»ö« | perl6.org/ | nopaste: paste.lisp.org/new/perl6 | evalbot usage: 'perl6: say 3;' or rakudo: / pugs: / std: , or /msg p6eval perl6: ... | irclog: irc.pugscode.org/ | UTF-8 is our friend!
Set by wolfe.freenode.net on 30 October 2009.
00:00 fridim_ left 00:01 pmurias joined, nihiliad left
Tene masak, jnthn: a standard solution to the intimidation problem is to separate into two channels, one being new-user-oriented. 00:01
pmurias is unsure if he should git svn dcommit mid migration to new STD mildew 00:02
Tene I remember that #perl6 used to be about Perl 6 in general, and #pugs was where pugs dev happened, iirc...
#perl6-devel, maybe...
#rakudo
masak but I like new users here! I just want them to be braver.
pmurias Tene: #pugs i don't remember 00:03
masak Tene: don't think it was ever #pugs.
it was #perl6 in early 2005 when I got here.
pmurias especially as autrijus created #perl6 00:04
Tene masak: i don't trust my memory for anything... not surprised I'm wrong.
masak: might be helpful to identify why we're intimidating, and consider whether moving some of the intimidating discussion to a back room would be helpful. 00:05
masak I'm not saying splitting into two channels would be a bad move... I'd probably hang out in both channels. but I like how #perl6 is all over the place, everything from lolspeak to compiler internals.
Tene masak: I agree, I don't like the idea of pushing new users to a different channel.
masak: Me too. :)
masak we even have some lolspeak in our compiler internals. :) 00:06
diakopter what? you mean all new Perl 6 users shouldn't have to pass the #perl6 gauntlet?
Tene There shouldn't be a gauntlet, no. 00:08
00:10 jferrero left
Tene Okay, going home now. The GF is dragging me out for some socializing, but I plan to lurk in the corner with my laptop the entire night hacking on Parrot. 00:12
00:12 dakkar left
Tene any requests for ng features? 00:12
diakopter heh 00:13
trailing while/until?
Tene Sure. :)
00:17 mberends left
diakopter pmurias, mberends: perl_vs_v8.pastebin.com/d2c12b8a6 a simple/dumb microbenchmark. v8 has higher startup time, but its loop scales much better 00:17
just missed mberends
masak don't worry, he backlogs. 00:19
pmurias diakopter: would would be more intresting is a v8 vs C comparision 00:21
diakopter :P
pmurias diakopter: and measuring the cost of abstraction is more important than numerics
00:21 hcchien left
diakopter yes, a Perl scalar is just about the same level of abstraction as a JavaScript variable... 00:23
has very similar semantics, I mean
sjohnson diakopter: is there a way to dump available methods for a type? 00:25
rakudo: say Array().methods; 00:26
p6eval rakudo 7ce13d: invoke() not implemented in class 'ResizablePMCArray'␤in Main (file <unknown>, line <unknown>)␤
pmurias diakopter: what i worry much is how having an extra layer (or two) of method calls will affect performance
* much more
jnthn sjohnson: .^
rakudo: say Array.^methods
p6eval rakudo 7ce13d: ScalarshiftArraypushspliceexistsunshiftpopitemvaluesdelete␤
jnthn rakudo: say Array.^methods>>.name.join(' ') 00:27
p6eval rakudo 7ce13d: pop item values delete Scalar shift Array push splice exists unshift␤
00:27 hcchien joined
zaslon lolperl6adventhazblogged! perl6advent++ 'Failure()<0xb78752d4>': Failure()<0xb785928c> 00:27
sjohnson thanks jnthn
jnthn: is there such thing as a List()? 00:28
colomon_ oh cool, zaslon is tracking perl6advent! (if awkwardly...)
I'm still working on the "table of contents" post....
masak yes, moritz_ added perl6advent earlier today.
colomon_++
zaslon lolperl6adventhazblogged! perl6advent++ 'Failure()<0xb785efe0>': Failure()<0xb7846fd0> 00:30
colomon_ It's short, but it's there now. (Must read to my boy a bit.) 00:32
I'm going to tackle the day 1 "How do you get rakudo working?" post in a bit.
zaslon lolperl6adventhazblogged! perl6advent++ 'Perl 6 Advent Calendar': perl6advent.wordpress.com/2009/12/0...-calendar/ 00:33
colomon_ btw, feel free to suggest changes or just make them if you have permission. :)
jnthn sjohnson: List is how you refer to the type object in the namespace. List() is how it stringifies. List(...) would be a coercion, but that's NYI. 00:37
sjohnson jnthn: my reason for curisoity is because i don't see pick and sort in that array method list
jnthn sjohnson: I was gonna say "yeah, they're inherited" but...hm. 00:38
.methods without :local is meant to show the whole load up the tree.
sjohnson hmm 00:43
not sure how that is accomplished
*scratches head*
jnthn rakudo: class A { method foo { } }; class B { method bar { } }; say B.^methods>>.name.join()
p6eval rakudo 7ce13d: ( no output ) 00:44
jnthn offs
I'd expect that to start "bar foo ..." anyway.
masak jnthn: it starts 'bar' locally, but 'foo' comes much later. 00:47
colomon_ jnthn: do the normal instructions for building rakudo work on windows? 00:48
jnthn masak: orly? 00:51
colomon_: Should.
masak oh, wait.
jnthn colomon_: Just remember to invoke make with the right name.
masak it should be 'B is A'... :) 00:52
colomon_ jnthn: what compiler would that be using?
jnthn masak: oh!
duh yes :-)
colomon_: I build with MS VC++
masak right. 'bar', 'foo'. :)
jnthn colomon_: e.g. the Visual C++ compiler
masak: Yeah, sorry. :-)
Perl 6 ain't *that* magical.
masak jnthn: apparently, I accepted it without reflecting. :) 00:53
jnthn masak: It *is* nearly 2am. ;-)
masak that doesn't say much. :) I'm not even a little tired... 00:54
00:54 pmurias left
jnthn masak: Yes, same. :-/ 00:54
nbrown colmon_: I use mingw on windows and that works too
jnthn masak: I gotta get up at 7am-ish for a flyg on Wednesday too.
masak a fokker? :) 00:55
jnthn I fokking hope not. :-P
colomon_ nbrown: Thanks.
jnthn masak: Ryanair don't have any of those.
masak maybe just as well. 00:56
jnthn masak: Just aluminum tubes painted blue and yellow inside with non-reclining seats and 2,50 EUR coffee.
nbrown colomon_: no problem, just thought more solutions is better
colomon_ nbrown: absolutely. there's more than one way...
jnthn masak: Yeah. I did a Fokker once.
masak: It was...an experience. :-)
masak it was a pleasant experience, my fokker I mean.
jnthn masak: I didn't especially dislike it. Sadly, every time I fly these days I seem to get more nervous about it, not less. :-/ 00:57
Meta-model compositon papers seem to help though. :-)
masak :)
jnthn Ooh...I should print another @interesting-paper off for the journey on Wed. 00:58
masak oh, definitely.
jnthn Or maybe 2 given I've got 4-5 hours of train afterwards.
masak for great justice.
01:06 cdarroch left 01:09 colomon_ left 01:15 cognominal joined 01:19 colomon joined
masak 'night. :) 01:22
01:22 masak left
colomon 'night. 01:22
phenny colomon: 30 Nov 15:18Z <masak> tell colomon that it may be a good idea to write an 0th blog post today, which introduces the whole Perl 6 Advent Calendar endeavour to the world.
colomon: 30 Nov 15:59Z <Wolfman2000> tell colomon username for wordpress is wolfman2000. I don't know if he got it previously due to the evil PEER
colomon apparently masak and I were thinking along the same lines... :) 01:23
01:26 agentzh joined
jnthn "Who is peer and why won't he stop resetting my connection?" 01:26
colomon Wow, cygwin takes forever to download their git.... 01:27
01:28 justatheory left 01:30 justatheory joined
jnthn oh noes, not cygwin. 01:31
sjohnson heh
i'm not a huge cygwin fan either
i think it is very clunky
colomon I've long considered it essential for working on Windows boxes. 01:38
But it sucks on Vista, which was part of my reason for getting my first Mac this time last year...
SirKay well there's your problem 01:39
you were using vista
colomon SirKay: Unfortunately, as a programmer who has to support Windows machines professionally, I cannot ignore it forever. 01:40
SirKay oh my.
01:40 zaphar_ps joined
jnthn *just* managed to avoid Vista. 01:40
colomon But I couldn't bear the thought of my main laptop being a Vista machine, and erasing it and installing XP seemed a little past its sell-by date. So I got a Mac instead. 01:41
SirKay meanwhile, I've been talking to a more experienced programmer, and he suggests that as a beginner, I should do Python instead.
colomon jnthn: I just tried to build rakudo using the standard instructions, and the --gen-parrot phase blew up on me.
jnthn colomon: On cygwin?
colomon As far as I know, I just used cygwin for git.
jnthn oh. 01:42
arnsholt SirKay: Perl vs. Python is largely a matter of preference
jnthn The msys git port works wonderfully for me. ;-)
And doesn't need cygwin.
colomon That system's Perl is active state.
jnthn colomon: How did she blow?
colomon: Yeah, that's what I have too...
arnsholt Some people prefer Python, others prefer Perl (I think you can guess where this channel stands =)
SirKay but hypothetically, if I did Python, what would I have for support? Perl has this channel, and it has perlmonks, and whatnot.
jnthn The Perl community is one of the things that keeps me going, that's for sure. :-) 01:43
SirKay my Perl-Fu is weak but not nonexistent. On the other hand, my Python-Fu is nonexistent.
arnsholt TBH, I don't know. Being a Perl guy, I've not paid that much attention to Python
lisppaste3 colomon pasted "How she blew" at paste.lisp.org/display/91315
jnthn SirKay: In many ways, Python and Perl are philosophically different. I can understand how they fit easier for different people. 01:44
arnsholt As jnthn points out, the main differences between Perl and Python are the philosophy behind it, and the syntax 01:45
SirKay I'm simply concerned with no longer sucking at programming and starting to make nontrivial programs.
jnthn SirKay: If you like what you've seen of Perl so far and it seems to fit for you so far, keep exploring. It's not always an easy language to learn, but I'm very glad I made the effort.
SirKay I could go either way, to be honest. 01:46
jnthn SirKay: Sure. And you should see which fits best for you and go for it. TBH, once you learn enough langauges, they all start looking the same anyway. Apart from Haskell. That always looks insane. :-) 01:48
arnsholt Lisp is kinda funky as well =)
jnthn (And when I say "looking the same" I really mean, you start seeing that "oh, that's just a lambda" and "that's just a class declaration" and "that's just a loop"...) 01:49
Well, I do anyway. :-)
SirKay I've been starting and stopping for 16 years.
jnthn But maybe the compiler writer in me just wants to deconstruct all languages into their primitives. ;-)
arnsholt jnthn: I'm not a compiler writer (not much of one, anyways =) and I have the same experience
Except, of course, for Haskell which melts my brain =) 01:50
jnthn arnsholt: lol
arnsholt: I did sorta figure it was likely a fairly shared experience.
:-)
SirKay: I guess I sorta started 16 years ago and got hooked somehow. 01:51
arnsholt Yeah. People who say Perl is hard can just sod off. If you wanna talk hard, let's talk Haskell =)
SirKay it's a shame, really.
if I had applied myself and started learning properly in all that time
I could be a computing demigod now
arnsholt (I really would like to learn Haskell though. The concept is cool)
SirKay Perl is not really any harder than anything else. In fact, I'm one of those freaks who loves that you denote scalars with a dollar sign. 01:52
jnthn SirKay: To be honest, the longer I do it, the more I realize it's good to have a mixture of things in my life.
SirKay so yeah...I'll do either perl or python, I just want whatever will give me the quickest route to programming power. 01:53
That is a vain and callow desire, but yeah.
jnthn SirKay: Well, I guess you mainly want to get there so you can Build Cool Stuff. :-)
Which sure isn't a vain and callow desire.
jnthn loves seeing cool tech in whatever language. 01:54
SirKay are you a troper?
jnthn ...troper?
I guess that means no. :-)
SirKay hmm, I guess not.
jnthn If I have to ask.
SirKay tvtropes.org
though the site is down now.
But alas. Yes, I want to get results now, since I've been on and off for 16 years with no progress, with only myself to blame. 01:55
jnthn SirKay: Heh. I just noticed...tvtropes.org (now up again) runs pmwiki! 02:05
SirKay orly
jnthn yarly
SirKay nowai
colomon Quick, gang, a cool one line p6 script? 02:06
SirKay but yeah...uh...
jnthn SirKay: ...you fit in hear pretty well, don't you? :-)
lolspeaking fluently already. :-)
SirKay if I could prevail upon you to share with me some cool secret that is the key to everything, if you know it, I'd like to know.
I try.
02:07 hercynium left
jnthn colomon: role Pint[::Beverage] { }; class Beer { method drink { say "glug" } }; my Pint of Beer @beers = Pint[Beer] xx 4; @beers>>.drink; 02:08
02:09 bryan[c1] joined
jnthn colomon: Parallel dispatch, parametric roles, classes, types, repetition operator... ;-) 02:09
colomon I should have specified "that runs on rakudo-master". :)
still, awesome code there!
bryan[c1] when i try to run a perl script in notepad off my desktop it says the file cant be found, why is that?
could it be my shebang line? 02:10
jnthn colomon: Er. I think that (or something very close to it) already does. ;-)
colomon Method 'drink' not found for invocant of class '' 02:11
jnthn colomon: Oh.
my Pint of Beer @beers = Pint[Beer].new xx 4; # forgot the .new call - needed to pun the role
bryan[c1]: Hi. Most likely you're interested in Perl 5 rather than Perl 6, so this isn't *quite* the place to ask. :-) On Windows, I dobut it's the shebang line.
Since those don't mean much to Windows in general.
bryan[c1] i got banned from #perl for asking a question and some dude started an argument 02:12
and of course, the n00b that just came in is always wrong
jnthn bryan[c1]: *sigh* Sorry to hear that. :-(
bryan[c1] but on any note
jnthn bryan[c1]: How are you trying to run the script?
colomon rakudo: role Pint[::Beverage] { }; class Beer { method drink { say "glug" } }; my Pint of Beer @beers = Pint[Beer].new xx 4; @beers>>.drink;
bryan[c1] cmd 02:13
im in the desktop directory and everything
p6eval rakudo 7ce13d: ( no output )
bryan[c1] activeperl installed
jnthn Just typing "perl thescript.pl"?
bryan[c1] yeah
jnthn Where thescript.pl is the name of your script?
Hmm
bryan[c1] weird right?
maybe i need to reinstall activeperl
jnthn Does just "perl -v" give you any output?
bryan[c1] yeah
sjohnson bryan[c1]: i would recommend Strawberry Perl over ActivePerl 02:14
jnthn bryan[c1]: Something like "This is perl v..."
OK.
Hmm. Does it say what file is not found? :-/
bryan[c1] yeah it worken jnthn
yeah
no such file or directory LOL 02:15
jnthn OK, so it's in your path alright.
dir myscript.pl shows it?
(where myscript.pl is file name of your script...)
Starting cmd doesn't actually dump you into your desktop directly by default, you'd need a "cd Desktop" first to do that...
bryan[c1] oh well i did that 02:16
im not THAT much of a n00b
jnthn :-)
bryan[c1] C:\Documents and Settings\Owner\Desktop>
jnthn Hey, me either, but we can all make mistakes now and then. :-)
OK, in that case...yes, it's...odd.
SirKay hmm. 02:17
bryan[c1] kk ill un install AP and install straw
jnthn k
bryan[c1] only answer i suppose
02:18 zamolxes joined 02:19 JimmyZ joined
SirKay jnthn 02:19
02:19 orafu left 02:20 orafu joined
jnthn SirKay: No, I don't really think there's some secret "key to everything". :-) 02:20
SirKay Okay. But perhaps something short of that, then. 02:21
jnthn Find topics and niches that you enjoy and pursue them. Accept that computing is too big to understand everything out there, so get a sense of the bigger picture and then specialize in what you like doing most. Expect to write code now and look back at it in a year and wonder what you were thinking - and realize that means you've learned and accept it as a good thing. Accept that some things are Just Hard and don't feel bad about finding them so. Look at w 02:25
SirKay Look at w
jnthn Look at what others have done to learn from it rather than re-inventing wheels, unless all the existing wheels for the task at hand are square-shaped, in which case make a round one. 02:26
SirKay Well.
I don't want to be one of those people who thinks they're a good programmer and just more or less cribs the entire CPAN.
bryan[c1] well its there for a reason 02:27
jnthn SirKay: Knowing when to re-use what's out there, and when to make your own thing, is a difficult - but IMHO important - skill.
bryan[c1] 2(jnthn2): i installed stawberry but its still not working
jnthn bryan[c1]: What's in your script?
bryan[c1] just the shebang and print hello world 02:28
well print "hello world";
SirKay I understand. I'm just saying a lot of people just write a little code, append some cpan libraries, and call it good
jnthn SirKay: And if it just solved their problem, why isn't it good? :-)
bryan[c1] then thats what your code is sirkay, you're not expected to 'reinvent the wheel', as he just said above.
SirKay it's good, but it doesn't make you good. 02:29
bryan[c1] its still YOUR code
SirKay if that makes sense.
bryan[c1] but finding the ins and outs and makes modules is advanced stuff
SirKay Forgive me for sounding elitist about this, but...one of the main reasons I've redoubled my resolve to learn is in response to people who think they're great programmers even though much of their shit is trivial.
jnthn SirKay: I dunno. I've seen plenty of bad reinventions of things that there are already good solutions out there. :-) 02:30
SirKay I'm sure that happens, but you know what I mean.
jnthn bryan[c1]: OK, I remain...puzzled.
SirKay People who claim to be "senior cadre of the internet" for no real reason, or "core linux developers" even when they don't have a single line of code in the kernel. 02:31
jnthn bryan[c1]: Maybe try and pull out the shebang line?
Just in case.
bryan[c1] k
the whole file is : print "hello";
on one line
jnthn Sure, that should be fine.
bryan[c1] nope 02:32
SirKay however, I was rightly admonished, by the person who recommended I learn Python instead of Perl, to not focus so negatively, so let's just say that right now my goal is to be able to make, at the very least, nontrivial programs. ideally I will make programs people actually use and are also nontrivial.
but nontrivial is good for a start.
jnthn OK, if there's no shebang line and that's all in the file, I guess it ain't finding the file you're trying to run. Notepad didn't by any chance follow it's annoying habbit of adding .txt to the filename when you saved it? 02:33
bryan[c1] no its test.pl
SirKay Am I making sense?
bryan[c1] sirkay, using modules isnt against any moral rules of programming
just use them when you need to, no one really cares.
jnthn SirKay: It's fine to want to make nontrivial programs. It's also fine to use existing libraries to help you make them. 02:34
bryan[c1] whats with the looking down on module use?
when you get deep into programming you're going to thank god for them
jnthn I sure do. 02:35
SirKay okay I have made myself misunderstood.
bryan[c1] DBI, LWP, any module with file handling
SirKay I am not saying "don't use modules" or "don't use cpan." that would be dumb.
I am saying "it would be nice if people actually understood what went into those modules and could recreate them on their own later on"
bryan[c1] you're wondering where the skill has gone since they exist 02:36
you can look at the source code for all of them
SirKay yes, I'm talking about the people who don't.
bryan[c1] well, they're thinking "all i need to learn is the syntax, i dont need to know how it works"
in that case, they're probably just using perl for a specific reason
or they are just a weird perl programmer
SirKay and that's why I don't really consider them programmers.
bryan[c1] then so be it 02:37
you're only entitled to your own opinion
SirKay I agree with jnthn that improperly reinventing the wheel is a waste of time at best
nbrown SirKay: I think that's fine if the scope of that person's interest is to only get past their problem, but you seem more interested in learning how to program and how things work 02:38
arnsholt ISO annoys me. They've got all these nice "open" standards. Except they cost a small fortune to get
SirKay now wait. bryan, what is the problem? I'm not talking about you or anyone in the channel here.
nbrown SirKay: in that case, the best way is to use what you need, write what you can and learn from both as best you can
SirKay indeed nbrown.
jnthn Nicely put. :-)
bryan[c1] with all do respect sirkay, you turned it into a discussion about programming into a spouting of your opinions
sorry that didnt make much sense 02:39
im kinda high
but that aside, i still stand by my point.
SirKay I could have said something snide, but I didn't.
jnthn SirKay: Seeking to understand stuff is fine, and a Good Thing. Please understand that people getting their job done by just using stuff is also fine. :-)
02:39 Exodist_ joined
bryan[c1] (SirKay) and that's why I don't really consider them programmers. 02:39
SirKay bryan, I admit that I sort of got on my soapbox there.
but there is a good reason for it. 02:40
bryan[c1] they dont consider YOU a programmer, you're whining about already created code
02:40 Exodist left
bryan[c1] try coding in c++ without the standard library 02:40
SirKay Hmm.
bryan[c1] have fun with that
quantumEd standard libraries suck
nbrown jnthn: dumb question, but where did the setting go in the ng branch? 02:41
I'm just looking at it for the first time and feel a little lost
SirKay Well man, as you say, you are entitled to your opinion.
bryan[c1] thats not my opinion
i dont have an opinion on you
heres what im saying 02:42
if you were to have this conversation about how modules are for n00bs with some experienced programmers, they would laugh at you behind your back as soon as you walked away
if they werent nice people at least
SirKay You misunderstood my point completely.
bryan[c1] no i didnt 02:43
im just trying to shorten it
SirKay My point is not, and would never be "modules are for noobs."
As you say, that would be like coding in C without libraries. 02:44
bryan[c1] i know that, it was a shortened version of your point
a generalized one
SirKay Except it isn't.
bryan[c1] you don't understand MY point, thats the problem
im not saying you're wrong
stop being so technical
i dont talk that way in chat rooms
jnthn ng_feed: It's in core now 02:45
SirKay shrugs.
bryan[c1] you think modules are for people that dont really care HOW it works, but just want to get a task done (i.e. not 'real' programmers, as you stated above)
jnthn gah
nbrown: it's in src/core/ now
02:45 Exodist_ left
SirKay And that is wrong. 02:45
bryan[c1] in more detail, people that dont look at the source of the module and understand whats really going on
02:46 Exodist joined
bryan[c1] you're saying that only mechanics should drive cars because they understand how it works 02:46
jnthn nbrown: Renamed it to fit a bit more with spec naming and STD...
bryan[c1] im not wrong, that was totally your demeanor this whole time, SirKay
nbrown jnthn: thanks, I felt a little lost poking around
SirKay Whatever. I'm dropping the argument. 02:47
bryan[c1] just use the modules
if you want to know how it works, look at the source code
jnthn needs to sleep...it's late
o/
02:52 JimmyZ left 02:55 Wolfman2000 joined
Wolfman2000 ...alright, 11 hours then. 02:55
phenny Wolfman2000: 30 Nov 17:07Z <colomon_> tell Wolfman2000 For some stupid reason, I need the e-mail address you used to subscribe to wordpress, rather than your wordpress id.
Wolfman2000 colomon: you there?
02:55 seano` joined
colomon yes. 02:55
just putting finishing touches on a "how to get rakudo" post. 02:56
Wolfman2000 So the rakudo post wll be the first gift then?
colomon yes.
Wolfman2000 Guess that means I'm getting bumped to Thursday at least
colomon Do you have a post ready to go as well?
Wolfman2000 colomon: just got online when not dealing with classes. 02:57
I was going to plan my stuff now actually.
colomon understood.
Wolfman2000 I'll be able to juggle gifts with my finals...they're all take home this semester
I know I earlier claimed the idea of multi subs with constraints...I'm wondering if Types should be introduced first before we go into multi subs and constraints. 02:58
There should be some logic and planning to the calendar.
colomon agreed that a general notion of easier to harder and building upon earlier posts would be a good thing. 02:59
Wolfman2000 ...now's a good of time as any to plan this.
If this is on the pugs repo, I think I can commit to it.
...man, wish masak was here. This way we can plan this better. 03:00
colomon our notes (what there are) are there.
Yeah, planning for more than a day or two will probably go better in the morning, when everyone else is awake again. 03:01
what is the e-mail address you used to create your wordpress account? just the id isn't good enough for something reason. 03:02
Wolfman2000 colomon: thought I /msg'ed you it
guess it didn't take
colomon it did.
I just completely failed to notice.
one sec.
bingo. you're on it now. 03:04
03:05 seano` left
Wolfman2000 colomon: got the email. Nice greet message. "If you don't care, just ignore this email. :)" 03:07
colomon :)
pugs_svn r29220 | jafelds++ | Add myself to the list of volunteers. 03:09
Wolfman2000 guess I'm definitely in now. Now we just have to figure out not only the schedule of when we speak...but the topics. 03:11
colomon: I'm guessing you've got the first day.
colomon yes. 03:12
it's all cued up (queued up, too), just hoping to have some of the rest of the gang proofread it. 03:13
I think you should be able to see the draft if you go to "Posts" in the dashboard. 03:14
Wolfman2000 ...we cna make drafts? wow
pugs_svn r29221 | jafelds++ | Add our gift to the schedule.
Wolfman2000 I just can't view yours. 03:15
colomon Really?
Wolfman2000 ...let me try something else
okay, maybe not. I'm not exactly used to wordpress that much here. 03:16
colomon I've never actually tried doing anything with wordpress before, so I don't quite know what its capabilities are.
Wolfman2000 I can view your published post. Don't think I can view your draft
colomon well hell.
okay, I'll proof it quickly and then post it. we can probably fix any mistakes before anyone else notices. :) 03:17
Wolfman2000 you're brave, that's for sure
anyway, I'm debating whether we should add a list of possible topics/gifts on the schedule file as well
colomon It's probably not a bad idea. Though I suspect masak doesn't have anything firm in mind for the days he picked (for example). 03:19
Wolfman2000 He can easily talk about Web.pm
colomon maybe should add a topic brainstorming file as well.
Well, day 1 is live. 03:22
zaslon lolperl6adventhazblogged! perl6advent++ 'Day 1: Getting Rakudo': perl6advent.wordpress.com/2009/12/0...ng-rakudo/
03:26 envi^home joined, KatrinaTheLamia left
pugs_svn r29222 | colomon++ | [perl6advent] Claim some dates, start a brainstorming file. 03:27
03:27 JimmyZ joined
Wolfman2000 afk 03:30
TimToady r/e 03:33
colomon r/e ? 03:35
JimmyZ ng: ~(Str.^methods).say;
p6eval ng f2ba53: succ ACCEPTS WHICH pred perl acotan roots lcfirst sec asec trim cotanh map cis log sech rand bytes sin sqrt asin cosh asinh acosech abs uc ceiling floor substr asech ord round does flip split match acosec acosh words cosech cotan atan2 reverse isa ucfirst comb tan atan cos acos fmt chop
..t…
JimmyZ ng: say ~(Str.^methods);
p6eval ng f2ba53: WHICH pred perl ACCEPTS succ asec trim chr acotanh floor sech asech rand round sqrt asin split exp cosh match acosh words can cotan atan2 lcfirst sec map cis log sin does flip p5chop lc abs acotan acosech chomp index ceiling cotanh comb atan ord acos chop tanh atanh cosec chars roots uc
..r…
JimmyZ rakudo: say ~(Str.^methods);
p6eval rakudo 7ce13d: encode succ WHICH ACCEPTS perl sprintf Scalar Complex Str pred polar ucfirst comb tan atan cos acos fmt first trans chop rindex tanh values subst elems chomp index keys chr acotanh pairs kv capitalize pick evalfile atanh cosec exp p5chop lc join chars sinh can acotan lcfirst roots 03:36
..reduce …
pugs_svn r29223 | colomon++ | [perl6advent] Another idea.
JimmyZ rakudo: ~(Str.^methods).say;
p6eval rakudo 7ce13d:
..sprintfScalarComplexStrpredsuccencodeperlACCEPTSWHICHelemschompindexkeyschracotanhpairskvcapitalizepickevalfileatanhcosecexpp5choplcjoincharssinhcanacotanlcfirstrootsreducetrimaseccotanhmap:dcis:e:flogsechminrandmaxtruncatebytessortsinsqrtasinIntcoshgrepasinhacosechabsenducceilingp5chomps…
quantumEd coool!
JimmyZ ng++ 03:38
03:42 KatrinaTheLamia joined 03:43 ShaneC left
diakopter rakudo: Str.^methods.sort.say 03:45
p6eval rakudo 7ce13d: Multiple Dispatch: No suitable candidate found for 'cmp', with signature 'PP->I'␤in Main (file <unknown>, line <unknown>)␤
diakopter rakudo: Str.^methods[0].PARROT.say 03:46
p6eval rakudo 7ce13d: Sub␤
03:52 envi^office joined
diakopter rakudo: Str.^methods.grep({ ~$^a eq 'capitalize' })[0]("hihi").say # okay, so that's like JavaScript's Function.prototype.call 03:54
p6eval rakudo 7ce13d: Hihi␤
nbrown rakudo: class A {}; role A {}; say "alive"; 03:55
p6eval rakudo 7ce13d: ( no output )
diakopter nbrown: unfortunately we can't figure out why the evalbot spins on class definitions (and others)
nbrown diakopter: yeah I thought that was true, but it was worth a shot 03:56
std: class A {}; role A {}; 03:57
p6eval std 29223: ok 00:00 105m␤
nbrown diakopter: I was just wondering if it was valid to name a class and a role the same thing
JimmyZ rakudo: Str.^methods.grep({ ~$^a eq 'foo' })[0]("hihi").say 03:59
p6eval rakudo 7ce13d: invoke() not implemented in class 'Undef'␤in Main (file <unknown>, line <unknown>)␤
JimmyZ ng: Str.^methods.grep({ ~$^a eq 'foo' })[0]("hihi").sayng:
p6eval ng f2ba53: Confused at line 1, near "Str.^metho"␤current instr.: 'perl6;HLL;Grammar;panic' pc 519 (src/stage0/HLL-s0.pir:336)␤
JimmyZ ng: Str.^methods.grep({ ~$^a eq 'foo' })[0]("hihi").say:
p6eval ng f2ba53: Confused at line 1, near "Str.^metho"␤current instr.: 'perl6;HLL;Grammar;panic' pc 519 (src/stage0/HLL-s0.pir:336)␤
diakopter nbrown: that's a good question 04:04
perigrin nbrown: I'm guessing here but I can't see how that would be valid given that in perl6 Roles and Classes are interchangeable-ish 04:05
it fails in Moose because a Role and a Class are each packages and you can't have two packages with the same name. 04:06
diakopter perigrin: I agree... 04:07
in fact, I suspect Role has a proper superset of Class's semantics 04:08
nbrown perigrin: I agree too, I was just trying to understand the IO portion of S32
diakopter (implying Class might be redundant)
perigrin diakopter: not exactly
nbrown and it has and IO class and role
I was trying to determine if that made sense or needed work 04:09
perigrin diakopter: I take that back ... in perl6 you may be right, I'm not entirely sure how the semantics there play out.
nbrown diakopter: I'm little confused, how might it be redundant? 04:10
perigrin nbrown: if you view a class as an set of Methods and Attributes 04:11
and a Role as a Set of Methods and Attributes
diakopter if any Class declaration means the same as a Role declaration with exactly the same content, then it's redundant
perigrin hmm yeah diakopter just said it much shorter 04:12
diakopter I don't know tho 04:13
nbrown diakopter: ok..... I guess I was a little confused by the superset comment. I don't really see the difference between the two except how you apply them
04:13 jeremiah left 04:14 jeremiah joined
diakopter ng: die (*) 04:14
p6eval ng f2ba53: too few positional arguments: 3 passed, 4 (or more) expected␤current instr.: 'perl6;Code;new' pc 9748 (src/builtins/List.pir:144)␤
04:14 justatheory left
diakopter perl6;Code;new ? 04:14
rakudo: ::die (4) 04:17
p6eval rakudo 7ce13d: Confused at line 2, near "(4)"␤in Main (file <unknown>, line <unknown>)␤
diakopter rakudo: ::die(4)
p6eval rakudo 7ce13d: 4␤in Main (file /tmp/E9vXTMcdLi, line 0)␤
diakopter ng: ::die(4)
p6eval ng f2ba53: Null PMC access in invoke()␤current instr.: '_block14' pc 29 (EVAL_1:0)␤ 04:18
04:20 alester joined, astrojp joined 04:23 astrojp left
diakopter ng: my Int $a = 78.8; say $a.WHAT 04:24
p6eval ng f2ba53: Rat()␤
04:25 beggars joined
Wolfman2000 I'll be back on...within an hour. even if I idle. 04:32
04:32 Wolfman2000 left
diakopter TimToady: S05:2291 04:33
should that list include something like "a reference to [a rule that causes a reference to] itself" 04:34
04:36 beggars left 04:37 xenoterracide joined
diakopter pmichaud: any idea? (does ltm include only regular languages) 04:37
04:51 quantumEd left 05:04 Wolfman2000 joined
Wolfman2000 evening/morning 05:10
diakopter , the first day 05:12
06:03 beggars joined 06:28 PZt left
diakopter [OT] blog.sigfpe.com/2009/11/programming...tions.html 06:30
06:36 stephenlb left
TimToady diakopter: S05:2258 already disallows recursion on LTM 06:52
07:08 JimmyZ left 07:09 rgrau` left 07:12 justatheory joined 07:16 PZt joined 07:20 rfordinal joined 07:25 kaare joined 07:26 kaare is now known as Guest87125, alester left 07:28 pnate joined 07:31 justatheory left
moritz_ good morning 07:34
colomon morning. 07:41
07:41 rlarson`a left, rfordinal left 07:42 rlarson8` joined
dukeleto moritz_, colomon: mornin' 07:46
colomon o/ 07:47
think I've got the boy asleep, back to bed for me... 07:49
moritz_ good night
07:49 envi^home left 07:59 colomon left 08:11 iblechbot joined
mathw Good morning 08:12
08:22 broquaint left
pugs_svn r29224 | moritz++ | [advent] volunteer for 4th Dec; brainstorming 08:25
08:27 dakkar joined 08:53 JimmyZ joined 09:01 meppl joined 09:05 colomon joined 09:26 pdcawley joined 09:27 xinming left 09:30 colomon_ joined, colomon left, colomon_ is now known as colomon, mariuz joined 09:34 xinming_ joined
sjohnson rakudo: my $x = "\t"; say split("\t", $x).perl; 09:40
p6eval rakudo 7ce13d: ["", ""]␤
mathw hi sjohnson
sjohnson y0 09:41
mathw: do you know if you can somehow return the same result in p5? it will return nothing
as opposed to two empty strings
moritz_ sjohnson: yes, write it with comb 09:42
sjohnson ... what is that *hides under pillow*
moritz_ $x.comb(/\T+/)
mathw comb is much like split, but you say what you want to keep
moritz_ comb is searching for what you want, split is search for what you don't want
sjohnson but in p5 though? 09:43
mathw rakudo: my $x = "\t"; $x.comb(/\T+/).perl.say;
p6eval rakudo 7ce13d: []␤
09:45 colomon_ joined, colomon left, SirKay left, colomon_ is now known as colomon
moritz_ in p5: grep length, split /\t/, $x; 09:47
sjohnson i tried Dumper on that 09:48
doesn't print anything...
*scratches head*
moritz_ oh wait 09:49
I misunderstood
print Dumper [split(/\t/, $x), $x =~ /\t\z/ ? '' : ()] 09:50
or so
sjohnson thanks 09:51
my brother came into my room and brought this to my attention
mathw prefers the Perl 6 way
sjohnson i guessed correctly that it would return nothing, but he thought it was better if it did it the perl6 way
mathw I find using split and join and so forth as method calls leads to much more readable code
sjohnson i think i sold him on p6 when i ran rakudo with him in my room
mathw yay!
sjohnson++ 09:52
sjohnson thanks!
he ran into this problem with Java, where it was doing only half of what p6 was doing 09:53
something along the lines of only returning the left empty string
moritz_ mathw: agreed
mathw Ah Java
some of its libraries do strange things
sjohnson yeah this do.this.that.thenthis is my kind of syntax 09:54
another reason why i thought ruby would be great, but then perl 6 came knocking at my door
moritz_ I'm sure I'll love the feed operators once they are (partially) implemented 09:55
writing multiple maps/greps in a statement feels always backwards
but with feeds you can write them in forward
sjohnson i need to lose my perl virginity, and start using map more often in my code 09:56
it's one of those things i know is very cool and powerful, but i'm afraid to use as i keep forgetting the syntax
moritz_ @list ==> map { 2*$x + 1 } ==> grep { $x !% 3 } ==> my @result
(assuming I understood how feeds are supposed to work) 09:57
mathw sjohnson: I take it you never did functional programming then? That's how I learned to use map.
09:57 rfordinal joined
sjohnson mathw: i've used it about 3 times, i kind of get the idea, but maybe my uses are a bit novice-level. i use it to do quick changes to every element of a list 09:57
mathw That's a perfectly acceptable use of it
sjohnson but there are probably uses that are beyond my wildest dreams 09:58
mathw It's so ingrained in my way of thinking that I get angry with C++ for not having anonymous functions so I can easily map one over a list
(bring on the next standard, which does have lambdas)
sjohnson i probably haven't done any funtional programming. i don't fully grasp even what the term means 10:00
i have heard it used to describe things like ocaml and haskell though
10:01 snearch_ joined
sjohnson perl is pretty cool. i use it a lot at work for my own use 10:02
unfortunately not much new stuff / excuses come up to use it at work, so i try to make time for it 10:03
i use it for all sorts of great things, like alternate vim frontends so that i quit editing 2 files at once in Vim by accident, or a handler for ls, so that if i type "ls l-" it knows what i want to do 10:06
or ls- l
10:16 envi^office left 10:19 riffraff joined
mathw Perl is very cool indeed 10:21
Functional programming would refer to Haskell and OCaml, since they're functional programming languages
Haskell being just about the purest one going, OCaml is a bit more of a mixture
sjohnson but does that mean that.. all you do is call functions?
mathw Perl supports many functional concepts
sjohnson i am not an expert with perl's OO syntax
mathw well in its purest form, it means that any function's result is entirely determined by its arguments 10:22
10:22 JimmyZ_ joined
mathw and before you say well of course 10:22
think about functions which look at other sources of state
frettled or take user input
mathw yes
unfortunately that does mean that in pure functional programming, user input and network or file I/O are difficult to fit into the overall structure 10:23
you have to cheat
frettled It's when you accept input that nearly all bets are off, muahaha.
(input from external sources, that is)
mathw But the advantage is that pure functional code is very easy to reason about
10:23 riffraff left
mathw Proof of correctness is a lot easier 10:23
frettled It is, that's why provers are often written in functional languages.
mathw Absolutely
Language researchers are very into it 10:24
Well, some of them - some of them are presumably researching other paradigms
frettled One of my favourite difficult classes from the university was called Verifiable Programming. 10:25
10:25 riffraff joined
mathw lots of preconditions, postconditions, invariants, bound functions and proofs? 10:25
frettled No, wait, that was the name of the book. The course was called Program Specification and Verification. 10:26
mathw I did one called Program Construction and Verification
which was probably very similar
frettled The first part of the course deals with how to specify your programs in order to avoid errors, the second part uses Hoare logic etc. for verification and proving.
10:27 broquaint joined
mathw Yes, it's basically the same course I did then 10:29
I really enjoyed it
frettled Yup. 10:30
The lecturer was Ole-Johan Dahl, quite a character.
mathw Our lecturer was writing his own textbook on the subject at the time
He used it as the course text, so we got handed out printouts of the late drafts of it
So I've got his entire textbook in almost-final form, for free
frettled :)
mathw Although have considered buying the actual book a few times 10:31
sjohnson mathw: do you have any preference of C over C++ or vice-versa?
mathw But it's him I got my thing for correct-by-construction programming from, even if I don't get a chance to exercise that very often 10:32
sjohnson: I prefer C++
10:34 colomon_ joined, colomon left, colomon_ is now known as colomon, JimmyZ_ left
sjohnson mathw: may i ask a couple questions? i was thinking of writing maybe an application someday, but i can never decide which one to try and learn 10:35
C or C++
mathw Sure
sjohnson though, 99% of the things i need coded can be best coded in Perl, so it's not a huge worry, but i am still curious
mathw I do C++ for a living, so I know a bit of it :) 10:36
sjohnson i do javascript and PHP and some perl for a living, but i'm probably not the best at it :)
mathw I'm definitely not the best at it, but there's always someone better 10:37
So ask away :)
jnthn o/ 10:38
mathw oh hai jnthn
By the way, if we're obstructing Perl talk with our other-languages talk, we'll take it somewhere else... 10:39
10:40 JimmyZ left
jnthn I was gonna say we'd happily steal things, but now I see PHP and C++ I'm not so sure that applies... ;-) 10:41
frettled hehe 10:44
mathw lol 10:52
no you probably don't want to
bryan[c1] 2(jnthn2): why doesnt me perl work still? 10:53
jnthn bryan[c1]: Well, because you didn't figure out why it won't...I kinda already gave you all my suggestions last night though. 10:58
bryan[c1] last night?
im still in the same day hahahah
seriously though, i uninstalled activeperl and installed strawberry perl 10:59
i dont understand how it could just sit there and say it doesnt exist
11:06 rfordinal left 11:20 mariuz left 11:21 agentzh left 11:47 SmokeMachine joined 12:06 envi^home joined 12:12 ruoso joined 12:14 bluescreen joined 12:16 rfordinal joined 12:25 bluescreen left, bluescreen joined 12:28 bluescreen left, bluescreen joined 12:32 iblechbot left 12:50 payload left 12:58 SmokeMachine left 13:00 patspam joined, patspam left 13:03 allbery_b joined 13:06 colomon_ joined, colomon left, colomon_ is now known as colomon, dakkar left 13:09 bryan[c1] left, SmokeMachine joined 13:11 takadonet joined
takadonet hey all 13:16
moritz_ wow, quiet in here today
hi takadonet
takadonet moritz_: how are you doing?
moritz_ stressed 13:17
takadonet moritz_: work? 13:18
mathw Hey
moritz_ kinda
takadonet: what about you? 13:20
13:22 snearch_ left
takadonet moritz_: Busy at work working on my large Perl 5 project (485 modules) while helping one of our co-op student speed up his Perl 5 script which working on a data set of a few million records 13:22
moritz_ takadonet: what's the project about? 13:23
some kind of data analysis?
takadonet moritz+: Yes, for bioinformatics. I'm been heavily modifying this application called Gendb: www.cebitec.uni-bielefeld.de/groups...endb_info/ 13:24
moritz_ sounds pretty interesting 13:26
sadly bielefeld is part of a huge consiparcy, and I can't take anything seriously that comes out of bielefeld :-) 13:27
takadonet moritz_ : I really love it . This going to be my project for the next few years
I heard that
but really have no idea what's it really about 13:28
<--- Canadian
moritz_ en.wikipedia.org/wiki/Bielefeld_Conspiracy
13:30 colomon left
takadonet wow hehe 13:31
13:32 tylerni7 joined 13:34 iblechbot joined 13:35 colomon_ joined
moritz_ wonders if pmichaud++ is doing a rakudo day or so soon 13:37
13:44 colomon_ left 13:45 payload joined
moritz_ std: undef 13:46
p6eval std 29224: ===SORRY!===␤Obsolete use of undef as a value; instead, in Perl 6 please use the most appropriate of:␤ Mu (the "most undefined" type object),␤ a more specific undefined type object such as Int,␤ Nil as an empty list,␤ *.notdef as a matcher or method,␤
..Any:U as a type con…
pugs_svn r29225 | masak++ | [misc/perl6advent-2009/schedule] nailed down a few topics 14:04
14:06 beggars left 14:08 colomon joined 14:12 mberends joined
colomon moritz_: I hope pmichaud++ is doing a rakudo day today. I think every day I'm the only one doing ng commits is probably two days later it gets.... ;) 14:12
mberends moritz_: Bielefeld: klasse :-) 14:13
Wolfman2000 *yawn* morning
mberends morning!
14:16 perlygatekeeper joined
dbrock` colomon: haha 14:20
moritz_ colomon: I hope you're not serious; so far your ng commits looked pretty good 14:21
14:23 iblechbot_ joined 14:24 iblechbot left, xinming_ left
pmichaud I'm surrfering from ISP fail at the moment :0| 14:24
*suffering
frettled Pain! Suffering! Torment!
pmichaud ISP currently giving about 30% packet loss
frettled That sounds like a good name for a TECO parser for Perl 6. 14:25
moritz_ ISP--
14:25 frettled sets mode: +oo colomon mberends
mberends It was off topic, but hopefully bryan[c1] of irclog.perlgeek.de/perl6/2009-12-01#i_1785868 got his Strawberry Perl to work. OTOH, something must have gone badly wrong between asking a simple question and getting banned from #perl. His argument with SirKay was something both of them should have been mature enough to avoid. It goes against the #perl6 spirit of friendliness and wasted a lot of our mindshare. Next time bryan[c1] shows up, let's giv 14:26
e him lots of hugs.
moritz_ ENOHUGME 14:27
mberends awww 14:28
14:28 hugme joined
moritz_ it seems hugme doesn't recover from netsplits :( 14:28
14:28 moritz_ sets mode: +v huf
mberends hugme: hug moritz_ 14:28
hugme hugs moritz_
14:28 moritz_ sets mode: -v huf, moritz_ sets mode: +v hugme
moritz_ sorry huf, tab fail :-) 14:29
huf ;) OH NO
14:35 pointme left, zaslon left 14:36 pointme joined, zaslon joined
frettled mberends: yikes, I'm glad I slept through that. 14:39
hugme: hug me
hugme hugs frettled
frettled smiles again.
mberends frettled: had to read it in the backlog. addicted to backlogging ;)
14:41 rodi joined
frettled mberends: I'm kindof addicted to backlogging, but I'm slowly getting over it. :D 14:41
So, does anybody else here want to join in implementing PST? 14:42
moritz_ PST?
frettled Pain, Suffering and Torment - a TECO parser in Perl 6 :D
14:43 xinming joined
moritz_ you mean parsing things like this: en.wikipedia.org/wiki/Text_Editor_a...e_programs ? 14:44
frettled yep
I usually use TECO as a «STFU»-thing for people who think Perl is like line noise. 14:45
14:48 rfordinal left
frettled On a more serious note, I think it might be an opportunity for learning some grammar and parsing skills. There is this thing about time, though, *Sigh*. 14:48
14:48 Guest87125 left
moritz_ well, my TODO list for Perl 6 is already quite long 14:49
pushing it to the end will do it no good
arnsholt frettled: The TECO parser should have three main components, named Pain, Suffering and Torment ;p 14:50
14:51 quantumEd joined
moritz_ (among them are: implement some builtins in ng; write more for the book; blog; write for the Perl 6 advent calendar; write a synopsis on testing; hack a bit on dev.perl.org/perl6/; convince TheDamian to finally put the current version of this S26 draft into the pugs repo; ...) 14:51
frettled arnsholt: :D
14:51 rgrau_ joined
frettled moritz_: Yes, that ought to suffice for April … 14:51
moritz_ continued: help proto; investigate more about putting Perl 6 modules on CPAN. 14:52
frettled: if I got a job as full-time Perl 6 hacker, I'd have a chance of getting 3/4 or so done until April :-) 14:53
frettled moritz_: the last thingy can be solved through forgiveness ;) 14:55
moritz_: eh, the last thing from the first list, that is.
frettled pats itself.
mathw hmm the perl 6 advent calendar 14:57
frettled yes, I should check that one out 14:58
mathw subscribes to it 15:03
Are authors required?
moritz_ yes 15:04
see svn.pugscode.org/pugs/misc/perl6advent-2009/
mathw excellent 15:06
pugs_svn r29226 | jafelds++ | Add other ideas for topics.
mathw I shall have to think of a topic or two I can write about usefully 15:07
and grab some days
and sort out anything else that needs sorting out
frettled too :D 15:08
moritz_ mathw: you don't even need to come up with ideas on your own, you can use some from the brainstorming list
mathw yes, but then there's the matter of precisely how much scope to use etc.
something about Perl 6's OOP is probably a good starting point
or maybe the joy of for loops
when combined with pointy blocks 15:09
and the Z operator
and prefix:<^>
frettled . o O ( against loops? Could we have that? )
Wolfman2000 rakudo: say (1..6).pick(2)
moritz_ zip is another good idea
p6eval rakudo 7ce13d: 61␤
Wolfman2000 rakudo: say (1..6).pick(2).join(" ")
p6eval rakudo 7ce13d: 5 4␤
Wolfman2000 ...I forgot. Does pick allow picking the same value twice?
15:09 colomon left
Wolfman2000 rakudo: say (1..6).pick(2).join(" ") 15:09
p6eval rakudo 7ce13d: 4 2␤
Wolfman2000 rakudo: say (1..6).pick(2).join(" ")
p6eval rakudo 7ce13d: 1 2␤ 15:10
mathw okay
frettled rakudo: say (1..2).pick(4).join(" ")
15:10 zloyrusskiy joined
p6eval rakudo 7ce13d: 2 1␤ 15:10
mathw I should sign myself up
shame my pugscode SVN details are at home
Wolfman2000 rakudo: say (1..2).pick(4, true).join(" ")
p6eval rakudo 7ce13d: Confused at line 2, near "(4, true)."␤in Main (file <unknown>, line <unknown>)␤
Wolfman2000 rakudo: say (1..2).pick(4, Bool::True).join(" ")
p6eval rakudo 7ce13d: No candidates found to invoke␤in Main (file <unknown>, line <unknown>)␤
frettled The rakudobot is annoyingly quick these days ;) 15:11
Wolfman2000 ...could have thought there was an option.
moritz_ rakudo: say (1..2).pick(4, :replace).perl
15:11 zloyrusskiy left
p6eval rakudo 7ce13d: [1, 2, 2, 1]␤ 15:11
Wolfman2000 ...I thought adverbs weren't implemented in rakudo yet
moritz_ that's not an adverb to an operator
15:11 quantumEd left
moritz_ it's just a named argument 15:12
pugs_svn r29227 | jafelds++ | Plan to discuss this topic.
Wolfman2000 ...oops. wonder why it didn't show that I uploaded schedule
Wolfman2000 doesn't understand pugs_svn
moritz_ pugs_svn: just reports revision number, author and log message 15:13
Wolfman2000 I could have thought it also reported files on occasion
moritz_ never
but people generally put a small moniker in square brackets into the commit message
to indicate the subsytem that was changed 15:14
Wolfman2000 ...yeah, that
moritz_ [advent] or [t/spec] or so
rgrau_ moritz_: how about announcing the perl6 advent on twitter account? (individually or every day) 15:24
frettled or on perl6.org :D
mathw and everywhere 15:26
everywhere!!
Wolfman2000 you guys are more fanatic about Perl 6 than I realized 15:27
frettled Huh? 15:29
15:29 colomon_ joined
Wolfman2000 ...don't worry about t 15:30
about it*
Wolfman2000 starts some of his final assignments.
mathw I have pinned my entire future sanity on Perl 6 15:31
moritz_ hugme: tweet rakudoperl The Perl 6 advent calendar: perl6advent.wordpress.com/
hugme hugs moritz_; tweet delivered
moritz_ hugme: tweet rakudoperl perl6 advent calendar day 1: get and install rakudo perl6advent.wordpress.com/2009/12/0...ng-rakudo/ 15:32
hugme hugs moritz_; tweet delivered
moritz_ rgrau_++ # good idea 15:33
15:33 alester joined 15:34 colomon joined
ng_feed rakudo-ng: (Jonathan Scott Duff)++ 15:35
rakudo-ng: Add while/until statement modifiers
15:36 colomon_ left
jnthn back 15:36
15:37 payload left
jeremiah Whoah, that perl 6 advent calendar is a great idea. :) 15:39
frettled Oooh, hugme-tweeting, that's neat. 15:41
Tene pmichaud: I bet you know what I'm going to ask about. ;) 15:43
pmichaud Tene: indeed, I do 15:44
I worked on it some last night but then ISP connection failed (and I was getting tired)
mathw I assume hugme only pays attention to certain people
o/ jnthn
pmichaud and this morning I'm trying to find a 50% slowdown in Parrot
(an as-yet-unexplained 50% execution cost, that is) 15:45
mathw ouch
jnthn pmichaud: huh?
mathw good luck
jnthn pmichaud: Parrot good 50% slower?
*got
lol
pmichaud jnthn: no
see nopaste.snit.ch/18908
15:45 masak joined
masak oh hai #perl6 15:45
jnthn lolitsmasak
masak omgitsjnthn 15:46
mathw \o/ masak
15:46 alexn_org joined
masak I'm thinking I might write the Dec 2 advent post about .fmt today. that way, I don't have to write it tomorrow, just post it. :) 15:46
jnthn pmichaud: huh...wtf.
mathw masak: good plan 15:47
pmichaud jnthn: that's exactly what I said :)
mathw I'm already thinking about my advent calendar post
I may even do two
or three
pmichaud did I get signed up for any dates this week?
mathw I haven't signed up yet
masak mathw: better grab the days before they're all occupied! :)
mathw I need my home laptop for my pugscode SVN access
moritz_ pmichaud: don't think I've seen you in the list
mathw I'd like December the 6th though 15:48
pmichaud I'll take Dec 5
masak pmichaud: do you have a pugs commitbit? :)
pmichaud masak: yes, but I have a very slow ISP connection today
I suspect it'll take an hour to get an up-to-date copy of the pugs repo 15:49
jnthn pmichaud: It's not in any way just that somehow we keep a reference to the parse tree around?
masak pmichaud: I'll add you, then.
moritz_ pmichaud: I'll volunteer you
jnthn pmichaud: And then GC keeps marking it and so kills us?
pmichaud jnthn: and then GC cost? might be
jnthn: I'm working on tracking it down now.... tht's going to be my rakudo day for vienna.pm grant
mathw Could somebody updating the schedule add me for the 6th as well please?
jnthn :-)
mathw I'll porobably do at least one other day as well, but I need to think about scheduling 15:50
pugs_svn r29228 | moritz++ | [advent] reorganize brainstorm list; volunteer pmichaud++ for Dec 5
moritz_ mathw: will do
masak dang, moritz_ beat me to it. :)
pmichaud moritz_++ # thanks 15:51
moritz_ nearly-free karma
pmichaud masak++ # just because
pugs_svn r29229 | moritz++ | [advent] mathw++ for Dec. 6
mathw hugme: hug moritz_++
hugme hugs moritz_++
masak in theory, I have made many excellent commits. :)
jnthn Ooh, I should volunteer myself too.
Not for this week though.
masak jnthn: ...before it's too late! :)
mathw I can see I'm going to have to thinka bout scheduling quickly
but for now, I have to think about getting my car fixed 15:52
mathw -> gone
jnthn masak: I only have checkouts of parts of the Pugs repo.
:-/
15:52 tylerni7 left
jnthn masak: Where is the bit I need? 15:52
masak misc/perl6advent-2009/schedule 15:53
pmichaud jnthn: misc/perl6-advent-2009
jnthn thanks
pmichaud what masak++ said
jnthn grabs that
Tene jnthn: if you want to see if it's GC, try running with GC disabled.
jnthn pmichaud: ^^
;-)
masak anyone want to post a comment to oylenshpeegul.vox.com/library/post/...=feed-atom about perl6advent?
pugs_svn r29230 | duff++ | [advent] volunteer for a couple of days
r29231 | duff++ | [advent] add a potential topic
PerlJam masak: I did, but it apparently didn't go through because of some problem setting up my vox account. 15:54
(good $localtime btw)
masak PerlJam: ah.
PerlJam++ anyway.
15:56 amackera joined
PerlJam doh! The vox email confirmation was flagged as spam. (it's been so long since this has bitten me that I totally forgot to even check for it until now) 15:58
Tene pmichaud: very well understood. I fell asleep as soon as I got home last night, and slept mostly through the night.
masak PerlJam: the curse of false positives.
15:58 tylerni7 joined
pmichaud Tene: yes -- at least I got a very good night's sleep... best one in a while 15:58
15:58 perlygatekeeper left 16:00 Psyche^ joined
Tene Same here. :) 16:01
pugs_svn r29232 | jnthn++ | [advent] Volunteer for two days. 16:02
r29233 | jnthn++ | [advent] You maded me a topic but I scheduled it.
pmichaud nopaste.snit.ch/18909
jnthn pmichaud: Hmm. That doesn't seem to account for all of it then. 16:03
pmichaud: How much gets blown by IMCC?
jnthn isn't so sure why it'd be different there though... 16:04
diakopter who doth disturb my slumber [with a hilite]?
jnthn pmichaud: Check that nqp fake executable doesn't start with a different runcore.
pmichaud jnthn: that's why I did the run of nqp.pbc 16:05
(second run)
doesn't appear to be significant difference between nqp.pbc and nqp fakecutable
Tene Oh, I know! I could implement macros in ng, and then write an advent post on them! Nothing could go wrong with that plan! 16:06
PerlJam pmichaud: Would you volunteer for the Dec 24 slot? Somehow I think it fitting that yours be the last post :)
pmichaud PerlJam: yes, I'll volunteer for that slot
also, note that Dec 17 is the release date
(it's also my wife's birthday :)
PerlJam are you volunteering for that date as well?
pmichaud not yet :) 16:07
16:07 payload joined
Tene PerlJam: no, he's volunteering his wife. 16:08
pmichaud what's the pir opcode to disable gc?
PerlJam I'm not sure she'd agree to that :)
pugs_svn r29234 | duff++ | [advent] pmichaud takes Dec 24 16:09
16:09 quantumEd joined
[particle] hrmm, there's sweep 0, mark 0, and one more.... 16:10
pmichaud I see "sweepoff" and "collectoff" 16:11
16:12 Patterner left, Psyche^ is now known as Patterner, pmurias joined
[particle] sweepoff will disable marking 16:12
collectoff will disable sweeping, iirc 16:13
don't quote me on that, the names are funny
arnsholt That sounds intuitive =p
pmichaud =item B<sweepoff>() 16:14
Disable GC runs. (Nestable)
=item B<collectoff>()
Disable GC runs (nestable).
16:14 perlygatekeeper joined
pmichaud docs appear to be LTA 16:14
[particle] op sweepoff() {
Parrot_block_GC_mark(interp);
}
op collectoff() {
Parrot_block_GC_sweep(interp);
}
the docs are definitely LTA
pmichaud code is far more informative and concise than the docs here :-|
[particle] the op names are LTA as well 16:15
arnsholt Whut? Since when do I have +o? (Or has my IRC client lost it?)
pmichaud well, it looks to me like the problem is indeed GC, and it's far more serious than we had imagined.
[particle] don't ask why, just bask in the glow of +o
i don't understand the big slowdown between executing directly and separate compile/load/execute 16:16
arnsholt Heh
pmichaud [particle]: it's gc
[particle]: it's the cost of marking the parse, past, and post structures, I suspect 16:17
[particle] shouldn't they disappear after compilation?
pmichaud if there aren't any references to them, probably
I'm not sure there aren't any references
[particle] hrmm, what if you put a sweep 1; collect 1; after pir generation 16:18
unconditionally run gc after compilation finishes 16:19
pmichaud I suspect it makes no difference
because it's the existence of the tree that is expensive, and you incur the cost on every GC run
forcing a GC run doesn't avoid the cost later
[particle] but generating the pir is fast
pmichaud right 16:20
[particle] so, in the interpreter teardown after generating pir, all objects are unconditionally destroyed
pmichaud I'm not following where you're headed with this.
[particle] but that's not necessarially timed, because it can happen outside the gc, in global
destruction
hrmm, no, you're using 'time' 16:21
pmichaud if you're claiming that arranging for the parse/past/post trees to be reclaimed will help us, I think we're missing the point.
[particle] i'm saying, when you write pir to a file and load it back in, you've destroyed the parse/past/post trees
pmichaud [particle]: yes. 16:22
which means that the difference in execution is due to the existence of parse/past/post
[particle] but not necessarially so when you compile/execute in the same interp
16:22 pdcawley left
PerlJam [particle]: you think the destruction takes so much time? 16:22
[particle] and you can't mark them read-only
no, i'm thinking of ways that those trees can be skipped by gc 16:23
generational gc would help
pmichaud it's a little worse than this...
these are not big trees
PerlJam looks like the two of you have something to bring up in #ps today :) 16:24
pmichaud this program is *tiny*
it has basically two subs
[particle] yes, the program is tiny, but has many pcc calls
which means many contexts and arrays
pmichaud [particle]: you're still missing the point
the pcc calls aren't the speed bottleneck we're aiming at
the number of pcc calls is the same when run combined and when run as separate pir 16:25
[particle] yeah, you're right
16:25 nihiliad joined
pmichaud I'm saying that fib.nqp produces very small trees 16:25
[particle] my jaw pain is getting in the way of clear thought.
pmichaud very small parse/post/past trees
and if a program this small can produce trees that result in a 10 second slowdown
(merely by their existence) 16:26
then we have a serious problem for any larger programs, regardless of the number of pcc calls involved
pmichaud prepares to write this up for parrot-dev 16:27
oh, first I want to add my instrumentation
(will let us see the time used for each stage of compilation)
[particle] wish you could show the # pmcs created for each stage 16:28
pmichaud is there an easy way to get that number from pir? 16:31
pmurias TimToady: would it be possible to make viv attach the unmangled operator name to nodes resulting from parsing an infix node? 16:37
TimToady that's what SYM is 16:39
or do you mean including the infix:<>?
pmurias checking... 16:40
16:40 ascent_ joined
pmurias TimToady: they don't seem to have a SYM 16:41
TimToady if I do viv -e '1 + 1' I get a SYM of +
pmurias what i want is to get * from VAST::infix__S_Plus
TimToady you mean +? 16:42
or is the * making it generic? :)
the SYM: + is in the infixish node, not in the . node 16:43
pmurias i mean + sorry 16:44
what exactly is the . node?
masak loliwroteadraft: gist.github.com/246434 # Perl 6 advent blog post about .fmt 16:45
TimToady the equivalent of .caps in the match object
it the nodes in textual order as a list 16:46
for semantic purposes you generally want to use one of the named entries instead of indexing .
16:47 cdarroch joined
TimToady the . entry is more for translators and pretty-printers 16:48
pmurias the reason i couldn't find the SYM is that i pruned the '.' before dumping the tree to yaml
16:49 ascent left
TimToady I guess there ought to be a name for the op to go with left and right 16:50
pmurias go with left = ? 16:51
TimToady huh? 16:53
masak TimToady: I asked a few days ago whether the PairValSet generated by SomeEnum.mapping guarantees anything about the order, when doing .pairs, .keys, etc. maybe I missed your reply in the backlogs. 16:54
TimToady it looks like I need to hoist the infix: pointer from inside its own node to the infixish node
masak: in my copy it now says PairValSeq instead, but thinking about whether that makes sense in general
pmurias TimToady: what does "go with left and right" mean in this context? 16:55
masak TimToady: also, the other question: what do we gain from talking about PairValSet and PerlValSeq rather than Set[PairVal] and Seq[PairVal]?
TimToady pmurias: I'm talking about the keys into the infixish hash
masak: been thinking about that too :) 16:56
16:56 mberends left
masak TimToady: ok, good. :) 16:56
from my point of view, the parameterized-roles type system looks more credible if Perl 6 dogfoods it, rather than doing the antipattern and inventing $m * $n classes. :P 16:58
TimToady well, orthogonality is good, especially where dehuffmanization is fine, but we can also have shortcut names, and a few shortcuts don't go $m * $n 16:59
they just give you N degrees of separation with a small N
masak nod. 17:00
TimToady part of what I'm trying to balance there is the utility of being able to extract the original declaration order, versus
allowing the programmer to set it up that way if needed 17:01
my @mapping = enum <foo bar baz>; enum FBB = @mapping # speculative syntax
well, s/my/consant/ 17:02
t
masak that second assignment feels odd.
TimToady the arg the other way is that it gives better introspection of an enum you didn't set up
but it pretty much precludes just using the Stash as the source of it, since that's unordered 17:03
so we'd have to have the mapping be stored duplicately
pmurias TimToady: it might be better to always have an args attribute instead of sometimes having left and right 17:04
TimToady or not use a stash, and fake a lookup with Foo::bar is used
17:04 zloyrusskiy joined 17:05 zloyrusskiy left
masak TimToady: I'm looking at gimme5 and LazyMap.pm, taking breaks when my brain explodes. what do 'B', 'C' and 'L' stand for in the blessed LazyMap hash? 'block', something and something? 17:06
TimToady block, called, and lazy 17:09
masak thanks.
TimToady thing is, a given call of the block may produce more than one value, and iter only returns one 17:10
masak I see.
TimToady but calling the block again on the processed value would be wrong, so they're stored separately
17:11 justatheory joined
TimToady it's actually pretty awful to try to debug code using it :/ 17:11
colomon masak++ # had no idea about .fmt before this.
TimToady which was the other reason for avoiding lazymaps whenever possible; it wasn't just about efficiency
masak I also looked a bit at the STD.pm5 output. it doesn't help me yet, except that I note that it's full of lazymap calls.
TimToady only where something might backtrack 17:12
masak colomon: I'm surprised about how unknown .fmt is. :)
zaslon lolmasakhazblogged! masak++ 'Failure()<0xb4ff6b24>': Failure()<0xb4ff605c>
TimToady unfortunately, it tends to propagate upward to the top of the rule
masak zaslon: no, I haven't!
zaslon Sorry, I don't understand that command
TimToady I wonder if zaslon restarts itself periodically and gets confused when it does 17:14
masak maybe I should blog more about failures... :) 17:15
zaslon lolmasakhazblogged! masak++ 'November 30 2009 -- gobsmacked but still in the loop': use.perl.org/~masak/journal/39966?from=rss
masak zaslon: oh, come on! that was yesterday.
zaslon Sorry, I don't understand that command
masak zaslon: don't be sorry, be quiet! </space-balls> 17:16
zaslon Sorry, I don't understand that command
17:16 alexn_org left
colomon feels slightly guilty about encouraging masak to blog again less than 24 hours after finishing his marathon... 17:16
masak colomon: not to worry. it was my own choice. :) 17:17
it will be a few days before I summarize the November blogging, though.
pugs_svn r29235 | pmurias++ | [mildew] start updating to newest STD 17:19
r29236 | pmurias++ | [mildew] t/01-sanity/01-tap.t passes using the new STD
17:19 justatheory_ joined 17:20 justatheory left
masak I think I'll withdraw now to make some tea, and stare a bit more at the LazyMap implementation. :) 17:20
17:20 justatheory_ is now known as justatheory
masak see y'all around. o/ 17:20
TimToady pmurias: you do get args for a list associative operator, but it seems like ->{left} is faster than ->{args}[0]
17:20 masak left
pmurias faster to type? 17:21
TimToady: having args for left and right associative operators would allow the same code branch to handle them 17:22
TimToady: btw, VAST is now much sainer then when i made the STD snapshot
TimToady I blame diakopter++ 17:26
17:27 iblechbot_ left
TimToady that's not left and right as in associative, that's just the left arg and the right arg 17:27
the op may or may not be associative at all
but possibly I should rename args to "parcel" anyway, in which case it makes more sense to do binaries and unarys that way too 17:29
*unaries # who invented this stupid orthography, anyway 17:30
17:30 stephenlb joined
pmurias was orthography designed or were the most common ways to write something codified? 17:34
TimToady mostly the latter
but Samuel Johnson could have used a course in morphophonemics before doing his dictionary :) 17:38
and Webster's fixes were mostly cosmetic huffmanizations 17:39
maybe we should all just switch to chinese characters, and then we wouldn't have to worry about how it's pronounced to read it. :D 17:40
17:44 payload left
zaslon lolfrettledhazblogged! frettled++ 'Oslo.pm Past and Future': http: 17:46
jnthn failurl
frettled heh 17:47
17:47 brrant joined
sjn frettled uses http as a protocol for his blog! NICE! :D 17:47
frettled :D 17:48
zaslon lolfrettledhazblogged! frettled++ 'Oslo.pm Past and Future': http:
frettled eyes zaslon suspiciously.
hmm, how did I use those bots again. *checks earlier blogpost* 17:49
pointme: zaslon 17:50
pointme Tracks blog posts. It can be instructed to follow new blogs from the channel if you are particularly brave. It is written in Perl 6: github.com/carlins/rssbot
moritz_ pointme: hugme
pointme Hugme hugs people! And gives commit access to various repos. It's source is at github.com/moritz/hugme/
frettled zaslon: link frettled
zaslon frettled's blog is at howcaniexplainthis.blogspot.com/
frettled So that works, but not the one that's extracted from RSS/Atom, hmm.
diakopter pmurias: hi 17:51
zaslon lolfrettledhazblogged! frettled++ 'Oslo.pm Past and Future': http:
lolfrettledhazblogged! frettled++ 'Oslo.pm Past and Future': http: 17:54
moritz_ lolzaslonfail 17:55
frettled I have a strong suspicion we'll see a repeat performance in thirty seconds. 17:57
zaslon lolfrettledhazblogged! frettled++ 'Oslo.pm Past and Future': http:
frettled oh
Argh.
17:58 fridim_ joined 17:59 zaslon was kicked by moritz_ (moritz_))
frettled Duh, good point, I could have done that, too. :) 18:00
TimToady I was just about to, but was waiting for one more time :)
pmurias diakopter: hi 18:01
frettled . o O ( I wonder if I can manage to sneak in a question to Barack Obama about Perl 6 when he's in Oslo next week ) 18:02
pmurias when will Perl 6 be founded? ;)
frettled «Will the US increase Perl 6 funding to secure peace?» 18:03
moritz_ 0.1% of their defense costs will do nicely
"defense" 18:04
TimToady given the email logs, Perl 6 climate research is obviously all faked
frettled this may cause some concern 18:05
TimToady wonders how often the phrase "just fake it for now" pops up...
18:06 snearch joined 18:07 barney joined
moritz_ echo 'select count(*) from irclog where line like "%just fake it%";' | mysql moritz5 18:07
count(*)
5
"fake it" 68 times :-) 18:08
diakopter Results 1 - 10 of about 2,750,000 for "just fake it"
TimToady and we've lost all the data from before 1980!!!
diakopter (google) 18:09
bing: 1-10 of 152,000,000 results
18:09 riffraff left
diakopter lol 18:09
TimToady (all the Perl 6 data)
moritz_ bing knows a lot more about faking than google 18:10
diakopter oh, and bing just found 3,000,000 more instances... now it says 155,000,000
way to go bing!
jnthn Yeah, but is it just faking it? :-) 18:11
TimToady and if so, is it really 155,000,001? 18:12
Wolfman2000 ...is bing really supposed to be that good, despite it being a Microsoft product?
TimToady vacuums have to be good at sucking 18:13
moritz_ I've banned bing from some of my sites, because the crawler referer-spammed 18:14
frettled I think we've router-blocked Yahoo's crawlers, because they insist on DDoS-ing our webservers. 18:16
diakopter Ω.suck
std: Ω.suck
p6eval std 29236: Undeclared routine:␤ 'Ω' used at line 1␤ok 00:01 103m␤
TimToady gee, where'd you find the horseshoe dingbat? :) 18:17
frettled pure luck
TimToady but as you can see, if you hang it upside down, all the luck runs out
moritz_ frettled: that one is curious too... three times more hits from yahoo crawler than from googlebot, but only about 1% of visitors coming from it 18:18
frettled yeah, that was clumsy, diakopter.
diakopter Ѡ
18:18 brrant left
TimToady what's that, a goat-shoe? 18:18
frettled moritz_: they typically send several parallel crawlers, visiting the same URLs, for reasons unknown to Man.
diakopter cyrillic omega, apparently 18:19
but yes, used as goat shoes in the #perl pen
frettled so it's not Ѡelsh?
TimToady I wonder how soon governments start outlawing fonts that contain indistinguishable glyphs... 18:21
frettled There has already been something about it for IDN. 18:22
The cyrillic version of a, for instance, cannot be used for at least some TLDs. 18:23
18:23 mberends joined
moritz_ the German NIC only allows 4 non-ascii characters that commonly used in .de 18:24
(in .de domains, that is)
TimToady whenever my cat sits on my lap between my laptop and the base station, my wifi connection drops; I must have a dense cat 18:28
moritz_
.oO( or the cat interferes? )
TimToady cats do seem to know a lot about wavefunctions 18:29
frettled Your cat obviously contains a lot of water. 18:30
It's water-fat.
TimToady saltwater, I believe
frettled You know how to use your krysknife, I suppose. ;) 18:31
Or perhaps you are not Fremen. Hmm. Wrong context? 18:35
allbery_b one of the major absorption frequencies of water is in the middle of the 2.4 wireless band 18:36
18:37 kthakore joined
allbery_b inconvenient around "ugly bags of mostly water" 18:37
kthakore hi
18:37 Avada joined
frettled allbery_b: yup, that was the hint, sortof. It's bloody annoying. 18:37
kthakore can some one answer this? www.reddit.com/r/programming/commen...it/c0giz5a
I am curious as well
what would the perl6 code of that look like? 18:38
jnthn say $foo[0]<bar> 18:39
kthakore jnthn: wat are the equivalents of the other line? 18:40
Wolfman2000 rakudo: my @foo = {bar => 42, hi => 32}, {baz => 38, bye => 71}, {qux => 99, yow => 11}; say $foo[0]<bar>;
p6eval rakudo 7ce13d: Symbol '$foo' not predeclared in <anonymous> (/tmp/IssprLTGsq:2)␤in Main (file <unknown>, line <unknown>)␤
Wolfman2000 rakudo: my @foo = {bar => 42, hi => 32}, {baz => 38, bye => 71}, {qux => 99, yow => 11}; say @foo[0]<bar>;
18:40 brrant joined
p6eval rakudo 7ce13d: ( no output ) 18:41
18:41 quantumEd left
Wolfman2000 ...wonder where I went wrong. 18:41
Wolfman2000 just goes back to his homework.
kthakore Wolfman2000: *g*
jnthn my @foo = {bar => 42, hi => 32}, ...;
rakudo: my @foo = {bar => 42, hi => 32}; say @foo[0]<bar> 18:42
p6eval rakudo 7ce13d: ( no output )
jnthn ...
Anyway, the main point is that you don't dereference stuff explicitly in Perl 6.
Wolfman2000 ...can't just be the both of us, can it jnthn?
jnthn Wolfman2000: Suspect it's the bot. It's been acting up for ages. 18:43
rakudo: class Foo { }; say "ok"
p6eval rakudo 7ce13d: ( no output )
18:45 ruoso left, dakkar joined 18:46 quantumEd joined
pmurias kthakore: that reddit post is very silly 18:46
TimToady but you repeat yourself 18:47
18:49 pmurias left
Wolfman2000 jnthn: I just ran my short program on Feather. It worked 18:50
kthakore p6eval: um ...
Wolfman2000: wth? 18:51
he just called me silly and left :p
hilarious
oops he didn't call me silly stupid me 18:52
18:52 Avada left
kthakore jnthn: anywho ... are there different way of accessing the @foo like in perl5 18:52
jnthn kthakore: Well...there are...but unlike Perl 5 I think there's One Obvious Right Way. :-) 18:55
kthakore: I mean, you could write @foo[0]{'bar'} too.
Wolfman2000 actually jnthn, that fails on rakudo 18:56
Could not find non-existent sub bar
jnthn Wolfman2000: I suspect you didn't write what I just did.
Wolfman2000 just compiled the master branch too
jnthn Wolfman2000: Note the quotes :-)
(around bar)
Wolfman2000 @foo[0]{'bar'}
lambdabot Unknown command, try @list
Wolfman2000 well aware
jnthn :-S 18:57
ng: my @foo = {bar => 42, hi => 32}; say @foo[0]{'bar'}
p6eval ng a77213: get_pmc_keyed() not implemented in class 'Block'␤current instr.: '_block14' pc 29 (EVAL_1:0)␤
jnthn oh, we didn't do hash composers in ng yet.
rakudo: my @foo = {bar => 42, hi => 32}; say @foo[0]{'bar'}
p6eval rakudo 7ce13d: ( no output ) 18:58
moritz_ ng: my @foo = {bar => 42, hi => 32}; say @foo[0]{'bar'}
p6eval ng a77213: get_pmc_keyed() not implemented in class 'Block'␤current instr.: '_block14' pc 29 (EVAL_1:0)␤
Wolfman2000 ng: my @foo = {bar => 42, hi => 32}; say @foo[0]<bar>
p6eval ng a77213: get_pmc_keyed() not implemented in class 'Block'␤current instr.: '_block14' pc 29 (EVAL_1:0)␤
kthakore jnthn: ah ok
Wolfman2000 ...sounds like a good thing to implement soon jnthn
moritz_ ng: my @foo = {bar => 42, hi => 32}.(); say @foo[0]{'bar'} 18:59
p6eval ng a77213: get_pmc_keyed() not implemented in class 'Undef'␤current instr.: '_block14' pc 29 (EVAL_1:0)␤
jnthn Wolfman2000: Yeah...can likely steal it straight from master. 19:05
19:05 iblechbot joined 19:08 ruoso joined 19:09 ingy joined
kthakore moritz_: so ... Perl6 cannot access an element from a array of hashes? 19:11
moritz_ kthakore: well, it can 19:18
rakudo has problems with p6eval, and ng with hash literals
frettled std: my @foo = {bar => 42, hi => 32}; say @foo[0]{'bar'}
p6eval std 29236: ok 00:01 108m␤ 19:19
moritz_ but in principle it works just fine
$ perl6 -e 'my @foo = {bar => 42, hi => 32}; say @foo[0]{"bar"}'
42
19:19 jaldhar left 19:20 masak joined
diakopter that's interesting 19:21
o wait; lagged
backlagged
masak what do we tell this person? twitter.com/dpcwollmann/status/6239724751 19:22
frettled ac
19:22 frettled sets mode: +o masak
frettled That you got op! 19:22
masak ooh, I'm scared. :)
oh wait.
frettled GNU readline support would be Really Neat. 19:23
masak is it doable?
frettled I suppose so, but it's not a straight plugin, because there are at least two ways of doing it.
mberends anything NCI is hard
frettled I haven't added readline to other code in several years, and I've only done it once. 19:24
mberends frettled: TAALTWTDI ?
masak dang. but doesn't the REPL alreadu use something like GNU ReadLine?
frettled I shot myself in the foot and didn't get it to understand all the cool commands that people - me included - usually expect from «readline support»
mberends: yep
19:24 frettled sets mode: +o mberends
mberends ooh! 19:25
moritz_ masak: parrot has a config probe for readline too
PerlJam doesn't rakudo grok readline if parrot does or something?
moritz_ I guess that's a good question for pmichaud
masak sometimes I've been toying with the idea of introducting &prompt and GNU readline to each other. 19:26
PerlJam I have one box where rakudo seems to be readline-aware (but that appears to have happened quite by accident)
masak maybe `prompt("> ", :@history)` or sump'n.
moritz_ PerlJam: the REPL, or prompt()? 19:27
PerlJam the REPL 19:29
moritz_ it is here, too
frettled Hmm, yes, running ./perl6 in the September 13 edition of rakudo _does_ yield a command history
it understands ESC-b and ESC-f for backward-word and forward-word, etc. 19:30
masak sushi & 19:41
19:41 perlygatekeeper left 19:47 barney left 19:48 SirKay joined 19:51 xenoterracide left 19:52 xenoterracide joined 19:56 zostay joined 20:04 alexn_org joined 20:05 sundar joined, alexn_org left, alexn_org joined
diakopter TimToady: does a non-greedy quantifier terminate ltm? pmichaud? 20:12
PerlJam HLL::Compiler handles the REPL and the command line, right?
20:15 brrant left
diakopter TimToady: duh, I see the answer right there (sry) 20:18
sweet 20:19
that makes the right-to-left conversion much easier :)
20:20 pdcawley joined
Tene PerlJam: Yes. 20:22
pugs_svn r29237 | colomon++ | [perl6advent] Add topic to Dec 7, minor brainstorming notes. 20:23
diakopter compiling leading LTM patterns is analogous to computing leading FIRST patterns, but a tad higher order. 20:24
masak pointme: masak
pointme masak submits rakudo bugs ;-)
masak pointme++ # spreading the meme that the bugs are MINE! :P 20:25
moritz_ pointme: moritz
pointme Sorry, I don't know anything about that project
masak we should definitely add something on project moritz.
moritz_ neither do I
20:30 bluescreen left
moritz_ colomon, masak: I think my access to perl6advent.wordpress.com doesn't work... I'd dump my post in the pugs repo instead and let one of you publish it, if that works for you 20:33
masak moritz_: sure, but it's too bad. better to have the post marked up by its author. 20:35
colomon moritz_: do you have a wordpress account? 20:36
moritz_ colomon: yes
20:36 moritz_ sets mode: +o colomon
colomon e-mail? 20:37
address, I mean?
for it?
moritz_ colomon: the one you sent the invitation for
s/for/to/
colomon okay, you're in now. 20:38
moritz_ works 20:39
colomon++
what format does wordpress want? html? 20:41
colomon html or it has a rich test editor.
20:41 bluescreen joined
moritz_ ok. So I'll write in pod, convert it to HTML and feed that into wordpress :-) 20:41
diakopter note to others who may ever follow this (reversing) strategy: alternations keep their left-to-right order, it's just sequence (both, concat, etc.) whose order needs switched (incl all literals more than 1 char in length) 20:42
colomon That appeared to be masak's plan as well? :)
I need to sit down and learn pod at some point. pod and grammars. 20:45
20:46 envi^home left
arnsholt is considering volunteering to write something on grammars 20:47
Except I should probably read for my exams instead =)
20:53 ruoso left
PerlJam colomon: er... how do I post to perl6advent? 20:53
colomon PerlJam: did you register with wordpress? 20:54
PerlJam aye
colomon pobox e-mail? 20:55
PerlJam (I apparently already had an account that I've never used)
yes, [email@hidden.address]
colomon okay, you're in there too.
PerlJam ah. okay 20:56
colomon++
ascent_ huh, any ideas for domain perl6.pl? :) 20:57
moritz_ forward it to perl6.org until you have a good idea :-) 20:58
hey, if you read "pl" as "people" it would be a great domain for a community site
Tene I could post something on... exceptions? 20:59
PerlJam Tene: go for it.
moritz_ Tene: sure, good idea
20:59 quantumEd left
colomon I'd take exception to that. 21:00
Tene Hm. I'd need useful examples of exceptions use, though...
colomon ;)
21:00 quantumEd joined
Tene see, colomon objects, I can't do it. 21:00
I'll think of something else, then.
PerlJam I think he was just punning.
(for the humor of it) 21:01
colomon exactly.
apologies, it was irresistible. I'm surprised no one beat me to it.
PerlJam Hmm. I like Wordpress' interface for editing posts much better than blogspots. 21:02
colomon so far I am very impressed with Wordpress, except that the invite thing seems not to work quite as I imagined. 21:03
moritz_ colomon: so you have to invite *and* add people to the blog?
colomon moritz_: yes.
PerlJam invite == invite people to use wordpress
moritz_ rakudo: my Int $x; $x *= 30; say $x 21:04
p6eval rakudo 7ce13d: 30␤
colomon The interface implied that it would be automatic, but as far as I could tell it never worked that way.
PerlJam moritz_: that's a feature! ;)
moritz_ PerlJam: I know, I just wanted to be sure it worked in rakudo
PerlJam rakudo: my Int $x; $x += 1; say $x;
p6eval rakudo 7ce13d: 1␤
21:05 ShaneC joined
PerlJam (just checking :) 21:05
colomon It has an "Add user to my blog as a contributor" tick box I ticked, but no one ever got added.
ng: my Int $x; $x *= 30; say $x
p6eval ng a77213: 0␤
colomon d'oh!
PerlJam colomon: well, a Mu Int is devoid of value. :) 21:06
colomon ng: say Int ~~ Mu
p6eval ng a77213: 1␤
21:08 takadonet left
masak ng: Mu[Mu] 21:08
p6eval ng a77213: Method 'postcircumfix:<[ ]>' not found for invocant of class ''␤current instr.: '!postcircumfix:<[ ]>' pc 243650 (src/gen/core.pir:25138)␤
masak jnthn: "invocant of class ''"? 21:09
moritz_ ng: 3[4]
p6eval ng a77213: Can't postcircumfix:<[ ]> foreign objects yet.␤current instr.: '!postcircumfix:<[ ]>' pc 243650 (src/gen/core.pir:25138)␤
masak foreign objects? 21:10
moritz_ alien Ints 21:11
masak ng: trois[quatre]
p6eval ng a77213: Confused at line 1, near "trois[quat"␤current instr.: 'perl6;HLL;Grammar;panic' pc 519 (src/stage0/HLL-s0.pir:336)␤
moritz_ changed his image on gravator.com and is surprised that github updated it immediately 21:12
masak nice new image.
colomon ack, too much perl 6 makes programming in C++ painful.
moritz_ masak: thanks. I put it on my CV today 21:13
colomon I mean, why can't I just say for trimmed_surfaces, breps -> $brep { ... } ? Requiring a new loop to iterate over each array is a PITA.
PerlJam colomon: too much perl6 makes programming in <insert almost any language> painful.
masak ooh! pmichaud-on-parrotdev says I can speed up my Perl 6 code by turning off GC. :P 21:15
sjohnson yo masak 21:17
masak sjohnson: hi there! 21:20
21:20 ShaneC left 21:21 ShaneC joined
sjohnson rakudo: my $x = "\t"; say split("\t", $x).perl; 21:23
p6eval rakudo 7ce13d: ["", ""]␤
colomon PerlJam: I almost thing that one of my projects for next year ought to be building a p6 interface for my $work C++ libraries, just to make writing tests easier... 21:24
sjohnson TimToady: is the above ^^ possible in p5, without getting complicated? or was its split not meant for this, where as p6 has comb and split
PerlJam sjohnson: my $x = "\t"; print split(/\t/, $x, -1), "\n"; 21:25
moritz_ I've written my advent post for Dec 04, and scheduled it for publishing
PerlJam sjohnson: (not sure what part you think might not be possible)
moritz++
moritz_ those of you who have admin access can probably view it already, and review/correct/bikeshed 21:26
colomon dude! scheduling ahead of time! awesome!
PerlJam colomon: yes, blogspot has that too, but it's way klunky
sjohnson PerlJam++ # awesomeness
colomon I see PerlJam has gotten to work, too. :) 21:28
PerlJam colomon: not really :)
But I plan to do what moritz_ has just done as well.
masak moritz_: ooh, scheduled posts! I gotta take a look at that. 21:29
colomon moritz_++ # nice post 21:30
moritz_ thanks colomon
colomon The newline formating is a little off, if you don't mind I will try to clean it up.
21:31 SmokeMachine left
moritz_ feel free 21:31
masak moritz_: can I change the dash in the title to a colon, to conform with today's post?
moritz_ masak: sure
masak I'll wait a bit, to avoid edit conflicts. :) 21:32
21:33 brrant joined
colomon "Warning: Moritz is currently editing this post" 21:33
PerlJam There have been 59 views for Day 1. Not bad for no advertising :) 21:35
moritz_ plans to do some advertising tonight
colomon how's that?
PerlJam moritz++
21:35 bluescreen left
colomon masak -- I take it this means you can see scheduled posts as well? 21:36
(I took care of the dash, btw.)
masak aye.
thanks.
moritz_ colomon: can you resubmit your changes? I fear I have hit reload accidentally 21:37
and undone your changes :(
moritz--
colomon not sure if I can resubmit, but I'll see.
PerlJam wonders if there's a widget for displaying code
(maybe syntax highlighted)
moritz_ I've now logged out so that it dosn't happen again
colomon yay, still had all the changes sitting in my browser. 21:38
PerlJam: I believe someone told me there was over the summer, but I don't remember the details.
(because I wasn't using wordpress at the time, so it was irrelevant to me.) 21:39
oh, moritz_'s test comments don't match the actual functions called. :)
moritz_ dammit 21:40
PerlJam Hmm. I wonder how/if we should use the categories? 21:42
moritz_ I don't see much benefit. tags > categories
colomon moritz_: shall I fix them? 21:43
moritz_ colomon: please do, if you feel like
PerlJam masak: you'll let us know when your entry is ready for critique? :) 21:44
colomon done.
masak PerlJam: oh, it has been all day. :)
moritz_ colomon++
colomon So far, I'm really liking the blog setup for the calendar. PerlJam++ for suggesting creating a blog for it. 21:45
PerlJam colomon++ for actually doing it! :)
masak colomon++ PerlJam++
moritz_ lolmoritzhazblogged: perlgeek.de/blog-en/perl-6/perl6-ad...endar.html
masak would somebody mind changing the theme of the perl 6 advent blog -- or show me how to do it? the current one feels a bit drab. 21:46
moritz_ masak: put a camelia in there somewhere!
PerlJam camelia +1
If it could be styled similar to perl6.org, that would be awesome 21:47
masak the blog has had 295 views today. I think it'll 10x or more before December is over. 21:48
colomon I think probably only PerlJam and I have permission to change the theme. 21:49
moritz_ colomon: then you need to hand out theme bits liberally :-) 21:50
PerlJam indeed.
The chance me changing the theme is close to zero right now. :)
masak either is fine by me. :) as long as the theme changes.
moritz_ just redditted it 21:51
PerlJam masak and moritz are now admins too
colomon PerlJam++ 21:52
masak yay, responsibility! \o/
PerlJam and jnthn
jnthn huh...what...
I have to be responsible?!
PerlJam you have the option to be responsible. 21:53
colomon I would definitely encourage someone other than me to work on the theme. My web pages are pretty uniformly boring. :)
masak heh. I just got called a 'true [Perl 6] believer' on Twitter. again. :) 21:54
PerlJam Where in pugs is perl6.org? 21:55
colomon (well, hopefully my content isn't that boring, but the style is very "1997 clean" .)
21:55 zaphar_ps left
sjohnson @karma 21:55
lambdabot You have a karma of 4
21:55 zaphar_ps joined 21:56 rgrau_ left 21:57 rgrau_ joined
moritz_ PerlJam: docs/feather/perl6.org/ 21:58
PerlJam yeah, just found it :)
thanks
moritz_ PerlJam: feel free to put it as the second item under Community on the front page 21:59
PerlJam: the IRC channel takes too much space there
pugs_svn r29238 | duff++ | [perl6.org] Add link to Perl 6 Advent calendar 22:04
PerlJam so ... how does the website get updated? 22:05
moritz_ by cron job 22:06
22:06 alexn_org left
colomon perl6advent got its first comment! \o/ 22:10
PerlJam doesn't believe it. 22:11
jnthn \o/ 22:13
Nice!
colomon ng: say (7/4+2/3).perl
p6eval ng a77213: 29/12␤
PerlJam ng: say (4/4+2/3).perl 22:17
p6eval ng a77213: 5/3␤
PerlJam rakudo++ 22:18
22:18 iblechbot left
avar Who turned Perl into Lisp?:) 22:19
sundar Hi.. Is it possible to enable readline support in the interactive perl6 interpreter?
masak sundar: it's enabled here. 22:20
moritz_ more specifically the readline header files need to be installed when parrot is configured 22:23
masak moritz_++ # a reply which actually helps
moritz_ libreadline5-dev on debian 22:25
sundar I'm on a Solaris machine.. I'll try to find if it has readline headers... Though, I'll have to then configure and build both parrot and rakudo, isn't it? 22:27
moritz_ right 22:28
22:28 fridim_ left
moritz_ rakudo: say %*VM<config><readline> 22:29
p6eval rakudo 7ce13d: Use of uninitialized value␤␤ 22:30
moritz_ on my computer that gives "define"
and the parrots Configure.pl says, among other things 22:31
auto::readline - Does your platform support readline.................yes.
22:36 dakkar left
PerlJam okay, so I said the chances of me changing the theme for perl6advent were zero ... turns out it was closer to one (such is being binary). I've just added a couple of Camelia to the header, but perhaps it inspires someone to do better 22:37
(the past 30 minutes was the most I've ever played with GIMP too :)
frettled :)
PerlJam++ - the website did improve 22:38
arnsholt Is the plural of Camelia Cameliae? =) 22:43
jnthn Depends on the language I guess. :-) 22:44
frettled Papilio Camelia?
arnsholt True, true 22:45
jnthn chooses Camelie ;-)
arnsholt Lepidotera celopardalis? =D
Er, Lepidoptera
And camelopardalis
Apparenly I fail at writing 22:46
frettled ng: say arnsholt.fail;
p6eval ng a77213: Could not find non-existent sub &arnsholt␤current instr.: '_block14' pc 29 (EVAL_1:0)␤
frettled ;)
arnsholt ^^ 22:47
22:50 snearch left 22:52 IllvilJa left 22:53 pmurias joined
pmurias kthakore: rehi 22:58
diakopter pmurias: g'morning? :) 23:02
pmurias diakopter: it four minutes past midnight here 23:04
diakopter oh
moritz_ 22:48 <@masak> the blog has had 295 views today. I think it'll 10x or more before December is over. 23:06
50x at least, if we continue to blog
masak moritz_: :)
do I hear 250x? 250x, anyone? 23:07
moritz_: I meant per day.
moritz_ oh. 23:08
I thought you meant in sum
23:08 IllvilJa joined
masak IllvilJa! \o/ 23:12
23:12 frew__ joined, frew__ left
IllvilJa Howdy! 23:13
Just did my first github commit and push! 23:14
Tiny patch, replaced 3 characters with 2 other characters in a Build.PL script. But still, baby steps!
masak IllvilJa++ 23:15
23:15 frew__ joined, frew__ left
IllvilJa Anyone knows if Russ Allbery is here at times? I submitted, by mail, a patch for his Term::ANSIColor module. 23:16
masak is it a Perl 5 module?
IllvilJa The purpose were to let the colors 8 through 15 (you know, those BRIGHT and GOOD LOOKING colors) be available in addition to the duller color 0 through 7...
Yes it is perl 5.
masak I don't recognize the name "Russ Allbery". 23:17
allbery_b this channel is for perl 6 development. perl 5 modules are somewhat irrelevant.
IllvilJa I asked about it on #perl, and I got a good response to ask the quesion on some mailing list for module maintainers on CPAN (but has not gotten around to do it yet)
masak IllvilJa: I'm very familiar with those bright and good looking colors, yes. :)
allbery_b (and no, I am not Russ)
masak IllvilJa: my teens were spent in BASIC. both text mode and EGA used those 16 colors. :) 23:18
IllvilJa (I know, I accidentally IMed you when I thought you were Russ... then realized I perhaps should use the 'info' command of Pidgin and thus understood I were mistaken...)
The thing is that I (probably incorrectly) think Term::ANSIColor is a good candidate to port to perl6. 23:19
masak just to be clear: I am not Russ, either.
23:20 meppl left
masak IllvilJa: I think it would, too. 23:20
IllvilJa Maybe I should decide to be Russ, accept my own patch and when I try to port Term::ANSIColor to Perl 6 include support for the bold colors numbered 8 through 15. 23:21
masak then we can brag that the Perl 6 module is better than the Perl 5 module. :P 23:22
IllvilJa Um.. rather have it as an example that the Perl5 module's patch should be implemented and get Term::ANSIColor 2.03 created :-D.
masak or that. :) 23:23
masak gets through the whole &LazyMap::iter subroutine, and actually understands it! 23:28
IllvilJa The nice thing with Term::ANSIColor is that it is a conceptually simple module, so I can follow the "you are not good enough" approach and create some very "baby perl 6"-ish module which works but otherwise is "full of fail".
Then I can improve it over time and learn some perl 6 in the process (or hope someone who is good enough tries to improve it ;-) ) 23:29
23:29 frew joined
moritz_ sounds like a very good plan 23:30
masak I don't speak for everybody, but the "you are not good enough" approach is the only one I've ever followed. it might be too early for anyone to do anything else with Perl 6. :)
colomon masak: LazyMap::iter ? 23:31
IllvilJa Worse is better. Fail is success. Absolutely useless is optimal. Hm... we start to sound a bit like Orwell's 1984 now...
masak colomon: ok, look, so I'm trying to steal STD's backtracking for GGE.
moritz_ IllvilJa: the amazing thing is that it actually works. I started perl6-projects.org, somebody contributed a nice design. I blogged about an XML grammar, somebody else made a module of it, and put it on github, improved it etc.
masak colomon: in so doing, I've been given a few pointers from TimToady where to look.
colomon masak: ah. carry on, then. ;) 23:32
moritz_ masak: why? doesn't PGE do backtracking in a module that fits better for GGE?
masak colomon: I think I just understood the core part of it. still don't really see where the backtracking takes place, though.
colomon still longs for iterator interface in ng...
masak moritz_: PGE cheats. :(
moritz_ masak: how so?
masak moritz_: it uses Parrot's coroutines and continuations.
moritz_: of course, if you do it *that* way, then it's easy! 23:33
moritz_ masak: well, then get real continuations into Perl 6, and do it that way :-)
masak moritz_: believe me, the thought has occurred to me. :) 23:34
cf my last talk.
frettled IllvilJa: What about 256-color while you're at it? :D (frexx.de/xterm-256-notes/) 23:35
pmurias masak: gather/take can be used for getting coroutines 23:36
moritz_ pmurias: the laziness of gather/take is not completely guarntueed by the spec 23:37
pmurias: does smop have continuations?
IllvilJa frettled: Well, the 16 color palette seem to be some sort of "standard" that many terminal emulators support.
masak pmurias: yes. that doesn't mean I (1) like that as my only option, or (2) see how to cross routine boundaries with gather/take.
pmurias moritz_: the could be easily implemented
IllvilJa With color 3 being dull yellow and color 11 is corresponding bright (and vivid and enlightening) yellow... 23:38
frettled IllvilJa: Yes, it's based on 16-color ANSI. The 256 color palette is based on 256-color ANSI, IIRC.
IllvilJa Yes. I think the gnome-terminal (and it's underlying widget, vte) only supports the 16 color palette.
pmurias moritz_: smop has one-shot continuations - one would have to add a .clone method to them
frettled IllvilJa: 256-color support is part of the xorg distribution, see e.g. www.opensource.apple.com/source/X11...colors2.pl for how to print those. 23:39
moritz_ so, let's summarize
actively developed Perl 6 compilers:
frettled IllvilJa: but your platform may not necessarily have it compiled into xterm, of course.
moritz_ rakudo, mildew, something JS based
rakudo can use parrot's continuations
IllvilJa Ok, 256 colors might show up in some future version of the Perl 6 Term::ANSIColor module :-).
pmurias diakopter: what is your current continuation?
diakopter: err compiler
frettled IllvilJa: \o/
moritz_ mildew would not be hardpressed to get support via smop 23:40
frettled IllvilJa: goodie, because it's been in rxvt for ages :D
IllvilJa frettled: I'm a bit old fashioned, I'm damaged from learning programming on (Microsoft!!) basic on a Commodore 64 back in the 1980.s :-).
moritz_ sprixel used something CPSy anyway, afaict
colomon IllvilJa: same here.
pmurias moritz_: it would be also be easy to add them to mildew-js
moritz_ so, what stops us from real, guarantueed continuations in Perl 6?
IllvilJa So 256, 4096 and 32 bit colors are really nice, but I'm a sucker for few-color (lack of) aesthetics.
frettled IllvilJa: I may have learned programming at least thrice, the second time in Microsoft BASIC on an IBM PC. :) 23:41
23:41 slavik left
masak moritz_: TimToady thinks continuations are like naked singularities. they should not be exposed directly in a programming language. 23:41
IllvilJa I were shocked a couple of years ago to learn that the Basic I "grew up on" on the Commodore 64 actually were written by Microsoft.
masak moritz_: but 'the event horizon is a lazy list'.
moritz_ masak: maybe it's one of those things that should be made possible, not necessarily easy? 23:42
colomon IllvilJa: almost every small computer's basic was microsoft's back then, as far as I know.
masak IllvilJa: yes. that's stuff that Bill Gates actually wrote, rather than copied. :)
moritz_ anyway, bed time here
'night 23:43
frettled IllvilJa: Anyway, people who _really_ need this can of course reuse the code in the 256colors2.pl script, but it's nice to have things in modules for some people :)
masak night, moritz_++.
IllvilJa Seriously, the good thing with having just support for 16 colors is that you can give each color a symbolic name ('yellow', 'bright_red' etc). Then of course 256-colors support can use some sort of parametric names like color_33 or such...
colomon goodnight, moritz_
frettled IllvilJa: I think it's cool that you're patching, though.
IllvilJa: uh, there are far more than 256 color names out there :D
IllvilJa: See rgb.txt (IIRC) in any X11 distribution.
pmurias moritz_: continuations limit your choice of implementation methods 23:44
frettled IllvilJa: sedition.com/perl/rgb.html
IllvilJa frettled: yes, but I would go even more nuts if I try to sort them out and map them to 256 specific colors. Then I think parametrized color names (specifying r, g and b of each color as part of the name) is more manageable.
frettled IllvilJa: those are well-defined in 8 bpc (16.7 million colors)
IllvilJa: the color codes are there, the mapping is already done by others. 23:45
IllvilJa frettled: ok. seem like some area to learn some more about in the future.
jnthn prints out two papers to read on his journey tomorrow 23:46
pmurias jnthn: which ones?
IllvilJa Well, it is fun to know that Bill Gates did SOME things that has benefit me directly. (Another thing I think Microsoft created were the DHCP standard which also is a very nice thing to have) 23:47
frettled IllvilJa: it's going to be out there on the intertubes for as long as we live, I suppose, so it should be easy to check out. And of course, it's in your Linux distribution if it uses Xorg, and if you use Emacs, you can easily look them up.
IllvilJa frettled: cool then.
frettled: but as stated before, ignorance is strength, failure is success and worse is better, so I go initially for a worse variant which only knows about 16 colors... 23:48
frettled IllvilJa: for instance, my favourite settings for an xterm or Emacs is cornsilk as the foreground color, black background. The reason is that white on black is too contrasty.
IllvilJa: hee-hee :D
IllvilJa: yep, better to actually implement a small improvement than none at all. 23:49
Anyway, enough bike-shedding on my part for tonight, goodnight!
IllvilJa frettled: I have a sick habit of creating one single colorscheme of foreground/background/cursor for every host I'm connecting to (and a matching terminal icon symbol). Makes it easy to instinctively track all different terminal windows I have lying around.
So yes, being able to use colors is not just a nice thing, it's critical to quickly manage all SSH connections. 23:50
At least for me.
jnthn pmurias: citeseerx.ist.psu.edu/viewdoc/summa....1.32.1784 and citeseer.ist.psu.edu/old/paepcke93userlevel.html 23:52
23:58 slavik joined