»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend! | Rakudo Star Released!
Set by diakopter on 6 September 2010.
00:00 ggoebel left 00:03 ggoebel joined 00:26 hudnix left 00:30 hudnix joined 00:36 hudnix left
frooh dukeleto: the link you put in your twitter requires a password :-( 00:37
dukeleto frooh: yeah, just realized that. It will be opened soon. I had it cached and forgot 00:39
01:35 cotto left 01:43 Chillance left 01:46 nymacro joined 01:49 Viestinik joined 01:50 ggoebel left 01:54 pythonian4000 is now known as pythonian4000afk, pythonian4000afk is now known as pythonian4000 02:02 nymacro left 02:06 nymacro joined 02:07 ggoebel joined 02:09 nym joined 02:12 Viestinik left, Viestinik joined 02:13 nymacro left 02:14 orafu left 02:30 whiteknight left 02:38 ggoebel left 02:42 skangas joined 02:43 pythonian4000 is now known as pythonian4000afk, pythonian4000afk is now known as pythonian4000 02:51 justatheory left 02:57 nym left 03:01 orafu joined, araujo left 03:07 meppl joined 03:10 nymacro joined 03:23 meppl left 03:26 meppl joined 03:29 knewt2 left 03:32 _twitch joined, sECuRE left 03:35 Limbic_Region left 03:36 knewt2 joined 03:37 sECuRE joined 03:42 araujo joined 03:45 justatheory joined 03:49 pjcj_ left
dalek ecza/master: 87a15a5 | sorear++ | / (3 files):
Implement abbreviated pair syntax :$foo
 03:52
ecza/master: e2b18ba | sorear++ | src/ (2 files):
Implement classes with enclosing superclasses
ecza/master: dfcce51 | sorear++ | src/ (3 files):
Allow superclasses with double colons
04:07 felipe joined
dalek ecza/master: 76bef32 | sorear++ | src/Metamodel.pm:
Implement simple need aliasing
 05:01
ecza/master: a5b470c | sorear++ | src/C (2 files):
Debugging aids for 'csharp' phase
ecza/master: b48c8db | sorear++ | src/Niecza/Actions.pm:
Improve subnames for grammars
05:02 REPLeffect left 05:15 REPLeffect joined 05:23 starcoder left 05:25 starcoder joined 05:34 sftp joined, sftp_ left 05:37 _kaare joined
sorear -rwxr-xr-x 1 stefan stefan 2802176 Oct 23 23:05 STD.dll 06:07
diakopter heh 06:09
06:10 lue left
diakopter curious to see how it works on the new p6eval 06:10
dalek ecza/master: 3268035 | sorear++ | src/ (2 files):
Implement <text=.sym>
ecza/master: 38e1e66 | sorear++ | src/ (2 files):
Allow $¢ inside :lang( )
ecza/master: 07fbac0 | sorear++ | v6/STD.pm6:
A few more small STD fixes
diakopter sorear: does niecza pass around a global environment/context object? 06:11
or is it globally referencable somehow? per root invocation?
(assuming there is one) 06:12
sorear niecza doesn't have a single coherent model of a "global environment"
you'll need to give more specifics 06:13
06:17 zulon joined 06:26 zulon left 06:28 PZt left 06:34 pythonian4000 is now known as pythonian4000afk, pythonian4000afk is now known as pythonian4000
sorear arghf. STD defines a 'Regex' which shadows SAFE::Regex and causes massive failure when the runtime tries to instantiate Regex objects. 06:46
06:48 PZt joined 06:54 cotto joined
sorear whee, actual runtime errors! 06:57
07:04 wormwood left 07:46 _twitch left, Axius joined 07:55 Axius left
sorear out 07:55
07:55 justatheory left 08:08 Axius joined 08:19 cafeblue joined 08:28 Ross joined 08:31 Axius left 08:33 cafeblue left
tadzik can compilers written with PCT be written in something else than NQP? 08:39
08:42 pythonian4000 is now known as pythonian4000afk 08:45 zulon joined 08:46 Axius joined 08:49 pythonian4000afk is now known as pythonian4000
dalek ecza/master: d1fae71 | sorear++ | / (3 files):
Implement cursor_fresh and mixin
 08:59
ecza/master: 6742210 | sorear++ | / (3 files):
Improve double fault reporting
ecza/master: 4e17da4 | sorear++ | lib/Kernel.cs:
Fix off-by-one in VarDeque grow
mberends tadzik: not easily. PCT is built in the grammars and actions structure, and NQP is the only language for executing the actions afaik. 09:03
09:03 Axius left
sorear didn't PCT predate NQP? 09:03
09:04 Axius joined
dalek ecza/master: 8fe1193 | sorear++ | lib/Kernel.cs:
It helps if you copy the elements into the new array.
 09:04
09:07 Chillance joined 09:12 envi joined 09:22 rgrau left 09:42 am0c^ joined, am0c^ left 09:46 Axius left 09:51 kjeldahl joined 09:52 szabgab left, szabgab joined 09:56 Axius joined 09:57 mavrc joined, lambdabot joined 10:03 kjeldahl left 10:08 wamba joined 10:19 Axius left 10:21 Axius joined 10:50 Axius left, jhuni left 10:54 whiteknight joined 10:55 shelling_lab left 10:58 icwiener joined 11:00 M_o_C joined 11:03 sbp is now known as [sbp] 11:10 tadzik left, nsh_ joined, sbp joined 11:11 sbp left, nsh_ left 11:13 nsh_ joined, sbp joined 11:14 sbp left 11:16 sbp joined 11:18 snarkyboojum left 11:19 snarkyboojum joined 11:22 pythonian4000 is now known as pythonian4000afk, pythonian4000afk is now known as pythonian4000, pythonian4000 left 11:23 pythonian4000 joined 11:28 MayDaniel joined 11:29 masak joined
masak \o/ 11:29
11:31 kjeldahl joined 11:32 tadzik joined
tadzik masak: oh hai 11:32
masak salutations.
11:35 Ross left 11:36 Ross joined, risou joined 11:37 ggoebel joined 11:39 nymacro left 11:40 rgrau_ left 11:44 ggoebel left
masak we're being compared against such distinguished things as IPv6 and Java 7: twitter.com/gleicon/status/28557567821 11:46
flussence so, comcast users had Java 7 several months ago! 11:47
11:48 ggoebel joined 11:49 ggoebel left 11:52 nsh_ left, sbp left 11:54 sbp joined 11:55 nsh_ joined 12:13 timbunce joined, Eevee left, Eevee joined 12:23 zostay left, felipe left, zostay joined 12:26 ggoebel joined 12:28 krunen left, krunen joined 12:29 Alias_ joined 12:30 sftp left, sftp joined 12:31 Alias left 12:34 ggoebel left 12:56 renormalist joined, mavrc left 12:57 tadzik left 13:02 am0c joined 13:07 M_o_C left 13:09 tadzik joined 13:14 Ross left, whiteknight left 13:20 Patterner left 13:26 mavrc joined 13:32 Psyche^ joined, Psyche^ is now known as Patterner 13:36 Alias_ left 13:39 zulon left 13:40 icwiener left 13:42 pythonian4000 is now known as pythonian4000afk 14:00 pjcj joined 14:09 dual left 14:53 zulon joined 14:54 rgrau joined 15:02 M_o_C joined 15:04 M_o_C_ joined 15:07 M_o_C left 15:08 M_o_C_ is now known as M_o_C, M_o_C left, M_o_C joined 15:10 Axius joined 15:18 Axius left 15:20 Axius joined 15:22 pjcj left, icwiener joined 15:23 zby_home_ joined 15:24 [sbp] left, nsh left 15:26 nsh joined, [sbp] joined 15:30 risou left 15:32 [sbp] left, nsh left 15:34 nsh joined 15:43 am0c left 15:52 dju joined 15:55 _jaldhar left 15:56 dju left 15:57 dju joined, lambdabot left
masak very quiet here today. 15:59
no jnthn here to pun the silence away :P
tadzik This silence is spunning me stiff 16:00
masak let's talk about something interesting, then. 16:02
tadzik contemplates fingernails 16:03
oh, Module::Tools as a backend for pls?
masak \o/
except that it'd have to be a local snapshot, then. pls is supposed to be dependency-less.
tadzik well, neutro has deps too, and they're bundled so it can be boostraped 16:04
masak nod.
I once planned to make a couple of changes to S26. maybe the quiet today makes for a good background to attempt a (possibly controversial) patch.
tadzik avoiding deps at all costs is silly imho. Especially if some of them are candidates to be the "3rd core" ones 16:05
what kind of patch?
masak well, deps are either avoidable, or essential for the installer to even bootstrap. I don't think we're in disagreement there. 16:06
the patch was about changing the way declarator blocks attach to declarations. 16:07
16:07 jaldhar joined
masak at present, they attach through clever magic a la TheDamian. I'd like to change them to attaching through a less clever but more robust method. 16:08
I'll try to locate the relevant backlog. 16:10
16:23 felipe joined 16:27 molaf joined 16:32 justatheory joined 16:37 tzhs joined, Ross joined, Ross left, Ross joined
masak ah! found it! 16:37
irclog.perlgeek.de/perl6/2010-06-18#i_2453615
tadzik oh, this thing 16:44
16:47 Trashlord left 16:48 icwiener_ joined 16:49 Axius left 16:51 icwiener left 16:52 Trashlord joined 16:54 zulon left 16:55 dju left 16:57 Axius joined 16:58 exodist is now known as Exodist
masak reads that backlog and tries to summarize 17:12
tadzik reads Pls.pm once again. . o O ( So implementing Pls is actually writing 8 methods... ) 17:14
masak :)
something like that.
there's still a yet-unrealized idea about making different multi methods correspond to different repository types. (GitHub, Gitorious, Google Code, etc) 17:15
tadzik App::Pls::ProjectsState is about a config file mostly, so deps-of should be known before fetching?
App::Pls::Fetcher is git in our case, and App::Pls::Builder/Tester/Installer is what M::Tools should do
masak no, it can only be known after fetching.
tadzik I see 17:16
17:16 dju joined
masak but pls is currently in a slightly inconsistent state wrt that. 17:16
17:16 meppl left
masak maybe that's what I should focus on tonight, rather than S26. 17:16
tadzik I think JSON can be avoided now 17:17
but right, that's a matter of implementation, not Pls itself 17:18
masak it's tricky -- I picked JSON, because proto had only done a poor man's YAML subset, and I wanted something more robust. on the other hand, the JSON module is quite slow. 17:19
17:19 Axius left, dju left
tadzik there is a in-parrot JSON implementation, which I partly got working 17:20
17:20 dju joined
tadzik well, it doesn't work for arrays 17:20
masak ooh
tadzik: I'm increasingly amazed at the number of things you have going on. :)
17:20 Axius joined
tadzik masak: I'm a lazy person, so I'm wasting my time on numerous things :) 17:21
masak :P
tadzik The semester has just started, and I'm the less-than-30-persons group of people who don't have any failed classes on their backs 17:22
masak kudos.
tadzik so, how do I get Parrot array from the Rakudo array? :)
s:2nd/array/Array/ 17:23
masak let's go to #parrot and ask :)
tadzik :)
by the way, it's astonishing how 54% of CS students on the year fail the Programming classes
17:23 Axius left
masak tadzik: every heard of the camel's two humps? 17:23
tadzik humps are those things on the camels' backs? 17:24
they sounds like those
17:24 Axius joined
masak www.codinghorror.com/blog/2006/07/s...goats.html 17:24
tadzik I image one seats on it with a "Hump" sound
s/seats/sits/ 17:25
masak also here: www.secretgeek.net/camel_kay.asp
tadzik interesting 17:28
about JSON: wklej.org/id/406539/
hrm, I just realized it's about producing, not parsing :) 17:30
masak pls needs to be able to do both.
tadzik To generate a PMC from a JSON string, see L<compilers/data_json>. 17:31
masak std: my @a; @a\ .\ foo 17:45
p6eval std a194beb: OUTPUT«ok 00:01 119m␤»
17:54 au|irc joined
au|irc rakudo: <1 2 3>.pick(4, :replace) 17:55
p6eval rakudo acce29: ( no output )
au|irc rakudo: <1 2 3>.pick(4, :replace).say
masak needs 'say'
p6eval rakudo acce29: OUTPUT«312␤»
masak oh, and .pick is now .roll
au|irc ok :)
masak er, .pick(:replace) is now .roll
tadzik what was changed to .roll... aha
au|irc rakudo: <1 2 3>.roll(4)
rakudo: <1 2 3>.roll(4).say
masak we should probably have a warning there.
p6eval rakudo acce29: ( no output )
rakudo acce29: OUTPUT«1232␤»
masak tries adding one
au|irc masak++
17:55 zby_home_ left
tadzik I remember filing a bug, or thinking about filing a bug for it 17:56
masak au: nice to see you here. I've been enjoying your tweets lately.
tadzik ah, there was a reason I didn't
masak tadzik: still not too late!
tadzik shame I don't remember why
no, I was suprised .pick eats this :replace argument, even though it's not in the signature 17:57
masak that's typical of methods.
see, um, "Interface consistency" in S12.
subs, however, throw an error on unrecognized named parameters. 17:58
rakudo: my @a = 1..3; pick @a
p6eval rakudo acce29: ( no output )
masak rakudo: my @a = 1..3; say pick @a
p6eval rakudo acce29: OUTPUT«␤»
masak rakudo: my @a = 1..3; say pick @a, 2
p6eval rakudo acce29: OUTPUT«2»
masak rakudo: my @a = 1..3; say pick @a, 2, :replace 17:59
p6eval rakudo acce29: OUTPUT«Unexpected named parameter 'replace' passed␤ in 'pick' at line 1786:CORE.setting␤ in main program body at line 22:/tmp/ex7H4r6bmo␤»
au|irc masak: thank you :) ditto - I was just writing a simple script for calcualting a md5 fixpoint 18:00
# gist.github.com/643748
masak ooh!
au|irc this goes into github.com/elliottkember/The-Kember-Identity soon - any stylistic suggestions? :) 18:01
tadzik au|irc: you could just use Perl6's Digest::MD5, no?
18:01 tzhs left
au|irc tadzik: indeed, it'll be part of 2010-10 - I was using HEAD checkout and forgot that. fixed 18:02
tadzik++
masak au|irc: why the infix:<===> instead of 'eq' and '==', respectively? 18:03
au|irc false symmetry? :) ok, fixed
tadzik au|irc: you can install it with neutro # self-advertising :)
18:03 whiteknight joined
masak tadzik++ # better at both self-advertizing and actually getting things done 18:04
(than I am, that is) 18:05
tadzik oh well
it's actually worse-but-good-enough, at least for now
18:05 Axius left
masak that counts for a lot. 18:05
now I have to compete on excellence, which is much more challenging than competing by being the only alternative! :P 18:06
au|irc ==> neutro depends on perl6-File-Tools, installing now
==> Building perl6-File-Tools
Could not load oplib `perl6_ops'
==> Building perl6-File-Tools failed
tadzik oh noes, wat do
but really, never encountered this. Will try to bootstrap it myself now
au|irc let's see. probably pwd has changed
was running it in a side-by-side checkout with rakudo
PERL6LIB=~/neutro/tmplib ../neutro/bin/neutro neutro
with pwd in ~/rakudo 18:07
tadzik I don't think it should matter
masak au|irc: please use a nopaste service when pasting many lines. thank you. :)
au|irc wilco :)
masak that's funny because you once taught me that. :)
au|irc =) 18:08
tadzik apprentice outgrew the master :)
masak I doubt it. :)
tadzik au|irc: it just flawlessly installed for me
wklej.org/id/406574/ 18:09
au|irc ok, it's a missing "make install" on my part. pebcak 18:11
tadzik I was thinking about this, but I thought rakudo won't even run w/o it
masak au|irc: seems your fixpoint thingy is about as efficient as bogosort. any particular reason you don't feed the old md5 back into $string? 18:13
I'm currently running a modified version that does that.
au|irc makes sense. fixed 18:14
masak gist.github.com/643769
it just passed 2e4 iterations, and hasn't converged yet... 18:15
oh! oh!
(++$count % 10) === 0
better written ++$count %% 10
also, consider ditching outer parentheses after all 'if' keywords.
au|irc ok, both fixed 18:16
masak gives a bit more air to the 风水 of the program. :)
au|irc added advertisement to neutro
tadzik \o/ 18:17
where?
au|irc gist.github.com/643748
line 4
use Digest::MD5; # available as "neutro perl6-digest-md5"
tadzik oh cute :) au|irc++
18:18 timbunce left 18:24 timbunce joined
tadzik fleeing 18:26
o.
o/
masak fly softly, tadzik. 18:27
tadzik heh, I could imitate flapping wings, but that'd flood the channel
masak the mental image is enough, I'm sure :)
tadzik wklej.org/id/406592/txt/ 18:28
18:28 tadzik left
au|irc ∽·~ 18:28
masak cute :)
in other news, both 'flee' and 'fly' originate from the same Old Norse root. 18:29
'fljuga' -- which is quite close to the modern Swedish word for 'fly'.
au|irc nice :) 18:31
18:32 Axius joined
masak the md5 fixpoint finder shows no sights of converging over here. 18:32
just passed 1.5e5 iterations.
au|irc well we're not even sure there is one 18:34
it's just yet another exercise in chrestomathy really :) 18:35
18:35 MayDaniel left
masak I see. :) 18:38
I feel better now about the program going on without having found a fixpoint.
maybe someone should contribute it to Rosetta Code, if it's chrestomanthy. 18:39
au|irc *nod* 18:40
masak I think it would be an interesting exercise to make Yapsi target C. 18:45
au|irc slowly lurks back to the background... :) & 18:47
masak phenny: tell isBEKaml that I'd like to conspire over Yapsi with him. :) 18:49
phenny masak: I'll pass that on when isBEKaml is around.
colomon awe, man, I just missed au|irc++ on here. 18:51
errr... aw
but I suppose awe is appropriate as well.
18:52 lue joined
lue .u ‽ 18:53
phenny U+203D INTERROBANG (‽)
lue unicode is fun :)
masak I'm sometimes reminded of a line from the Matrix: "Take a look at his neurokinetics, they're way above normal." that's what it feels like when au is applying changes to code. 18:54
lue (o hai o/ as well)
masak lollue, hai \o
lue: I've been thinking a bit more about some kind of game written in Perl 6. 18:59
18:59 M_o_C left
masak lue: my latest idea is a fairly simple ASCII-based RPG map-traversing thing, in which levels persist and in which a big part of the game is building the levels. 19:00
lue me as well, but more in a passing though format.
s/gh/ght/
masak it's just an idea so far. I'll happily discuss it further.
cxreg masak: does it teach you perl 6 as you play it? 19:01
masak cxreg: I guess it could. I hadn't thought of putting that in as an essential part.
cxreg it could be like nethack, but each of the monsters is a p6 operator
lue It sounds like a good idea. The most interesting part would be refreshing the screen.
cxreg "You're fighting a Whatever"
masak cxreg: I thought of tying it to Perl 6 by first making a quick draft of the game, and then building it from scratch over the course of 30 days or so.
flussence ooh 19:02
cxreg masak: i'm down
masak \o/
flussence imagines a game like minecraft, but with a REPL instead of blocks
masak lue: some curses library would be needed.
cxreg flussence: > dig. dig. dig. dig. dig.
lue
.oO(curses, if only we had a library like that)
 19:03
masak :)
lue: there's one in Parrot.
oh, and it would be multiplayer and networked. I'm not sure how realistic that is at this point. 19:04
but it'd be fun to try.
cxreg "You're fighting an infinite lazy list. But you win easily, because it was lazy."
lue "You're fighting infinity! Next time don't divide by zero"
flussence cast eager on it to make it explode...
masak lue: a division by zero doesn't result in infinity.
lue rakudo: say 1/0
p6eval rakudo acce29: OUTPUT«Inf␤»
lue QED 19:05
masak ...except in some IEEE standards :)
but not in real maths.
the thing that becomes infinity is the limit of division by ever smaller numbers. and even then, only the limit from the right. 19:06
lue
.oO(Although graphs of f(x)=1/x could be used to suggest 1/0=Inf)
masak lue: no, they would act as counterevidence.
because the limit from the left would be -Inf.
and hence the "actual" limit is undefined.
lue I think we can agree x/0 is Very Interesting™ 19:07
masak it's not like it's a concept in need of deeper study in analysis. 19:08
though many crackpots would have you think so.
lue
.oO(Let's see how far down the Layers of the Internet we have to go for networking :P)
masak here's a common crackpot "explanation" of 1/0: www.bbc.co.uk/berkshire/content/art...ture.shtml
sadly, such nonsense is treated as news because people don't know enough maths. 19:09
lue I don't know, 'nullity' is a very mathy-sounding term, so I believe it. [JOKE] 19:10
masak wow, wikinews has great coverage of the whole thing: en.wikinews.org/wiki/British_comput...ematicians 19:11
lue I knew I saw that piece of news somewhere before! (on wikinews) 19:12
"Imagine you're landing on an aeroplane and the automatic pilot's working," he suggests. "If it divides by zero and the computer stops working - you're in big trouble. If your heart pacemaker divides by zero, you're dead." Isn't that why IEEE has it equal Inf instead of an error? 19:13
masak I believe I might have learned about it through Mark C. Chu-Carroll's blog.
lue: I don't know why IEEE has it that way. 19:14
lue
.oO(A Doctor Who story where everthing bad happened because the TARDIS divided by zero... *evil grin/laugh*)
 19:15
masak lue: I'm enjoying some of the new Doctor Who episodes now, by the way. lovely stuff. 19:17
lue wikinews: "Anderson's proof that one can divide by zero ('written [...] on the back of an envelope for added authenticity')" :D
did you see the season 31/series 5 finale? 19:18
masak no, I'm only at the beginning of series 4 so far. 19:19
lue I'll shut up then :) 19:20
masak thank you :)
19:20 Axius left
lue I, on the other hand, need to watch the first 26 seasons (all of the old series). I started during the 2009 gap year. 19:21
dalek ast: 4f15296 | masak++ | S32-list/pick.t:
[S32-list/pick.t] test for deprecated :replace
lue When I kept reading about the old series, I thought "For Last of the Time Lords, there sure were a lot of them in the old series" before discovering the Time War was a new Who invention 19:22
masak building/testing a deprecation error for C<.pick>.
lue continues considering a P6 RPG 19:23
arnsholt As an aside, division by zero is an exceptional state according to IEEE floating point, according to Wikipedia =) 19:24
masak let me see what it gets me in Java. 19:26
masak .oO( System.out.println( 1.0 / 0.0 ); ) 19:27
gives me 'Infinity'. so Rakudo is not alone in doing that. 19:28
pugs: say 1/0
p6eval pugs: OUTPUT«*** Illegal division by zero␤ at /tmp/MYjMw49tpR line 1, column 5 - line 2, column 1␤»
masak huh 19:29
rakudo: say 1/0
p6eval rakudo acce29: OUTPUT«Inf␤»
masak rakudo: say (-1)/0
p6eval rakudo acce29: OUTPUT«-Inf␤»
lue rakudo: say 1/Inf # just for kicks 19:30
p6eval rakudo acce29: OUTPUT«0␤»
lue "You meet Bob, a hyperoperator. He'll give you the Important Thing, but only after you get him some decaf. Pleeeeease?" 19:34
masak it sounds like customizing monster appearance and behavior is perhaps more important than building levels. 19:36
and maybe some kind of per-level networking is more realistic than real-time networking.
lue levels are important too. And we can abuse unicode on the way :)
masak heh.
19:37 molaf left
lue "[Level 32] You come to the base of Parrot Mountain. Be careful, as there's something at the summit hurling PMC's at you" 19:37
I think it's a great idea, but the networking should wait until the game works :) 19:42
19:46 pjcj joined
masak you're most likely right. 19:47
and until Rakudo provably supports it.
(not just on paper, but in practice, too) 19:48
cxreg "Steve Kramarsky lamented that Dr Anderson is not from the "University of 'Rithmetic". # lol 19:50
19:51 pjcj left
sorear wonders if and how perl 6 can ever match perl 5's efficient list context implementation 19:52
laziness is, on the surface at least, a huge hurdle 19:53
19:53 dual joined, pjcj joined
flussence I think the answer there is more laziness, doing nothing until something asks you for a .Array or .Num etc. 19:54
19:54 MayDaniel joined
flussence though that only helps if everything that depends on the context is called lazily... 19:55
having to learn functional programming to get the most efficient code does seem a bit suboptimal 19:57
sorear laziness only helps if the value is never needed
lue masak: I'll think more about that RPG. If it weren't for my drastically reduced P6 time (school :/), I'd have more chances to help as well. 20:00
afk
masak we're all on some kind of time budget.
20:02 molaf joined, Chillance left 20:13 hudnix joined
diakopter au|irc: can an md5 prog detect loops of any length? (oh look, we've seen this hash before, x hashes ago) 20:24
20:25 molaf left
diakopter even if there's not a loop with size 1, there could be others with higher sizes 20:28
masak: want to tackle it? :)
20:28 hudnix left
masak diakopter: that's a nice idea. Floyd's cycle-finding algorithm can probably be made to do that, even if the linked list in question is generated on the spot. stackoverflow.com/questions/2663115...nked-list/ 20:28
diakopter masak: cool, good idea 20:29
20:29 rvrgs joined
masak ooh, that means that the memory consumption can be kept constant. 20:29
nice!
20:29 rvrgs left
simon compiling rakudo on a netbook. this is the first time I regret my choice of CPU. 20:30
diakopter plus, the list only needs traversed if the md5 bitmapspace (16 GB?) shows a hit
masak simon: how much memory does it have? 20:32
RAM, I mean. 20:33
20:35 mavrc left
dukeleto simon: are you using ccache? 20:38
masak those are some inexplicable sortings: mail.pm.org/pipermail/moscow-pm/201...09063.html 20:45
and fairly suspious uses of the Whatever star, too.
Moscow.pm seems to have an active list. 20:46
simon masak, 1G. 20:49
dukeleto, probably not. I just hit 'make'.
it's stalled for a long time on /usr/local/bin/parrot src/gen/perl6.pbc --target=pir src/gen/core.pm > src/gen/core.pir
not stalled, it seems to be working
masak simon: I think 1G is just on the limit. many people with less memory than that have been complaining about excessive slowness (due to swapping). 20:50
simon masak, I'm using <800M currently, so should be fine.
20:50 hudnix joined
masak yes. it's the compilation of src/gen/core.pm that's the memory bottleneck, by the way. 20:51
flussence fwiw I can compile on my 1G machine without any swap at all
simon I should've timed it. :)
20:52 PacoLinux left
flussence hmm, I should check this cron output on my other atom box... 20:52
simon ah yes, atom is the name of the processor.
dalek kudo: d49eea2 | masak++ | src/core/Any-list.pm:
[core/Any-list.pm] deprecation note .pick :replace

Prior to this change, the C<.pick> method accepted the :replace parameter (which no longer applies and has been replaced by the C<.roll> method) without complaining. Now it throws an error.		 20:54
flussence rakudo: say (DateTime.new('2010-10-24T05:16:00Z').Instant - DateTime.new('2010-10-24T03:00:00Z').Instant) - (1*3600 + 54*60 + 34) 20:57
p6eval rakudo acce29: OUTPUT«1286␤»
flussence 21 minutes to compile from git, more or less
(there's probably something I'm missing in that code, it's not as nice as I'd like...) 20:58
21:05 envi left 21:20 wamba left 21:22 sftp left 21:23 sftp joined, _kaare left 21:25 Ross left 21:38 timbunce left 21:39 Trashlord left
sorear aloha, karma sorear 21:45
aloha sorear: sorear has karma of 150.
masak sorear++ # definitely more than that
sorear things that frustrate me: 11 minute compile cycles that give me back the same error. 21:46
21:48 pythonian4000afk is now known as pythonian4000
masak I remember similar delays with GGE. that's when things like tote and a good movie come in handy. 21:54
ooh! haven't released tote to the world yet, have I? :)
always things to do...
dalek ast: 794519e | masak++ | S32-list/pick.t:
[S32-list/pick.t] removed TODO fudging
 21:59
masak well, that's it for me. time for my regeneration cycle.
sleep. I mean sleep.
'night, #perl6.
21:59 masak left 22:04 spacebat_ is now known as spacebat
dalek ecza/master: 1e0277e | sorear++ | src/RxOp.pm:
Fix subrule-induced backtracking under :ratchet
 22:09
ecza/master: 9895e91 | sorear++ | src/RxOp.pm:
<!!{ }> blocks should create Bool context
ecza/master: 967b37e | sorear++ | src/Niecza/Actions.pm:
Implement augment class Foo::Bar
22:30 rvrgs joined 22:42 bluescreen left 22:50 risou joined, icwiener_ left 22:55 bluescreen joined 23:01 MayDaniel left 23:06 sftp left
sorear well, I finally got STD.parse("2 + 2") to not *crash* 23:19
it does, though, panic("Malformed subset")
wtf.
23:25 envi joined
sorear oh, there's an insidious precedence bug. 23:29
<sym> [ ... ... ] || <.panic: "Malformed subset">
23:30 risou_ joined 23:33 risou left 23:58 rgrau 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!