pugscode.org/ | nopaste: sial.org/pbot/perl6 | ?eval [~] <m oo se> | We do Haskell, too | > reverse . show $ foldl1 (*) [1..4] | irclog: irc.pugscode.org/
Set by diakopter on 11 July 2007.
00:02 dvorak joined, lambdabot joined 00:06 ofer joined 00:08 justatheory left 00:09 xdg joined 00:11 xdg left 00:12 xdg joined
pugs_svnbot r17097 | Darren_Duncan++ | ext/Muldis-DB/ : updated Language.pod concerning transaction specifying and scope 00:14
diff: dev.pugscode.org/changeset/17097
lambdabot Title: Changeset 17097 - Pugs - Trac
dduncan can I assume that pugscode.org is not on feather? 00:15
or is it?
actually, probably not
or the commit wouldn't have worked
diakopter it is on feather.. 00:18
Juerd Routing still works
I'm in the process of building a new router
So far, I've screwed up twice already (I blame fatigue) 00:19
00:19 drupek12 left
Juerd First forgot to zero out the partition table 00:19
And the second time I forgot that newer knoppixes use newer mdadm's, that default to a newer superblock version, that lilo cannot handle yet 00:20
It's now booting for the third time.
00:20 drupek12 joined
pugs_svnbot r17098 | Aankhen++ | * [irclog] a few more abbrs. 00:23
diff: dev.pugscode.org/changeset/17098
lambdabot Title: Changeset 17098 - Pugs - Trac
Juerd If I do everything right (probably not) you will experience not more than a network glitch 00:24
If I'm fast enough (probably not) existing ssh sessions won't even die
MAKEDEV-- # takes 55 seconds! 00:25
Dear whoever developped MAKEDEV, please don't do all logic in shell scripting. It's incredibly slow.
Dear myself, make a tarball of whatever MAKEDEV makes, and re-use that. It's the same every time anyway. 00:26
Base system installed; installing kernel 00:27
Dear Debian, why the hell is lilo executed before the kernel is installed? 00:28
Restoring config 00:32
Aankhen`` Dear Juerd, I CAN HAZ CHEEZBURGER? Love, Aankhen. 00:35
Juerd No, sorry, but I can make you a cheese sandwich 00:37
Come to Einsteinstraat 67, 3316 GG Dordrecht, The Netherlands 00:38
Here goes 00:41
If you want to ping, ping 193.200.132.2 and feather
:)
00:44 devogon left 00:46 d4l3k_ joined, ruoso left, devbot6 left, devbot6 joined 00:47 d4l3k_ left
diakopter my two ssh sessions remained up WHEE!!!!!!!!!!!!!!! 00:47
Juerd re
That was...
...scary :)
00:47 d4l3k_ joined, dalek left 00:48 d4l3k_ is now known as dalek, buetow left 00:51 nipotaway is now known as nipotan 00:55 Schwern left 01:05 ofer left 01:11 ofer joined 01:19 ashelyb_ left
diakopter Juerd: I was exaggerating my elation. sorry to frighten you. 01:19
Juerd I was frightened mostly by the PCI NIC that wouldn't fit. 01:21
Most of the downtime was actually caused by mechanical problems!
Fortunately, I carry a leatherman tool with pliers.
01:21 REPLeffect left
devbot6 planet6: perl6.announce: Parrot 0.4.14 "Now, with Seat Belts!" Released by jerry gay <www.nntp.perl.org/group/perl.perl6....1.html> 01:22
lambdabot www.nntp.perl.org/group/perl.perl6....1.html>
diakopter heh 01:23
devbot6: thank you
devbot6 diakopter: Error: "thank" is not a valid command.
diakopter devbot6: help 01:24
devbot6 diakopter: (help [<plugin>] [<command>]) -- This command gives a useful description of what <command> does. <plugin> is only necessary if the command is in more than one plugin.
01:31 Schwern joined 01:39 Schwern left 01:44 nperez joined 01:47 Blwood left 01:54 Schwern joined, Schwern left 01:56 ashleyb joined 02:04 weinig left 02:12 Alias_ joined, xinming left
Aankhen`` I may have asked this before, but I can't find where 02:16
ā€¦ what is {*} for in STD?
I think I've seen a few explanations but I couldn't really follow any of them.
diakopter Aankhen``: backlog a bit; it was discussed earlier today 02:17
Aankhen`` OK, thanks. 02:18
diakopter unless you were talking about the explanations from earlier today :) (sorry)
Aankhen`` Er, where was it discussed? 02:20
Oh, possibly while I was asleep.
Aankhen`` hits up the online los.
Logs, even.
Ohhhhhh. 02:23
[particle]++ # the ABC::Grammar stuff really helps. 02:24
02:45 ashleyb left, thorat left 02:46 ashleyb joined 02:47 mako132_ joined 02:48 jedai left 02:54 Schwern joined, bonesss left 02:55 weinig joined, jhorwitz left 02:57 ashelyb_ joined 02:59 Neophyte[Away] joined 03:04 ashleyb left 03:18 Schwern left 03:23 Alias_ left 03:45 amnesiac joined 04:07 mako132_ left 04:09 kanru joined 04:25 Schwern joined 04:28 Squern joined 04:29 Squern left, Squern joined 04:35 Alias_ joined 04:37 pro21ms4 joined 04:43 monomorph joined 04:49 Schwern left 05:11 devogon joined 05:14 Squern left 05:20 dduncan left 05:28 sunnavy joined 05:30 jisom joined 05:33 dduncan joined
moritz_ Aye, it does 05:38
05:39 dduncan left 05:46 stevan_ left 05:50 stevan_ joined 05:51 ashleyb joined 05:58 ashelyb_ left 05:59 icy-lisper joined
icy-lisper ?eval sub postfix: <!> ($n) { return [*] 1 .. $n }; say 8!; 06:01
evalbot_r17041 Error: ā¤Unexpected ": <!>"ā¤expecting "::", "handles", "is", bare trait, subroutine parameters, trait or block
icy-lisper ?eval sub postfix:<!> ($n) { return [*] 1 .. $n }; say 8!;
evalbot_r17041 OUTPUT[40320ā¤] Bool::True
moritz_ ?eval sub postfix :<!> { [*] 1..$^a }; 4! 06:04
evalbot_r17041 Error: ā¤Unexpected ":<!>"ā¤expecting "handles", "is", bare trait, subroutine parameters, trait or block
moritz_ ok, no whitespace allowed at all
?eval sub postfix:< ! > { [*] 1..$^a }; 4!
evalbot_r17041 Error: ā¤Unexpected "!"ā¤expecting "_", fraction, exponent, term postfix or operator
06:06 ashelyb_ joined
bloonix good morning 06:08
moritz_ good morning bloonix ;)
06:13 ashleyb left 06:24 elmex joined 06:34 amnesiac left 06:37 DarkWolf84 left 06:40 franck___ left 06:45 amnesiac joined, amnesiac left 06:48 iblechbot joined 06:51 hakobe is now known as hkb_lab 06:52 Neophyte[Away] left 06:59 franck___ joined 07:03 pro21ms4 left 07:23 ashelyb_ left, ashleyb joined 07:42 weinig left 07:43 Daveman left 07:47 flokuehn joined, franck____ joined 07:52 Schwern joined 08:04 franck___ left
moritz_ S02/"Immutable types" # irc log test 08:05
YaY, it works ;) 08:06
08:19 Schwern left 08:24 xinming joined
thoughtpolice oi, morning 08:27
hey moritz_
08:27 masak joined
moritz_ hey thoughtpolice ;) 08:28
08:28 moritz_ is now known as moritz 08:29 Schwern joined 08:35 pmurias joined
pmurias hi 08:35
08:38 pro21ms4 joined
masak hi pmurias 08:39
08:44 franck___ joined 09:01 Schwern left
thoughtpolice @tell [particle] i submitted my dragonfly smokes for 0.4.13; they're out of the tarball and not the svn repo, although I'm most likely going to get around to pulling a svn copy pretty soon and after which I'll set up a cron or somesuch to get the latest revision and run a smoke for you guys 09:01
lambdabot Consider it noted.
09:01 franck____ left
thoughtpolice damn, I'm behind the times. 0.4.14 was just released? 09:02
now i feel obsolete
moritz ;)
www.nntp.perl.org/group/perl.perl6....sg571.html
lambdabot Title: Parrot 0.4.14 "Now, with Seat Belts!" Released - nntp.perl.org, tinyurl.com/ypggma
moritz cool name, I have to admit 09:03
fitting for all the lint/splint things they have done
masak moritz: the name makes you not want to try the earlier versions
thoughtpolice i think i'll be fine
moritz masak: aye ;)
thoughtpolice i just need to drive safely :)
09:03 flokuehn left
masak thoughtpolice: :) 09:04
09:04 flokuehn joined
masak why don't you upgrade? 09:04
thoughtpolice masak: i'm actually running that smoke right now
moritz as someone on #parrot put it: "with seat belts, but not everone is buckled yet" ;)
thoughtpolice and it's out of a tarball
masak: after I get this smoke done I'll pull the svn repo and from there I can start doing regular smokes
masak thoughtpolice: ah, ok 09:05
thoughtpolice gah 09:06
09:07 buetow joined
thoughtpolice * Sending data to smokeserver "smoke.parrotcode.org/smoke/"... error: The submitted smoke has an invalid format! 09:07
screw it; I'll pull the SVN repo and see if I can get it done from there.
only now I also have to install svn. 09:08
oh, the woes
masak :)
thoughtpolice of having a brand new system, that is
moritz aptitude install subversion # debian++ 09:09
thoughtpolice my dfbsd box is somewhat in its infancy :(
cd /usr/pkgsrc/devel/subversion; bmake install
pkgsrc++
:)
09:09 franck____ joined
thoughtpolice i would use git-svn for doing it since I already have git on there 09:10
moritz I'm just compiling pugs on maschine by bloonix++ ;)
thoughtpolice but I'm not exactly sure of its overall mechanics in case I ever want to submit code
(since I'd probably want to use a branch for it) 09:11
moritz that's really cool, installing all dependencies within minutes
thoughtpolice i have to compile my stuff 09:13
on both my bsd boxes
but I really have no problem with it; I don't quite know why people complain about compilation so much
moritz i takes time 09:14
that's all
09:14 Aankhen`` left
thoughtpolice i guess the time of the compilation isn't too life-altering for me; I normally have other things I can partake in at the time :> 09:14
moritz and when I want ghc installed, I don't want to wait 3 hours until it finishes with bootstrapping itself ;)
thoughtpolice hah. i don't think it took quite 3 hours for 6.6.1 to install here 09:15
moritz but on my (slow) maschine
thoughtpolice overall I'd say it would have taken around there though, since I had to compile 6.2.4 from C first
moritz still, it will be more than an hour for you
thoughtpolice and then compile 6.6.1
moritz: this thing has 256mb ram with a p4 and like 09:16
8gb of hdd space (openbsd box)
moritz thoughtpolice: sounds like no fun to compile pugs on it ;)
thoughtpolice kinda
:x
09:18 dmq joined
thoughtpolice hm. i feel like coding something. 09:18
a lot.
moritz hack pugs. or kp6 09:19
thoughtpolice no, something more low level.
parrot mayhaps?
moritz yes
or perl6-on-parrot
thoughtpolice I haven't licked my C chops in a while
09:20 idiotmax joined 09:24 franck___ left, franck___ joined, chris2 joined 09:26 Schwern joined 09:27 nothingmuch joined, beppu joined 09:28 bloonix left, bloonix joined 09:36 franck____ left 09:44 zamolxes joined 09:56 agentzh joined
agentzh *** Cannot find GHC 6.6.1 or above from path (we have 6.6). 09:57
*** Please install a newer version from haskell.org/ghc/.
lambdabot Title: The Glasgow Haskell Compiler
masak agentzh: well, have you?
moritz agentzh: audreyt++ recently added the new dependency
agentzh moritz: feather still uses GHC 6.6, i'm afraid. 09:58
masak oh, it's feather
09:58 buetow left
moritz agentzh: then it's just wating for an update - feather is unstable, and I know that testing has 6.6.1 09:59
agentzh if anyone with root access to feather, please update the GHC there. :)
s/with/has/
masak who does have root on feather?
agentzh me :)
and a few others. 10:00
masak agentzh: so... if you have it...
moritz diakopter, Juerd , TimToady , adureyt
s/du/ud/
agentzh well, i'm currently on holiday and the windows box i'm using doesn't have putty... 10:01
10:01 aindilis left
moritz windows-- 10:01
masak agentzh: putty is downloadable on the web
moritz it should be made a crime to ship an OS without ssh 10:02
agentzh masak: i know that part...but i'm just too lazy ;)
masak ah, ok
10:02 aindilis joined
agentzh the web connection is very bad here, especially while using CGI:IRC...i'm off :) & 10:04
10:04 agentzh left 10:13 iblechbot left
thoughtpolice @karma windows 10:13
lambdabot windows has a karma of -6
thoughtpolice hah.
10:14 xdg left 10:18 hcchien joined 10:24 franck____ joined 10:40 masak left, franck___ left, franck___ joined 10:50 franck____ left 10:54 iblechbot joined 10:56 awwaiid left, awwaiid joined 11:04 masak joined
meppl gugu 11:10
TimToady I always carry a copy of putty on my USB stick... 11:11
moritz: the postfix:< ! > form ought to work
since it should canonicalize to postfix:{'!'} 11:12
moritz I thought so
TimToady but the other forms don't admit whitespace
moritz the <...> is just the list constructor, irght?
TimToady it's the qw// naked form as used to represent constant subscript slices 11:13
hash slicing syntax, essentially
so circumfix:< ( ) > turns into circumfix:{'(',')'} 11:14
moritz ok
TimToady I use extra whitespace in STD for instance
thoughtpolice @tell [particle] scratch that last message; I just pulled parrot's SVN repo and the smokes went great and they've been submitted; I'll continue smoking future revisions for you guys on dragonfly.
lambdabot Consider it noted.
moritz TimToady: did you read [particle]'s suggestions about using grammar subclassing for perlhints? 11:15
TimToady darn my compose key is broken...
thoughtpolice 7255 OK from 7262 tests (99.90% ok)
almost there. :>
moritz buy a new one? *g*
thoughtpolice i need a new keyboard and chair
TimToady I think that would help in I18n eventually
or some other kind of indirection
moritz and do you think it works? 11:16
thoughtpolice anybody have any recommendations?
i've heard microsoft's natural ergonomic keyboard 4000 is totally awesome
TimToady moritz: would probably depend on how the data structures actually attach 11:17
it's an interesting question whether a subclass can turn dumb comments into smart comments after the fact in a used module... 11:18
moritz well, the subclass would hold the data, and not as comments 11:19
TimToady sorry, was going off on a tangent
moritz never mind ;)
and should all tokens that actually match text (not assertions) actually get a {*}? 11:21
I noticed lambda has none, so I guess a compiler couldn't determine if -> or <-> was parsed
TimToady yes, lambda is an example of a missing {*} 11:23
it should be a separate token
moritz ok
TimToady from a grammar point of view, it could be processed in the rule above as $<lambda>, but you might want a hook in the actual rule 11:24
11:31 franck____ joined 11:32 mncharity left
masak I did a bit of grepping in t/ for tests relating to the bug we uncovered in pugs yesterday 11:43
?eval my $a = 'word'; --$a; $a
evalbot_r17041 \-1.0
masak didn't find anything relating to ++ and -- of string values 11:44
TimToady and did you find any such tests?
masak TimToady: no
there's t/operators/inc.t
11:44 nipotan is now known as nipotaway
masak but it almost exclusively does numbers 11:44
so I'm adding some string ++ and -- tests if that's ok 11:45
TimToady masak++
moritz more tests are always good ;)
masak should anyone ever realize they're duplicates, I guess that's easily fixed :)
moritz masak: the smartlinks should show that easily 11:46
masak moritz: how? I know about them, but not where they are 11:47
11:48 franck___ left
moritz masak: when you look at perlcabal.org/syn/S02.html, you see those links like "- Show t/regex/smartparse.t lines 9ā€“14 (0 āˆš, 1 Ɨ) -" 11:48
lambdabot Title: S02
masak moritz: thx 11:49
moritz you're welcome
masak ah, found it 11:51
it's t/operators/auto.t
but all tests pass in that one, so I'm sure I could add something :) 11:53
sunnavy hi, does anyone encounter this: when `make test' about pugs, your box will be shutted down surprisingly. my os is Gentoo, Pugs is up to date and GHC is 6.6.1.
masak sunnavy: sounds very unfortunate. do you get any message about it? 11:54
moritz what do you mean by "shutted down"? as if you typed "halt"?
sunnavy moritz: yes
TimToady in the compile or in the smoke?
moritz sunnavy: are you compiling as root?
sunnavy: a normal user shouldn't have the right to shutdown 11:55
sunnavy moritz: I don't think so.
TimToady and if in the compile, how much memory on the box?
I'm thinking running out of swap space maybe
sunnavy TimToady: I think it's in the smoke.
moritz sunnavy: does that happen every time?
TimToady though should just give "out of memory" 11:56
sunnavy I tried 3 times :-(
moritz :(
TimToady same place in the smoke?
masak ?eval my $foo = 'A99'; ++$foo;
evalbot_r17041 "B00"
masak TimToady: is the reverse still true in p6?
i.e -- gives back A99
or are both false?
sunnavy it's suddenly shutted down, I even don't have time to look. 11:57
TimToady should go to A99, yes
masak ok
sunnavy Next time I'll stare at this
TimToady and A00 should fail
masak ok
moritz sunnavy: try to run 'script pugs_smoke.log' and then 'make smoke' 11:58
sunnavy: then you have a transcript after reboot ;)
TimToady you might force an fsck 11:59
masak why's it called "autoincrementing" anyway?
sunnavy thanks, moritz. I'll try again.
TimToady PDP-11
moritz masak: auto = self
masak moritz: ah, 'course
thought of it as 'automatic' there for a while
moritz what's "matic"? *g* 12:00
TimToady mov @R0++,@R1++ or some such
masak moritz: :)
TimToady: cool
sunnavy If I'm out of irc soon, you'll know what happened ;-)
moritz ;)
masak TimToady: did PDP-11 have pre- as well?
TimToady I think so, good for stack ops 12:01
pushes did --( 12:02
--(SP) and pops (SP)++
so stacks grew downward 12:03
masak as they should :)
TimToady which lets buffer overflows clobber your return address. :( 12:04
masak oops
maybe an argument for the harvard architeture after all
en.wikipedia.org/wiki/Harvard_architecture 12:05
lambdabot Title: Harvard architecture - Wikipedia, the free encyclopedia
moritz sunnavy: still there? *g*
pugs_svnbot r17099 | masak++ | [t/operators/auto.t] 12:06
r17099 | masak++ | * added tests for auto-decrementing of string variables
diff: dev.pugscode.org/changeset/17099
lambdabot Title: Changeset 17099 - Pugs - Trac
TimToady www.anvari.org/fortune/Laws_2/481_h...other.html
lambdabot Title: Harvard Law Under the most rigorously controlled conditions of pressure, tempera ..., tinyurl.com/yrvkoj
masak TimToady: :) 12:07
12:08 r0bby left 12:10 Daveman joined 12:15 jhorwitz joined 12:16 ashelyb_ joined 12:17 ashleyb left 12:18 sunnavy left, pro21ms4 left 12:27 DarkWolf84 joined
moritz is there a git repository clone of pugs anywhere? 12:31
masak www.ohloh.net/projects/3300 12:37
lambdabot Title: Pugs - Ohloh, the open source network 12:38
masak how come it says on that page that the Pugs project is mostly written in Perl?
wolverian it doesn't grok haskell 12:40
masak that explains it
who does, by the way? :)
moritz lol
wolverian try #haskell.. :)
moritz ghc and audreyt, I guess ;)
TimToady um, quite a few folks in this room...
masak true 12:41
wolverian just ask, don't ask to ask! (;
masak wolverian: is it just meta-asking that's not ok, or are questions like this fine? 12:42
wolverian I have a headache. 12:43
masak ehm, s/ or//
TimToady "Anything you can do I can do meta; I can do anything meta than you!"
thoughtpolice no you can't! 12:44
12:45 ashleyb joined
TimToady not me. that was from "Annie get your pun!" 12:45
thoughtpolice moritz: me kinda sorta. :(
moritz thoughtpolice: do you have a clonable URL for me?
thoughtpolice moritz: i meant in relation to 'grok haskell'
moritz: git has built in support for svn though 12:46
man git-svn
moritz thoughtpolice: ah, ok
thoughtpolice you'll need perl libs for SVN though.
but yeah it works
moritz thoughtpolice: I no about git-svn, but I don't want to stress feather's network more than necessary
thoughtpolice ah.
in that case i don't know of a git repo. :(
TimToady
.oO(Annie git your pun)
moritz maybe I should get a feather account to set one up 12:47
thoughtpolice moritz: i could probably see if the guy who hosts my website would let me use tailor and put a git repo there
moritz: the thing is getting in contact with him (he lives in sweden, I live in texas. damn geography and time zones :( ) 12:48
if not i could just put one on my feather account
moritz thoughtpolice: feather would be preferable, because then the initial checkout is not such a pain for feather
thoughtpolice moritz: then we could just use my account. 12:49
masak TimToady: that was a bit meta, too 12:50
thoughtpolice I've been meaning to also use it to host some pugs haddock documentation, but right now the haddock comments all aren't valid (although it wouldn't take much effort to make them valid I assume)
masak there should be a direct link to the Wayback Machine on 404 pages 12:51
12:52 mncharity joined
thoughtpolice moritz: when Juerd's around I'll ask him about it. 12:52
moritz thoughtpolice: I'm sending him an email right now to get a feather account 12:53
thoughtpolice ah cool
moritz thoughtpolice: i guess it's handy after all
12:54 ashelyb_ left
thoughtpolice moritz: a git repo would be nice i'd think. :> 12:55
git rocks pretty nicely. and plus I think I found my new favorite command: git-bisect
12:55 pmurias left
moritz ;) 12:56
binary search to trac regressions, right?
s/trac/locate/
thoughtpolice binary search to find which commit blob introduced a bug
which as you'd think is pretty useful, and using it is really really easy 12:57
just mark a good commit and a bad one, and start a bisection. it'll put you at various points in the commit history and you just tell git whether that commit at that point was good or bad. eventually you'll narrow it down to the commit that gave you your bug 12:58
12:58 stevan_ left, stevan_ joined 13:02 sunnavy joined
sunnavy uh, after halt for another 3 times, I got the reason: my box was too hot to bare the tests. :-( 13:04
13:05 drupek12 left 13:07 masak left, masak joined
szbalint sunnavy: it's always something simple isn't it? :) 13:08
13:10 Alias_ left, drupek12 joined, funktio joined
sunnavy szbalint: yeah. oh, my poor box 13:12
masak sunnavy: so once again it's confirmed: Perl 6 is hot 13:18
thoughtpolice @quote masak so once again it's confirmed: Perl 6 is hot 13:20
lambdabot No quotes for this person. Listen, broccoli brains, I don't have time to listen to this trash.
thoughtpolice er
@remember masak so once again it's confirmed: Perl 6 is hot
lambdabot Done.
thoughtpolice lambdabot insults hurt :(
masak thoughtpolice: thx. the quote sort of loses its punch outside of the context of sunnavy++'s box, though
diakopter thoughtpollice: re: "no you can't!" - "Yes, I can!" ;) 13:22
thoughtpolice diakopter: no you can't!
diakopter Yes I can! 13:23
thoughtpolice you win. :(
diakopter :P
too bad I don't know any more lyrics to that song
thoughtpolice i remember that part because I see commericals with it in them when I actually watch TV 13:24
which is somewhat rare at best. tv really sucks. :(
13:24 franck___ joined 13:25 mj41 left, nperez left 13:26 ashelyb_ joined
diakopter I had *thought* I'd already updated feather to 6.6.1... but I'll upgrade it now if not. 13:26
13:27 nperez joined 13:28 nperez left
diakopter oh. it's b/c someone put a 6.6 build in /usr/local/bin 13:28
yes, MS Keyboard 4000 rocks, in reply to whoever said that. 13:29
13:29 mj41 joined
diakopter k.... 1;0 root@feather:~# which ghc 13:31
1;0 root@feather:~# ghc --version
The Glorious Glasgow Haskell Compilation System, version 6.6.1
13:35 ashleyb left, sunnavy left
thoughtpolice diakopter: me 13:37
yeah that's what I've heard about the 4000 (some say it's the best keyboard they've ever used)
and at $40 I'd be willing to pay for that, I've heard a lot of praise
masak diakopter: IBM Model M keyboards rock too, but in an orthogonal way
13:38 xdg joined
thoughtpolice i used to have a legendary IBM clicky 13:38
unfortunately when we got rid of our old DOS system my father used it was also taken away. :(
13:39 franck____ left
diakopter I dunno if I agree with that. The 4000 is kinda loud; I learned to type on a rusty typewriter and I'm a guitarist, so I tend to strike each key with about 80N. 13:39
thoughtpolice hah.
diakopter I had an old compaq ps/2 keyboard that was whisper quiet, even with that style of typing.
thoughtpolice well I'm mainly looking for something nice and ergonomic and I've heard lots of praise for the NE 4000 13:40
diakopter old meaning 1990
thoughtpolice I also need a chair too, because this one really sucks.
any recommendations on that note?
i've never had a really good keyboard or chair and everybody tells me it makes a world of difference. to be honest, I'm pretty sick of this chair/keyboard anyway. 13:42
diakopter I knew a guy during college who had a keyboard with the LHS and RHS on opposite sides of a vertical plane. 13:43
thoughtpolice i also probably need to switch to dvorak one of these days 13:44
13:44 franck____ joined
masak thoughtpolice: amen. me too 13:47
dvorak.nl/ 13:48
lambdabot Title: dvorak.nl (aoeu.nl)
diakopter I still say someone should produce/sell a keyboard optimized for Perl 6. :)
13:49 ashelyb_ left
diakopter sure, the market would be small, at first. 13:49
masak diakopter: design one and put the design notes online 13:50
somebody will be tempted enough to build one
and then the ball is rolling
moritz @seen sunnavy 13:55
lambdabot I saw sunnavy leaving #perl6 19m 20s ago, and .
masak I was never sure what lambdabot wanted to say after the "and" 13:56
moritz masak: probably the parting message, if there was one
diakopter I *hate* these keys: Shift, Ctrl, Alt. Why can't the space bar be much smaller (or elsewhere) and let the thumbs operate Shift, Ctrl, Alt toward the center of the bottom row? continually torquing the wrists or 4th fingers to reach Shift to Shift the number row, or to reach Ctrl, Alt, or Apple/Command/Windows is torturous. 13:57
13:58 iblechbot left 13:59 franck___ left
diakopter rants about people who use irc channels as their personal rant blogs. 13:59
masak
.oO( modular rebuildable keyboards )
 14:00
Patterner my space key is the same size as the enter key. and I use thumbs for ctrl, alt and backspace...
szbalint I simply have big hands. :) 14:01
diakopter Patterner: does your enter key touch the ;: key? or is the '" key there?
14:01 xdg left
Patterner the enter key is surrounded by page down, page up, alt, ctrl and space 14:01
diakopter I haven't seen that keyboard 14:02
awwaiid I was just pondering about the use of ';' to separate statements and how it is so conveniently located under the pinky
moritz just a minute ago I read about a german startup where you can design and order your own keyboard layout...
awwaiid and what a waste that key is for non-code
Patterner diakopter: kinesis ergo elan
moritz and that's printed and attached as a USB keyboard
Patterner the russion optimus keyboard? 14:03
masak moritz: URL?
moritz masak: flippress.com/ 14:04
masak: I think they just started, I don't know if you can already order the keyboards
diakopter Patterner: *wow* those Kinesis ones look cool. 14:05
with a 3-pedal foot switch! 14:06
Juerd I have a Kinesis Ergo Contoured keyboard
I love the form factor, but the cherry switches are too heavy.
(require much force. I don't know if "heavy" is the proper term.) 14:07
diakopter I almost suspect that a pipe organ rank would enable faster typing than a keyboard. and sound better.
masak :) 14:09
diakopter use v6::mixolydian;
14:09 TJCRI joined
Patterner Juerd: i have no problem with mine 14:10
14:10 franck___ joined
diakopter of course, that would have to be editor-time enabled. 14:11
Juerd Patterner: Your hands are probably not as ruined as mine
Patterner probably... although I bought mine because I started to get joint aches... 14:12
14:15 icy-lisper left
thoughtpolice Patterner: help much? 14:20
Patterner yes. no problems since I got it. 14:21
14:22 rhandom joined
moritz thoughtpolice: if you have much space, you could use an old car seat 14:22
thoughtpolice: they are designed so that you can sit quite a long time in them, and you can get them rather cheaply
thoughtpolice: ... at your nearest scrap yard 14:23
14:24 rashakil left 14:26 franck____ left
[particle] Juerd: did you get my testimonial? 14:26
lambdabot [particle]: You have 2 new messages. '/msg lambdabot @messages' to read them.
thoughtpolice moritz: hah
14:26 ashleyb joined
moritz thoughtpolice: I read that once on /. and it sounded like a good idea to me ;) 14:26
thoughtpolice i'd get some awesome seat covers then
[particle] thoughtpolice++ # the parrot team thanks you for smoking parrot 14:27
[particle] showers &
thoughtpolice [particle]: np. i also did a smoke for the languages
[particle] yay
thoughtpolice on the regular tests you have 99.9% passing
on the languages it was about 90% flat
so pretty good. :)
14:28 rdice joined
moritz the parrot team has pretty good QA ;) 14:31
I think I never say a parrot smoke on my maschine with more than 15 tests failing
and only once or twice one with more than 5
rdice moritz: kid51 (Jim Keenan) and Andy have really buckled down on that over the past few months. And good on 'em. 14:32
Juerd [particle]: That depends on your name :)
afk
thoughtpolice moritz: for the regular tests dfbsd only failed 7
moritz kid51++, Andy++
thoughtpolice (0.4.14) 14:33
on the languages ~200 failed
but it's getting there at least. :)
moritz and I have a rather common system ( debian linux on i386 )
thoughtpolice ouch
14:33 franck____ joined
thoughtpolice on fbsd 0.4.14 the languages only passed with 37% ok 14:34
almost all of the lua, apl, tcl and regex tests failed. 14:35
14:35 r0bby joined
diakopter there are studies that conclude seat belt usage generally causes individuals to drive more daringly (because of the increased sense of "safety"), but without the increased alertness that more daring drivers utilize, thence more wrecks/injuries for seat belt users and their victims. 14:37
thoughtpolice people are dumb. 14:38
really.
moritz thoughtpolice: FULL ACK ;)
diakopter is a daring driver.
people are zombies 14:39
thoughtpolice getting eaten by zombies would suck. :(
diakopter hopes the increased seat belts in parrot cause parrotfolks to be more daring (and alert). :) 14:40
moritz thoughtpolice: unless they are attractive female zombies ;)
diakopter (not that they are lacking in dare/alertness currently; I'm just trying to salvage the metaphor.
thoughtpolice in any case, I need to go mow my lawn and shower before it gets too blazingly hot outside and i turn into proverbial fried chicken 14:41
thoughtpolice --status=away --reason=chores &> /dev/null &
[particle] diakopter: just cause we have seatbelts doesn't mean folks are using 'em. that's next. then airbags and abs. :) 14:43
Juerd: [email@hidden.address] # i'll let you guess my name :)
great, now i'm hungry for some pfc. thanks, thoughtpolice. 14:44
Juerd [particle]: Then I got it :)
[particle] great 14:45
Juerd Indeed :)
[particle] i suspect frequency analysis of the perl 6 test suite would reveal that many sigiltwigil and other metacharacters belong in unshifted kb positions 14:48
14:48 franck___ left
[particle] how'd you like <shift-left index> := 4 14:48
moritz I want a foot pedal for $ and @ 14:50
[particle] and ;
args you have only two feed, haven't you 14:51
moritz [particle]: that's better reachable
[particle] i'm heavily biased towards left-pinky-shift
i've probably used right-pinky-shift 20 times in 10 years
14:52 cerridwen left
moritz thinks about a second layer keyboard above the first one that can be operated by pushing the hands upwoards 14:52
14:54 Patterner left 15:02 kanru left 15:09 flokuehn left 15:10 daxim_ left 15:15 pro21ms4 joined, DarkWolf84 left
moritz can anybody over his/her p6 OO wisdom on that thread: www.perlmonks.org/?node_id=627130 15:26
lambdabot Title: &#91;Perl 6&#93; Object methods on the fly?
masak interesting question 15:28
15:40 REPLeffect joined 15:45 franck____ left 15:58 masak left 16:04 ashelyb_ joined 16:09 Blwood joined 16:13 franck___ joined 16:14 ashleyb left 16:17 weinig joined 16:25 TJCRI left 16:26 TJCRI joined 16:37 justatheory joined 16:45 TJCRI left 16:46 randomity joined 16:47 aindilis left 16:51 TJCRI joined 17:10 buetow joined 17:19 bonesss joined 17:21 aindilis joined 17:29 dduncan joined 17:31 Blwood left 17:39 randomity left, pro21ms4 left 17:41 Psyche^ joined, Psyche^ is now known as Patterner 17:56 iblechbot joined 18:00 pmurias joined 18:01 buubot joined 18:04 ashelyb_ left 18:09 joao joined 18:20 rho joined 18:21 offby1 joined
rdice I've a perl6 technical/philosophical question I'd like some help with, in case anyone out there has insights I don't on something. 18:27
What are references (well, captures) good for in p6? I mean, when I think about what I use them for in p5 it's a) param passing of arrays and hashes to subs, and b) creating complex data structures. 18:28
But in p6 we've got a) real param definitions for subs and b) "flattening" of arrays and hashes isn't the behaviour, so you can stick your arrays and hashes into each other directly. 18:29
18:29 REPLeffect left
rdice So what's left? (Captures are cool for currying and co-routines and such I guess.) 18:29
18:30 REPLeffect joined
TimToady basically, whenever you need to delay binding an argument list 18:31
suppose you want to pass an argument list to a switch statement
in order to to serial matching rather than unordered matching like mmd does 18:32
mostly you get capture objects in wrappers that don't want to bind
18:32 bonesss left
TimToady sub wrappush (|$capture) { ... push |$capture; ... } 18:33
that doesn't need \ of course
Match objects are also derived from Capture
so you could fake a match with $/ := \($foo, $bar, :baz(42) ) 18:34
(wouldn't have .from or .to attrs tho) 18:35
18:37 crem_ left
rdice Thanks TimToady. 18:37
18:37 crem_ joined
rdice As suspected, references/captures exist and can be used in p6 as they are used in p5, but if you're using them that way then you're not doing idiomatic p6. And the idiomatic p6 uses of captures barely even have analogs in p5. 18:39
TimToady they're almost like (shh!) anonymous typeglobs 18:41
18:43 SportChick joined
rdice That thought had occurred to me. :-) 18:45
TimToady which is another argument for not using them much. :) 18:46
rdice Captures seem to serve as an awesome "grand unified theory" of several underlying aspects of p6, to the point where they're basically an object of semantics made concrete, and bordering on implementation. In that sense I'm astonished by their beauty. 18:52
But as just another perl hacker, I'm still struggling to see how I might ever use them. But I felt that way about closures once too.
Maybe when I finally bootstrap myself into functional programming I'll have a renewed appreciation of what I might do with them.
18:52 offby1 left, funktio left
TimToady they're essentially the complement to the signature matching that is pervasive in FP 18:53
It's kinda strange to be able to say $foo ~~ $signature and not have a type for $foo 18:54
and my (...) := $foo is essentially the same 18:55
well, I guess it'd have to be |$foo there 18:56
18:57 ashleyb joined, ashleyb left
TimToady dinner & 18:59
dduncan way off topic, but if there are any Mac users among you (as am I), or you know others who are, I recommend checking out this time-limited bundle deal (of non-free software) at www.macupdate.com/ 19:09
lambdabot Title: MacUpdate: Macintosh Software & Games
19:13 penk joined, chris2 left 19:19 bonesss joined, jisom left 19:31 joao left
wolverian ?eval map { "$a -- $b" }, "a".."c", "d".."f" 19:33
evalbot_r17041 (" -- ", " -- ", " -- ", " -- ", " -- ", " -- ")
wolverian erm.
?eval map { "$a -- $b" }, "a".."c" Z "d".."f"
evalbot_r17041 ((" -- ", "d"), (" -- ", "e"), (" -- ", "f"))
wolverian ?eval map -> $a, $b { "$a -- $b" }, "a".."c" Z "d".."f"
evalbot_r17041 (("a -- b", "d"), ("c -- ", "e"), (undef, "f"))
wolverian right. I don't get what's doing on here.
s/doing/going/
19:34 Blwood joined 19:38 rho left
Tene wolverian: you want $^a and $^b 19:39
Or perhaps that's not the part you were confused about.
?eval 'a'..'c' Z 'd'..'f' 19:40
evalbot_r17041 (("a", "d"), ("b", "e"), ("c", "f"))
wolverian shouldn't map -> $a, $b { ... } work equally well? 19:41
(I tried the ^ variables in pugs shell, which failed to parse them altogether, so I didn't try them here.) 19:42
Tene Yes, it should. I suspect a grouping issue.
?eval map -> $a, $b { "$a -- $b" }, ("a".."c" Z "d".."f")
evalbot_r17041 ("a d -- b e", "c f -- ")
wolverian heh, it's not doing structural expansion into the tuples 19:43
I suppose we want map -> [$a, $b]?
(or ()?)
Tene ?eval map -> [$a, $b] { "$a -- $b" }, ("a".."c" Z "d".."f")
evalbot_r17041 Error: ā¤Unexpected "[$"ā¤expecting subroutine parameters, trait or block
Tene ?eval map -> ($a, $b) { "$a -- $b" }, ("a".."c" Z "d".."f")
evalbot_r17041 ("a d -- b e", "c f -- ")
wolverian maybe that's the ticket, pugs just doesn't parse it yet.
thoughtpolice it seems as though what's happening is map doesn't seem to taking a tuple such as ('a','d') and being able to feed it appropriately into the given sub 19:49
19:53 dduncan left 19:54 Aankhen`` joined 20:07 dduncan joined
diakopter ?eval map -> ([$a, $b]) { "$a -- $b" }, ("a".."c" Z "d".."f") 20:15
evalbot_r17041 Error: ā¤Unexpected "[$"ā¤expecting formal parameter or ")"
wolverian good try :) 20:16
diakopter ?eval map -> ($a => $b) { "$a -- $b" }, ("a".."c" Z "d".."f")
evalbot_r17041 Error: ā¤Unexpected "$a"ā¤expecting ")"
20:17 devogon left 20:18 franck___ left 20:19 mr_ank left, devogon joined
rhr ?eval map -> $a { "$a[0] -- $a[1]" }, ("a".."c" Z "d".."f") 20:35
evalbot_r17041 ("a -- d", "b -- e", "c -- f")
rhr I think -> [$a, $b] should work 20:36
wolverian right.
TimToady Z is not supposed to add [...] by default
20:37 jedai joined
wolverian I misread S29 then, I think 20:37
TimToady S29 might be wrong
wolverian would it be map ..., $ .. Z ... then?
(that is, if I really wanted the tuples) 20:38
(for some unspecified reason)
TimToady I think you want @@ or slice
or just take them two at a time from the flat list
wolverian ah, right. 20:39
hm. from S03: $ map { $_, $_*2 }, ^4 # [\(0,0),\(1,2),\(2,4),\(3,6)]
that confused me. :)
TimToady only one arg there, but spits out captures 20:40
wolverian right.
what's the name of the [] there? capture? reference?
TimToady it's an array 20:41
wolverian good.
TimToady a list in scalar context makes one
and inside of [] provides @ context, so flattens
wolverian Z makes me want an operator version of map 20:42
a la mathematica
TimToady which side does the closure go on? 20:43
wolverian mathematica's goes on the right side
I was thinking the other way around, I think
TimToady @array.map: {...} works
lambdabot Unknown command, try @list
wolverian maybe have two of them ;)
right. it was just for consistency in map ..., ... Z ..., which is not a very good motivation. 20:44
TimToady hmm, yeah, Z eats lists on both sides
well, {...} for ... Z ... is supposed to work the same eventually 20:46
and ... Z ... ==> map {...}
wolverian good point 20:47
multi zipwith (&f, *@ls) { { .reduce(&f) } for @@ zip @ls } # look right? 20:48
20:50 thorat joined, TJCRI left
TimToady I think zip @ls is a noop 20:53
wolverian hm 20:54
makes sense
zip @@ls?
wolverian reads
TimToady yes, if @ls is a list of something listy
wolverian well, the *@ls was meant to imply that, though I realise I didn't specify it explicitly 20:55
(...surely the type inferencer will do this for me! or so)
been doing haskell.
TimToady the trouble with a really smart compiler is that it can tend to require the programmer to be just as smart to outsmart it 20:56
wolverian I don't need to outsmart haskell, I just need to understand what it's doing. which is not a trivial task, indeed.
TimToady heh, I was just sitting next to Simon Peyton-Jones at dinner...
wolverian and the space-time continuum didn't have an epileptic seizure? hm. 20:57
TimToady nope. though one of the other hardcore FP folks here told him earlier today that Haskell is the Perl of FP. :)
20:58 lisppaste3 left
TimToady meaning somewhat accretive, I believe. 20:58
thorat at least the way Oleg Kiselyov uses it :)\
wolverian ghc is accretive, at least
the haskell standard is pretty conservative (boring) 20:59
lumi So ghc is the perl of FP?
20:59 ashleyb joined
TimToady maybe of fp 20:59
SamB why is it the perl? because Baughn doesn't like the syntax? 21:00
wolverian there isn't much syntax to speak of, though the lack itself can certainly be hateful 21:01
TimToady I suspect it has more to do with trying to solve all the problems of the world.
wolverian I particularly don't like how everything in haskell looks the same.
perl does this much better.
TimToady I think we're better at conserving identifiers too
thorat (we're back to chipped fingernails in oatmeal metaphor)
TimToady little of this for vs forM business 21:02
wolverian except haskell doesn't have the fingernails...
TimToady they're invisible
thorat right
wolverian TimToady, right, though perl doesn't have that division, to be fair.
SamB TimToady: wth is for?
TimToady I meant map vs mapM 21:03
SamB oh, that
TimToady adding monads tends to increase the alphabet soup
thorat (I guess I meant simile)
SamB well, they have rather different types
wolverian I thought ghc 6.6 added for/forM, but, yeah.
TimToady admittedly, inheriting from different classes in OO tends to use up names too 21:04
wolverian (being just flip . map)
SamB I mean mapM f = sequence . map f, doesn't it?
21:04 lisppaste3 joined
wolverian SamB, that's the division... sequence being the monad 21:04
TimToady but the fact that most everything is definitional means you have to name a lot more definitions
wolverian namespaces would work too 21:05
M.map instead of mapM 21:06
TimToady but maybe I'm just going crosseyed at all the CamelCase
wolverian (though M. is usually used for Data.Map, so...)
SamB M is like a sigil at the wrong end of the variable ;-P 21:07
TimToady so Haskell is really the BASIC of FP :)
SamB heh
wolverian hm, how about enforcing a sigil for monads
SamB I thought that was called Visual Basic 21:08
wolverian $monadAction instead of monadAction
I'm sure that'd fly... (except $ is taken)
SamB $ is already taken
twice!
wolverian no $$ :)
SamB also, what are the typing rules for your proposal? 21:09
wolverian I don't know if it has any that differ from the status quo, though I see how it would apply to uniqueness typing too
visually anyway 21:10
meppl good night
wolverian please don't tell anyone I proposed sigils for haskell though, I'm sure they'd kill me.
SamB I won't, because your proposal is incomplete ;-P
Tene /join #haskell
wolverian I'm there. 21:11
Tene I was joking suggesting that I was going to /join #haskell and mention what you just jokingly asked not be mentioned. 21:12
wolverian ah.
TimToady, what'd be the right way to mark the zipwith as operating on a list of listy things?
lazily, I suppose.. 21:13
TimToady @@ should do it 21:14
or zip([;] @ls) in a pinch 21:15
maybe even @ls.zip
wolverian I meant exposing that to zipwith's users 21:16
but that's a good one, hm.
TimToady well, any zippy thing probably wants to have a slicey interface that separates lists with ; 21:17
wolverian right 21:18
diakopter "implicit things need explicit (possible) denotations"?
TimToady it really depends on whether the typical use case is zip(1,2,3; "a","b","c") or zip [;] @foo 21:19
generally it's probably zip somefunc() 21:20
wolverian I think so
I don't get what [;] @foo does 21:21
TimToady which currently requires zip @@ somefunc(), which might get tiresome
all reductions are defined syntactically
so equiv to @foo[0] ; @foo[1] ; etc
wolverian agh, sorry. I don't get what ; does. ;) I'll read. 21:22
TimToady it separates lists, for bindings that care
wolverian hrm
that makes me think it turns 1..3 into ([1],[2],[3]) 21:23
now I found the section in S9 21:25
TimToady it's mostly for multidimensional subscripts 21:26
wolverian hm, it carries the context into the array 21:27
yeah, I can see me putting @@ everywhere, just to make sure..
or just not using ; 21:31
21:32 Blwood left
TimToady yes, there's always the [[],[],[]] approach, which many languages take 21:32
wolverian or the ; approach, where ; is always chunks
21:33 ashleyb left
TimToady we're just doing that lazily depending on context 21:33
because p5 programmers will expect flat
esp from things like map 21:34
not to mention return
wolverian could it be a pragma then?
TimToady possibly
21:35 ashleyb joined
TimToady all functions that return 1,2,3 would end up returning [1,2,3] 21:35
instead of a Capture
wolverian hrm, no. that's ugly.
?eval my @foo = [1,2,3] 21:36
evalbot_r17041 [[1, 2, 3],]
wolverian right.
TimToady there's much to be said for lazy binding
wolverian in principle I like it. in practice, I'm worried about the extra responsibility on the receiving end
TimToady well, if we made it only a receiving end pragma 21:37
then it would do @@ map for you
(by default)
and you'd have to @ it explicitly to get flat, I guess
and return would still just return a capture 21:38
that seems saner with foreign funcs
wolverian yeah, that's my expectation of a sane language, I think.
derived from haskell, I suppose, but it's pretty generic.
TimToady alternate seems to be two versions of map, two versions of gather, etc. 21:39
which drives it towards contextual in PerlThink 21:40
wolverian so captures are perl6's monads
TimToady hmm 21:41
I'd think the monads are more like the binding contexts
the captures are the enforcees, not the enforcers 21:42
wolverian well, yes, if you think about it.. but I was making a point about the dualism.
(thinking is hard)
I suppose a true haskeller would be scared of runtime context operations 21:44
TimToady likely
wolverian it's a bit iffy to me, but then again, I can think of it as implicit typeclass operations
TimToady monads are mostly about escaping from pure value semantics, which essentially means how do you discipline something that is not a value, that is a container of values 21:45
wolverian capture would be an instance of all that they can transform into.
was that a question? :)
21:48 pmurias left
TimToady it's the binding that decides the eventual container 21:49
I suppose print() is how you bind values to the IO monad. :)
wolverian well, right 21:50
same thing in haskell, except the binding is decided at compile time
and there can only be one :)
TimToady well, a filehandle is just a lifted IO, maybe 21:51
but yes about compile time
wolverian sleep time 21:52
cheers
filehandles as monads is a weird thought
the monad api is incredibly flexible
TimToady likewise on sleep
21:53 iblechbot left, mj41 left 21:54 demq joined 22:00 aindilis left 22:03 dmq left 22:09 crem_ left, crem_ joined 22:14 aindilis joined 22:17 BooK joined 22:18 noobist joined
noobist Hello. 22:18
22:18 aindilis left
thorat hi 22:18
22:19 aindilis joined 22:35 REPLeffect_ joined, REPLeffect left 22:39 BooK left 22:41 theorbtwo left 22:48 theorbtwo joined 22:49 monomorph left, weinig_ joined, weinig_ left 22:50 weinig_ joined 23:02 noobist left 23:03 rdice left 23:04 weinig left 23:06 REPLeffect_ left 23:10 weinig_ left 23:11 weinig_ joined 23:13 rhandom left 23:18 ashleyb left 23:20 Limbic_Region joined 23:26 felipe left 23:37 weinig_ is now known as weinig 23:38 elmex left

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!