svn switch --relocate svn.openfoundry.org/pugs svn.pugscode.org/pugs/ | run.pugscode.org | spec.pugscode.org | paste: sial.org/pbot/perl6 | pugs.blogs.com
Set by avar on 16 November 2006.
Juerd :| 00:00
Caelum what about Parrot? how's that coming along? 00:02
[particle] caelum: hey, you're named after my wireless network!
parrot's coming along well. we recently had a release, and both design and implementation are ongoing 00:03
Caelum cool :) 00:19
obra Juerd: audrey is moving house. 00:27
It sounds like things will start clicking along soon
anatolyv Sounds nice.
obra "It's almost christmas" 00:30
Juerd obra: I think Audrey's really great, but the truck number is dangerously low... 00:33
audreyt I thought it's bus 00:34
lambdabot audreyt: You have 5 new messages. '/msg lambdabot @messages' to read them.
audreyt as in bus error
Juerd audreyt: Same thing.
obra Busses. Definitely busses
I see no "Truck Error" in my OS ;)
Juerd Could be trains. Let's hope not.
obra ponders Karenina Failure 00:35
anatolyv "Karenina interrupt", perhaps. 00:36
Juerd goes to bed 00:37
Good night or whichever localtime, everyone
anatolyv good night.
Juerd Watch out for busses, trucks, and trains. And all other big moving things.
z 00:38
audreyt packets!
anatolyv electrons
obra "that would have to be a big packet"
[particle] perks up 00:46
pmichaud I have a question about t/01-sanity/07-tailcall.t
07-tailcall.t has the line: &baz.goto("param1", "param2");
[particle] deja vu
audreyt [particle]: heya. tree transformation with MP6.
pmichaud it's pretty obvious what this does, but is the 'goto' method defined anywhere in the Synopses?
[particle] audreyt: yes! 00:47
audreyt pmichaud: I believe not; &baz.call(\("param1", "param2")) is the specced form though it doesn't explicitly tailcall 00:48
[particle] pm: i think we've concluded the answer is 'no'. the question becomes, is this syntax legal?
pmichaud well, the syntax is legal, it's just a question of whether the method exists or not :-) 00:49
audreyt pmichaud: you can safely replace it with a .call and a \() capture literal
pmichaud audreyt: what should 07-tailcall.t be testing, then? i.e., what sanity is it checking?
audreyt or just do &baz("param1", "param2")
audreyt oh wait, that can't do 00:50
[particle] do you need capture syntax for Test.pm?
audreyt [particle]: no
pmichaud: it's testing tailcall optimization
the idea of .goto is that it's replacing perl5's
goto &foo
[particle] well then, i don't see a need for it in sanity tests
audreyt which we need a replacement anyway
pmichaud would we lose anything by taking it out of 01-sanity?
audreyt nope. please rm ahead.
[particle] audreyt: goto &$AUTOLOAD is referenced in S10, so if the syntax changes that needs update
audreyt or mv, as the case may be 00:51
Limbic_Region mv to unspecced ?
pmichaud audreyt: excellent, many thanks
audreyt [particle]: not at all.
[particle]: look at context again
""Perl 5's goto &$AUTOLOAD..."
[particle] ah, well that's what i get for not paying attention
audreyt so, tree transformation. 00:52
[particle] hopes pmichaud sticks around
audreyt relevant chapters: S06:1177
pmichaud I'll stick around if I can -- family is interrupting at times :-|
audreyt all the way till :1211 or so 00:53
pmichaud but I'll keep the log open :-)
audreyt that is the matching part. the construction part in MP6 is
::Tree(left => ..., right => ...)
fglock advocates that we use explicit :$left in matching part
sub f (::Tree(:$left, :$right)) {...} 00:54
which has this good property that construction and deconstruction looks the same.
cf. ML and its descendents
which is ideal (designed for) tree transformations, I hope
any particular questions about the syntax, semantics or pragmatics?
[particle] as an aside, i think the syns need s/\t/ /g translation 00:55
pmichaud (kids are calling, bbl)
audreyt I'll be offline in 1hr too
audreyt [particle]: I'll do that. 00:56
[particle] reads
[particle] wonders if there's an inverse of dor/err 01:00
$<k> !// say "k = $<k>";
...that looks *ugly*
audreyt uhm 01:02
$<k>.defined and say "..."
or if you are reasonably sure $<k> is an object
$<k> && say "..."
or you want a \\ operator? 01:03
[particle] (phone) 01:04
TreyHarris wow, saw a 79K synopsis diff land in my inbox and was wondering if Audrey had decided Perl 6's syntax was waaaay too stable until i opened the mail :) 01:09
eviltwin_b heh
audreyt :D
eviltwin_b tends to assume, from years of experience, that a diff that big is a reformatting diff 01:10
TreyHarris audreyt: i don't have a feather account, how do i get one again?
audreyt TreyHarris: mail your realname to juerd and desired account name 01:10
TreyHarris: or we can try dcc...
TreyHarris audreyt: ok, i'm uploading the file to this machine, but it's slow going, it's DV quality video 01:11
audreyt >1G? 01:12
TreyHarris no, 400M about 01:13
audreyt oh. that's fine 01:14
diotalevi So does anyone here have an idea how I can type << in emacs and have « inserted instead? There's an abbrev-mode but it appears to only work if the shortcut is word-like which << isn't. 01:15
Chewie[] I'd almost just do something like A<< to trigger the insertion. 01:16
unless that fails too.
eviltwin_b take a look at iso-accents-mode
you can define your own accents, IIRC 01:17
Tene diotalevi: in vim, I have << mapped to «. I also have my caps lock key set to act as 'compose', so I can hit caps, <, < anywhere and have « inserted 01:19
setting a compose key is probably the most widely-useful solution
diotalevi I don't know what a "compose key" is. 01:20
Tene diotalevi: what Desktop Environment are you using?
diotalevi Windows XP, and Gnome. *Currently* Gnome. 01:21
Tene in gnome, system->preferences->keyboard 01:22
'Layout Options' tab, 'Compose key position' section
eviltwin_b with stock iso-accents-mode ~< is « and ~> is »
diotalevi Aha, thanks eviltwin_b. Strangely, it displays like {<<}, not the « I expected. 01:23
eviltwin_b that suggests it doesn't know how to display «. you may need to tweak something somewhere but I'm not sure how to do it for fsfemacs (I use xemacs) 01:24
(and unfortunately xemacs and fsfemacs have diverged quite a bit) 01:25
diotalevi Like I said, it's odd. I can do ucs-insert to say « and that works fine. It displays nicely and everything.
eviltwin_b hm. it works as expected here --- but again, xemacs. and probably with some customization I've forgotten but could dig up 01:27
audreyt pmichaud: sorry for the timeout - I need to run now :/ 01:37
pmichaud audreyt: no problem -- I was afk anyway 01:39
we'll catch up another time :-)
audreyt k.
pmichaud I'll do more background reading in the meantime 01:40
diotalevi I just found that the iso-accent-mode thing I was pointed a before is now considered obsolete. Current state of the art is to switch into an "input method" mode where typing ~< and ~> becomes « and » respectively. 01:59
diotalevi That's the latin-1 input method anyway. 01:59
Or.... with the utf-8 mode then &<< becomes «. 02:00
Hmm.
eviltwin_b not surprising. emacs' handling of non-ascii is still evolving 02:01
unfortunately, it also still sucks pretty badly :>
diotalevi It says it's following rfc1345. 02:02
wolverian shift+altgr+8 inserts « here. :) 02:03
diotalevi At least there's an RFC, I guess. So hopefully not just completely random. 02:05
eviltwin_b option-backslash here :>
gaal diotalevi: see docs/quickref/unicode in the pugs tree 05:46
diotalevi_ Are there any unicode operators other than «, », and ¥? Why is "zip" using the yen symbol anyway? It seems weird that it's using a currency symbol. 06:26
Tene well... « is "much less than" in mathematics. 06:28
just as < is "less than". 06:29
So is it weird to use a mathematical comparison operator for quoting lists?
TimToady actually, that's a separate character in Unicode: ≪
Tene Hmm. 06:30
TimToady ¥ is used only for its picture of a zipper.
Tene What is « then?
TimToady french/german quote
Tene Okay then, is it weird to use quotes for -«(list) ? 06:31
TimToady no weirder than using $ for scalars 06:32
Tene ... yes, that's a very good comparison. Much clearer than mine.
diotalevi_ Thanks all y'all for those explanations of Yen, etc. 06:47
Vspirit from the perl6 "So Perl6 will be exactly like every other compiled language: it will be possible to protect your source code, but only from those who don't know how to run a disassembler." and I'm wondering, to what degree, will disassembling a compiled perl program be easier because of parrot, than say disassembling a C program ? 08:53
anatolyv Vspirit: perhaps about as easy as disassembling a compiled Java program. 09:02
Vspirit anatolyv: yeah my guess is the same, but thats just it, only a logical guess, would be nice to know if anyone have had the skill and time to put in some proper research to clarify on subject, as that would make it an easier sell to the powers that be if confirmed. 09:04
araujo @tell Limbic_Region Thanks! 09:20
lambdabot Consider it noted.
gaal clkao: ping 09:47
clkao gaal: pong 09:59
svnbot6 r14849 | andara++ | [runpugs] 10:13
r14849 | andara++ | Proper closing of slave filehandles. Should stop running out of them.
r14850 | andara++ | [runpugs] 10:16
r14850 | andara++ | -New user interface with continuous screen presence and other goodies -- "Ajax du pauvre"
r14850 | andara++ | Try it out at feather.perl6.nl:8080/runpugs/
lambdabot Title: Run Perl 6 Now -- in your browser!
andara ?seen anatolyv 10:22
lambdabot anatolyv is in #perl6. I last heard anatolyv speak 1h 19m 23s ago.
Juerd andara: It does weird things when I try to input an ampersand (&) 10:33
andara: And I think development version may be a better default
andara: I'm amazed at how this thing works. This ajaxified version is really much better :) 10:34
I don't seem to be able to input an open paren
Ah, & behaves like the up arrow 10:35
And ( like down
andara Juerd: oops, that sounds bad. I catch the arrow keys, but your keyboard must have a different mapping or something 10:38
gaal andara++ 10:39
this is teriffic
andara Juerd: the ajaxy version is 'devel' as I haven't tested it much. Seems to work on FF on linux, Camino on Mac and IE on windows
gaal feather.perl6.nl/%7Eandara/runpugs/ contains contradictory claims about command history :-)
lambdabot Title: Run Perl 6 Now -- in your browser!
andara ha, a doc bug as well :-) 10:40
gaal: does command history work for you?
gaal yes
ff/windows
andara cool 10:41
gaal :q should probably not start a new sessino immediately though
Juerd andara: My keyboard uses dvorak mapping, but that's exactly the same as US qwerty for & and (
andara Juerd: what happens when you use the arrows? 10:42
Juerd andara: Works perlfectly
gaal btw, the ajax version can probably be moosed to close the pugs when the browser window is closed
andara catching "window close" events is tricky. Anyway, that's not a big issue as the server cleans up every now and then 10:45
andara Juerd: I can reproduce your bug. I'll try to figure out what causes it 10:47
andara Juerd: I fixed it for FF, but by breaking it for IE/win. FF/Mac is not quite right too. *sigh* 11:12
have to go to $work now. bbl. 11:15
Richar1 Hello 11:29
svnbot6 r14851 | andara++ | [runpugs] changed devel version URL 15:07
svnbot6 r14852 | andara++ | [runpugs] 15:08
r14852 | andara++ | "ajaxified" runpugs:
r14852 | andara++ | -fixed ( & -disabling JS bug
r14852 | andara++ | -better cross-platform JS support (command history, block backspace on prompt)
r14852 | andara++ | FF/Camino on OS X still have a few quirks, Safari is quite broken.
pmichaud another pugs 01-sanity test question 16:26
07-ref.t has a '.ref' method on items, but afaict '.ref' is only mentioned once in A04 16:27
fglock pmichaud: .ref is obsolete afaik, may use .WHAT instead 16:28
pmichaud does the pugs test need .ref?
fglock Test.pm seems to use .WHAT only 16:30
pmichaud ah, good
so I just need to change 07-ref.t
or update it
many thanks
fglock np 16:32
Limbic_Region araujo - did you get my @tell msg from yesterday 16:46
lambdabot Limbic_Region: You have 1 new message. '/msg lambdabot @messages' to read it.
Limbic_Region @messages
lambdabot araujo said 7h 25m 44s ago: Thanks!
Limbic_Region ah
good
araujo Limbic_Region, yeah :-) 16:52
Limbic_Region araujo - so after you got the extra-libs everything just worked right? 16:53
araujo Limbic_Region, next to test really 16:57
bit a busy :-(
but i think that should do it...
Limbic_Region araujo - there is likely 1 test that will hang on your smoke 17:04
t/perl5/modify_inside_p5_p6.t 17:05
on Win32, I need to hit ctrl-z for it continue past that test - presumably ctrl-d would do it on *nix 17:06
Tainted` can i use this chan for perl help? 21:17
PerlJam Are you using perl6?
Limbic_Region Tainted` - not really
Tainted` - you likely want #perl
Tainted` know any good servers i can get perl help?
ok
PerlJam try #perl or #perlhelp
Tainted` thx
Limbic_Region welcome 21:17

Powered by Cro and the Raku® Programming Language.

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

Thank you!