notviki This doesn't look like a bug to me. Someone with more understanding of itemized things cares to reply? rt.perl.org/Ticket/Display.html?id=130395 06:38
samcv well they're both Arrays 07:19
so I can see why it's saying they are the same… i mean item forces to item context and I think for an array it doesn't change it enough for is-deeply to fail. so either need to tweak is-deeply or make it identify as its own type of object i think 07:26
m: use Test; is-deeply (1,3,(1,3)).List, (1,3,(1,3)).Seq 07:31
camelia rakudo-moar a7bf1e: OUTPUT«ok 1 - ␤»
samcv m: use Test; is-deeply (1,3,(1,3)).List, (1,3,(1,3)).item
camelia rakudo-moar a7bf1e: OUTPUT«ok 1 - ␤»
samcv m: use Test; is (1,3,(1,3)).List ~~ Seq, True
camelia rakudo-moar a7bf1e: OUTPUT«not ok 1 - ␤␤# Failed test at <tmp> line 1␤# expected: 'True'␤# got: 'False'␤»
samcv m: say item ~~ Array 07:32
camelia rakudo-moar a7bf1e: OUTPUT«===SORRY!=== Error while compiling <tmp>␤Expected a term, but found either infix ~~ or redundant prefix ~␤ (to suppress this message, please use a space like ~ ~)␤at <tmp>:1␤------> say item ~~⏏ Array␤»
samcv m: say so item ~~ Array
camelia rakudo-moar a7bf1e: OUTPUT«===SORRY!=== Error while compiling <tmp>␤Expected a term, but found either infix ~~ or redundant prefix ~␤ (to suppress this message, please use a space like ~ ~)␤at <tmp>:1␤------> say so item ~~⏏ Array␤»
samcv oh item is a sub not a type 07:33
notviki, there is not any bug
notviki, i think what happens is is-deeply examines them in the same context and they are the same context. 07:39
m: use Test; is-deeply item([1,2,3]), [1,2,0]
camelia rakudo-moar a7bf1e: OUTPUT«not ok 1 - ␤␤# Failed test at <tmp> line 1␤# expected: $[1, 2, 0]␤# got: $[1, 2, 3]␤»
samcv i am thinking if we changed is-deeply we'd probably fail some roast tests 07:40
notviki, RE this rt.perl.org/Ticket/Display.html?id=117683 did we used to use icu on mac? 07:44
"Note that these failures are also happening on rakudo-moar" [Coke]_ didn't say JVM but still pretty sure we didn't use icu? unless it's a mac specific thing which no clue why 07:45
[Tux] This is Rakudo version 2016.12-107-ga7bf1eab6 built on MoarVM version 2016.12-19-ge2ffc358 07:50
csv-ip5xs 3.035
test 13.899
test-t 5.523
csv-parser 14.309
samcv [Tux], help
[Tux] 42 07:51
samcv :D
what system do you run that benchmark on?
[Tux] Linux 3.16.7-53-desktop [openSUSE 13.2 (Harlequin)] HP Z220CMT Xeon(R) CPU E3-1245 V2 @ 3.40GHz/1963(8) x86_64 11913 Mb
samcv will doesn't-warn catch die's? 07:55
i'm kind of wondering why the spec says we return <control-0000> for \0 since it technically has no name atm 08:14
well prolly never will
but unicode has it listed as <control>
codepoints aren't usually labeled with same number based on their codepoint though
that's what the synopsis says and it's in roast 08:15
seems odd though
but i guess maybe it stems from no repeat names? but we have it return CJK_Ideograph for those unless we also specify to return the codepoenit in the name too
m: "\0".uniname.say 08:17
camelia rakudo-moar a7bf1e: OUTPUT«NULL␤»
samcv i can fix that though, at least returning the unicode 1 names 08:22
lizmat, can I just send email to [email@hidden.address] and if it has the right title it will add to the ticket? 08:42
lizmat samcv: yes, that's the idea
basically a reply to all, and removing all of the other ones, is what I usually do 08:43
samcv cool
how do i get special powers for RT 08:44
trying to look through some unicode bugs and would be nice to do some things with them other than just reply 08:46
ugh why is it not instant sending by email. oh well 08:47
lizmat samcv: I think [Coke]_ can help you there
samcv cool
lizmat and/or masak perhaps
samcv not sure if perl foundation sent any email to whoever they should. they just told me to notify perl 6 admins they got my CLA 08:48
didn't imply them sending anybody anything. should i reply back and say to notify X person?
whoever that is
m: 7 ~ "\x[308]" x 100_000 08:51
camelia rakudo-moar a7bf1e: OUTPUT«WARNINGS for <tmp>:␤Useless use of "~" in expression "7 ~ \"\\x[308]\" x" in sink context (line 1)␤»
dalek kudo/nom: 005166e | lizmat++ | src/core/List.pm:
Sorting an uninited List should return a new List
 08:59
kudo/nom: 340bc90 | lizmat++ | src/core/Rakudo/Internals.pm:
Make sorting on 0/1 element lists faster

  - by adding a single check on # elements
  - makes them 10% (), 15% ({$^a cmp $^b}) and 40% ({$^a.Str}) faster		 09:35
lizmat afk&
samcv .ask [Coke]_ can i get special powers for rt.perl.org? need to triage some unicode bugs and rename some titles to be the real cause of them 09:39
yoleaux2 samcv: I'll pass your message to [Coke]_.
samcv .ask [Coke] can i get special powers for rt.perl.org? need to triage some unicode bugs and rename some titles to be the real cause of them
yoleaux2 samcv: I'll pass your message to [Coke].
samcv lizmat, the emails take so long to show up :( 09:41
night everybody. see you in the morning 10:47
timotimo night samcv 11:01
samcv night! 11:02
dalek kudo/nom: 54cc06b | lizmat++ | src/core/Supply.pm:
Make Supply.sort a multi for faster dispatch
 12:08
kudo/nom: 4724bd6 | lizmat++ | src/core/Rakudo/Internals.pm:
Make .sort on 2 elem lists about 50% faster

  - because we only need to compare once and not set up a lot of stuff
p: 4a950e1 | samcv++ | tools/build/MOAR_REVISION:
Bump Moar version to get fix for RT #122471 and #122470

return <control-0000> for \0 and other controls RT: rt.perl.org/Ticket/Display.html?id=122471
  rt.perl.org/Ticket/Display.html?id=122470
We now pass several tests we were not passing before in uniname.t		 13:55
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=122471
Link: rt.perl.org/rt3//Public/Bug/Displa...?id=122470
samcv pull to bump the version for rakudo too: github.com/rakudo/rakudo/pull/976 14:00
dalek kudo/nom: 8163113 | samcv++ | tools/build/NQP_REVISION:
Bump NQP to get the latest Moar fixing RT #122471 and #122470

return <control-0000> for \0 and other controls
  rt.perl.org/Ticket/Display.html?id=122471
  rt.perl.org/Ticket/Display.html?id=122470
We now pass several tests we were not passing before in uniname.t		 14:01
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=122471
Link: rt.perl.org/rt3//Public/Bug/Displa...?id=122470
kudo/nom: 9594a3e | lizmat++ | tools/build/NQP_REVISION:
Merge pull request #976 from samcv/bump_nqp

Fix RT #122471 and #122470 return <control-0000> for \0 and other controls
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=122471
Link: rt.perl.org/rt3//Public/Bug/Displa...?id=122470
samcv lizmat, so i never got an answer who the perl foundation to CC for my CLA, they told me they got it and to tell the rakudo admins, but 14:05
i don't think they told anybody but me
lizmat ah, intriguing
perhaps jnthn masak moritz [Coke]_ could help you 14:06
that force is not that strong with me
MasterDukeLaptop anyone have an idea why changing 'has $!name' here ( github.com/perl6/nqp/blob/master/s...or.nqp#L33 ) to 'has str $!name' would compile fine for nqp and rakudo, but die during rakudo's 'make install'? 14:56
with: X::Multi::NoMatch exception produced no message in block <unit> at tools/build/create-moar-runner.pl line 10 14:57
and i did fix the relevant getattrs and isnulls in the rest of the file 15:08
samcv m: 0x0640.uniprop('Joining_Type').say 15:10
camelia rakudo-moar 9594a3: OUTPUT«C␤»
JimmyZ MasterDukeLaptop: try --ll-exception to run create-moar-runner.pl to find the bug? 15:16
MasterDukeLaptop JimmyZ: i tried once, but it didn't work. create-moar-runner.pl is being run by moarvm directly 15:17
or, that's how it appears to me. where should i put the --ll-exception 15:18
this is the exact command that was run: /home/dan/Source/perl6/install/bin/moar --libpath="/home/dan/Source/perl6/install/share/nqp/lib" --libpath="/home/dan/Source/perl6/install/share/nqp/lib" perl6.moarvm --nqp-lib=blib tools/build/create-moar-runner.pl "/home/dan/Source/perl6/install/bin/moar" perl6.moarvm perl6-valgrind-m . "valgrind" --nqp-lib=blib "/home/dan/Source/perl6/install/share/nqp/lib" "/home/dan/Source/pe 15:19
JimmyZ MasterDukeLaptop: after perl6.moarvm
MasterDukeLaptop ah, i think i tried it earlier. i'm in the middle of another build, but i'll give that a shot in a couple min 15:20
*earlier in the command
|Tux| This is Rakudo version 2016.12-113This is Rakudo version 2016.12-113This is Rakudo version 2016.12-113-g9594a3ece built on MoarVM version 2016.12-21-g3dc56470 15:21
csv-ip5xs 2.902
test 13.245
test-t 5.137
csv-parser 13.606
-g9594a3ece built on MoarVM version 2016.12-21-g3dc56470
csv-ip5xs 2.902
test 13.245
test-t 5.137
csv-parser 13.606
-g9594a3ece built on MoarVM version 2016.12-21-g3dc56470
csv-ip5xs 2.902
test 13.245
test-t 5.137
csv-parser 13.606
sorry for my bouncing mouse
JimmyZ 5.1!
MasterDukeLaptop is csv-ip5xs usually <3? 15:22
JimmyZ: putting --ll-exception after perl6.moarvm worked (it ran and didn't complain about an unknown option), but the actual error message produced was exactly the same 15:50
JimmyZ MasterDukeLaptop: --ll-exception is not fixing the problem, just a way to explode it 16:15
MasterDukeLaptop JimmyZ: yeah, but i did hope for more information from --ll-exception 16:16
JimmyZ MasterDukeLaptop: maybe try rakudo-debugger? 16:24
japhb buggable: speed 16:35
buggable japhb, ▅▅▅▅▅▅▆▅▅▆▆▄▅▅▅▄▅▅▅▅▅█▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▃▃▃▂▂▂▂▂▂▂▁▁ data for 2016-12-06–2016-12-27; variance: 5.137s–7.592s
AlexDaniel m: say Any().WHAT 17:05
camelia rakudo-moar 9594a3: OUTPUT«(Any(Any))␤»
AlexDaniel star: say Any().WHAT
camelia star-m 2016.10: OUTPUT«(Any(Any))␤»
AlexDaniel commit: 2015.12 say Any().WHAT
committable6 AlexDaniel, ¦«2015.12»: Method 'shortname' not found for invocant of class 'Perl6::Metamodel::CoercionHOW'␤ in block <unit> at /tmp/JskmUT0nxT line 1␤ «exit code = 1»
dalek ast: ea2ac7f | moritz++ | S05-capture/match-object.t:
Tests for non-Str Match.orig
 17:27
notviki samcv: He'd need your account's email address
oh, I was scrolled up
And I don't even know what icu is 17:28
moritz a library that implements many Unicode-y algorithms 17:29
b2gills Parrot uses ICU for its Unicode support 17:34
moritz as far as Parrot still does anything, yes 17:35
notviki .ask can the two mentioned tickets be closed or do we still need tests for them? github.com/rakudo/rakudo/pull/976 18:18
yoleaux2 notviki: I'll pass your message to can.
notviki eeeeh
Thanks, robot.
.ask samcv can the two mentioned tickets be closed or do we still need tests for them? github.com/rakudo/rakudo/pull/976
yoleaux2 notviki: I'll pass your message to samcv.
samcv . 19:18
yoleaux2 18:18Z <notviki> samcv: can the two mentioned tickets be closed or do we still need tests for them? github.com/rakudo/rakudo/pull/976
[Coke]_ . 19:20
yoleaux2 09:39Z <samcv> [Coke]_: can i get special powers for rt.perl.org? need to triage some unicode bugs and rename some titles to be the real cause of them
notviki [Coke]_: account can be seen on this ticket rt.perl.org/Ticket/Display.html?id=130384 19:21
samcv thanks notviki 19:22
[Coke]_ (t/spectest.data) you can add annotations; we can come up with an annotation that means "run on 6.c but nowhere else" 19:23
samcv talking to me [Coke]_ ? 19:24
notviki Ah, true
[Coke]_ icu was for parrot, not moar or jvm. 19:28
samcv: ^^
.tell samcv I need your email address or login id to make you a bug admin. 19:30
yoleaux2 [Coke]_: I'll pass your message to samcv.
samcv kk
yoleaux2 19:30Z <[Coke]_> samcv: I need your email address or login id to make you a bug admin.
[Coke] . 19:31
yoleaux2 09:39Z <samcv> [Coke]: can i get special powers for rt.perl.org? need to triage some unicode bugs and rename some titles to be the real cause of them
[Coke] catching up.
samcv, you're a bugadmin now 19:33
samcv cool
yeah replying to bugs I have never been in the conversation of (by email) is not adding the replys 19:34
lame
[Coke] the CLA list hasn't been updated since July 2016.
notviki that'd be me
[Coke] man this loaner laptop is just a dog. :| 19:35
samcv bark bark
notviki wonders if "a dog" is good or bad...
samcv ^
dogs are really good at running 19:36
work like a dog: to work very hard. though also 19:39
something undesirable or worthless; merchandise that no one wants to buy. Put the dogs out on the sale table so people will see them.
so kind of odd 19:40
i like this one though: “to stay with one and haunt one. Will this memory dog me all the days of my life?”
hahaha
memory dogs
This is an older reference to doing something without announcing it. “I gotta see a man about a dog.” Most often used by men to excuse themselves out of a conversation. Their exact destination remains unclear, whether they are heading to the bathroom or to their bookie, one simply doesn’t ask. It was used widely during the prohibition years to refer to buying or consuming alcohol. 19:43
XD
notviki, how do I edit title? 19:47
notviki title of what?
samcv RT 19:48
notviki go to "Jumbo" and it's the first field on top left
moritz or "basics", for that matter
samcv m: say ("\c[REGIONAL INDICATOR SYMBOL LETTER G]" xx 2).Str.ords 19:49
camelia rakudo-moar 9594a3: OUTPUT«(127468 32 127468)␤»
samcv this is a weird one
the bug was a bug for xx, but the bug is in the stringification of the array actually 19:50
but. that's really odd
m: say ("\c[REGIONAL INDICATOR SYMBOL LETTER G]" xx 2).elems.say
camelia rakudo-moar 9594a3: OUTPUT«2␤True␤»
samcv m: say ("\c[REGIONAL INDICATOR SYMBOL LETTER G]" xx 2).ords
camelia rakudo-moar 9594a3: OUTPUT«(127468 32 127468)␤»
samcv m: say ("\c[REGIONAL INDICATOR SYMBOL LETTER G]" xx 2)[0].say
camelia rakudo-moar 9594a3: OUTPUT«🇬␤True␤»
samcv m: ("\c[REGIONAL INDICATOR SYMBOL LETTER G]" xx 2)[0].say
camelia rakudo-moar 9594a3: OUTPUT«🇬␤»
samcv m: ("\c[REGIONAL INDICATOR SYMBOL LETTER G]" xx 2)[1].say
camelia rakudo-moar 9594a3: OUTPUT«🇬␤»
samcv m: ("\c[REGIONAL INDICATOR SYMBOL LETTER G]" xx 2)[1].ords.say
camelia rakudo-moar 9594a3: OUTPUT«(127468)␤»
samcv actualy happens with other characters too 19:52
m: ("a" xx 2).Str.perl.say
camelia rakudo-moar 9594a3: OUTPUT«"a a"␤»
samcv adds a space.
is this not a bug? 19:53
notviki nope
m: <a b>.Str.say
camelia rakudo-moar 9594a3: OUTPUT«a b␤»
samcv yeah
japhb m: ('a', 'a').Str.say # samcv
camelia rakudo-moar 9594a3: OUTPUT«a a␤»
japhb yup
moritz does this apply a combining character to the space?
samcv no
it only adds a space
japhb samcv: Are you looking for .join ? It stringifies and joins *without* a space (unless you ask for one, of course) 19:54
samcv yes i was about to post that
at first I thought the bug was related to stringification of two joining characters which the flag sequences are (they don't have anything between them but are one character) and maybe us doing something weird. but luckily no bug 20:00
moritz we still have enough of those :-)
japhb repeat { hack-hack-hack() } while bugs; 20:03
dalek kudo/nom: 796d6e1 | lizmat++ | src/core/Any.pm:
Fix for RT #130416

Please note this is *only* dd output, not generic .perl output for lazy lists.		 20:30
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=130416
b2gills lizmat: sort should be a stable sort. that is things that cmp the same should be in the same order as the original list 20:45
m: $_ = ("0",0); say $_[0] cmp $_[1]; say .sort eqv $_
camelia rakudo-moar 9594a3: OUTPUT«Same␤False␤»
b2gills m: $_ = (0,"0"); say $_[0] cmp $_[1]; say .sort eqv $_ # It isn't even just sorting Int before Str 20:46
camelia rakudo-moar 9594a3: OUTPUT«Same␤False␤»
lizmat b2gills: afaik, the sort *is* stable 20:48
notviki m: dd [0, False].sort 20:49
camelia rakudo-moar 9594a3: OUTPUT«(Bool::False, 0)␤»
notviki m: dd [0, "0"].sort
camelia rakudo-moar 9594a3: OUTPUT«("0", 0)␤»
b2gills m: $_ = (0,"0"); say $_[0] cmp $_[1]; say .sort.sort eqv $_ # then why does it always swap the order
camelia rakudo-moar 9594a3: OUTPUT«Same␤True␤»
lizmat notviki: ok, I can work with that :-)
notviki m: dd ["0", 0].sort
camelia rakudo-moar 9594a3: OUTPUT«(0, "0")␤»
notviki lizmat: \o/
b2gills m: dd ["0", 0].sort.sort 20:50
camelia rakudo-moar 9594a3: OUTPUT«("0", 0)␤»
lizmat m: say 0 cmp "0"
camelia rakudo-moar 9594a3: OUTPUT«Same␤»
lizmat ok, I guess I was working on the premise that cmp would never return Same for such a cmp
b2gills notviki fixing nw 20:52
b2gills m: $_ = ("0",0); dd .sort({ $^a cmp $^b || Less }) # If you treat Same as being Less it works 20:53
camelia rakudo-moar 796d6e: OUTPUT«("0", 0)␤»
notviki .tell samcv seems '2' ~~ /<:Digit>/ now returns False. Bisectable points to github.com/rakudo/rakudo/commit/81...da1c643156 would you know what's causing it? 21:00
yoleaux2 notviki: I'll pass your message to samcv.
samcv yes 21:01
yoleaux2 21:00Z <notviki> samcv: seems '2' ~~ /<:Digit>/ now returns False. Bisectable points to github.com/rakudo/rakudo/commit/81...da1c643156 would you know what's causing it?
samcv m: '2' ~~/<:Digit>/
camelia ( no output )
samcv m: say '2' ~~/<:Digit>/
camelia rakudo-moar 796d6e: OUTPUT«Nil␤»
samcv yeah
notviki it's normal?
samcv uhm
not what we want 21:02
do we have roast tests for that?
notviki ¯\_(ツ)_/¯
Guess not
samcv m: say '2' ~~/<Digit>/
camelia rakudo-moar 796d6e: OUTPUT«No such method 'Digit' for invocant of type 'Cursor'␤ in block <unit> at <tmp> line 1␤␤»
samcv m: say '2' ~~/<digit>/
camelia rakudo-moar 796d6e: OUTPUT«｢2｣␤ digit => ｢2｣␤»
samcv digit isn't a unicode property
m: say '2' ~~/<:Numeric_Type>/ 21:03
camelia rakudo-moar 796d6e: OUTPUT«Nil␤»
samcv m: '2'.uniprop('nt').say
camelia rakudo-moar 796d6e: OUTPUT«Decimal␤»
notviki star: say '2' ~~ /<:Digit>/
camelia star-m 2016.10: OUTPUT«｢2｣␤»
notviki so that was a bug?
samcv possible bug yeah 21:04
i mean :whatever matches unicode properties though
and digit isn't one? so uh
though it could alias to <:Nd>
m: say '2' ~~ /<:Nd>/ 21:05
camelia rakudo-moar 796d6e: OUTPUT«｢2｣␤»
samcv that works fine
for Decimal_Number (also Digit)
so it looks like <:Digit> should do the same as <:Nd>
m: say '2' ~~ /<:Numeric_Type>/ 21:07
m: say 'test'
camelia rakudo-moar 796d6e: OUTPUT«Nil␤»
rakudo-moar 796d6e: OUTPUT«test␤»
samcv :\
m: '1'.uniprop('Numeric_Type').say 21:08
camelia rakudo-moar 796d6e: OUTPUT«Decimal␤»
samcv notviki, there is a bug in nqp
notviki m: '1'.uniprop('Numeric_Type').say; say now - INIT now
camelia rakudo-moar 796d6e: OUTPUT«Decimal␤0.0035704␤»
samcv and how it checks regexes
if you saw the conversation between me and jnthn 21:09
notviki didn't
samcv it checks VALUES not just property names
so uhm
numbers don't have a property named digit i think so it fails somethin glike that
something aweful
which is what is causing tests to fail even when i fix things in moar
u: { .unimatch('Digit') } 21:10
b2gills I think where ASCII/POSIX doesn't (and isn't likely in the future to) interfere with Unicode it should be allowed to work.
unicodable6 samcv, U+0000 NULL [Cc] (control character) 21:11
samcv, U+0001 START OF HEADING [Cc] (control character)
samcv, U+0002 START OF TEXT [Cc] (control character)
samcv isn't that what ascii does though it's 8 bytes that are unicode compatible. and unicode is defined IN ascii
:\
m: say "\0" ~~ /<:Digit>/
camelia rakudo-moar 796d6e: OUTPUT«｢␀｣␤»
b2gills m: dd "\c[NULL]"
camelia rakudo-moar 796d6e: OUTPUT«===SORRY!=== Error while compiling <tmp>␤Unrecognized character name NULL␤at <tmp>:1␤------> dd "\c[NULL⏏]"␤»
samcv :\\\\
notviki star: say "\0" ~~ /<:Digit>/ 21:12
camelia star-m 2016.10: OUTPUT«｢␀｣␤»
notviki :S
samcv m: say Rakudo::Internals.PROPCODE('Digit')
camelia rakudo-moar 796d6e: OUTPUT«27␤»
samcv let me see what 27 is
MVM_UNICODE_PROPERTY_NUMERIC_TYPE 21:13
yeah so that is fine. digit works fine
no problem in moar
notviki, github.com/MoarVM/MoarVM/blob/mast...code_gen.h if there's a problem with some match 21:14
check its propcode and then check that linked file in moar and see what it matches. if it's the right one then things are fine as long as uniprop-bool/str/int returns fine 21:15
notviki doesn't know C
samcv since the just 'uniprop' one resolves names
it's just a list of like
names and values
MVM_UNICODE_PROPERTY_HEX_DIGIT = 53,
like that
m: say Rakudo::Internals.PROPCODE('Hex_Digit')
camelia rakudo-moar 796d6e: OUTPUT«53␤»
samcv m: say Rakudo::Internals.PROPCODE('Hex')
camelia rakudo-moar 796d6e: OUTPUT«53␤»
samcv m: say Rakudo::Internals.PROPCODE('hex')
camelia rakudo-moar 796d6e: OUTPUT«53␤»
samcv m: say Rakudo::Internals.PROPCODE('space')
camelia rakudo-moar 796d6e: OUTPUT«23␤» 21:16
samcv m: say Rakudo::Internals.PROPCODE('White_Space')
camelia rakudo-moar 796d6e: OUTPUT«92␤»
samcv interesting how that nqp bug persisted due to just chance… of having it work most of the time
coincidentally
committable6, say "\0" ~~ /<:Digit>/
committable6 samcv, ¦«say»: Cannot find this revision
samcv committable6, all say "\0" ~~ /<:Digit>/ 21:17
committable6 samcv, gist.github.com/c7b9ca0fc22633ab09...0ee5c8c111
samcv :\
dalek kudo/nom: d810d44 | lizmat++ | src/core/Rakudo/Internals.pm:
Make sure .sort() is stable

TIL that 0 cmp "0" is Same, which I hadn't anticipated. b2gills++ for pointing this out.		 21:18
samcv notviki, github.com/perl6/specs/issues/118 21:19
the code matches property values too which it should probably only do for the Script, Script Extensions, General Category 21:20
and otherwise just match property names
dalek ast: e8b9f7e | lizmat++ | S15-unicode-information/uniname.t:
Unfudge now passing tests
 21:22
samcv lizmat, working on a fix so that \c[NULL] will work again 21:26
and still pass those tests
lizmat samcv: you mean the ones I just unfidged ? 21:27
samcv yeah
lizmat ah, so they were passing for the wrong reason ?
samcv no they are passing for the right reason :) 21:28
lizmat ok, *phew* 😅
samcv but the old moarvm put unicode 1 names if the current names were <control>
which made \c work fine
b2gills Assuming that Unicode will never reuse the 1 names, allowing them to work in "\c[…]" should probably be allowed 21:35
samcv they did though
u: BELL
unicodable6 samcv, U+0007 BELL [Cc] (control character)
samcv, U+237E BELL SYMBOL [So] (⍾)
samcv, U+2407 SYMBOL FOR BELL [So] (␇)
samcv, gist.github.com/995e5e7d56af989727...b7b7fa3e7d
samcv well. err 21:36
U+1F514 BELL [So] (🔔)
yep
they did
b2gills Perl 5 had the same problem with that one
geekosaur U+237E looks more like a space invaders sprite in this font >.>
samcv m: "\c[BELL]".say 21:37
camelia rakudo-moar d810d4: OUTPUT«🔔␤»
samcv evalable: "\c[BELL]".say
evalable6 samcv, rakudo-moar d810d44: OUTPUT«🔔»
samcv bisectable: "\c[BELL]".say
bisectable6 samcv, Bisecting by output (old=2015.12 new=d810d44) because on both starting points the exit code is 0
samcv, bisect log: gist.github.com/36f3ece3929499193a...1d146ed251
samcv, (2016-12-27) github.com/rakudo/rakudo/commit/81...da1c643156
samcv m: say "\c[BACKSPACE]" 21:39
camelia rakudo-moar d810d4: OUTPUT«===SORRY!=== Error while compiling <tmp>␤Unrecognized character name BACKSPACE␤at <tmp>:1␤------> say "\c[BACKSPACE⏏]"␤»
samcv m: say "\c[BEL]"
camelia rakudo-moar d810d4: OUTPUT«===SORRY!=== Error while compiling <tmp>␤Unrecognized character name BEL␤at <tmp>:1␤------> say "\c[BEL⏏]"␤»
samcv we could use \c[BEL] or something 21:40
# Note that no formal name alias for the ISO 6429 "BELL" is
# provided for U+0007, because of the existing name collision
# with U+1F514 BELL.
lizmat b2gills: how did Perl 5 fix that ?
samcv if we are going to support these characters
we need to use NameAliases.txt to do it
because they don't collide
b2gills They warned for a release or two then upgraded
samcv see ftp://ftp.unicode.org/Public/UCD/latest/ucd/NameAliases.txt 21:41
this is what we should use not the unicode 1 names
cause no collisions
looks like we do process this file 21:42
but it doesn't end up in the moarvm unicode data structure :\
luckyly there's no "BELL" in roast 6.c-errata 21:44
bisectable6, say "\c[BEL]" 21:46
bisectable6 samcv, On both starting points (old=2015.12 new=d810d44) the exit code is 1 and the output is identical as well
samcv, Output on both points: ===SORRY!=== Error while compiling /tmp/EYp39qnckE␤Unrecognized character name BEL␤at /tmp/EYp39qnckE:1␤------> say "\c[BEL⏏]"
samcv kk
looks like it never worked properly with that 21:47
dalek kudo/nom: 9d4cdd4 | lizmat++ | src/core/Rakudo/Internals.pm:
Introducing R:I.ReifiedListIterator

  - abstracted from List.iterator for more general usage
kudo/nom: ca919d9 | lizmat++ | src/core/List.pm:
Use R:I.ReifiedListIterator in List.iterator

Because now we can
b2gills I think that ROAST would have been updated to return the non-control bell anyway
samcv yeah
for sure
though at least we have a good path going forward 21:48
none of this 'unicode 1 names' only the _Official_ name aliases
and will be future portable
moritz m: say "\n".uniname 21:50
camelia rakudo-moar d810d4: OUTPUT«<control-000a>␤»
samcv ack should be uppercase
m: say "\c[LF]" 21:54
camelia rakudo-moar ca919d: OUTPUT«␤␤»
samcv that works tho
m: say "\c[LINE FEED]"
camelia rakudo-moar ca919d: OUTPUT«===SORRY!=== Error while compiling <tmp>␤Unrecognized character name LINE FEED␤at <tmp>:1␤------> say "\c[LINE FEED⏏]"␤»
b2gills I think LF might even predate ASCII 21:55
samcv m: say "\c[NULL]"
camelia rakudo-moar ca919d: OUTPUT«===SORRY!=== Error while compiling <tmp>␤Unrecognized character name NULL␤at <tmp>:1␤------> say "\c[NULL⏏]"␤»
samcv m: say "\c[CR]"
camelia rakudo-moar ca919d: OUTPUT«
samcv hmmmm
spooky
must have that programmed in somewhere…
b2gills m: say "boooo\c[CR]hell" # too bad camilia isn't special cased for this 21:58
camelia rakudo-moar ca919d: OUTPUT«boooo
notviki m: say "boooo\c[CR]QUIT :bye" 22:02
camelia rakudo-moar ca919d: OUTPUT«boooo
samcv m: say "\c[EOH]" 22:05
camelia rakudo-moar ca919d: OUTPUT«===SORRY!=== Error while compiling <tmp>␤Unrecognized character name EOH␤at <tmp>:1␤------> say "\c[EOH⏏]"␤»
lizmat good night, #perl6-dev! 22:06
notviki night 22:07
samcv ok so 0x7's control name is ALERT 22:50
and its abbreviation is BEL
so i will just use the control name 22:53
until there is a better method of doing this
the only 6.c-errata test this will affect is: is("\c[LINE FEED (LF)]", "\c10", '\c[LINE FEED (LF)] works'); 22:55
it also tests is("\c[LINE FEED]", "\c10", '\c[LINE FEED] works'); 22:56
is("\c[LF]", "\c10", '\c[LF] works');
so if i make this change we will pass the \c[LINE FEED] test
but break the LINE FEED (LF) one. but. i mean who actually would type that all out?
b2gills, thoughts? 22:57
LINE FEED is in unicode name aliases, but LINE FEED (LF) is not, there is LF and LINE FEED
it's a bit silly somebody typing \c[LINE FEED (LF)] =\ and i'm not totally sure the (LF) is actually part of the name tbh 22:58
well unicode 1 was a long time ago
we can still decide what the 'correct' name to return for uniname(:one ) for unicode 1 names, but should use the official alias names 22:59
the only things that will break will be CARRIAGE RETURN (CR), and LINE FEED (LF) and BELL. but will fix CARRIAGE_RETURN and LINE FEED which are much more common 23:00
and BELL will be ALERT for now, and will also be able to do \c[BEL] when we get proper name alias support working
travis-ci Rakudo build errored. Elizabeth Mattijsen 'Use R:I.ReifiedListIterator in List.iterator 23:02
travis-ci.org/rakudo/rakudo/builds/187075702 github.com/rakudo/rakudo/compare/d...919d96c97e
buggable [travis build above] ☠ Did not recognize some failures. Check results manually.
notviki :/ "The job exceeded the maximum time limit for jobs, and has been terminated" 23:21
samcv 47 minutes

Powered by Cro and the Raku® Programming Language.

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

Thank you!