»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg camelia perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend!
Set by moritz on 3 May 2013.
masak 'night, #perl6 00:03
rjbs Ummm. I ran "perl Configure.pl --gen-parrot
and after doing some "make stuff" I eventually end up in less, showing the document "src/packfile.c - Parrot PackFile API"
If I quit that, I get into an ops document. I quit that, then: Cannot chmod 0644 ops/bit.pod:No such file or directory at /Users/rjbs/perl5/perlbrew/perls/19.0/lib/5.19.0/ExtUtils/Command.pm line 274. 00:04
and the whole thing fails.
sorear Ummm. indeed 00:05
rjbs: in parrot/docs/Makefile, what is the PERLDOC variable set to? 00:06
rjbs PERLDOC = /Users/rjbs/perl5/perlbrew/perls/19.0/bin/perldoc 00:07
sorear do you have a PERLDOC environment variable?
the makefile is using perldoc -ud to extract pod from files 00:08
rjbs -n/opt/local/bin/groff
sorear perhaps you have an environment variable that tells perldoc to do something interesting 00:09
00:09 potatogim joined
sorear or perhaps perldoc5.19.0 behaves in a novel way 00:09
dalek kudo/nom: af50a6e | pmichaud++ | src/core/Str.pm:
Add Str.succ for codepoints \x2581..\x2588 (RT #118519).
 00:16
00:21 grondilu left
rjbs Worked okay once I unset PERLDOC. 00:33
00:35 robinsmidsrod left 00:37 robinsmidsrod joined
rjbs has a perl6 again! 00:39
Given an object, I can get methods with .^methods. How do I see from where each method has come? 00:40
.HOW? 00:41
sorear look over .^mro / .^methods(:local) will definitely work 00:42
there might be a better way
rjbs Huh. enum surprises me 00:44
(everything surprises me)
r: enum X <1 2 3>; X ~~ Enum 00:45
camelia rakudo b2072f: ( no output )
rjbs r: enum X <1 2 3>; say X ~~ Enum
camelia rakudo b2072f: OUTPUT«False␤»
rjbs which means that X.roll's source is local and Enum isn't in the mro for X
sorear Enum is the type of immutable key/value pairs. 00:46
Don't ask me why.
Enumeration is a role which is mixed into all enum objects
r: enum X <x1 x2 x3>; say X ~~ Enumeration 00:47
camelia rakudo b2072f: OUTPUT«False␤»
sorear r: enum X <x1 x2 x3>; say X.^does(Enumeration)
camelia rakudo b2072f: OUTPUT«False␤»
sorear hmm
00:47 abnorman left
sorear r: enum X <x1 x2 x3>; say X.^find_method("roll") === Enumeration.^find_method("roll") 00:48
camelia rakudo b2072f: OUTPUT«False␤»
sorear r: enum X <x1 x2 x3>; say X.^find_method("roll")
camelia rakudo b2072f: OUTPUT«roll␤»
sorear r: enum X <x1 x2 x3>; say X.^find_method("roll").perl
camelia rakudo b2072f: OUTPUT«method roll(X : *@pos, *%named) { ... }␤»
sorear r: enum X <x1 x2 x3>; say Enumeration.^find_method("roll").perl
camelia rakudo b2072f: OUTPUT«method roll(Enumeration : *@pos, *%named) { ... }␤»
sorear oh right, ::?CLASS genericity
00:49 abnorman joined 00:56 benabik left 00:57 benabik joined 01:00 fgomez left 01:12 BenGoldberg left 01:18 ldthien0 joined 01:46 ldthien0 left 01:52 fgomez joined 01:56 betterworld left 01:58 FROGGS left 02:03 betterworld joined
labster r: < . 0 >».ord.say 02:08
camelia rakudo b2072f: OUTPUT«46 48␤»
labster r: say eager ".".."0"
camelia rakudo b2072f: OUTPUT«(timeout)»
labster yeah, that would still be a bug. 02:09
n: say eager ".".."0" 02:10
camelia niecza v24-77-g2b14288: OUTPUT«===SORRY!===␤␤Undeclared routine:␤ 'eager' used at line 1␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1502 (die @ 5) ␤ at /home/p6eval/niecza/src/STD.pm6 line 1147 (P6.comp_unit @ 36) ␤ at /ho…
labster n: say (".".."0").join
camelia niecza v24-77-g2b14288: OUTPUT«.␤»
02:13 FROGGS joined 02:21 dayangkun_ joined 02:22 dayangkun_ left, mrallen1 joined 02:25 dayangkun left 02:41 bonsaikitten left, atroxaper joined, bonsaikitten joined 02:58 SamuraiJack joined
colomon rn: say ((set <a b c>), (bag <a b c>)).grep(* ~~ Baggy) 03:22
camelia rakudo b2072f, niecza v24-77-g2b14288: OUTPUT«␤»
colomon rn: say ((set <a b c>), (bag <a b c>)).grep({$_ ~~ Baggy}) 03:23
camelia niecza v24-77-g2b14288: OUTPUT«bag("a" => 1, "b" => 1, "c" => 1)␤»
..rakudo b2072f: OUTPUT«␤»
colomon errr.... what am I missing with the first one? and rakudobug?
lue .tell jnthn nqp-moar-cc.nqp:68 and MASTTesting.nqp:11 seem to be using a Windows specific command (the same one) without some OS-based conditional, leading to "sh: del: command not found" occurring in temp.output 03:25
yoleaux lue: I'll pass your message to jnthn.
lue .tell jnthn I'm still able to use nqp-moar-cc.nqp though, so it doesn't seem to immediately break things, at least to me. 03:26
yoleaux lue: I'll pass your message to jnthn.
03:32 preflex_ joined, ChanServ sets mode: +v preflex_ 03:33 preflex left, preflex_ is now known as preflex 04:00 raiph joined 04:27 potatogim left 04:29 potatogim joined 04:30 Patterner left 04:32 Psyche^ joined, Psyche^ is now known as Patterner, Ayiko joined 04:54 birdwindupbird joined 04:55 araujo left 04:59 havenwood joined 05:03 ldthien0 joined 05:18 tomyan joined 05:24 potatogim left 05:25 potatogim joined 05:31 ssutch left 05:36 twigel joined 05:37 REPLeffect left 05:41 dmol joined
skids star: gist.github.com/skids/5802914 05:47
camelia star 2013.02: OUTPUT«1 2 3 4␤1 2 3 4␤»
skids hmmm. Oh it isn't 05. 05:48
r: gist.github.com/skids/5802914
camelia rakudo b2072f: OUTPUT«1 2 3 4␤1Block.new()␤No exception handler and no message␤␤current instr.: 'throw' pc 347557 (src/gen/CORE.setting.pir:151689) (src/gen/CORE.setting:8887)␤called from Sub 'sink' pc 379739 (src/gen/CORE.setting.pir:164183) (src/gen/CORE.setting:10169)␤called from Sub…
05:55 PacoAir joined 05:56 raiph left 05:59 PacoAir left 06:02 FROGGS left 06:18 dmol left, domidumont joined 06:19 domidumont left 06:36 PZt left
dalek p: ca52770 | sorear++ | src/vm/jvm/runtime/org/perl6/nqp/runtime/Ops.java:
Add a nqp::debugnoop opcode, which does nothing but you can put a breakpoint on it.
 06:41
p: c9738ae | sorear++ | src/vm/jvm/ (3 files):
Multi cache should decontainerize before taking the type for JVM/Parrot consistency. Still a little dodgy because it assumes hashCode values are unique.
sorear sorear$ ./perl6 -e 'use v6; say "hi"' 06:42
hi
06:42 tomyan left
moritz \o/ 06:43
06:44 havenwood left 06:48 ldthien0 left 06:49 domidumont joined
labster good morning 06:50
r: "eighty" <= 80 06:51
camelia rakudo b2072f: OUTPUT«Cannot call 'Real'; none of these signatures match:␤:(Mu:U \v: Mu *%_)␤ in method Real at src/gen/CORE.setting:870␤ in method Real at src/gen/CORE.setting:2596␤ in method Real at src/gen/CORE.setting:870␤ in sub infix:<<=> at src/gen/CORE.setting:3104␤ in sub …
sorear labster: hey, I thought you were in my time zone 06:52
labster it's morning in 10 minutes :P
and it's always morning in UGT
07:01 mrallen1 left 07:08 domidumont left 07:09 domidumont joined
moritz labster++ # UGT 07:12
07:24 FROGGS joined, kaleem joined
TimToady
.oO(The two hardest problems in Computer Science are indirection coherence, and off-by-one levels of indirection...)
 07:41
I suppose a cache is an off-by-one level of indirection, so maybe they're the same problem... 07:42
08:02 araujo joined, araujo left, araujo joined 08:03 domidumont left
dalek ast: 0b9ec37 | (Elizabeth Mattijsen)++ | S32-list/categorize.t:
Added some typed hash tests to categorize
 08:06
08:15 kresike joined
kresike hello all you happy perl6 people 08:15
08:15 domidumont joined
sorear o/ kresike 08:30
08:33 pochi left
dalek kudo/nom: 9fae045 | (Brent Laabs)++ | src/core/IO.pm:
add IO::Path methods rename, chmod
 08:38
kudo/nom: 74ddcd4 | (Brent Laabs)++ | src/core/IO.pm:
move dir()'s logic into IO::Path.contents; it's working with paths already
kudo/nom: 0c78d26 | (Brent Laabs)++ | src/core/IO.pm:
add IO::Path.succ and .pred
kudo/nom: 92c09eb | (Brent Laabs)++ | src/core/IO.pm:
undo rename for now
rakudo/nom: c90fcf7 | (Brent Laabs)++ | src/core/IO.pm:
labster sorry dalek
dalek p: fff73cf | sorear++ | src/NQP/Actions.nqp:
Need to use savecapture when calling MMD dispatcher for reentrancy on JVM
 08:39
08:46 dakkar joined 08:50 kaleem left 08:51 brrt joined
brrt hi all 08:51
yoleaux 10 Jun 2013 15:35Z <diakopter> brrt: frankly, I think those project types are actually superhuman, at least in the attention to detail and skill and training required, if not actual code size
brrt wow
old discussion :-) 08:52
08:53 kaleem joined
brrt .tell diakopter again, they are hard, like surgery is hard, but it is doable 08:53
yoleaux brrt: I'll pass your message to diakopter.
brrt for the record, i don't intend to knock people who do spend their time writing efficient compilers / vm *at all* 08:54
i wish to knock down the conception that anything which isn't RoR autogenerated CRUD is superhuman, because I believe it holds us down
and articles like the wired' one - which idolize the developers that do harder work - reinforce the conception 08:55
bonsaikitten brrt: looking at really "hard" problems like compilers - most are incomplete / crap 08:57
that suggests that either people lack the attention span for it or settle for a "good enough" solution
brrt thats a possibility
tadzik most are crap? Most of everything is crap, but there are good compilers around 08:58
as compared to, say, email clients, which _all_ suck
sorear CLA is in the outbox
brrt lol :-) the other possibility is that outsiders such as you and i do not understand the same problems the authors do
bonsaikitten tadzik: I've not been able to find a reliable C compiler that can be built with only a C compiler
brrt i.e. 'why does GCC does X?' well, probably there is a good reason 08:59
08:59 pr_ joined
sorear 1..1 08:59
ok 1 - hi
this is promising
09:00 fhelmberger joined
moritz sorear: is this rakudo? 09:01
09:03 pr_ left
sorear moritz: yes 09:04
moritz \o/
09:05 domidumont left
sorear trouble is it only wants to look for modules in . and blib 09:06
while test.pm lives in lib
09:06 domidumont joined
sorear r: use Bob 09:07
camelia rakudo b2072f: OUTPUT«===SORRY!===␤Could not find Bob in any of: /home/p6eval/nom-inst/lib/parrot/5.2.0-devel/languages/perl6/site/lib, /home/p6eval/nom-inst/lib/parrot/5.2.0-devel/languages/perl6/vendor/lib, /home/p6eval/nom-inst/lib/parrot/5.2.0-devel/languages/perl6/lib, /home/p6eval…
sorear r: say %*ENV.keys
camelia rakudo b2072f: OUTPUT«PERLBREW_VERSION MAIL USER SHLVL HOME OLDPWD PERLBREW_ROOT PERLBREW_MANPATH LOGNAME _ TERM PATH PERLBREW_HOME PATH_WITHOUT_PERLBREW PERLBREW_BASHRC_VERSION LANG LS_COLORS PERLBREW_PATH SHELL PERLBREW_PERL PWD MANPATH MANPATH_WITHOUT_PERLBREW␤»
sorear r: say @*INC 09:08
camelia rakudo b2072f: OUTPUT«/home/p6eval/nom-inst/lib/parrot/5.2.0-devel/languages/perl6/site/lib /home/p6eval/nom-inst/lib/parrot/5.2.0-devel/languages/perl6/vendor/lib /home/p6eval/nom-inst/lib/parrot/5.2.0-devel/languages/perl6/lib /home/p6eval/.perl6/2013.05-30-g9557da2/lib␤»
moritz sorear: that's why Tests are run with PERL6LIB=lib or -Ilib 09:10
sorear moritz: PERL6LIB is being ignored 09:13
as is -Ilib 09:14
moritz :( 09:16
nwc10 sorear: when does the USPS empty the outtray? "Tomorrow" morning? 09:19
sorear nwc10: I think so, but I'm not actually sure 09:20
not at 2am, that's for sure :) 09:21
r: say nqp::cwd() 09:24
camelia rakudo b2072f: OUTPUT«/home/p6eval␤»
09:39 daxim joined 09:49 atroxaper left 09:52 atroxaper joined
dalek : 8221c3a | (Tobias Leich)++ | t/test_summary:
print summary when tests are done too
 09:59
: 42ba0fa | (Tobias Leich)++ | / (3 files):
make `time` an identifier with proto (), rather than a term
10:00 rindolf joined, brrt left
daxim lang designers, have you seen <philsturgeon.co.uk/blog/2013/01/php...-wind>? 10:01
this is not the un-php6 from <schlueters.de/blog/archives/128-Fut...6.html> and <lwn.net/Articles/379909/> 10:02
masak good midday, #perl6 10:06
sorear o/ masak 10:07
10:07 daxim left
sorear masak: I've convinced rakudo-jvm to accept both use v6 and use Test. also, my CLA is in the mail. kind of. 10:07
10:08 daxim joined
nwc10 but, presumably, unabmigiously CLA-enabled Rakudo committers could merge your work right now? 10:09
just like any regular pull request
masak sorear: nice.
nwc10 that too :-)
masak nwc10: oh yes. I am sure of that. 10:10
sorear masak: how much time do I have left to surprise jnthn++ with a clean spectest run? 10:16
10:18 jaldhar left
FROGGS where is he anyway? 10:20
lizmat I think jnthn++ is spending some well earned days afk 10:27
masak teaching. 10:28
sorear: you can expect him to resurface in about 28 hours.
sorear excellent 10:35
masak sorear++ 10:40
dalek p: 9625921 | sorear++ | src/vm/jvm/ (2 files):
Add jvmgetproperties op for access to JVM VM info
 10:41
p: 9acdf7a | sorear++ | src/vm/jvm/runtime/org/perl6/nqp/runtime/Ops.java:
Implement CCLASS_PRINTING on jvm
10:50 snearch joined
dalek kudo/nom: 6620d3b | (Elizabeth Mattijsen)++ | src/core/Hash.pm:
Get rid of unneeded local lists in categorize
 11:02
moritz who does the star modules smoking on current rakudo? 11:05
was that [Coke]?
lizmat I think so, yes
sorear gist.github.com/sorear/5804522 11:11
(that's for everyone :D) 11:15
moritz that's quite impressive 11:18
sorear btw, -Ilib works now 11:19
FROGGS cool!
sorear and on that note, sleep& 11:20
lizmat goodnight sorear++
sorear i've pushed everything to sorear/rakudo/nom; beware that I haven't tested the changes on parrot 11:21
11:22 atroxaper left
dalek kudo/nom: 003f346 | (Timo Paulssen)++ | src/Perl6/Grammar.nqp:
allow a dot before hyper postfix (also unspace).
 11:23
kudo/nom: 2765030 | moritz++ | src/Perl6/Grammar.nqp:
Merge branch 'add-dotty-hyper-postfix' of github.com/timo/rakudo into nom
11:24 benabik left
dalek ast: 629ff33 | moritz++ | S03-operators/set.t:
fudge set.t for rakudo
 11:24
woolfy sorear++ 11:26
11:38 pr_ joined
pr_ Hi, is the :nl parameter in IO::Handle::open not functional? eg, open $file, :w, :enc<Latin-1>, :nl<"\r\n\r\n"> 11:39
11:41 yves left
moritz pr_: seems it isn't 11:42
pr_: the proper way to supply it would be :nl("\r\n\r\n") 11:43
:nl<...> already implies quoting, so the double quotes would be part of the string passed to the function -- not what you want
(but it's not yet implemented)
pr_ When nl("\r\n\r\n") is implemented, will this be the best way to specify Windows line endings? 11:45
masak pr_: I should think in the ordinary case, Windows line endings will be handled correctly by default. 11:46
pr_: what you want to do looks like a combination of that and paragraph breaks.
pr_ masak: it seems a bit complicated. I miss :crlf :-) 11:47
masak again, Windows line breaks Should Just Work without you doing anything. just like in Perl 5. 11:48
I don't think that seems complicated ;)
pr_ masak: understood. But I was thinking of writing to Unix/Mac files on Windows, and vice-versa. I will just have to practice the new technique. 11:50
lizmat r: my $a=1; my @a; @a.push($a); say $a === @a[0]; $a=2; say $a; say @a[0] # apparently a===b is not enough to check for bound variables 11:53
camelia rakudo b2072f: OUTPUT«True␤2␤1␤»
11:54 yves joined
lizmat Q: how would I check whether two values are bound to the same container? 11:54
Q: how can I push a bound value onto a list
FROGGS I thought arrays dont do binding
r: my $a=1; my @a; @a.push($a); say $a === @a[0]; $a=2; say $a; say @a[0] # hmmm 11:55
camelia rakudo b2072f: OUTPUT«True␤2␤1␤»
FROGGS r: my $a=1; my @a; @a.push($a); say $a === @a[0]; $a=2; say $a; say @a[0] # this is hmmm 11:56
camelia rakudo b2072f: OUTPUT«True␤2␤1␤»
FROGGS damn it
r: my $a=1; my @a; @a[0] := $a; say $a === @a[0]; $a=2; say $a; say @a[0]
camelia rakudo b2072f: OUTPUT«True␤2␤2␤»
FROGGS r: my $a=1; my @a; @a[0] = $a; say $a === @a[0]; $a=2; say $a.WHERE; say @a[0].WHERE 11:57
camelia rakudo b2072f: OUTPUT«True␤-1160124824␤-1174207627␤»
FROGGS r: my $a=1; my @a; @a[0] := $a; say $a === @a[0]; $a=2; say $a.WHERE; say @a[0].WHERE
camelia rakudo b2072f: OUTPUT«True␤-581111009␤-581111009␤»
FROGGS that is the answer to Q1
lizmat $a.WHERE ~~ $a[0].WHERE you mean? 11:58
or $a.WHERE == $a[0].WHERE ?
FROGGS == 11:59
moritz =:= checks for container equivalence
r: my $x = 3; my @a; @a[0 12:00
camelia rakudo b2072f: OUTPUT«===SORRY!===␤Unable to parse expression in subscript; couldn't find final ']'␤at /tmp/6_bV39Wdq4:1␤------> my $x = 3; my @a; @a[0⏏<EOL>␤ expecting any of:␤ postfix␤ statement end␤ statement modifier␤ …
lizmat r: my $a=1; my $b:=$a; say $a =:= $b # indee
camelia rakudo b2072f: OUTPUT«True␤»
lizmat d
FROGGS r: my $a=1; my @a; @a[0] = $a; say $a =:= @a[0];
camelia rakudo b2072f: OUTPUT«False␤»
moritz r: my $a=1; my @a; @a[0] := $a; say $a =:= @a[0];
camelia rakudo b2072f: OUTPUT«True␤»
FROGGS answer to Q2: bind_pos might what you want 12:01
lizmat moritz++, FROGGS++ thanks for the pointers 12:02
masak .oO( "thanks for the pointers", said the c2.com/cgi/wiki?ThreeStarProgrammer ) 12:04
lizmat I'm so glad I don't have to program in C 12:05
masak: our as the previous pope would say "denk you four ze blumen" 12:06
masak .oO( "Did Beethoven compose eight symphonies?" -- "Nein!" ) 12:08
moritz groans 12:09
tadzik ehehe
masak mission accomposed :)
lizmat goes out to cool down 12:10
masak ...something that's not possible in Austin, Texas... :) 12:12
nwc10 not very possible in Vienna currently
lizmat well, that was the dead pan part of my joke: it's 31 celsius in the shade outside now 12:17
masak :) 12:18
sorry for being so slow. it must be the heat.
12:22 JimmyZ joined
lizmat I wonder how the description of pop at S32/Containers:697 can be combined with the reverse example at S32/Containers:386 12:23
in other words: 697 describes as pop failing on an empty array, while the reverse example seems to assume it will not? 12:24
12:26 atroxaper joined 12:28 Kelder joined, Ayiko left
masak rn: say (my @a).pop ~~ Failure 12:29
camelia niecza v24-77-g2b14288: OUTPUT«===SORRY!===␤␤Undeclared name:␤ 'Failure' used at line 1␤␤Potential difficulties:␤ @a is declared but not used at /tmp/5Kj7kwG9_c line 1:␤------> say (my ⏏@a).pop ~~ Failure␤␤Unhandled exception: Check failed␤␤ at /ho…
..rakudo b2072f: OUTPUT«True␤»
masak r: my @a = 1, 2, 3; say "yay" while @a.pop; say "alive"
camelia rakudo b2072f: OUTPUT«yay␤yay␤yay␤alive␤»
masak lizmat: no, the reverse example only seems to assume that 'pop @values' will return something falsy. 12:30
lizmat so the while eats the failure and interpretes it as false
so it never gets thrown?
masak right.
that's Failures for you.
lizmat: which works (see above) but is stupid for a different reason -- it would stop on 0 and "" and empty arrays.
I dislike pseudocode like this in the spec. it almost always turns out to be too naïve. 12:31
lizmat or if there is any other (internal) failure in pop, one would never know
masak right. 12:32
lizmat hmmmm....
masak worse, code like that in the spec makes it seem like that's how the routine *should* be implemented (instead of just providing a hopefully useful hint on how it could be implemented) 12:33
lizmat well, that could be fixed with a "For example"
masak it's a *spec*. it should be conservative about what it specs.
lizmat so why couldn't pop return Mu on an empty array 12:34
and have it throw any Failures?
masak er. you throw Exceptions, not Failures.
moritz if you do nothing with the returned Failure, it blows up 12:35
masak and Failure seems to be the right thing to return from a pop on an empty array.
moritz so the chances of hiding a bug are smaller when returning a Failure
lizmat r: my @a; say pop @a
camelia rakudo b2072f: OUTPUT«Element popped from empty list␤ in method gist at src/gen/CORE.setting:10161␤ in method gist at src/gen/CORE.setting:893␤ in sub say at src/gen/CORE.setting:11047␤ in block at /tmp/_boKEm_mAL:1␤␤»
timotimo isn't a Nil more appropriate for pop on an empty array?
lizmat so I guess "say" is stringifying the Failure
moritz .gist actually
lizmat ack 12:36
moritz and it blows up, because it's not yet handled
timotimo r: my @a; print pop @a
camelia rakudo b2072f: OUTPUT«Element popped from empty list␤ in method Str at src/gen/CORE.setting:10160␤ in method Str at src/gen/CORE.setting:876␤ in method print at src/gen/CORE.setting:11231␤ in method print at src/gen/CORE.setting:11225␤ in sub print at src/gen/CORE.setting:11041␤ i…
lizmat and "while" is silently "handling" failure
moritz right
testing truthiness and .defined-ness on a Failure sets its 'handled' bit 12:37
moritz is not a big fan of that one
lizmat for now, I tend to agree
with not being a big fan of that
.oO{wish we had a fan now} 12:38
12:38 konundra left
FROGGS gah, don't make so much wind about it 12:38
nwc10 don't worry, it will all blow over soon
timotimo lets out a big sigh 12:40
[Coke] sorear++ #cla
masak so, I caught wind of these air-based puns you guys are doing. 12:41
frankly, I think the whole thing is a bit overblown.
[Coke] Anyone wants rakudo-jvm smoked for star (needs more infrastructure) or for the spec tests, we need a box we can run java on. feather1 is out, per Juerd.
lizmat can one create a MMD candidate depending on the value of a method / attribute of the object? 12:42
[Coke]: am about to retire my 3 year old 17" MacBook Pro, with 8G of RAM, would that be able to do this? 12:43
[Coke] lizmat: aside from the difficulty of building pugs the first time, likely.
masak r: class C { has $.attr is rw = 5; multi method foo($x where { $x == $.attr }) { say "OH HAI" }; multi method foo($x) { say "OH default" } }; given C.new { .foo; .attr = 42; .foo }
camelia rakudo b2072f: OUTPUT«Cannot call 'foo'; none of these signatures match:␤:(C : $x, Mu *%_)␤:(C : $x, Mu *%_)␤ in method foo at src/gen/CORE.setting:440␤ in block at /tmp/AvLKXp2MGv:1␤␤»
moritz would be very happy with such a machine :-)
masak r: class C { has $.attr is rw = 5; multi method foo($x where { $x == $.attr }) { say "OH HAI" }; multi method foo($x) { say "OH default" } }; given C.new { .foo(5); .attr = 42; .foo(5) } 12:44
moritz [Coke]: where are the star-on-current-rakudo smoke results?
camelia rakudo b2072f: OUTPUT«OH HAI␤OH default␤»
masak lizmat: like that? :)
[Coke] github.com/coke/rakudo-star-daily/...odules.log
12:44 pr_ left
lizmat how would that work performance wise? Thinking about adding MMD candidates for so List.infinite 12:44
moritz [Coke]: thanks 12:45
[Coke] lizmat: if you'd be willing to leave the lights on, I would definitely be happy to give the box a whirl.
lizmat [Coke]: not yet, I still need to order the new one (delivery should be 1-3 working days)
[Coke] (or I could host it somewhere.) 12:46
lizmat: no rush.
lizmat so sometime next week
12:46 domidumont left
moritz the only test failing in star with current rakudo is perl6-File-Tools:t/01-file-find.t test 10 12:46
[Coke] so, I'm wondering if it make sense to have the daily rakudo star builds always go for latest rakudo and latest modules. 12:47
(as opposed to the versions marked in the repo) 12:48
lizmat I would think so… that was the whole idea, no?
moritz and panda fails to bootstrap :(
tadzik :|
lizmat so that problems are found early, rather than at Rakudo release time?
[Coke] lizmat: when we started it a few weeks ago, even the selected versions of everything failed to work.
moritz ==> Bootstrapping Panda
==> Fetching File::Tools
Failed to remove the file '.work/1371559686_1': unlink failed: No such file or directory
tadzik fff
dagurval ._. 12:49
tadzik easy to fix at least
dagurval test 10 is the one I just added ._.
tadzik I'll look at it after $job
12:52 konundra joined
lizmat seemed to have introduced a recursion while saving the restricted settings :-( 12:53
12:53 Timbus left
FROGGS .oO( s.o.s - save our setting ) 12:54
12:56 panchiniak_ joined, Timbus joined 12:57 ajr joined, ajr is now known as Guest4236 12:58 Guest4236 is now known as ajr_
JimmyZ :D 12:59
13:01 domidumont joined
lizmat r: class C { has $.attr = 5; multi method foo($x where { $.attr == 5 }) { say "OH HAI" }; multi method foo($x) { say "OH default" } }; given C.new { .foo(5); .attr = 42; .foo(5) 13:05
camelia rakudo b2072f: OUTPUT«===SORRY!===␤Unable to parse expression in block; couldn't find final '}'␤at /tmp/TU2gMDHRY6:1␤------> ven C.new { .foo(5); .attr = 42; .foo(5)⏏<EOL>␤ expecting any of:␤ postfix␤ statement end␤ statement …
lizmat excuse, ww :-)
timotimo github.com/rakudo/rakudo/pull/160 - is there something wrong with this that prevents merging? 13:06
lizmat spec changes generally need to be looked over by pmichaud / TimToady ? 13:08
<plug> a bit like github.com/perl6/specs/issues/54 </plug> ?
timotimo oh, i forgot. 13:09
13:09 atroxaper left
timotimo in that case: is there something wrong with it that prevents a discussion from taking place? :) 13:09
moritz timotimo: nothing wrong, just that nobody has looked over it yet, it seems 13:10
timotimo: and I'd like to wait until after the release, for fear that it might break some module and thus star
timotimo S06 at least doesn't mention types on slurpies 13:11
13:12 PZt joined
timotimo it doesn't seem like the spec says anything about typed slurpy arguments; so there should be an addition somewhere that says "you can't do that." 13:13
so not really a spec change per se 13:14
moritz sure, but it still might be used somwhere
timotimo also, if it does break modules, it would only mean they were broken already, just waiting for compile time to actually explode :)
moritz maybe even in a multi candidate that is never reached
timotimo should i perhaps just merge locally and run a full star test run?
moritz so the code might work right now, even for the wrong reasons
timotimo actually, i could even just grep this 13:16
13:24 tidux joined, PacoAir joined
tidux so is there any big software that's been built in perl6 yet? 13:25
JimmyZ how big is big?
dalek ecza: 9f58f84 | (Solomon Foster)++ | lib/CORE.setting:
Rewrite infix:<∪> and infix:<∩> to take arbitrary inputs.

Previous versions of these all took two arguments, but they are list-ops, and so may be passed any number of arguments.
tidux bigger than simple scripts and the usual perl fare
not necessarily Frozen Bubble :P 13:26
JimmyZ how simple is simple?
tidux <500 lines
JimmyZ rakudo is big enough? it's written in Perl 6
tidux lol
point taken 13:27
13:27 tidux left
dalek ast: ee5778e | (Solomon Foster)++ | S03-operators/bag.t:
Fudge for Niecza.
 13:27
colomon tidux: ABC module is 1600+ lines at the moment, not counting test files. 13:29
masak the biggest I ever wrote is probably GGE... 13:30
daxim www.ibiblio.org/pub/linux/search/ke....54.tar.gz # behold! I bring you code from that last millenium! 13:35
JimmyZ hmm, looks like mbc file is half size of the pbc one 13:36
FROGGS mbc is moarvm bytecode? 13:37
timotimo it probably doesn't store all the line numbers in 64bit integers :P 13:38
dagurval r: use MONKEY_TYPING; augment class Cool { method bark() { say "woof!" } }; 5.bark
camelia rakudo b2072f: OUTPUT«No such method 'bark' for invocant of type 'Int'␤ in block at /tmp/sKY0qVp_fw:1␤␤»
dagurval shouldn't my Int bark since Int is Cool?
timotimo isn't cool a role, though?
JimmyZ yes, but after to mbc.zip, mbc file is 9/10 of the pbc.zip one
moritz it's not
timotimo i think retroactively enhancing roles won't do it
moritz dagurval: yes, known bug
timotimo mhm
dagurval moritz: ok
FROGGS dagurval: inherited classes are not affected yet 13:39
moritz the problem is that method caches are wrote at class composition time
dagurval I see :)
moritz but a class can't just keep a reference to all subclasses, because then anonymous subclasses wouldn't be garbage-collected
so how should a class know which method caches from subclasses need to be invalidated? 13:40
timotimo r: say dir()
camelia rakudo b2072f: OUTPUT«IO::Path<star> IO::Path<src> IO::Path<.subversion> IO::Path<.bashrc> IO::Path<nom-inst1> IO::Path<toqast> IO::Path<test3.pl> IO::Path<.profile> IO::Path<t> IO::Path<nom-inst2> IO::Path<nom-inst> IO::Path<toqast-inst> IO::Path<toqast-inst2> IO::Path<examples> IO::Pa…
moritz r: say ~<<dir()
camelia rakudo b2072f: OUTPUT«star src .subversion .bashrc nom-inst1 toqast test3.pl .profile t nom-inst2 nom-inst toqast-inst toqast-inst2 examples Makefile .gitignore lib test2.pl .local bin rakudo p1 VERSION simple-tests .perlbrew std obj main.pl .lesshst nom niecza test.pl .cpanm log .bash_…
timotimo having a bit of trouble with panda at the moment 13:41
rather, with file::tools
File::Find::Result is cool, but doesn't implement .path, just .Str, so the dir function breaks; perhaps dir should call $thing.Str.path instead of $thing.path? 13:42
timotimo investigates
13:42 brrt joined
tadzik I think F::F::Result is mostly obsolete, since we have now a sensemaking Path objects 13:42
I thought there's even a patch for it 13:43
timotimo ah. cool doesn't prescribe a .path to exist, so dir() isn't actually correct.
thoughts? should Cool get a .path that runs .Stringy.path? 13:44
moritz aye 13:45
timotimo prepares a patch & PR 13:47
13:50 btyler joined
timotimo runs spectests 13:52
13:54 bluescreen10 joined 13:59 ajr_ left 14:00 ajr joined, ajr is now known as Guest78348 14:01 Guest78348 is now known as ajr_ 14:04 kaleem left 14:05 fgomez left
dalek kudo/nom: 5feb260 | (Timo Paulssen)++ | src/core/Cool.pm:
give Cool a path method to run self.Stringy.path
 14:08
kudo/nom: 8bbebc9 | moritz++ | src/core/Cool.pm:
Merge pull request #167 from timo/give-cool-path-method

give Cool a path method to run self.Stringy.path
14:15 domidumont left 14:16 domidumont joined
dalek ast: b5485fa | (Timo Paulssen)++ | S0 (2 files):
unfudge now-passing unspace and dotty hyper tests
 14:17
14:19 gabriel__ joined 14:32 konundra left, mrallen1 joined 14:34 p5eval left 14:38 p5eval joined 14:46 araujo left 14:49 mrallen1 left 14:54 araujo joined 14:55 konundra joined, JimmyZ left 14:56 kaare_ joined, ajr_ left
kresike bye folks 14:58
14:58 kresike left, ajr_ joined 15:00 brrt left 15:01 panchiniak_ left 15:10 potatogim_ joined 15:14 btyler left 15:20 potatogim_ is now known as potatogim_H, potatogim_H is now known as potatogim_LT 15:25 konundra left, konundra joined 15:31 bruges left 15:32 dmol joined 15:33 btyler joined, bruges joined 15:39 REPLeffect joined, potatogim_LT left 15:41 SamuraiJack left 15:42 FROGGS left, domidumont left 15:56 konundra left 16:04 vk joined, vk left, spider-mario joined
daxim r: sub by(Int $n, @a) { my @r; push @r, splice @a, 0, $n; return @r; }; by(5, [1..15]).perl.say; 16:06
camelia rakudo b2072f: OUTPUT«Array.new(1, 2, 3, 4, 5)␤»
16:06 FROGGS joined
daxim r: sub by(Int $n, @a) { my @r; while (@a) { push @r, splice @a, 0, $n; }; return @r; }; by(5, [1..15]).perl.say; 16:06
camelia rakudo b2072f: OUTPUT«Array.new(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15)␤»
daxim r: sub by(Int $n, @a) { my @r; while (@a) { push @r, [splice @a, 0, $n]; }; return @r; }; by(5, [1..15]).perl.say;
camelia rakudo b2072f: OUTPUT«Array.new([1, 2, 3, 4, 5], [6, 7, 8, 9, 10], [11, 12, 13, 14, 15])␤»
masak r: sub infix:<by>(@a, Int $n) { my @r; while (@a) { push @r, [splice @a, 0, $n]; }; return @r }; .say for [1..15] by 5 16:08
camelia rakudo b2072f: OUTPUT«1 2 3 4 5␤6 7 8 9 10␤11 12 13 14 15␤»
masak \o/
daxim that's neat. can you do method "by" on List literals, too? 16:09
masak if you're up for MONKEY_TYPING, sure. 16:10
daxim I'm always up for it
perigrin how do you think we write our code *now*?
masak :)
FROGGS r: sub by(Int $n, @a) { gather { take [splice @a, 0, $n] } xx +@a/$n }; by(5, [1..15]).perl.say;
camelia rakudo b2072f: OUTPUT«([1, 2, 3, 4, 5], [6, 7, 8, 9, 10], [11, 12, 13, 14, 15]).list␤»
masak I forgot you are Perl 5 people :P
daxim "we need typewriters. lots of typewriters."
colomon perigrin++ 16:11
[Coke] "THIS IS PERL 5!!!" -khan 16:12
perigrin daxim++
masak r: use MONKEY_TYPING; augment class Range { method by(Int $n) { my @a = self; my @r; while (@a) { push @r, [splice @a, 0, $n] }; return @r } }; .say for (1..15).by(5)
camelia rakudo b2072f: OUTPUT«1 2 3 4 5␤6 7 8 9 10␤11 12 13 14 15␤»
perigrin masak: worse, increasingly I am business people. 16:13
masak r: say Range.^mro; say Array.^mro; say List.^mro
camelia rakudo b2072f: OUTPUT«(Range) (Iterable) (Cool) (Any) (Mu)␤(Array) (List) (Iterable) (Cool) (Any) (Mu)␤(List) (Iterable) (Cool) (Any) (Mu)␤»
masak hm, should probably go on Iterable.
perigrin I happen to write code because I can't afford programmer people to do it for me at the moment.
daxim is Iterable something concrete or just dangling in there for abstraction?
16:13 domidumont joined
masak r: use MONKEY_TYPING; augment class Iterable { method by(Int $n) { my @a = self; my @r; while (@a) { push @r, [splice @a, 0, $n] }; return @r } }; .say for (1..15).by(5) 16:14
camelia rakudo b2072f: OUTPUT«No such method 'by' for invocant of type 'Range'␤ in block at /tmp/IJcgrBhBZZ:1␤␤»
masak :(
r: class A {}; class B is A {}; use MONKEY_TYPING; augment class A { method foo { say "OH HAI" } }; B.new.foo
camelia rakudo b2072f: OUTPUT«No such method 'foo' for invocant of type 'B'␤ in block at /tmp/eSG1qAjhpW:1␤␤»
masak shouldn't this trivially work?
is this a known bug?
daxim: Iterable is a concrete class in the CORE setting. 16:15
16:15 ajr_ left
daxim I'm having fun already 16:16
doc.perl6.org/type/Iterable#gist is delightful
[Coke] masak: I can't say I'm surprised that you're not seeing a change in your already instantiated class. 16:17
I can see where we'd want it to work, though.
r: class A {}; class B is A {}; use MONKEY_TYPING; augment class A { method foo { say "OH HAI" } }; A.new.foo;
camelia rakudo b2072f: OUTPUT«OH HAI␤»
FROGGS r: List.HOW.add_method( List, sub ($self, $n) { gather { take [splice $self, 0, $n] } for ^3 } ); [1..15].by(5).perl.say; 16:19
camelia rakudo b2072f: OUTPUT«too few positional arguments: 3 passed, 4 (or more) expected␤ in any add_method at src/gen/Metamodel.nqp:391␤ in block at /tmp/DEdlWqEz0i:1␤␤»
FROGGS meh
masak [Coke]: class wasn't already instantiated.
at least not according to the normal definition of "instantiate".
FROGGS masak: that is the caching bug mentioned earlier this day, no?
masak [Coke]: it feels like a fairly standard assumption that deriving class see new methods. I realize it takes some kind of method cache update, which takes some kind of "backlinks" from ancestor classes to descendant classes, which is conceptually problematic. 16:20
FROGGS r: List.HOW.add_method( List, 'by', sub ($self, $n) { gather { take [splice $self, 0, $n] } for ^3 } ); [1..15].by(5).perl.say;
camelia rakudo b2072f: OUTPUT«No such method 'by' for invocant of type 'Array'␤ in block at /tmp/LvaAvwy5t6:1␤␤»
masak but Perl 6 doesn't shy away from conceptually problematic things :)
FROGGS: oh, most likely.
masak goes hunting through the backlog
FROGGS r: Array.HOW.add_method( Array, 'by', sub ($self, $n) { gather { take [splice $self, 0, $n] } for ^3 } ); [1..15].by(5).perl.say;
camelia rakudo b2072f: OUTPUT«([1, 2, 3, 4, 5], [6, 7, 8, 9, 10], [11, 12, 13, 14, 15]).list␤»
FROGGS \o/
masak FROGGS: I don't find it in the backlog. url? 16:21
FROGGS masak: maybe search for moritz + cache
hold on
masak ok
found some joking about caching yesterday, but it doesn't seem related. 16:22
16:22 diederich joined
FROGGS irclog.perlgeek.de/perl6/2013-06-18#i_7211743 16:22
masak yes, moritz correctly identifies the problem. 16:23
ISTR that jnthn has a solution in mind, though.
FROGGS cool
that would be sweet
masak "how can anonymous classes be GC'd?" -- seems to me like a classic use case for weak references. 16:24
but then TimToady would point out that weak references are a design smell. :) 16:25
daxim good enough
masak dangit, I've spent so much time here that I can basically enact people's replies without them actually being around! :P
anyway, I don't see an insurmountable conceptual problem there. 16:26
flussence aw crap, rakudo-parrot isn't compiling for me today. (gist.github.com/anonymous/5806904) 16:27
masak conceptually, it feels to me like deriving classes:base class :: consumers:producer, and as long as subscriptions are set up as weak references, there are no GC issues. 16:28
FROGGS hmmm, I guess it would help if I knew what weak refs are...
masak FROGGS: it's like a normal reference, except that the GC chooses to ignore it in counts.
FROGGS: with the effect that the reference may at any point become invalid because the pointee was GC'd. 16:29
FROGGS ahh
masak FROGGS: basically, it's a way for someone to say "I point to these things, but I don't care *enough* to keep them alive".
16:29 kaleem joined
FROGGS but that doesnt sound too good 16:30
masak which helps prevent cycles (and thus memory leaks) sometimes.
well, you'd better have a story for what happens when something disappears. :) most of the time it's "meh, who cares?", though. 16:31
16:32 kaleem_ joined
masak the reason it's a design smell is that you need the opt-in from the GC to do this. so it's user code instrumenting the GC; that's a dependency going very much the wrong way. 16:32
FROGGS well, if I have a class in a scalar, and want to use it later, and it is gone... I think I would care it it got GC'd already
masak yeah. which is why you wouldn't use a weak reference for a scalar :)
FROGGS ahh, I see
16:32 kaleem left
masak "doctor, it hurts when I do this" 16:33
FROGGS I said pretty much that last wednesday
16:34 daxim left
FROGGS ( tendinitis ) 16:34
[Coke] I apologize for basically re-asking my last tcl question each day, but:
(wait, let me check the test suite)
16:36 stevan_ left
[Coke] does anyone have an example of nqp's lexdyn? nqp has it in nqp's own source, but not as QAST. 16:36
masak I'm not sure it's a big problem that anonymous classes wouldn't get GC'd. I should think they're often quite long-lived anyway. 16:37
FROGGS dinner & 16:38
masak r: my @a; for ^2 { push @a, Any.new but role {} }; say @a[0].WHAT === @a[1].WHAT
camelia rakudo b2072f: OUTPUT«True␤»
16:38 benabik joined
masak there you go. that anonymous class in there is program-wide, not scope-wide. 16:38
thinking about scopes during compilation and runtime has been one of the most interesting things for me in the past year or two. 16:39
hm, maybe call it "lexical-scope-wide" for clarity.
moritz www.perlmonks.org/?node_id=1039484
16:40 fhelmberger left
masak moritz: I half-expected some insightful reply by you in that thread. 16:41
moritz masak: I'm now thinking about writing one 16:42
16:43 ajr joined
masak I hadn't really thought of () and [] being the "wrong" syntax match for that semantics. 16:43
it still feels pretty right to me.
16:43 ajr is now known as Guest7443, konundra joined
masak & 16:44
16:49 kaleem_ left
moritz www.perlmonks.org/?node_id=1039484 # my reply 16:49
16:51 Guest7443 is now known as ajr_ 16:52 Chillance joined
dalek : ba34a16 | (Tobias Leich)++ | / (4 files):
better handling for ord/chr

This also handles sub call with parens better.		 16:55
16:56 dakkar left 16:58 domidumont left 16:59 domidumont joined 17:03 twigel left 17:04 birdwindupbird left
[Coke] next qast question past had a :scope<keyed> for Var's - what's the equivalent for QAST? 17:28
17:33 xenoterracide joined
xenoterracide so I was wondering... is it possible to provide a grammar and then reverse it? so I provide the perl6 equivalent of the grammar and then I want to output that as the grammar could read it back in (if that makes remote sense) 17:35
[Coke] ah. perhaps an atkey call on a Var.
Didn't jnthn just demo something like that at YAPC?
xenoterracide possibly but I missed half of that talk 17:36
17:37 stevan_ joined, vk_ joined
lizmat jnthn's talk at YAPC::NA had this new concept of "backtions" 17:41
but I don't think it did what you're describing here
lizmat must admit it was going over her head :-) 17:42
17:43 vk_ is now known as vk
[Coke] lizmat: I feel like that often when sixperling. 17:45
invoke() not implemented in class 'QAST::SVal'f.. hurm. 17:46
moritz [Coke]: QAST::SVal() vs. QAST::SVal.new() 17:47
[Coke]: you typed the former, the latter is correct
(most likely) 17:48
colomon actually, I think what xenoterracide describes is pretty much exactly what jnthn++ did
masak yeah. 17:49
[Coke] wow, moritz++'s builtin qast compiler is prety accurate.
lizmat see, I didn't get it at all then
colomon github.com/jnthn/grammar-generative
masak lizmat: conceptually, a grammar is a function that takes a string and returns a match-tree. jnthn wrote some stuff to also make a grammar conceptually into a function that takes a match-tree and returns a string. 17:50
moritz [Coke]: I've made the same error more than once 17:51
(which is an euphemism for "many times")
masak oh, *that* error. 17:52
yes, me too.
it's very tempting, especially when you're writing big nested things.
[Coke] I feel like I am blindly hacking at this half baked nqprx->nqp code. "what did this do? eh, just droip it. maybe it's the default"
lizmat looking at jnthn's grammar-generative: "sub" is by default "our", is it not? wondering why the "sub collect" doesn't have a "my" in front of it 17:53
[Coke] gah. this is such a pita. 17:55
17:55 abnorman left
masak lizmat: 'sub' is by default 'my'. 17:56
...in Perl 6.
in Perl 5, it is certainly by default 'our'.
lizmat ok, another false friend bites the dust
17:56 stevan_ left
masak lizmat: the rationale here is something like this: lower-case-letter things, like variables and sub names and keywords, are lexically scoped. 17:56
lizmat I was thinking of adding some more tests for the Hash[TValue,TKey].new format 17:57
[Coke] 20e
17:57 lucasb joined
lizmat and realize there is still no way to find out the TKey of an existing hash :-( 17:57
masak whereas upper-case things, like classes and grammars and roles, are package-scoped by default.
lizmat masak++
masak lizmat: having worked with lexically scoped subs for several years, I guarantee it's the right default :> 17:58
17:58 stevan_ joined
masak I'm not as sure about classes being 'our'-scoped by default, but hey. 17:58
lizmat I think I agree with that
but that is mainly an implementation problem right now 17:59
class Foo { class Bar } }
the Foo::Bar doesn't really exist (yet), if I remember jnthn correctly, and Bar doesn't know it has a "parent" class Foo 18:00
r: class A { class B {} }; say A::B.new 18:01
camelia rakudo b2072f: OUTPUT«B.new()␤»
lizmat ok, it can find A::B, but since B doesn't know it lives inside of A, the object is blessed with the wrong package
*name
masak well, right and wrong... 18:02
r: class A { class B {}; method foo { B.new } }; say A.foo
camelia rakudo b2072f: OUTPUT«B.new()␤»
masak that's completely right, IMO.
it created a B, because "B" is what it's called inside of that package.
lizmat r: class A { class B {} } say B.new 18:03
camelia rakudo b2072f: OUTPUT«===SORRY!===␤Two terms in a row␤at /tmp/JkJH9jqVtG:1␤------> class A { class B {} } ⏏say B.new␤ expecting any of:␤ statement list␤ horizontal whitespace␤ postfix␤ statement end␤ statement mo…
lizmat r: class A { class B {} }; say B.new
camelia rakudo b2072f: OUTPUT«===SORRY!===␤Undeclared name:␤ B used at line 1␤␤»
masak ...but not outside of it.
cf. relative file paths.
lizmat r: class A { class B {} }; say A::B.new # created outside
camelia rakudo b2072f: OUTPUT«B.new()␤»
masak *nod* 18:04
thing is, it's the same class both time.
[Coke] ok. have partcl so that it can get through test_more.tcl again - have to put the lexical handling for variables back together, but this is progress.
lizmat on the other hand, it shouldn't make a difference where an object is created, or should it?
masak feels weird to have stringification of the name depend on the caller's position in the code...
lizmat: right.
lizmat and I guess it doesn't for execution right now. It's just a problem for introspection, I guess 18:05
masak it feels like a conceptual problem to me.
what *is* the right answer?
lizmat I feel that it should always be A::B, regardless where the object is created 18:06
class A { class B {}} # B feels like syntactic sugar for A::B to me 18:07
masak ok, yes. at least that's a consistent answer. 18:08
lizmat r: class A { class B {}}; class Z { class B {}; say A::B.new; say Z::B.new
camelia rakudo b2072f: OUTPUT«===SORRY!===␤Unable to parse expression in block; couldn't find final '}'␤at /tmp/2UXSclFl2U:1␤------> { class B {}; say A::B.new; say Z::B.new⏏<EOL>␤ expecting any of:␤ method arguments␤ postfix␤ stateme…
lizmat r: class A { class B {}}; class Z { class B {}}; say A::B.new; say Z::B.new
camelia rakudo b2072f: OUTPUT«B.new()␤B.new()␤»
masak lizmat: I spoke with TimToady about this once, and he pointed me to S10:145. 18:10
...yes, S10. a very uncommon synopsis :)
xenoterracide colomon: thanks for the link, basically seems like an awesome idea 18:11
lizmat masak: I don't think this is about globalness, because the classes are in OUR:: anyway 18:12
xenoterracide returns to the dark, musing about how much better and easier it woud have been to solve this problem in p6 18:14
lizmat xenoterracide: take care! 18:15
masak lizmat: hm, maybe. 18:19
lizmat to me, it feels like a NYI… because B still does not have a link back to A internally (jnthn told me), it can not tell the world it is actually A::B yet 18:20
although practically, it already is 18:21
18:22 REPLeffect left, MrMeek-afk left, cosimo left, dalek left, kst left, Woodi left, jferrero left, simcop2387 left, jferrero joined, Woodi joined, cosimo joined, Rotwang joined, MrMeek joined, panchiniak joined, dalek joined, ChanServ sets mode: +v dalek, logie joined 18:23 logie_ left, kst joined
masak lizmat: yes, you might be right. 18:23
lizmat: it would certainly be clearer if it said A::B.
lizmat: I'm not as sure about this case, though:
r: class A { my class B {}; method foo { B.new } }; say A.foo 18:24
camelia rakudo b2072f: OUTPUT«B.new()␤»
masak because the '::' should mean "it's in the 'our' scope of the package".
lizmat even with the my, it should report as being A::B, I think 18:25
the my is just so that outside, you shouldn't be able to make that object
because A::B doesn't exist outside
r: class A { my class B {} }; say A::B.new # should fail 18:26
camelia rakudo b2072f: OUTPUT«Could not find symbol '&B'␤ in method <anon> at src/gen/CORE.setting:10166␤ in any at src/gen/Metamodel.nqp:2504␤ in any find_method_fallback at src/gen/Metamodel.nqp:2492␤ in any find_method at src/gen/Metamodel.nqp:939␤ in block at /tmp/cD3ImzFIg0:1␤␤»…
lizmat r: class A { my class B {} say A::B.new } # should work
camelia rakudo b2072f: OUTPUT«===SORRY!===␤Two terms in a row␤at /tmp/JSl5tAtjfp:1␤------> class A { my class B {} ⏏say A::B.new } # should work␤ expecting any of:␤ statement list␤ horizontal whitespace␤ postfix␤ infix stopper…
lizmat r: class A { my class B {} ; say A::B.new } # should work
camelia rakudo b2072f: OUTPUT«Could not find symbol '&B'␤ in method <anon> at src/gen/CORE.setting:10166␤ in any at src/gen/Metamodel.nqp:2504␤ in any find_method_fallback at src/gen/Metamodel.nqp:2492␤ in any find_method at src/gen/Metamodel.nqp:939␤ in block at /tmp/EiEJbwgu72:1␤␤»…
lizmat hmmm….
18:26 simcop2387 joined
lizmat r: class A { my class B {} ; say B.new } # should work 18:26
camelia rakudo b2072f: OUTPUT«B.new()␤»
masak lizmat: no, I disagree. 'A::B' literally means "an our-scoped B package inside an A package". 18:27
lizmat: note that back around 2008, the Perl 6 design was confusing our-scope and has-scope, using '::' for both. that was a bad idea. 18:28
I think it's a bad idea to muddle 'our' scope and 'my' scope, too.
'::' means 'our' scope.
18:29 sqirrel joined
lue couldn't my class B just be invisible from the outside? (That code example would work but an additional A::B.new on the outside wouldn't) 18:29
lizmat ah, in that sense… then it should be specced like that ?
lue: indeed, that was my point
but I can live with :: in a name indicating ourness 18:30
masak lue: yes, it's already that way.
lue: question is about how to stringify the class name.
lizmat r: class A { my class A::B {} } # does this bomb?
camelia rakudo b2072f: ( no output )
lue masak: OK. I thought your disagreement was about something else. 18:31
lizmat r: class A { my class A::B {} } ; say A::B.new # does this work?
camelia rakudo b2072f: OUTPUT«A::B.new()␤»
lizmat intriguing, I guess the "my" is superseded by the ourness of ::
masak I'd consider that a bug. 18:32
PerlJam that looks like a bug to me
masak submits rakudobug
lue r: class A { my class B {}; say B.WHAT}
camelia rakudo b2072f: OUTPUT«(B)␤»
masak lizmat: in a sense '::' is the ourness accessor. '.' is for has-scoped methods. '!' (within the same class) is for attributes and private methods. we have no twigil/separator for lexical access... maybe because as a rule, looking into other lexical scopes than one's own suggests a gross misunderstanding of lexical scoping, to a first approximation. 18:34
18:35 census joined
masak lizmat: there's an exciting special syntax: $object.MyRole::foo() where '::' doesn't mean "our-scoped". but it's clearly a special case, and quite a forgivable one, IMO. 18:35
(it's for disambiguating when several methods with the same name might have been mixed into a class)
PerlJam masak: ::($foo) doesn't imply "our" either.
masak PerlJam: no, but that's prefix '::' ;) 18:36
I was talking about infix '::'
PerlJam $::($foo) then.
Looks infix to me :)
masak that's a twigil :P
census twigil sigil ... aaah
masak census: "twigil" is short for "secondary sigil". 18:37
census: the sigils are the one you're used to by now: $ @ % and to a lesser extent &
PerlJam masak: I wonder if the specs say as clearly as you did that ":: is the ourness accessor" ? If not, it probably should.
census i know. i get excited when i see all of these new words to learn :)
masak PerlJam: it probably doesn't.
census you taught me sigil so i know that word :)
masak \o/ 18:38
pmichaud I never thought of it as being strictly "ourness"; it just refers to the B symbol in the A package. (I guess that's "ourness".)
lue
.oO(I do believe $::($foo) contains a sigil, twigil, and the elusive trigil)
masak pmichaud: yes :)
pmichaud I think of :: as a postfix operator at times.... A:: is A's package.
thus A::B is "B within A's package"
lizmat pmichaud: but does that constitute "ourness" ? 18:39
pmichaud well, symbols within a package are "our" 18:40
or, put another way, "our" says that a symbol is to be defined within a package, iiuc.
18:42 kaleem joined
lizmat but if we don't specify "our", but *do* specify an infix :: such as A::B, would *that* make it our, or just the fact that it is a symbol defined inside a package 18:44
pmichaud it makes it "our" with respect to "A"
i.e., it's the same as an "our" declaration within A's package scope. 18:45
well, an "our declaration of B"
in the case of: 18:46
18:31 <lizmat> r: class A { my class A::B {} } ; say A::B.new # does this work?
masak oh!
pmichaud I think Rakudo has it right.
masak yes, so notabug.
masak unsubmits rakudobug
pmichaud you created a class B, but you specified that it should be installed in package "A" by using "A::B"
PerlJam Hmm.
pmichaud thus A::B.new works, because there's a B class inside of A's package scope.
nwc10 gosh, masak has backtions too 18:47
lizmat r: class A { my class B {} } ; say A::B.new # but this doesn't work
camelia rakudo b2072f: OUTPUT«Could not find symbol '&B'␤ in method <anon> at src/gen/CORE.setting:10166␤ in any at src/gen/Metamodel.nqp:2504␤ in any find_method_fallback at src/gen/Metamodel.nqp:2492␤ in any find_method at src/gen/Metamodel.nqp:939␤ in block at /tmp/xOrWFC9S4p:1␤␤»…
pmichaud correct, that doesn't work because nothing was installed in A's package.
s/installed/bound/ # more accurately
lizmat ok, I think I get it now… you need to think of classes as objects ;-) 18:48
masak nwc10: "undo" != "reverse" :P
18:48 kaleem left
pmichaud lizmat: yes. 18:48
masak lizmat: definitely.
pmichaud and also note that each class has a separate "package" component, accessed via '::'
thus "A" is the class, while "A::" is the package associated with the class. 18:49
masak r: sub are-you-an-object($o) { "yes" }; class SomeClass {}; say are-you-an-object SomeClass
camelia rakudo b2072f: OUTPUT«yes␤»
masak :)
18:49 lucasb left
pmichaud well, "SomeClass" in this instance is really a type object. :) 18:50
we used to call them protoobjects :)
moritz accidentally his lilbc6
lizmat his what, where, why, how?
moritz I installed valgrind 18:51
and didn't notice I had 'testing' in my sources.list
so it took a new valgrind from testing, and the libc6 along with it
and now my perlbrew'ed perls don't work anymore 18:52
oh, and locales and stdio.h were missing too, for some reason
masak r: sub are-you-a-type-object($o) { !$o.defined }; say are-you-a-type-object class SomeClass {}
pmichaud I'm afk again a while.
camelia rakudo b2072f: OUTPUT«True␤»
lizmat pmichaud++ 18:53
masak there's supposed to be a newer way to test for type-object-ness, but I've forgotten what it is exactly.
pmichaud DEFINED
I think
mebbe not
lue :D and :U, perhaps?
18:54 ssutch joined
lizmat .DEFINITE 18:54
pmichaud .DEFINITE is it.
lizmat++
afk
masak r: sub are-you-a-type-object($o) { !$o.DEFINITE }; say are-you-a-type-object class SomeClass {} 18:55
camelia rakudo b2072f: OUTPUT«True␤»
masak \o/
ssutch got some actions that work w/ my protobuf grammar! github.com/samuraisam/p6-pb/blob/m...Actions.pm 18:58
diakopter ssutch: lolwut 19:03
yoleaux 08:53Z <brrt> diakopter: again, they are hard, like surgery is hard, but it is doable
19:03 pochi joined
ssutch diakopter: ? 19:03
19:04 rindolf left
diakopter ssutch: just quoting from your link :) 19:04
19:05 rindolf joined
ssutch ah, my debug code is not very elegant with any language 19:05
19:14 tomyan joined
dalek ecs: dc6770f | (Elizabeth Mattijsen)++ | S10-packages.pod:
Perl 5 now *does* allow packages with block scoping
 19:17
PerlJam "Perl 5"?!? 19:20
oh
so strange to be talking about perl 5 stuff here :)
19:22 rindolf left, rindolf joined, snearch left
timotimo is it sacrilegious to drill a hole through a round tuit? 19:23
masak you did WHAT?!
:P
19:23 domidumont left
timotimo not yet 19:23
lizmat it's fine as long as you keep the tuits rolling :-) 19:24
19:24 xenoterracide left
tadzik timotimo: to carry it as a necklace? 19:24
timotimo would attach it to my rucksack 19:25
19:25 tomyan left, stevan_ left
PerlJam timotimo: The wooden nickel company sells them with holes already drilled. You could order some like that :) 19:26
19:26 DarthGandalf left
timotimo i gotmine as a present though ;) 19:27
19:27 DarthGandalf joined
masak lizmat: would it be OK to file off the edges of a tuit, making it square? 19:28
lizmat as I said, as long as you keep them rolling :-) 19:29
masak ok, so a square is out, then, I guess. :) 19:30
but a en.wikipedia.org/wiki/Reuleaux_triangle should be OK...
lizmat yup, that should work
19:31 panchiniak left
timotimo k :) 19:31
19:34 vk left
Util FYI, #ps meeting has started 19:35
labster saw your commit, thanks timotimo++ 19:36
19:37 vk joined
timotimo er, which? 19:38
labster Cool.path
timotimo oh yeah :)
labster the other ones are good too
all of them
timotimo i love pleasing people with little bits of work
thanks 19:40
19:40 xenoterracide joined 19:43 tgt left, rindolf left, rindolf joined
dalek rl6-roast-data: dfead28 | coke++ | / (4 files):
today (automated commit)
 19:45
[Coke] S02-lexical-conventions/unspace.rakudo 15 - unspace with postfix pre-dot hyperops 19:46
timotimo [5~compiler release is in a few days, right? 19:47
sorear good * #perl6
[Coke] niecza has climbed back to > 100 failures.
looks like most of the new ones are related to classify & categorize 19:48
dalek kudo-star-daily: b072e16 | coke++ | log/ (5 files):
today (automated commit)
 19:49
moritz not surprising, since lizmat++ has cleaned up some tests there
timotimo i am confused. what hyperopp test fails? i only see it in the summary not on the diff above? 19:50
[Coke] timotimo: not ok 15 - unspace with postfix pre-dot hyperops 19:52
# got: '3 4 5'
# expected: '7 8 9'
colomon [Coke]: yes, lizmat thoroughly broke the classify and categorize tests for niecza
I started to fudge it, then decided it would probably be easier just to implement the missing functionality. 19:53
as I result, you should expect them to be broken for a few days. 19:54
:)
[Coke] niecza has been dirty for 336 days. No rush. :)
coming up on a year.
colomon is pondering taking this, but has no idea where he'd find the time: www.coursera.org/course/optimization 19:55
masak timotimo: yes, Rakudo release is on Thursday.
[Coke] is there a planned star release, to quiet the people who expect such things? 19:56
timotimo what a silly mistake 19:57
i will fox it in about fifteen minutes
masak .oO( because the most important reason we make R* releases is to keep people quiet? ) :P 19:59
lue
.oO( "Is Perl 6 ready in this R* release?" )
 20:01
huf spec an is-ready? sub that shall forever return false 20:04
ssutch is there somewhere that describes how string escapes work in perl 6
moritz ssutch: perlcabal.org/syn/S02.html#Literals
masak huf: '?' is not valid in an identifier. 20:05
huf _damn_
masak also, *hug*.
huf i'll have to look for another character that is, and resembles ? :)
hug?
moritz ssutch: perlcabal.org/syn/S02.html#" is probably the better link
masak hugs huf 20:06
huf :)
moritz hugme: hug masak and huf
hugme hugs masak
huf incidentally, "hug" means "younger sister" in hungarian, and i happen to have one :)
masak .oO( hugarian notation )
huf :)
also <3
PerlJam so ... "hugme sisters masak"? That's some interesting verbing 20:07
lue wishes git had a "timeline" command of some sort, where you could easily get the repo as it was in any commit without a series of "git checkout <hash>;git checkout master;git checkout <other hash>;..." 20:08
masak r: class HeartFactory { method make { "<3" } }; loop { say HeartFactory.make }
flussence er... bisect?
camelia rakudo b2072f: OUTPUT«(timeout)<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤<3␤
huf looks like your heart will go on 20:09
masak forreva
huf not 5?
masak 5 is right out.
benabik lue: Why are you git checkout-ing master between hashes?
lue because you can't go forward in time in a detached HEAD, so far as I could tell :( 20:10
moritz sure you can
you can just 'git log master' to find the hashes
lue (without keeping a copy of `git log` from master on hand to get the SHA1 hashes)
masak sure you can :)
just go 'master~10' or whatever.
moritz or you can git checkout master~15, master~10 etc. 20:11
lue I think what I'd like is the opposite of HEAD^, perhaps HEAD$ :)
moritz masak++ is faster :-)
masak .oO( master++ is fasak )
tadzik :D
ssutch what is .oO()
moritz if git history was always linear, going ahead could work.
masak ssutch: thoughts.
moritz ssutch: comic thought bubbles
ssutch aha
diakopter fastak is maser. mastak is faser.
PerlJam git is .oO( what shoud we use for serious thought bubbles? ) 20:12
masak lue: problem is, "first parent" is highly meaningful in git. "first child", not so much.
r: .say for <m f> X~ <ast as> X~ <ak er> 20:13
benabik master~N, master~N-1, etc is probably the best option.
camelia rakudo b2072f: OUTPUT«mastak␤master␤masak␤maser␤fastak␤faster␤fasak␤faser␤»
lue (it would appear to me the best option for what I want is gitk :/ . I guess I'm just being too lazy to want to count commits from master when I want to start from commit 1, or to look up SHA1 hashes)
benabik Although if you're trying to find some kind of bug, git bisect is really your friend.
ssutch whoa X~
masak ssutch: :D 20:14
20:14 twigel joined
moritz the only thing where I need repeated checkouts is chasing through git blames 20:14
masak ssutch: you can do Xop for any op.
ssutch: any infix op, that is.
ssutch dashot
masak yah
moritz r: say (1, 2, 3) X* (10, 11, 12)
camelia rakudo b2072f: OUTPUT«10 11 12 20 22 24 30 33 36␤»
masak r: .say for 1 X+ 100 X+ 10_000
camelia rakudo b2072f: OUTPUT«10101␤»
masak er :)
r: .say for 1, 2, 3 X+ 100, 200, 300 X+ 10_000, 20_000, 30_000 20:15
camelia rakudo b2072f: OUTPUT«10101␤20101␤30101␤10201␤20201␤30201␤10301␤20301␤30301␤10102␤20102␤30102␤10202␤20202␤30202␤10302␤20302␤30302␤10103␤20103␤30103␤10203␤20203␤30203␤10303␤20303␤30303␤»
masak \o/
ssutch hehehe
diakopter r: say (my $a = 4 X= 5); say $a
camelia rakudo b2072f: OUTPUT«5␤4␤»
20:15 benabik left
diakopter masak: ^ 20:15
masak diakopter: um.
flussence that looks evil
masak diakopter is evil. :)
diakopter: I think it's a precedence thing. 20:16
diakopter: = is item is tighter than X= is list.
so, notabug.
flussence n: say (my $a = 4 X= 5); say $a # I half-expect this one to explode...
camelia niecza v24-78-g9f58f84: OUTPUT«===SORRY!===␤␤Cannot cross with = because list assignment operators are too fiddly at /tmp/uCi_u1U6Gw line 1:␤------> say (my $a = 4 X=⏏ 5); say $a # I half-expect this one to ␤␤Unhandled exception: Check failed␤␤ at /…
moritz r: say (my $a = 4) X= 5; say $a
camelia rakudo b2072f: OUTPUT«5␤4␤»
flussence
.oO( 50% of the time, I'm 100% right )
moritz r: say (my $a = 4) = 5; say $a
camelia rakudo b2072f: OUTPUT«5␤5␤»
masak diakopter: unless you wanted that to say 5 twice...
moritz just plain old rw-ness not being carried correctly
diakopter r: say (my $a = (4 X= 5)); say $a
camelia rakudo b2072f: OUTPUT«5␤5␤» 20:17
masak so, yesabug?
moritz yes
masak but filedalready?
moritz r: 4 X= 5
sorear pmichaud: The CLA has been picked up
camelia rakudo b2072f: ( no output )
moritz masak: I'm nearly sure :-)
iirc with X+= or so
masak ok, good.
20:17 tgt joined
diakopter r: 4 X:= 5 20:19
camelia rakudo b2072f: OUTPUT«Nominal type check failed for parameter 'op'; expected Any but got Mu instead␤ in sub METAOP_CROSS at src/gen/CORE.setting:13994␤ in block at /tmp/cz8I3OQI1e:1␤␤»
nwc10 sorear: yay! www.collectgbstamps.co.uk/images/gb...1098_l.jpg
colomon \o/
20:20 stevan_ joined
dalek ast: e121e0e | (Timo Paulssen)++ | S02-lexical-conventions/unspace.t:
fix tests for unspace/postfix-hyper.
 20:22
20:23 vk left
masak I'd like to make y'all aware of mathwithbaddrawings.com/2013/06/16/...c-tac-toe/ 20:28
for kicks, today's mini-challenge: implement this game in Perl 6.
HN discussion: news.ycombinator.com/item?id=5898506
20:28 tgt left 20:30 preflex left
lue
.oO(The markers alternated *almost* every time in those 9 first pictures)
 20:31
20:31 preflex joined, ChanServ sets mode: +v preflex
timotimo so, afaict "☀".succ isn't specced to be "☁", right? 20:33
rn: "☀".succ 20:34
camelia rakudo b2072f, niecza v24-78-g9f58f84: ( no output )
timotimo rn: "☀".succ.say
camelia niecza v24-78-g9f58f84: OUTPUT«WTF␤»
..rakudo b2072f: OUTPUT«☀␤»
FROGGS WTF? 20:35
[Coke] niecza++
lue what happened to niecza? o.o
flussence it became sentient
tadzik :D 20:36
this is awesome
niecza: do something funny 20:37
camelia niecza v24-78-g9f58f84: OUTPUT«===SORRY!===␤␤Undeclared routines:␤ 'funny' used at line 1␤ 'something' used at line 1␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1502 (die @ 5) ␤ at /home/p6eval/niecza/src/STD.pm6 line 1147 …
moritz masak: nice game :-)
timotimo r: "f".ord.say 20:38
camelia rakudo b2072f: OUTPUT«102␤»
timotimo r: "f".ord.succ.chr
camelia rakudo b2072f: ( no output )
timotimo r: "f".ord.succ.chr.say
camelia rakudo b2072f: OUTPUT«g␤»
timotimo r: "foo9".succ.say 20:43
camelia rakudo b2072f: OUTPUT«fop0␤»
[Coke] ponders a MathSymbols library that makes ÷ etc. work.
r: "÷".succ 20:44
camelia rakudo b2072f: ( no output )
[Coke] r: "÷".succ.say
camelia rakudo b2072f: OUTPUT«÷␤»
masak writes the simulation scaffolding for ELEVATOR 20:45
lue Ah, just when I thought cperl-mode was manageable, it throws a lisp error when I try to type a colon (something about electric terminators, it seems) 20:47
20:48 tomyan joined
FROGGS .oO( ecleptic you say? with a plug? ) 20:50
20:51 araujo left 20:52 araujo joined, araujo left, araujo joined
ssutch how is it again that you tell Grammar.parse to start from a single rule? is it Grammar.parse($text, :rule(name)) 20:52
masak ssutch: yeah, where name is a string.
r: grammar G { rule foo { { die "it worked!" } } }; G.parse("", :rule<foo>) 20:53
camelia rakudo b2072f: OUTPUT«it worked!␤ in regex foo at /tmp/cTkz_7LFCr:1␤ in method parse at src/gen/CORE.setting:10493␤ in block at /tmp/cTkz_7LFCr:1␤␤»
20:54 bluescreen10 left
ssutch r: "\0xc3".perl 20:58
camelia rakudo b2072f: ( no output )
ssutch r: "\0xc3".perl.say
camelia rakudo b2072f: OUTPUT«"\x[0]xc3"␤»
ssutch r: "\0xc3".gist 20:59
camelia rakudo b2072f: ( no output )
ssutch r: "\0xc3".gist.say # derp
camelia rakudo b2072f: OUTPUT«␀xc3␤»
timotimo i have a local patch that makes diverse ranges-with-unicode-things work
will do another spectest run.
sorear r: say "\xC3"
camelia rakudo b2072f: OUTPUT«Ã␤»
sorear r: say "\x[C3]"
camelia rakudo b2072f: OUTPUT«Ã␤»
sorear r: say "\c[0xC3]"
camelia rakudo b2072f: OUTPUT«Ã␤»
20:59 tgt joined
ssutch just trying to get a string literal with unicode/octal escapes from my grammar into the same thing in perl 21:00
21:00 Rotwang left
timotimo ssutch: i suggest "eval" 21:01
sorear ssutch: i suggest "chr" 21:02
masak rn: sub normal { ([+] rand xx 12) - 6 }; say normal
camelia niecza v24-78-g9f58f84: OUTPUT«-0.25535927119448765␤»
..rakudo b2072f: OUTPUT«0.0460164515684127␤»
lue r: say "octal: \o[52] unicode char: \c[ASTERISK] or \c[42] (decimal number)"
camelia rakudo b2072f: OUTPUT«octal: * unicode char: * or * (decimal number)␤»
ssutch this is what im doing: make chr(:16($<xdigit>.join)); 21:03
21:03 pmurias joined
masak this is how Wikipedia suggests I generate (an approximation of) normally distributed values. 21:03
pmurias pmichaud: ping
ssutch and this is what im doing for oct: make chr(:8($<digit>.join));
masak easier than I thought.
timotimo seriously, you have already established the text is harmless and correctly escaped. eval it to get the actual string from it
sorear timotimo: except that eval is slow as hell
timotimo oh?
well, yeah, that's true.
lue are the escape sequences the same as in P6 though? 21:04
sorear and leaks memory
ssutch lue: no
sorear because packfile objects can't be garbage collected
ever
lue so eval() wouldn't help too much anyway.
ssutch right
timotimo i'm sorry in that case 21:05
carry on :)
ssutch heh
sorear r: grammar G { token char { <[0..7]>+ { make chr(:8($/)) } } }; say G.parse("176").ast 21:06
camelia rakudo b2072f: OUTPUT«No such method 'TOP' for invocant of type 'G'␤ in method parse at src/gen/CORE.setting:10493␤ in block at /tmp/uFjTnb7NT1:1␤␤»
sorear r: grammar G { token char { <[0..7]>+ { make chr(:8($/)) } } }; say G.parse("176", :rule<char>).ast
camelia rakudo b2072f: OUTPUT«Nominal type check failed for parameter '$str'; expected Str but got Match instead␤ in sub unbase at src/gen/CORE.setting:5164␤ in regex char at /tmp/mFeP4yBvYh:1␤ in method parse at src/gen/CORE.setting:10493␤ in block at /tmp/mFeP4yBvYh:1␤␤»
sorear r: grammar G { token char { <[0..7]>+ { make chr(:8(~$/)) } } }; say G.parse("176", :rule<char>).ast
21:06 konundra left
camelia rakudo b2072f: OUTPUT«~␤» 21:06
sorear what's not working?
ssutch it's working, far as i can tell 21:07
21:09 sqirrel left 21:12 spider-mario left 21:15 zby_home joined 21:18 woolfy left 21:20 woolfy joined
ssutch ah, so i was doing make $<str-contents-double>>>.flat.join; but i needed make $<str-contents-double>>>.flat>>.ast.join; 21:22
21:22 kaare_ left
timotimo i shattered one of my tuits with a prick punch ~_~ 21:24
[Coke] awl hell. 21:26
timotimo i was told to apply a bit of a hit with a prick punch before applying the drill. turns out the grain was very unfortunately aligned with the surface 21:28
n: say '▁' ... '█' # works locally on rakudo as well 21:29
camelia niecza v24-78-g9f58f84: OUTPUT«▁ ▂ ▃ ▄ ▅ ▆ ▇ █␤»
FROGGS masak: not many attendees tmw :/ 21:30
timotimo++
timotimo github.com/rakudo/rakudo/pull/168 - this is the corresponding pullrequest
FROGGS: i didn't notice there was different kinds of tuits so i shattered the "perl community" tuit ... very symbolic :(
FROGGS hehe
timotimo: np, I maybe have spare tuits :o) 21:31
timotimo s/3/2/ in my zip code, btw ~_~ 21:32
s:g/3/2/ even
FROGGS ohh
masak FROGGS: I guess most people got their fill the first time :)
FROGGS: also, we didn't pull the reddit lever this time around.
FROGGS: I think tomorrow will be very nice, though. 21:33
FROGGS masak: yeah, I recognized
masak all the more attention for those who'll show up ;)
FROGGS true
masak right now, I have an elevator which completely refuses to pick people up, so they all take the stairs :) 21:34
sorear ach sentient elevators
I read a book about those once D: 21:35
FROGGS okay, so it is at 8pm in our tz?
21:35 pecastro joined
masak like this: gist.github.com/masak/3ca7e6d6b06c325bdf03 :) 21:35
FROGGS: aye. till midnight.
hm, person Y does get tired a lot. that doesn't look right :)
21:36 pecastro_ left
sorear floor 0? o_O 21:36
masak ah, that's better. :)
sorear oh, right
masak sorear: no, not sentient elevator. I simply haven't implemented the elevator yet, so conceptually it's just an empty elevator shaft. 21:37
lue masak: may I ask what you're working on?
dalek kudo/nom: f0ff25b | (Timo Paulssen)++ | src/core/operators.pm:
implement sequences between 1-codepoint strings
kudo/nom: ce5be8f | (Tobias Leich)++ | src/core/operators.pm:
Merge pull request #168 from timo/one-codept-str-sequences

implement sequences between 1-codepoint strings
masak lue: github.com/masak/workshop/issues/16
mst sorear: .uk does 'ground floor, 1st floor, 2nd floor' where .us does '1st floor, 2nd floor, 3rd floor'
sorear mst: yeah, I remember that now (.de, but same principle)
masak mst: weirdly, though, the floor numbering in my stairwell is: -1, 0, 2..6 21:38
21:38 rindolf left
timotimo does anyone else remember some LHF i could try to shove into the next compiler release? 21:38
perigrin mst: there is one building here in town that has 1st floor, ground floor, 3rd floor ... 21:39
but it sank after they built it so ... it's unique.
21:39 tomyan left
masak timotimo: maybe rt.perl.org/rt3/Ticket/Display.html?id=118339 ? 21:39
sorear how are we going to handle compiler releases with jvm?
'1 eqv 3' # works 21:40
'use v6; 1 eqv 3' # dies in eqv
masak timotimo: or rt.perl.org/rt3/Ticket/Display.html?id=118325 , or rt.perl.org/rt3/Ticket/Display.html?id=118231 , rt.perl.org/rt3/Ticket/Display.html?id=118063 , or rt.perl.org/rt3/Ticket/Display.html?id=117997 21:41
FROGGS hehe
masak I could go on.
but at least one of those ought to be LHF.
timotimo thank you, masak :)
masak kowtows
[Coke] sorear: I'd like to imagine the jvm and parrot releases are cut at (near) the same time. 21:43
masak me too. 21:45
21:54 vk joined
sorear i'm going to blame the multi dispatch cache again 21:56
FROGGS :/
labster timotimo is fixing all my bugs
sorear yeppp
and it's a bug I introduced, too
labster rn: say "あ" ... "お" 21:57
camelia niecza v24-78-g9f58f84: OUTPUT«あ ぃ い ぅ う ぇ え ぉ お␤» 21:58
..rakudo b2072f: OUTPUT«(timeout)»
timotimo what does it mean for an operator that its .prec<prec> has @ or : in it?
FROGGS sorear: ohh, so you are the one who does bugs?? :P
sorear
.oO( use iroha; )
21:59 vk left
sorear FROGGS: i've been fixing bugs in rak-jvm 21:59
FROGGS timotimo: example?
sorear: I know, t'was a joke
timotimo FROGGS: i have none, that's my problem :)
dalek p: f3c828f | sorear++ | src/vm/jvm/runtime/org/perl6/nqp/sixmodel/reprs/MultiCacheInstance.java:
Multi cache should be remembering the definedness of the value, not its container
 22:00
sorear that fixes t/01-sanity/06-eqv.t
timotimo yays! :)
FROGGS: i'm not sure why i got moarvm tuits from you; were you gently nudging me in the direction of moarvm work? :3 22:01
FROGGS *g*
timotimo: I just gave you one from every kind 22:02
timotimo oke :)
actually, two moarvm, two rakudo :)
FROGGS true 22:03
timotimo i don't understand how the prec thing works, why it would replace = with @= for tighter and = with := for looser 22:04
22:04 pjcj left
sorear nqp: my $int := 1; say(nqp::isnanorinf($int)) 22:04
camelia nqp: OUTPUT«0␤»
sorear timotimo: because '@' lt '=' and ':' gt '=', or maybe the other way around
22:04 pmurias left
sorear timotimo: precs are compared as strings 22:04
masak r: say sort * leg *, '@=', '=', ':=' 22:05
camelia rakudo b2072f: OUTPUT«:= = @=␤»
timotimo oh, okay
sorear nqp: my $int := 1; nqp::unbox_n($int) 22:06
camelia nqp: OUTPUT«Can only use repr_unbox_num on a SixModelObject␤current instr.: '' pc 53 ((file unknown):169684796) (/tmp/8Cmdy3NYj3:1)␤»
22:06 census2 joined
timotimo i'd like to make div 0 create a Failure object instead of dieing, but it comes from C code and uses Parrot_ex_throw_from_c_args and I don't know about it 22:08
FROGGS r: try nqp::div_i(42, 0) 22:13
camelia rakudo b2072f: ( no output )
sorear just pushed a fix for array indexing. we are now successfully running all of the files in t/01-sanity, and failing only four tests all in 99-test-basic.t 22:14
FROGGS r: my $value = try nqp::div_i(42, 0); say "alive"; say $value
camelia rakudo b2072f: OUTPUT«alive␤Nil␤»
sorear all related to eval
suspect eval is broken, will investigate later
timotimo r: say &infix:<%>.pred; multi infix:<%>($lhs, Str $rhs) is looser(&infix:<xx>) { $lhs.join: $rhs }; say &infix:<%>.pred;
camelia rakudo b2072f: OUTPUT«No such method 'pred' for invocant of type 'Sub+{<anon>}+{Precedence}'␤ in block at /tmp/m6bsiqZ1s7:1␤␤»
timotimo r: multi infix:<%>($lhs, Str $rhs) is looser(&infix:<xx>) { $lhs.join: $rhs }; say &infix:<%>.pred; 22:15
camelia rakudo b2072f: OUTPUT«No such method 'pred' for invocant of type 'Sub+{<anon>}+{Precedence}'␤ in block at /tmp/Ow7R6fA_Tz:1␤␤»
FROGGS r: my $value = (try nqp::div_i(42, 0)) || $!; say "alive"; say $value # timotimo, something like that?
camelia rakudo b2072f: OUTPUT«alive␤Divide by zero␤ in block at /tmp/EFc61I6Nm7:1␤␤»
timotimo sounds like a good idea.
FROGGS dunno if it is
that is some quick hack where a little pmichaud-ish voice in the back of my head screams words I dont understand 22:16
22:16 zby_home left
timotimo heh. 22:17
labster looking at the IO stuff, it looks there is usually a catch block to produce a typed error.
timotimo a spectest shall show.
ah, a typed exception may be cleverer.
r: my $value = (try nqp::div_i(42, 0)) || $!; say "alive"; say $value.WHAT; 22:18
camelia rakudo b2072f: OUTPUT«alive␤(X::AdHoc)␤»
timotimo should it do the same for /, too?
r: 10 / 0;
camelia rakudo b2072f: ( no output )
timotimo no, of course not. that just gets NaN
FROGGS r: 10.1 / 0 # there was another issue 22:19
camelia rakudo b2072f: ( no output )
FROGGS hmmm
labster r: 10.1 / 0.0
camelia rakudo b2072f: ( no output )
FROGGS r: 10.1 / 0e0 # there was another issue
camelia rakudo b2072f: OUTPUT«Divide by zero␤ in sub infix:</> at src/gen/CORE.setting:3892␤ in sub infix:</> at src/gen/CORE.setting:3054␤ in sub infix:</> at src/gen/CORE.setting:3242␤ in sub infix:</> at src/gen/CORE.setting:3054␤ in block at /tmp/oP5GddVEuy:1␤␤»
timotimo r: try 10.1 / 0e0; say $!.WHAT 22:20
camelia rakudo b2072f: OUTPUT«(X::AdHoc)␤»
timotimo now i get a type check for return value
because the div is written to "returns Int" 22:21
FROGGS from a performance perspective: is it wiser to check for zeroness and create the failure instead on try-ing?
timotimo hm, you think try is too much overhead?
labster X::Numeric::DivideByZero ?
FROGGS timotimo: I'd think so when looking at its ast 22:22
timotimo maybe try with $! ~~ "Divide by zero" is indeed less performant than checking for 0
labster I thought Failure was allowed to pass those type checks.
timotimo yeah, apparently not :(
FROGGS r: sub a returns Int { die "abc" }; say a.WHAT 22:23
camelia rakudo b2072f: OUTPUT«abc␤ in sub a at /tmp/fUuvoyzCZQ:1␤ in block at /tmp/fUuvoyzCZQ:1␤␤»
sorear nqp::if, nqp::eq_i
timotimo oh, i'll have to die $! instead of returning $! perhaps?
FROGGS r: sub a returns Int { Failure.new }; say a.WHAT
camelia rakudo b2072f: OUTPUT«Not enough positional parameters passed; got 1 but expected 2␤ in method new at src/gen/CORE.setting:10138␤ in sub a at /tmp/c5J5M2FBkL:1␤ in block at /tmp/c5J5M2FBkL:1␤␤»
sorear actually just nqp::if
wait. this will break inlining, won't it.
timotimo sorear: don't i have to nqp::decont, too?
is nqp::if and nqp::decont faster than just == 0? 22:24
sorear timotimo: the binder automatically decontainerizes your arguments
unless they're marked rw, \, or copy
timotimo in this case, they are \
how do i nqp::if? 22:25
nqp::if(b, {}, { die "foo" }); ?
FROGGS r: nqp::if( 1, say(42) ) 22:26
camelia rakudo b2072f: OUTPUT«42␤»
FROGGS r: nqp::if( 0, say(42) )
camelia rakudo b2072f: ( no output )
labster sorear: does that mean that this line has redundant unboxing: nqp::copy(nqp::unbox_s($from), nqp::unbox_s($to));
timotimo r: nqp::if( 0, say(42), say("what") )
22:26 dmol left
camelia rakudo b2072f: OUTPUT«what␤» 22:26
timotimo neat.
r: nqp::if( 0, 0, say("what") )
camelia rakudo b2072f: OUTPUT«what␤»
sorear labster: no, boxing and containerization are different 22:27
labster Huh. I was trying to figure out why nqp::mkdir doesn't unbox its arguments, and yet still works. 22:28
Well, maybe it does in nqp, but I mean in mkdir() in rakudo.
sorear QAST automatically adds unbox opcodes when necessary for type control 22:29
so I guess technically you can do nqp::copy($from, $to)
but the generated code contains calls to Ops.unbox_s()
...I'm talking about the JVM version of QAST::Compiler, take this with a grain of solt
salt
sure enough, eval("2") dies on jvm 22:31
timotimo hm
sorear huh. eval is failing for the same reason the backtraces are failing: CodeRef not getting hllized 22:33
timotimo r: say 10.1 / 0.0 22:34
camelia rakudo b2072f: OUTPUT«Divide by zero␤ in sub infix:<div> at src/gen/CORE.setting:3431␤ in sub infix:<div> at src/gen/CORE.setting:3059␤ in method floor at src/gen/CORE.setting:7981␤ in method Str at src/gen/CORE.setting:7999␤ in method Str at src/gen/CORE.setting:876␤ in method gi… 22:35
timotimo r: my $val = 10.1 / 0.0; say $val.WHAT; say "alive"
camelia rakudo b2072f: OUTPUT«(Rat)␤alive␤»
timotimo r: my $val = 10.1 / 0e0; say $val.WHAT; say "alive"
camelia rakudo b2072f: OUTPUT«Divide by zero␤ in sub infix:</> at src/gen/CORE.setting:3892␤ in sub infix:</> at src/gen/CORE.setting:3054␤ in sub infix:</> at src/gen/CORE.setting:3242␤ in sub infix:</> at src/gen/CORE.setting:3054␤ in block at /tmp/9t914SYaRH:1␤␤»
timotimo ah, ok, that's because 0.0 is a Rat, but 0e0 is a Num, yes?
sorear yes
timotimo i can't find a test file where that could fit 22:38
there isn't even a DivideByZero exception
FROGGS well, you can add one 22:39
timotimo sure
FROGGS ahh, you mean the text you get now?
this can be in nqp or parrot
22:39 pjcj joined
timotimo i didn't mean that. that comes from nqp.ops, which is C/parrot code aiui 22:40
FROGGS I'm sorry, I'm almost sleeping 22:44
timotimo FROGGS: how much did shipping the tuits cost you btw? 22:45
FROGGS I dont know, 1.45€ ? something like that
timotimo hm. that's not much for deutsche post i believe 22:46
dalek : d1ae650 | (Tobias Leich)++ | / (3 files):
refactor identifier's ast
 22:47
FROGGS well, it is just a letter
timotimo how do i return a Failure from an exception? Failure.new(X::Numeric::DivideByZero.new)?
oh, with fail it seems
FROGGS r: say 78900 / 40451 22:48
camelia rakudo b2072f: OUTPUT«1.950508␤»
FROGGS r: sub a { X::AdHoc.new( :payload<test> ) }; say a.WHAT
camelia rakudo b2072f: OUTPUT«(X::AdHoc)␤»
FROGGS r: sub a { X::AdHoc.new( :payload<test> ) }; say a
camelia rakudo b2072f: OUTPUT«test␤␤Error while creating backtrace: No such method 'backtrace' for invocant of type 'Any'␤ in method new at src/gen/CORE.setting:8731␤ in method new at src/gen/CORE.setting:8728␤ in method backtrace at src/gen/CORE.setting:8865␤ in code at src/gen/CORE.setti…
timotimo r: sub a { fail X::AdHow.new( :payload<test> ) }; say a 22:49
camelia rakudo b2072f: OUTPUT«Could not find symbol '&AdHow'␤ in method <anon> at src/gen/CORE.setting:10166␤ in any at src/gen/Metamodel.nqp:2504␤ in any find_method_fallback at src/gen/Metamodel.nqp:2492␤ in any find_method at src/gen/Metamodel.nqp:939␤ in sub a at /tmp/aJ_tV7Gcuu:1␤ i…
timotimo r: sub a { fail X::AdHoc.new( :payload<test> ) }; say a.WHAT
camelia rakudo b2072f: OUTPUT«(Failure)␤»
timotimo r: sub a { fail X::AdHoc.new( :payload<test> ) }; say a
camelia rakudo b2072f: OUTPUT«test␤ in method gist at src/gen/CORE.setting:10161␤ in method gist at src/gen/CORE.setting:893␤ in sub say at src/gen/CORE.setting:11047␤ in block at /tmp/ePH1OeQI22:1␤␤»
timotimo r: sub a { fail X::AdHoc.new( :payload<test> ) }; say a; say alive
camelia rakudo b2072f: OUTPUT«===SORRY!===␤Undeclared routine:␤ alive used at line 1␤␤»
timotimo r: sub a { fail X::AdHoc.new( :payload<test> ) }; say a; say "alive"
camelia rakudo b2072f: OUTPUT«test␤ in method gist at src/gen/CORE.setting:10161␤ in method gist at src/gen/CORE.setting:893␤ in sub say at src/gen/CORE.setting:11047␤ in block at /tmp/A8Svh5MpFw:1␤␤»
timotimo mhm.
dalek p: 6a5fb75 | sorear++ | src/vm/jvm/runtime/org/perl6/nqp/runtime/Ops.java:
foreign_transform_code was omitted from hllconfig setup, fixed.
 22:53
sorear unless I've broken something new, we are now passing all of t/01-sanity/*.t
timotimo oh yeah! :)
and backtraces will now work, too?
FROGGS wow, sorear++
sorear timotimo: testing now
no.
timotimo aaw. well, still impressive :) 22:54
FROGGS gnight all 22:55
22:55 FROGGS left 22:56 btyler left
pmichaud was there discussion on pull request 168? 23:00
timotimo pmichaud: only on irc
pmichaud looking
timotimo well, only about the things it should do; yesterday at about the same time as now today 23:01
not between my proposing the PR and it being merged
pmichaud I had already fixed it.
timotimo you ... had?
timotimo looks 23:02
pmichaud well, I had fixed the block case.
timotimo ah
pmichaud The single-character case... I want to see the spec for it.
timotimo ah, my pull request was about any-one-codepoint-unicode-strings for both endpoints 23:03
i'll dig it up for you real quick
there are test cases for it, fwiw
pmichaud looking
okay, found the spec instance 23:04
23:04 fgomez joined
timotimo ah, good 23:04
sorear pmichaud: hi
pmichaud I don't like the "unisuccpred" method, though.
seems like the sequence operator ought to just use the closure directly.
for that matter, I'm not a fan of 'succpred' either. 23:05
pmichaud looks
timotimo it's definitely not optimized (it shouldn't go through .ord.succ.chr
but it does use the closure that's returned from succpred
rather than go through it on every iteration
if i understand you correctly
pmichaud hmmmm
okay, I read it wrong the first time. 23:06
still looking.
okay, on re-look I can accept this patch. 23:07
still might want some refactoring but it'll work for now :)
timotimo do you have an intuition where the "current" ordinal value could be stored, so that internally it can just increase the ords and create the chrs only each step of the way? 23:08
23:08 ajr_ left
timotimo that's the refactoring that would be obvious to me 23:08
is there a way in perl6 to specify "when decoding this file using utf8, please insert the replacement character on error rather than exploding"? 23:10
pmichaud (current ordinal value) at the moment I think that'd be a premature optimization 23:11
timotimo OK
pmichaud the sequence operator needs some reworking anyway, and perhaps we need to refactor given/take, so I'd not put too much effort into making it overly efficient.
but clearly the closure that is returned by unipredsucc can reference any local variable it wants :) 23:12
timotimo i'm confused about my failure things. if i call fail from a sub, it should return from the sub, giving a Failure object as result, yes? 23:13
sorear pmichaud: how paranoid should I be about not breaking the parrot build while I hack jvm? 23:14
23:14 census2 left
timotimo could "returns Int" cause a Int.new to be called on the return value? 23:14
sorear should I make a jvm branch and merge it daily after testing on parrot, or so?
timotimo r: sub foo returns Int { fail "what" }; foo;
camelia rakudo b2072f: OUTPUT«what␤ in method sink at src/gen/CORE.setting:10169␤ in method BUILDALL at src/gen/CORE.setting:801␤ in method bless at src/gen/CORE.setting:743␤ in method new at src/gen/CORE.setting:728␤ in method new at src/gen/CORE.setting:726␤ in sub foo at /tmp/DzU42RSGh…
timotimo r: sub foo { fail "what" }; foo;
camelia rakudo b2072f: OUTPUT«what␤current instr.: 'throw' pc 347557 (src/gen/CORE.setting.pir:151689) (src/gen/CORE.setting:8887)␤called from Sub 'sink' pc 379739 (src/gen/CORE.setting.pir:164183) (src/gen/CORE.setting:10169)␤called from Sub 'MAIN' pc 381 (src/gen/perl6.pir:147) (src/main.nqp …
timotimo yeah, that seems to be it 23:15
i don't want to get rid of "returns Int" :(
pmichaud sorear: be fairly paranoid about breaking the parrot build.
timotimo: there's a (known) bug there. Type checking return values is supposed to be transparent to failure objects. 23:17
timotimo mhm
should i remove the "returns" and add an # XXX?
pmichaud if you like.
sorear pmichaud: mmk
timotimo does it seem like a big undertaking to fix the return value typechecks?
pmichaud (parrot build) we have to make sure that that the compiler continues to pass spectests on parrot; i.e., we don't want any regressions in the main branch
23:18 konundra joined
timotimo unfortunately i have no idea where to look >_> 23:19
sorear will TPF tell me when they get the CLA? 23:22
pmichaud: does it seem reasonable to make a jvm-bugfixes branch, then spectest it and merge it every day or so?
timotimo what is 0/0 and 0 div 0 supposed to give? 23:25
rn: say 0/0;
camelia rakudo b2072f: OUTPUT«Divide by zero␤ in sub infix:<div> at src/gen/CORE.setting:3431␤ in sub infix:<div> at src/gen/CORE.setting:3059␤ in sub DIVIDE_NUMBERS at src/gen/CORE.setting:8088␤ in sub infix:</> at src/gen/CORE.setting:8220␤ in sub infix:</> at src/gen/CORE.setting:3054␤ …
..niecza v24-78-g9f58f84: OUTPUT«0␤»
timotimo rn: say 0 div 0;
camelia niecza v24-78-g9f58f84: OUTPUT«Unhandled exception: System.DivideByZeroException: Division by zero␤ at Builtins.divop (Niecza.Constants c, Int32 opc, Niecza.Variable a1, Niecza.Variable a2) [0x00000] in <filename unknown>:0 ␤ at CORE.cbe06ccd-d537-4516-b9ac-0ee3fe3f1675.C1472infix:<di…
..rakudo b2072f: OUTPUT«Divide by zero␤ in sub infix:<div> at src/gen/CORE.setting:3431␤ in sub infix:<div> at src/gen/CORE.setting:3059␤ in block at /tmp/i3yfqxPuaw:1␤␤»
sorear timotimo: how about 31415926 23:27
23:27 benabik joined
sorear hmm, small problem with the lexotics. they get null stables, so decontainerization blows up when you try to call one 23:28
timotimo i wonder if it would be beneficial to add multi candidates where the second argument is fixed at 0 23:31
rather than check for 0 inside the body of the sub
sorear no 23:32
timotimo oke
sorear that would completely disable the multidispatch cache :/
timotimo well, anyway... that's what my code does now and i'm ready to issue a PR
pmichaud sorear: (jvm-bugfix-branch) sure, branch+merge is generally a sound strategy
timotimo i'd be interested for someone to tell me where exactly to place tests for that exception & behavior 23:33
is it okay to use X::Numeric::DivisionByZero for both div and /? because one is for Num and the other for Int ...
sorear pmichaud: TPF-CLA-feedback? 23:35
timotimo: yes, imo
actually I'd kind of prefer dividing a Num by zero to return Inf
in accordance with the standard IEEE 754 semantics
timotimo r: say 10 / 0.Num; 23:36
camelia rakudo b2072f: OUTPUT«Divide by zero␤ in sub infix:</> at src/gen/CORE.setting:3892␤ in sub infix:</> at src/gen/CORE.setting:3054␤ in sub infix:</> at src/gen/CORE.setting:3242␤ in sub infix:</> at src/gen/CORE.setting:3054␤ in block at /tmp/KfqEoXKkaR:1␤␤»
timotimo i can make it do that, sure. +Inf and -Inf i suppose?
r: say (-0).perl
camelia rakudo b2072f: OUTPUT«0␤»
timotimo r: say (-0e0).perl
camelia rakudo b2072f: OUTPUT«-0e0␤»
sorear r: say 1 / 1e-309
camelia rakudo b2072f: OUTPUT«Inf␤»
sorear r: say 0
camelia rakudo b2072f: OUTPUT«0␤»
sorear r: say 1 / 0
camelia rakudo b2072f: OUTPUT«Divide by zero␤ in sub infix:<div> at src/gen/CORE.setting:3431␤ in sub infix:<div> at src/gen/CORE.setting:3059␤ in method floor at src/gen/CORE.setting:7981␤ in method Str at src/gen/CORE.setting:7999␤ in method Str at src/gen/CORE.setting:876␤ in method gi… 23:37
timotimo 0 is a Rat though
sorear r: say 1 / 0e0
camelia rakudo b2072f: OUTPUT«Divide by zero␤ in sub infix:</> at src/gen/CORE.setting:3892␤ in sub infix:</> at src/gen/CORE.setting:3054␤ in sub infix:</> at src/gen/CORE.setting:3242␤ in sub infix:</> at src/gen/CORE.setting:3054␤ in block at /tmp/sPwuYk4Ne5:1␤␤»
sorear / already returns Inf if you overflow a Num
timotimo indeed. it would be sensible.
sorear (see what I did there, passing a denormal? >_>)
pmichaud sorear: there's not normally direct feedback from TPF to you, but they'll often tell me when a CLA arrives and update the shared spreadsheet, so I'll let you know when it comes in. 23:38
(assuming I hear about it)
timotimo pmichaud: ideas about division by num-0?
sorear pmichaud: great
ssutch r: say -Inf
camelia rakudo b2072f: OUTPUT«-Inf␤»
timotimo should it also return Inf if i divide by rat-0?
pmichaud timotimo: I think the spec indicates it returns failure.
ssutch r: say Nan
camelia rakudo b2072f: OUTPUT«===SORRY!===␤Undeclared name:␤ Nan used at line 1␤␤» 23:39
ssutch r: say NaN
camelia rakudo b2072f: OUTPUT«NaN␤»
pmichaud actually, the spec is somewhat specific here.
timotimo do you have a linky for me? 23:40
pmichaud perlcabal.org/syn/S03.html#Multipli...precedence
23:41 mrallen1 joined
pmichaud if either arg is Num, then a divisor of zero results in +/- Inf or NaN 23:41
timotimo ah, good.
i need to handle +/-0 there specifically?
i mean, 1 / +0 vs 1 / -0?
pmichaud from S03: "Note however that these are not actually the native IEEE non-numbers, they are undefined values of the "unthrown exception" type.... 23:42
timotimo r: say NaN ~~ Failure;
camelia rakudo b2072f: OUTPUT«False␤»
timotimo so ... fail NaN?
sorear bah Failure
sorear puts the jvm stuff on hold for today
pmichaud r: say NaN.defined 23:43
camelia rakudo b2072f: OUTPUT«True␤»
timotimo how about i inch the implementation a bit closer to the spec now by turning the div by zero into a soft exception and later (probably tomorrow, perhaps after the compiler release) turn it into proper +/-Inf and NaN? 23:44
23:44 mrallen1 left, cooper left
pmichaud soft exception is okay 23:44
not exactly sure what a "proper +/-Inf and NaN" are, or if Rakudo has those yet :)
r: say -Inf.defined 23:45
camelia rakudo b2072f: OUTPUT«-1␤»
timotimo wait what? :)
pmichaud uhhhhhhhhhh
timotimo hahaha :)
r: say (-Inf).defined
camelia rakudo b2072f: OUTPUT«True␤»
pmichaud thus my point :)
ah, yes.
timotimo r: say (-1...1).perl
camelia rakudo b2072f: OUTPUT«(-1, 0, 1).list␤»
timotimo mhm, mhm.
pmichaud anyway, there's still some slushiness in the implementation and definition of Inf and NaN that needs some freezing up 23:46
masak 'night, #perl6 23:48
timotimo mhm 23:49
sorear o/ masak 23:50
timotimo looking forward to coming home to an apartment heated up by the repeated compiles and spectests run on the machine in it >_< 23:53
now i get a division by zero when building the RESTRICTED.setting m) 23:55
23:56 risou_awy is now known as risou

Powered by Cro and the Raku® Programming Language.

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

Thank you!