»ö« | perl6-projects.org/ | evalbot: 'perl6: say 3;' | irclog: irc.pugscode.org/ | UTF-8 is our friend!
Set by moritz_ on 17 June 2009.
00:02 japhb left 00:07 aindilis joined, bacek left 00:12 stDavid left 00:18 Limbic_Region joined 00:23 jaldhar joined 00:24 jferrero left 00:29 s1n1 joined 00:38 DemoFreak left, payload joined 00:41 nErVe left 00:42 japhb joined, rindolf left
wayland76 bots: ping 00:42
00:43 cdarroch left 00:44 athomason left
wayland76 pmichaud: ping? 00:47
00:49 ihrd joined 00:50 ihrd left 01:00 stDavid joined 01:13 s1n1 left 01:32 stDavid left 01:38 s1n1 joined 01:46 jaldhar left 01:47 jaldhar joined 01:51 mycelium joined 01:56 agentzh left, agentzh joined 02:03 molaf_x joined 02:04 molaf_xx left 02:05 Chillance left 02:06 hercynium joined 02:07 last joined 02:14 stDavid joined 02:16 eternaleye left 02:32 mycelium left 02:43 stDavid left 02:44 KyleHa joined
KyleHa I'm looking at [perl #66868] Zero-arg sub params interpreted as parameterless sub... 02:52
I came up with this to put in S06-signature/arity.t: dies_ok { a_zero( 'hello', 'world' ) }, 'no matching sub signature';
02:54 twigil joined
KyleHa That seems right, if this is a run time error, but I'm not sure that it is. Anyone have an opinion? 02:54
02:54 twigil left 02:58 cotto joined
wayland76 perl6: a_zero( 'hello', 'world' ) 03:05
p6eval pugs: OUTPUT«*** No such subroutine: "&a_zero"␤ at /tmp/628bzGrqTY line 1, column 1 - line 2, column 1␤»
..rakudo 10f223: OUTPUT«Could not find non-existent sub a_zero␤»
..elf 27178: OUTPUT«Undefined subroutine &GLOBAL::a_zero called at (eval 124) line 3.␤ at ./elf_h line 5881␤»
wayland76 std: a_zero( 'hello', 'world' )
p6eval std 27178: OUTPUT«Undeclared routine:␤ a_zero used at 1 ␤ok 00:02 37m␤»
03:05 azawawi joined
KyleHa The test file I'm working in defines: sub a_zero () {}; 03:06
wayland76 KyleHa: Well, STD errors on it, and, while my compilers-fu is somewhat sketchy, it seems to me that that indicates a parser error
03:06 last left
wayland76 ok, let me try again :) 03:06
std: sub a_zero () {}; a_zero( 'hello', 'world' )
p6eval std 27178: OUTPUT«ok 00:02 37m␤»
03:07 skids left
wayland76 perl6: sub a_zero () {}; a_zero( 'hello', 'world' ) 03:07
p6eval pugs: OUTPUT«*** No compatible multi variant found: "&a_zero"␤ at /tmp/R4WRwR6UpN line 1, column 19 - line 2, column 1␤»
..elf 27178, rakudo 10f223: ( no output )
wayland76 rakudo: sub a_zero () {}; a_zero( 'hello', 'world' )
p6eval rakudo 10f223: ( no output )
azawawi hi
KyleHa The bug report says the TimToady says it should error out with "no matching sub signature." 03:08
wayland76 Well, STD likes it, and pugs fails
Yeah, I'd say that's run time, but I'm pretty much randomly guessing
KyleHa I'm glad I'm not the only one guessing! 03:09
wayland76 We're approaching the time of day when it's hard to get answers, because many of the appropriate people are asleep. But sometimes TimToady is still around at this time
KyleHa That's good to know. 03:11
This is about the time I'm most free since the kids are in bed, but I haven't passed out myself yet.
03:13 Casan left
wayland76 Well, it's 1pm here (Australia), and I find that from about 2pm to about 4pm, it's not a good time to ask questions 03:16
Which timezone are you in?
azawawi wayland76: good morning from +2 timezone :) 03:17
KyleHa I'm in central time (Chicago). It's 10pm here. 03:18
azawawi 6:18am here
wayland76 azawawi: Yes, I know you're around at this time. bacek often is too. But jnthn and pmichaud and moritz_ and masak and TimToady and the like seem to not be available as much
Oh, and ruoso and pmurias 03:19
And it seems to me like they're the people with the best handle on the situation
(nothing personal against the rest of us, of course) 03:21
KyleHa: Anyway, my guess is that you'll have to just make assumptions, and ask for forgiveness if you're wrong :) 03:22
(forgiveness rather than permission)
KyleHa wayland76: Yeah, I'm going that route now. 03:23
I'll commit the tests I think are right and answer the bug emails with notes about that and pleas for assistance.
03:23 frew joined
KyleHa (Stop me before I screw up your test suite even more!) 03:24
03:30 mycelium joined 03:31 Limbic_Region left, Patterner left
pugs_svn r27179 | kyle++ | [t/spec] Add a test to show bug #66868 03:36
r27180 | kyle++ | [t/spec] add test to show bug #66854 03:40
03:51 FurnaceBoy left 03:52 azawawi left 04:00 azawawi joined
azawawi std: ''; 04:00
04:00 justatheory left
p6eval std 27180: OUTPUT«ok 00:02 35m␤» 04:00
azawawi std: module Foo::Bar { }; module Moo::Bar; 04:02
p6eval std 27180: OUTPUT«ok 00:02 36m␤»
azawawi std: 04:04
std: ;
p6eval std 27180: OUTPUT«ok 00:02 35m␤»
04:06 s1n1 left 04:07 s1n1 joined
azawawi TimToady: thanks for the fix; for STD->parse(''), im getting "Not a reference at lib/Cursor.pm line 203". It seems related re-init setting on re-parse. 04:12
04:16 hercynium left 04:19 s1n1 left 04:25 justatheory joined 04:26 justatheory left 04:30 meppl joined 04:35 agentzh left 04:41 azawawi left 04:42 ihrd joined 04:53 xinming_ joined
amoc rakudo: enum Color <foo bar>; sub foo { 'sub' }; say foo; #63650 05:02
p6eval rakudo 10f223: OUTPUT«Redefinition of routine foo␤0␤»
05:04 frew left
amoc rakudo: class A { ... }; 05:06
p6eval rakudo 10f223: OUTPUT«Can't return outside a routine␤in Main (/tmp/9k1eH40Qly:2)␤»
05:11 xinming left
amoc rakudo: enum Color <black white pink>; sub pink { 'sub' }; say pink; #63650 05:16
p6eval rakudo 10f223: OUTPUT«Redefinition of routine pink␤2␤»
05:17 ejs joined
amoc rakudo: enum Color <black white pink>; sub pink { 'sub' }; say pink(); #63650 05:17
p6eval rakudo 10f223: OUTPUT«Redefinition of routine pink␤invoke() not implemented in class 'Integer'␤in Main (/tmp/cgkJudknVS:2)␤»
05:21 xinming_ left 05:22 xinming joined 05:23 ejs left 05:24 ejs joined 05:30 molaf_x left 05:32 agentzh joined 05:35 finanalyst joined 05:40 ejs left 05:41 mikehh_ joined 05:51 mikehh left 05:55 finanalyst left
pugs_svn r27181 | kyle++ | [t/spec] Add Greek tests to autoincrement 05:57
wayland76 Yay kyle :) 06:04
KyleHa Don't cheer too loud.
I quit playing chess an hour ago because I was feeling too dumb, but that didn't stop me from going into the pugs repo... 06:05
amoc non-loud-yay kyle :D
KyleHa Heh, thanks. 06:06
06:07 xinming_ joined 06:11 xinming left 06:14 azawawi joined
moritz_ good morning 06:15
azawawi moritz_: good morning
where can i find TimToady's 'Perl 6 errors' talk at YAPC? is there any video/slides?
06:25 KyleHa left 06:32 sw1sh1 joined
amoc moritz_: good morning! 06:38
moritz_ oh hai 06:39
06:42 nihiliad left 06:49 xinming joined, PZt joined
moritz_ phenny, tell KyleHa, yes, it should be a runtime error 06:50
phenny moritz_: I'll pass that on when KyleHa is around.
06:50 finanalyst joined 06:59 DemoFreak joined, rfordinal joined 07:02 payload left 07:03 iblechbot joined, meppl left 07:05 xinming_ left
sw1sh1 hi, i just started to read wiki to learn perl6, and already found at least four things that dont work in pugs, is it still so incomplete or what? 07:10
moritz_ yes 07:11
and it's not developed anymore
we mostly use rakudo these days
sw1sh1 ok, i'll try rakudo then 07:12
moritz_ rakudo.org/how-to-get-rakudo
sw1sh1 thanks 07:13
07:14 payload joined 07:17 mikehh_ left 07:29 ejs joined 07:39 ejs1 joined 07:48 ejs left 07:49 barney joined 07:50 ihrd left 07:56 twigil joined 07:59 mjk joined
Matt-W Morning 08:01
08:03 xinming_ joined
amoc morning! 08:04
08:17 meppl joined 08:21 andreasg_ left, literal joined, andreasg_ joined 08:22 agentzh left 08:24 xinming left 08:25 agentzh joined 08:28 donaldh joined 08:30 masak joined
masak morning, lurkers! 08:30
huf lurkers, morning! 08:31
Matt-W o/ 08:34
azawawi masak: yeah it is a good morning :)
Matt-W No it's not 08:35
It was only good once I got into somewhere with air conditioning
moritz_ www.perlmonks.org/?node_id=773757 # the yapc seems to spawn some interest in Perl 6
Matt-W The air outside feels like someone's been boiling slugs in it
08:35 amoc__ joined
Matt-W moritz_: cool 08:36
08:37 synthEEEE left
masak it's a good morning, albeit a slightly late one. 08:37
wohoo! a hurkle! 08:38
Matt-W also my car's been written off 08:39
so now I have to trawl through the used car market to find a suitable replacement
masak still, that sounds like good news.
Matt-W well it's news 08:40
which is a good start
masak rakudo: say [?^] 1, 1, 1 08:41
p6eval rakudo 10f223: OUTPUT«1␤»
azawawi masak: another attempt at picture blogging, ahmadzawawi.blogspot.com/2009/06/pa...ction.html :) 08:42
masak likes picture blogging
08:43 clintongormley joined
masak azawawi++ # nice! 08:43
azawawi thx and that's just the beginning; p6 var/sub/method refactoring is next :) 08:44
Matt-W Hmm 08:52
Ctrl+~ is not an easy keystroke on a UK keyboard
moritz_ easier than on a German keyboard :-)
Matt-W (poking around the entries about Padre)
huf how do you code on anything but the US layout anyway? :)
08:53 amoc left
Matt-W moritz_: is it? 08:53
azawawi Matt-W: really?
moritz_ Matt-W: on the German keyboard you need AltGr + another key to create the ~
Matt-W azawawi: ~ requires shift on the UK layout
azawawi Matt-W: what about Ctrl-1?
moritz_ Matt-W: so you'd need Crtl+AltGr+Q or so
08:53 mjk left
Matt-W moritz_: sounds about the same as Ctrl+Shift+# 08:53
which is what it is here 08:54
azawawi Matt-W, moritz_: so is Ctrl + 1 easier to use on a uk, german keywords? 08:55
moritz_ azawawi: it sure is
azawawi s/keywords/keyboard/ :)
Matt-W definitely
huf: coding on a UK keyboard is easy, because the punctuation is mostly in the same place as the US layout so we don't have quite so many insane key combinations as some of the other european layouts have to suffer 08:56
azawawi then i should 'borrow' it from Padre's core since it is using it :)
huf Matt-W: ah 08:57
szabgab that's the brace jumper!
huf i took one look at our national layout and forever said no... :)
azawawi szabgab: lol
moritz_ speaking of weird keyboards... writing perl on a keyboard without a dollar sign is a real pain
Matt-W I bet 08:58
fortunately we have one of those
although it'd be easier if it wasn't shift-4, I suppose
szabgab huf, whz that kezboard lazout is really nice
moritz_ at least I couldn't find it on the Norwegian layout
szabgab reallz 08:59
sorrz, that was a tzpo
huf szabgab: altgr to get ';'? *real* nice ;)
masak moritz_: I've been doing most of my computer work with a US layout for over two years now. not going back. I say this as someone whose mother tongue has a strange alphabet. it's a matter of having good software to compensate.
Matt-W German keyboards seem pretty horrible for coding to me
huf isnt german the one that has the "paired" chars "unpaired"? 09:00
Matt-W the extra letter keys do nasty things to the availability of your punctuation
moritz_ masak: agreed
huf as in, [] and {} are not next to each other... or was it <>?
moritz_ Matt-W: right
huf: [] and {} are next to each other, but you need AltGr to reach it
masak moritz_: I can still switch, but the Swedish layout now seems quaint and slightly unnatural.
moritz_ < and > are on the same key, separated by shift 09:01
(sigh)
huf oh yeah, that was it ;)
moritz_ you get used to it, but it makes switching between US and German layout a real pain. 09:02
masak so don't. that's my point. 09:03
use US, and emulate the rest.
Emacs is your friend. even vim helps.
moritz_ well, I also use other programs than vim 09:04
like icew{weasel,dove}, the shell etc. 09:05
masak aye, Firfox/Gmail is about the only time I switch, too.
moritz_ so I really need an input method that's handled either by X or the OS
masak for longer emails, I simply type things out in Emacs and copy/paste.
moritz_ plus, I have a job as an administrator, and have to cope with whatever keyboard the poeple I work with use 09:06
so I can't get around switching anyway
masak no, it's probably not possible to get around it altogether. 09:08
I just try to minimize it, and to make my default as pleasant as possible.
huf i just dont use the extra chars, so the US layout is fine
masak huf: and you're... from the US? 09:09
huf hungary
masak which means you skip some accents and the like? 09:10
huf we've got a few accented chars, but it's perfectly legible without them
plus, code is in english anyway
masak huf: that's not an acceptable solution for Germans and Swedes, I'd say.
huf: plus, it feels like a bad compromise.
moritz_ I just find it impolite to transliterate other people's names
masak that, too.
moritz_ in emails or so it's fine to substitute &auml; with ae, but for example not in websites I write 09:11
or æ with ae 09:13
.øØ( thought bubbles for wrong thoughts ) 09:16
maybe I should use capslock as the compose key 09:17
09:18 payload left
wayland76 moritz_: But do you have to type non-English on the keyboards of people you administer? 09:18
09:18 bacek joined
krunen I use US-International with no-dead-keys for norwegian. AltGr-[ZLW] is [ÆØÅ]. That also has «» and some other snacks. 09:19
moritz_ wayland76: usually not
wayland76 That's what I thought
I have two keyboards
One is US-ASCII or something
The other is US-International + Greek + Hebrew + APL 09:20
(although it doesn't have all those symbols printed on it)
moritz_ sounds pretty scary :-)
wayland76 And I use Scroll lock to switch layouts
And the scroll lock led lights up when I'm not in US-International :) 09:21
moritz_ what kind of tool do you use for the switching?
wayland76 I press Scroll lock :) 09:22
moritz_ on the backend, I mean
wayland76 (just in case I didn't make myself clear, I have two physical keyboards)
moritz_ but the switching between Greek and US-International, for example, is handled in software, no? 09:23
Matt-W I wish I had a compose key on windows
on linux I can type virtually anything from a latin-ish european language easily
wayland76 setxkbmap
moritz_ wayland76: thanks
wayland76 I have a perl script that finds the second keyboard, and runs setxkbmap on it
setxkbmap is configured with a bunch of undocumented but moderately straightforward config files in /usr/share/X11/xkb with all the combining being done by the files in /usr/share/X11/locale/<locale>/Compose and the like 09:24
I'm also planning to make a keyboard that I might be able to sell that does US-International + Greek + Maths/APL 09:25
moritz_ locales. gah. 09:26
I try to avoid those like hell
wayland76 Well, yes, but the compose sequences are kinda cool :)
moritz_ right 09:27
and perl5 doesn't handle Unicode correctly if the wronge locale is set.
wayland76 Well, this isn't the generic locale stuff, it's X-specific stuff, which pretty much means a huge list of compose sequences mapping to characters 09:28
Unfortunately my second keyboard doesn't have a right windows key
moritz_ a different locales system.
oh wow
well, off to lunch 09:29
wayland76 But it does have keys that produce no scancodes :(
Matt-W hah
wayland76 well, not a different locales system
09:29 bionoid joined
Matt-W PuTTY has compose key support built in 09:29
that'll do for me :)
%hash«key» :D
(did that render right?)
wayland76 But just that the locale stuff in XWindows doesn't do a lot of the stuff that's done by other software 09:30
ie. a different set of config files that do different things.
Matt-W: French rather than Texan :)
PuTTY++ ++ ++
I avoid Windows as much as possible, but PuTTY is a good bit of software for which I have been thankful often enough 09:31
huf yeah, frustrating movie. the end is nice tho
wayland76 For those wondering about the locale files, but too lazy to look, I'll paste one line from them 09:32
<Multi_key> <o> <r> : "®" registered # REGISTERED SIGN
bionoid Also there is a google code mod called "puttycyg" that allows you to use putty as a local cygwin terminal if you _have_ to do some actual work on Windows ;)
wayland76 I haven't seen "PuTTY: the movie" yet :)
huf i meant windows the movie :)
wayland76 I take it by "nice", then, than you mean that Windows dies a violent death, but somewhere offscreen where we don't have to watch? 09:33
(afk for a bit, but will backlog)
huf ;)
09:36 azawawi left 09:50 amoc__ left, amoc joined
masak I have a Model M keyboard here at work. I have to bribe my co-workers constantly so they won't complain about the clattering too much -- but it's well worth it. 09:52
wayland76 masak: I just got an Avant Stellar :) 10:01
Buckled spring, but the keyboard can be reprogrammed from the keyboard, without software :)
masak (springs)++
wayland76 My only gripes are that it doean't have n-key rollover, and that I haven't figured out a way to make it send any keycode that isn't already on one of the keys 10:02
I wonder if the Optimus Maximus has n-key rollover :) 10:03
masak rakudo: enum A <b>; say A::c
p6eval rakudo 10f223: OUTPUT«Null PMC access in invoke()␤in Main (/tmp/eZqi9x5plT:2)␤»
masak submits rakudobug
rakudo: say q<< <<woot>> >> 10:04
p6eval rakudo 10f223: OUTPUT«Statement not terminated properly at line 2, near ">>"␤in Main (src/gen_setting.pm:0)␤»
masak submits rakudobug
rakudo: say (map -> $a, $b { $a }, 1..5).perl; for 1..5 -> $a, $b {} 10:05
p6eval rakudo 10f223: OUTPUT«[1, 3]␤StopIteration␤in Main (/tmp/2qZBAsxsOE:2)␤»
masak submits rakudobug
rakudo: sub foo(--> Array of Str) { my Str @a = <foo bar baz>; @a }; foo
p6eval rakudo 10f223: OUTPUT«Use of type object as value␤Type check failed on return value␤in sub foo (/tmp/gJcBE0Ht4P:2)␤called from Main (/tmp/gJcBE0Ht4P:2)␤»
masak submits rakudobug
what? I had a few lying around. :) 10:06
10:07 last joined
masak lunch & 10:14
10:15 sw1sh joined 10:16 sw1sh1 left 10:17 AzureStone_ joined 10:20 last left, AzureStone left 10:28 last joined 10:29 amoc left 10:30 amoc joined 10:34 ejs2 joined 10:43 ejs1 left 10:45 payload joined 10:49 payload left, payload joined
moritz_ is there a cheap test somehow to figure out if an integer is a square number, without calculating sqrt($x)? 11:05
last moritz_: I don't think so. 11:08
barney only test positive integers
11:09 sri_kraih_ left
last moritz_: There are certainly pretty good first-pass filters you could use. Like if it's even, it has to be divisible by four. 11:11
moritz_: www.mathpages.com/home/kmath265.htm 11:13
11:13 s1n1 joined 11:14 sri_kraih joined
huf btw, will rakudo be self-hosting eventually? 11:16
11:16 s1n2 joined
moritz_ last: thanks, that looks pretty interesting 11:17
huf: well, it'll always need parrot, presumably. 11:18
huf self-hosting on parrot is what i mean
entirely written in perl6 and able to compile itself
moritz_ some of the runtime will probably always remain in PIR, for speed reasons 11:19
but it's pretty much written in Perl 6 already (the parser, the action methods (ie transforming the parse tree into a syntax tree, and the setting (ie builtins))
11:20 donaldh left
moritz_ it might become self-hosting if there's something to gain from it 11:20
but at some point we'll surely get rid of the Perl 5 dependency for building and testing
11:20 donaldh joined 11:21 ejs1 joined, barney left
moritz_ last: if it were a pressing problem I'd probably invent something along these lines for binary numbers, but it doesn't seem to warrant the effort for project euler :-) 11:21
11:22 szabgab joined, mizioumt joined
last mortiz_: I believe a quick (?) binary filter is to count the number of zero digits at the end of the number. That number has to be even for a square. 11:25
11:26 payload left
last moritz_: Seems like there ought to be a fast hex equivalent to the sort of tests on that page, but I should probably do paying work rather than try to figure it out. :) 11:27
moritz_ last: :-) 11:28
masak moritz_: you could calculate [+} 1, 3, 5 ... $n until you go above your number. :)
11:30 ejs2 left, s1n1 left
moritz_ masak: :/ 11:30
masak no good? :)
moritz_ not particularl efficient 11:31
*particularly
masak no, definitely not.
but it avoids sqrt.
moritz_ I should've asked for something *faster* than sqrt() :-) 11:32
masak yes...
last How are bignums stored internally?
masak ok, I'd like to talk about StrPos, one of those nice ideas, which so far hasn't seen an implementation anywhere (AFAIK).
in my opinion, the substr multi in S32/Str.pod is missing a variant whete $start is an Int. 11:33
same with index and rindex.
I'll willingly add those; just seeking confirmation first that I'm not insane.
um, that what I'm proposing is not insane. :P 11:34
moritz_ last: dunno, I just use them :-) 11:35
masak: I think the only one how maybe has a faint idea of how StrPos works is TimToady 11:36
masak moritz_: it's described fairly well in S02.
the important point is that StrPos is _not_ and Int.
nor does an Int coerce into a StrPos.
moritz_ in my mind they are multi-valued objects 11:37
ie a StrPos could know "I'm 5 graphs, or 8 codes, from the start"
masak moritz_: not sure you're seeing my point. the way substr works now (with Ints) will still be supported in Perl 6.0.0. 11:38
the current signatures in S32/Str don't allow for that.
(same with index and rindex)
moritz_ masak: I didn't even get to your point yet :-)
masak Rakudo and the spec are at odds on that point right now. 11:39
moritz_: ok, I'm too eager. :) go on.
(I do agree about the multivaluedness)
moritz_ the point is, if you have an integer and want it to be interpreted as a StrPos (like when using it with substr) you have to be very clear about your units, again
masak moritz_: S02 mentions that, too.
an Int defaults to the current scope's string abstraction level. 11:40
moritz_ maybe that's what's controlled by 'use codes;' etc.
masak aye.
moritz_ though I don't quite understand the talk about a StrLen knowing its length also in bytes 11:41
because I thought that Str's don't have a byte level, only Buf's 11:42
masak arguably the spec is self-contradictory in places.
there are geological layers of insight involved.
moritz_ masak: but IMHO you can go ahead and add the Int candidates of these multis 11:43
masak goes ahead
as far as I can see, that's one Int variant for every StrPos variant.
11:43 cognominal left 11:44 Whiteknight joined
moritz_ be careful for example with substr 11:44
11:45 ejs2 joined
moritz_ given substr(Str $s, StrPos $start = 0, StrLen $len = *, Str $replacement?) 11:45
masak the defaults? yes.
just realized that.
moritz_ will you allow (Str, Int, StrLen)?
masak I think so.
moritz_ so you need Int|StrPos and Int|StrLen?
masak ah, I can write it like that? great. 11:46
moritz_ I'm not sure.
11:46 azawawi joined
moritz_ we had so much back and forth on that one 11:46
azawawi yellow 11:47
masak rakudo: sub foo(Int|Str $a = "foo") { say $a }; foo(); foo(5); foo("bar")
p6eval rakudo 10f223: OUTPUT«sh: ./perl6: No such file or directory␤»
masak not supported in Rakudo, at least.
I'll write it the long way for the time being. 11:48
moritz_ std: foo(Int|Str $a = "foo") { say $a }; foo(); foo(5); foo("bar") 11:49
p6eval std 27181: OUTPUT«##### PARSE FAILED #####␤Unable to parse argument list; couldn't find final ')' at /tmp/yijaDp463J line 1:␤------> foo(Int|Str $a = "foo") { say $a }; foo(); foo(5); f␤ expecting any of:␤ infix or meta-infix␤ infix stopper␤ standard stopper␤
..terminator␤FAILED 00:02…
11:50 jdv79_ left
masak same error as Rakudo, incidentally. 11:50
11:51 frew joined
masak oh! and won't the signatures (Str $string: StrPos $start, StrLen $length?) and (Str $string: StrPos $start, StrPos $end?) tie if no third argument is provided? 11:51
should I remove one of the question marks? 11:52
last moritz_: Reason I ask about bignums is because that is going to determine how to efficiently look for large squares. If your potential squares fit into a 32-bit Int, you're probably better off just building a table of all the 32-bit squares (should take 256K of memory) and doing a binary search into it. (At least, assuming you're doing a enough lookups to make it worth optimizing.) (Hey, I'm compiling in another window.) 11:53
11:53 ejs1 left 11:54 jdv79 joined 11:56 wayland76 left
pugs_svn r27182 | masak++ | [S32/Str] Added Int variants to index/rindex/substr 11:57
r27182 | masak++ |
r27182 | masak++ | There has to be a way to send in an Int to these Str methods. Arguably,
r27182 | masak++ | StrPos and StrLen are not Ints -- that's their point. Also, removed
r27182 | masak++ | optionality from the $end param in substr, because IMO that would lead
r27182 | masak++ | to MMD ties when no third argument is provided.
12:03 azawawi_ joined 12:04 wayland76 joined
masak like this: 12:04
rakudo: multi foo(Str $a, Str $b?) {}; multi foo(Str $a, Int $b?) {}; foo("bar")
p6eval rakudo 10f223: OUTPUT«Ambiguous dispatch to multi 'foo'. Ambiguous candidates had signatures:␤:(Str $a, Str $b?)␤:(Str $a, Int $b?)␤in Main (/tmp/kKYhDPWV05:2)␤»
masak you don't know if it's a Str or an Int that wasn't provided. :) 12:05
12:05 azawawi left 12:06 ejs1 joined 12:07 azawawi_ left 12:08 jdv79 left 12:12 Whiteknight left
moritz_ last: I expect them to be 64bit ints, but I also estimated the current run time is sufficient. It's not a program I have to run often :-) 12:14
12:15 ejs2 left 12:20 Whiteknight joined 12:21 lichtkind joined
last moritz_: Ah, obviously building a 16GB table to speed up your calculations is probably a bit impractical. 12:22
moritz_: I mean 32GB -- 4 GB 64-bit numbers. 12:25
12:25 skids joined
wayland76 ...unless he can keep it on disk and it's still faster 12:26
12:28 zamolxes joined 12:37 FurnaceBoy joined 12:38 Util joined, s1n2 left 12:45 Util left 12:46 mizioumt left 12:48 broquaint joined 12:49 s1n1 joined 12:50 twigil1369 joined 12:54 cognominal joined 12:56 synth joined
pmichaud Good morning, #perl6 12:58
wayland76 Good morning 12:59
Did you get my message asking about the branch with the make files?
masak mornin'. 13:00
pmichaud I'm way behind on email
(and other messaging environments)
today's the first "free" day I've had since last Thursday
13:01 ejs1 left
pmichaud wayland76: was the message sent via email, irc, rt, or ... ? 13:03
wayland76 ok. The basic question was, can you give me a link to a diff (or the branch) that has the changed makefiles in it?
(sent via lambdabot and phenny)
13:04 Gothmog_ left, twigil left
pmichaud branch is the "ins" branch in github 13:04
wayland76 Ok, thanks.
pmichaud it's out of date already, but the basic issues are addressed there
wayland76 Ok 13:05
pmichaud current problem is that pbc_to_exe still depends on the build tree
wayland76 Ok. I´m going to need to make my RPM patches follow it, that´s all :)
pmichaud sounds good. :-) 13:06
13:06 jaldhar left
pmichaud But I really hope the whole install process gets a good refactor rather than us just plastering more slop on top of it to make things work. 13:06
(It's a hope, not an expectation, alas.) 13:07
wayland76 Well, I´ve been talking to kid51
pmichaud excellent 13:08
wayland76 My hope all along has been to make the Rakudo install process use any relevant and useful bits from the parrot install process
pmichaud Please, no.
wayland76 And he said that you said that the Parrot install process needs reworking
pmichaud In some ways I think that's the wrong direction.
should we switch this discussion to #parrot ? 13:09
wayland76 Ok, although I´ll need to go to bed sometime soon. I´ll be there in a minute
13:11 skids_ joined 13:18 pmurias joined
pmurias wayland76: re keep on disk, doing two multiplications is likely (way) faster then doing a disk lookup 13:19
wayland76 pmurias: Yes, I wasn´t following the thread closely enough :)
13:23 stDavid joined 13:29 molaf_x joined, Util joined
Util What is Perl 6 equiv of Perl 5's $::foo and &::bar? Is the bare :: shortcut for main:: still around? 13:32
13:33 mizioumt joined
masak Util: I don't believe so, but safest bet is to check the synopses. 13:34
moritz_ maybe the $*foo thing?
13:36 c9s joined
masak surely not an exact equivalent? 13:37
moritz_ no
but it used to mean "look in the global namespace"
13:38 stDavid left 13:42 jaldhar joined
pmichaud GLOBAL:: 13:42
masak THERMONUCLEAR:: 13:44
Util PEACE::
masak :)
pmichaud I don't think there's a shortcut for GLOBAL:: .
13:45 bacek left
pmichaud and GLOBAL:: is really CORE::GLOBAL 13:45
13:46 araujo left
Util Maybe binding (:=) symbols in the current package to the same-named symbols in GLOBAL:: at the top of a module will be a good equivalent Perl 6 idiom? 13:47
13:48 azawawi joined
azawawi hi, is ctrl-shift-1 an easy shortcut to use on uk/german keyboards? 13:49
moritz_ better than Ctrl+~ in any case
azawawi moritz_: thx 13:50
13:51 alester joined
alester Anyone know who texted me about admin access for pmichaud from 201 area code? 13:51
13:52 stDavid_ joined 13:54 stDavid_ left
pmichaud alester: particle 13:54
alester aha 13:55
13:55 Util left
alester I have never seen his phone # 13:55
pmurias azawawi: ctrl-shift-1 is ctrl-! ?
azawawi pmurias: yeah.. 13:56
13:56 Gothmog_ joined
rgs on french keyboards you need a shift to type 1, so shift-1 isn't possible. (or hit both shift keys ?) 13:57
pmurias azawawi: re quick fix animated gifs, $foo = -1 should be changed into my $foo = -1; and if(1) {...} into if 1 { }
rgs: what's more important then 1?
moritz_ 0 13:58
azawawi pmurias: sure but what if you had $foo == 1... it is a bit generic i know. 13:59
13:59 REPLeffect joined 14:01 jaldhar left
masak perldoc.perl.org is one of the nicest-layouted Perl pages on the Web. 14:03
but what happened to the bullet list here? perldoc.perl.org/perlsyn.html#Switch-statements
moritz_ =item o 14:04
...
rgs fixed in bleda
blead
14:04 jaldhar joined
rgs (with =item *) 14:04
pmurias azawawi: fixing $foo == 1 is hard as it's need some value in $foo for the comparision to be usefull 14:05
masak oh, good.
moritz_ rgs: btw I found your post on the future of perl 5 very interesting, thank you 14:08
rgs thanks
masak URL? 14:09
14:10 last left
moritz_ consttype.blogspot.com/2009/06/futu...erl-5.html 14:10
masak dz
14:12 masak left
wayland76 'night all 14:12
14:12 wayland76 left 14:13 finanalyst left
azawawi pmurias: here is the new try { } quick fix which only appears when you have no std errors - feather.perl6.nl/~azawawi/quick_fix...ix_ani.gif 14:19
moritz_ azawawi: $foo is a variable, not an attribute (in your animation) 14:26
azawawi moritz_: i know. :) 14:27
moritz_: but im lazy :)
pmurias use !!! instead of #XXX-implement 14:30
azawawi pmurias: for # (TODO|XXX...etc) - implement; i think we can have that configurable in Padre somewhere 14:32
pmurias rakudo: !!!
p6eval rakudo 10f223: OUTPUT«Syntax error at line 2, near "!!!"␤in Main (src/gen_setting.pm:0)␤»
pmurias pugs: sub foo {!!!};foo(); 14:33
p6eval pugs: OUTPUT«*** Unimplemented stub called␤ at /tmp/BxPjH9moyb line 1, column 15-20␤ /tmp/BxPjH9moyb line 1, column 10-13␤»
azawawi pmurias: what's !!! ? 14:34
pmurias azawawi: die, as this thing is not implemented yet
moritz_ it's the "this is a stub" marker
pmurias the lethal variant
moritz_ right
... is fail()
!!! is die() 14:35
and ??? is warn()
or so.
azawawi really? :)
azawawi learnt a lot of Perl 6 while doing implementing quick fixes...
s/learnt/learned/ 14:36
14:38 dj-dongdot joined 14:39 cottoo joined, cotto left 14:41 ruoso joined, azawawi left 14:42 nihiliad joined 14:57 justatheory joined
jaldhar take a look at this code: repeat { state $count = 0; } until ++$count == 5; 14:58
rakudo gives: Symbol '$count' not predeclared in <anonymous>
is that right? 14:59
14:59 sri_kraih_ joined
moritz_ that's right 14:59
because the scope of $count is limited to the block it's declared in 15:00
spx2 ~
s1n1 am i reading that snippet right, an infinite loop? or are state variables only assigned like that once?
moritz_ s1n1: it's a compile time error, nothing else.
15:01 s1n1 is now known as s1n_yapc
s1n_yapc moritz_: but i was referring to the intent of the repeat {} loop 15:01
moritz_ rakudo: while 1 { state $x = 0; last if $x++ > 5 }; say $x
s1n_yapc in that it does it repeatedly reassign $count to 0?
p6eval rakudo 10f223: OUTPUT«Symbol '$x' not predeclared in <anonymous> (/tmp/zvUygAeK4O:2)␤in Main (src/gen_setting.pm:3225)␤»
jaldhar moritz_: so the repeat...until is not "part" of the block? but why is for or while different? (or is it?)
moritz_ rakudo: while 1 { state $x = 0; say $x; last if $x++ > 5 };
p6eval rakudo 10f223: OUTPUT«0␤1␤2␤3␤4␤5␤6␤»
moritz_ jaldhar: it's not different
jaldhar s1n_yapc: no only the first time 15:02
s1n_yapc moritz_: okay
moritz_ jaldhar: a block extends from its opening brace to its closing brace
no Perl 5-like exceptions
jaldhar moritz_: ok. I was thinking about things like for < 1 2 3 4 5 > -> $count { say $count; } 15:03
moritz_ jaldhar: that's something else 15:05
there -> $count is a signature
which introduces $count as a formal parameter for that block
jaldhar I see 15:06
pmichaud (the arrow is part of the block, not part of the 'for') 15:08
moritz_ right 15:09
huf how do blocks differ from anonymous subs?
moritz_ rakudo: my $x = -> $a { $a * $a }; say $x.WHAT
p6eval rakudo 10f223: OUTPUT«Block()␤»
moritz_ huf: return() stops at sub boundaries, but not at block boundaries
pmichaud rakudo: my $x = sub ($a) { $a * $a }; say $x.WHAT; 15:10
p6eval rakudo 10f223: OUTPUT«Sub()␤»
huf so you cant do for @list sub { .. } ?
moritz_ huf: no
pmichaud std: for @list sub ($_) { ... }
p6eval std 27182: OUTPUT«##### PARSE FAILED #####␤Missing block at /tmp/GGe2qS4QtC line 1:␤------> for @list sub ($_) { ... }␤ expecting any of:␤ infix or meta-infix␤ infix stopper␤ parameterized block␤ standard stopper␤ terminator␤Other potential difficulties:␤ Variable
..@list is not pred…
moritz_ huf: 'for' has a grammar rule that expects a block afterwards 15:11
huf mhm
pmichaud it wants a block, not a sub declaration.
moritz_ you can do it with map, though
rakudo: say <a b c>.map: sub ($x) { uc $x }
TimToady rehi
p6eval rakudo 10f223: OUTPUT«ABC␤»
TimToady is in despair at the backlog...
15:11 moritz_ sets mode: +o pmichaud
huf but $a = -> {}; and the you can do $a(); correct? 15:12
moritz_ huf: correct
TimToady correct
15:14 sri_kraih left 15:18 rfordinal left 15:19 rfordinal joined 15:20 donaldh left, donaldh joined, donaldh left, dj-dongdot left 15:30 masak joined 15:31 jaldhar left
masak rakudo: given 42 -> $foo { say $foo } 15:33
p6eval rakudo 10f223: OUTPUT«42␤»
masak rakudo: my @a = 1..3; given @a -> $a, $b, $c { say $a } 15:34
p6eval rakudo 10f223: OUTPUT«too few arguments passed (1) - 3 params expected␤in Main (/tmp/mYGyO3r5rL:1)␤»
TimToady rakudo: my @a = 1..3; given @a -> [$a,$b,$c] { say $a } 15:35
p6eval rakudo 10f223: OUTPUT«Statement not terminated properly at line 2, near "-> [$a,$b,"␤in Main (src/gen_setting.pm:0)␤»
TimToady that one should work
masak std: my @a = 1..3; given @a -> [$a,$b,$c] { say $a }
p6eval std 27182: OUTPUT«ok 00:02 38m␤»
moritz_ well, rakudo doesn't do any parameter unpacking yet
masak looks forward to parameter unpacking 15:36
TimToady parameter unpacking is the road to tree matching
masak std: my [$a,$b,$c] = [1..3];
p6eval std 27182: OUTPUT«##### PARSE FAILED #####␤Malformed my at /tmp/MYGV8rIMyu line 1:␤------> my [$a,$b,$c] = [1..3];␤ expecting scoped declarator␤FAILED 00:02 35m␤»
TimToady [] is not a signatre 15:37
*ture
15:37 jaldhar joined 15:39 s1n_yapc left, rfordinal left 15:40 Psyche^ joined, Psyche^ is now known as Patterner
masak TimToady: so the thing between '->' and '{' is not a signature? 15:43
15:43 frew left 15:45 Whiteknight left
TimToady it is a signature, by context 15:45
it wouldn't be a signature without the -> in front
15:47 Chillance joined, last joined
PerlJam TimToady: so, how do you do the array unpacking as masak intended? 15:52
TimToady basically you bind the inside of [] or () as a subsignature 15:53
they're really much the same
15:53 s1n1 joined
TimToady the [] is more like documentation that you only expect a list 15:53
masak that's all very well, but how to write it? 15:54
TimToady you already have signature binding, how hard can it be to call itself recursively?
masak follows not 15:56
15:57 s1n1 left
PerlJam yeah, I think I need some syntax or something. 15:57
masak yes, my thought exactly. 15:58
we're syntax-deprived simpletons, here. :)
moritz_ Perl 6: syntax driven development 15:59
PerlJam at a guess, :([$a,$b,$c]) = [1..3]; # not sure if I need := though 16:02
16:02 cdarroch joined
PerlJam (not sure of anything really, this is just what my memory tells me right now) 16:02
PerlJam checks the synopsis 16:03
huf can you do something like erlang and its ilk have?
take apart a complex data structure in one line
16:03 cottoo left
moritz_ like haskell's pattern matching? 16:06
PerlJam huf: probably. That sounds like what we're talking about now.
masak yes. it's especially attractive in for loops, I think. 16:07
std: my @a = [[1, 2], [3, 4]]; for @a -> [$a, $b] { say ($a => $b).fmt }
p6eval std 27182: OUTPUT«ok 00:03 39m␤»
masak \o/ 16:08
PerlJam std: :([$a,$b,$c]) = [1..3]
p6eval std 27182: OUTPUT«ok 00:02 37m␤»
masak std: my :([$a,$b,$c]) = [1..3]
p6eval std 27182: OUTPUT«##### PARSE FAILED #####␤Malformed my at /tmp/KJ7D9vzRZq line 1:␤------> my :([$a,$b,$c]) = [1..3]␤ expecting scoped declarator␤FAILED 00:02 35m␤»
masak :(
moritz_ my ($a, $b, $c) = 1..3;
PerlJam moritz_: he wants the more complicated version of that though :) 16:09
masak: I think :($a,$b,$c) doesn't need the my 16:10
masak moritz_: given @a = [[1,2],[3,4]], how do I assign directly from @a into four new variables?
PerlJam: oh. that's good news, in that case. 16:11
PerlJam: if they don't need the my, how do I express an 'our' equivalent? :)
16:11 viklund joined
PerlJam masak: can you do that in subroutine signatures? 16:11
huf is this just in sub signatures? 16:12
viklund o/
PerlJam doesn't know if :() can be used in an lvalue context though 16:14
jaldhar come on guys you need to implement :-) somehow
PerlJam all of the examples in the synopses I've found have :() as rvalue 16:15
masak jaldhar: write a module!
amoc std: sub foo ([$a, $b]) { ... }
p6eval std 27182: OUTPUT«ok 00:02 37m␤»
16:17 azawawi joined
azawawi home sweet home :) 16:17
PerlJam rakudo: sub postfix:< :-) >($x) { say "$x : happy?" }; "jaldhar" :-) 16:18
p6eval rakudo 10f223: OUTPUT«jaldhar : happy?␤»
masak huh, don't postfix ops require unspace nowadays? 16:19
PerlJam oh .. maybe. If so, bug.
masak an already-reported one, I suspect.
std: sub postfix:< :-) >($x) { say "$x : happy?" }; "jaldhar" :-) 16:20
p6eval std 27182: ( no output )
PerlJam rakudo++
masak note that STD.pm chickened out on that one.
16:20 M_o_C joined
masak it didn't say "ok". 16:20
PerlJam it didn't complain either. That makes rakudo right ;) 16:21
masak I'm not saying you're right. :P
jaldhar PerlJam: awesome! 16:22
16:23 rfordinal joined
PerlJam btw, when I was writing that example for jaldhar, I wanted to write "suffix" rather than "postfix". Am I the only one who thinks that way? 16:23
masak PerlJam: yes. :P 16:24
amoc rakudo: sub postfix:< PFIX >($x) { say "$x : happy?" }; "jaldhar" PFIX
p6eval rakudo 10f223: OUTPUT«jaldhar : happy?␤»
TimToady that's no good 16:25
jaldhar fwiw postfix sounds better to me
suffix means eg -ing -ed etc.
TimToady std: sub postfix:< PFIX >($x) { say "$x : happy?" }; "jaldhar" PFIX
huf you could impement those too :D
p6eval std 27182: OUTPUT«##### PARSE FAILED #####␤Syntax error (two terms in a row?) at /tmp/5Z3yTfkubI line 1:␤------> >($x) { say "$x : happy?" }; "jaldhar" PFIX␤ expecting any of:␤ infix or meta-infix␤ infix stopper␤ standard stopper␤ statement modifier loop␤
..terminator␤FAILED 00:05 45…
PerlJam jaldhar: prefix/suffix make a natural pair for me.
TimToady std: sub postfix:< :-) >($x) { say "$x : happy?" }; "jaldhar" :-) 16:26
p6eval std 27182: OUTPUT«##### PARSE FAILED #####␤Syntax error (two terms in a row?) at /tmp/cOYifZWMBl line 1:␤------> >($x) { say "$x : happy?" }; "jaldhar" :-)␤ expecting any of:␤ infix or meta-infix␤ infix stopper␤ standard stopper␤ statement modifier loop␤
..terminator␤FAILED 00:05 45m…
last moritz_: I'm too lazy to prove this, but it appears that just as there are only 22 possible last two digits (base 10) for squares, there are 44 possible last two hex digits for a square. That should provide a very fast test to filter out 82% of all positive integers as non-squares (presuming you can easily get at the binary representation of your number).
TimToady std: sub postfix:< :-) >($x) { say "$x : happy?" }; "jaldhar"\ :-)
p6eval std 27182: OUTPUT«ok 00:05 45m␤»
PerlJam okay, two bugs then
TimToady lunch &
masak I'm not submitting, because I believe both are known/submitted. 16:27
PerlJam yeah, I was going to look just in case
16:28 aindilis left, s1n1 joined 16:29 clintongormley left
masak PerlJam: good idea. 16:30
rakudo: "foo" ~~ /foo/; sub bar() { say $/ }; bar 16:31
p6eval rakudo 10f223: OUTPUT«Use of uninitialized value␤␤»
masak what if I want $/ to be $OUTER::/ in a sub? 16:32
PerlJam isn't OUTER for lexical scopes anyway? I thought $/ was dynamically scoped. 16:33
(well, supposed to be dynamically scoped even if rakudo doesn't think so) 16:34
16:37 amoc left
PerlJam oh no, looks like $/ is lexically scoped. 16:42
PerlJam really needs to use perl 6 more
moritz_ isn't it context<rw>?
16:42 s1n1 left
masak in the interests of release-early, let me announce Grampa, an XPath matcher for Match objects: github.com/masak/grampa 16:42
s/me/us/, viklund++ is in it too. 16:43
azawawi masak: hi
moritz_ grampa = grammar paths?
masak azawawi: greetings.
moritz_: aye.
16:43 icwiener joined, Chillance left
azawawi masak: I am looking for Larry's recent YAPC speech. Any links/slides/video would be appreciated. 16:44
acajou me would also be interested... 16:45
masak azawawi: I'm as in the dark as you about that.
I'm not on YAPC|NA.
acajou s#me#/me# :)
moritz_ usually they take much more time to appear somewhere
azawawi masak: so we're in the same ship; cool :)
acajou I wish some were Webcast...
masak azawawi: yes, so let me know if you find something... :) 16:46
acajou too!
azawawi will do for sure
masak afk # food 16:47
azawawi hmmm... PerlCritic quick fixes or Perl 6 rename var / extract method :)
16:47 frew joined 16:48 Psyche^ joined
masak azawawi: the totally unbiased answer of this channel is "Perl 6 rename var". :) 16:49
16:56 mj41 left 16:57 s1n1 joined 16:59 mj41 joined 17:04 icwiener_ joined 17:05 Patterner left, Psyche^ is now known as Patterner 17:08 ruoso left 17:09 cottoo joined 17:11 zamolxes left 17:13 M_o_C left 17:18 icwiener left 17:20 [particle] left, [particle] joined
pmichaud azawawi/masak: Larry's talk did not have slides this year. 17:21
It was completely live demo.
azawawi pmichaud: thx 17:22
pmichaud: who wrote the 'padre' dumper for PGE highlighter? 17:24
17:24 perlmongo joined
pmichaud azawawi: some hacker called "pmichaud" 17:24
s/called/nicknamed/
pugs_svn r27183 | lwall++ | [CORE] add Cat type
pmichaud TimToady: is the "Cat" type meant to chase down the Rats? 17:25
TimToady it's meant to be a virtual string 17:26
pmichaud forgot my smiley :-|
TimToady cats play with string, after all
azawawi pmichaud: after enabling it in Padre Perl 6; i found that it does not generate any 'pod_comment' or 'hash_comment' records.
pmichaud I think we're not currently capturing pod_comments 17:27
azawawi pmichaud: it is really super-fast... pmichaud++
17:28 nErVe joined 17:31 justatheory left
masak what's a virtual string? I thought we had enough issues with our physical strings... :P 17:31
pmichaud The physical strings had insufficient virtuality.
:-P
which is yet another issue with physical strings, which is why the virtual strings are addressing that issue.... 17:32
obviously I have been too long at yapc|10 and my brain is now tapioca.
moritz_ and you learned about all these new layers of indirection that you have to implement? ;-) 17:34
pmichaud it's abstraction levels, all the way down.
masak what's wrong with this picture? news.perlfoundation.org/2009/06/
PerlJam it's not a picture at all? :)
masak PerlJam: well, the thing that hits your retinas is. but s/picture/page/, if that makes you see it. 17:35
PerlJam But the word "no" appears far too often in regards to progress.
masak I was referring to the posts numbering. 17:36
PerlJam oh, yeah
that too
masak OTOH, this explains why I thought posts IX and X were missing last time.
szabgab azawawi, do you think the parrot based syntax highlighter is fast enough for use ? 17:42
pugs_svn r27184 | moritz++ | [t] simplify a test to not use outer lexicals in classes 17:44
r27185 | moritz++ | [t] move oo/class_inheritance_become_infinite_loop.t to spec/
17:44 s1n1 left
moritz_ the file that I just moved also causes errors in rakudo 17:44
azawawi szabgab: yeah it is very fast; but ofcourse we lose STD's error reporting if we use it.
moritz_ with an error I haven't seen before
17:44 s1n1 joined
moritz_ "Null PMC access in get_string() 17:44
current instr.: 'parrot;P6metaclass;register' pc 654 (runtime/parrot/library/P6object.pir:434)
17:45 s1n1 is now known as s1n_yapc
moritz_ " 17:45
szabgab azawawi, then we should try to generalize it to be able to highlight any parrot based syntax
and we are on the right track to world domination
azawawi szabgab: sure once we get p6 refactoring working first... 17:46
szabgab ok, so tomorrow?
or next week ?
azawawi szabgab: no idea really :)
17:48 explorer joined 17:49 s1n_yapc left
huf is anyone working on p6 vim integration? 17:50
viklund huf: how do you mean?
moritz_ huf: literal++ did some very good work
huf: I don't know what else you'd need 17:51
masak huf: there's a perl6.vim in the Pugs repository.
moritz_ github.com/hinrik/vim-perl is the up-to-date location
huf hmm.. padre has its own editor, right? no way to embed vim/emacs?
szabgab huf, not likely but we have plugins for both vi and emacs mode 17:52
huf eh :*
:( i mean
szabgab just neither of them really for production use
moritz_ but perl6.vim is pretty good for syntax hilighting
szabgab yeah, I guess it is better for the vim addicts among us :-) 17:53
moritz_ rakudo: say (-8) % 5
p6eval rakudo 10f223: OUTPUT«2␤»
17:53 jaldhar left
huf i just dont see why we need another editor when there are two that can do everything :D 17:53
pugs_svn r27186 | lwall++ | [CORE] add Test::EXPORT::DEFAULT entries
r27186 | lwall++ | [Cursor] import from DEFAULT to fix 'use Foo::Bar; use Foo;' problem
17:54 s1n1 joined
PerlJam if padre were more vim-like, I'd probably use it more. 17:54
moritz_ huf: they don't make coffee.
huf moritz_: is that a challenge? :D
17:54 jaldhar joined
masak huf: writing your own plugins in Perl is quite an incentive. 17:54
17:54 barney joined
moritz_ huf: no, but you're surely invited to view it as such ;-) 17:54
huf :))))
szabgab huf, the simple answer is that neither vim nor emacs fits most of the people out there
huf odd people
;)
17:55 justatheory joined
huf o well, that means no padre and automatic refactoring and whatnot for me 17:55
szabgab yeah, but that's about 99.37% of the (potential) user
=-
+-
masak "szabgab turned to speak entirely in symbols" 17:56
szabgab but we would welcome someone to fix the vi plugin to make it really awesome
huf oh, i understand the need for a full-featured IDE by the time p6 is ocmplete
PerlJam szabgab: it sounds like you just said padre isn't for power users as they all use vim/emacs anyway
huf i just dont like to use more than one program for one purpose
editing text -> vim 17:57
szabgab power users can use Padre but I think most of them can also use vimacs
huf yeah, we'll probably cope
szabgab huf, funny I used to use similar explanation why I am using gvim on windows
PerlJam szabgab: I was just thinking that many of padre's cool features are geared towards power users, but maybe they're geared more towards making "ordinary" people power users. 17:58
szabgab but then I decided I can try to stop using it in order to build padre
pmichaud a thought: perhaps the purpose of an ide should be to help people become more powerful :-)
PerlJam unix is still my ide of choice :) 17:59
szabgab I think my primary intention is to make Padre really good for beginners. As I am not good enough to build such a tool, the way to do that is to make Padre interesting for power users who in turn will make Padre really awesome for everyone
PerlJam szabgab++
huf nah. they'll make another vim :D 18:00
18:00 Molaf left
szabgab huf, in a way yes I guess 18:00
18:00 Molaf joined
szabgab but with a default editing mode similar to notepad and co 18:00
PerlJam huf: vim is nice, but there are some things I wish it had that it does not.
huf PerlJam: yep
PerlJam huf: If textmate and vim had a kid, that would be the perfect editor.
pugs_svn r27187 | moritz++ | [t] merge syntax/decl_vs_assign_prec.t into spec/ 18:01
moritz_ Q: how do you know you're in a geek channel? A: the people around you talk about text editors having kids
szabgab PerlJam, I'd really like to see a list of all the awesome features of Textmate, preferably in the form of open tickets on padre.perlide.org/ 18:02
PerlJam szabgab: padre seems to have all of the awesome features I care about (except that it's not enough like vim :)
szabgab and/or we have a page listing features of padre that are linked to tickets
oh, nice to hear that 18:03
huf just make it an option to embed vim instead of padre's own editor
there's a simple ide that does that, dunno if it could be done in this case
szabgab huf, rindolf shopped around for embedding vim but did not come up with any solution
so for now the Padre::Plugin::Vi is the best bet 18:04
it has serious missing features but it does have the two modes of vim
18:05 bacek joined 18:07 jferrero joined, jferrero left
moritz_ 621 files in t/spec/, 30 left to be moved. 18:08
18:08 explorer left, nErVe left
masak moritz_: I just want to take this opportunity to say that I think your work on the test suite has been amazing. 18:09
18:09 explorer joined
moritz_ thanks 18:09
moritz_ happy
dalek kudo: 0e0671a | moritz++ | t/spectest.data:
another passing integration test
pugs_svn r27188 | kyle++ | [t/spec] Fix the Greek autoincrement test I wrote past my bed time 18:10
18:10 KyleHa joined
masak KyleHa: yes, never autoincrement Greeks past your bedtime. :) 18:11
18:11 fdgkldgjfg joined
KyleHa "x" means "hex", folks. 18:11
phenny KyleHa: 06:50Z <moritz_> tell KyleHa yes, it should be a runtime error
fdgkldgjfg where is perl 6? 18:12
KyleHa phenny: Thank you!!
moritz_ fdgkldgjfg: wherever you install it ;-)
PerlJam rakudo: "Perl 6 is right here!".say
p6eval rakudo 10f223: OUTPUT«Perl 6 is right here!␤»
masak rakudo: say "fdgkldgjfg: I'm right here!"
p6eval rakudo 10f223: OUTPUT«fdgkldgjfg: I'm right here!␤»
masak PerlJam: great minds think alike.
fdgkldgjfg rakudo fails, i want a perl6 that can pass spec
PerlJam indeed :)
masak fdgkldgjfg: yeah, know the feeling. 18:13
moritz_ too
huf o/
masak \o
fdgkldgjfg i know, i'll make my own
huf i want a time-machine
masak yes, you do that.
I want a cat.
huf among other problems, it'd solve this
PerlJam fdgkldgjfg: just wait in a cryogenic chamber for a while.
huf altho it'd stifle p6 development
if we could just bring back future versions
fdgkldgjfg freeze myself up in the mountains?
moritz_ fdgkldgjfg: feel free, and tell us where your project lives
fdgkldgjfg i will call it perlthon. 18:14
masak fdgkldgjfg: I'd love to have another Perl 6 implementation.
huf masak: get a cat. seriously. they are better than a million hotdogs.
masak huf: I know!
fdgkldgjfg masak: why, is rakudo bad iyo?
KyleHa Spec says "Decrement of 'A00' should fail". What does that mean? How do I test for that "fail"?
masak fdgkldgjfg: "bad" is a strong word.
PerlJam fdgkldgjfg: we just don't have enough of them. There's only 3 (I think)
fdgkldgjfg make a motivational post out of it?
masak: explain (i'm interested) 18:15
moritz_ KyleHa: I assume it will return some undef value
huf dont more implementations mean more implementation-dependent clusterfucks? like we have with c/c++ compilers?
;)))
moritz_ fdgkldgjfg: read masak's blog, he often points out some weaknesses
masak fdgkldgjfg: Rakudo keeps getting more and more great with every day. for real.
moritz_ fdgkldgjfg: as do I on my blog
masak fdgkldgjfg: but there are times when it still really sucks.
fdgkldgjfg please post a link, i'm bored at work
18:15 xinming joined
moritz_ use.perl.org/~masak/journal/ 18:15
perlgeek.de/blog-en/perl-6/ 18:16
masak why, thanks. :)
fdgkldgjfg sweet. i tried reading yegge's blog earlier... it actually made me go back to working
masak fdgkldgjfg: we'll get you un-bored at work in no time! 18:20
pmichaud moritz_: (RT #66826) the patch contains some executable code at the bottom... is that intentional, or is it to be stripped out when applying the patch?
moritz_ fdgkldgjfg: also planetsix.perl.org/ is a nice blog aggregator, covering most Perl 6 blogs out there
fdgkldgjfg a king among men, you are 18:21
and when my boss catches me i am merely researching
moritz_ pmichaud: to be stipped (actually forgot to remove it)
KyleHa fdgkldgjfg: How do you pronounce your nick?
masak fdgkldgjfg: yes, "future technologies in computing". :)
pmichaud moritz_: okay, thanks.
fdgkldgjfg fitzkerpatrick
masak KyleHa: as it sounds, of course. :P
pmichaud I suspect the problem is that the code for operator overloading tries to adjust the optable
viklund fdgkldgjfg: :D 18:22
moritz_ I also tried it with an already-parsed op (infix:<eqv>, same result)
pmichaud right
I think that's the same issue
PerlJam masak: I just saw your post regarding 3d svg from Jun 19. Very cool.
masak fdgkldgjfg: I think I'll call you "Fitz" for short.
PerlJam: thank you! I'm halfway to doing the next level of cool with regards to that. interactive 3d SVG. 18:23
moritz_ masak: do you know if current browsers support that at all?
masak moritz_: yes, I think most do.
SVG is fairly deployed nowadays. 18:24
and screw the rest. :P
sbp who maintains p6eval again?
(no information from /whois)
moritz_ mostly
masak very little 18:25
viklund not at all
sbp moritz_, masak, !viklund: would it be possible to have it inhabit another channel?
masak thinks so
moritz_ sbp: sure, if the inhabitants of the other channel don't oppose that
masak unless all the inhabitants are also bots.
sbp it's my channel so they'll like it or face the jolly fist of doom
or both, more than likely. and yes, they're mainly bots 18:26
masak bots don't have a say.
18:26 p6eval left
sbp don't be so sure; they can already form unions 18:27
18:27 p6eval joined
masak read 'they can already form unicorns' 18:27
PerlJam masak: that proves it ... you live in a fantasy land.
(or you've done too many drugs ;) 18:28
masak that reminds me. I've forgotten to feed my unicorn today.
also, I hear Perl 6 is very big on the Unicorn standard. 18:29
sbp hehe 18:30
viklund haha, ha, aaa... 18:31
pugs_svn r27189 | kyle++ | [t/spec] Added some more greek autoincrement tests
masak slaps viklund
@slap viklund
lambdabot pokes viklund in the eye
viklund hey! 18:32
masak hides
18:33 xinming_ left 18:35 KyleHa left 18:38 eternaleye joined 18:39 _jaldhar joined, stDavid joined 18:40 jaldhar left
moritz_ rakudo: say "\x[391]\x[3c9]"; 18:40
p6eval rakudo 10f223: OUTPUT«Αω␤»
18:41 barney left
moritz_ rakudo: say "\x[392]\x[3b1]" 18:41
p6eval rakudo 10f223: OUTPUT«Βα␤»
18:41 s1n1 left
pugs_svn r27190 | lwall++ | [STD] unbreak the check for do {} while 18:42
18:45 stDavid left 18:46 athomason joined 18:47 perlmongo left 18:51 alester left 18:55 eternaleye left
sbp confusing error message: 18:58
rakudo: my @a = 1...{ $_ >= 3 ?? () !! $_ + 1 }
p6eval rakudo 0e0671: OUTPUT«Use of uninitialized value␤Use of uninitialized value␤Method 'postcircumfix:{ }' not found for invocant of class 'Failure'␤»
sbp code from t/spec/S03-operators/series.t
masak sbp: a very common error message, if you ask me. 19:00
it means you're trying to index undef.
moritz_ sbp: that's because rakudo doesn't parse infix:<...> correctly
sbp: I have a patch here that changes that, but it's not helpful either 19:01
because it then says "null PMC access in find_method"
sbp masak: where does the indexing take place?
moritz_: ah, fair enough. no problem that ... isn't implemented
was just entirely baffled by the error
19:01 bacek left
sbp I'm mainly waiting for the sequence guessing 1, 3, 5 ... and so forth 19:02
just because it's going to be a fun thing to demonstrate to people
masak sbp: I don't know. 19:03
sbp depending on whether the level of smarts it has exceeds people's boredom limitation... :-)
which, thinking about it, is probably quite unlikely
especially due to www.research.att.com/~njas/sequences/
19:04 sw1sh left
moritz_ sbp: the sequence guessing will be very boring, because it covers only constant differences and factors 19:04
sbp so, not even prime numbers? 19:05
moritz_ no. 19:06
sbp so ... isn't even going to be as powerful as ^1?$|^(11+?)\1+$? :-)
moritz_ that's a good task for a module
sbp :-)
TimToady std: do { s/.// } while $_; 19:07
p6eval std 27190: OUTPUT«##### PARSE FAILED #####␤Obsolete use of do...while; in Perl 6 please use repeat...while instead at /tmp/8bLJkBM9Ea line 1:␤------> do { s/.// } while $_;␤FAILED 00:04 37m␤»
19:08 azawawi left 19:11 stDavid joined 19:19 stDavid left 19:25 stDavid joined
TimToady n 19:26
masak ∞ 19:27
19:31 meppl left
lichtkind isnt :M file modification time - script start? 19:31
moritz_ it is, see S32::IO 19:32
TimToady well, that's just how it was copied over from P5; not sure forcing it into that syntax really makes sense in P6 19:34
lichtkind moritz_: i red that Syn but oh wait its numers not timepoints 19:36
19:45 justatheory left 19:49 ejs joined 19:56 fdgkldgjfg left 19:57 jan_ left 20:01 explorer is now known as Guest88545 20:02 Guest88545 left, jferrero joined 20:07 stDavid left, justatheory joined, jan_ joined 20:11 ejs left 20:18 REPLeffect left 20:21 alester joined 20:26 jaldhar_ joined 20:27 _jaldhar left 20:35 stDavid joined, mizioumt left 20:39 _jaldhar_ joined, jaldhar_ left 20:41 frew left 20:56 frew joined
masak [particle]: ping 20:59
sjohnson masak: you there? 21:07
masak sjohnson: yes, for a while longer.
then I'm going to bed.
lichtkind moritz_: cached trait bedeited memoized? 21:08
sjohnson masak: isn't perldoc somewhat close to perls commandline help? 21:09
PerlJam "bedeited"?
moritz_ lichtkind: ENOPARSE
masak sjohnson: not sure I understand the question.
perldoc is a command-line tool for getting Perl help, yes.
it's also a subdomain to perl.org 21:10
sjohnson oh, it was the blog i read
about how you thought 1) an irb thing would be nice and 2) a command-line help like ruby
lichtkind moritz_: eno? I HAz found it now in syn6 now anyway :)
21:10 stDavid left
masak sjohnson: yes, but perldoc is Perl 5 only. I'm talking about Perl 6. 21:10
sjohnson gotcha
masak sjohnson: and my vision for u4x is only partly overlapping with perldoc.
sjohnson well, just so you know, i agreed with everything you wrote 21:11
i would also like to see those things
moritz_ lichtkind: that meant "I couldn't parse your sentence"
masak good. let's build it together.
sjohnson well, i would offer my services in any way i can
lichtkind moritz_: understood but i dunno the call ENOPARSE, i asked for meaning of sub trait cached 21:12
masak sjohnson: do you have Pugs commit access? 21:13
sjohnson nope
masak sjohnson: could you /msg me your email address?
sjohnson i can't write in Haskell either
moritz_ there's much more than pugs to the pugs repo 21:14
including documentation project(s)
masak not necessary to know Haskell.
sjohnson: there should be an email in your inbox now. 21:17
sjohnson: welcome aboard!
sjohnson thanks
!
masak sjohnson: if you're free at 19 UTC tomorrow, there will be an inspirational meeting on #perl6-soc that you could come to.
we'll be talking about grok, the tool that will deliver u4x. 21:18
now I've got to dash.
21:18 masak left 21:19 cottoo left 21:26 _jaldhar_ left 21:35 frew left 21:44 viklund left, icwiener_ left 21:46 bacek joined 21:47 jferrero left 22:04 cotto joined 22:05 bacek left 22:07 hatseflats left, pmurias left
lichtkind laydies: Appendix A has now 400 entries 22:09
22:10 eternaleye joined 22:11 eternaleye left 22:16 hatseflats joined 22:22 bacek joined 22:25 frew joined 22:30 cotto left 22:32 lichtkind left 22:36 iblechbot left 22:37 jferrero joined 22:53 nihiliad left 23:03 szabgab left 23:04 szabgab joined 23:21 xomas left 23:25 zamolxes joined 23:27 [hrx] joined 23:29 payload joined 23:32 tarpsocks joined
tarpsocks hey 23:32
sjohnson hi 23:33
pugs_svn r27191 | lwall++ | [STD] more error errors 23:35
TimToady ho 23:37
23:37 Limbic_Region joined 23:39 xomas joined 23:42 cdarroch left 23:44 [hrx] left
sjohnson so... TimToady, is Perl community suggestions still in the works? 23:44
i think string editing via $string[3] type access, would stop the Python-based whiners
23:44 twigil1369 left
sjohnson also also, the Perl-based whines (me) too ;) 23:44
23:46 Whiteknight joined
sjohnson using subst to grab 1 char, using s// for 1 char, and split (//, $string) is kind of tedious for certain applications :[ 23:47
23:59 bacek left, drbean left

Powered by Cro and the Raku® Programming Language.

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

Thank you!