»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!
Set by sorear on 4 February 2011.
00:01 sftp left 00:02 sftp joined 00:06 sftp left 00:08 sftp joined, carlin joined 00:15 am0c left 00:17 replore joined
dalek gs.hs: 9782a01 | (Ingy dot Net)++ | install-haskell-platform-ghc7-and-pugs-on-debian-or-ubuntu-from-scratch.sh:
This script will install GHC7 and Pugs on Debian.

If you are very lucky. :)
00:22
ingy moritz: ^^ 00:24
moritz: That should do the trick.
moritz: and what a holy pain it was...
00:25 ab5tract left
ingy anyone want to try to install pugs on debian? 00:26
:) 00:27
masak: ? 00:29
[Coke] ah, I can try to install haskell platform on OS X now. 00:31
ingy moritz: even though the script above will take a long time, the building of pugs in cron using Pugs-hs/Pugs/Makefile should only take a few seconds
[Coke]: you'll need to tweak the script 00:32
one sec
[Coke] ingy: download is going to take another 22 minutes. 00:33
ingy actually the script is pretty debian specific 00:34
but please try getting going on mac
[Coke] ingy: i'm looking at the haskell platform .pkg ; I can only imagine it's OS X specific.
ingy :)
00:35 whiteknight left
ingy I think linux was the hardest 00:35
I was even working with a haskell guy and he was sighing how ridiculous the process was
[Coke]: how do we give out commit bits? 00:38
shachaf needs a bit for pugs 00:39
hugme: shachaf
guess
[Coke] ingy: I think perl6/ is pretty liberal. 00:45
hugme: help?
hugme [Coke]: (add $who to $project | list projects | show $project | hug $nickname | tweet $twittername $message )
[Coke] hugme: add shachaf to Pugs.hs
hugme [Coke]: sorry, I don't know anything about project 'Pugs.hs'
[Coke] hugme: list projects
hugme [Coke]: I know about Math-Model, Math-RungeKutta, MiniDBI, bench-scripts, book, gge, hugme, ilbot, java2perl6, json, modules.perl6.org, november, nqp-rx, nqpbook, perl6, perl6-examples, perl6-wtop, proto, pugs, rakudo, star, svg-matchdumper, svg-plot, tardis, try.rakudo.org, tufte, ufo, web, yapsi
[Coke] I think the pugs there is old pugs. 00:46
diakopter Pugs.hs is under the perl6 project I thought
[Coke] but iunno
ingy right
[Coke] oh! that's easy, then.
ingy do it
[Coke] I leave it as an excercize for the reader.
diakopter hugme: add shachaf to perl6
[Coke] ;)
00:46 hugme left
ingy shachaf: ^^ 00:46
diakopter nice
I crashed it
[Coke] BWUAAHAH.
ahem.
ingy gah 00:47
moritz: ^^
rather
sigh
what a day of FAIL
beer time
&
gfldex nom: my Rat $sum = 0.Rat; $sum += 24537130 * 3.37; $sum += 5678 * 4.28; say $sum; 00:48
p6eval nom 3fd353: OUTPUT«82714429.94␤»
gfldex anybody still awake who could try that on 32bit ?
[Coke] sure. 00:49
gfldex on my 32bit host it says: -3184915.98
[Coke] gfldex: 82714429.94 00:50
I'm on OS X 00:51
10.6.8, 32 bit, intel
gfldex Linux dexhome 3.0.0-1-686-pae #1 SMP Sat Aug 27 16:41:03 UTC 2011 i686 GNU/Linux
[Coke]: what is your perl6 --version ? 00:53
sorear shachaf is a proper p6er now? \o/
[Coke] This is perl6 version 2011.09-66-gabc6874 built on parrot 3.8.0 revision RELEASE_3_8_0-130-gfad30ea 00:55
few days old.
I'll kick off a rebuild. 00:56
sorear I've added Shachaf to perl6
made sure it was the right one, too.
gfldex mine is a few hours old
[Coke] gimmie a minute.
shachaf sorear: Oh, hello. 00:59
ingy: What have you gotten me into?
ingy trouble
with two capital S's 01:00
shachaf So Pugs has been resurrected? 01:02
[Coke] trouSSle ?
shachaf moves #perl6 down from irssi window 87.
[Coke] ugh. compiling rakudo just kills my machine. :P 01:08
shachaf What's the oldest version of GHC that people actually compile Pugs with? I'd like to get rid of all these deprecation warnings.
[Coke] shachaf: I suspect that all 3 of them are here. 01:09
01:09 plobsing_ joined
[Coke] I haven't started yet, but am installing whatever the latest version on haskell platform is. 01:10
shachaf [Coke]: People who compile Pugs?
[Coke] ingy just resurrected it, I thought. 01:11
shachaf: yes, them.
I'm installing 2011.2.0.1 , apparently.
01:12 tyatpi_ joined 01:13 plobsing left, donri left 01:27 daniel-s joined 01:31 tyatpi_ left
[Coke] does "cabal install Pugs" install from the local checkout? 01:34
shachaf [Coke]: No, it installs from Hackage. 01:37
[Coke]: Just "cabal install" should run from the current directory, though.
[Coke] I'm thinking that "cabal install Pugs" shouldn't be the first thing in the README, then. 01:38
01:39 soh_cah_toa_ left
[Coke] gfldex: same result. 01:42
gfldex i will rebuild myself while i ./snore & 01:43
01:43 awwaiid left
dalek ast: 167c741 | Coke++ | S16-io/bare-say.t:
rakudo fudging
01:49
kudo/nom: 3e66bdf | Coke++ | t/spectest.data:
track failures, run fudges
[Coke] wonders why Pugs.hs/README says "Perl 5" 01:50
[Coke] finds HACKING, and goes down that path instead. 01:51
sorear hello shachaf
[Coke] "at least the following dependiences are missing"... "cabal install <dep>" ? 01:56
looks like. 01:57
02:17 jaldhar joined 02:18 benabik joined 02:28 wolfman2000 joined
[Coke] ingy, shachaf: nopaste.snit.ch/87090 02:35
;(build failure on pugs on OS X)
02:38 Khisanth left, Khisanth joined
shachaf [Coke]: Hmm, what version of GHC/base are you using? 02:38
sorear wonders if shachaf has set an /ignore 02:40
shachaf sorear: No, why? 02:41
shachaf is confused. 02:45
sorear shachaf: ignoring hellos, and our past 02:47
shachaf: apologies for the hasty conclusion
[Coke] shachaf: um... 02:48
how would I tell? ;) 02:49
02:49 drbean left
shachaf [Coke]: ghc --version; ghc-pkg list base; ghc-pkg list stm 02:49
02:49 drbean joined
[Coke] ah, which ghc is showing me the macports one, not the cabal one. checking... 03:08
aaaaargh. all the prereqs I installed were apparently tied to other-pugs. 03:10
er, other-ghc.
starting over...
(or, are wiped with a 'make clean')
colomon b: say 10.rand 03:23
p6eval b 1b7dd1: OUTPUT«2.59589085196243␤»
sorear o/ colomon 03:24
colomon \o
I'm trying to see if I can get a quick version of pick up and running before bedtime 03:25
nom: say 10.rand
p6eval nom 3e66bd: OUTPUT«3.306615122865␤»
[Coke] new error, nopaste coming.. 03:27
gist.github.com/1283291 03:32
(that's only a fragment of the build)
shachaf [Coke]: Looks like an OS X issue. 03:33
[Coke]: As far as I know only 32-bit GHC is available for OS X. Maybe libperl.a is 64-bit?
[Coke] I've only got 32 bits. 03:38
zzzz soon. will investigate more tomorrow. Thanks! 03:39
colomon sorear: bother. no splice yet either? ;) 03:42
[Coke] (and after removing ports, I'm using ghc 7.0.3, btw.) 03:43
colomon afk #bedtime
03:57 satyavvd joined
ingy shachaf: I think you can feel free to make pugs use fairly modern ghc 04:03
I had it running on 6.12.4? for a bit but now I am going to stick with 7+ 04:05
what version did the deprecations start with?
time to zzzz
04:17 dukeleto left, Khisanth left 04:18 dukeleto joined, dukeleto left, benabik left 04:19 dukeleto joined 04:38 molaf left 04:42 lestrrat left 04:43 lestrrat joined 04:47 envi_ joined 04:48 Khisanth joined 05:01 Exodist left
moritz good morning 05:03
05:05 Exodist joined
moritz ingy: I alrady tried to install GHC 7.0.3 myself. It failed because of GHC bug 5005. 05:07
05:11 orafu left, orafu joined
sorear nom: role A[::B] { class C { method foo() { say B } }; method bar { C.foo } }; class D { }; A[D].bar 05:12
p6eval nom 3e66bd: OUTPUT«Cannot call a method on type variable B␤ in find_method at src/gen/Metamodel.pm:267␤ in sub say at src/gen/CORE.setting:4957␤ in method foo at /tmp/B2F_HHl8Pe:1␤ in method bar at /tmp/B2F_HHl8Pe:1␤ in <anon> at src/gen/Metamodel.pm:1125␤ in block <anon> at /tmp/…
sorear I'm looking for a way to emulate ML functors
parametric roles would do the job if that example worked. :)
05:14 sftp left
sorear essentially, modules (with EXPORTed functions and zero or more types) parametrized over one or more other modules 05:18
moritz nom: role A[$B] { class C { method foo() { say $B } }; method bar { C.foo } }; class D { }; A[D].bar 05:19
p6eval nom 3e66bd: OUTPUT«Null PMC access in find_method('gist')␤ in sub say at src/gen/CORE.setting:4957␤ in method foo at /tmp/IvqRxAPI_7:1␤ in method bar at /tmp/IvqRxAPI_7:1␤ in <anon> at src/gen/Metamodel.pm:1125␤ in block <anon> at /tmp/IvqRxAPI_7:1␤ in <anon> at /tmp/IvqRxAPI_7:1␤…
moritz I guess it doesn't like the inner class being composed at compile time, before the role is being concretized
05:22 bazqux joined 05:52 kaleem joined 05:56 wtw joined 06:06 am0c joined 06:08 bazqux1 joined, bazqux left 06:13 bazqux1 left
masak good morning, #perl6. 06:28
www.modernperlbooks.com/mt/2011/10/...ility.html # chromatic++ 06:29
moritz good masak, morning
06:30 Woodi left
masak in some ways, it's not a fun read. but it keeps close to reality. I'm saddened that in 2011, a Perl 6 backer like chromatic can't realistically use Perl 6 for daily tasks. 06:31
06:31 Woodi joined 06:33 MayDaniel joined 06:37 MayDaniel left
masak I agree with a commenter that good Perl 5 integration would be a big help. 06:37
moritz "I worry that once you release a programming language with minimum viable utility, you ossify early decisions of design and implementation such that you can't quite yank the floorboards out from underneath any users you have, lest they stop being users and stop providing feedback." 06:39
I can't believe this sentence comes from somebody who complains about the frequent rewrites of rakudo
well, makes a very big deal out of those rewrites 06:41
chromatic makes a good job describing the problem. If he only offered a solution... 06:44
masak I don't blame him for not doing that. he's one of the few people, maybe the only one, who describes the problem. 06:45
as to yanking the floorboards, I've always considered the phase Perl 6 has been in from 2005 up until now to be this "fluid" phase where we have the privilege to make spec course corrections without it hurting anyone. 06:46
that would change overnight if someone took Perl 6 on as their "production platform".
of course, it's not a one-way causality arrow there. not entirely. some prospective clients may be scared off by all the fluidity. 06:47
"it's 2011 and you're still mucking around with the spec!?"
I've gathered that mberends argues, or at least has argued, for freezing the entire spec in order to attract such clients. 06:48
moritz that sounds nice, unless you actually plan to implement it :-) 06:49
masak right. :) 06:50
my counter-point at the time was something along the lines of the fluidity of the spec being important to the implementors, and the benefits of freezing the spec uncertain. 06:51
moritz it would make us fall into the same kind of trap as the parrot deprecation policy
06:51 PacoLinux joined
mberends masak: I no longer argue so. I wrongly suggested freezing when I believed the spec was largely complete, and merely being extended. I now understand better that the spec is still very patchy. 06:51
moritz ie it would make sense iff we had a horde of users out there. Maybe.
jnthn It doesn't make sense to talk about "the spec" as if it's all at the same level of maturity.
The bar for changes to most of S03, S06, S12 and S14 should be very high. 06:52
To S11, pretty low.
mberends aye
masak right. there's a city center of the spec that's stable and high quality. the suburbs are so-so.
jnthn: I'd include S05 too in the "bar should be high", even though it's seen some changes lately. 06:53
jnthn er, yes, that to 06:54
I wasn't trying to be exhaustive.
At a personal level, I no longer have appetite for big changes to things I consider stable any moer.
*more
It's demotivating.
OK, teaching time. o/
masak that's an important factor too. implementor burnout. 06:58
but with any such change I would prefer to argue against change on merits of the current spec rather than "jnthn quits if you change this again!", if possible ;)
sorear's recent proposal for changing the MMD semantics is a good example. 06:59
as I understand it, there are really good reasons Perl 6 shouldn't/can't emulate C# semantics there.
moritz I didn't follow that discussion very closely -- can you give a quick summary of the reasons? 07:08
masak only a garbled version of what jnthn told me. 07:09
essentially, C# has different distributions of compile-time/runtime dispatch decisions. 07:10
sorear the C# spec is very close to what P6 is doing already. I do not beleive jnthn's answer to be correct at all.
but I'm not going to convince him except with working code 07:11
masak I do take it seriously that jnthn has implemented the algorithm for dispatch resolution, and he says that we're in a model sweet spot.
yes, that's arguing from authority. but in this case, the authority is what makes my MMD calls work ;) 07:12
sorear after I get /serialize working I'll make the necessary changes to niecza. I don't think it will change the behavior of any sane code, but it does a better job of carping in insane cases 07:14
but for now, *sleep* 07:15
masak 'night, sorear. dream of working serialization and sane code.
07:19 pochi joined
masak ingy: I'll try installing Pugs when I get home tonight. 07:21
nom: role A[$B] { class C { method foo() { say $B } }; method bar { C.foo } }; class D { }; A[D].bar 07:24
p6eval nom 3e66bd: OUTPUT«Null PMC access in find_method('gist')␤ in sub say at src/gen/CORE.setting:4957␤ in method foo at /tmp/xCHvZokqR7:1␤ in method bar at /tmp/xCHvZokqR7:1␤ in <anon> at src/gen/Metamodel.pm:1125␤ in block <anon> at /tmp/xCHvZokqR7:1␤ in <anon> at /tmp/xCHvZokqR7:1␤…
masak moritz: did you RT this one?
moritz masak: no
masak submits rakudobug
moritz after yesterday's discussion, I'm not convinced anymore that it makes sense to ticket each weird corner case 07:25
masak wow, that's not at all what I took away from yesterday's discussion. 07:27
07:27 wooden left
masak and even if it were, I'd probably consider the above case well within the set of worthy tickets. 07:27
moritz well, it was more the result of the non-discussion part 07:28
where I asked what value RT provides, and nobody had really convincing cases 07:29
masak I'd rather have a longish RT queue and people complaining that it's longish, than a situation where someone suffers from a Null PMC access in production and we go "oh yeah, that one. we stumbled upon it, and then forgot about it."
moritz: wut? I thought I replied to just that question.
RT, in the best case, represents what we know to be wrong with Rakudo and/or Perl 6. 07:30
moritz masak: you did, and it didn't convince me of the practical value of RT
masak :(
on balance, if I find something wrong, I submit it to RT. here's why: over 50% of the tickets I've ever submitted have been fixed. there's no doubt in my mind that a smaller quantity of the same bugs would've gotten fixed had those tickets not been in the queue. 07:32
I don't see how that's not value.
moritz we had a huge amount of tickets where parrot objects leaked through to p6 space 07:33
masak (I'm not talking about just my tickets, of course. I just happen to know what proportion of my tickets have been closed.)
moritz switching to nom made those tickets really uninteresting 07:34
masak so what?
07:34 mj41 joined
masak it's like saying "this refactor made your feature requests moot" 07:34
there was no way for us to know *at the time* that those tickets would be uninteresting. 07:35
and each one probably represented some amount of pain for the submitter.
moritz so they turned out to be an (RT, roast) maintenance burden, and not useful
masak what you're saying is maybe cause for overseeing process somewhere. not for throwing out RT.
07:38 drbean left
masak the relevant question is perhaps "could we have known, foreseen, or prepared for those tickets becoming really uninteresting?" 07:41
07:42 uvtc joined
masak maybe pre-emptively tagging them all with a parrotobjectleakingthrough would've been enough to help combat the subsequent sense of futility. 07:42
uvtc Interesting blog post from Chromatic. 07:43
I've been thinking that what Perl 6 indirectly needs to reach "minimum viable utility" stage is Perl 6 user docs. perldocs for Perl 6.
I wrote an article about it, which could be a blog post, I suppose, except that I don't have a blog at the moment. It's temporarily here, in case anyone wants to take a gander: www.unexpected-vortices.com/temp/pe...needs.html
The last section addresses Perl 6.
07:43 drbean joined
masak takes a gander 07:43
frettled you give me goose bumps ;) 07:46
uvtc Who, me?
Why?
frettled No, masak. He's like that. :)
(gander = adult male goose, BTW) 07:47
frettled is a horrible punster.
uvtc Hahaha. Nice. :)
masak "and yet Perl 5 survives" -- the principle of "worse is better" in action, I guess. 07:48
frettled yup 07:49
uvtc: Nice write-up.
uvtc Thanks, frettled. 07:50
masak I agree. Perl 6 needs docs.
moritz the real problem is that Perl 6 needs so many things
it needs features I can use 07:51
it needs docs for those features
masak and Perl 6 is always going to be compared with Perl 5 for the things it lacks.
moritz and it needs the stability to make the documentation not outdated after half a year
and it needs to be not so slow
and it needs to be robust, and so on
masak we just have to converge slowly on all of those things.
uvtc My understanding is that, although it needs all those things, feedback from users drives the process. 07:52
I'm guessing it also provides encouragement to developers.
masak someone who takes on a major doc writing project needs to be prepared for some amount of spec bitrot, and not be discouraged by that. 07:53
uvtc And to get users (hence, feedback), you need those docs. It's just a necessary evil that they're going to have to be updated a lot.
masak the Perl 6 Advent runs have showed that exposing the workings of Perl 6 to users is a really good thing. docs would help with that, too.
07:54 koban joined
masak uvtc: we're waiting for a champion. 07:54
uvtc My impression of the advent articles was that they generated interest. They showed off features. But they left me wanting real meat and potatoes perldocs.
masak: I don't think it needs to even be a "major" doc writing project. Just some basic readable user-level basics. 07:56
moritz uvtc: then write them. I'll support you wherever I can
in fact I've started writing docs multiple times, and got starved by lonliness
uvtc moritz: "starved by lonliness"?
moritz: do you mean it was turning into more work than you anticipated? 07:57
masak moritz: me too.
moritz uvtc: nobody seriously joined my effort, and I lost focus
masak it's hugely thankless work, with little feedback on the way.
jnthn Null PMC access is always wrong really...tickets on it are fine. In the role case above, we should do something better...
moritz same thing with the book; I can't seem to motivate my co-authors to work more on it
commute& 07:58
uvtc I see. A possible solution might be to institute some kind of policy where, when spec changes happen that disagree with the docs, someone pings the doc champion. 07:59
For me, the Synopses are difficult to follow. If I were writing perldocs, and the Synopses were changing under my feet, and I didn't know where to look to see what's changing, 08:03
and users were complaining that the docs were out of date, 08:04
that would probably spiral into the docs getting abandoned. 08:05
masak the desired traits of a doc champion are rare. it has to be someone with more than a decent understanding of the spec, the implementations, and how to write docs.
uvtc If no one experienced here is working on perldocs in stealth-mode, then I don't think you're going to find someone with the traits you desire. 08:06
masak me neither.
uvtc What I think you're going to find is a volunteer who's good at technical writing, but who has trouble making sense of the specs. 08:07
If you can think of a system for that person to succeed, I think you'll have a workable solution.
masak I'm wondering if I don't consider the book of higher short-term importance/benefit, though. and *it*, for all intents and purposes, has currently stalled.
mberends uvtc: I like your post. I've tried to be a perl6doc champion for some time, and have been set back and disillusioned by the overambition of the design and the impracticality of the resulting tools.
masak it's just hard to write a lot of text and keep it consistent and updated.
08:07 johs39 joined
johs39 just reading about perl6. anyone know of its current performance vs python? 08:08
mberends uvtc: I'd like to take a chainsaw to S26 and severely prune it. 08:09
masak johs39: if it's performance you're looking for, take a look at Niecza.
niecza: say "OH HAI johs39!"
p6eval niecza v10-43-gf5298c8: OUTPUT«OH HAI johs39!␤»
johs39 thanks
masak johs39: can't say anything offhand about how it stacks up against Python, but I'd guess pretty well except for perhaps startup time.
johs39 because of mono? 08:10
masak johs39: if you'd be willing to do some measurements of your own, I'm pretty sure we'd be interested.
uvtc masak: the book is good for those with minimal experience, who want a nice leisurely walk-though. However, one of the most useful programming language docs I've read is the first or 2nd chapter of the Python nutshell book.
masak johs39: because of a number of things.
johs39 alright
masak johs39: github.com/sorear/niecza
mberends johs39: yes. programming Niecza is like programming C# whilst wearing Perl 6 gloves. 08:11
johs39 already there
nice
what are the main differences between this implementation and rakudo? 08:12
uvtc mberends: regarding S26, I agree. Though, for me, I actually prefer Markdown. Seems like the path of least resistance to getting docs written.
johs39 i see the feature comparison, but i mean methodology maybe or maybe something else guides it?
masak johs39: Niecza started out focusing on optimizations. also, it's on .NET, not Parrot. 08:13
uvtc mberends: regarding "disillusioned by the overambition of the design", you're talking about S26 or Perl 6 in general?
johs39 alrigth that sums it up 08:14
masak johs39: I'm using Niecza currently and finding it pleasingly fast and relatively feature-complete. 08:15
snarkyboojum also don't forget perl6.org/compilers/features :)
masak it doesn't have all of what Rakudo has, for sure. but it's getting there, and what's not there can often be worked around.
uvtc masak, mberends: Perhaps if perldocs are too ambitious, then a perl101.org/ , with an easy way for core devs to provide feedback on bitrotted docs, 08:16
is the way to go to get started.
masak, mberends: I meant, a perl101 for Perl 6, of course.
masak a perl110 ;) 08:17
uvtc masak: Heh. Though, I don't like the '1' next to the 'l'. I'd prefer perl-6-101.org.
Oh, man. Why do I stay up so late. Thanks for the food for thought, all. 08:20
08:21 uvtc left
masak #perl6. the soup kitchen for thought. 08:22
08:22 mj41 left
johs39 niecza built, up and running 08:29
masak johs39: nice. let us know how it goes. 08:32
johs39 alright
will do
thanks for the info. i'm heading out 08:34
08:34 johs39 left, amkrankruleuen left
moritz if somebody starts writing docs, I'll happily take the role of pinging the maintainers about spec changes 08:34
08:35 mj41 joined 08:39 uvtc joined
uvtc masak, wait. I missed your comment. You wrote: "the principle of "worse is better" in action, I guess." 08:40
masak re smosher's p6c message about .reduce and higher arities: it always seemed to me that MMD almost-but-not-quite applies when what we're binding against is a prefix of unknown length of a stream of arguments.
uvtc: yes. I don't mean it in any derogatory way.
uvtc masa: My point was, if you have those docs in order -- if you have a way to bring that new user up to speed and productive in a very short amount of time --
masak uvtc: I mean that everyone can agree that Perl 6 is superior on paper, and still elect Perl 5 in practice, partly for the arguments you outline.
frettled These days, they call it «agile» ;)
uvtc then you beat "worse is better". 08:41
masak not sure you can "beat" worse is better :P
08:41 amkrankruleuen joined
frettled It's theoretically possible to beat "worse is better", but in practice, noone has. ;) 08:41
masak frettled: no, agile is about development methodology. "worse is better" is about design philosophy. 08:42
08:42 [particle] left
uvtc A new user has a new task and is sitting in front of a choice of a few new languages they might use. Which language do they choose? As chromatic points out, 08:42
frettled masak: technicalities, besides, Perl in either incarnation isn't quite worse-is-better. 08:43
(www.jwz.org/doc/worse-is-better.html)
uvtc it's the one with the minimum "distance between zero and just getting something done."
08:43 dakkar joined
uvtc If your "not worse" language has the shorter distance, then you beat "worse is better". 08:44
frettled masak: I'd rather argue that Perl 5 is from the bazaar, while Perl 6 is from the Cathedral. 08:45
masak that's an interesting point, for sure. 08:46
I wouldn't say that it's that clear-cut, but in relative terms, sure.
moritz masak: I'm writing a reply to smosher
masak moritz++
moritz ... and p6c
frettled masak: mm, absolutely not clear-cut.
masak moritz: do you see what I mean about streamish MMD? it's like the dispatch is "slanted". in fact, it looks a bit more like LTM than the regular MMD. 08:47
08:48 DarthGandalf left
moritz masak: yes 08:48
uvtc moritz: your offer is noted. I think it's a good step.
moritz masak: that's exactly what I'm writing about :-)
masak \o/
uvtc moritz: re. irclog.perlgeek.de/perl6/2011-10-13#i_4561649
08:48 [particle] joined
moritz uvtc: I think one of the mistakes of the previous doc efforts was that they were too humble in going public, thus receiving too little feedback 08:49
uvtc: so my second offer is that I can set up a docs.perl6.org/ system that people can easily hack on
masak I'd like to add to that that a "spike" solution would probably go a long way.
i.e. a working workflow of editable documentation consisting of only one entry. 08:50
moritz so that it sounds "official" from the start, and we can point people to it. Hopefully others will contribute a nice design, as has happened with several Perl 6 related sites already
masak and yes, public.
uvtc moritz: I was writing some docs on my own. I was trying some things out, and remember asking here about lists of lists -- about the difference between (1,2,3) and [1,2,3], I think.
This seemed like a very basic thing to ask, but I recall that the answer seemed convoluted (something about one flattening and one not). 08:51
masak lichtkind had some hopes for November in that regard, but November is unfortunately stuck in bitrot-land.
moritz uvtc: I guess that our understanding of that topic is now much better than it was half a year ago
at least mine is
uvtc moritz: The kind of docs I'm thinking of would be basic bread-and-butter, "here's how you do this and here's how it works" type docs.
moritz (though still not perfect)
uvtc moritz: I'm not sure a docs.perl6.org would be motivating. I know, for me anyway, I'd rather do something on my own server -- in a definitely-unofficial capacity -- and have it mature a bit before going for docs.perl6.org... 08:53
masak I vote for docs.perl6.org 08:54
moritz uvtc: whatever you prefer. I just found that when things looked official, people looked at it, and provided feedback, and that motivated me
masak if nothing else, it might scare us into maintaining the quality of the docs :P
moritz if they don't look official, people say "why isn't this official, if it's the best you have?"
uvtc Ok. I think I need to leave this soup kitchen of thought, as I'm becoming delirious from lack of sleep. G'night. :) 08:56
08:56 uvtc left
masak we make it look official. and we choose a cooler name for the effort than "u4x". or better yet, no name at all ;) 09:00
moritz mail to p6c sent. 09:02
09:03 DarthGandalf joined 09:10 mberends left 09:12 ab5tract joined 09:22 PacoLinux left 09:24 PacoLinux joined 09:26 satyavvd left 09:27 ab5tract left 09:31 ab5tract joined 09:32 risou_awy is now known as risou 09:39 wk left 09:40 ab5tract left
masak it would be fun/interesting to prototype a stream binder as a Perl 6 module. 09:47
09:49 risou is now known as risou_awy
moritz challenging to do right :-) 09:50
and would need some badass signature introspection
09:51 risou_awy is now known as risou
masak badass signature introspection -- isn't that what we want, anyway? 09:52
this might smoke out missing bits.
moritz aye 09:53
masak with my limited vision of the problem domain, it seems to me it's "MMD but longest match wins".
dalek ast: a179684 | moritz++ | S05-mass/rx.t:
remove bogus use of eval as exception catcher from S05-mass/rx.t
09:55
moritz one thing that needs care is to not eagerly evaluate the argument stream 09:56
09:58 satyavvd joined
masak well, one needs to lookahead as far as the longest signature demands. but no further. 10:10
slurpy arrays would perhaps work (and always win) with finite streams. they seem downright anti-social with infinite or not-known-to-be-finite streams. 10:11
moritz you can pass a lazy/infinite stream to a slurpy param
s/pass/bind/
dalek ast: 1de96d9 | moritz++ | S05-mass/rx.t:
some initial rakudo fudging
10:12
masak oh!
10:13 PacoLinux left
masak for some reason I had it as slurpy arrays always being eager. 10:13
moritz nom: sub f(*@a) { say @a[10] }; f 1 ... * 10:14
p6eval nom 3e66bd: OUTPUT«11␤»
moritz so you just have to be careful to only do that binding when nothing before it in the signature failed to bind 10:15
10:17 drbean left, risou is now known as risou_awy 10:18 PacoLinux joined 10:21 drbean joined 10:23 wk joined
masak right. 10:24
moritz std: sub f(*@a, $b?) { } 10:25
p6eval std 580b69a: OUTPUT«===SORRY!===␤Cannot put optional positional parameter after variadic parameters at /tmp/Cgx_sJ37xz line 1:␤------> sub f(*@a, $b?⏏) { }␤Potential difficulties:␤ $b is declared but not used at /tmp/Cgx_sJ37xz line 1:␤------> …
moritz is "variadic" the new "slurpy"?
masak STD has always had it as that, I think.
moritz std: sub f(|$a, $b?) { }
p6eval std 580b69a: OUTPUT«===SORRY!===␤Cannot put optional positional parameter after variadic parameters at /tmp/zuCVdW6uj3 line 1:␤------> sub f(|$a, $b?⏏) { }␤Potential difficulties:␤ $b is declared but not used at /tmp/zuCVdW6uj3 line 1:␤------> …
masak ah, that might be why.
moritz ok, that isn't slurpy. but variadic
I see the point. 10:26
masak right.
10:28 replore left
moritz on a totally off-topic note I've learned some interesting rules about the convergence of fourier series 10:30
if you fourier-transform a continuous function, the coefficients drop at least as fast as 1/omega 10:31
if it's also a smooth function, they drop at least as fast as 1/omega^2 10:32
(I forgot, needs to be periodic too)
</off-topic>, self -> lunch 10:33
tadzik can anyone compile Grammar::Debugger? 10:54
jnthn Worked last time I tried it...note it needs the ANSI::TermColor around. 10:55
tadzik: If it fails in some way, file an issue 10:56
tadzik okay
jnthn: ===SORRY!===
PAST::Compiler can't compile node of type InterventionPoint
10:57 kaleem left
jnthn tadzik: Hm...thought that got fixed a while ago (the enum fix I did) 10:57
Does it work non-precompiled?
back to $class... 10:58
10:58 satyavvd left, PacoLinux left 11:00 kaleem joined
TiMBuS nom: sub f($a, $b, *@c) { say $b; say @c[10] }; f 1 ... * 11:02
p6eval nom 3e66bd: OUTPUT«Not enough positional parameters passed; got 1 but expected at least 2␤ in sub f at /tmp/fOtGpZ2QKK:1␤ in block <anon> at /tmp/fOtGpZ2QKK:1␤ in <anon> at /tmp/fOtGpZ2QKK:1␤»
TiMBuS oh yeah 11:03
lesse here
nom: sub f($a, $b, *@c) { say $b; say @c[10] }; f(|1 ... *)
p6eval nom 3e66bd: OUTPUT«Could not find sub &prefix:<|>␤ in block <anon> at /tmp/pDpXjUlrSD:1␤ in <anon> at /tmp/pDpXjUlrSD:1␤»
TiMBuS not it..
11:04 mj41 left 11:05 PacoLinux joined
TiMBuS ehhn it doesnt matter, if i force it to flatten to an arglist it will run forever won't it. 11:05
11:07 drbean left, im2ee joined
im2ee Hello 11:08
flussence does nom have any equivalent of ng's <?DEBUG(1)> for regex? 11:10
11:10 f00li5h left
tadzik nope 11:13
11:13 drbean joined
tadzik just stumbled upon it 11:13
well, not that I know of
nom: class A { method perl { "foobar" } }; A.new.perl.say 11:16
p6eval nom 3e66bd: OUTPUT«foobar␤»
tadzik hrm 11:17
11:19 mj41 joined 11:21 koban left
colomon flussence: look for jnthn++'s regex tracer / debugger -- that only works with nom. 11:27
flussence thanks, but you said that about 10 seconds after I figured out what was wrong in my code :) 11:28
Str.match doesn't set $/ apparently :(
moritz should it? 11:30
flussence dunno... I was trying to use it for an if {} and couldn't figure out why it returned true, but $/.perl showed a failed match 11:31
masak yay, Karen replied via email about my grant application. 11:32
tadzik ....aaaand? 11:33
masak turns out there's been a bit of a delay due to the zombie Web.pm grant.
moritz are you going to finish it? or GC it? 11:34
masak actually, now that I think about it, that might very well be why Alberto contacted me and Tene about the Web.pm grant the other week.
moritz: I'm/we're going to finish it according to its grant proposal before the end of November, or fail trying. 11:35
that's what we decided with Alberto.
moritz sounds sane.
masak Karen also asked what I think about D4, and how much of a risk it is. I'll need to mull over that one today before I reply. 11:36
11:40 wamba joined
moritz maybe you could decouple it from the grant in some way 11:40
masak D1..D3 are essentially feature additions, whereas D4 is a refactor. refactors are difficult to assess.
11:40 Patterner left, benabik joined
masak moritz: maybe. right now I don't feel that's necessary, though. 11:41
11:42 Psyche^ joined, Psyche^ is now known as Patterner
masak I could try to outline D4 in as much detail as I can discern from my current vantage point. 11:43
sort of recursively find milestones within D4. 11:44
I thought of doing that in reply to my commenters on the grant proposal post anyway.
11:45 [particle] left 11:46 benabik_ joined, benabik_ left, benabik_school joined, [particle] joined, benabik_school left
moritz wouldn't hurt, I think 11:48
masak in fact, that might be the only way to assess the risk involved; to break it down into subtasks. 11:50
interesting article about Dart: arstechnica.com/business/news/2011/...e-them.ars 11:52
11:55 am0c left
dalek p: 15be997 | mls++ | src/HLL/Compiler.pm:
Do not let control exceptions slip through. Also adds a handle-control hook.
12:02
12:04 mkramer1 joined 12:13 SHODAN joined, amkrankruleuen left 12:18 lestrrat left 12:19 lestrrat joined 12:25 SHODAN left 12:27 replore joined 12:29 SHODAN joined 12:34 kaleem left 12:38 kaleem joined 12:41 ab5tract joined 12:44 mtk left 12:45 f00li5h joined 12:47 imarcusthis left
masak perl6: my $N = Inf; say join " on the wall, ", "$N bottles of beer"; say "take one down, pass it around"; $N--; say "$N bottles of beer on the wall" 12:48
p6eval rakudo 3e66bd: OUTPUT«Inf bottles of beer␤take one down, pass it around␤Inf bottles of beer on the wall␤»
..niecza v10-43-gf5298c8: OUTPUT«Infinity bottles of beer␤take one down, pass it around␤Infinity bottles of beer on the wall␤»
..pugs: OUTPUT«sh: /home/p6eval/.cabal/bin/pugs: not found␤»
12:48 drbean left
masak sorear: is it LHF to make 'Infinity' read 'Inf' above? 12:49
niecza: say Inf
jnthn INFINITE BEER!!!
p6eval niecza v10-43-gf5298c8: OUTPUT«Infinity␤»
12:50 mtk joined
masak jnthn: most of it's room temperature, though, since there isn't a fridge large enough to hold it. 12:50
perl6: my $N = Inf; say join " on the wall, ", "$N bottles of beer" xx 2; say "take one down, pass it around"; $N--; say "$N bottles of beer on the wall" 12:51
p6eval rakudo 3e66bd: OUTPUT«Inf bottles of beer on the wall, Inf bottles of beer␤take one down, pass it around␤Inf bottles of beer on the wall␤»
..niecza v10-43-gf5298c8: OUTPUT«Infinity bottles of beer on the wall, Infinity bottles of beer␤take one down, pass it around␤Infinity bottles of beer on the wall␤»
..pugs: OUTPUT«sh: /home/p6eval/.cabal/bin/pugs: not found␤»
masak that's better.
12:53 wolfman2000 left 12:54 drbean joined
dalek albot: e94ed5d | moritz++ | evalbot.pl:
exclude pugs from "perl6:" target while its installation is busted; include b for the lulz
12:55
12:57 p6eval left, p6eval joined, ChanServ sets mode: +v p6eval
masak for the lulz, eh? 13:03
maybe I should switch to this 'b' implementation. it seems to be on the rise in popularity...
13:06 ggoebel left 13:08 awwaiid joined
masak perl6: say "b, are you there?" 13:11
p6eval nom 3e66bd, b 1b7dd1, niecza v10-43-gf5298c8: OUTPUT«b, are you there?␤»
masak \o/
moritz++ # lulz
13:11 Holy_Cow joined 13:17 drbean left
masak perl6: say .perl for (gather for 1, 2, 3 -> $a { for <a b c> -> $b { take ($a, $b) } }) 13:19
p6eval nom 3e66bd, b 1b7dd1, niecza v10-43-gf5298c8: OUTPUT«1␤"a"␤1␤"b"␤1␤"c"␤2␤"a"␤2␤"b"␤2␤"c"␤3␤"a"␤3␤"b"␤3␤"c"␤»
masak perl6: say .perl for (gather for 1, 2, 3 -> $a { for <a b c> -> $b { take [$a, $b] } })
p6eval b 1b7dd1, niecza v10-43-gf5298c8: OUTPUT«[1, "a"]␤[1, "b"]␤[1, "c"]␤[2, "a"]␤[2, "b"]␤[2, "c"]␤[3, "a"]␤[3, "b"]␤[3, "c"]␤»
..nom 3e66bd: OUTPUT«1␤"a"␤1␤"b"␤1␤"c"␤2␤"a"␤2␤"b"␤2␤"c"␤3␤"a"␤3␤"b"␤3␤"c"␤»
masak submits rakudobug
13:20 donri joined 13:21 drbean joined
moritz what's wrong, the .perl? 13:21
13:22 PacoLinux left
moritz nom: say <a b>.perl 13:22
p6eval nom 3e66bd: OUTPUT«("a", "b")␤»
moritz oh, it flattens
masak nom: say [<a b>].perl 13:23
p6eval nom 3e66bd: OUTPUT«["a", "b"]␤»
masak no, the problem isn't .perl, it seems, but &take
moritz I think we have a ticket about take flattening 13:25
masak searches for it 13:26
benabik perl6: say .perl for (gather for 1, 2, 3 -> $a { for <a b c> -> $b { take $a, $b } })
p6eval nom 3e66bd, b 1b7dd1, niecza v10-43-gf5298c8: OUTPUT«1␤"a"␤1␤"b"␤1␤"c"␤2␤"a"␤2␤"b"␤2␤"c"␤3␤"a"␤3␤"b"␤3␤"c"␤»
benabik Right.
moritz rakudo: say (<a b c d e> Zxx 0,1,0,1,0) 13:27
p6eval rakudo 3e66bd: OUTPUT«b d␤»
masak moritz: searches for 'take' and 'flatten' didn't turn up a ticket except the one I just submitted.
moritz: oh, cool! 13:28
moritz rakudo: say (<a b c d e> Zxx 0,1,0,1,0).Str
p6eval rakudo 3e66bd: OUTPUT«b d␤»
moritz so rt.perl.org/rt3/Ticket/Display.html?id=78026 is fixed it seems
masak: I might have confused it with a test or something 13:29
masak I'll mark #78026 as needstest
13:29 wk left
moritz S04-statements/gather.t around line 87 13:30
rakudo: my @list = gather { take [1,2,3]; take [4,5,6];}; say @list.perl
p6eval rakudo 3e66bd: OUTPUT«Array.new(1, 2, 3, 4, 5, 6)␤»
13:31 localhost left
moritz rakudo: sub foo {my @a = (1,2,3,4,5);my $val ;gather {while @a {$val = @a.shift();take $val;}}};say ~( foo() ) ; say foo(); 13:32
p6eval rakudo 3e66bd: OUTPUT«1 2 3 4 5␤1 2 3 4 5␤»
13:32 eiro joined
eiro hello perl6 13:32
moritz and rt.perl.org/rt3/Ticket/Display.html?id=77302 is teh same as #78026 it seems 13:33
masak heiro!
13:33 localhost joined, sayu joined
masak ah, decontainerization. 13:34
13:34 GlitchMr joined
dalek ast: 7b2fd5a | moritz++ | S04-statements/gather.t:
tests for RT #78026, RT #77302, gather/take value decontainerization
13:39
13:43 PacoLinux joined
dalek ecza: 82a2206 | (Solomon Foster)++ | lib/CORE.setting:
Add method form of rand, pick.
13:44
13:44 PacoLinux_ joined
masak oh look, Real has a .rand method! :) 13:49
colomon++ # .pick
13:49 espadrine left
colomon masak: needed it for my project. :) 13:50
13:50 espadrine joined
frettled colomon++ v.nice 13:50
masak is it... a sekkrit project? :>
colomon no, it's the project I've been talking about for a week or two. ;)
not to mention repeatedly asking for help with the Gtk bits of it on here. 13:51
ingy moritz: how did you try to install 7.0.3?
moritz ingy: download tarball, extract it, ran ./configure --prefix=/some/path && make install
masak colomon: oh, oh, something with TreeIter, right? right? ;) 13:56
ooh! www.harmonyware.com/pictures/gtk-treeview.png
Niecza did that? 13:57
colomon yes
masak \o/
that's pretty awesome.
colomon current version also has a (working) column of checkboxes at the front, and records which selections were checked at exit in a log file 13:58
now I just need working pick to so it's easy to randomly select tunes which haven't been checked recently
14:00 imarcusthis joined
colomon masak: the Gtk code is at github.com/colomon/TuneReminder/bl...ractice.pl 14:02
masak ooh
masak devours it
colomon other than a couple of minor weirdnesses, it's very close to what the equivalent C# would be (in terms of Gtk calls)
ie this is a practical, real-world UI that works in Niecza now. 14:03
masak heh, both you and mberends go for sub names in CamelCase ;)
moritz does it feel responsive?
ingy moritz: would you mind running it again and sending me the output?
masak and method names. 14:04
moritz ingy: is it not enought to know that it hits hackage.haskell.org/trac/ghc/ticket/5005 and dies with the same error as in that ticket?
masak colomon: it's refreshing to see Perl 6 code used in this way.
for a GUI, I mean. not the CamelCase :P
ingy moritz: i've looked at all that, I just want to see the full run failure 14:05
14:05 drbean left 14:06 jferrero joined
colomon moritz: I'm not sure the actual interaction parts of the UI are enough to be representative, but certainly the checkbox checking portion of it feels completely responsive. 14:07
masak: I've always used CamelCase for subs in C++, and since I started with a hunk of mberends' Gtk code, I just followed his conventions without thinking about it. 14:08
moritz ingy: moritz.faui2k3.org/tmp/ghc-7.0.3.build.log 14:09
ingy: that's the transcript from script(1) 14:10
14:10 drbean joined
colomon masak: The extremely awesome thing (even better than having a working UI for Perl 6!) is that (as far as I know) there's nothing magic going on to specifically support Gtk. I think any mono/.NET/C# library can be called from Niecza this way. Which means there's actually a chance I may be working on a Perl 6 CAD program before the year is out.... 14:10
masak woo 14:13
that's great news!
frettled wow! 14:14
colomon hopes he's not over-promising what niecza can do... 14:15
14:20 simcop2387 left, alastyr left 14:23 _sri left, simcop2387 joined, dukeleto left 14:24 dukeleto joined 14:26 _sri joined, PacoLinux_ left
ingy moritz: it looks like 5005 was resolved on March 12th and 7.0.3 was released March 27th :\ 14:30
not sure how this could be happening :\
moritz ingy: but in which branch was it resolved?
ingy: I've also tried it with 7.0.4, and it's the same error
ingy I'll ask around 14:31
14:34 drbean left, SHODAN left
TimToady BBC says Dennis Richie has died 14:36
masak aye :/ 14:37
14:37 pmurias joined
masak so does Twitter. 14:37
TimToady *Ritchie
moritz and hackernews. And reddit.
14:38 PacoLinux_ joined
pmurias colomon: what is the CAD software currently written in? 14:38
masak rip dmr
14:40 drbean joined, GlitchMr42 joined, jack-ji joined 14:43 GlitchMr left 14:47 GlitchMr42 is now known as GlitchMr
PerlJam what? 14:48
masak Requiescat in pace, Dennis MacAlistair Ritchie. 14:49
PerlJam :( 14:50
The one thing about getting old that I'm not sure I'm going to handle too well is everyone dying. 14:51
14:52 tyatpi_ joined
PerlJam (particularly when it's me :) 14:52
moritz PerlJam: oh, when it's you, you won't have to handle it :-)
masak PerlJam: over the years, I've gotten the feeling that the ones who are best at handling their own death are Tibetans. they consider it an acquired skill. consider checking it out. 14:55
14:56 imarcusthis left 14:59 alastyr joined 15:01 kaleem left, imarcusthis joined 15:03 mkramer1 left
colomon pmurias: for $work I do CAD libraries in C++. One of the geometry engine companies I work with has a C# version of their geometry engine. 15:05
15:06 PacoLinux_ left 15:09 sayu left
ingy moritz: I just talked to folk on #haskell and they say the fix did not go in until 7.2 and suggested 7.2. 15:10
15:10 PacoLinux_ joined
ingy moritz: then au suggested 7.2 as well 15:10
moritz: you don't get platform but supposedly we don't need it 15:11
moritz ingy: problem is, I don't get Cabal either
ingy moritz: it's just sugar modules basically
15:11 wtw left
ingy moritz: soo... I'll script it all for you 15:11
bbl 15:12
moritz ingy: irclog.perlgeek.de/perl6/2011-10-12#i_4557943
ingy: that's what happen when I try to install Cabal from tar ball on top of 7.2.1
ingy after nuking .cabal? 15:13
moritz yes
15:16 risou_awy is now known as risou
sjohnson perl6.pals 15:16
masak greetings, sjohnson 15:19
sjohnson yo masak
tadzik ( `ー´) 15:22
or, ( `ー´)/
masak ;)
15:22 loloiiiijj joined 15:25 kaare_ joined 15:26 jkeats_ joined
masak perl6: class perl6 { method pals { <sjohnson tadzik masak> } }; say so 'sjohnson' eq any perl6.pals 15:27
p6eval nom 3e66bd, b 1b7dd1, niecza v10-44-g82a2206: OUTPUT«Bool::True␤»
sjohnson tadzik: :)
masak++ # programming powar
masak just goofing off. 15:28
sjohnson (=゚ω゚=) 15:32
15:32 PacoLinux_ left
sjohnson kind of interesting this syntax though 15:33
what does "so" do?
moritz it so boolifies your result
opposite of 'not'
15:33 Trashlord left
masak sjohnson: same as prefix:<?> does, but with looser precedence. 15:34
rakudo: say ?42; say so 42
p6eval rakudo 3e66bd: OUTPUT«Bool::True␤Bool::True␤»
sjohnson that's handy
masak well, it lets me skip the parentheses in the above case :) 15:35
moritz adds "what is &so?" to his FAQ list
masak perl6: class perl6 { method pals { <sjohnson tadzik masak> } }; say ?('sjohnson' eq any perl6.pals)
15:35 Trashlord joined
p6eval nom 3e66bd, b 1b7dd1, niecza v10-44-g82a2206: OUTPUT«Bool::True␤» 15:35
benabik I wonder if we can make "does so" valid syntax. :-D 15:36
masak moritz: "...and why do I seem to need it when I'm playing around with junctions?"
15:36 tokuhiro_ left
sjohnson perl6: say 4? 15:36
moritz masak: that's not frequently asked here
p6eval nom 3e66bd: OUTPUT«===SORRY!===␤Confused at line 1, near "say 4?"␤»
..niecza v10-44-g82a2206: OUTPUT«===SORRY!===␤␤Confused at /tmp/yAykPloVwx line 1:␤------> say 4⏏?␤␤Parse failed␤␤»
..b 1b7dd1: OUTPUT«===SORRY!===␤Confused at line 22, near "say 4?"␤»
sjohnson perl6: say ?4
masak moritz: people ask "what's so?" quite a lot.
p6eval nom 3e66bd, b 1b7dd1, niecza v10-44-g82a2206: OUTPUT«Bool::True␤»
moritz nom: my $x = 3; $x does so Whatever; say so $x
sjohnson perl6: say ?(4)
p6eval nom 3e66bd, b 1b7dd1, niecza v10-44-g82a2206: OUTPUT«Bool::True␤»
nom 3e66bd: OUTPUT«No applicable candidates found to dispatch to for 'infix:<does>'. Available candidates are:␤:(Mu $obj, Positional @roles)␤:(Mu $obj, Mu $role)␤␤ in sub infix:<does> at src/gen/CORE.setting:7364␤ in block <anon> at /tmp/O7qJUQqxuh:1␤ in <anon> at /tmp/O7qJUQqxuh:1␤»…
moritz masak: that's why I put it on my FAQ list :-)
masak moritz: we give them the immediate semantic answer, but don't often tell them why we put it there.
moritz: as long as you explain why it's needed, I'm happy :) 15:37
moritz masak: currently I'm collecting questions, not answers
masak right.
consider it a reminder to myself, then.
there's often a surface question and a deeper question.
moritz sure 15:39
masak "what's &so ?" -- surface answer: "it boolifies a value" -- deep answer: "it causes the junctional value there to collapse"
moritz masak: if you recall my 5-to-6 series, you might remember that I tried really hard to answer the deep questions too :-)
masak you did indeed.
15:40 jedai left
masak that's why I keep linking people to it ;) 15:40
15:40 jedai joined 15:43 dukeleto left 15:44 dukeleto joined, dukeleto left, dukeleto joined 15:45 dukeleto left, dukeleto joined, dukeleto left, dukeleto joined 15:46 dukeleto left
masak by the way, it seems good practice to me to do &so or equiv on any junction before returning it from a routine. 15:46
I call it the principle of "don't return junctions" :P
15:46 dukeleto joined, dukeleto left
masak may need future refinement, but so far it's working out just fine. 15:46
moritz so far I've never been tempted to return junctions from subroutines
15:46 jedai_ joined 15:47 dukeleto joined
masak right. exactly. 15:47
moritz it's mostly not what the caller expects, and thus bad
15:47 Trashlord left, dukeleto left
moritz I use them only as matchers 15:47
so either I smart-match against them, or I pass them to a routine or callback that does
15:47 dukeleto joined, dukeleto left, jedai left
PerlJam that sounds like a challenge for someone to build a "junction oriented" program where every sub returns a junction 15:47
15:48 dukeleto joined, dukeleto left
masak I use them (1) in conditionals, (2) as matchers, (3) in 'where' clauses. 15:48
though arguably (3) is a special case of (2).
moritz PerlJam: in roast there's a small program that solves integer equations by brute-forcing them with junctions
15:48 dukeleto joined, dukeleto left 15:49 dukeleto joined, dukeleto left, dukeleto joined, dukeleto left
moritz it's not exactly a program that returns junctions all over, but it's the closest form of junction abuse in that direction which I've seen so far 15:49
15:50 dukeleto joined, dukeleto left 15:51 dukeleto joined, dukeleto left, dukeleto joined, dukeleto left, Trashlord joined, pmurias left 15:52 dukeleto joined, dukeleto left, dukeleto joined, dukeleto left
jferrero TimToady, Did you meet Dennis Ritchie? 15:53
15:53 dukeleto joined, dukeleto left 15:54 dukeleto joined 15:55 Mowah_ joined
masak decommute & 15:55
15:55 dukeleto left
PerlJam corresponded with dmr via email almost exactly 1 year ago 15:56
15:56 dukeleto joined, dukeleto left 15:57 dukeleto joined
TimToady jferrero: yes, we were at the same conference quite a few times 15:57
PerlJam (I was asking him what he thought about how long C would live back in the 1970s because on #perl6 we were talking about Perl as a hundred year language)
15:58 mj41 left
mls_ niecza: while 0 {LAST {say "hello"}} 15:58
p6eval niecza v10-44-g82a2206: OUTPUT«===SORRY!===␤␤Action method statement_prefix:LAST not yet implemented at /tmp/tTWWOYi1Ii line 1:␤------> while 0 {LAST {say "hello"}⏏}␤␤␤Unhandled Exception: Unable to resolve method statement_level in class Str␤ at /h…
moritz mls_: interesting edge case :-) 15:59
PerlJam and back then, he had no idea that C would even *exist* in 2010
mls_ I think it should say hello
moritz what's the last element of an empty list? :-) 16:00
PerlJam mls_: whyfor? The loop is never entered, ergo it's never executed a"last" time
moritz but I can see your reasoning
TimToady it has to be outside of the while test in any case
which is officially outside of the block
moritz PerlJam: I think LAST is more of an AFTER -- you can't know if an iteration is the last 16:01
16:01 GlitchMr left
moritz PerlJam: you can just save a reference to the LAST closure, and execute it *after* the last iteration 16:01
TimToady so while cond { LAST stuff } turns into something more like { while cond {}; stuff }
16:01 GlitchMr joined
GlitchMr I was disconnected 16:01
16:02 Trashlord left
moritz TimToady: so in plain English, the LAST phaser fires even if the body wasn't ever executed. Right? 16:02
mls_ except is sees the lexicals from the block?
TimToady yes, it also has to act as a closure
GlitchMr ... and in log I don't see question which I've asked...
sjohnson wow, dennis ritchie is dead
GlitchMr split //, "AAA"
How can I do it in Perl 6?
Just asking...
moritz GlitchMr: "AAA".comb
TimToady but if the loop is never entered, the closure isn't closed :)
PerlJam That seems like an artifact of implementation more than anything else. 16:03
16:05 PacoLinux_ joined
colomon Won't the lexicals from the block be kind of weird? 16:05
moritz PerlJam: it's not just an artifact, unless you count computability questions as artifacts too
TimToady the thing is, it may be very inefficient to keep reclosing the closure each time through the loop
colomon TimToady: true, I guess 16:06
TimToady if it's being used as a "finally" without access to lexicals, then there's no problem, I suppose
"ya pays for whatcha uses" 16:07
PerlJam moritz: so ... would you also expect LEAVE phasers to fire on loops that never execute? 16:08
perhaps we should rename START to BEFORE and create a corresponding AFTER 16:09
moritz PerlJam: no, because LEAVE is per invocation (which means "per iteration" in a loop)
PerlJam (START has always looked a little lonely to me)
Though I guess changing START to BEFORE does muck with the semantics. 16:10
TimToady there is nothing corresponding to START because we can't know the future 16:11
START does something at first use, and there's no way to determined "last use"
*mine
at least, no way to determine that till END time, at which point you might as well use END 16:12
16:12 mj41 joined
moritz
.oO( the END is nigh )
16:13
TimToady it's sort nice that "end" can be construed as the antonym of both "begin" and "start"
*sorta
PerlJam no way to know the future? I thought this was #perl6?!? Impossible things become possible all the time! ;-)
TimToady though the words "begin" and "start" don't indicate their asap vs alap relationship in p6 16:14
PerlJam in any case, having a LAST phaser fire on a loop that is never entered would be tres surprising.
TimToady maybe we should have "INIT", "EAGERINIT", and "LAZYINIT" instead ;) 16:15
ingy moritz: can someone else host the pugs p6eval ? 16:16
PerlJam
.oO( INIT to WINIT )
moritz ingy: I can't make sense of that question. Care to rephrase?
16:16 lateau_ joined
ingy moritz: the eval bots or whatever you call them, run on your machine? 16:17
moritz ingy: well, on diakopter's
it could run on a different machine, but it would be lots of effort migrate
ingy i see
moritz of course you can run a second copy that only handles pugs 16:18
feel free to do that
ingy ok. is there a writeup?
moritz no, just a repo. perl6/evalbot/ on github.
ingy ok
moritz -> decommute
16:21 mj41 left 16:24 slavik1 joined
PerlJam reads chromatic 16:25
[Coke] (tickets) as someone who has gone through a LOT of tickets this week, I'd rather have 4 tickets that all have the same root cause but we didn't know it that are all then easy to close later than 0 tickets. 16:28
16:31 sftp joined 16:47 mj41 joined 16:48 jrockway left 16:49 jrockway joined 16:57 kaleem joined 16:59 mj41 left 17:00 GlitchMr left 17:01 mj41 joined, saaki left 17:02 wk joined 17:03 GlitchMr joined 17:04 dakkar left
colomon is reading chromatic 17:05
17:06 mj41 left, ggoebel joined 17:09 Kivutarrr joined 17:10 dukeleto left, dukeleto joined, dukeleto left
GlitchMr rakudo: say 0.1+0.2-0.3 17:11
p6eval rakudo 3e66bd: OUTPUT«0␤»
GlitchMr ...
17:11 dukeleto joined
moritz perl6: ... 17:12
p6eval nom 3e66bd, b 1b7dd1: ( no output )
..niecza v10-44-g82a2206: OUTPUT«␤Unhandled Exception: >>>Stub code executed␤ at /tmp/KTMxvXfKYX line 1 (MAIN mainline @ 1) ␤ at /home/p6eval/niecza/lib/CORE.setting line 2160 (CORE C1034_ANON @ 2) ␤ at /home/p6eval/niecza/lib/CORE.setting line 2161 (CORE module-CORE @ 58) ␤ at /home/…
GlitchMr This is why floats are weir
d
moritz what's weird about 0.1 + 0.2 - 0.3 being 0? 17:14
TimToady those aren't floats
rakudo: say (0.1 + 0.2 - 0.3).WHAT 17:15
p6eval rakudo 3e66bd: OUTPUT«Rat()␤»
TimToady rakudo: say (0.1 + 0.2 - 0.3).perl
p6eval rakudo 3e66bd: OUTPUT«0/1␤»
GlitchMr rakudo: e**pi-pi 17:16
p6eval rakudo 3e66bd: OUTPUT«Could not find sub &pi-pi␤ in block <anon> at /tmp/ORVDFFg1Io:1␤ in <anon> at /tmp/ORVDFFg1Io:1␤»
GlitchMr rakudo: e**pi - pi
p6eval rakudo 3e66bd: ( no output )
GlitchMr rakudo: say e**pi - pi 17:17
p6eval rakudo 3e66bd: OUTPUT«19.9990999409208␤»
masak TimToady: I'm surprised at the notion that LAST runs after a 0-iteration loops. if nothing else, it creates a strange assymmetry with FIRST.
TimToady yes, I'm thinking it shouldn't run on 0-times, since the closure makes no sense
masak I'm in full support of it not running on 0-times. 17:18
if you want AFTER semantics, put it after the loop ;)
17:18 molaf joined
masak if you wanted access to lexicals in the loop without ever running the loop, see a shrink :P 17:18
17:19 dukeleto left, dukeleto joined, dukeleto left 17:20 dukeleto joined, dukeleto left
TimToady it also, coincidentally, allows the LAST block to be rewritten as a goto, if the conditional test can be inlined 17:20
17:20 jack-ji left
masak oh man. in a year or two, I'll take those phasers for granted as part of my programmer vocabulary. and then I'll miss them greatly in all other languages. :) 17:21
17:21 dukeleto joined, dukeleto left
TimToady an important optimization will be figuring out which loop blocks don't actually have to become closures 17:21
masak aye.
17:21 dukeleto joined, dukeleto left 17:22 dukeleto joined, dukeleto left, dukeleto joined 17:23 dukeleto left
moritz masak: there are several solutions to that problem which are all not really desirable/possible 17:23
17:23 dukeleto joined, dukeleto left
moritz :-) 17:23
17:24 dukeleto joined, dukeleto left, dukeleto joined, dukeleto left
masak to what problem? 17:24
GlitchMr: 0.1 isn't a Num (floating-point) in Perl 6, it's a Rat. 17:25
moritz masak: missing phasers in other languages
17:25 dukeleto joined, dukeleto left
masak WORLD DOMINATION 17:25
17:26 tyatpi_ left, bluescreen10 joined, dukeleto joined, dukeleto left
moritz for example :-) 17:26
masak er, I mean, successful employment of PR to increase the mindshare of Perl 6.
of course.
17:26 dukeleto joined, wooden joined, wooden left, wooden joined, dukeleto left
moritz that's what I read from the start :-) 17:27
17:27 dukeleto joined, jdhore left, dukeleto left, bluescreen10 left, dukeleto joined
masak dukeleto: would you stop doing that? :/ 17:27
17:27 dukeleto left, bluescreen10 joined
masak backlog space doesn't exactly grow on trees. 17:28
17:28 dukeleto joined, dukeleto left
ashleydev gosh think if it *was* paper... 17:28
17:28 dukeleto joined
moritz maybe it is paper, and somebody prints out every day of #perl6 backlog 17:30
masak if that printer ever runs out of ink, we're hosed. o.O
oh wait
17:32 Kivutarrr left, jack-ji joined, Kivutarrr joined 17:34 arthur-_ left, ab5tract left, GlitchMr42 joined
[Coke] following up on the "use perl6 now" thread from earlier: for me to use perl6 for projects, I'd need it to either be a browser scripting language or an awesome middle tier web language. Or maybe possibly a PL/SQL replacement. And then I'd have a hard sell internally unless someone was selling support. 17:36
So I've pretty much ruled it out in that niche.
dukeleto masak: sorry, my irc proxy gets into fits occasionally
masak dukeleto: I'm just glad you're OK :) 17:37
17:37 GlitchMr left 17:38 GlitchMr42 is now known as GlitchMr
dukeleto masak: i use dircproxy, and sometimes I forget to register my nick and then freenode and dircproxy start dueling with each other 17:40
masak: or, i forget to identify for my registered nick
17:41 jdhore joined 17:42 donri left 17:44 donri joined, jack-ji left 17:48 tyatpi_ joined, PacoLinux_ left
masak ah. 17:50
17:51 molaf left
diakopter was the "use perl6 now" thread on irc or email? 17:52
17:52 wk left 17:53 arthur-_ joined
[Coke] discussion on IRC about chromatic's latest blog posting. maybe his last 2. 17:54
about 10 hours ago.
17:54 wk joined 17:55 Chillance joined 17:56 jack-ji joined 17:58 daniel-s left 17:59 am0c joined
masak diakopter: tl;dr: it's 2011. why aren't there more overwhelming reasons to use Perl 6 rather than some other language? 18:00
18:03 wk left, jkeats_ left
diakopter blackogs 18:03
sjohnson diakopter: hi 18:04
diakopter hi 18:06
sjohnson how's it going? 18:13
18:14 Elizabeth_ joined, risou is now known as risou_awy 18:16 wk joined
[Coke] moritz: is S05-mass/rx.t now legitimately unfudgable? 18:26
(it's no longer crashing, just failing some tests (and passing some TODOs) 18:27
18:34 jack-ji left 18:39 slavik1 left
ingy types his first irc msg from oneiric ocelot 18:40
[Coke] phenny: ask moritze if S05-mass/rx.t now legitimately unfudgable? (it's no longer crashing, just failing some tests (and passing some TODOs) 18:41
phenny [Coke]: I'll pass that on when moritze is around.
[Coke] phenny: ask moritz if S05-mass/rx.t now legitimately unfudgable? (it's no longer crashing, just failing some tests (and passing some TODOs)
phenny [Coke]: I'll pass that on when moritz is around.
18:42 PacoLinux left, slavik1 joined
ingy unity seems better than on natty 18:42
progress++
TimToady spent >< that much time futzing with unity before switching
ingy spent > < that much 18:43
TimToady: are you still on natty?
masak why are you all doing Cartman impressions? 18:44
ingy masak: wanna pair on a pugsbot?
masak tempting. please specify task some more.
ingy masak: masak you ssh into my server and join my screen session and we install pugs and p6eval 18:45
diakopter ingy: what do you mean by pugsbot 18:46
ingy pugs: say "fail"
p6eval pugs: OUTPUT«sh: /home/p6eval/.cabal/bin/pugs: not found␤»
diakopter surely that is fixable.
flussence perl6: fail "say"
p6eval niecza v10-44-g82a2206: OUTPUT«===SORRY!===␤␤Undeclared routine:␤ 'fail' used at line 1␤␤␤Unhandled Exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 766 (CORE die @ 2) ␤ at /home/p6eval/niecza/src/STD.pm6 line 1136 (STD P6.comp_unit @ 36) ␤…
..nom 3e66bd, b 1b7dd1: ( no output )
TimToady ingy: yes, but I hit the upgrade button > < minutes ago :) 18:47
masak ingy: do you simply mean fixing p6eval?
sounds easier than writing a whole new bot for Pugs. 18:48
TimToady ingy: does the upgrade switch back to unity if you switched away?
ingy TimToady: I'll give you a unity report next week
in person
moritz [Coke]: yes, you can (un)fudge rx.t if you have the necessary patience :-)
phenny moritz: 18:41Z <[Coke]> ask moritz if S05-mass/rx.t now legitimately unfudgable? (it's no longer crashing, just failing some tests (and passing some TODOs)
ingy masak: moritz cannot install pugs on diakopter's machine and after 24 hours of investigation I am convinced it cannot easily be done 18:49
moritz I've fudged out the missing <?before> and fixed up the evals, but didn't want to do the rest
ingy masak: so moritz said we could just install p6eval elsewhere
masak ah.
moritz (at least a copy of p6eval that powers pugs) 18:50
masak pugseval :)
ingy ♥ lightbulbs
ingy hates the default terminal font in unity
masak: if you are interested it should be really easy... 18:52
diakopter moritz: did you try installing older version of ghc?
masak it sounds like fun but 18:53
I'm afraid I don't have time to pair up tonight. :/
ingy diakopter: moritz tried EVERYTHING
moritz diakopter: yes; there pugs' dependencies don'T build
diakopter: I've tried 4 different GHC versions
ingy it's an insane set of fail
masak nice. I've never seen "don'T" before, but it kinda makes sense.
18:53 Holy_Cow left
ingy even au gives up 18:53
masak ingy: but you installed Pugs locally, didn't you? 18:54
moritz masak: it's the "shift key still pressed from typing '"
ingy yes 18:55
moritz masak: problem is that the kernel on p6eval's host is old-ish
masak oh!
moritz masak: and while GHC would probably build, but the GHC that is used for bootstrapping has been compiled a newer kernel...
ingy haskell is having some growing pains
masak upgrading the kernel not an option? 18:56
moritz and has thus selected a newer version of a syscall which my old kernel doesn't help
diakopter nope :(
moritz masak: not so easy on a virtualized machine
masak ah, right.
sounds like Mr Different Server to the rescue, then.
ingy moritz: which is fixed in 7.2, but then cabal fails, yes?
moritz there's a fix for GHC, but it's on in The New Version (7.2) 18:57
but the rest of the toolchain doesn't work on 7.2
ingy masak: let me set up my pair station
bbiab
moritz (there's a big warning on the GHC download page that you should rather use 7.0.3. Which I can't. Thank you)
masak ingy: I... I don't have time... :/ 18:58
ingy: I'm in the middle of doing some residual $dayjob and working on the macros branch and doing a thing with Niecza and looking for a new apartment... 18:59
ingy ok 19:01
TimToady: gnome classic is not installed by default :) 19:03
TimToady: sudo apt-get install gnome-session-fallback
seems to be the trick
19:03 kaleem left 19:04 loloiiiijj left
ingy masak: ping 19:05
masak pong
[Coke] moritz: it will be a welcome relaxation after working on CF all day. 19:06
ingy masak: get back to doing some residual $dayjob and working on the macros branch and doing a thing with Niecza and looking for a new apartment...
:)
masak yeah... 19:08
sorry for not being more available right now.
it's fun to pair up with you, ingy :)
I'm hoping we'll have some fun together in PugsClub in the near future. 19:09
ingy TimToady: ironically (for some definition of ironic) when you try to install that it will try to get the flash plugin from a special url, because the flash plugin will fail during your upgrade, and that url is being ddossed at the moment, so have fun using your forced Unity trial :)
moritz that's just because flash is nasty non-free software, and can't be distributed by Debian|Ubuntu directly 19:10
TimToady well, I guess I'm gonna be helping with that ddos
moritz sometimes I believe that RMS is more right than he has any right to be 19:11
TimToady 'course, I use the noflash plugin... 19:12
sjohnson someone from #gnu asked why a program i was working on wasn't free
TimToady but it's still gonna try to install it, I'm sure
sjohnson my answer: cause i'd lose my job?
</gpl rant> 19:13
ingy oh right, Unity takes over the Super key. Therefore useless...
tadzik if you didn't have a job you'd have to eat stuff from your feet *SCNR*
TimToady yes, that griped me last time
ingy sjohnson: what's wrong with losing your job? 19:14
sjohnson tadzik: if i didn't have a job... ( `ー´)would become -> (`-') 19:15
tadzik :)
sjohnson ingy: i'd die of starvation?
tadzik that'd be sad
sjohnson well, i don't think i'd die.
ingy sjohnson: I doubt it
sjohnson but i'd be in financial woe
i'm already broke enough as it is!
ingy sjohnson: give away all your possessions and join the Cult of Perl 6 19:17
We hang out in airports
sjohnson heh
masak preaching about butteryflies 19:18
argh
battery flies.
sjohnson showering ourseves with wet paper towels in the restrooms 19:19
ourselves*
masak living the dream.
moritz ingy: if you give away all of your posessions, how do you use Perl 6, without a computer? :-) 19:20
ingy moritz: you think I'm using a computer? 19:21
slavik1 moritz: is a computer required to compile and use perl? why not do it in your head?
moritz ingy: I'm pretty sure
slavik1: memory constraints :-) 19:22
ingy tell me about it
slavik1 moritz: need more paper
moritz ingy: how do you use IRC, without any form of computer?
ingy oneiric has 5.12.4
masak moritz: abacus.
tadzik fair enough, I'm staying on arch
19:23 envi_ left
ingy moritz: it's hard to explain... very mystical... 19:24
GlitchMr niecza> Bool::False::False 19:25
Funny merge failure ::CORE:30::False (Bool, False)
lolwhat?
niecza: Bool::False::False
p6eval niecza v10-44-g82a2206: OUTPUT«␤Unhandled Exception: Funny merge failure ::CORE:30::False (Bool, False)␤ at line 0 (boot-MAIN @ 0) ␤ at line 0 (ExitRunloop @ 0) ␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 766 (CORE die @ 2) ␤ at /home/p6eval/niecza/src/NieczaBackendDotnet.…
GlitchMr rakudo: Bool::False::False
p6eval rakudo 3e66bd: OUTPUT«===SORRY!===␤Confused at line 1, near "Bool::Fals"␤»
GlitchMr ...
std: Bool::False::False
p6eval std 580b69a: OUTPUT«===SORRY!===␤Undeclared name:␤ 'Bool::False::False' used at line 1␤Check failed␤FAILED 00:01 118m␤»
GlitchMr perl6: Bool::False::False
p6eval niecza v10-44-g82a2206: OUTPUT«␤Unhandled Exception: Funny merge failure ::CORE:30::False (Bool, False)␤ at line 0 (boot-MAIN @ 0) ␤ at line 0 (ExitRunloop @ 0) ␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 766 (CORE die @ 2) ␤ at /home/p6eval/niecza/src/NieczaBackendDotnet.…
..b 1b7dd1: OUTPUT«Cannot find sub Bool::False::False␤ in main program body at line 1:src/metamodel/RoleToInstanceApplier.nqp␤»
..nom 3e66bd: OUTPUT«===SORRY!===␤Confused at line 1, near "Bool::Fals"␤»
GlitchMr OK... 19:26
ingy TimToady: I survived the ddos
brb
TimToady "Funny merge failure" is a funny error message :) 19:27
sjohnson is Fals on purpose?
TimToady but I imagine it really means "failure to unify different notions of global"
moritz niecza seems to full of those
sjohnson 4 letter opposite of True?
moritz slavik1: there is no Fals, except where 'False' was cut off 19:28
erm sorry, meant sjohnson
masak "C3 wedgie"!
sjohnson :)
oic
GlitchMr perl6: /[([])]/
moritz masak: that's the one I was thikning of
p6eval nom 3e66bd: OUTPUT«===SORRY!===␤Unrecognized regex metacharacter (must be quoted to match literally) at line 1, near ")]/"␤»
..niecza v10-44-g82a2206: OUTPUT«===SORRY!===␤␤Null pattern not allowed at /tmp/EUMfm0n7pI line 1:␤------> /[([⏏])]/␤␤Parse failed␤␤»
..b 1b7dd1: OUTPUT«===SORRY!===␤Confused at line 22, near "/[([])]/"␤»
TimToady niecza has that one nailed 19:29
moritz aye 19:30
GlitchMr perl6: Bool->new
masak worth rakudobug reporting?
p6eval niecza v10-44-g82a2206: OUTPUT«===SORRY!===␤␤Unsupported use of -> as postfix; in Perl 6 please use either . to call a method, or whitespace to delimit a pointy block at /tmp/1g4FWnF5xg line 1:␤------> Bool->⏏new␤␤Parse failed␤␤»
..b 1b7dd1: OUTPUT«===SORRY!===␤Unsupported use of -> as postfix; in Perl 6 please use either . to call a method, or whitespace to delimit a pointy block at line 22, near "new"␤»
..nom 3e66bd: OUTPUT«===SORRY!===␤Unsupported use of -> as postfix; in Perl 6 please use either . to call a method, or whitespace to delimit a pointy block at line 1, near "new"␤»
TimToady not really
moritz masak: nope, known
GlitchMr ... ok...
perl6: Bool.new
p6eval nom 3e66bd, b 1b7dd1, niecza v10-44-g82a2206: ( no output )
TimToady it's just the Non-convergence-with-STD bug
moritz masak: nqp/rakudo don't really recognize null patterns
TimToady perl6: say Bool.new 19:31
p6eval nom 3e66bd, b 1b7dd1, niecza v10-44-g82a2206: OUTPUT«Bool::False␤»
GlitchMr Think negatively... with new falses!
TimToady ooh, widespread agreement for something completely unspecced!
moritz TimToady: exactly my thought :-) 19:32
GlitchMr I was trying to do .new just because of Perl 5 :P.
I was wondering of reaction for this... 19:33
moritz probably because an Int holding the boolean value internally is initlialized to 0 or so
ingy TimToady: they seem to have made gnome classic feel really crappy
it's definitely not the same
TimToady GlitchMr: but Perl 5 has no Bool type, so I don't know what you were expecting :)
GlitchMr Well, it's object :P
moritz what does it object to?
masak TimToady: it's spec'd implicitly. 19:34
GlitchMr So I was wondering what would happen if I would do .new on it
masak Bool::False is isomorphic to 0, the default default.
ingy TimToady: you should study ubuntu's techniques to push people from Perl Classic to Perl Unity\w6
GlitchMr perl6: Int.new
perl6: say Int.new
right...
masak Perl Plurality
p6eval nom 3e66bd, b 1b7dd1, niecza v10-44-g82a2206: ( no output ) 19:35
niecza v10-44-g82a2206: OUTPUT«␤Unhandled Exception: System.InvalidCastException: Cannot cast from source type to destination type.␤ at Niecza.CtxIntStr.Get (Niecza.Variable obj) [0x00000] in <filename unknown>:0 ␤ at Niecza.CtxBoxify`1[System.String].Get (Niecza.Variable obj) [0x0000…
..nom 3e66bd, b 1b7dd1: OUTPUT«0␤»
GlitchMr ...?
It's ok for Niecza until it has to actually say it... O_o
masak GlitchMr: when you don't spell out your questions, it's hard to know what you're asking.
yeah, that's odd.
GlitchMr I'm actually surprised about it
masak maybe it's evaluating lazily. 19:36
:)
niecza: Int.new; say "alive"
p6eval niecza v10-44-g82a2206: OUTPUT«alive␤»
masak huh.
moritz niecza: Int.new.gist; say 'alive'
GlitchMr masak, you're not using that value
masak niecza: my $a = Int.new; say "alive"
TimToady perl6: say 1 + '2x'
p6eval niecza v10-44-g82a2206: OUTPUT«␤Unhandled Exception: System.InvalidCastException: Cannot cast from source type to destination type.␤ at Niecza.CtxIntStr.Get (Niecza.Variable obj) [0x00000] in <filename unknown>:0 ␤ at Niecza.CtxBoxify`1[System.String].Get (Niecza.Variable obj) [0x0000…
niecza v10-44-g82a2206: OUTPUT«Potential difficulties:␤ $a is declared but not used at /tmp/1YdrBdc5NW line 1:␤------> my ⏏$a = Int.new; say "alive"␤␤alive␤»
niecza v10-44-g82a2206: OUTPUT«␤Unhandled Exception: System.FormatException: Unknown char: x␤ at System.Double.Parse (System.String s, NumberStyles style, IFormatProvider provider) [0x00000] in <filename unknown>:0 ␤ at System.Double.Parse (System.String s, IFormatProvider provider) […
..nom 3e66bd: OUTPUT«No applicable candidates found to dispatch to for 'Numeric'. Available candidates are:␤:(Mu, Mu %_)␤␤ in method Numeric at src/gen/CORE.setting:601␤ in sub infix:<+> at src/gen/CORE.setting:1941␤ in sub infix:<+> at src/gen/CORE.setting:1941␤ in block <anon> at /t…
..b 1b7dd1: OUTPUT«3␤»
masak ah :)
GlitchMr It's ok to do 1 + '2x'... until you have some weird characters 19:37
moritz nom doesn't really handle failures gracefully
GlitchMr "x" is unexpected... weird...
TimToady and niecza doesn't handle it at all :)
moritz do you expect an 'x' when parsing a string as a number?
GlitchMr Nope
19:38 PacoLinux joined
moritz so why do you find it weird that x is unexpected? 19:38
GlitchMr C:\Users\Konrad\Desktop\run\run>php
<?php echo 1 + '2x';
^Z
3
But that's probably not reason
Strict typing is better :).
PerlJam GlitchMr: this is #perl6, not #php6 :)
GlitchMr lol
There is no PHP 6 :P. 19:39
That was PHP 5.3.8...
Perl 5 also does it like PHP 19:40
ingy TimToady: I feel like gnome has been made non-acceptable option. So looks like I'll have to futz > <
GlitchMr But then, I'm not sure if it's good
And PHP was made for parsing data coming from Internet which might be intentionally invalid to make errors.
ingy TimToady: I'll be interested to get your take
GlitchMr So yeah... 19:41
Except Perl 5 seems to make warnings in strict mode
Argument "2x" isn't numeric in addition (+) at - line 1.
3
So yeah....
PHP doesn't care...
moritz warnings and strict are orthogonal
wow, cabal seems to repeat the mistakes from CPAN.pm 19:42
Pugs-6.2.13.20111008 depends on haskeline-0.6.4.3 which failed to install.
haskeline-0.6.4.3 failed during the building phase. The exception was:
ExitFailure 1
GlitchMr PHP is probably only programming language which doesn't care at all about 1 + '2x';...
PerlJam yet warnings and strict always seem to be mentioned in the same utterance ;)
moritz so it caches build failures, but doesn't tell me what went wrong the last time
PerlJam moritz: it doesn't write a log somewhere? 19:43
masak PerlJam: that's because they're both good ideas, not because they're the same idea :)
moritz PerlJam: dunno. If it does, it doesn't tell me about it
PerlJam masak: sure ... like PB&J or beans & rice or pots & kettles :)
19:50 kaare_ left
masak PerlJam: I had to look up "PB&J". you must be from the US. :P 19:52
PerlJam masak: stroopwafels & icing then ? 19:54
(what goes with stroopwafels?)
masak: nah, I'm not from the US ... I'm from Texas ("it's like a whole other country" according to the tourism ads I see on TV)
TimToady misses BLTs
Texans *think* they are from a different country, but Californians really are... 19:55
masak PerlJam: do they say "whole other" or "whole 'nother"? the latter sounds more genuine to my ears.
ingy TimToady: Gnome will be going the way of the BLT for you very soon 19:56
PerlJam masak: you know ... I'm not really sure. I never listen that closely
TimToady in which case I will preach dis-unity with Ubuntu
ingy I for one welcome our new Unity overlords :\
actually it's not called Unity anymore
it's called Ubuntu 19:57
hmm
what does that remind me of...
PerlJam masak: www.traveltex.com/ "It's like a whole other country" (at least in print ;)
ingy begins with a P, I think
TimToady maybe I should cancel my upgrade
ingy TimToady: not a bad idea
PerlJam unity -- just counting the number of people that actually like it 19:58
ingy you will not have what you had before, and it seems to have thrown away many settings
gnome is ass
masak PerlJam: :)
ingy Uwhatever is the only way to go 19:59
masak PerlJam: I think it's mostly in spoken language that it happens. or written language wanting to mimic spoken language.
ingy but I'm already tripping and applying bandaids
awwaiid xmonad works pretty good in ubuntu :)
ingy then again, that's the spirit of Perl 6 in general 20:00
TimToady ah well, Ubuntu was good while it lasted
awwaiid (but I switched back to debian in protest anyway)
ingy TimToady: the new stuff looks a whoole lot better than 6 months ago 20:01
kinda like Perl 6
TimToady I'll let other people guinea pig it for me
not interested in losing my key mappings
ingy I didn't lose all of them :) 20:02
20:03 kfo_ joined
TimToady bbl & 20:05
djanatyn rd |<automatous> exit [13:01]
Ermm, sorry about that.
20:07 kfo left
diakopter at least it was only one line :D 20:07
20:08 GlitchMr left
djanatyn yep :) 20:09
dalek ecs: 1770286 | diakopter++ | S05-regex.pod:
S05 typo
20:14
20:14 Kivutarrr left 20:15 huf left 20:18 wamba left 20:24 huf joined, PZt left, djanatyn is now known as secret-djan 20:25 secret-djan is now known as djanatyn 20:32 wamba joined, wamba left 20:34 soh_cah_toa joined 20:35 Lma joined
[Coke] S05-mass/rx.t is slooow 20:37
20:45 dual joined
masak [Coke]: probably not unusually slow, just the single largest test file in roast. 20:53
ingy sudo apt-get install compizconfig-settings-manager
sorry
but you WILL need that
masak :)
ingy so, you're welcome ;)
I almost got things back to normal except I can't stop Super+t opening trash 20:54
I use super for all my term settings
so super+t opens tab 20:55
oh well
super+cntl+t for now 20:56
20:57 saaki joined 20:59 wamba joined 21:07 lateau_ left 21:12 wamba left 21:13 PZt joined 21:16 mj41 joined 21:19 bluescreen10 left
im2ee Dobranoc! :) 21:20
o/
21:20 mj41 left, im2ee left
saaki reception supposed to start at 2:30, but there's already like 20 people here 21:25
oops
21:27 wamba joined 21:29 wk left 21:31 mj41 joined 21:40 bluescreen10 joined, Mowah_ left
soh_cah_toa perl6: say $?FILE 21:43
phenny soh_cah_toa: 11 Oct 05:55Z <moritz> tell soh_cah_toa that source_url in raw.github.com/soh-cah-toa/p6-dige.../META.info points to the old repo name (without the p6- prefix)
p6eval b 1b7dd1: OUTPUT«===SORRY!===␤Symbol '$?FILE' not predeclared in <anonymous> (/tmp/MxUFFgiRRD:22)␤»
..nom 3e66bd: OUTPUT«===SORRY!===␤Symbol '$?FILE' not predeclared in <anonymous> (/tmp/l8gVicUfia:1)␤»
..niecza v10-44-g82a2206: OUTPUT«/tmp/U5JBE8cATX␤»
soh_cah_toa lame... 21:45
21:45 tyatpi_ left
colomon huh. it gave you the result in all three, but spurious warnings in both rakudo versions? 21:45
21:46 wamba left 21:47 Lma_ joined, Lma_ left, dorlamm joined 21:48 wolfman2000 joined
masak well, I got residual $dayjob and apartment search done today. at least that's something ;) 21:49
tomorrow I'll do some Perl 6, I promise. 21:50
'night, batteryflies.
21:50 Lma left 21:54 wamba joined 21:58 wamba left 22:00 wamba joined, wamba left 22:02 wamba joined 22:04 cognominal_ joined, jevin left 22:05 mj41 left
soh_cah_toa phenny: tell moritz good eye. i took care of it. thanks :) 22:06
phenny soh_cah_toa: I'll pass that on when moritz is around.
22:07 cognominal left 22:12 aqz left 22:14 jevin joined 22:15 mikemol left, Exodist left 22:16 mikemol joined, aqz joined 22:17 Exodist joined 22:21 benabik left
diakopter niecza: print "{ "foo" ~~ /f(<perl>)o/ }" 22:24
p6eval niecza v10-44-g82a2206: OUTPUT«␤Unhandled Exception: System.InvalidCastException: Cannot cast from source type to destination type.␤ at RxFrame.IncorporateChild (System.String[] names, Boolean passcap, Niecza.P6any match) [0x00000] in <filename unknown>:0 ␤ at RxFrame.IncorpShift (Sys…
22:24 colomon left 22:34 floyy joined 22:36 whiteknight joined
dalek ast: a3f34bf | Coke++ | S05-mass/rx.t:
rakudo fudging
22:37
22:37 donri left
[Coke] fudges S05-mass/rx.t 22:38
dalek kudo/nom: ce15be4 | Coke++ | t/spectest.data:
run fudged test
diakopter phenny: ask TimToady is the following supposed to work (S05 Named scalar aliasing to subpatterns) std: "foo" ~~ ms/ $<key>=( f o o ) /; say $/<key> 22:43
phenny diakopter: I'll pass that on when TimToady is around.
22:44 tokuhiro_ joined 22:50 replore left
sjohnson bc == BNU Calculator 22:51
22:52 sivoais left 22:53 donri joined, wolfman2_ joined 22:56 wolfman2000 left 22:57 dorlamm left 23:01 wamba left 23:07 DarthGandalf left 23:09 bluescreen100 joined 23:13 DarthGandalf joined 23:16 lestrrat left 23:17 lestrrat joined, soh_cah_toa left
diakopter std: "foo" ~~ rx<a> 23:19
p6eval std 580b69a: OUTPUT«ok 00:01 122m␤»
23:26 Tene left, Tene joined, Tene left, Tene joined, djanatyn left 23:27 djanatyn` joined
TimToady this vv can't work because identifiers can't have spaced in 'em 23:28
phenny TimToady: 22:43Z <diakopter> ask TimToady is the following supposed to work (S05 Named scalar aliasing to subpatterns) std: "foo" ~~ ms/ $<key>=( f o o ) /; say $/<key>
TimToady *spaces
diakopter std: ms/ $<key>=( (<[A..E]>) (\d**3..6) (X?) ) /; # ok I'll paste the example exactly 23:29
p6eval std 580b69a: OUTPUT«===SORRY!===␤Two terms in a row at /tmp/p8by_hMWLE line 1:␤------> ms/ $<key>=( (<[A..E]>) ⏏(\d**3..6) (X?) ) /; # ok I'll paste the␤ expecting any of:␤ bracketed infix␤ infix or meta-infix␤ statement modifier loop␤Undeclared
..r…
diakopter (that's what it has in S05) 23:30
23:34 cognominal joined 23:35 cognominal_ left 23:48 jkeats_ joined 23:49 Elizabeth_ left 23:50 Trashlord joined
TimToady probably shouldn't be an ms 23:54
23:55 jkeats_ left, Trashlord left
diakopter what's an ms 23:55
TimToady m:s 23:56
diakopter match sigspace?
TimToady nod
diakopter whee; I might actually learn S05 this time through
23:57 Trashlord joined, Trashlord is now known as Guest34898
diakopter what does "match null before a letter, don't capture" in S05 mean 23:58
for "A leading ? ..."
23:59 wolfman2_ left, shinobicl_ joined
diakopter is that like ?= ? 23:59
TimToady means it's a 0-width assertion, and doesn't end up in $<alpha>