»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend!
Set by sorear on 4 February 2011.
jnthn morning o/ 00:20
phenny jnthn: 25 Sep 18:45Z <[Coke]> ask jnthn how to run a QAST CompUnit in a standalone NQP program when you're testing QAST.
jnthn [Coke]: See t/qast/qast.t which has code you can take 00:21
TimToady o/ 00:27
japhb has now had the experience of submitting a functionality patch for a project written in a language he doesn't actually know 00:45
felher 'night, #perl6 :)
[Coke] jnthn++ for putting up with my questions. 01:16
[Coke] ok, I have a standalone qast that is compilable and does what I need. now to figure out why the thousands of lines of nqp aren't working. ;) 01:22
japhb jnthn++ # QAST docs 03:28
More please! ;-)
japhb sorear, why doesn't niecza compile the setting when you `make`? 03:48
Or is there a different make invocation now that I should use instead?
sorear japhb: it does, but niecza by default can't find the setting you just compiled
japhb Oops
sorear if you run niecza without arguments it assumes it is in installation mode and puts all its writable data in $HOME 03:49
sorear phenny: tell moritz irclog.perlgeek.de has seemed very slow the past day or two, is something up? 05:20
phenny sorear: I'll pass that on when moritz is around.
moritz sorear: the whole server has seemed slow the past few days 05:25
phenny moritz: 05:20Z <sorear> tell moritz irclog.perlgeek.de has seemed very slow the past day or two, is something up?
moritz good morning 05:26
GlitchMr .tell sirrobert or * + * for the anonymous function you want 05:35
phenny: tell sirrobert or * + * for the anonymous function you want
phenny GlitchMr: I'll pass that on when sirrobert is around.
diakopter std: ***************************** 05:41
p6eval std 77327a4: OUTPUT«ok 00:00 42m␤»
diakopter rn: ***************************** 05:42
p6eval rakudo 16f22b: OUTPUT«===SORRY!===␤HyperWhatever (**) is not yet implemented. Sorry. ␤at /tmp/fXzV2TfSMR:1␤»
..niecza v22-3-g9840f57: ( no output )
diakopter n: say *****************************()
p6eval niecza v22-3-g9840f57: OUTPUT«Unhandled exception: System.NullReferenceException: Object reference not set to an instance of an object␤ at Niecza.StashCursor.Core (System.String key, Boolean final, Niecza.StashCursor& sc, Niecza.Variable& v, Niecza.Variable bind_to) [0x00000] in <filen…
diakopter whee; nieczabug
sorear n: say ** 05:44
p6eval niecza v22-3-g9840f57: OUTPUT«Unhandled exception: System.NullReferenceException: Object reference not set to an instance of an object␤ at Niecza.StashCursor.Core (System.String key, Boolean final, Niecza.StashCursor& sc, Niecza.Variable& v, Niecza.Variable bind_to) [0x00000] in <filen…
sorear n: **
p6eval niecza v22-3-g9840f57: OUTPUT«Unhandled exception: System.NullReferenceException: Object reference not set to an instance of an object␤ at Niecza.StashCursor.Core (System.String key, Boolean final, Niecza.StashCursor& sc, Niecza.Variable& v, Niecza.Variable bind_to) [0x00000] in <filen…
sorear n: *
p6eval niecza v22-3-g9840f57: ( no output ) 05:45
diakopter sorear++ golfo
sorear should just print something to the effect of "HyperWhatever NYI"
diakopter n: say *****************************; # why is this ok 05:46
p6eval niecza v22-3-g9840f57: OUTPUT«{ ... }␤»
sorear n: -> $x { ** } 05:47
p6eval niecza v22-3-g9840f57: OUTPUT«Potential difficulties:␤ $x is declared but not used at /tmp/ehGqO53UZE line 1:␤------> -> ⏏$x { ** }␤␤»
sorear diakopter: I guess currying means that the ** is not actually executed
n: say '*****************************'.chars
p6eval niecza v22-3-g9840f57: OUTPUT«29␤»
sorear 7*4+1, that's 7 HyperWhatever, 7 infix:<**>, and 1 Whatever 05:48
the Whatever prevents execution
sorear OT: what are the well-supported Perl 5 module build/metadata tools these days? 05:51
like what I used Module::Install for five years ago
doy dist::zilla 05:53
sorear heh 05:55
diakopter r: say X.WHAT # what is X2 06:10
p6eval rakudo 16f22b: OUTPUT«X()␤»
diakopter r: say X.WHAT # what is X
p6eval rakudo 16f22b: OUTPUT«X()␤»
diakopter r: X XX X 06:12
p6eval rakudo 16f22b: ( no output )
sorear diakopter: X is a package object 06:18
r: X.WHO.<NYI>.new(feature => "HyperWhatever").throw 06:19
p6eval rakudo 16f22b: OUTPUT«HyperWhatever is not yet implemented. Sorry. ␤ in block at /tmp/UIy1xRNaDx:1␤␤»
FROGGS good morning 06:51
sorear good morning FROGGS 06:55
dalek rlito: 2c9a199 | (Flavio S. Glock)++ | / (2 files):
Perlito5 - js3 - fix @_
 07:57
dalek ecza: 9acb21f | sorear++ | lib/ (3 files):
STable knows own setting, more Top removal
 08:48
ecza: 219f822 | sorear++ | lib/ (2 files):
Refactor Frame to always have a valid sub
masak mo ning, #perl6 08:55
moritz \o masak
masak notices with interest that his brain helpfully fills in an 'r' between 'mo' and 'ning'... 08:56
sorear o/ masak 09:03
dalek ecza: 9e5350d | sorear++ | / (4 files):
Top removal: JSYNC, Bool boxing, some list bits; unrot JSYNC class support
 09:14
sorear sleep& 09:16
masak SWADOL[Niecza improvements] 09:22
moritz what does the L stand for? 09:24
masak "Large". 09:26
seems the "sleep well and dream of" meme has its roots in a movie that both [Coke] and I have seen, and that you should watch too. 09:27
moritz which movie? 09:28
masak www.imdb.com/title/tt0093779/
really a tropefest like few others.
moritz ooh, I know the music 09:29
it's by Mark Knopfler
masak "Rodents Of Unusual Size? I don't think they really exist." <gets attacked by one>
moritz makes a mental note to obtain the movie 09:30
dalek rlito: 30ca26b | (Flavio S. Glock)++ | TODO-perlito5:
Perlito5 - TODO - add error report
 09:31
masak you won't regret it, I think. 09:32
"Hello. My name is Inigo Montoya. You killed my father. Prepare to die."
daxim ewww, it's got kissing in it.
tadzik gross 09:33
masak I guess it does. but I don't especially remember any kissing, so it must be explipsed by all the other things in the movie :)
eclipsed*
by the way, is Eclipse called that because when you start it, the lights dim in the rest of the house? :P 09:34
moritz masak: I guess because they wanted to conquer Sun
now they should call it 'well', because an Oracle can fall into a well 09:35
:-)
masak :D
I thought it was philosophers that fell into wells.
Oracles are more likely to be tarred and feathered, or to have their eyes poked out. 09:36
daxim warning: do not look into stick with remaining eye
masak *lol* 09:37
I want to spend the rest of my life in the company of people smart enough to produce snowclones on existing memes.
moritz wonders when Ronja will start doing that 09:40
masak the thought is dizzying. 09:41
moritz speaking of her, she usually communicates with single words or two-word-sentences
moritz but sometimes she surprises us with fully formed sentences 09:41
one she surely picked up verbatim, "willst du noch mehr haben?"
masak moritz: I remember when my nepwhew did two-word sentences. that was cool to observe.
moritz but one she must have crafted on her own: "Papa, da ist die Nase!" 09:42
masak :P
moritz it's amazing how expressive you can be with one or two words, if that's all you have
kresike hello all you happy perl6 people 09:42
masak moritz: at one point, he wanted to express the absence of a dog in some context. so he said first "inte hund" and then "hund inte". which I found interesting from a language development perspective. 09:43
moritz \o kresike
masak hello happy kresike! \o/
moritz: we could all learn from that two-word expressivity.
moritz is sv:inte the same as no:ikke?
masak moritz: yeah, we are the ones diverged on that 'un.
daxim masak, I wonder whether you knew about the website TMOL, a prototypical q&a-site (modeled after enwp.org/Conversatron and the like), but with its unique blend of video games, oughties' pop culture and buddhism, the answers snocloned quite a bit 09:44
moritz btw Ronja recognizes herself in the mirror, but doesn't yet have a word for herself
masak daxim: I didn't know about it. thank you. 09:45
& 09:47
moritz Su-Shee++ # always really good, funny links in the twitter feed 11:13
Su-Shee moritz: I need my laugh per day and I like good cartoons of all kinds :) 11:14
moritz: did you see my ultimate hug picture a few days ago? I pasted it here :)
moritz Su-Shee: yes, I did
Su-Shee: just now I've reposted the "life of a software engineer" to two other channels :-) 11:15
Su-Shee *hihi*
moritz www.bonkersworld.net/images/2011.11..._a_swe.png # for reference
dalek rlito: e8d1c6d | (Flavio S. Glock)++ | / (2 files):
Perlito5 - js3 - fix join()
 12:27
[Coke] Su-Shee: my family loved that hug pic. Thanks. ;) 12:41
[Coke] ponders a ③rd cup of coffe. 13:54
brrt disadvises 3 cups 13:56
not really dangerous in itself, but bad for sleep
and the effects on the sympathetic nervous system are... vague 13:57
timotimo almost every person who ever did coffee has died, so ...
brrt :-) 13:58
[Coke] brrt: it's a choice between coffee or coke zero. the caffiene isn't going to change. 14:00
[Coke] switches to his namesake. 14:00
*caffeine 14:01
brrt well, if it doesn't matter it doesn't matter 14:01
sevvie snorts some ground coffee beans.
Su-Shee added sugar and butter to the caffein intake. ;) 14:06
[Coke] ow.
... butter!?
Su-Shee [Coke]: yes. cookies. :)
[Coke] oh, I thought it was a single thing with sugar, butter, AND caffeine!
Su-Shee [Coke]: I could do that, too. Mokka truffles for example. 14:07
sevvie o.o That sounds delicious. 14:11
Su-Shee easy. cook a lot of espresso with the cream before you add it to the truffle ganache. if you like, you can also crunch coffeebeans and caramlize them and add those too. 14:14
Su-Shee tea truffle made accordingly, just don't add caramelized tea leaves ;) 14:15
moritz sounds like something that'll keep me awake the whole night if I eat one at tea time :/
Su-Shee moritz: no, you won't get that much caffeine out of it, one or two truffles would be fine. just don't eat a handful and don't add crunched coffeebeans, those are really dangerous :) 14:16
kresike bye folks 15:04
pmichaud good morning, #perl6 15:54
diakopter hi 15:55
felher o/
jnthn o/ pmichaud 15:56
heh, it's sleep time here :)
pmichaud jnthn: well, that's what you get for being on the wrong side of the world again. :-P
jnthn Yeah, but this side is from the future :P 15:57
pmichaud Meh, the future isn't what it used to be. 15:58
jnthn It was at least tasty this evening... :) 15:59
pmichaud works on 2012.09 star release. 16:00
jnthn \o/
I've got some local patches towards getting capture ~~ signature working.
pmichaud \o/
jnthn Will maybe finish 'em up tomorrow.
jnthn Anyways, sleep... & 16:01
dalek ar: 3d6c3a4 | pmichaud++ | modules/ (10 files):
Update submodules to latest commits ("git pull origin master").
 16:04
ar: 0474ace | pmichaud++ | tools/star/Makefile:
Makefile updates for 2012.09 release.
dalek ar: 67cc1c0 | pmichaud++ | tools/star/release-guide.pod:
Add information about updating submodules.
 16:07
masak dang, I missed jnthn :) 16:11
he's going to sleep really early. it's almost like he's in East Asia or something.
dalek ar: 27a8b20 | pmichaud++ | docs/announce/2012.09:
Add draft release announcement for 2012.09 .
 16:13
ar: 263648e | pmichaud++ | README:
Bump release number in README.
 16:14
leont He wouldn't go there, he hates traveling :-p 16:17
masak yeah, true. jnthn hardly ever travels. nor does he like beer. 16:21
ggoebel perl6: 'o/'.say 16:48
p6eval rakudo 16f22b, niecza v22-6-g9e5350d: OUTPUT«o/␤»
not_gerd could someone please run github.com/perl6/perl6-examples/bl...tide.p6.pl on an up-to-date Parrot/Rakudo? 16:55
I've got some kind of Bohr bug going on - it switches unpredictably between wrking and not working...
masak not_gerd: why do you $line.subst(/\n/,'')? do you expect lines read using $*IN.get to contain newlines? 16:57
masak also, you seem to declare @keys and $frame_size on line 21 but never use them. 16:58
not_gerd masak: dcarrera++ wrote that code 3 years ago, wasn't me ;) 16:59
masak ah. :)
not_gerd I did update some of the benchmrks, though...
masak $lines on line 10 is unused, too.
not_gerd anyway, that file used to run, and now it doesn't
masak :/
not_gerd at least not consistently - every few tries, it sometimes works 17:00
no idea if my system is to blame or someone introduced a regression
not_gerd tries again with --gen-parrot - will take a while 17:02
masak from en.wikiquote.org/wiki/Linus_Torvalds#1991 -- "The Hurd 0.0 was released in August 1996 and as of 2011, is still not complete." 17:08
we need a name for the fallacy of "is still not complete". 17:09
because, as we all know, the nice thing about Linux is that it is complete and that people have stopped improving it. :P
FROGGS GNU Hurd still today is W.I.P, maybe even R.I.P
hehe
masak I'm not disagreeing.
I just think the "is still not complete" description is low-bandwidth enough to be positively misleading. 17:10
just as it is with Perl 6.
FROGGS ya, about Perl 6 I'd say the improvement is going on
masak I rush to add that I think there are a lot of things Perl 6 doesn't do well enough yet for people to consider it ready for them. 17:11
we need to identify those areas and focus on them.
I think we at least partially are succeeding in doing that.
not_gerd see news.ycombinator.com/item?id=4576487 ;)
FROGGS masak: the major problem IMO is that it is so damn slow 17:12
If I had to replace my SOAP stuff at work using perl6, I guess it will be at least 20 times slower, even if it was precompiled 17:13
masak not_gerd: I would agree more with "is not a very good marketing tactic", if I hadn't seen a number of big Perl 5 names come around to Perl 6 and start considering it seriously recently.
FROGGS the second major thing is the lack of important modules
masak FROGGS: yeah. speed is an ongoing concern.
and we absolutely *need* good Perl 5 integration. that will be a watershed thing. 17:14
oh, I suspected this thread would be found under the "Python is doing just fine" article ;) 17:15
[Coke] masak: are you suggesting that Hurd is a useful and usable toolset? (and used?)
not_gerd masak: I mainly posted that because I just read it a few minutes ago, not because I share that opinion
masak I don't think it's a proper use of frames to give a post that title. the comments seem to corroborate that. "You know what? The very fact that you are having to write this post means Python isn't doing just fine."
not_gerd incidentally, the thread is about *Python* not being dead
not_gerd fails at reading before writing 17:16
masak [Coke]: no! I just have an unhealthy researcher-like interest in other long, drawn-out development efforts that a smallish number of people keep contributing to.
masak [Coke]: the Hurd was *probably* killed off by the advent of Linux. I consider the likelihood that it'll play any major role in the future very, very low. 17:18
it could also be that microkernels are actually worse than monolithic kernels, in practice, even though they look more pure on paper.
masak that seems to be what Linus argued against Tannenbaum back in the day. 17:19
but I'm not actually savvy enough to weigh in on that.
not_gerd: well, both Mystery Science^W^WPython 3k and Perl 6 are "failures" each in their own way, on different scales. both may still prove to be influential. 17:20
masak saying "Python is doing just fine" is a bit like saying "There is absolutely no cause for concern". :) 17:21
[Coke] masak: like my HR rep saying "This doesn't mean we're looking at a headcount reduction" 17:22
doy it is a nice contrast to the constant flow of "perl is dead"/"perl is not dead" articles from a couple years ago
(:
masak [Coke]: man, "headcount reduction" sounds like you wanted to make "firing people" sound like "chopping off people's heads". 17:23
doy: it's Python's turn not to be dead. :P 17:24
masak pre-empts all "pining for the fjords" jokes
not_gerd that's appropriate for Parrot as well...
Su-Shee well once there was tcl and mostly perl and now we share the realm with perl, python, ruby and recently javascript..
masak Tanenbaum* 17:25
pmichaud I'm getting module errors in the star tarball. 17:26
not_gerd popular languages don't really die, they just find their niche and continue in zombie-mode
even COBOL and Fortran are still around
pmichaud gist.github.com/3789352 # errors from "make modules-test" 17:27
pmichaud looks like failures in perl6-lwp-simple and panda 17:27
pmichaud release candidate (containing module errors) at pmichaud.com/sandbox/rakudo-star-20...rc1.tar.gz 17:29
masak &
pmichaud Looks like there are a couple of "Obsolete use of | or \ with sigil on param $" warnings in the rakudo-debugger, too. 17:32
not_gerd k-nucleotide from perl6-example/shootout still fails for me using a --gen-parrot Rakudo 17:48
not_gerd bye, #perl6 18:02
sorear good * #perl6 18:14
GlitchMr std: say join ' ', <*.jpg> 18:24
p6eval std 77327a4: OUTPUT«ok 00:00 43m␤»
GlitchMr std: eval { print 42 } 18:29
p6eval std 77327a4: OUTPUT«ok 00:00 42m␤» 18:30
dalek rlito: a185d00 | (Flavio S. Glock)++ | / (3 files):
Perlito5 - js3 - use exists() method
 18:43
masak I find Bret Victor's ideas fascinating, and I want to subscribe to his newsletter: worrydream.com/LearnableProgramming/ 18:47
PerlJam masak: you saw his light-table stuff? 18:56
masak I saw the talk where he applied his ideas to all kinds of things, if that's what you mean.
dalek rlito: b5576f2 | (Flavio S. Glock)++ | / (3 files):
Perlito5 - js3 - use fetchsize() method
 19:00
masak I'd like to bring discussion of github.com/perl6/specs/issues/24 onto the channel. 19:06
I've already expressed my opinion in the comments of that issue.
interested what others think.
moritz I'd like to mention that Bool is an Enum too 19:08
if all enum elements should be True in boolean contenxt, that would be very bad for False
masak heh :)
I agree. I didn't think of that case because I think it's more fundamental than that. 19:09
geekosaur I would think random enums should not have Bool instances. Bool itself should, obviously
masak geekosaur: sorry, what?
moritz that said, I can see why the current state of affairs could be seen as LTA 19:10
masak "have Bool instances"?
moritz maybe we should make enuma values start with 42 (or 1 or so)
FROGGS If someone starts a vote, I vote for the current (C-like) implementation, the first enum (if it is zero) should be treated as false
moritz: can I declare another start value like in C?
masak moritz: I think sirrobert is arguing for making his preferred use case (so = defined) the default.
moritz FROGGS: sure
sorear I have a gut feeling that the currnent impl is just a little bit too magical 19:11
masak moritz: (and I don't like that)
moritz r: enum A (b => 42, <c d e>); say A.enums.perl
p6eval rakudo 16f22b: OUTPUT«===SORRY!===␤Method 'returns' not found for invocant of class 'Parcel'␤»
FROGGS good, so somebody who wants to use 'so' to check for definedness should use enums starting with 1
moritz rakudobug
geekosaur hm, wait, this is perl6. I'm misremembering what else boolifies automatically though; if other types do, then enums probably should
moritz r: enum A (b => 42, 'c', 'd', 'e'); say A.enums.perl
p6eval rakudo 16f22b: OUTPUT«("b" => 42, "c" => 43, "d" => 44, "e" => 45).hash␤»
FROGGS very good
r: enum A (b => 42, 'c', 'd' => 77, 'e'); say A.enums.perl 19:12
p6eval rakudo 16f22b: OUTPUT«===SORRY!===␤Method 'returns' not found for invocant of class 'Pair'␤»
FROGGS r: enum A (b => 42, 'c', d => 77, 'e'); say A.enums.perl 19:12
p6eval rakudo 16f22b: OUTPUT«("b" => 42, "c" => 43, "d" => 77, "e" => 78).hash␤»
FROGGS good
masak moritz: shall I submit the above as a rakudobug? 19:13
r: enum A (b => 42, <c d e>)
p6eval rakudo 16f22b: OUTPUT«===SORRY!===␤Method 'returns' not found for invocant of class 'Parcel'␤»
masak looks like something that oughta work.
moritz masak: please submit
sorear masak: and it clearly should not fail like that
masak submits
moritz I think I know how to fix it, though jnthn might not like the fix. sorear will though :-)
sorear masak: I think it's trying to treat a parcel as QAST node 19:14
masak moritz: you have too much knowledge of the internals of our implementors :P
sorear is curious :-)
theoretically you should be able to do enum A < :b<42> c d e >
might need to use «» instead
masak yes, I think it does. 19:15
sorear n: say val(':b<42>')
p6eval niecza v22-6-g9e5350d: OUTPUT«:b<42>␤»
sorear n: say val(':b<42>').WHAT
p6eval niecza v22-6-g9e5350d: OUTPUT«Str()␤»
moritz masak: the fix is to call $*W.compile_time_evaluate on the expression, instead of trying to obtain the return values from the AST
masak moritz: sounds sane to me.
moritz and doing some kind of half-baked constant folding, as we do now
s/and/instead/
but last time I did that for some values, jnthn didn't like the approach 19:16
whereas sorear uses it in niecza for a variety of cases already
so, that's what my quasi-cryptic statement was about :-)
masak all noted in the rakudobug ;) 19:18
Tene worrydream.com/LearnableProgramming/ -- another great article from Bret Victor
masak Tene: I beat you to it ;) 19:18
Tene :D 19:19
masak <masak> I find Bret Victor's ideas fascinating, and I want to subscribe to his newsletter: worrydream.com/LearnableProgramming/
moritz don't beat a dead hor... oh wait
masak but yeah, Bret Victor gets dupes on the channel because he has really good points. 19:26
when I read this one (and sometimes with other things), I wonder how well his ideas would scale. 19:27
I guess that's the big challenge, and the reason we don't already have more of what he's proposing.
Tene That's exactly what I was thinking; these microexamples are nice, but I'm having trouble thinking of their applicability to any programming I've done recently. 19:30
masak there's another point which bothers me about it.
it's not scale so much as concreteness. 19:31
in order to get as good visualisations as he gets, you can't stray very far from being very concrete and un-generalized.
and sometimes you want to abstract over, say, a coordinate or a color.
his examples never do that, mostly because they can't. they're all loops and the like in mainline code. 19:32
Tene Although, the title *is* "Learnable Programming" 19:35
masak indeed. and the ideas are still interesting as such.
cognominal Perl 6 as a "natural" language is not optimized for learnability. So any thought about making an environment to make Perl 6 learnable is good. Even if learning a language is different from learning to programming. On the other hand, one needs a language to program. 19:39
moritz cognominal: Perl 6 is very much optimized for learnability
GlitchMr say 'Hello, world!';
It's easy to learn
Compare "Hello, world!" in COBOL or Java
You don't have to describe types. 19:40
You don't have to use MAIN sub
All you do is just type statement
And it just works.
Isn't it easy to use/
?
Tene "easy to use" is very different from "optimized for learnability" 19:41
cognominal I have not a TimToady quote ready but learnability comes second to expressivity.
Tene Perl 6 is optimized for people who know the language first 19:42
GlitchMr Yeah... surely if you know let's say Python, learning Perl 6 is easier 19:43
Tene The goal is to have a great tool; if you make sacrifices to optimize for someone's first few days using the language, they're going to keep paying that cost for the entire rest of their use of the language.
There are *also* plenty of considerations for new users, but experienced users tend to take priority, from what I've seen.
GlitchMr Why they would pay that cost for the entire rest of their use of that language?
And what those "sacrifices" are? 19:44
I surely don't think that *%*@!%(@!&*^(!^*!*@%* being syntax error would be harmful
std: *%*@!%(@!&*^(!^*!*@%*
p6eval std 77327a4: OUTPUT«===SORRY!===␤Confused at /tmp/lxUoaAXepe line 1:␤------> *%*⏏@!%(@!&*^(!^*!*@%*␤ expecting any of:␤ POST␤ infix or meta-infix␤ infixed function␤ postcircumfix␤ postfix␤ postfix_prefix_meta_operator␤ postop␤ statement
..modifier l…
Tene I don't have a reference for an actual description of this policy, it's just what I've seen from conversations here 19:45
GlitchMr: they're not directly opposing forces, just different priorities in optimization.
cognominal Tene: "experienced users tend to take priority". Agreed, but that's why we end up with PHP. People growing with a brain dead language for fear of learning what is there.
GlitchMr Tene: Perl 6 makes it easy to make good programs
When you know language, that is
sorear I am amazed by the variety of opinions on #24 19:46
GlitchMr Let's say you have BASIC background
He could even use gotos
And yes, Perl 6 allows him to use that rope
It's rarely neccessary, but it's there when you need it
sorear GlitchMr: perl 6 does not let you use numbers as line labels :D
GlitchMr perl6: #`(1) say 'this is line 1' 19:47
p6eval rakudo 16f22b, niecza v22-6-g9e5350d: OUTPUT«this is line 1␤»
GlitchMr It doesn't?
sorear um, that's not a label
n: 1: say "this is line 1"
p6eval niecza v22-6-g9e5350d: OUTPUT«===SORRY!===␤␤Illegal use of colon as invocant marker at /tmp/1CQTpU9NbT line 1:␤------> 1:⏏ say "this is line 1"␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1437 (die @ 5) ␤ …
sorear n: foo: say "this is line foo"; goto "foo"
GlitchMr n: _1: say "this is line 1"
p6eval niecza v22-6-g9e5350d: OUTPUT«(timeout)this is line foo␤this is line foo␤this is line foo␤this is line foo␤this is line foo␤this is line foo␤this is line foo␤this is line foo␤this is line foo␤this is line foo␤this is line foo␤this is line foo␤this is line foo␤this is line foo␤this is li… 19:48
niecza v22-6-g9e5350d: OUTPUT«this is line 1␤»
GlitchMr n: "1": say "this is line 1"
p6eval niecza v22-6-g9e5350d: OUTPUT«===SORRY!===␤␤Illegal use of colon as invocant marker at /tmp/hyRT_XmLvv line 1:␤------> "1":⏏ say "this is line 1"␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1437 (die @ 5) ␤…
masak cognominal: TimToady has said something to the effect of that we're not aiming to make Perl 6 people's first language, but people's last language :)
cognominal indeed 19:49
GlitchMr masak: After COBOL, Brainfuck, PHP 6 and Python 4 19:50
Am I right?
masak um. 19:51
GlitchMr It's "last language", isn't it
So you would have to learn Python 4 first
cognominal there is also the part about baby Perl. It is ok to talk baby Perl to begin with; evenif that limited knowledge does not help much to grasp "adult" programs 19:55
cognominal as much as I like haskell. There is no such things as baby haskell and monads for babies. 19:55
GlitchMr Prelude> (+) 3 4 19:56
7
I can do that in Haskell
I really don't know how to do more
sorear 3 + 4 -- :D 19:57
cognominal on the other hand, as a grown up, it is difficult to learn language that starts with different basics. There is much unlearning in learning. Or more exactly one must know what is relevant to the context of a different language. 20:00
[Coke] r: enum Days «:Mon(1) Tue Wed Thu Fri Sat Sun »; say +Tue;
p6eval rakudo 16f22b: OUTPUT«1␤»
dalek href="https://glitchmr.github.com:">glitchmr.github.com: 9635772 | GlitchMr++ | _posts/2012-09-23-perl-6-changes-2012W38.md:
`cmp` can be used to compare hashes in Niecza
[Coke] would the enum issue on list go away if this syntax (from the specs) worked?
(and output 2 instead of 1) ?
moritz yes 20:00
it's just that rakudo doesn't «...» smart enough yet 20:01
what you can do is
r: enum Days (Mon => 1, 'Tue', 'Wed'); say Tue
p6eval rakudo 16f22b: OUTPUT«Tue␤»
moritz r: enum Days (Mon => 1, 'Tue', 'Wed'); say +Tue
p6eval rakudo 16f22b: OUTPUT«2␤»
sorear [Coke]: I would still think the behavior is weird, but honestly at this point I've lost interest in fighting it
masak sorear: please be more specific. 20:02
(even if you've lost interest in fighting it)
sorear masak: enums other than Bool with numeric value 0 being considered false feels excessively magical to me 20:03
[Coke] on the contrary, it feels entirely logical to me. 20:04
but then, I think of the enums as a stand in for a bunch of constants.
FROGGS C-like
moritz aye, that's my understanding too 20:05
and the stringification to the enum name is just a bit of extra magic to make it nicer
sorear my reasoning is that 80% of the time, my enums don't actually have semantically relevant ordering 20:06
enum PhaserType < INIT END BEGIN START CATCH CONTROL ... > 20:07
if $some-sub.phaser { ... } # oops, this should have been .defined
masak "doctor, it hurts when I do this"
I can't stop people from using the language without knowing how things boolify. 20:08
moritz having a different default value than 0 would solve that problem without adding special magic
masak it would. but the default should clearly be 0.
[Coke] yah. I see no real issue in making the default start value 1.
masak: why?
sorear it's a gratuitous sharp edge.
masak because everything else starts at 0 in Perl 6.
just like moritz pointed out in the Github issue. 20:09
[Coke] masak: when are you checking the numeric values of an enum?
moritz forcing people to provide a start valeu would also work
leont Why wouldn't they start with 0 but true?
masak moritz: I like that better.
[Coke] that also seems vaguely reasonable, leont.
masak leont: I think '0 but True' is basically considered a code smell at this point. 20:10
leont The numeric value shouldn't matter, even if it was randomly distributed
[Coke] r: say ?(0 but True)
moritz leont: because all uses of "but"... what masak said
p6eval rakudo 16f22b: OUTPUT«True␤»
moritz what about making the defualt value not numeric, but string?
masak '0 but True' is an idea which sounds great, if it's the year 2000, and no-one has tried it in a language :)
sorear doctor, it hurts when I step on the exposed nail in my door threshold
we can either: tell people to be more careful, or: fix the problem 20:11
moritz enum A <B C > # same as enum A (B => 'B', C => 'C');
[Coke] that's not horrible.
masak sorear: I see what you're saying. I just don't consider the 0 an exposed nail in my door threshold.
moritz: I could live with that, I think.
that would force people like me who like the numbers to declare them explicitly. 20:12
moritz and if you need a numeric enum, you can always make that happen by using a first value
masak and everyone else gets their truthy strings.
leont Hmmm, that may actually make sense too
sorear but now that I've explained my position, I'd rather live with the status quo than spend more energy debating a replacement
masak that works for me, too ;)
moritz added the default stringy suggestion to the github issue 20:16
masak moritz++ 20:19
I would advise patching the spec directly with the stringy suggestion... but I would like to hear if TimToady opines something first. 20:21
FROGGS moritz: I like it 20:27
skids Hrm. In 09-rc1 there is something different going on with Failure. For example if you include a file that has a "fail('foo')" in it, you get "get_attr_str() not implemented in class 'Failure'' 20:29
But if it is in the mainline file, it doesn't do that.
Oh. Also if you run 'my $a = fail("foo"); +$a;' using -e onthe commandline, it throws the failure, but the same in a .pm6 file does not. 20:31
skids Hrm... it seems to abort the run at the throw, but does not print anything. 20:33
masak 'night, #perl6 23:00

Powered by Cro and the Raku® Programming Language.

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

Thank you!