»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend! | Rakudo Star Released!
Set by diakopter on 6 September 2010.
TimToady but only if successful 00:00
flussence will that return false if someone stuck a >/dev/null on the command line?
TimToady no
flussence /dev/full? 00:01
TimToady that always succeeds
masonkramer perl6: my $rv = say("hi"); say $rv.WHAT; say $rv;
p6eval rakudo cae7f9: OUTPUT«hi␤Bool()␤Bool::True␤»
..pugs: OUTPUT«hi␤Bool␤1␤»
TimToady but yes, output to a full disk should fail.
masonkramer perl6: say("hi").WHAT
p6eval pugs, rakudo cae7f9: OUTPUT«hi␤»
flussence hm, it returns true there. Just tried.
TimToady or to a shutdown socket
that's weird 00:02
masonkramer so if I capture the return value of &say in a variable, I get that expected behavior.
if I do it immediately,
jah
sorear um, shouldn't I/O errors throw a (delayed) exception?
00:02 Eth4n joined
TimToady perl6: "0".say.WHAT.say 00:03
p6eval rakudo cae7f9: OUTPUT«0␤Bool()␤»
..pugs: OUTPUT«0␤Bool␤»
TimToady perl6: "42".say.WHAT.say
p6eval rakudo cae7f9: OUTPUT«42␤Bool()␤»
..pugs: OUTPUT«42␤Bool␤»
TimToady oh, it's probably the listop bug
oh, no, it's my brane 00:04
duh
masonkramer yeah
say("hi").WHAT.say
works fine
neeeeevermind
is there a special case in the perl6 repl that suppresses printing of the return value of say? 00:05
TimToady sorear: it should fail with an I/O error, yes, but not necessarily throw it
just like any other failure
sorear: did you see my speculation about @! the other day?
and calling CATCH for each error in @! 00:06
masonkramer: dunno
sorear TimToady: no 00:11
TimToady: I don't really have a deep understanding of the failures system
TimToady the idea is to replace $!.pending with @!, more or less, and treat all unhandled exceptions equally, other than feeding them to CATCH in one order or the other 00:13
00:15 nymacro joined
sorear Mostly I think seeing how p6exceptions are used in practice would help 00:16
TimToady irclog.perlgeek.de/perl6/2010-12-18#i_3102010 and following
sorear but I probably need to implement them before that can happen :) 00:17
thanks.
TimToady which was, in fact, a response to your irclog.perlgeek.de/perl6/2010-12-18#i_3100836
my ruminations are scattered over about an hour 00:19
00:19 yarrak left, yarrak joined, yarrak left 00:20 yarrak joined, yarrak left, yarrak joined 00:21 icwiener left 00:22 araujo left, coldhead joined
sorear TimToady: when a function returns, does @! become a thrown exception, or is it appended to the caller's @! ? 00:22
00:23 yarrak left, yarrak joined 00:24 risou left
TimToady I suspect it should throw if it reaches the end of a function with non-empty @! 00:24
(after allowing CATCH to weed it)
00:25 yarrak left, yarrak joined
TimToady but perhaps the user could pick the other behavior somehow if wanted 00:25
fail if @! or some such
colomon clojure.blip.tv/file/4503162/ -- wondering how to do this in p6. But brainpower very low at the moment...
TimToady (which would presumably have to be done in a LEAVE) 00:26
yes, LEAVE { @!.fail if @! } or some such could probably be made to propagate to the outer @!, if default is to throw 00:28
since LEAVE is outside of CATCH
or use UNDO { fail } 00:29
I suspect any fail should also copy out any @! to the outer @!
fail should not hide inner exceptions; only CATCH should be able to handle exceptions 00:31
I think...
sorear TimToady: how does 'try' interact with @! ? 00:32
TimToady try that supplies a default CATCH would handle all the @! 00:33
it would just handle all the exceptions one at a time now, instead of throwing away $! and its $!.pending 00:34
00:34 soroush left, yarrak left
sorear TimToady: how does try communicate @! to the surrounding code? 00:35
00:35 soroush joined
TimToady try doesn't have to do that 00:35
try delegates all that to the CATCH and its block
try with an explicit CATCH is a no-op
sorear there's a disquieting amount of redundant code between niecza's implementations of 'class', 'role', and parameteric role 00:36
TimToady: so try { } with no CATCH throws the exceptions completely out and they can't be accessed?
TimToady correct
hmm, or not
I think we're confusing two different uses of @! here... 00:37
sorear Can I tell if try { } threw an exception?
Tene I thoguht that try set its exception as handled, and then pushed it in $!
TimToady Tene: we're talking about a slightly different model that gets rid of .pending (in response to some of your questions of Dec 18th) 00:39
Tene nods. 00:40
TimToady but sorear++ is correct that we need to detangle "last handle exception(s)" from in-flight exceptions (either unthrown or thrown)
*handled
Tene Yes.
TimToady the only time $! can really mean an in-flight exception is inside of CATCH (and maybe in LEAVE-like blocks) 00:41
perigrin sorear: there is a line of thought that a class is simply a (set of) role(s) applied to the empty class
TimToady but such blocks are also normal code that might want to refer to a handled exception, so we can't just overload $! for that 00:42
whoops, holiday visitors, will think on this as we drive to Seattle tomorrow...
colomon have a nice holiday visit! 00:43
Tene Have nice holiday visitors, too! 00:44
lue afk
Tene try could return the @! of its block.
try {} == { ...; CATCH { @!.handled; return @! } } 00:45
approximately
sorear Tene: what does it return to?
Tene my @success = try { ... } 00:46
erm, @fail
colomon rakudo: say +({ Any } ... *.defined)
p6eval rakudo cae7f9: OUTPUT«0␤»
colomon stupid, stupid programmer
rakudo: say !Any 00:47
p6eval rakudo cae7f9: OUTPUT«Bool::True␤»
colomon stupid, stupid, stupid programmer
00:49 lichtkind joined 00:52 Trashlord joined 00:59 araujo joined, araujo left, araujo joined, QinGW joined 01:01 redicaps joined
sorear 220, wow 01:04
we're even higher now than we were on the cusp of Star
perigrin 7 more and you beat #moose 01:08
01:09 cdarroch left 01:10 Trashlord left
colomon making sure you deal with the marginal cases in p2 is a brain-bender. 01:16
01:16 masonkramer left
sorear p2? 01:16
01:16 masonkramer joined
colomon or perhaps that's the vicodin talking. 01:16
masak's contest, problem 2.
sorear vicodin? what happened? D: 01:17
colomon surgery today, to fix an umbilical hernia.
I'm tired and a little sore, fuzzy and tired. 01:19
pretty sure I didn't mean to type tired twice there, either. :) 01:20
01:23 nymacro left
arnsholt colomon: If it helps, masak has said that points on the border of the polygon can be handled in an implementation-defined manner =) 01:23
colomon arnsholt: sure, the issue is how to robustly handle things.
arnsholt Heh. True
I think I've got a pretty simple and robust solution though 01:24
01:30 roen joined, roen left 01:31 roen joined, roen left 01:32 roen joined, roen left 01:34 roen joined 01:41 araujo left 02:03 araujo joined 02:07 noganex left 02:09 noganex joined 02:27 ggoebel left 02:32 ggoebel joined 02:38 molaf left 02:40 ggoebel left 02:45 lichtkind left 02:46 agentzh joined 02:51 molaf joined 03:01 mtk left 03:18 GinoMan joined 03:22 araujo left 03:32 BaggioKwok joined 03:37 yuki joined 03:38 BaggioKwok left
GinoMan can't wait for perl6 03:39
03:39 BaggioKwok joined 03:40 BaggioKwok left
kthakore p6eval: say ' GinoMan it's already here ' 03:42
flussence rakudo: say 'a' Zx 1..10 03:43
p6eval rakudo cae7f9: OUTPUT«a␤»
flussence rakudo: say 'a' Xx 1..10
p6eval rakudo cae7f9: OUTPUT«aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa␤»
03:45 yuki left 03:47 yarrak joined
sorear So many new people! 03:47
GinoMan Ok, I mean, stableizes to where someone can write a book about it, it's used in production, and it's stable 03:54
kthakore GinoMan: there is a perl6 book.
diakopter there was a book being written 5-6 years ago by brian d foy & merlyn
kthakore diakopter: Hmm my impression was perl6 has changed alot since then? 03:55
diakopter true
they were updating it heavily 2-3 years ago, but it's on hiatus again I think 03:56
dalek rixel: d464ada | diakopter++ | / (10 files):
YAWIPC; I'm too good to write descriptive commit messages (joking of course). The real reason is that I'm not good at coding in small increments; sorry!
 03:58
GinoMan or maybe I should put it this way, "I can't wait until I can put #!/usr/bin/perl at the top of my script and it's perl 6 on most systems" 04:03
what's the perl6 book? 04:05
sorear That will never happen. /usr/bin/perl6 is incompatible with Perl 5 and will stay that way
As far as being available on most systems - I don't think that will happen, either 04:06
It was a historical accident that all the major Linux distributions wound up depending on Perl in the base system
(also, many distributions are forking perl these days. If you want a Real Perl, you need to tell users to install it.) 04:07
diakopter sorear: re "That will never happen. /usr/bin/perl6 is incompatible with Perl 5 and will stay that way" - that's wrong 04:10
sorear Any program installed as /usr/bin/perl MUST interpret Perl 5 by default 04:11
diakopter sorear: according to the spec, /usr/bin/perl6's default behavior is (near-)perfect Perl 5 compatibility
sorear It may be able to switch into Perl 6 mode
allbery_b didn't the spec have to change on that though?
sorear But that won't be the default, because it's incompatible
diakopter ccording to the spec, /usr/bin/perl6's default behavior is (near-)perfect Perl 5 compatibility
allbery_b wa originally "use v6;" but that caused some "time travel" issues
diakopter according to the spec, /usr/bin/perl6's default behavior is (near-)perfect Perl 5 compatibility. to switch to Perl 6 mode, 'use v6;' is near the top, or 'package something... '
in S01 04:12
allbery_b I thionk you need to check the latest spec. I seem to recall that got killed because it caused some serious issues with (1) having to restart parse (2) inability to use perl6 -e sanely (3) some other similar warts
diakopter The perl interpreter will assume that it is being fed Perl 5 code unless the code starts with a "class" or "module" keyword, or you specifically tell it you're running Perl 6 code in some other way, such as by:
^^ in S01
today
perlcabal.org/syn/S01.html 04:13
sorear The "perl" interpreter
diakopter referring to a Perl 6 compiler/interpreter 04:14
otherwise the first part of that bullet "Migration is important." wouldn't make sense 04:18
lue So does a Perl 6 program run in P5 mode by default? [or P5 compatible?] 04:19
kthakore lue: Perl 6 and perl 5 have different interperter and are considered different languages
lue I know. /me looks at S01 to get a better idea of the discussion 04:20
[S01] from what I'm getting, A Perl 6 interpreter *does not* take the input as P6 until told so. 04:23
diakopter which agrees with discussion here on #perl6 on many many occasions over the past few years 04:25
allbery_b: what do you think got killed? 04:26
04:26 soroush_ joined, soroush left
lue I'm someone who has never touched Perl 5 [ well, I tried it for 5 minutes, then decided P6 was better :) ] : Why would a Perl*6* interpreter want to assume Perl*5* by _default_ ? 04:26
sorear A *Perl* interpreter must continue to take input as Perl 5 04:27
GinoMan so then... the book on Perl6
?
kthakore GinoMan: ah sorry
sorear Rakudo can get away with installing as 'perl6'
GinoMan lue: lots of old code using 5.10
allbery_b I recall a big fight on the mailing list about how it was wrong to require people to 'use v6;' in one-liners, but also wrong to make -e (or however *that* is now specced) behave differently from regular perl
diakopter sorear: it's referring to a Perl 6 interpreter/compiler
GinoMan the latest books on perl all discuss the current version to learn as perl 5.10
allbery_b 5.12 came out pretty recently
"latest" for print is ... relative 04:28
GinoMan true
they all mention perl 6 but don't cover it in depth
as far as I can tell
ok.... to bed i go
lue goodnight GinoMan o/ 04:29
GinoMan kthakore: did you find the title of that book?
goodnight lue
kthakore yeah
GinoMan: sorry
internet did
died*
GinoMan that's ok
I'll look it up later 04:30
gnight everyone!
04:31 GinoMan is now known as GinoSleeps
lue This is the first time I heard of Perl6 given this idiosyncracy [Although it was mentioned a couple days ago]. I think it's pretty... unexpected, to say the least. 04:32
kthakore github.com/perl6/book GinoSleeps see download after sorry
diakopter lue: it originates from the emphasis on migration, as that bullet says 04:34
and it generally seemed achievable since pugs was able to operate in that mode 04:36
04:38 _twitch joined
lue I know. And I'm not saying it's a bad thing, it's just unexpected. 04:38
04:39 Eth4n left, Eth4n joined
lue I haven't used P5, so to me making it run P5 by default isn't sensible. Although I do see why someone else would need it. 04:40
diakopter :) like the millions of P5 users and billions of [copies of] P5 scripts :D ... 04:41
but such an ideal scenario does seem much less achievable in light of the fact that Perl 5 is such a moving target (especially the past 4-5 years) 04:42
lue lemme see the update history of S01... 04:43
04:43 [Coke] left
diakopter on that note, here's yet another p6 book: svn.perl.org/viewvc/perl6/doc/trunk.../tutorial/ 04:45
lue I don't think that line about migration has been updated in a long time (of course the history only goes back to 26 Nov 2008)
diakopter but here's the S01 history from 2006-pugscode svn.perl.org/viewvc/perl6/doc/trunk...d?view=log
lue thank you, diakopter /me looking 04:46
diakopter 2005, I mean
lue diakopter: were you around in the days of pugs? 04:50
diakopter variously 04:51
04:52 orafu left
lue Would you say that the view of Perl 6 in relation to Perl 5 has changed over the years? I'm theorizing that view influenced the bit about P5-by-default. 04:53
diakopter lue: yes. 04:54
04:54 orafu joined
lue My idea is that, if way back when P6 was considered a normal successor to P5, like P5 to P4 and so on, this sort of default would make sense. 04:55
diakopter yeah 04:56
lue But now that P6 is parallel with P5, I'm not so sure that's a sensible option anymore. 04:57
diakopter that expectation has kinda morphed into a both forking ("different/separate but equal-ish languages") and interoperating ("linkable/usable from each other")
lue begins creating a detailed blog post. 05:00
allbery_b it's not clear that parallel is the right word; rather, perl 5 is steadily migrating toward perl 6 05:02
starting with 5.10
diakopter there are many facets to the discussion around that assertion 05:03
allbery_b I imagine so
especially given that p6 i still in flux
diakopter I think no one would dispute that the 'use v6' requirement is quite optimistic, but it begs the question, how optimistic in comparison to the other ideals of the language 05:05
05:07 [Coke] joined
diakopter lue: particularly relevant is the second-to-last bullet of S01 05:10
though I might posit that the 'use v6' feature is a higher-order feature 05:11
05:15 estrabd_xmas left
lue I've always see use v6; as a way for a P5 interpreter to tell the user the code he tried running is actually P6. 05:17
diakopter pugs' embedded/bridged p5 was quite useful, and I think its model of symbol/namespace-sharing/translation (and lexical scope-sharing/translation) is a good one that ought to be considered strongly by other implementations wishing to achieve the similar effect pugs could.
lue: yep, it does that too (perhaps merely conveniently, now) 05:18
lue Even though I don't see the need for embedding P5 code into P6 code (again, I have never used P5), I see that being useful for others. I just don't think it should parse as P5 by default. 05:20
diakopter lue: true, I agree with you, if it's installed as "/prefix/bin/perl6". but the Aug 2004 view was that it would be installed as "/prefix/bin/perl" 05:21
so perhaps a clarification/reduction to S01 would be to specify exactly that, since it resolves those difficulties 05:22
(as sorear mentioned above; rakudo can get away with installing as perl6)
lue I'm wondering: would it be wise for a P6 installation to install as `perl' nowadays? 05:24
diakopter right, that's what I'm proposing (as a spec clarification)
TimToady it's already clarified by "or you specifically tell it you're running Perl 6 code some ohter way" 05:26
diakopter "If an implementation doesn't implement the 'p5-as-default' mode, then it shouldn't be installed to /prefix/bin/perl, but should instead be installed to /prefix/bin/rakudo (and/or /prefix/bin/perl6 etc)
TimToady naming your executable perl6 is one way
diakopter oh
TimToady any executable named "perl6" is allowed to assume Perl 6; likewise any file with a .p6 or .pm6 extension is assumed Perl 6 05:28
05:28 cotto joined
TimToady that was the phrase we added in after the initial formulation 05:28
when it became more obvious that not all implementations would be interested in "replacing" Pelr 5
*rl 05:29
but yes, the original assumption was that we'd want /usr/bin/perl to be able to run either 05:30
that seems less likely since p5 development picked up again
so everyone's right :) 05:31
lue Another interesting blast from the past for me. This is better than government-sanctioned history class! :) 05:33
wait, when P6 was created (around 2000 IIRC), was P5 at a low point of activity? 05:34
TimToady it was perceived to be stagnating 05:35
lue Doesn't look like it today :) 05:36
TimToady and the perl5-porters mailing list was extrememly, um, feisty in the period before that
diakopter Is the general opinion that if a language isn't adding (breaking?) features somewhat frequently, then it's stagnating? 05:37
lue
.oO(feisty as in "excited, happy-to-code-Perl-5", or as in "nervous about the future" ?)
TimToady feisty as in fight over new features that nobody sane wanted to add to Perl 5 05:38
diakopter though Perl 5 was already essentially infitely extensible, so it really couldn't "un stagnate" unless one counted the enormous (but largely hidden because the curve of the earth isn't visible from the surface) growth of CPAN wasn't visible
infinit*ely 05:39
I guess my point is that languages generally have to make new paradigms available *by default* in the default installation.. and/or make backward-incompatible changes ... in order to be perceived as non-stagnating 05:42
that's my perspective, anyway. I'd be glad to be corrected. 05:43
TimToady cpan was also part of the stress on the core language, especially as source filters were proving to be non-scalable 05:47
lue source filters? 05:53
sorear source filters were the first attempt to add macros to Perl 5 05:54
they proved a failure as originally conceived
VXZ speaking of failure...I just almost burnt my apartment down 06:31
diakopter eek 06:34
VXZ it's all good; at least the annoying smoke detector doesn't have a battery in it 06:35
diakopter dukeleto excess flood? 06:42
06:42 dukeleto joined
sorear diakopter: no doubt secondary to a network hiccup 06:43
sometimes they cause an hour's worth of PONG messages to arrive at the server all at once, or something like that
VXZ question about 'use v5'...does that allow me to use the same variables between perl 6 and perl 5 code? What does it do with things that are represented differently between the two, like multidimensional data structures? And what about lazily evaluated things like 0..Inf? 06:45
sorear yes, undecided, undecided 06:49
VXZ Is any of use v5 implimented right now? 06:51
06:51 kaare joined, kaare is now known as Guest94684
sorear yes but not in Rakudo 06:51
VXZ Hm. What in? Besides p5 :P 06:53
TimToady pugs: use v5; print ord($"),"\n"; 06:54
p6eval pugs: OUTPUT«Error eval perl5: "sub { use ops (':default', 'binmode', 'entereval');; print ord($"),"\n";␤␤}"␤*** 'print' trapped by operation mask at (eval 2) line 1.␤␤Undefined subroutine &main:: called.␤»
lue Is there a way to link to a specific version of a file in github (e.g. S01 as it stands right now), for future-proofing purposes? 06:55
VXZ Thanks guys
lue Nevermind, I found out. 07:00
07:00 molaf left 07:02 molaf joined
dalek ecza: 0e836f6 | sorear++ | / (3 files):
Refactor handling of attributes and methods in backend
 07:02
ecza: d4638e1 | sorear++ | / (10 files):
Implement attribute initializers
ecza: 25d6c5c | sorear++ | test (2 files):
Test mergeback
ecza: 710b329 | sorear++ | docs/announce.v1:
Describe today's changes for v1
07:07 yarrak left 07:19 sukria joined
dalek rixel: ce47b50 | diakopter++ | / (9 files):
4-function arithmetic operators, among other things. YAWIPC. next: eval STRING special form.
 07:27
07:33 sukria left 07:36 snearch joined, Layla joined
Layla helo! 07:37
07:45 LeifD joined 07:47 sukria joined 07:48 _twitch left
lue blog post! rdstar.wordpress.com/2010/12/28/th...the-specs/ 07:52
07:56 glow joined
sorear Layla: HI! 07:58
Hi.
Layla: What I meant yesterday was that nobody here can understand how to us variable lookbehind 07:59
Layla sorear: hey you remember me! ^_^ 08:04
sorear TimToady: Someone just private-mailed me to say that github.com/audreyt/pugs/tree/master/t/01-sanity is part of the official test suite and I should be using it instead of my home-grown file. True or false? 08:07
Layla sorear: You also told me that you are the implementors of the language, so I guess there is no variable length lookbehind in perl6 yet.. it is ok since many programming languages do not have it too..
sorear: can you tell me how who to know if a specific feature is implemented or not yet in a rakudo imp of perl6? 08:08
sorear: A link to some site would be great! :) 08:09
sorear We have the spectest fudges, although that's rather too low level for most people
There's also the t/spectest.data file in your Rakudo checkout, at a higher level 08:10
08:11 Trashlord joined
colomon rakudo.org/status 08:11
for a very high level overview, easy to read
lue goodnight #perl6 o/ [ whenever your night comes :) ] 08:14
sorear bye lue.
08:14 envi left
Layla colomn/ sorear: thanks I will check them. 08:15
colomon rakudo: say pi.unpolar(10) 08:19
p6eval rakudo cae7f9: OUTPUT«-2.63602095158294 + -1.7090927253678i␤»
colomon rakudo: say 10.unpolar(pi)
p6eval rakudo cae7f9: OUTPUT«-10 + 1.22464679914735e-15i␤»
colomon ah
rakudo: say (0 .. 2 * pi).roll 08:20
p6eval rakudo cae7f9: OUTPUT«1␤»
colomon rakudo: say (0 .. 2 * pi).rand
p6eval rakudo cae7f9: OUTPUT«3.08020446862091␤»
colomon still someone out of it, I fear.
*somewhat 08:21
08:25 justatheory left 08:26 araujo joined, araujo left, araujo joined
dalek ecza: dc2fc3b | sorear++ | / (2 files):
Fix INIT block test
 08:31
ecza: c2857c1 | sorear++ | / (2 files):
Allow passing initial attribute values to .new
08:36 Layla left
colomon rakudo: say Complex.new(1, 2) 08:40
p6eval rakudo cae7f9: OUTPUT«1 + 2i␤»
08:57 Layla_89 joined 09:00 Layla_89 left
colomon rakudo: sub foo($a) [ $a // return Any; }; say foo(10); say foo(Complex); 09:03
p6eval rakudo cae7f9: OUTPUT«===SORRY!===␤Confused at line 22, near "sub foo($a"␤»
colomon rakudo: sub foo($a) [ $a }; say foo(10); say foo(Complex); 09:04
p6eval rakudo cae7f9: OUTPUT«===SORRY!===␤Confused at line 22, near "sub foo($a"␤»
colomon rakudo: sub foo($a) { $a // return Any; }; say foo(10); say foo(Complex);
p6eval rakudo cae7f9: OUTPUT«10␤Any()␤»
colomon woah, that allows for some nice golfing 09:05
09:29 sftp_ joined, sftp left 09:33 adu joined
colomon rakudo: say Complex.new(0, 1/2) - 1 - 1i 09:37
p6eval rakudo cae7f9: OUTPUT«-1 + -0.5i␤»
colomon rakudo: say Complex.new(0, 1/2) - (1 - 1i) 09:38
p6eval rakudo cae7f9: OUTPUT«-1 + 1.5i␤»
colomon rakudo: say Complex.new(0, 1/2) + (1 - 1i)
p6eval rakudo cae7f9: OUTPUT«1 + -0.5i␤»
09:49 cognominal joined, snearch left 09:53 lao294 left 10:03 agentzh left 10:07 adu left 10:23 roen left 10:29 QinGW left
colomon masak: I'm starting to be happy with my p2 solution. Worked out a couple of subtle bugs. Currently have about 100 lines of code and 200 lines of tests. :) 10:41
moritz_ the actual decision logic is 38 lines here, including comments and blank lines 10:46
tests are 141 lines, including some ascii art visualizing the polygons 10:47
10:47 pmurias joined
pmurias why gives commit bits to the github perl6/roast repo? 10:49
moritz_ did you mean "who"?
sorear, pmichaud an me (for now) 10:50
pmurias yes
moritz_ since github doesn't have an API for organizations, we can't yet do it with hugme (which is the ultimate plan)
pmurias moritz_: could i have one please?
moritz_ sure
pmurias my github nick is pmurias
moritz_ done 10:51
pmurias thanks
moritz_ pmurias: you're now in the 'perl6' team, which means you have access to roast, specs, perl6.org, modules.perl6.org etc.
pmurias adding fuge markers for niecza is ok?
moritz_ sure
sorear hello pmurias 10:54
glad I'm staying up an extra hour today
10:55 snearch joined
pmurias hi 10:55
i'll try to get up ealier this week
sorear: 4**20 is causing an error in the CLRBackend 10:56
any use of the ** operator
sorear there is no ** operator 10:57
what you're seeing is a STD buglet which causes the parser to fail to reject undefined standard infix operators
this couldn't happen with a full Perl 6 setting 10:58
pmurias what was infix:<**> replaced with?
sorear it wasn't 10:59
there is no &infix:<**> in Niecza's setting
it's not that I've replaced it; I just haven't gotten around to adding it
pmurias i'll add it then
pmurias didn't notice the top of the huge error 11:00
sorear I don't think we really need a backtrace for that one 11:01
pmurias standard user errors should have one 11:03
sorear there are 4 different ways errors can get reported in niecza
1. Perl 6 exception handling - generates nice backtrace
2. CLRBackend throws an exception - generates usually useless backtrace 11:04
3. compiler Perl code throws an exception
4. compiler calls sorry()
2 and 3 are mostly assertions; I want to have all user errors caught in 1 and 4 but I'm not there yet 11:05
11:05 coldhead left, coldhead joined 11:08 glow left 11:09 agentzh joined 11:10 daxim joined, jhuni left
sorear moritz_: github.com/blog/758-new-organizations-api ? 11:14
dalek ecza: 711b4c6 | sorear++ | / (6 files):
Reimplement $?FOO in new metamodel
 11:23
ecza: 396dce8 | sorear++ | / (3 files):
Fix backtrace file names
moritz_ sorear: oooh, I didn't know that
sorear out
11:25 redicaps left 11:27 LeifD left 11:28 wamba joined, redicaps1 joined 11:30 agentzh left 11:43 LeifD joined
colomon pmurias++ # was just thinking it was probably time to start looming at spectest on niecza, even if sorear++ isn't focusing on it. 11:44
tadzik I think he is, just she eval seems to be a problem 11:45
11:46 redicaps joined 11:49 envi joined
dalek ast: e75b0e8 | pmurias++ | S32-num/power.t:
fudge power.t for niecza
 11:57
pmurias what should i do if fuding breaks the test plan? 12:03
12:03 coldhead left
moritz_ how would it do that? 12:08
the usual #?skip doesn't
ah, if you remove some tests with #?emit impl # 12:09
then you can also #?emit skip(...)
pmurias moritz_: i skipped a block containing a for loop which generates tests 12:16
moritz_ pmurias: then add a #?DOES $number_of_tests before the block 12:18
12:19 Guest94684 is now known as kaare_ 12:20 masak joined
masak oh hai, zebras. 12:20
rakudo: say 42 ?? 1, 2, 3 !! 5 12:21
p6eval rakudo cae7f9: OUTPUT«===SORRY!===␤Confused at line 22, near "say 42 ?? "␤»
masak std: say 42 ?? 1, 2, 3 !! 5
p6eval std 625303c: OUTPUT«===SORRY!===␤Precedence of , is too loose to use between ?? and !!; please use parens around inner expression at /tmp/6F00FgKzoH line 1:␤------> say 42 ?? 1,⏏ 2, 3 !! 5␤Parse failed␤FAILED 00:01 120m␤»
pmurias moritz_: why do we keep the plans, we have done_testing? 12:22
masak pmurias: there has been some back-and-forth on that in the spectest suite. 12:23
pmurias: the chief argument for keeping plan numbers seems to be that the statistics become more reliable. 12:24
moritz_ which IMHO is not a good point, since they don't mean anything
masak nevertheless, people like statistics. 12:25
and I can certainly see it from the point of view of those collecting the statistics.
pmurias how many fourth roots of 8 are there? 12:27
moritz_ 4
there are 4 fourth roots of any finite complex number 12:28
pmurias and 8i?
moritz_ rakkudo: print ~8i.roots(4)
rakudo: print ~8i.roots(4)
p6eval rakudo cae7f9: OUTPUT«1.55377397403004 + 0.643594252905583i -0.643594252905582 + 1.55377397403004i -1.55377397403004 + -0.643594252905582i 0.643594252905583 + -1.55377397403004i»
moritz_ en.wikipedia.org/wiki/Fundamental_t...of_algebra 12:29
masak \o/
moritz_ en.wikipedia.org/wiki/Properties_of...mial_roots is even more fitting 12:30
moritz_ marks the day in the calendar which is reserved for theoerm usage
abstrusegoose.com/133 12:31
*theorem
pmurias DOES sets affects all following things not just one 12:32
moritz_ DOES should affect a block
if not, you've found a bug in fudge
12:33 bluescreen joined
masak fudgebug. 12:33
12:33 bluescreen is now known as Guest54734
pmurias looking at the code there seems to be a bug there 12:34
should it be done_testing or done-testing? 12:35
tadzik _ 12:37
masak ...which is also against Perl 6 "practices", come to think of it. 12:38
but it's inherited from Perl 5 which only has underscores, of course.
pmurias where does the Test.pm test suit live nowdays? 12:39
moritz_ in each compiler's repo 12:41
wait
moritz_ misread the question 12:42
mu
t/02-test-pm/
12:42 am0c^ left
masak lue: nice post with nice pictures -- rdstar.wordpress.com/2010/12/28/th...the-specs/ 12:43
lue: the "tuning fork" metapohor was devised around this point: strangelyconsistent.org/blog/the-wi...-true-heir 12:44
pmurias lue: you are aware that the default only applies if the perl6 interpreter is invoked as 'perl'? 12:47
masak sorear++: congrats on the v1 announcement! 12:48
pmurias where was the v1 announced to? 12:52
moritz_ p6a
12:52 ggoebel joined
moritz_ planetsix 12:52
tadzik p6u 12:55
masak maybe a status report is in order: I've started using niecza to make a Markdown converter. it will take on the order of two weeks to build, I think. I'm running into quite a lot of minor issues, which is good. my goal is to eventually replace the Perl 5 Markdown converter with the Perl 6 one in my blog software psyde, and not loose much performance doing so. 13:07
that may require me to precompile the Markdown source code somehow, because currently compilation is slow in niecza.
blocker: I'm caught up in HPatMoR and unable to code. ;) 13:09
13:09 lidden joined
lidden Has Rakudo Star 2010.12 been cancelled? 13:11
masak hi. no, it's usually released a week after the montly Rakudo release. 13:12
and that one was... last Thursday.
masak checks that
yes.
dalek ast: f65933f | pmurias++ | S32-num/power.t:
add #?DOES to S32-num/power.t to fix fudging for niecza
 13:14
lidden Ok, Thanks. 13:15
13:16 masonkramer left, masonkramer joined 13:21 redicaps left
masak s/loose/lose/ # argh! 13:23
I hate it when I see that typo in print :(
VXZ I hate it when I see any typo in print
13:36 GinoSleeps left 13:38 fhelmberger joined, fhelmberger left 13:39 fhelmberger joined
shortcircuit I realize it's out of order, but with operators like &&=, &=, ||=, += and such being so commmon, wouldn't it make more sense to define a metaoperator to apply to &&, &, ||, +, etc. instead? 13:42
13:42 tomaw joined
colomon shortcircuit: that's generally what is done 13:43
13:43 roen joined
masak shortcircuit: what gave you the impression that they're not defined that way? 13:43
shortcircuit I just noticed the "New operators" line in the Niecza v1 announcement on planet perl six. 13:44
For new ?= operators to be defined, it would seem necessary that the general case is that they're not already.
Which is why I suggested the metaop. 13:45
colomon ah, I cannot speak for what Niecza does
masak I was also a bit surprised that && was defined in niecza but &&= wasn't. 13:48
colomon hmmm, where does STD.pm live these days?
to be far, I think &&= is a bit of a special case.
masak colomon: github.com/perl6/std
colomon because it shortcircuits 13:49
masak++
shortcircuit groans
masak right. && is a special case, because it short-circuits.
shortcircuit just realized he's not in #haskell any longer, because he hasn't been pinged by 'shortcircuiting', et al in a while.
13:50 smash joined
smash hello everyone 13:50
masak smash! \o/ 13:51
colomon \o
shortcircuit: anyway, it definitely is a meta-op in theory and in Rakudo, but I dunno what Niecza does with it. 14:01
rakudo: sub infix:<splat>($a, $b) { $a ~ " splat " ~ $b; }; say "a" splat "b"; my $a = "fred"; $a splat= "barny" 14:02
p6eval rakudo cae7f9: OUTPUT«a splat b␤Null PMC access in invoke()␤ in main program body at line 1␤»
colomon std: sub infix:<splat>($a, $b) { $a ~ " splat " ~ $b; }; say "a" splat "b"; my $a = "fred"; $a splat= "barny"
p6eval std 625303c: OUTPUT«ok 00:01 125m␤»
masak colomon: might help to make it an 'our' sub. 14:03
colomon masak: is that one known? seems like I might have stumbled across it six months ago...
rakudo: our sub infix:<splat>($a, $b) { $a ~ " splat " ~ $b; }; say "a" splat "b"; my $a = "fred"; $a splat= "barny"
masak it's known.
p6eval rakudo cae7f9: OUTPUT«a splat b␤»
colomon rakudo: our sub infix:<splat>($a, $b) { $a ~ " splat " ~ $b; }; say "a" splat "b"; my $a = "fred"; $a splat= "barny"; say $a
p6eval rakudo cae7f9: OUTPUT«a splat b␤fred splat barny␤»
colomon masak++
14:05 mtk joined 14:12 araujo left 14:13 adu joined 14:14 plainhao joined 14:18 araujo joined, araujo left, araujo joined
masak nom & 14:21
14:25 shi joined, adu left, masak left 14:31 MayDaniel joined 14:40 roen left 14:44 xinming joined 14:45 shi left 14:48 kaare_ left 14:49 fhelmberger left 14:50 vastak joined 14:51 roen joined 14:56 fhelmberger joined 14:57 sftp joined, sftp_ left 15:00 icwiener joined
colomon is starting to worry he's too fuzzy in the head to make progress on p1. 15:01
LeifD where is rakudos bug reports stored?
15:04 wamba left
colomon rt.perl.org/rt3/Public/?nossl=true 15:04
LeifD been there. The bugs seem to be so old?
[Coke] LeifD: some of them are, sure. 15:05
LeifD Is there a way to order them, newest to old
[Coke] If you are not a guest, you can click on "Created" header to sort by. 15:07
LeifD Im a guest. That solves it ;).
[Coke] You might be able to self-register for a bitcard account off the main page. 15:09
LeifD Already on it.
Works good now :) 15:15
colomon oooo, stack machine! 15:16
15:16 LeifD left 15:26 roen left 15:28 am0c left 15:29 jest joined
colomon 's Forth heritage still comes out every now and then... 15:30
15:54 thepler joined 16:00 _twitch joined 16:03 justatheory joined 16:09 vastak left, JimmyZ joined, vastak joined 16:10 Intensity joined
rjbs p1? 16:12
16:15 hercynium joined, risou joined
colomon masak's quiz, problem one 16:16
16:25 risou left 16:26 risou joined 16:30 JimmyZ left 16:35 WonTu joined 16:36 spq1 joined, WonTu left
moritz_ strangelyconsistent.org/blog/masaks...ng-contest 16:39
colomon by george, I think I've got it. 16:42
moritz_ »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! | tinyurl.com/p6contest 16:42
moritz_ colomon: who is george? 16:42
colomon I presume the phrase originally comes from one of the English kings. (which is to say, I have no clue.) 16:43
16:43 snearch left 16:44 JodaZ joined
colomon Seems like there are only three of us actually working on the contest, if we go by the chatter on #perl6 16:44
(and T-135 to #phasers?) 16:45
moritz_ won't make it to #phasers today
having breakfast with $rest_of_family and then a board game :-)
colomon what game? 16:46
colomon will be at #phasers unless he is napping.
colomon hopes moritz_ has better luck than he has with playing board games once he has a little one around... 16:48
moritz_ colomon: en.wikipedia.org/wiki/Carcassonne_%...rd_game%29
colomon Oh sure, we have Carcassonne
16:48 lichtkind joined
lichtkind rakudo: 4 and 5 16:49
p6eval rakudo cae7f9: ( no output )
lichtkind rakudo: say 4 and 5
p6eval rakudo cae7f9: OUTPUT«4␤»
[Coke] t/spec/S29-context/exit.rakudo
colomon Carcassonne is one of those lovely games that two can play. 16:50
[Coke] (fails 2nd test with latest parrot, even with whiteknight's fix)
looks like "exit" and "exit 5" are not handled the same in Rakudo, as whiteknight's fix makes them both exit quietly in parrot. 16:51
moritz_ colomon: we celebrate New Year's Eve with a family where the parent manage to invite people for playing board games, even though they have a toddler. So I do have hope :-)
[Coke] rakudo: say "hi" ; exit
p6eval rakudo cae7f9: OUTPUT«hi␤»
[Coke] rakudo: say "hi" ; exit 3 16:52
p6eval rakudo cae7f9: OUTPUT«hi␤»
[Coke] (works fine against parrot 2.11.0)
moritz_ it seems that the rakudo on p6eval is the last before the bump to 2.11.0
moritz_ curses at the bytecode incompatibilities
colomon moritz_: our guy now sleeps well enough in the evening that we will probably be able to have people over to play once or twice a month when the holiday season is over. 16:53
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....dex_tablet
moritz_ runs wipe-and-rebuild-rakudo.sh 16:54
[Coke]: does S02-magicals/pid.t pass for you?
16:56 kanishka joined 16:57 MayDaniel left
[Coke] moritz_: yes. 16:57
lichtkind rakudo: say 4 xor 5
p6eval rakudo : OUTPUT«sh: ../p/bin/perl6: not found␤»
lichtkind rakudo: say 0 xor 5
p6eval rakudo : OUTPUT«sh: ../p/bin/perl6: not found␤»
lichtkind rakudo: say 5 16:58
p6eval rakudo : OUTPUT«sh: ../p/bin/perl6: not found␤»
[Coke] uhoh.
moritz_ lichtkind: the rakudo on p6eval is currently being rebuilt
sorry for the inconvenience
lichtkind k
its ok
moritz_: i have fresh os here doesnt came around to install rakudo 16:59
flussence I was gonna suggest using try.rakudo... but it looks like I'll have to fix that again now :(
huh, screen session's gone 17:00
oh, box was rebooted... 17:01
hm... 17:02
→ say $*PERL
namerakudo
version2010.09-3-ga820a44
seems a bit stale...
shortcircuit lichtkind: Installing rakudo's pretty easy. Finding the directions I used seems to be hard, though. 17:03
17:04 soroush_ left, redicaps1 left
shortcircuit github.com/rakudo/rakudo/ 17:04
Looks like the inlined README is the instruction set I used. I went with the latest release from github.com/rakudo/star/downloads , though 17:05
flussence seen ash_
aloha ash_ was last seen in #perl6 25 days 21 hours ago saying "sorear: i think you have to evaluate a constant at compile time".
flussence aw, jeez.
moritz_ rakudo: say $*PERL
p6eval rakudo : OUTPUT«name rakudo␤version 2010.12-3-g3750262␤␤»
flussence tryrakudo's using a perl6 in his user dir, I'd install a newer one myself for the site but I dunno how well that'd go on feather3; it seems a bit tight on RAM 17:06
moritz_ flussence: I think he copied a compiled perl6.pbc from his own machine, or something
flussence hm, that could work... 17:07
well, site's functioning now anyway. 17:09
(or it would be, if I'd noticed there were two separate source trees to update. argh!) 17:16
daxim mascot page: duke.kenai.com/
17:18 soroush joined
dalek ok: 7be2120 | (Przemysław Wesołek)++ | lib/ (2 files):
Nicely wrap long text in tables, due to tabulary.

Closes #41		 17:25
moritz_ the book, the book
we have to finish it
jest do it, do it, so I can read it ;) 17:30
17:32 kanishka left 17:35 Eth4n left, Eth4n joined, drbean joined
jest I can help with some LaTeX knowledge and building, but I'm a Perl 6 newbie 17:37
dalek ok: a9caeb2 | moritz++ | src/images/title.svg:
add vector graphic version of the cover image
 17:41
17:41 lue joined
moritz_ jest: using this .svg instead of the .jpeg for the title would be nice 17:41
needs to be converted to pdf first, I believe 17:42
moritz_ decides to put the authors in alphabetical order 17:43
jest sure, I'll convert it to PDF and add the file to the repo. 17:44
lichtkind rakudo: say 5
p6eval rakudo : OUTPUT«5␤»
jest I don't see the required fonts in the system. Have to look for them on the web
lichtkind rakudo: say 0 xor 5 17:45
p6eval rakudo : OUTPUT«0␤»
lichtkind rakudo: say 2 xor 5
p6eval rakudo : OUTPUT«2␤»
moritz_ jest: I think i have the fonts, I can do the conversion too
jest (I mean the ones used in SVG, generated PDF won't need them)
17:45 cdarroch joined, cdarroch left, cdarroch joined
jest hm, Avenir Next Pro, by Linotype; doesn't look like a free face... 17:47
moritz_ not free, no 17:48
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....dex_tablet 17:49
jest moritz_: go ahead, if you can. Exporting to PDF (from Inkscape?) please turn texts into paths, and maybe rasterize filter effects
moritz_ inkscape is sooo slow to start up
17:50 espadrine joined 17:51 kanishka joined
dalek ok: 3e3427b | moritz++ | src/images/title.pdf:
add pdf-ized cover image
 17:51
17:51 kanishka left 17:53 risou left
tadzik ~~ 17:54
jest moritz_: how about inkscape src/images/title.svg -D -T -A src/images/title.pdf, for me it runs pretty fast (preloaded, though)
moritz_ jest: that's what I just committed... took felt 10s
jest _moritz: the text in PDF doesn't look like Avenir
tadzik carcassone++
moritz_ is pretty ignorant to font questions :( 17:55
seems I only have Avenir LT Std and not Avenir Next Pro 17:56
jest moritz_: are you sure in inkscape the SVG looks exactly like (fonts...) the frontpage of the book?
moritz_ jest: not sure, no
jest moritz_: OK, I'll change the code in LaTeX to use PDF, which may be corrected later. 17:57
shortcircuit is stuck in a holding pattern at work. Could compare screenshots and do other mindless things.
lichtkind shortcircuit: do you know dep repo that has rakudo?
debian
shortcircuit lichtkind: No, I don't. Running Lenny at home. I used the linked instructions and the latest tarball to install under my user directory. 17:58
huf there was an ancient version in debian experimental if you feel like using an old rakudo with an old parrot AND potentially breaking your debian while doing so :) 17:59
17:59 espadrine left
diakopter moritz_++ github.com/perl6/book/raw/3e3427b8.../title.pdf 18:00
I guess the orange moth signifies a Camelia that hasn't blossomed into her full rainbow-colored Egyptian coat 18:01
moritz_ diakopter: sri++ came up with it, I merely did some conversion
dalek ok: 51c28cd | moritz++ | lib/book.sty:
alphabetic order of contributors
 18:03
moritz_ diakopter: it's certainly an allusion to camelia, and at the same time less controversial 18:04
shortcircuit I'd try put together a .deb using checkinstall, but 'make install' doesn't place things in, e.g. /usr/local/*, so I don't know quite how it'd behave. (I also don't know that much about changing Rakudo's installer.) 18:05
diakopter to follow the Joseph (in Genesis) analogy, I guess she's still ruled/persecuted by her older siblings
shortcircuit *try to
moritz_ shortcircuit: rakudo is installed to the same prefix as parrot
shortcircuit: so once you have a good parrot package
it's a piece of cake 18:06
shortcircuit SSHes home and pokes around
How old is Parrot 2.0.0-1? 18:07
diakopter mummified?
shortcircuit apt-cache show indicates [email@hidden.address] for the Maintainer of the Parrot package. *checks mailing list archives for undead behavior* 18:09
18:10 M_o_C joined
shortcircuit is reminded of the parrot sketch from Monty Python. 18:11
It may apply to that list.
Wow.
dalek rixel: 1572a27 | diakopter++ | / (7 files):
make the recursive operator precedence actually work correctly
flussence debian's effort looks a fair bit more lively than gentoo's atm 18:12
(if I get sufficiently irritated, I might do something about that at some point)
[Coke] 2.11 just came out, monthly releases. 18:18
18:23 masak joined
masak ahoy! 18:23
tadzik ahoy! 18:24
masak sits down writing a bit of Markdown converter in Perl 6 18:25
shortcircuit I'm curious. How many Perl 6 developers in here are paid to work on implementations of Perl 6? (Either explicitly so, or on something that they can treat such work as a dependency) 18:27
18:29 LeifD joined 18:33 kanishka joined
[Coke] several people are receiving grants. 18:35
but I wouldn't expect those to be == a salary or anything.
pretty sure everyone has a day job that takes precedence. 18:36
lichtkind shortcircuit: , huf thanks
masak the grants aren't salary-level, but they're not insignificant either.
[Coke] colomon: holy crap, sir.
huf lichtkind: for what? i *hope* you didnt install what i suggested ;)
masak I should have made this one of the problems in the Perl 6 Coding Contest: en.wikipedia.org/wiki/Firing_squad_...on_problem :) 18:38
shortcircuit masak++....Now I need to figure out how to work that into a Rosetta Code task. 18:41
masak ooh
shouldn't be too difficult. :)
shortcircuit masak: Although...you're welcome to get sucked into the glorious timesink that is RC, if you like. :) 18:42
masak I'll certainly think about it :)
though I've created quite a number of personal glorious timesinks already, and I tend to have them take precedence :)
masak hacks a bit more on Markdown 18:43
shortcircuit First hit's free. All subsequent ones, too. :)
masak Free in the Linux sense, "only if your time isn't worth anything" :P
man, using niecza is worth it just for the heredocs... 18:44
18:44 MayDaniel joined
dalek ok: be7ea6f | (Przemysław Wesołek)++ | lib/book.sty:
Switched to PDF version of the logo.
 18:45
ok: f3ea109 | (Przemysław Wesołek)++ | src/images/title.pdf:
PDF logo version more resembling previous one.
18:46 jest left 18:47 shi joined 18:51 MayDaniel left
Util In Perl 6, can you think of any circumstance that this: 18:51
if some-multi-term-expression {...}
would differ from this?: 18:52
if ?(some-multi-term-expression) {...}
In other words, is ?() redundant in *every* if-conditional-clause? 18:54
flussence rakudo: my $x = 1 but role { method Bool { !!! } }; if $x { say 'a' }; if ?$x { say 'b' }
p6eval rakudo : OUTPUT«a␤Stub code executed␤ in 'Bool' at line 22:/tmp/4IvEvrnu8I␤ in 'prefix:<?>' at line 468:CORE.setting␤ in main program body at line 22:/tmp/4IvEvrnu8I␤»
flussence not sure what the difference is there, but there is one. 18:55
masak Util: if the .Bool method is overridden, boolifying twice might have arbitrary consequences.
pmurias masak: any feature missing in niecza you would like me to add? 18:58
Util Hmmm. Since this is in regards to P5-to-P6 translation, .Bool and roles should not be factors. Thx!
18:58 Eth4n left
masak pmurias: do you think you could put in multi dispatch? :P 18:58
shortcircuit likes multi dispatch 18:59
pmurias i could move the one i did for mildew over
but i doubt it would be very fast
masak fast is a later concern for multis.
shortcircuit notes that as he was writing his first Perl 6 programs last week, almost all of his subs became multis. 19:00
masak sorear, pmurias: in my Markdown code, I've copy-pasted Test.pm6 and modified it so that it prints "got" and "expected" information 19:02
niecza: say "foo\n".perl
p6eval niecza 406e042: OUTPUT«Can't locate JSON.pm in @INC (@INC contains: /home/p6eval/niecza/src /etc/perl /usr/local/lib/perl/5.10.1 /usr/local/share/perl/5.10.1 /usr/lib/perl5 /usr/share/perl5 /usr/lib/perl/5.10 /usr/share/perl/5.10 /usr/local/lib/site_perl .) at /home/p6eval/niecza/src/NAMBackend.pm
..line 9…
masak hm.
"Unable to resolve method perl in class Str" locally.
pmurias: if you want, you could put in .perl for the CORE classes. 19:03
pmurias masak: got and expected should be merged in
masak they'd be even better with .perl, though.
pmurias testing .perl properly requires eval :( 19:06
masak I haven't felt a pressing need for &eval yet. but I suspect I will, any day now.
pmurias eval doesn't tend to be very usefull in everyday programming 19:08
masak no, but it's wonderful in the workaround toolset.
shortcircuit Isn't it needed for REPL?
masak not really. 19:10
just compilation.
lichtkind huf for the info
masak alternately, "only if you're writing the REPL in the implementation language" :)
shortcircuit masak: I believe RC has a task for that. :)
pmurias REPL's usually need some extra magic so that lexicals persist between calls
pmurias uses REPL mostly as calculators 19:11
masak people: is there a more efficient way to write C<< [<!before \n\n> .]+ >> ?
sorear [ \N+ || \n <!before \n> ]+ maybe? 19:12
masak nice, thanks.
19:12 Chillance joined
pmurias sorear: hi 19:14
sorear: should i rename .dump to .perl?
with fixes to make it return perl code
sorear shortcircuit: I have a metaop which can make but=, +=, -=, ~= etc, but && is a special case as remarked 19:16
shortcircuit: currently $x ~= $y turns into assignop(&infix:<~>, $x, $y) - can't make that work with && easily 19:17
shortcircuit: esp. since there is no &infix:<&&>
19:18 kaare joined, kaare is now known as Guest55231 19:19 Guest55231 is now known as kaare_
sorear pmurias: yes 19:21
masak rakudo: grammar G { token TOP { [<para> \n*]* }; token para { [ \N+ || \n <!before \n> ]+ } }; for G.parse("foo\nbar\n\nbaz")<para> { say "|", $_ }
sorear I think I lost that battle
p6eval rakudo : OUTPUT«|foo␤bar baz␤»
masak sorear: you might want to look into why the above produces a different output in niecza.
pmurias sorear: which battle?
sorear pmurias: trying to get .perl to stop being called that
masak sorear: it says "|bar" rather than "bar".
sorear ow. 19:22
pmurias sorear: what is the reasoning behind that?
masak sorear: sorry, "|baz" rather than "baz". 19:23
sorear: that is, it implicitly listifies that expression when it probably shouldn't.
sorear pmurias: it's not particularly useful for serializing, and people should be using JSON or whatever anyway, so I think it makes more sense to optimize for readability and simplicity than Perl 19:24
masak sorear: TimToady has said that an expression should only be listified if there's something syntactical to cause it to listify, either at the beginning of the expression (like '@') or at the end (like '.[]')
masak submits nieczabug 19:25
dalek ecza: c1a6e23 | sorear++ | lib/Cursor.cs:
Fix listification of degenerate captures
 19:26
masak whoa!
19:26 Patterner left
masak I didn't have a chance to press "Submit"! :) 19:26
sorear: it's a pleasure to work with you. 19:27
19:28 envi left
pmurias rakudo: class Foo {};say Foo.perl; 19:30
p6eval rakudo : OUTPUT«Foo␤»
pmurias rakudo: class Foo {};say Foo.new.perl;
p6eval rakudo : OUTPUT«Foo.new()␤»
19:31 flussence_ joined 19:33 MayDaniel joined, flussence_ is now known as flussence 19:38 Psyche^ joined, Psyche^ is now known as Patterner 19:39 estrabd joined
pmurias sorear: i get a compile error on the current niecza 19:46
19:48 kanishka left
pmurias it might be an overly old mono 19:48
pastebin.com/idWnc1Li 19:49
[Coke] if I can get a spectest run against a newer version of parrot, any reason not to update the revision? 19:53
tadzik how do I 'make clean' niecza using xbuild? 19:55
masak sorear: github.com/sorear/niecza/issues#issue/17 is a bother when things go wrong and debugging is needed. :/ 19:58
nom & 20:01
20:02 masak left 20:05 Eth4n joined, Eth4n left, Eth4n joined 20:11 Eth4n left, Eth4n joined, Eth4n left, Eth4n joined 20:12 Eth4n left, Eth4n joined, Eth4n left, Eth4n joined
colomon [Coke]: please feel free. :) 20:14
20:15 shi left 20:19 daxim left 20:21 _twitch left
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....tor_tablet 20:23
20:29 molaf left
dalek rixel: 149c5fc | diakopter++ | / (11 files):
nested (every parent lexical scope included) eval EXPR working.
 20:35
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....tor_tablet 20:37
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....dex_tablet
20:40 vastak left 20:43 plainhao left
dalek rixel: b1df5f8 | diakopter++ | perlesqueTreeParser.g:
delete unneeded file
 20:43
rixel: 6055def | diakopter++ | sprixel/src/ (13 files):
delete more unneeded files
20:43 LeifD left 20:44 a_Pat joined 20:55 MayDaniel left 20:56 Eth4n left
tadzik hmm. I was thinking about writing a chapter for The Book, with easier grammar examples. Now I'm wondering, is there a need, or could the manpower be focused on something more interesting? 20:57
maybe writing about packages and modules? 21:00
or defining own operators in the operators section 21:01
21:03 jest joined 21:07 Eth4n joined, Eth4n left, Eth4n joined 21:12 rgrau joined
dalek rixel: a6cfc57 | diakopter++ | sprixel/ (3 files):
demo of doubly-nested string eval and named sub direct recursion
 21:12
21:16 hercynium left 21:19 hercynium joined, hercynium left, hercynium joined 21:20 ShaneC joined 21:21 cognominal left
diakopter perl6: say((my $a) ~ ($a = "moo")) 21:21
p6eval rakudo : OUTPUT«moomoo␤»
..pugs: OUTPUT«moo␤»
diakopter niecza: say((my $a) ~ ($a = "moo"))
p6eval niecza 406e042: OUTPUT«Can't locate JSON.pm in @INC (@INC contains: /home/p6eval/niecza/src /etc/perl /usr/local/lib/perl/5.10.1 /usr/local/share/perl/5.10.1 /usr/lib/perl5 /usr/share/perl5 /usr/lib/perl/5.10 /usr/share/perl/5.10 /usr/local/lib/site_perl .) at /home/p6eval/niecza/src/NAMBackend.pm
..line 9…
diakopter urp
21:23 ShaneC left
sjohnson oink! 21:23
colomon mooooooooo 21:25
21:25 [hercynium] joined, [hercynium] left, [hercynium] joined
tadzik zezezezezebra! 21:25
or, well, whatever
shortcircuit should see what's in The Book. 21:26
I wonder if finding language analogies for concepts would be a good addition. 21:27
21:27 hercynium left, [hercynium] is now known as hercynium
shortcircuit For example, when I blogged about writing my first Perl 6 program, someone commented that 'multi' reminded him of guards in Erlang. That kind of conceptual bridge might be useful. 21:28
Ok, obviously, I come from a biased background in such things... 21:29
pmurias what's your background? 21:31
21:31 mtk left
shortcircuit pmurias: I run rosettacode.org 21:31
21:32 mtk joined 21:34 Guest54734 left
pmurias multi's and erlang guards are a different conceptually but have a similiar role 21:35
21:39 coldhead joined 21:42 simcop2387 joined 21:45 fhelmberger left 21:47 shi joined 21:49 mtk left 21:53 silug joined 22:02 nadim joined 22:03 a_Pat left 22:05 vert2_ joined
moritz_ mmol-6453.livejournal.com/257585.html # don't know if it has been noticed yet 22:10
22:10 Eth4n left 22:23 GinoMan joined 22:24 kaare_ left 22:26 kfo joined 22:34 simcop2387 left 22:38 MayDaniel joined 22:39 M_o_C left 22:41 hercynium left 22:44 pmurias left 22:46 ggoebel left
colomon ..... aiiiiieeee, what's the method to make something lazy unlazy? 22:50
colomon has too much vicodin in his system to remember, he fears.
Tene eager? 22:51
colomon \o/
Tene++
23:00 jest left 23:08 soroush left, soroush joined 23:10 simcop2387 joined 23:11 simcop2387 left
dalek rixel: d7da935 | diakopter++ | / (9 files):
named sub invocation (without &) & beginnings of use statement,

lexical block, save its location, and eval the rest of that lexical block as a lexical scope contained within the scope of the compilation unit resulting from the 'use' statement.		 23:11
23:13 araujo left, simcop2387 joined
diakopter BEGIN blocks will of course be available, trivially, as well 23:17
23:18 spq1 left
colomon I don't suppose someone out there has a good set of test cases and answers for masak++'s p1? ;) 23:20
afk 4 noms 23:21
23:27 araujo joined 23:34 vm joined 23:39 vm left 23:47 masak joined
masak <colomon> I don't suppose someone out there has a good set of test cases and answers for masak++'s p1? ;) 23:47
I must warn everyone considering borrowing test cases from each other that I will consider all signs of plagiarism -- even non-obvious ones -- a strong disadvantage. 23:48
I'm not going to disqualify anyone for it, I'll just not consider it to be elegant solutions. 23:49
the rational thing for contestants to do is to be fairly discreet about their own code. 23:50
masonkramer What's the metaclass of < a b c >? What's the metaclass of Array? How can I find this out programmatically? 23:56
masak masonkramer: generally, all normal classes have the same metaclass. 23:58
masonkramer: somehow I sense that you're really asking about something else...
23:58 yarrak joined
masonkramer You're right 23:58
23:58 kst left
masonkramer I'll ask a better question then 23:59
What happens when I type Array.^methods

Powered by Cro and the Raku® Programming Language.

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

Thank you!