory on the system?
Wolfman2000 256MB
pmichaud oh.
very likely that's not enough to build Parrot quickly
TimToady that's small even for 32-bit compile 04:37
pmichaud so I'm guessing thrashing
how much swap?
Wolfman2000 unless I want to take my websites down for a few hours, I know
double
pmichaud yeah, definitely pretty small
just a sec, there's an option that might help
04:37 meppl left, petdance joined
Wolfman2000 hard to argue that for $20/month, even in this bad economy 04:37
04:37 quietfanatic left
pmichaud try running Rakudo's Configure.pl again as follows 04:38
perl Configure.pl --parrot-option='--cgoto=0' --gen-parrot 04:39
that will probably get you there.
Wolfman2000 considering I haven't read much about Configure.pl, I'm not exactly aware of what that option did
pmichaud at least, it's much more likely.
Wolfman2000 at the very least, that new cgoto thing
pmichaud that passes an option to Parrot's build system saying "don't compile the computed goto option", which is what is taking so long
from Parrot's Configure.pl: 04:40
--cgoto=0 Don't build cgoto core - recommended when short of mem
Wolfman2000 ...why is there goto in Perl 6 anyway?
Or is this something I don't want to know?
pmichaud it's not in Perl 6
TimToady that's a C goto
pmichaud it's in the Parrot VM, in C
TimToady so yes, you don't want to know :) 04:41
pmichaud even I don't want to know (and I don't, really :-)
TimToady it only just occurred to me that _cg means "computed goto"
Wolfman2000 ...I wonder why I never learned of computed gotos in class 04:42
okay, on the first troubling file: core_ops_switch.c 04:44
04:45 patspam joined, patspam left
Wolfman2000 ...it's advanced quickly 04:46
or at least, quicker than before
now on core_ops_cg.c 04:49
04:50 pnate joined 04:51 quietfanatic joined 04:52 xp_prg joined
Wolfman2000 ...I think this is the big switch statement area. I will look forward to when the src/ops/* files are split up 04:55
...wait a second...
GAH, I can't scroll up fast enough to stop screen 04:57
pmichaud: I just tried starting it again. Configure.pl complained of the parrot-option thing
pmichaud oh I had it wrong
sorry about that
perl Configure.pl --gen-parrot-option="--cgoto=0" --gen-parrot 04:58
my bad
(I forgot the --gen on --gen-parrot-option)
Wolfman2000 ...this is certainly an adventure 04:59
bpetering Wolfman2000: on what scale are you talking?
Wolfman2000 been awhile since I felt I could...just chuckle at my poor attempts at compiling code 05:00
bpetering ah :)
TimToady we all chuckle at ourselves frequently here, so you fit right in :) 05:02
Wolfman2000 here we go...on src/ops/core_ops.c
relatively quick 05:03
and...the switch file is at it again. I think parrot hates me 05:06
05:09 cdarroch left
Wolfman2000 ...the big problem with screen is that I can't easily scroll up. so I have no clue when my history goes out of whack 05:09
Wolfman2000 goes back to normal ssh mode to be sure he can copy and paste the output as required 05:11
...good thing I can stay up for awhile tonight
05:13 cotto joined
Wolfman2000 ...I think I'm seeing new files being compiled in now 05:23
compilers/imcc/imcparser.c <-- that's definitely new.
Perl6, I will compile you yet! 05:24
05:25 mepplock is now known as meppl
Wolfman2000 TimToady, bpetering, pmichaud, et al: parrot made 05:26
bpetering Wolfman2000: \o/ 05:27
05:34 sifusam left
Wolfman2000 ...at this point, I'm not going to be surprised if rakudo doesn't want to make now. 05:36
off to bed now. I can mess with this in the morning...and take down some of my websites if required. 05:45
bpetering good night Wolfman2000 - again, all the best :) 05:48
pmichaud once you get past the parrot build, things tend to work.
at least, until you get to the parts that don't. =-)
carlin rakudo: loop { my $a = 'OH HAI' }; 05:49
p6eval rakudo 74f561: TIMED_OUT 05:50
carlin In 657d55cce1f1 that's fine - the process sits at ~50MiB. As of d4f03670df0af the process slowly consumes more and more memory. 05:51
(Couldn't test anything between there because it segfaults during build)
05:52 pnate2 joined 05:58 petdance left
carlin so I think I'll roll back the rakudo on my server, which will hopefully stop my IRC bots from leaking memory 06:02
06:02 zaslon left 06:05 pnate left 06:12 pnate joined 06:13 pointme joined, zaslon joined 06:14 bpetering left 06:16 envi^office left 06:27 pnate2 left 06:31 araujo joined 06:34 justatheory left 06:36 NorwayGeek joined 06:37 pnate2 joined 06:38 pnate2 left 06:39 pnate left 06:47 akl left 06:51 hirschnase joined 06:57 agentzh joined 07:07 JimmyZ__ left 07:08 NorwayGeek left 07:12 viklund left 07:24 pnate2 joined, Su-Shee joined
Su-Shee good morning. 07:24
07:28 kidd` left 07:31 justatheory joined 07:33 hirschnase left 07:34 NorwayGeek joined
moritz_ good morning 07:37
Su-Shee of course you are awake. ;) 07:38
07:41 justatheory left
pmichaud good morning 07:42
and yes, it's morning here also :)
moritz_ Su-Shee: just got up ;-)
Su-Shee pmichaud: welcome to the morning shift. ;) 07:44
moritz_ pmichaud: did you change time zone? ;-)
pmichaud no
it's just that your morning is a lot later than my morning 07:45
or something like that
at any rate, locally right now the hour portion of the time is less than 12h, so that officially makes it morning, I think :)
moritz_ ;-) 07:46
spinclad still erev, tho, i venture; not yet boqer 07:49
spinclad is an hour dawnward from pmichaud 07:52
08:05 TiMBuS left 08:07 iblechbot joined 08:14 mariuz joined, ihrd1 joined 08:19 TiMBuS joined 08:20 hirschnase joined 08:22 hirschnase left, hirschnase joined 08:27 abra joined 08:37 abra left 08:40 drry joined 08:41 drry left 08:42 ihrd1 left 08:47 xenoterracide left, IllvilJa left, agentzh left 08:48 jferrero joined 08:49 xenoterracide joined 08:50 xenoterracide left 08:51 xenoterracide joined 08:55 rfordinal joined, xenoterracide left 08:58 xenoterracide joined 09:04 IllvilJa joined 09:07 NorwayGeek left 09:10 rfordinal left 09:11 reid04 left 09:12 reid04 joined
moritz_ pmichaud, [particle]: the rakudoconsulting website still links to parrotcode.org (and has one CSS validation error) 09:12
pmichaud thanks 09:16
particle will need to fix it
09:17 viklund joined, NorwayGeek joined, SmokeMachine left 09:28 dakkar joined 09:38 frodwith left 09:47 frodwith joined 09:48 rgrau joined 10:06 frodwith left 10:10 frodwith joined 10:17 slavik1 joined 10:28 khisanth_ joined, khisanth_ left 10:39 NorwayGeek left 10:40 NorwayGeek joined 10:52 SmokeMachine joined 11:06 rfordinal joined 11:10 rfordinal left 11:13 broquaint joined
jnthn yay snow! \o/ 11:18
mathw it's just grey and damp here 11:26
don't know why the cat was so keen to go outside this morning
he'll have already been rained on twice
Su-Shee snow? I'm envious. 11:28
mathw better move to bratislava then
Su-Shee around the corner anyway :)
jnthn Yes, I suspect it's not all so far. 11:29
jnthn forgets which city Su-Shee is in
mathw: It doesn't matter what the cat wants to do, just that the cat is Always Right. :-)
mathw even when he changes his mind thirty seconds later
jnthn Even then. :-) 11:30
It's amazing what cuteness does for ones always-rightness.
mathw yes and my cat is almost universally agreed to be utterly adorable
he even won over my Dad, who doesn't like cats 11:31
moritz_ so the perfect source for lolcat pics? :-)
mathw class Cat does Infallible { ... }
www.flickr.com/photos/mattofwalton/...820239428/ <- judge for yourself 11:32
jnthn aww 11:33
mathw That's what everyone says! 11:34
he's charmed the entire PLANET
even the vet thinks he's gorgeous
even there he won't stop trying to cuddle people - including the woman with the giant needle she's about to stick in him 11:35
moritz_ www.flickr.com/photos/mattofwalton/...820239428/ <-- nice for illustrating laziness 11:36
jnthn is taken by the cuteness 11:37
mathw My sister kept saying 'but he's so soft and cuddly!' 11:40
and then shrieking when he stuck his claws in somewhere sensitive
jnthn lol! 11:41
mathw I don't think she was quite prepared for him to jump into her lap and demand stroking 11:44
fortunately he had the sense not to try that with my parents
12:11 envi^home left 12:14 envi^home joined 12:17 hanekomu joined 12:20 astrojp left 12:21 astrojp joined 12:22 wolverian left 12:23 wolverian joined 12:39 masak joined
jnthn lolitsmasak! 12:39
masak good afternoon, #perl6.
jnthn oh hai, masak
mberends oh hai you both 12:40
masak mberends! \o/ 12:41
mberends: I thought I'd make Tuesdays "Temporal Tuesday" during the November month. 12:42
mberends hee
masak it's not directly related to November, but it would definitely be good for Rakudo Star.
12:43 NorwayGeek_ joined
jnthn is glad masak++ is making time for this 12:43
*cough*
masak ah. time jokes. 12:44
been a long time since we did those.
jnthn I'm sure I can come up with more in an instant.
And sustain them for the duraction of your hacking.
12:48 NorwayGeek left 12:49 petdance joined 12:50 iblechbot left 12:51 NorwayGeek joined, NorwayGeek_ left
masak jnthn: :P 12:54
takadonet morning all 12:59
13:01 FCO joined
masak takadonet: o/ 13:03
takadonet how are you doing masak?
masak takadonet: I'm... well-rested today. that's good, I guess.
it's perhaps indicative when I get to work and people say "Ah, masak! *there* you are!" 13:04
takadonet hehe
masak takadonet: how are you?
takadonet masak: feeling sick again. Got a nice perl 'script' from another bioinformatics group that taken some time to figure how it works 13:06
masak ok. 13:07
takadonet I hate that half of the input file need to be in 'dos' format and other half in unix/linux!
stupid carriage return and line feed....
masak needs to be? why? 13:08
13:08 frettled sets mode: +ooo masak mberends takadonet
frettled good something :) 13:09
takadonet: my sympathies, btw.
13:10 Patterner left
takadonet Otherwise the 'parser' will fail to read the file and crash the problem. I do not want to change too much of the script since not sure if we are ending up using it or not. Depends how well it works 13:11
jnthn heh, reminds me of a talk at the Italian Perl Workshop. 13:13
Theme was essnentially, "The data is always shit!" :-)
13:14 TiMBuS left, Psyche^ joined, SmokeMachine left 13:15 Psyche^ is now known as Patterner
takadonet you can NEVER trust data from users. 13:15
jnthn s/data from // # ;-) 13:16
cognominal what am I doing wrong with the ng branch? /Users/stef/git/rakudo-ng/parrot_install/bin/nqp --output=src/gen/signature_pm.pir --encoding=utf8 \
--target=pir src/Perl6/Compiler/Signature.pm
make: /Users/stef/git/rakudo-ng/parrot_install/bin/nqp: No such file or directory
takadonet hehe 13:17
jnthn Failing to install nqp-rx, at first guess.
cognominal probably something in my path
*PATH
13:17 NorwayGeek left 13:19 tak11 left 13:21 Psyche^ joined, Patterner left, Psyche^ is now known as Patterner 13:25 wolverian left 13:26 wolverian joined
masak anyone here speaks Japanese? :) 13:31
Su-Shee badly.
masak sounds like this Twitterer is looking for information of some kind. twitter.com/__gfx__/status/5379551557
something about 'strict constructors'?
Su-Shee oh, you want japanese _read_ ...
masak ah, yes. sorry. 13:32
anyone here reads Japanese? :)
13:32 nbrown left 13:33 colomon left
hanekomu masak: a bit 13:36
what do you need?
ah
that link
can't find a description about Perl6's constructor re strict 13:38
13:38 petdance left
hanekomu (not sure what "de aru koto ni tsuite" means - will look up the grammar 13:38
but ask in irc.freenode.net #soozy, where japanese perl hackers hang out
moritz_ tsuite = test suite? 13:40
hanekomu no
ni tsuite = about, regarding
de aru koto = thing exists
moritz_ very bad at guessing
hanekomu i'm asking on #soozy
kazuho on #soozy told me: "can't find description stating that constructors are strict in perl6" 13:43
jnthn "strict"? 13:45
hanekomu no idea, i just tried to translate the thing :)
jnthn :-)
13:45 colomon joined
hanekomu ah, "X de aru koto" = "being X", as, so more literally it is "Perl 6's constructor being strict - I can't find a description about it." 13:46
moritz_ maybe related to www.cpanforum.com/posts/7293 13:47
so if anybody cares to answers: it is an error to set a non-existing attribute in a constructor in Perl 6 13:48
13:49 NorwayGeek joined
moritz_ I don't know if that qualifies as a "strict constructor" 13:49
mathw Well it's some kind of strictness I suppose 13:50
does it count as an unknown named parameter i.e. ignored?
moritz_ well, there are two things you can wrong 13:51
the first is calling it with wrong arguments
that still fails silently in Perl 6
hanekomu afterwards, gfx tweets twitter.com/__gfx__/status/5379559571
moritz_ the second is mis-spelling an attribute in a constructor
that one is caught by Perl 6
mathw umm 13:52
how is that different to wrong arguments?
hanekomu something like "strict constructors strike me as being fairly high-speed. Before long I intend to implement it in Mouse"
(if i'm not too mistaken)
moritz_ mathw: making the error in the class itself vs. on the caller side
13:55 colomon left
mathw doesn't follow 13:55
moritz_ mathw: the main difference is that in Perl 6 you have a default constructor 13:57
masak I replied to __gfx__. twitter.com/carlmasak/status/5391649893
moritz_ mathw: if you have a class A { has $!quuox }; and call A.new(quuuuox => 3), this is silently ignored
masak argh, meant 'non-existent argument'.
moritz_ mathw: however if you write a custom constructor in which you try to assign to $!quuuuox, you'll get an error 13:58
masak moritz_: oh. then my reply to ___gfx__ was wrong, I guess.
13:59 patspam joined
mathw moritz_: okay, yes of course. I thought that was so obvious that you were talking about something else :) 13:59
moritz_ masak: just point him to our IRC discussion in the logs
masak moritz_: nod.
moritz_ it's not something that fits in 140chars if you're new to the subject
masak he doesn't seem to be, though. :) 14:00
14:00 ejs joined
masak Wolfman2000: greetings! not sure if you're awake now, but I'm currently catching up on your backlog trails. it's cool that you're showing an interest in Web.pm. 14:08
Wolfman2000 *yawn* morning
masak Wolfman2000: I'm sure we can think of a way for you to help the project along. :)
Wolfman2000 masak: The website in question shouldn't involve...too much work. It's just slightly horribly outdated in terms of content though. 14:09
masak which website?
Wolfman2000 www.jasonfelds.com <-- that's the Catalyst page. View it quickly: I may be tempted to bring it down so I can actually get rakudo to build
masak views it quickly
'is For Sale' is all I get. 14:10
Wolfman2000 ...for...sale?
that's a new one
masak oh, sorry. 'felds', not 'fields'.
trying again. 14:11
14:11 Psyche^ joined
masak ah. here we go. 14:11
Wolfman2000 tested another page of mine that's not on catalyst: the pages load right now
masak Wolfman2000: you know, I'm looking at your cretentials and thinking that you'll fit right in here. welcome aboard! :) 14:13
Wolfman2000 My resume is more updated...on my mobile page actually
www.jasonfelds.mobi
not on Catalyst, but it works on cell phones
masak why two different URLs? 14:14
14:14 iblechbot joined
Wolfman2000 masak: fair question. mobi is for those on mobile devices: com is for everyone else. 14:14
not all phones can do com
mobi is specifically meant for mobile devices. Not just the iphones, but the other ones
masak hm.
Wolfman2000 This is more of my web development background coming in 14:15
masak nod.
sbp I worked on CC/PP for a bit, which was supposed to make this redundant
man did THAT fail...
masak Wolfman2000: I greatly prefer the style sheet on your .mobi site, for what it's worth. :) 14:16
Wolfman2000 masak: That stylesheet was copied from another stylesheet without me really thinking about it
The desktop part is a slight rush job admittedly
masak Wolfman2000: there's two of us (me and Tene) who are currently finishing up our Web.pm grant. it involves providing basic tools in Web.pm for building web applications. 14:18
Wolfman2000: after that, I imagine we'll be focusing on a Catalyst-like framework.
Wolfman2000 how much is Web.pm usuable in its current state?
masak Wolfman2000: it runs. let me dig out a link for you to make it concrete. 14:19
Wolfman2000: github.com/masak/druid/blob/master/.../Webapp.pm
that's about the state of the art right now. 14:20
what you see on that page is quite a small program which, when run, acts as a web application.
14:20 colomon joined
Wolfman2000 Something about moving others around a simple board. 14:21
If it helps, my .com is...nowhere near that complicated.
masak apart from that, there are other interesting tools on the way. database layer, html templates, and simple dispatching.
Wolfman2000 As long as I can figure out Routes.pm, I think I will have a lot covered.
masak Routes.pm in the Web.pm project? 14:22
Wolfman2000 yeah
That's the only heavy thing I'm using.
masak hm, haven't looked at that one in a while.
Wolfman2000 I like /resume versus /resume.html
masak ihrd++ wrote Routes.pm. he has prior experience with Rails.
maybe Catalyst too. I don't remember.
but he couldn't continue on the project, due to the deteriorating global economy. :/ 14:23
PerlJam greeble
Wolfman2000 ...if it helps, Catalyst probably has a working implementation of Routes 14:24
so you could just borrow that code
masak aye.
I plan to be inspired by Catalyst a great deal.
14:24 Patterner left, Psyche^ is now known as Patterner
masak I've bought the book. 14:24
and I have good relations with people working on Catalyst. 14:25
PerlJam Catalyst++
moritz_ I just think it takes too long to figure things out in catalyst
masak I'm also looking at a number of other frameworks, to widen my perspective.
moritz_ getting started is non-trivial
masak moritz_: nod.
I want simplicity.
I think Jifty does a better job there.
Wolfman2000 masak: in about...10 minuts, I'm going to pull some of my sites down to try to get the memory I require to compile rakudo properly 14:28
PerlJam masak: you want the python to our perl? (Catalyst is "optimized" for customizability everywhere where as something like Jifty (or Rails) are optimized to getting things going quickly. This seems very much like python and perl)
masak PerlJam: I want simplicity for small things and customizability for big things. 14:29
PerlJam (easy things easy, hard things possible)++
masak PerlJam: that may mean two different frameworks, but I'm hoping it means one, where the defaults are tailored towards getting started easily. 14:30
PerlJam: I'm facing the same sorts of issues with Web.pm in general.
Wolfman2000 I think Catalyst was meant to be one framework for all
masak oh, for sure.
mathw Catalyst does make easy things fairly easy, I thought
I played with it for a while
PerlJam mathx: Rails' opinionated software gets you going quicker than Catalyst (at the expense of some customizability and some other design tradeoffs) 14:32
er, mathw
mathw yes
I played with Rails too
I discovered that you can write a blog engine in 15 minutes
But anything else will drive you insane
masak :)
I don't want that. 14:33
PerlJam mathw: anything significantly more complex you mean
If you have need of simple database driven sites with small relationships between the tables, rails and jifty are perfect and simple
jnthn A lot of the time, that's enough, though. 14:34
Wolfman2000 My problem with a lot of blog packages...and many other packages in general...
PerlJam jnthn: yep, that's why the rails hype machine was able to take off.
Wolfman2000 I prefer Postgresql over MySQL. Not all of them support the good features of PG
jnthn The things that put me off e.g. Jifty are its dependencies and that it's too magical.
I don't *want* magic I can't understand. 14:35
Or that takes me weeks to grok.
PerlJam jnthn: Hmm. s/jnthn/random perl user/ and s/Jifty/Perl 6 OOP/ and that might be a problem for Perl 6 too ;) 14:36
jnthn Perl 6 OOP has magic?
moritz_ I think the basics are rather easy to grok
PerlJam jnthn: magic that might take people weeks to grok
Wolfman2000 I think Perl has more magic than we want to believe
moritz_ Wolfman2000: it has, but we try very hard to hide it :-) 14:37
jnthn It's got a bunch of sugar, but I'm not convinced the underlying model is all that bad. :-)
Wolfman2000 alright...time to see if I can take websites down properly (and hope my new 50x webpage shows) 14:38
14:38 KyleHa joined
Wolfman2000 ...nope. wrong file 14:38
gah, what's the unix way of searching a file for a specific phrase again? 14:39
moritz_ grep word $file
Wolfman2000 moritz_: and if you don't recall the file the phrase is in?
moritz_ Wolfman2000: rgrep word $dir
or rgrep -r word $dir
PerlJam I think you all mean "ack" :)
mathw find $dir | xargs grep word 14:40
moritz_ or if you're a perl hacker you install 'ack' from CPAN and use that instead :-)
Wolfman2000 I need to see which file contains the phrase "502 Bad Gateway"
I don't recall if I have ack from cpan installed on here
moritz_ /usr/share/apache2/error/HTTP_BAD_GATEWAY.html.var 14:41
on my debian box
Wolfman2000 moritz_: nginx
14:42 ejs left
Wolfman2000 okay, pages are down, decent looking placeholder up. 14:49
Parrot, Rakudo...you're mine
moritz_ all ur birds r belong 2 us! 14:50
Wolfman2000 lots of memory and swap free. Let's do this
masak \o/ 14:51
PerlJam Wolfman2000: Are you really on a system that is t hat starved for resources?
Wolfman2000 PerlJam: Virtual Private Server, 256 MB
PerlJam ah. 14:52
good luck!
Wolfman2000 I used perl Configure.pl --gen-parrot-option="--cgoto=0" --gen-parrot to configure it. The computed gotos was part of the problem
now...it's just the slow waiting for something to happen
I'll assume that once Rakudo is up, I can put my websites back online without a problem 14:53
14:53 envi^home left 14:54 kidd` joined
masak Wolfman2000: I'd assume that, too. 14:55
[particle] until you try and run rakudo...
moritz_ sends Wolfman2000 a spare 1GB ram module 14:56
[particle] then you might have problems, depending on your source.
Wolfman2000 moritz_: money would help more
[particle]: rakudo shouldn't require that much resources to run a simple website. at least...no more than double what Catalyst did
moritz_ Wolfman2000: I haven't yet met anybody with spare money :-)
in terms of resource usage we're Less Than Awesome, I fear 14:57
Wolfman2000 How Less Than Awesome?
moritz_ 2 to 3 orders of magnitude in terms of run time 14:58
Wolfman2000 compared to Perl 5? 14:59
moritz_ yes
at least last I measured
since then run time improved, parsing got slower
Wolfman2000 If I remember my fuzzy math on that correctly...
If Perl 5 runs in 2 cycles, you're saying Perl 6 runs in 200 - 2000 cycles?
moritz_ yes 15:00
pmichaud good morning, #perl6
moritz_ good second morning pmichaud :-)
mathw oh hai pmichaud
Wolfman2000 ...the things I'm doing to help test an up and coming language
jnthn good less-likable part of the morning pmichaud :-) 15:01
mathw we've all done similar stuff
jnthn the early hours of the mroning are SO nicer :-)
mathw just remember, it's got to work right before it's really worth trying to make it go fast
Wolfman2000 Well, if anyone is willing to contribute to the "Help Wolfman get Perl 6 running while not breaking his other stuff by upgrading his virtual server" fund, donations are accepted.
mathw: Of course.
moritz_ Wolfman2000: I can give you access to a server 15:02
Wolfman2000: where you can host your Perl 6 things
(for free)
Wolfman2000 The Perl 6 stuff isn't ready yet. I was hoping to experiment on my own server. Still...this offer is tempting
moritz_ and that one builds parrot in less than two minutes, if properly parallelized :-) 15:03
masak mornin', pmichaud++.
Wolfman2000 /home/wolfman2000/repos/rakudo/parrot_install/bin/parrot perl6_s1.pbc --target=pir src/gen_setting.pm > src/gen_setting.pir <-- this seems to be the trouble file on the normal make process BTW
I'm guessing this one is also one big switch statement? 15:04
moritz_ yes, that's the slowest step
mathw Wolfman2000: It's the entire setting, in one file
moritz_ it's the builtins that are written in Perl 6
mathw so it's quite big
moritz_ I worked on splitting the compilation up
but it was blocking on other things
mathw stubbed functions and classes wasn't it?
masak Wolfman2000: apart from the server moritz_ is offering you, there's also feather.
moritz_ stubbed types, yes 15:05
Wolfman2000 Alright, I see what feather is about 15:06
moritz_: any fancy or simple web page to make your cause?
moritz_ feather.perl6.nl/
cognominal reads Test.pm. The is "export(:Default)" for each sub is visual noise. Is it possible to declare default properties once to avoid these nasty repetitions?
moritz_ Wolfman2000: make what cause?
cognominal: actually 'is export' is enough
Wolfman2000 www.perlfoundation.org/perl6/index.cgi?feather This is the...declaration, for lack of a better word. It encourages people to host Perl 6 stuff, as long as they give back. 15:07
I don't know how much I can give back honestly
moritz_ Wolfman2000: don't worry about that
masak Wolfman2000: what moritz_ said.
Wolfman2000 Especially since I'm in the process of changing career paths at thie moment
Trying to become a high school math/programming teacher
moritz_ Wolfman2000: if you tell masak "Web.pm is nice but $feature is a bit rough" you already gave back
masak Wolfman2000: I don't know if it shows, but we like you already. we plan to keep you. :)
Wolfman2000 As long as I'm able to make a career in doing what I need to do to put food on my table, fair enough. 15:08
masak Wolfman2000: if you don't have the tuits, that's fine. but it's often surprising how tuits keep showing up.
15:08 Exodist left
moritz_ having a constant influx of newbies who are willing to talk to us is very helpful, so that we can see what the entry hurdles are 15:08
and how they change over time 15:09
masak nod.
15:09 NorwayGeek left
Wolfman2000 The tuits? Sorry, that term is unfamiliar 15:09
moritz_ time units
"I can't get around_to_it"
Wolfman2000 I've gained time units for this week, so I'm able to pull off something. 15:10
moritz_ short tuit
Wolfman2000 ...if this big file doesn't compile itself by the time The Price Is Right ends, I may have to look into that server of yours.
masak Wolfman2000: this is a tuit economy.
Wolfman2000 Since my website is down, and therefore my resume...who needs my email?
or even...gulp...my Facebook?
moritz_ either reads faces or books, but not both at the same time 15:14
Wolfman2000: www.mail-archive.com/perl6-language...20060.html for getting a feather account 15:15
Wolfman2000 moritz_: I'm torn between feather and yours
moritz_ if you want an account on my machine, just /msg me your desired username
15:16 NorwayGeek joined
moritz_ feather is a bit more reliable, but also a bit slower and more crowded 15:16
masak crowded in a good way. :)
15:16 NorwayGeek left
moritz_ you can also use both, of course :-) 15:16
Wolfman2000 I only plan on having one website use Perl 6, not two 15:17
masak Wolfman2000: oh, so you actually plan to set up a website with Perl 6? then our paths will surely cross again soon. :) 15:21
Wolfman2000: what kind of website? 15:22
Wolfman2000 I want to turn my Catalyst website (simple resume + bonus stuff site) into Perl 6
it's right now down so I can attempt to compile Rakudo
...attempt and fail at this point it seems
masak Wolfman2000: seems to me those pages are fairly static. 15:23
Wolfman2000: i.e. you could generate the HTML pages once (using Perl 6 if you want), and then just serve them staticly.
Wolfman2000 They use routes, a common header with browser detection, and a common footer. They were written with clean HTML, CSS, JS
masak ah, so basically HTML templates, then. 15:24
Wolfman2000 Not much more than that, no.
My main website doesn't use postgresql
masak you could take a look at Hitomi.
it's a Perl 6 fork of a Python-based HTML templater.
it works already, but it's not very well-tested. 15:25
I'd love it if someone gave it a go.
Wolfman2000 If I can find the webpage for hitomi, that would be useful.
Hopefully it works with 楽度
(yeah, felt like typing rakudo in kanji) 15:26
masak :)
it works only on 楽度 as far as I know.
hold on, I'll get you an URL.
or two. 15:27
Wolfman2000 I don't even know what rakudo means properly
masak "happy earth". "paradise".
TimToady that's the wrong "do" 15:28
15:29 Psyche^ joined
TimToady that would be happy degree 15:30
Su-Shee I thought, there's more than one do to do it? ;) *scr*
masak TimToady: oh, you're right.
15:30 Patterner left, Psyche^ is now known as Patterner
TimToady I'm always right, except when I'm not (yet). 15:31
masak Wolfman2000: github.com/masak/web/blob/master/dr...xample.xml use.perl.org/~masak/journal/39137 use.perl.org/~masak/journal/39351 github.com/masak/web/raw/master/tal...9/talk.pdf
Wolfman2000 huh? there's a pl namespace? 15:32
masak Wolfman2000: there is now. :)
TimToady: you're right to a happy degree, perhaps. :P
Wolfman2000 Will that page work if sent with the text/html mime type, or does it have to be application/xhtml+xml?
Still...I think I like what I'm seeing with that first link 15:33
TimToady afk & # ~2hr
Wolfman2000 goes to check the others
masak Wolfman2000: all the transforming is done server-side, if that answers your question. 15:34
15:34 dj_goku left
masak Wolfman2000: you can output HTML. 15:34
Wolfman2000 masak: that does answer
also, I added myself to the github thing if it makes it easier to reach me 15:35
masak Wolfman2000: it does. what's the nick?
Wolfman2000 <--
masak good.
15:35 justatheory joined
masak hugme: add Wolfman2000 to proto 15:35
Wolfman2000 well, lower case w, but...still
hugme masak: github reported success, but it didn't work anyway - WTF?
masak ah.
hugme: add wolfman2000 to proto
hugme hugs wolfman2000. Welcome to proto!
masak hugme: add wolfman2000 to web 15:36
hugme masak: sorry, I don't know anything about project 'web'
masak huh.
Wolfman2000 add wolfman2000 to Web.pm
hugme: add wolfman2000 to Web.pm
hugme Wolfman2000: sorry, I don't know anything about project 'Web.pm'
masak does it manually
Wolfman2000 ...I'll bite
hugme: help
hugme Wolfman2000: (add $who to $project | list projects | show $project | hug $nickname)
jnthn hugme++ # error has WTF
Wolfman2000 hugme: list projects
hugme Wolfman2000: I know about book, json, november, nqp-rx, perl6-examples, proto, svg-matchdumper, svg-plot, temporal-flux-perl6syn, tufte
jnthn ah, it doesn't know about web. 15:37
masak fortunately, I do.
moritz_ didn't it say that in the first place? :-)
masak: I can add it, if you want
masak hugs Wolfman2000. Welcome to Web.pm!
Wolfman2000 At least I confirmed it with help
masak moritz_: please do. :)
15:38 nihiliad joined
jnthn Add lolsql! Add lolsql! ;-) 15:38
akshually don't...
Wolfman2000 I've heard of sql, mysql, and postgresql
what's lolsql?
jnthn Wolfman2000: Erm. A couple of people being silly. :-) 15:39
masak Wolfman2000: www.aaronbassett.com/2009/i-can-haz-lolsql/
mathw was tempted to write a LOLSQL parser
jnthn mathw: github.com/jnthn/lolsql
mathw: join the fun!
Wolfman2000 ...no thanks
jnthn :-) 15:40
Wolfman2000: It mostly serves as a small example of Perl 6 grammars.
Or at least, that's why I wrote the thing... 15:41
Wolfman2000 Lots of the pages you provided, masak, mention what's going on with Web.pm and Hitomi. I'm guessing there aren't a lot of example pages out there...live. That are easily viewable.
masak Wolfman2000: no. none.
Wolfman2000: it's very much an ongoing thing.
Wolfman2000: I had a Web.pm app online for a few minutes last week. 15:42
it kept segfaulting, though.
Wolfman2000 ah
so I shouldn't just flat out move my page and abandon the old one
masak no.
Wolfman2000 ...still, nothing wrong with a Perl 6 mirror I guess. 15:44
15:45 dj_goku joined
Wolfman2000 Especially since gen_setting.pm is STILL not done with its work 15:45
...I said I'd give it until noon EST
Hopefully my patience won't snap until then.
masak Wolfman2000: if you get yourself a feather account, you can compile Rakudo in your homedir. it's definitely faster than that. 15:47
Wolfman2000 ...I am starting to think my slice is incapable of it 15:49
So I have two options. feather and moritz_
More community support, more obligations with feather
masak don't think of it as obligations.
Wolfman2000 faster speed, fewer tech support with moritz_
masak Wolfman2000: I'm on feather. I even have a home page :) feather.perl6.nl/~masak/ 15:50
Wolfman2000 ...oh yeah. The other thing I forgot about shared web hosting 15:51
I have to get used to lack of server control
moritz_ I can give you a virtual host with AllowOverride All if that helps :-)
Wolfman2000 moritz_: I shouldn't need that much control. 15:52
If I'm writing in Perl, I can set up browser overrides anyway.
Either way, it will feel...different for me to go from having all control to not much.
PerlJam Wolfman2000: nearlyfreespeech.net 15:54
Wolfman2000 PerlJam: already have stuff in other languages besides PHP and Perl. Also have Python
I don't want to rewrite all of them 15:55
Plus, they all use Postgresql
thanks anyway
moritz_ moritz@ds80-237-201-115:~/rakudo$ time ( perl Configure.pl --gen-parrot && make -j 4 )
...
real 5m2.716s
that's a full rakudo rebuild after git clean -xdf :-)
much faster if you can parallelize the parrot build 15:56
Wolfman2000 I should have 2 cores at least
Either way...it's all on that big file
moritz_ yes, that's the largest bottleneck 15:57
16:00 vmbrasseur joined, hcchien joined 16:10 hicx174 joined 16:14 hicx174 left
Juerd Wolfman2000: feather accounts come without obligations 16:17
Wolfman2000 Modules and programs written in Perl 6 can also have their home on feather, if their authors work with the Perl 6 community to report and resolve bugs.
That's the obligation I'm seeing.
...wait, I'm doing that already anyway 16:18
Juerd That's everyone's default mode of operation, isn't it? 16:19
masak exactly.
Wolfman2000 Alright, point
masak Wolfman2000: there's no need to go looking for obligations.
they will find you. :)
Juerd hehe
Wolfman2000 Jason Felds. [email@hidden.address] I hope to eventually bring one of my websites to Perl 6...that, and I have some personal redemption to take care of. Even my little bits of help will go to that goal. 16:20
Juerd If that doesn't encourage you, nothing can ;)
Wolfman2000 I believe that's the information you require.
Juerd Wolfman2000: mail it please 16:21
include preferred login nawe
name
moritz_ that way it gets into Juerd's all-encompassing ticket system :-)
Juerd haha
Wolfman2000 I would need to know your email. All the page says is to request an account by e-mail to Juerd. Don't see an email address.
16:21 envi^home joined
Juerd unread=todo 16:21
16:22 nbrown04 joined
Juerd That's the system 16:22
Wolfman2000: that's the captcha/turing test of feather 16:23
Wolfman2000 Juerd: sorry, didn't fully understand that.
I'll just register myself to the perl foundation first 16:24
moritz_ Wolfman2000: you know Juerd's nickname. You know a host he owns. Try to combine that to an email address
16:24 am0c joined
Wolfman2000 ...oh 16:24
Juerd see it as a minor challenge, to find my mail address, to prove your worthiness ;) 16:25
actually feather email is probably broken
so try juerd.nl :) 16:26
Wolfman2000 well, I just sent an email to what I think is your address
16:27 nbrown04 left
Wolfman2000 ...and I failed 16:27
sbp wow, esperanto
Wolfman2000 It's not [email@hidden.address]
moritz_ Wolfman2000: $nickname at $nickname.nl
Wolfman2000 Trying again 16:28
16:28 zaphar_ps left
Wolfman2000 I never remember if email addresses are case sensitive or not 16:28
Juerd they're not
Wolfman2000 ...no mail delivery failure message yet 16:29
I think I did it right
sbp phenny: "Maar ik hoor 'm wel liever dan m'n officiële naam, die ik nauwelijks nog gebruik."?
phenny sbp: "But I hear him do better than my official name, which I hardly use." (nl to en, translate.google.com)
Juerd yay!
Wolfman2000 looks like you got it then 16:30
Juerd that's not esperanto
Wolfman2000: let's see
16:31 NorwayGeek joined, nbrown04 joined, nbrown04 left
Juerd working on my mobile phone so everything is slow 16:31
16:32 nbrown04 joined, nbrown04 left
Wolfman2000 long as the email got there, then I ddi what I was supposed to 16:32
masak phenny: "Sed mi aŭdas ke li fartas pli bone ol mia oficiala nomo, kiun mi apenaŭ uzas."?
phenny masak: The eo to en translation failed, sorry!
masak phenny: at least you tried. :)
16:36 envi^home left
Wolfman2000 alright, I'm in feather. 16:38
Juerd ssh on a phone keyboard sucks but i created Wolfman2000 an account
jnthn masak: What language? :-)
Wolfman2000 I just got in with the first name you gave me
masak jnthn: Esperanto. 16:39
jnthn masak: Ah, ok.
masak jnthn: I'm surprised phenny detected even that.
jnthn Shoulda guessed. :-)
masak didn't expect her to succeed in translating it.
Wolfman2000 ...I don't think feather likes me trying to do unicode on nano 16:40
I wanted to put in the square root symbol in for prefix testing
Juerd welcome on board, Wolfman2000++ 16:41
Wolfman2000 √ <-- when I try to put that in, it prints as ?^?^?
Juerd Wolfman2000: set LANG
ie your locale
Wolfman2000 right...how do I set that permamently again? I think I need to set it to en.utf8 or something like that
Juerd eg en_US.UTF-8 16:42
Wolfman2000 put LANG in my profile, source'd it, nano still hates me 16:43
masak export LANG
Wolfman2000 ...right, forgot that step
Juerd echo export LANG=en_US.UTF-8>>$HOME/.bashrc
found a limitation of this on screen touch keyboard: no tilde 16:44
Su-Shee not LANG. you need LC_CTYPE. 16:45
Juerd ~~~~~~ oh there it is
Wolfman2000 Seems like I need a lot
Juerd Su-Shee: LANG suffices
Wolfman2000 LANG is not sufficing for me 16:46
16:46 carlin_ joined, carlin_ left
Juerd Wolfman2000: run .bashrc after adding to it 16:46
Su-Shee Juerd: LANG sets only messages and menus and stuff like that on my system. CTYPE is what I need and if I want sorting, I need COLLATE
16:46 abra joined
Wolfman2000 Juerd: permission denied 16:47
Juerd Su-Shee: current objective is getting utf8 support. LANG fixes that
Wolfman2000: . .bashrc 16:48
Wolfman2000 I know that
I was told permission denied
Juerd not ./
Wolfman2000 ...oh
huh
Su-Shee Juerd: yes, I'm talking about the same thing and LANG is just _one_ element of possible locale settings.
Wolfman2000 Su-Shee is correct: LANG is not enough, even after running . .bashrc 16:49
Juerd sorry. i should have said 'source', not 'run'
16:49 takadonet left, abra left 16:50 krunen left
Su-Shee Wolfman2000: check first with locale -a what you actually got. 16:50
Juerd Su-Shee: I've never explicitly set any LC_ yet I get full utf8 everywhere
Wolfman2000 ...it's utf8, not UTF8
Juerd isn't ctype implied by lang?
Su-Shee no. 16:51
LANG is the "weakest" of all settings.
Juerd Wolfman2000: iirc it is UTF-8 with capitals and hyphen
16:51 ssm left
Su-Shee Wolfman2000: export LC_CTYPE=en_US.utf8 16:51
Juerd I could be mistaken of course 16:52
Wolfman2000 Juerd: according to the list, it's en_US.utf8, not en_US.UTF-8
Juerd okay
Su-Shee Juerd: depends on your system, check with locale -a
Wolfman2000 I got it workign now
Su-Shee I usally generate both versions, because who knows what needs which setting. 16:53
Wolfman2000: you should COLLATE as well and maybe NUMERIC
Juerd Weird. nl_NL.UTF-8 works for me
and setting lang apparently did also get me that ctype 16:54
Wolfman2000 ...do I have to compile Rakudo on my feather box first before using it? I just got an error while testing my √ prefix 16:55
perl6: error while loading shared libraries: libparrot.so.0.9.1: cannot open shared object file: No such file or directory
p6eval elf 28978: OUTPUT«Parse error in: /tmp/vuzzsW15ZI␤panic at line 1 column 50 (pos 50): Statement not terminated properly␤WHERE: shared libraries: libparrot.so.0.9.1: cannot open shared obj␤WHERE: /\<-- HERE␤ STD_red/prelude.rb:99:in `panic'␤ STD_red/std.rb:355:in
..`eat_t…
..pugs: OUTPUT«*** ␤ Unexpected "0"␤ expecting ".", "\187", ">>", "=", "^", operator name, qualified identifier, variable name, "...", "--", "++", "i", array subscript, hash subscript or code subscript␤ at /tmp/bNaxYhu3ya line 1, column 52␤»
..rakudo 74f561: OUTPUT«Confused at line 2, near ": libparro"␤in Main (file <unknown>, line <unknown>)␤»
Wolfman2000 ...
Juerd lol
rakudo: say 'hi'
p6eval rakudo 74f561: OUTPUT«hi␤»
Wolfman2000 This will take awhile. 16:56
~jafelds/sqrt.pl contains what I'm after. I should have it set up decently...I think.
I guess the big question...am I doing something wrong with it? 16:57
Juerd gotta go 16:58
ttyl
Wolfman2000 *nods*
moritz_ Wolfman2000: yes, you have to compile locally...
Wolfman2000: the error message indicates an ancient rakudo version
Wolfman2000 I guess I should try git and make again? 16:59
moritz_ aye 17:00
Wolfman2000 ...well, may as well reactivate my websites right now
alright, pages back up 17:02
I wonder if I should skip the make install step when I get there 17:03
17:03 h1rschnase joined
moritz_ no, you shouldn't 17:05
make install installs into $your_rakudo_dir/parrot_install/
17:05 pmurias joined
moritz_ and then you just add $your_rakudo_dir/parrot_install/bin/perl6 ot your path, and be happy 17:05
Wolfman2000 ...so it's alright to make install? Wait...I don't even know my sudo 17:06
moritz_ you don't need sudo for that 17:07
it's in your home where it will be installed
Wolfman2000 Funny, I didn't tell it where to PUT the files on make install. It will...just work?
moritz_ yes 17:08
it defaults to the same dir where your parrot_config binary is
and that's iinstalled locally by Configure --gen-parrot
17:09 mberends left 17:10 hicx174 joined
Wolfman2000 alright...let me wait for the spec test to get done first, then I'll try to contribute more math prefix and postfix stuff before I have to do real life things 17:11
I wonder if my directory is visible to others
17:15 fax joined
masak Wolfman2000: I'm in ur directory. sqrp.pl 17:15
er, sqrt.pl
Wolfman2000 I couldn't run the file due to my compiling of rakudo right now
I plan on making mini files like this right now until I feel ready to try using Web.pm 17:16
masak Wolfman2000: good plan.
Wolfman2000: it runs fine here.
√25 = 5
Wolfman2000 so we can all access each other's files
masak aye.
Wolfman2000 I will keep that in mind and don't bring in my database websites over to the server
masak :)
I wouldn't do anything production-critical on feather. or with Perl 6, for that matter. 17:17
at least not without some kind of fallback.
Wolfman2000 masak: My main websites are back up on my original host
This will allow me to experiment for awhile
masak carlin: ping 17:19
17:23 mariuz left, hicx174 left, hicx174 joined
masak is there any way to say "I don't want this method to be overridden" in Perl 6? like 'final' on methods in Java. 17:24
moritz_ there used to be a 'final' pragma or something 17:25
but all in all it doesn't really make sense
Wolfman2000 the spectest has failed, BTW
moritz_ except maybe for optimization that will fire backwards somme day 17:26
masak Wolfman2000: it sometimes does. can you nopaste the summary somewhere?
lisppaste3 wolfman2000 pasted "spectest on feather" at paste.lisp.org/display/89733
moritz_ hugme: reload data
hugme moritz_: reloaded successfully
masak moritz_: I'm pretty sure I read a good argument for 'final' on methods once. don't remember it offhand, though.
moritz_ masak: you can always bypass the 'final' by a proxy object 17:27
masak moritz_: I guess...
maybe Perl is too liberal for 'final' to make sense.
Wolfman2000: some of those look like segfaults. IO::Socket::INET is ever problematic (in the spectest suite). sinh.t I don't think I've seen before. 17:28
moritz_ sinh.t is probably a segfault or resource problem
17:28 h1rschnase left
moritz_ unicode.t is known 17:29
Wolfman2000 and make install...just worked
KyleHa I see those big trig test files bomb every so often.
Wolfman2000 just have to set the path to that and I'm good. got it
moritz_ the junction tests where changed recently, and I haven't got around to fix them up again
hugme: list projects
hugme moritz_: I know about book, hugme, json, november, nqp-rx, perl6-examples, proto, svg-matchdumper, svg-plot, temporal-flux-perl6syn, tufte, web
17:30 jaffa8 joined
Wolfman2000 yay, my program works 17:30
masak \o/
moritz_ next step: world domination 17:31
Wolfman2000 <M Bison> OF COURSE!
...sorry: like Nostalgia Critic a little much there
masak rakudo: class A { sub foo() { say "OH HAI" } }; class B { has B $!b handles <foo>; method bar { foo } }; B.new.bar 17:35
p6eval rakudo 74f561: TIMED_OUT
masak here it says "Could not find non-existent sub foo". 17:36
should that work?
17:36 xp_prg left
jnthn masak: Did you meant "has A $!b" ? 17:36
masak oops. yes. 17:37
KyleHa Yeah, what jnthn++ said.
masak same result, though.
KyleHa Wouldn't $!b have to get initialized with an A.new somewhere? 17:38
masak KyleHa: even when it isn't, it contains the A type object. 17:40
rakudo: class A { sub foo() { say "OH HAI" } }; class B { has A $!a handles <foo> = A.new; method bar { foo } }; B.new.bar 17:41
17:41 linguini joined
masak still the same result locally. no such sub. 17:41
p6eval rakudo 74f561: TIMED_OUT
moritz_ shouldn't it complain that it can't find *method* foo? 17:42
masak no, because I'm calling the sub.
my question boils down to "does 'handles' concern itself with subs?" 17:43
moritz_ in my understanding 'handles' is for methods
Wolfman2000 I think I can see why perl6 requires more power.
moritz_ masak: somehow your code doesn't make sense to me
masak moritz_: what about submethods?
Wolfman2000 Hmm...what's the proper type of "fix" for a standalone character? Something similar to the standalone +?
jnthn masak: no no no no 17:44
masak: $.foo
moritz_ masak: it just does a normal dispatch
jnthn There *is* no foo sub.
masak jnthn: there is, in A.
moritz_ but you're calling a sub foo in B
jnthn oh wtf
masak moritz_: yes. I was wondering whether 'handles' could do that.
jnthn I read masak's code as what I thought it was meant to do.
masak: No, it won't.
moritz_ masak: it can't 17:45
jnthn: me too
jnthn It shouldn't.
moritz_ ;-)
masak :)
jnthn :-)
masak Wolfman2000: I don't understand your question.
moritz_ thinks masak++ wants to import things from A, not delegate
masak moritz_: yes, how do I do that?
moritz_ masak: in Rakudo, or in Perl 6? ;-)
Wolfman2000 ½ = 0.5. I want ½ as its own operator so to speak, and have that automatically converted to 1.0 / 2
masak moritz_: in Rakudo. 17:46
Wolfman2000 I thought infix would work...probably did something wrong
masak Wolfman2000: sounds like a term to me.
moritz_ Wolfman2000: no, you'd have to define it as a constant, which only works for identifiers
Wolfman2000: or as a term, which is not yet implemented
Wolfman2000 so we haven't hit that point yet. got it
moritz_ rakudo: class A { sub foo { say 'bar' } }; class B { my &foo := &A::foo; method t { foo() } }; B.new.t 17:47
p6eval rakudo 74f561: TIMED_OUT
moritz_ doesn't work :( 17:48
Wolfman2000 alright, most of my working math pl files are in feather/jafelds/math 17:50
Not too much in there right now, but...it's a start
moritz_ masak: sub foo(*@_, *%_) { A::foo(|@_, |%_) } # SCNR
17:52 kidd` is now known as rgrau`
masak moritz_: :) 17:53
17:54 barney joined
moritz_ I'm sure you'll also find a "solution' involving eval() ;-) 17:54
masak indubitably. 17:55
moritz_ "eval -- it's Turing complete"
dukeleto moritz_: good tshirt idea :) 17:56
moritz_ dukeleto: indeed ;-) 17:57
KyleHa Every program should start with: eval q{
17:57 stephenlb joined
KyleHa There'd be no more death! 17:57
moritz_ KyleHa: or with 'use Inline::PERL;'
KyleHa: for those dubious PERL programmers out there
Wolfman2000 bah, my attempt at infixing failed
lisppaste3 wolfman2000 pasted "Infix failure." at paste.lisp.org/display/89735 17:58
moritz_ rakudo: multi sub infix:<^>(Num $x, Num $y) { $x**$y }; say "10^2 = {10^2}"
p6eval rakudo 74f561: OUTPUT«push_pmc() not implemented in class 'Sub'␤in Main (file <unknown>, line <unknown>)␤» 17:59
17:59 cdarroch joined
moritz_ Wolfman2000: presumably infix:<^> isn't overloadable yet :( 17:59
KyleHa "Your Inline::Perl program will run slowly the first few times that you run it. After that you will get used to it." Har.
moritz_ must be moved to the setting to wrok
Wolfman2000 Does ^ already have a function I'm forgetting?
moritz_ has to leave for a table tennis match, TTFN folks
Wolfman2000: it constructs a junction
rakudo: say (3^5).WHAT
TimToady an XOR junciton
p6eval rakudo 74f561: OUTPUT«Junction()␤»
TimToady *tion 18:00
Wolfman2000 rakudo: say 3^5 18:01
p6eval rakudo 74f561: OUTPUT«one(3, 5)␤»
Wolfman2000 ...unfortunately, I don't recall reading what junctions are
Hmm...I need a reminder 18:02
rakudo: say 5 / 2
p6eval rakudo 74f561: OUTPUT«2.5␤»
Wolfman2000 rakudo: say 5 // 2
p6eval rakudo 74f561: OUTPUT«5␤»
Wolfman2000 rakudo: say (//).WHAT 18:03
p6eval rakudo 74f561: OUTPUT«say requires an argument at line 2, near " (//).WHAT"␤in Main (file src/gen_setting.pm, line 2593)␤»
Wolfman2000 so // is an error?
TimToady std: say (//).WHAT
p6eval std 28978: OUTPUT«===SORRY!===␤Null regex not allowed at /tmp/juFzoLXOOq line 1:␤------> say (//⏏).WHAT␤FAILED 00:01 107m␤»
TimToady STD will usually give you a better error message
but only for syntax errors
Wolfman2000 ...I wasn't trying to use regex
TimToady well, but / where a term is expected is always a regex in Perl 18:04
Wolfman2000 Just trying to do integer division
similar to how Python does it
TimToady rakudo: say 5 div 2
p6eval rakudo 74f561: OUTPUT«2␤»
Wolfman2000 ...
nevermind
...right, I never understood that. There is a separate symbol for the minus sign 18:06
TimToady in p6 // is like ||, but tests for defined instead of true
(which got borrowed back to 5.10 to)
*too 18:07
masak: re "final", see S12:1814 18:11
18:13 xp_prg joined
frettled That's one of my favourite enhancements in P6. 18:13
Wolfman2000 I have to get into the real life now. Hopefully I can mess with more of P6 later.
18:13 akl joined
Wolfman2000 Look in my feather account to see what I've gotten in there. 18:13
masak TimToady: so finality works on a class level, not on a method level? 18:21
Wolfman2000 .away 18:22
err
masak Wolfman2000: whoa, you've been productive!
Wolfman2000 hard to tell if those mini fixes are productive or not
unsure how many of them are useful to the public
the only one that doesn't work right now is half.pl 18:23
18:23 justatheory left, rgrau left, rgrau joined
masak Wolfman2000: that's because ½ is not an infix operator. 18:24
Wolfman2000 it's a term, and terms can't be defined
yet
masak right.
Wolfman2000 I'll get this at some point 18:25
for now, *AWAY*
masak o/
TimToady std: sub term:<½> () {...}; ½ + ½
p6eval std 28978: OUTPUT«ok 00:03 124m␤»
masak someone should write an implementation that can do all the cool things STD.pm can do :) 18:27
zaslon lolpmichaudhazblogged! pmichaud++ 'A brief report on progress': use.perl.org/~pmichaud/journal/39834?from=rss
TimToady masak: yes, well, non-finality is about classes. the optimizer can do what it likes on the method level for any class that is required to be non-final 18:28
*that is not required
and most of the things you might want final methods for are really infrastructural, and therefore the proper domain of submethods 18:29
masak ah, true. 18:30
that's probably the answer to the question that I couldn't phrase.
TimToady and our classes are "open", but that doesn't mean we don't have final classes, which we call "roles" :)
pmichaud btw, jnthn, I think that the new parcel and list stuff means that multiple return value semantics fall out naturally 18:31
TimToady s/final/closed/
masak pmichaud++ # blog post
food & 18:32
18:33 masak left, r0bby left
jnthn pmichaud: Me too 18:33
TimToady doesn't seem to have shown up on planetsix yet
18:33 r0bby_ joined
jnthn pmichaud: I didn't want to make it sound like part of my grant was now a no-op though. :-P 18:33
Well, OK, extracting the design from TimToady++'s brain counts as effort I guess. :-)
pmichaud jnthn: you've more than made up for it in other are.... oh, wait, I'm the grant manager. "Yes, that's a no-op now. You have to do some additional work to make up for it." :-)
oops 18:34
you're right, you're the person who successfully extracted the needed details from TimToady. That's an accomplishment. :)
jnthn ;-)
pmichaud anyway, I think "done."
well, as soon as we demonstrate that it works :)
jnthn Yay. 18:35
\o/
18:36 TimToady sets mode: +vv hugme iblechbot, TimToady sets mode: +vvvv ilbot2 ilogger2 IRSeekBot lambdabot, TimToady sets mode: +vvvv lisppaste3 p6eval phenny pugs_svn, TimToady sets mode: +v zaslon 18:37 abra joined 18:38 abra left 18:40 NorwayGeek left
PerlJam anyone happen to know how to enable cookies in FF 3.5.4 ? 18:40
18:42 quietfanatic left, justatheory joined
awwaiid They're not enabled by default for you? 18:43
PerlJam nope
awwaiid Well some prefs are in: Edit->Preferences->Privacy-> set "use custom settings for history" -> check "accept cookies" 18:45
or you can go to "about:config", search for "cookie", and then change some settings at random. 18:46
But I just upgraded from FF 3.0 and then 3.5 -> 3.5.4, and cookies were enabled the whole way. So maybe something else is wrong :)
18:47 Morpheus` joined
PerlJam looks like a buglet. The "accept cookies" box was already checked, but until I viewed the checkbox and closed the prefs window, it wouldn't accept cookies from anywhere. 18:48
frettled There are no errors but user errors. However, it varies who the user is.
Argh. I get so frustrated with Blogger/Blogspot sometimes; I started writing a blog post yesterday evening, but didn't finish it, so I saved it for later. 18:51
I then edited it this afternoon, published it, and lo and behold: it's posted with yesterday's date. *growl* 18:52
Fortunately, I can fixify the date after the fact, but that may mess up people's RSS subscriptions.
Today's frustration was sponsored by Google. :)
18:53 jaldhar left
frettled pmichaud++ - my oh my how you've been busy! 18:56
Reading the summary, I get the distinct impression that you're at the SMOP stage. :D 18:57
18:58 am0c left
PerlJam He's not implementing anything in C and he's got much more implemented than SMOP 18:58
*rimshot* :-)
okay, it was weak and vague and only makes sense to very few people, but I'm one of those people, so it made me smile a little and that's all that matters :) 18:59
jnthn -> store, ENEEDFOOD 19:00
frettled PerlJam: hee-hee :) 19:01
19:01 frettled sets mode: +o PerlJam 19:13 h1rschnase joined, FCO left 19:14 FCO joined, h1rschnase left
cognominal plumage? 19:18
purl answered me :) 19:19
19:20 Morpheus joined 19:21 Morpheus` left 19:26 Morpheus is now known as Morpheus` 19:27 lichtkind joined 19:29 athomason joined 19:31 mberends joined 19:39 jferrero left 19:42 stephenl1 joined 19:47 NorwayGeek joined 19:52 Wolfman2000 left 20:05 stephenlb left 20:13 jferrero joined 20:20 crythias joined 20:25 barney left 20:28 frew__ joined 20:30 desertm4x joined
crythias has anyone done any Win32::Ole perl? 20:34
I just started working on it with Microsoft MapPoint... 20:35
20:36 jferrero left
mberends a colleague who does not use irc has written some Excel automation using Win32::Ole 20:37
20:38 aufrank joined
aufrank Hi all! 20:38
I just saw pmichaud++'s gist comparing NQP to JSON and had to come ask what's going on with that example 20:39
obra_ crythias: you probably #perl
pmichaud we're just looking at alternate serialization formats for Parrot's PAST output format
lots of people have been asking about a yaml-based version, we thought we might try json as a simpler first step 20:40
aufrank I see. And is a declaration of &infix:<+> really embedded within a declaration of &say? What would that even do? 20:41
pmichaud those are ast representations of calls
call &infix<+>, pass the result to &say
aufrank ah
pmichaud i.e., it's the PAST representation of say($a + 1)
aufrank thx 20:42
I've really enjoyed observing as the pace of development has increased in the past couple of weeks. You folks are doing great. 20:44
20:46 justatheory left 20:47 Su-Shee left, mberends left
pmurias pmichaud: wouldn't it be harder to serialize to a simpler structure, json can't serialise cyc 20:47
pmichaud: wouldn't it be harder to serialize to a simpler structure, json can't serialise cycStąd nk=0k(x)1 .
sorry 20:48
pmichaud pmurias: we're just playing at the moment
pmurias * serialize cyclic datastructures
pmichaud pmurias: it's not intended for anything "real"
pmurias json is yaml
pmichaud yes, true.
20:49 masak joined 20:54 frew__ left 20:56 pnate joined, mberends joined
crythias ah. yes. sorry. 20:57
masak mberends: Temporal time! :) 20:58
mberends ooh!
masak mberends: I'll do this publicly, in a branch in a Rakudo fork.
mberends the old temporal-flux one? 20:59
masak might as well.
unless it's easier to delete that one and rebranch.
21:00 vmbrasseur left
mberends dunno, git can be surprising in tasks like that 21:01
masak nod.
21:02 icwiener joined
masak moritz_: ping. 21:03
mberends is watching a BBC programme about black holes
masak mberends: they're a fascinating subject. it's easy to get sucked in. 21:04
mberends did
masak :)
I'm still a bit surprised that I understand the concept of Hawking radiation.
sjohnson Q: are \b for backspace / word boundary (in p5) separated in Perl 6? 21:05
21:05 pnate2 left
masak sjohnson: when does \b mean backspace? 21:05
sjohnson masak: according to P5 camel book 21:06
masak sjohnson: not saying you're wrong, just asking.
I've never used \b in that sense myself.
sjohnson: I don't see \b in perldoc perlop, for example. 21:07
sjohnson: belay that; it's because I'm blind.
sjohnson hands masak a cloth for his glasses ;)
masak thanks.
sjohnson: I guess the answer is the same for P6 as for P5, then: \b means backspace in strings, and word boundary in regexes. :) 21:08
either that, or the former meaning got an early retirement in Perl 6. I don't know.
rakudo: say "OH HA\bI"
p6eval rakudo 74f561: OUTPUT«OH HAI␤»
masak no, there it is.
sjohnson i believe it also means backspace in regex's too 21:11
Camel book p161
masak sjohnson: no camel book on my shelf; sorry. 21:12
sjohnson \b is under the table "Regex Metasymbols" twice :[
masak sjohnson: but you might be right -- it might mean it in char classes, for example.
just a guess.
sjohnson masak: to see the behaviour in P6, should i consult / dig thru the S## sheets?
masak sjohnson: S05. 21:13
sjohnson thanks!
masak but playing with p6eval might be just as profitable.
rakudo: say "\b" ~~ s/<[\b]>/
p6eval rakudo 74f561: OUTPUT«s/// not implemented, try .subst as workaround at line 2, near ""␤in Main (file <unknown>, line <unknown>)␤»
masak um.
rakudo: say "\b" ~~ /<[\b]>/
p6eval rakudo 74f561: OUTPUT«␤»
frettled lolitsmasak :D
masak lolitsfrettled \o/ 21:14
sjohnson masak: it's changed in p6
21:14 frettled sets mode: +ooo masak mberends sjohnson
masak ugh. doing two parallel builds of Rakudo slows down my computer. don't try this at home. 21:14
frettled masak: evidently, you need a better computer. 21:15
masak I doubt it.
frettled «Here's a nickel, kid, get yourself a better computer.» (STR)
masak :P
mberends: actually, there's no temporal-flux branch in the rakudo repo. and thinking about it, I feel like a branch in a fork might be overkill. I'll just work in master, unless I'm trying out something experimental. 21:17
mberends initially it's pimped up constructors, right? 21:18
mberends was just yanked back from the event horiz^W^W^W dozing off 21:20
21:20 Wolfman2000 joined
masak right. a perfect job for TDD, actually. :) 21:20
Wolfman2000 ...alright, you guys couldn't keep me away
frettled Dozing off is a perfect job for me. 21:21
That, and inane blogging.
mberends frettled: you made blogging an S&M art form 21:22
masak mberends: I also killed off TimeRange in temporal-flux-perl6syn.
mberends as overkill for a foundation layer? 21:23
masak as reinventing the wheel when a range of DateTime would do the trick.
and I want to express my slight dissatisfaction that the most common class in the whole hierarchy is called Gregorian::DateTime. Huffman cries in his grave.
frettled mberends: thank you, I think 21:24
masak: It looks like there is room for some improvement in naming, yeah 21:25
mberends masak: Huffman would have called it Cal or something like that
frettled sorry for not being overly involved in this work, life's been way too busy, but at least you guys get useful stuff done :)
masak mberends: I've proposed Time before. even DateTime would be fine by me.
frettled I like Time because it's short, to the point, and covers it all. mberends likes DateTime for other reasons. ;) 21:26
masak mberends: having the Gregorian:: prefix there feels very correct, in the sense the Java time classes are correct.
21:26 justatheory joined, justatheory left
mberends it's understandable to avoid being Gregorian-centric, when StarDate might be adopted as the standard during the lifetime of Perl 6 21:27
frettled Oooh, cool feature on a wish list near you: automatic class expansion if there are no name space collisions.
masak mberends: yes, but must avoiding being Gregorian-centric amount to being long-winded?
frettled: enums already do that, in a sense. 21:28
mberends no, a brief label would be so much better
21:28 justatheory joined
frettled Example: Let's say we have the two modules Gregorian::DateTime and Granny::Weatherwax. Because there are no other modules with similar names, you can simplify to Gr::DT and Gr::W 21:28
masak frettled: as long as it doesn't lead to some horrible action-at-a-distance, I would be fine with it.
oh, belay that. I just read your clarification. 21:29
frettled Ideally, Gre::DT and Gra::W
masak no please no.
frettled >:->
masak frettled: you'll just have to do it as a module like everybody else. :P
mberends the 'use' spec (S11) allows for abbreviated aliases
frettled My goodness, Perl 6 is becoming the EMACS of programming languages! ;) 21:30
In other words: every feature you ever think of is already there. 21:31
Wolfman2000 frettled: I can't override infix operations yet. :(
21:31 IllvilJa left
frettled Wolfman2000: it's probably on the ToDo list, then. 21:32
21:33 stephenlb joined
mberends frettled: What's the date? Some date without a time. What's the time? Some time without a date. It's intuitive to keep those separate. 21:34
frettled mberends: except that «time» is the concept of, well, time.
mberends: «time of day» is the one that's without specific meaning without a date 21:35
«Time» is the all-encompassing thing, the fourth dimension in spacetime.
mberends frettled: sure, but what we commonly call stuff is as simple as possible 21:36
frettled And no, I don't think we need a SpaceTime class, although that would be kindof cool.
mberends: yes, and using «Time» would make it simpler than «DateTime»
Time of day, if that should be subclassed somehow, could be e.g. Time::TimeOfDay, Time::TOD, or somesuch. 21:37
mberends that needs more Huffman contraction, being very frequently used 21:38
21:38 NorwayGeek left
frettled Then perhaps it shouldn't be subclassed, but rather be methods in the Time class. Subclassing only for specific features which are less frequently used. 21:39
mberends right. keep it simple as far as possible 21:40
frettled If there is a strftime method, for instance, that clearly belongs to the parent class.
masak here's how Ruby does it. ruby-doc.org/core/classes/Time.html
discuss.
21:40 desertm4x left
frettled Well, from a brief glance, Ruby's approach seems sensible enough. 21:41
However, I don't see anything about localization. 21:42
masak frettled: ZoneOffset?
frettled Nopes.
Localization also implies translations of month names, day names, abbreviations, standard formatting etc.
masak ah. 21:43
mberends strftime() was always my favourite. duct tape :)
frettled Leaning on the operating system's libraries may be a very good idea.
mberends: yeah, I like that one, too.
arnsholt frettled: Might result in the same function differing across platforms though 21:44
21:44 hudnix joined
masak my plan is to use one day a week during November to put together a temporal-flux proposal that p6l can't affort to refuse. 21:44
frettled arnsholt: Yes, but that's life.
arnsholt: There's always going to be a lot of things that are platform specific. I don't know which is worse; implementing localization yourself, or leaning on the OS's implementation. 21:45
moritz_ re
masak: pong
21:45 stephenl1 left
masak moritz_: oh hai. I accidentally reviewed your last Rakudo commit. it doesn't work. 21:45
frettled heh 21:46
masak rakudo: my $c = Complex.new(1,2); say $c.perl
p6eval rakudo 74f561: OUTPUT«Complex.new(1.perl, 2.perl)␤»
frettled «accidental review» - that's a wonderful way of putting it.
moritz_ dumb
masak moritz_: it would work in ideal Perl 6.
Wolfman2000 ...complex numbers? wow
masak moritz_: did you perchance not write tests for this? :P 21:47
*SCNR*
moritz_ will fix it later
masak moritz_++ 21:48
21:48 hudnix left 21:49 xenoterracide left, xenoterracide joined
pmurias frettled: automatic generation of abbreviations would lead to new modules making old code ambigious 21:50
frettled pmurias: yep
masak jnthn: if a where clause in a param list is a regex match, does $/ get set so that it's accessible from within the routine? if not, is there any way to do that? 21:52
jnthn heh
masak: I dunno off hand. I guess you can try it. :-)
It may well work. 21:53
masak oh, you can be sure I'll try. :)
21:53 hudnix joined
jnthn Since the where clause is invoked relative to the context of the routine you're in. 21:53
moritz_ masak: even in perfect Perl 6 my last patch is wrong 21:54
+ "Complex.new($.re.perl, $.im.perl)";
doesn't end in a postcircumfix
so only the $.re is interpolated
masak moritz_: oh, you're right.
jnthn: sounds promising. :)
jnthn rakudo: sub f($x where /\d+/) { say $/ }; f("abc123def") 21:55
p6eval rakudo 74f561: OUTPUT«Use of uninitialized value␤␤»
jnthn bah!
fail!
sjohnson rakudo: use justdoit; 21:56
p6eval rakudo 74f561: OUTPUT«Can't find ./justdoit in @*INC␤in Main (file <unknown>, line <unknown>)␤»
masak jnthn: could you... make it happen? :)
frettled Make it so.
masak and then we'll spec it! \o/
jnthn masak: It may turn out that it just fixes itself in ng anyway. Not sure.
moritz_ it's already specced
jnthn I'll keep it in mind.
I think it probably should work anyway. 21:57
masak \o/
moritz_ if the RHS is not a block, it's parsed as the RHS of a smart match
masak moritz_: yes, but...
moritz_: that doesn't automatically mean that $/ is propagated into the routine, right?
21:58 FCO left
jnthn oh hmm, that's true. 21:58
masak just got the error "set_pmc() not implemented in class 'default'" from my fork'd Rakudo.
jnthn Sounds very fork'd to me. ;-) 21:59
masak wonder what I did? :)
moritz_ masak: not sure which scope it is in
masak: or what happens if you have multiple such things
masak moritz_: the last one, in case of multimples.
provided this makes sense at all.
moritz_: parameters are evaluated LTR anyway. 22:00
moritz_ masak: yes, but binding time != evaluation time 22:01
jnthn Yes, order of parameter binding is LTR.
moritz_ masak: and where clauses are executed at binding time
doesn't calling positionals by name change that?
masak I was worried something like that might spoil the fun. :) 22:02
oh well.
moritz_ doesn't have to be a show-stopper
jnthn moritz_: Calling positionals by name does *not* change the order in which the parameters are bound. 22:04
22:04 jaffa8 left
jnthn rakudo: sub foo($x where { say 1 }, $y where { say 2 }) { }; foo(1,2); foo(x => 1, y => 2); foo(y => 2, x => 1); 22:04
moritz_ so :($a, $b) := \(3, a => 4) binds $b first, and then $a?
p6eval rakudo 74f561: OUTPUT«1␤2␤1␤2␤1␤2␤»
jnthn moritz_: No. 22:05
moritz_ wow
jnthn moritz_: It's *always* bound in the order the parameters are declared in the signature.
moritz_ genuinely suprised
jnthn Yes, that was one of the fun parts of writing the binder. ;-)
moritz_ do we have tests for that? ;-)
jnthn It's actually not a big deal to do really though.
moritz_: Not yet. ;-)
It is spec though.
It's a line of spec that caused me to toss my first binding algorithm when I was designing the new binder. :-) 22:06
moritz_ is it also spec that constraints have to be evaluated as seldom as possible?
if not, it'll make testing with side effects in the constraints much harder 22:07
jnthn moritz_: Just make a constraint that requires the compiler to solve the halting problem in the test. ;-) 22:08
I think not doing them would count as an optimization.
But it'll be a hard opt to do.
moritz_ I think I know a decent way to test it 22:09
jnthn For any non-trivial case.
:-)
masak I agree with jnthn. it feels strange to spec 'as seldom as possible'.
better to leave it up to the implementations.
moritz_ I don't need that constraint for valid testing 22:10
masak I'm getting strange errors. 22:12
jnthn "Monkey eating beach ball can not handle banana."? 22:13
masak ok, slightly strange errors. 22:14
masak slowly inches away from jnthn
jnthn At least now your error doesn't feel so strange any more.
masak true.
first, it was that set_pmc() error. 22:15
now, it's "Can only transform an Object to p6opaque"
jnthn Ouch. 22:16
Those are gutsy errors.
masak: How're you achieving them?
moritz_ I've seen that with outdated files lying around in the build or install dir
masak jnthn: I don't know yet.
jnthn The only time you'll run into "default" normally is if some dynpmc type name looking went pear-shaped, which really should not happen. 22:17
pugs_svn r28979 | moritz++ | [t/spec] binding happens in parameter order
moritz_ regexes ftw
this test is a bit permissive, but it doesn't blow up if the dispatcher does more checks than necessary 22:18
Wolfman2000 Hmm...alright, feather question. Where exactly is pugs/docs/feather/~?
moritz_ Wolfman2000: where did you get that reference form?
s/form/from/
Wolfman2000 feather.perl6.nl/~/
moritz_ might refer to the pugs svn repo
22:19 crythias left
moritz_ Wolfman2000: if you put something in the public_html folder in the home dir, it'll show up on feather.perl6.nl/~wolfman2000/ 22:19
22:19 Whiteknight joined
Wolfman2000 moritz_: actually, ~jafelds 22:20
it does show
but I want to update the listing
...and yeah, looks like I need pugs as well
alright, what's the proper way to get pugs?
moritz_ there's an INSTALL file in the pugs repo that describes that 22:22
ah
the ~ dir in the pugs repo
how surprising ;-)
in pugs/docs/feather/~/
22:23 Morpheus` left 22:27 tak11 joined
frettled moritz_: originally there was a nothing dir, but then Emacs auto-saved a backup copy ;) 22:30
Wolfman2000 and...I don't have any $PAGER variables, and thus can't perl Makefile.PL pugs 22:31
frettled hrm, that's buggyish
moritz_ Wolfman2000: you shouldn't be doing that anyway 22:32
jnthn frettled: Sometimes Emacs is nearly as buggy an OS as Windows, eh? ;-)
Wolfman2000 so someone else will have to update the website listings to see my username
frettled jnthn: Yeah, but with fewer remove execution vulnerabilities ;)
moritz_ Wolfman2000: if you want pugs installed, you need to follow the "INSTALLING PUGS VIA CABAL" instructions in the INSTALL file 22:33
masak here's my progress for the evening: github.com/masak/rakudo/blob/master/temporal.t github.com/masak/rakudo/blob/master...emporal.pm 22:45
all tests pass.
now blogging about it.
moritz_ do we have a book meeting tomorrow? 22:46
masak yes, I think so.
but I'll be unable to attend. :/
moritz_ ok
jnthn oh book...I forgot to write this week :-( 22:47
jnthn <- fail
moritz_ you wrote code that week
that's fine too
jnthn Yeah, and the week before...
:-/ 22:48
moritz_ so we can have something to write about ;-)
anyway, I'd also welcome reviews of what I wrote over the weekend
frettled masak: where's your yak? 22:50
masak frettled: sorry? 22:51
frettled masak: I have a razor, but I see no yak :(
masak frettled: is this a very contrived way of saying you want to contribute?
very well. first read this: github.com/masak/temporal-flux-perl...mporal.pod 22:52
then write tests. kthxbai.
the interesting things start at line 209.
frettled masak: No, it's a very contrived way of saying that it looks fine enough that there is no need for yak shaving
masak frettled: thanks. but there's a lot of redundancy in there. 22:53
frettled masak: yes, a bit
But at the moment, I don't really see how you can easily get around that. 22:55
22:55 nihiliad left
masak me neither. 23:00
frettled But if I for some reason should have an epiphany, I'll be sure to let you know, perhaps even in the form of a patch. :) 23:01
masak excellent. I'd expect nothing less. :) 23:02
23:03 KyleHa left
zaslon lolmasakhazblogged! masak++ 'November 3 2009 -- doing it with style and sophistication': use.perl.org/~masak/journal/39836?from=rss 23:03
frettled Wotzthiszaslonthing? 23:05
masak it's a bot written in Perl 6. carlin++ 23:06
diakopter zaslon: hi
zaslon Sorry, I don't understand that command
moritz_ zaslon: help 23:07
zaslon Sorry, I don't understand that command
moritz_ zaslon: you're repeating yourself
zaslon Sorry, I don't understand that command
diakopter zaslon:
zaslon Sorry, I don't understand that command
masak zaslon: Sorry, you don't understand that command? 23:09
zaslon Sorry, I don't understand that command
masak thought so 23:10
23:11 pmurias left 23:13 iblechbot left
fax Sorry, I don't understand that command 23:13
jnthn fax: get me a beer
...bah, useless bot. 23:14
23:14 dj_goku left
moritz_ hugme: hug jnthn 23:14
hugme hugs jnthn
jnthn hugme++
frettled maybe it can fax you a beer drawing
masak 'night, #perl6.
23:14 masak left
jnthn night masak++ 23:14
moritz_ 'night
frettled yep, nighty-night!
jnthn frettled: That'd be an amazing feat, given I don't actually have a fax machine. :-)
frettled jnthn: what century are you living in, anyway? :) 23:15
jnthn 21st. Fax machines were so 20th century. ;-)
moritz_ does anybody know the URL to the svn repo that holds the perl.org website? 23:19
TimToady I used to edit the specs via svn.perl.org/perl6/doc/trunk/design but I suspect it was all indirect via combust 23:24
frettled Not via the pugs repo? 23:25
TimToady not till we moved the specs there
frettled D'oh :D
moritz_ it seems svn.perl.org/perl.org/ is the correct URL, but it won't let me in
TimToady std: D'oh:D 23:26
p6eval std 28979: OUTPUT«Undeclared name:␤ D'oh used at line 1␤ok 00:01 104m␤»
moritz_ not even with the svn.perl.org account info
23:27 ronny_ joined
diakopter std: D'oh :D 23:27
TimToady looks like a new cert as of August
p6eval std 28979: OUTPUT«Undeclared name:␤ D'oh used at line 1␤ok 00:02 103m␤»
ronny_ hi
TimToady hmm, why isn't that two terms in a row
but it works for me after I allow the new cert 23:28
diakopter std: :D :D :D
moritz_ because D'oh is a listop, expecting a term?
p6eval std 28979: OUTPUT«ok 00:02 102m␤»
TimToady oh, it's parsed as a sub call, D'oh
moritz_ hi ronny_
TimToady std: class D'oh {...}; D'oh:D 23:29
p6eval std 28979: OUTPUT«ok 00:02 105m␤»
TimToady std: class D'oh {...}; D'oh :D
p6eval std 28979: OUTPUT«ok 00:02 105m␤»
moritz_ ah, I only have access to the live/ subdir
diakopter std: :D:D:D:D:D:D:D
p6eval std 28979: OUTPUT«ok 00:01 102m␤»
frettled ooohkaaay.
TimToady okay, :D in adverbial position
23:30 tak_ joined
ronny_ whats the current state of the different perl6 implementations? didnt find any information on completeness of the language and other interesting things like bindings 23:30
23:30 justatheory left 23:31 tak_ left
moritz_ rakudo.org/status 23:31
that's about rakudo at least
Tene ronny_: that looks like several questions...
moritz_ pugs is mostly sleeping for several years now
diakopter std: (::C:(Int))
moritz_ kp6 and elf also look rather dormant
p6eval std 28979: OUTPUT«ok 00:02 110m␤»
23:32 Whiteknight left, tak_ joined
moritz_ mildew and sprixel are being developed actively 23:32
but both aren't very usable yet
TimToady ooh, I have an idea!
it might even be a good idea
jnthn uh-oh :D
diakopter takes responsibility in that case
TimToady we have bare infixes, and infixes in [op]
but there's no special reason to say $a [eqv] $b 23:33
what if that were special, and set $! on failure?
moritz_ set $! to what?
23:33 dakkar left
TimToady whatever you want for testing 23:33
moritz_ to the thing that fail() retunredß
s/ß/?/
ronny_ hmm, yes, so rakudo is the one to beet on 23:34
TimToady ok $a [eqv] $b, 'message including $!';
diakopter zaslon: beet on rakudo
zaslon Sorry, I don't understand that command
diakopter ronny_: yep, these days.. 23:35
TimToady that is, [op] where infix is expected would capture its arguments sufficiently well to implement is, is_deeply, etc
the parser would just do that for you somehow
ronny_ diakopter: i suppose i'll grab it and test if i might like the actual language better than python 23:36
TimToady or set some other variable than $! maybe
moritz_ you can like and use more programming languages than just one ;-)
frettled TimToady: perhaps set any variable by programmer's choice? *g,d&r* 23:37
ronny_ moritz_: yeah, but its certainly distracting for single projects to use more than one, (sometimes 2 is acceptable tho)
TimToady and for parsers that aren't up to implementing [op] that way, fudge could easily translate to some other form
moritz_ ronny_: yes
diakopter std: say \|\|\|&eval 23:38
p6eval std 28979: OUTPUT«ok 00:03 130m␤»
moritz_ TimToady: so you want to diag($!) on failed tests?
or something similar?
ronny_ moritz_: and since im unhappy with pythons library situation im taking a look at other things every now and then 23:39
TimToady the [op] form would just capture all the data that we wish we could with "ok $a op $b" now, but can't
moritz_ ronny_: well, Perl 6's library situation isn't bright either at the moment
ronny_: though we're working on improving that ;-)
ronny_ moritz_: well, i suspect that will change
TimToady and it would be useful for run-time testing of assertions, as well
moritz_ I'm sure it will
TimToady: sounds useful 23:40
diakopter std: $%
moritz_ maybe that also answers the questions how error messages propagate from exceptions
p6eval std 28979: OUTPUT«===SORRY!===␤Obsolete use of $% variable; in Perl 6 please use Form module instead at /tmp/ea0lcjGPdj line 1:␤------> $%⏏<EOL>␤FAILED 00:01 103m␤»
diakopter std: ::Form 23:41
p6eval std 28979: OUTPUT«ok 00:01 102m␤»
ronny_ moritz_: perl5 already has shown the emphasis on good libs in perl, i just never liked the language itself, perl6 seems to be more close to my liking as a language
moritz_ then 'foo' [~~] /f ~ f o*/ would set $! to "Unable to match regex, can't find final 'f' at $thing"
frettled hmm, interesting
diakopter std: \|\|\|\|\|() 23:44
p6eval std 28979: OUTPUT«ok 00:01 110m␤»
23:44 tak11 left
TimToady well, maybe not that much, but certainly 'expected /f ~ f o*/ but got "foo"' or some such 23:45
diakopter std: \|/\|/|/\|\|\//|()
23:45 justatheory joined
p6eval std 28979: OUTPUT«ok 00:03 110m␤» 23:45
moritz_ I'd expect ~ to explicitly tell me what it was looking for 23:46
TimToady it could be made to happen, if individual [op] checkers can work differently, but I was looking for something that would work with any relational 23:47
diakopter std: ::foo [[[eqv]]] ::bar
p6eval std 28979: OUTPUT«ok 00:02 104m␤»
TimToady and that could easily be fit into fudge for implementations that aren't up to treating [op] specially yet 23:49
I guess if we want the failure mode of // we need $/ to contain interesting values of false 23:50
diakopter std: ::foo ***** ::bar
p6eval std 28979: OUTPUT«ok 00:01 104m␤»
TimToady so maybe [op] sets $/ when it fails
$/ could be false, while $<left> and $<right> could interesting 23:51
or $<lastpanic> or whatever 23:52
ltext op rtext larg rarg errmsg 23:53
diakopter std: $<> 23:54
p6eval std 28979: OUTPUT«ok 00:01 104m␤»
TimToady std: $<><><><><><><> # diamonds are forever 23:55
23:55 rjbs joined
p6eval std 28979: OUTPUT«ok 00:02 104m␤» 23:55
TimToady more like
rjbs First off, thanks, #perl6, for introducing me to U+2424. What an awesome glyph.
TimToady std: $$$$$<><><>
p6eval std 28979: OUTPUT«ok 00:02 104m␤»
TimToady bows 23:56
moritz_ rjbs: ;-)
23:56 icwiener left
dukeleto TimToady: wow. 23:56
rjbs Nextly, I was wondering if I would be extremely foolish to hope that we'll have persistent lexicals in the repl for R*.
moritz_ rjbs: not foolish, it's on the ROADMAP
jnthn rjbs: We very much want them. 23:57
rjbs: Highly likely we'll have them for R*.
rjbs That's great. It's been a sort of annoying stumbling block for just messing about.
TimToady there's probably some way to fake it now with nested evals of closures :)
jnthn Yeah, fully agree.
TimToady: Yes yes, it'll just involving some cheating and lies. :-)
rjbs TimToady: If I wanted to cry in my beer, I'd stay in Perl 5. ;) 23:58
Hm. I have no small beer at home. Curses!
23:58 nihiliad joined
jnthn rjbs: Whatever's wrong with the big beer? ;-) 23:58
TimToady one large works as well as several small
rjbs no...
23:58 frew__ joined
rjbs I have one 12 flox (ISB) of beer at home. 23:58
IT is 20 percent alcohol by volume. 23:59
23:59 Gruber is now known as Grrrr
jnthn Whoa. 23:59
moritz_ even Starkbier usually has less
jnthn That's some serious beer.
TimToady That's no moon! That's a space station!!!
rjbs This is the "famous" Dogfish Head 120 Minute Pale Ale.
I'm going to split it with a friend, I think.

Powered by Cro and the Raku® Programming Language.

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

Thank you!