»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend! Set by moritz on 22 December 2015. |
|||
00:02
rindolf left
00:07
pierre_ joined
00:10
Actualeyes left
00:13
labster left
00:16
nicq20_ left
00:20
rpburkholder joined
|
|||
timotimo | bazzaar: i see your pull request came through to us | 00:31 | |
we should be able to upload renders from travis for pull requests so you could view the final render on doc.perl6.org somewhere | 00:32 | ||
like doc-1012.perl6.org maybe? | |||
bazzaar | timotimo: oh, that's great ... been flying by the seat of my pants :), just now staring at the merge button, thinking am I supposed to push it? | 00:33 | |
timotimo | well, if you're confident the pull request is good, sure | ||
bazzaar | timotimo: but am I supposed to wait for others to approve before doing so? | 00:34 | |
seatek | is there any difference invoking perl6 as perl6 vs perl6-m ? | 00:36 | |
viki | seatek: cat them both and see :) They're just wrapper scripts | 00:38 | |
timotimo | usually you'd wait for others, yeah | ||
seatek | viki: i did notice they were the same size ;) | ||
timotimo | i see a colon with a space in front of it | ||
was that in error? | |||
bazzaar | timotimo: ok, I'll wait | 00:39 | |
viki | bazzaar: if it's just a small change you're confident with you can just commit directly. It's only for changes with large impact that folks tend to open a PR and wait for feedback. | ||
geekosaur | seatek, you could (although right now I would not) specify the default to be jvm instead of moar | 00:40 | |
if you want to enforce moarvm, use the -m one. if you don't care which then omit the suffix | |||
timotimo | generally we operate on "it's easier to ask for forgiveness than for permission" | ||
viki | Yeah, the revert feature works :P | 00:41 | |
seatek | geekosaur: gotcha! thanks :) | 00:42 | |
bazzaar | viki: Thanks, I will bear that in mind :) this one is a rewrite of zip and roundrobin, and because it's my first PR maybe I should wait. | 00:44 | |
00:45
cdg left
00:59
sufrostico left
01:02
user10 joined,
user9 left,
user10 is now known as user9
01:03
Actualeyes joined
01:04
AlexDaniel left
01:05
BenGoldberg joined
01:06
sufrostico joined
01:07
aries_liuxueyang left
01:11
aries_liuxueyang joined,
bazzaar left
01:16
labster joined
01:19
kyclark joined
01:25
sufrostico left,
kyclark left
01:30
nadim left,
nadim_ joined
01:32
sufrostico joined
01:41
nadim_ left
01:50
flaviusb left
01:58
flaviusb joined
|
|||
Xliff | \o/ | 02:00 | |
Made it to the next stage of this freaking silly SSO process. | 02:01 | ||
Xliff does a silly dance. | 02:09 | ||
02:12
girafe left
02:27
kalkin- joined
|
|||
kalkin- | hi | 02:27 | |
How can i list all the variables in the current scope? (I'm trying out in REPL) | 02:28 | ||
So when i declare my $foo = "Bar"; $::{'$foo'} should be the same var, according to docs | |||
so i should be able to do $::.kv | 02:29 | ||
to list the contents | |||
but REPL says $ is not declared | |||
docs.perl6.org/language/packages#Direct_lookup | |||
MasterDuke | m: my $foo = "Bar"; my \Bar = "baz"; say ::(q|$foo|); say ::($foo); | 02:33 | |
camelia | rakudo-moar 58a482: OUTPUT«Barbaz» | ||
02:35
kalkin- left,
kalkin-_ joined
|
|||
MasterDuke | well, it should be my $foo = "Bar"; $::{'foo'} | 02:36 | |
according to the doc | |||
m: my $foo = "Bar"; say $::{'foo'} | |||
camelia | rakudo-moar 58a482: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Variable '$' is not declaredat <tmp>:1------> 3my $foo = "Bar"; say 7⏏5$::{'foo'}» | ||
MasterDuke | bisectable6: my $foo = "Bar"; say $::{'foo'} | ||
bisectable6 | MasterDuke, On both starting points (old=2015.12 new=58a4826) the exit code is 1 and the output is identical as well | ||
MasterDuke, Output on both points: ===SORRY!=== Error while compiling /tmp/TmRkwhWsrzVariable '$' is not declaredat /tmp/TmRkwhWsrz:1------> my $foo = "Bar"; say <HERE>$::{'foo'} | |||
MasterDuke | m: my $foo = "Bar"; say ::{'$foo'} | 02:38 | |
camelia | rakudo-moar 58a482: OUTPUT«Bar» | ||
kalkin-_ | thanks! | ||
but why does ::.kv not list the var? | |||
MasterDuke | but it looks like something is wrong about $::{'foo'}, either the docs or it's a bug | 02:39 | |
m: my $foo = "Bar"; .say for ::.kv | 02:40 | ||
camelia | rakudo-moar 58a482: OUTPUT«$=pod[]$fooBar!UNIT_MARKER(!UNIT_MARKER)EXPORT(EXPORT)$_(Any)$!Nil::?PACKAGE(GLOBAL)GLOBALish(GLOBAL)$¢Nil$=finish(Mu)$/Nil$?PACKAGE(GLOBAL)» | ||
samcv | trying to work on Perl6FE highlighting for atom right now, get it to recognize fancy quotes and more styles of #`{{ multiline comments | 02:41 | |
02:42
Rotwang left
|
|||
MasterDuke | m: my $foo = "Bar"; for ::.kv -> $k, $v { say "$k => $v.perl()" } | 02:42 | |
camelia | rakudo-moar 58a482: OUTPUT«$=pod => $[]$foo => "Bar"!UNIT_MARKER => !UNIT_MARKEREXPORT => EXPORT$_ => Any$! => Nil::?PACKAGE => GLOBALGLOBALish => GLOBAL$¢ => Nil$=finish => Mu$/ => Nil$?PACKAGE => GLOBAL» | ||
MasterDuke | kalkin-_: ^^^ see it there? | ||
kalkin-_ | This doesn't work in repl | 02:43 | |
MasterDuke | there are known problems with the repl | ||
kalkin-_ | ahh i see | ||
in repl is $foo missing | 02:44 | ||
it's only in it if i declare it on same line | |||
02:45
ilbot3 left
|
|||
kalkin-_ | thank you guys | 02:46 | |
MasterDuke | sounds similar to rt.perl.org/Public/Bug/Display.html?id=78068 | 02:47 | |
02:47
ilbot3 joined
02:50
richi235 left
02:51
richi235 joined
|
|||
dj_goku | anyone notice colabti.org searching is slow/broken/never returns? | 02:54 | |
02:57
kyclark joined
03:18
pierre_ left
03:22
kyclark left
03:30
kyclark joined
03:33
kyclark left
03:39
noganex_ joined
03:41
noganex left
03:48
kyclark joined
03:50
pierre_ joined
03:54
cibs left
03:55
cibs joined
04:00
troys joined
|
|||
Xliff | m: my $a = do for ^4 { $_ * 2 }; say $al | 04:02 | |
camelia | rakudo-moar 58a482: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Variable '$al' is not declaredat <tmp>:1------> 3my $a = do for ^4 { $_ * 2 }; say 7⏏5$al» | ||
Xliff | m: my $a = do for ^4 { $_ * 2 }; say $a; | ||
camelia | rakudo-moar 58a482: OUTPUT«(0 2 4 6)» | ||
04:06
kyclark left
04:21
aborazmeh joined,
aborazmeh left,
aborazmeh joined
|
|||
Xliff | m: my @a = ^6; for @a:kv -> $k, $v { say "$k $v" } | 04:21 | |
camelia | rakudo-moar 58a482: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Variable '@a:kv' is not declaredat <tmp>:1------> 3my @a = ^6; for 7⏏5@a:kv -> $k, $v { say "$k $v" }» | ||
gfldex | seatek: zef problem already solved: github.com/ugexe/zef/commit/4a9745...7e4039859c | ||
Xliff | m: my @a = ^6; for @a.hash:kv -> $k, $v { say "$k $v" } | ||
camelia | rakudo-moar 58a482: OUTPUT«4 5 0 1Too few positionals passed; expected 2 arguments but got 1 in block <unit> at <tmp> line 1» | ||
Xliff | Well crap. I thought you could get array index and value via :kv adverb from arrays. | ||
m: my @a = ^6; for @a.kv -> $k, $v { say "$k $v" } | 04:22 | ||
camelia | rakudo-moar 58a482: OUTPUT«0 01 12 23 34 45 5» | ||
Xliff | Oh. Not adverb. LOL! | ||
04:24
kyclark joined
04:25
Actualeyes left
04:37
labster left,
ggoebel joined
04:38
rpburkholder left
04:40
skids left
|
|||
dj_goku | what are valid block delimiters? | 04:41 | |
a code block that is. | |||
04:46
aborazmeh left
|
|||
seatek | gfldex: wow that was fast! i'm going to have to try zef eventually | 04:47 | |
04:49
rpburkholder joined
04:55
kyclark left
04:58
redhands left
04:59
khw left
|
|||
seatek | core! hmm. double free or corruption? double free please! | 05:14 | |
05:17
Cabanossi left
05:22
Cabanossi joined
|
|||
seatek | anyone know of any known memory eating bug issues related to Callables? | 05:23 | |
In Crust... a Callable is used and passed up to the connected server from the application. | 05:26 | ||
Multiple repeated calls can create some extra memory use at first, but it levels off | |||
If you use the Crust::Builder to create "middleware" -- it temporarily assigns this Callable to a new local Callable, passing to to some constructed classes, then returning. I'm wondering if those locally created Callables might linger | 05:28 | ||
05:33
Actualeyes joined
|
|||
seatek | if you pass an Callable to a method, and return that callable time and again, memory use goes up (in Crust here). | 05:37 | |
if you pass in a Callable to a method, then assign that Callable to a local $_app var, then return that local $_app var instead of the Callable passed in, memory usage noticably increases | 05:38 | ||
(and the rate at which it climbs) | 05:39 | ||
gfldex | seatek: are you trying to tell us that you want to golf that? :-> | ||
seatek | i don't know what that means. i just want to stomp my foot and make everything better :) | 05:40 | |
does that mean I should keep trying to isolate down what the cause is? | |||
or that i'm trying to get someone else to fix it? | 05:41 | ||
golfing it... putting it down further | |||
gfldex | a code golf is the shortest prossible program that can trigger a bug | ||
seatek | i need a memory profiling tool :) | ||
ah ok | |||
ugh | |||
it's hard extracting that out from something like Crust | |||
I don't even like Callables | 05:42 | ||
they're nothing but trouble anyway | 05:43 | ||
05:43
troys is now known as troys_
|
|||
gfldex | You could file a ticket on your own module and come back to it later. Very well possible that the bug is already known. | 05:43 | |
seatek | i'll go a little further into the chaos. or i'll forget what i did and didn't do | 05:44 | |
i should file the bug in Crust::Builder cuz that's where it comes up, even using none of the middleware modules (like mine) | 05:46 | ||
but i'm betting it's a Callable in general memory issue relating to scopes of some kind... freeing stuff | 05:47 | ||
$20 | |||
05:51
rpburkholder left
05:53
troys_ is now known as troys,
BenGoldberg left
|
|||
seatek | what the heck is the MONKEY-SEE-NO-EVAL - not in modules listing, not in --doc | 05:57 | |
ah on the site | 05:58 | ||
gfldex | docs.perl6.org/language/pragmas | ||
06:08
troys left
|
|||
dj_goku | anyone build the perl6/doc html on a mac and is on sierra? | 06:11 | |
gfldex | dj_goku: doubt it will work. There are still a few things that are case sensitive | 06:12 | |
dj_goku | gfldex: haha ok, well I guess I am done for the night. | 06:13 | |
I was going to spin up a ubuntu system and try it thee. | |||
s/thee/there/ | |||
06:44
djbkd left
|
|||
Xliff | m: my $a = do for ((1, 'a'), (2, 'b'), (3, 'c')) -> $c { $_.flat }; dd $a; | 06:45 | |
camelia | rakudo-moar 58a482: OUTPUT«List $a = $((Any,).Seq, (Any,).Seq, (Any,).Seq)» | ||
Xliff | m: my $a = do for ((1, 'a'), (2, 'b'), (3, 'c')) -> $c { $c.flat }; dd $a; | ||
camelia | rakudo-moar 58a482: OUTPUT«List $a = $((1, "a").Seq, (2, "b").Seq, (3, "c").Seq)» | ||
Xliff | m: my $a = { do for ((1, 'a'), (2, 'b'), (3, 'c')) -> $c { $c.flat } }; dd $a; | ||
camelia | rakudo-moar 58a482: OUTPUT«Block $a = -> ;; $_? is raw { #`(Block|65663120) ... }» | ||
Xliff | Damn. | ||
seatek | gfldex: where's a good place to paste code into for viewing here? | ||
Xliff | fpaste | 06:46 | |
gfldex | gist.github.com/ | ||
Xliff | fpaste.scsys.co.uk or gist.github.com | ||
gfldex++ was faster on the gist. | |||
m: my $a = { do for ( 1 => 'a', 2 => 'b', 3 => 'c') -> $c { $_ } }; dd $a; | 06:47 | ||
camelia | rakudo-moar 58a482: OUTPUT«Block $a = -> ;; $_? is raw { #`(Block|64771344) ... }» | ||
gfldex | there is gist-paste to paste and update directly from the shell, provided by packages.debian.org/jessie/gist | ||
seatek | gist.github.com/adaptiveoptics/3df...y-wrappers | ||
Xliff | m: my $a = { do for ( 1 => 'a', 2 => 'b', 3 => 'c') -> $c { $c } }; dd $a; | ||
camelia | rakudo-moar 58a482: OUTPUT«Block $a = -> ;; $_? is raw { #`(Block|78349744) ... }» | ||
Xliff | Hrrm.... | 06:48 | |
seatek | That's basically what Crust with middlewear (from Builder) is doing | ||
and it grows and grows in memory use. and i think it may be a bug in the way callables are being passed around more than callables themselves | 06:49 | ||
maybe not though - they're all self-contained | 06:50 | ||
06:50
RabidGravy joined
|
|||
seatek | even when you put a $c=C.new in there to clear it out each time, memory grows and grows | 06:53 | |
gfldex | seatek: if you remove line 8 and 9, does it still blow up? | ||
seatek | yup | 06:54 | |
that was my first thought before too ;) | |||
06:54
Woodi joined
|
|||
gfldex | could be RT#77644 | 06:55 | |
synopsebot6 | Link: rt.perl.org/rt3//Public/Bug/Displa...l?id=77644 | ||
seatek | i'll try it without evals | 06:56 | |
could i just substitue a code ref for that and it's still Callable? | 06:57 | ||
06:59
neuraload joined
|
|||
Xliff | m: @a = ^4; @b = ^3; for @a, @b { .say } | 06:59 | |
camelia | rakudo-moar 58a482: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Variable '@a' is not declaredat <tmp>:1------> 3<BOL>7⏏5@a = ^4; @b = ^3; for @a, @b { .say }» | ||
Xliff | m: my @a = ^4; my @b = ^3; for @a, @b { .say } | ||
camelia | rakudo-moar 58a482: OUTPUT«[0 1 2 3][0 1 2]» | ||
Xliff | m: my @a = ^4; my @b = ^3; for |(@a, @b) { .say } | 07:00 | |
camelia | rakudo-moar 58a482: OUTPUT«[0 1 2 3][0 1 2]» | ||
Xliff | m: my @a = ^4; my @b = ^3; for |(@a, @b).flat { .say } | ||
camelia | rakudo-moar 58a482: OUTPUT«0123012» | ||
seatek | gfldex: yes that's exactly the problem -- the EVAL. dammit | 07:01 | |
who fixes that? | |||
ha! Nobody is the owner :) | |||
07:01
u-ou left
|
|||
seatek | clear back when Parrot was the thing | 07:02 | |
07:06
nadim_ joined
|
|||
seatek | wait -- i don't think it gets eval'ed each loop in Crust | 07:07 | |
a flaw in my model | |||
curses | |||
07:08
domidumont joined
|
|||
gfldex | perl6 -e 'loop { say ((slurp "/proc/$*PID/stat").lines[0].split: " ")[22]; EVAL q[]; }' | 07:08 | |
problem cleary still persists | |||
07:10
darutoko joined
|
|||
gfldex | actually it stops growing after ~30sec | 07:11 | |
seatek: and that's why we golf :) | 07:12 | ||
seatek | i see :) | 07:14 | |
yeah i noticed earlier that it tended to level off a bit | |||
masak .oO( when the bug hits your eye / like a big pizza pie / that's a-golf-e ) | 07:15 | ||
seatek | so who fixes it? can you assign it to masak? | ||
masak | doubtful | ||
seatek | he'll fix it. he can fix anything :) | 07:16 | |
masak | I think you might be thinking of some other masak | ||
seatek | hehe | ||
07:16
wamba joined
|
|||
seatek | alright, if it can't be fixed I'll just have to hard-code my application into Crust. | 07:18 | |
brrt | good morning #perl6 | ||
RabidGravy | marning! | ||
to be honest that's the way I use Crust anyway | 07:19 | ||
seatek | seriously? love it | 07:21 | |
07:21
go|dfish joined
|
|||
RabidGravy | infact I tend to implement the actions as "class Foo does Callable { method CALL-ME(%env) { } }" and put the action in the CALL-ME | 07:22 | |
seatek jots things down | 07:23 | ||
RabidGravy | I think there is an example of this kind of usage in Lumberjack::Application | 07:24 | |
anyway off to the grind-stone | |||
gfldex | seatek: it would not hurt if you would write your adventures down. We got the question, what web framework to use, quite often. | 07:27 | |
seatek | I gave up on that question -- that's why I'm messing so deeply with Crust ;) | 07:28 | |
07:28
RabidGravy left
|
|||
seatek | I really don't like web frameworks. I spend more time learning them than I do making stuff work myself | 07:29 | |
that's why i was so happy to see sessions with Crust | |||
just your basic stuff. | |||
i'm wanting to incorporate grammars into what i'm doing with this one. can't imagine anything better for an api | 07:33 | ||
07:35
domidumont left
07:38
imcsk8 left
07:40
cgfbee left
07:42
xinming left,
imcsk8 joined
07:45
labster joined
07:46
cgfbee joined
07:49
brillenfux joined
08:01
abraxxa joined
08:04
domidumont joined,
domidumont left
08:08
domidumont joined
|
|||
samcv | i still need to lead grammars | 08:10 | |
08:10
abraxxa left
08:11
abraxxa joined
|
|||
samcv | they look like they could be fun though | 08:12 | |
moritz | they aren't really fun, but less painful than other approaches to parsing, IMHO | 08:13 | |
08:13
jferrero joined
|
|||
samcv | that's fun enough for me :) | 08:13 | |
i guess fun is relative | |||
08:14
imcsk8 left,
domidumont left
08:15
ufobat joined,
imcsk8 joined,
domidumont joined
08:25
Actualeyes left
08:26
zakharyas joined
08:29
b2gills1 joined
08:30
Actualeyes joined
08:32
b2gills left
08:42
domidumont left
08:44
domidumont joined
08:52
domidumont left
08:53
domidumont joined
08:54
ocbtec joined
09:02
dakkar joined
09:04
g4 left
09:09
Actualeyes left
09:14
g4 joined,
g4 left,
g4 joined
09:34
labster left
09:35
domidumont left
09:36
maybekoo2 joined,
domidumont joined
09:41
rindolf joined
09:55
khw joined
09:57
TEttinger left
10:03
dextertzu joined,
kurahaupo__ joined
10:04
xinming joined
10:09
xinming_ joined
10:10
xinming left
10:17
xinming_ is now known as xinming
10:35
Actualeyes joined
10:38
maybekoo2 left
10:43
domidumont left
10:53
gregf_ joined
10:54
kurahaupo__ is now known as kurahaupo
10:58
canopus left
11:04
canopus joined
11:11
pmurias joined
11:16
seatek left
11:18
pierre_ left
11:20
pierre_ joined
11:21
itaipu joined
11:24
brillenfux_ joined
11:27
brillenfux left
11:36
Actualeyes left
11:38
Actualeyes joined
11:43
Actualeyes left
11:45
imcsk8 left,
Actualeyes joined
11:47
imcsk8 joined
11:54
geekosaur left
11:57
geekosaur joined
12:09
Actualeyes left
12:15
AlexDaniel joined
|
|||
timotimo | hmm. do we want to redirect people from perl6.org to perl6.org? | 12:22 | |
moritz | timotimo: if you want that, please grep the ecosystem for tests that try to fetch from HTTP URLs from perl6.org | 12:23 | |
timotimo | oh, that's a very good point indeed | ||
jonadab | Redirecting onto https really only makes sense for sites that most users are going to log into with a username and password or the like, and even then really only if it's the kind of login where it _matters_ for something other than leaving blog comments. | 12:25 | |
IMO. | |||
timotimo | i just saw PoisonTap, which injects backdoors into pretty much everything the user's browser might ever request | ||
moritz | or hotstops that inject ads, and then some of these ads contain malware | 12:28 | |
AlexDaniel | timotimo: generally, yes | 12:29 | |
jonadab | timotimo: That explains why you are thinking about this now; but I don't see how it's really an argument in favor of redirecting all site traffic to https. | ||
AlexDaniel | jonadab: and you are wrong, sir | ||
jonadab | (It _might_, _arguably_, be an argument in favor of web browsers trying https first in preference to http when the user inputs a URL without a protocol. Possibly.) | 12:30 | |
moritz | jonadab: people then have the HTTPS URL in their browser history or bookmarks, and that's where they will go in a hostile environment | ||
jonadab | (Not that most users ever input URLs these days, except into search engine boxes.) | ||
moritz | well, and that's the next argument in favor | 12:31 | |
arnsholt | If we want to move perl6.org to HTTPS, but also make it possible to download some things over HTTP for testing, maybe we could have a dedicated endpoint/virtual host that doesn't redirect to HTTPS and redirect everything else? | ||
moritz | if there's a redirect to HTTPS, that's where the search engines will send you | ||
arnsholt: we can make an insecure.perl6.org or so | |||
or nohttps.perl6.org | |||
arnsholt | Yeah | ||
jonadab | moritz: There are more direct ways to influence where the search engines send users. | 12:32 | |
(Granted, they aren't entirely foolproof.) | |||
moritz | jonadab: somehow the canoncial URL meta tags never really worked for me | ||
12:33
AlexDani` joined
12:35
AlexDaniel left
|
|||
moritz | www.reddit.com/r/IAmA/comments/5c9...e/d9ur4a2/ # the let's encrypt team's take on whethere you should use HTTPS everywhere | 12:35 | |
arnsholt | Yeah, on the whole I think it's the right direction to move in | 12:36 | |
It introduces an additional OpenSSL dependency (more or less de facto) on many things, which isn't great, and introduces a bootstrapping problem of getting HTTPS working when you don't already have working HTTPS | 12:38 | ||
But the gains are worth it I think | |||
AlexDani` | github.com/supernovus/perl6-http-c...1-get.t#L8 | ||
github.com/supernovus/perl6-http-c...post.t#L11 | |||
github.com/supernovus/perl6-http-c...part.t#L11 | 12:39 | ||
these are the only tests you have to fix I think (I'm not sure if it works with https or not) | |||
12:39
AlexDani` is now known as AlexDaniel
|
|||
arnsholt | I'd be surprised if it works over HTTPS | 12:39 | |
AFAIK we don't have any bindings to an SSL library, much less a pure Perl 6 SSL suite | 12:40 | ||
moritz | the client that sergot++ worked on has openssl integration | 12:41 | |
AlexDaniel | these are the only tests I see in our ecosystem that use perl6.org domain in this way | 12:42 | |
arnsholt | moritz: Oh, neat! | ||
12:43
itaipu left
|
|||
dalek | c: f73aca6 | bazzaar++ | doc/Type/List.pod6: Attempt to improve the documentation of routines zip, and roundrobin. |
12:44 | |
c: 3ca220b | (Zoffix Znet)++ | doc/Type/List.pod6: Merge pull request #1012 from bazzaar/roundrobin-edits Improve the documentation of routines zip, and roundrobin. |
|||
synopsebot6 | Link: doc.perl6.org/type/List | ||
arnsholt | Speaking of something completely different: Have any of you experimented with proportional fonts for code? | 12:47 | |
In the SmallTalk environment at $dayjob, there's no highlighting and proportional fonts by default, and I'm finding it to be pretty nice actually | 12:48 | ||
AlexDaniel | arnsholt: um… that's some interesting point of view | 13:02 | |
arnsholt: so what's nice about it? | 13:03 | ||
timotimo | maybe it's pleasant to the eye? | ||
13:11
b2gills1 is now known as b2gills
|
|||
AlexDaniel | but the question about fonts is a good one | 13:14 | |
arnsholt | At least for reading normal text, I'm pretty sure proportional fonts score better on legibility than fixed-with | ||
AlexDaniel | are there any studies about this? | ||
arnsholt | I think so | ||
There're studies comparing serif and sans-serif at least | |||
13:15
user9 left
|
|||
arnsholt | But that's not my area of expertise | 13:15 | |
Only annoying thing with no highlighting is that comments (especially with code inside) are a bit too prominent | |||
timotimo | "where the heck is the code" | 13:16 | |
13:17
brillenfux_ is now known as brillenfux
13:18
domidumont joined
13:19
domidumont left,
domidumont joined
|
|||
pmurias | arnsholt: aren't SmallTalk environments a hate it or love it kind of thing? | 13:32 | |
timotimo stumbled over subresource integrity, but doesn't really think it applies for perl6.org or something in that space | 13:40 | ||
13:47
kyclark joined
13:49
lizmat left
13:50
kyclark left
13:55
user9 joined
|
|||
nine just booked his flight for LPW2016 | 14:03 | ||
DrForr | Yay! I'm coming in Friday morning, returning Monday morning. | 14:04 | |
tadzik | I'm coming wednesday night, leaving sunday night | 14:05 | |
14:05
lizmat joined
|
|||
nine | Sooo...on to finding a hotel | 14:05 | |
256 Euros for 2 nights? In London? With 4 stars? I don't trust this | 14:09 | ||
timotimo | they'll probably harvest one of your kidneys | ||
14:10
kyclark joined
|
|||
DrForr | I gave up and found a place in Croydon; it's a little outside of London proper but all I could find in town was those damn GBP60/night walkups that are just above a hostel. | 14:11 | |
14:18
lizmat left
|
|||
nine | Bloomsbury Palace Hotel it is. It cannot be any worse than the last one I picked. That's physically impossible, so I may as well give it a try ;) | 14:19 | |
14:19
lizmat joined
|
|||
tadzik | nine: hmmm, not entirely unreasonable I think | 14:20 | |
I pay ~480 for 3 nights, 2 people | |||
nine | Now all I need to do is get really started working on my talk. Got swamped with work lately which kicked me really out of my Perl 6 habits. | 14:21 | |
14:22
domidumont1 joined,
domidumont left
14:24
iH2O joined
14:33
itaipu joined
|
|||
iH2O | z... | 14:34 | |
14:36
pmurias left
|
|||
viki | ? | 14:36 | |
14:43
kyclark left
14:48
pmurias joined
14:54
cdg joined
14:55
xtt joined
15:00
titsuki joined
|
|||
arnsholt | pmurias: Yeah, I suspect ST is love it or hate it. But I'm digging it | 15:02 | |
At work we're stuck with a pretty old commercial ST, but apart from that it's pretty neat | |||
15:03
iH2O left
15:06
handle_ joined,
handle_ left
15:14
lizmat left
15:15
itaipu left,
pierre_ left
15:16
pierre_ joined
15:19
jonas2 left
15:20
pierre_ left
15:38
pmurias left
15:41
neuraload left
15:42
brrt left
|
|||
titsuki | bisect: my $a = [email@hidden.address] my $b = / <:L>+\.<:L>+\@<:L+:N>+\.<:L>+ /; say $/ if $a ~~ $b; | 15:45 | |
bisectable6 | titsuki, On both starting points (old=2015.12 new=85c7072) the exit code is 0 and the output is identical as well | ||
titsuki, Output on both points: 「john.doe@perl6.org」 | |||
15:46
FROGGS joined
|
|||
titsuki | bisect: old=2016.10 my $a = [email@hidden.address] my $b = / <:L>+\.<:L>+\@<:L+:N>+\.<:L>+ /; say $/ if $a ~~ $b; | 15:47 | |
bisectable6 | titsuki, Bisecting by output (old=2016.10 new=85c7072) because on both starting points the exit code is 0 | ||
FROGGS | o/ | ||
bisectable6 | titsuki, bisect log: gist.github.com/18df2dd79a0a2e1e8f...6ca12fa7ab | ||
titsuki, (2016-10-27) github.com/rakudo/rakudo/commit/5a...925c544ea4 | |||
titsuki | hi | 15:48 | |
AlexDaniel | NeuralAnomaly: when? | 15:50 | |
NeuralAnomaly: status | |||
NeuralAnomaly | AlexDaniel, [✘] Next release will be in 1 day and 13 hours. Since last release, there are 32 new still-open tickets (0 unreviewed and 0 blockers) and 12 unreviewed commits. See perl6.fail/release/stats for details | ||
AlexDaniel | titsuki: yeah, seems like it will be solved in the next release. Still an “oops” :) | 15:51 | |
committable6: 6c my $a = [email@hidden.address] my $b = / <:L>+\.<:L>+\@<:L+:N>+\.<:L>+ /; say $/ if $a ~~ $b; | 15:52 | ||
committable6 | AlexDaniel, ¦«2015.12,2016.02,2016.03,2016.04,2016.05,2016.06,2016.07.1,2016.08.1,2016.09,HEAD»: 「john.doe@perl6.org」¦«2016.10»: Nil | ||
AlexDaniel | there was no ticket for that? | 15:53 | |
meaning that there are still no tests? | |||
titsuki | AlexDaniel: I'm not sure, but I can't find the test code for that bug. | 15:55 | |
AlexDaniel | titsuki: perhaps you can contribute some? | ||
15:57
pmurias joined
|
|||
titsuki | AlexDaniel: Should I RT this bug or write some test cases ? | 15:58 | |
viki | Isn't it fixed already? | 15:59 | |
AlexDaniel | titsuki: it is fixed already, so I don't think a ticket is required | ||
viki | committable6: 2016.10 my $a = [email@hidden.address] my $b = / <:L>+\.<:L>+\@<:L+:N>+\.<:L>+ /; say $/ if $a ~~ /<$b>/; | 16:00 | |
committable6 | viki, ¦«2016.10»: 「john.doe@perl6.org」 | ||
viki | I recall this bug and lizmat fixing it. | ||
16:03
pmurias left
16:04
committable6 left,
pmurias joined
16:05
committable6 joined,
ChanServ sets mode: +v committable6
|
|||
viki | a glance through commits doesn't show tests | 16:05 | |
viki adds | |||
just in case | |||
dalek | c: c40ccb5 | coke++ | doc/Type/List.pod6: remove trailing whitespace |
16:06 | |
synopsebot6 | Link: doc.perl6.org/type/List | ||
viki | s: &infix:<~~>, \($, $) | 16:07 | |
SourceBaby | viki, Sauce is at github.com/rakudo/rakudo/blob/85c7...Mu.pm#L818 | ||
16:08
Actualeyes joined
|
|||
titsuki | Isn't it the test case for this bug ? github.com/perl6/roast/commit/40b8...24a1521566 | 16:08 | |
viki | Ah, OK. That gives me the commit that fixed it :) github.com/rakudo/rakudo/commit/5ac593e | 16:09 | |
titsuki: not really, basically the fix made it so you get truthy values instead of Bool | 16:10 | ||
So the above (old) tests needed an adjustment | |||
The issue itself is regex stored in a variable used with a smartmatch and so the bug assumed that meant the ~~ was chained and so it returned just a Bool | 16:11 | ||
16:14
g4 left
16:15
kyclark joined
16:16
kyclark left
|
|||
[Coke] | jnthn: does your file finder module deal with this corner case: stackoverflow.com/questions/406492...-directory (last comment) | 16:19 | |
16:22
pmurias left
|
|||
titsuki | m: ("a" ~~ /a/ ~~ /abc/).say | 16:27 | |
camelia | rakudo-moar 85c707: OUTPUT«Nil» | ||
16:27
kyclark joined,
xtt left
|
|||
titsuki | viki: thanks for your explanation. I understand that. | 16:28 | |
16:29
pmurias joined
|
|||
viki | Tests for the bug have been added in github.com/perl6/roast/commit/839566cfd6 | 16:33 | |
AlexDaniel | viki++ | 16:41 | |
16:51
skids joined
16:53
xtt joined
16:54
zakharyas1 joined
16:56
abraxxa left
17:03
ufobat left
17:05
zakharyas1 left
|
|||
titsuki | viki: thx. I could solve this issue: github.com/hankache/perl6intro/issues/118 | 17:07 | |
viki | sweet | ||
17:08
titsuki left
17:14
lizmat joined,
xtt left
17:16
girafe joined
17:31
zakharyas left
|
|||
[Coke] | google alert gave me: infogalactic.com/info/Perl_6 ... TIL someone is just copying wikipedia... | 17:35 | |
17:36
itaipu joined
17:38
kyclark left
|
|||
viki | Except they seem to be copying it as it were a year ago: "Stable release Rakudo Star / November 28, 2015" | 17:38 | |
geekosaur | that's sadly common | 17:42 | |
(copying old revs of wikipedia, that is. usually weaponized with adware or malware) | |||
17:43
dakkar left
|
|||
jnthn | [Coke]: D'oh, I missed the negation case. | 17:49 | |
17:49
ufobat joined
|
|||
jnthn | [Coke]: Left myself a reminder at github.com/jnthn/p6-file-ignore/issues/1 :) | 17:50 | |
17:51
domidumont1 left
17:52
kyclark joined
17:54
domidumont joined
17:59
domidumont left
18:20
Actualeyes left
18:37
domidumont joined
18:42
dextertzu left
|
|||
viki | What is the benefit in writing 2*$fract >= 1 rather than $fract >= .5 when deciding which way to round? | 18:44 | |
.oO( or is there any... ) |
18:46 | ||
mst | possibly the 2* approach is cargo-cultish floating point bug avoidance? | ||
viki | hm. OK, I'll leave it in then | ||
DrForr | One's a shift and int comparison, the other is a floating-point comparison. | ||
18:47
kyclark left
18:49
kyclark joined
18:50
dextertzu joined
|
|||
El_Che | why: $15,563 already? www.kickstarter.com/projects/14228...ing-perl-6 | 18:53 | |
1 10k pledge | |||
DrForr | Liz & Wendy, I'm pretty sure. | 18:54 | |
18:54
nnunley joined
|
|||
viki | I was thinking the same :) | 18:54 | |
El_Che | it sounds like them, indeed :) | ||
viki | lizmat++ wolfy++ | ||
El_Che | don't guilt-shame them in case it's not them :) | ||
I was having a look which pledge to take, but 10k is not happening :) | 18:55 | ||
[Coke] | oh, there was a 10K pledge? that explains how it jumped 10K between me leaving work and getting home yesterday! :) | ||
El_Che | well, the good thing about Trump ois that the dollar is cheaper now :) | 18:57 | |
DrForr | And thanks to Brexit so's the pound :) | 18:58 | |
El_Che | well, it will be handy when someone in the uk kickstarts a perl 6 book | ||
viki | So what you're saying the value of my monopoly Canukistan dollars goes us by as much as the rest of the world is getting trashed with bad decisions? | 18:59 | |
El_Che | about that? is Timtowdy or someone else working on the Camel book succesor? | ||
viki | Sweet. | ||
Damian promised a book "in a year" from last April. | 19:00 | ||
...but maybe he just got scared with the audience of 30 people starting at him, damanding a book :P | |||
El_Che | haha | ||
just throw them some bones and run away | |||
(the pledges are not cumulative, right? If you go for the t-shirt you won't get the ebook, I guess) | 19:02 | ||
(I mean the early version pdf's) | |||
DrForr | El_Che: Read the comments. | ||
El_Che | ah cool | ||
19:07
jdv79 left
19:10
kyclark left
19:12
jdv79 joined
|
|||
kalkin-_ | Why in Perl6 if I'm using pipes i.e @array ==> sort() ==> unique() …, do i have to write ()? | 19:13 | |
Why can't i just do @array ==> sort ==> unique … | |||
19:14
kyclark joined
|
|||
viki | They're called feeds. | 19:15 | |
The … has very low precedence | 19:16 | ||
(why? dunno) | |||
kalkin-_ | viki: ahh feeds, thanks | ||
viki | The Ancients who designed the language may know why it's so low. | ||
kalkin-_ | Viki … was just a replacement for etc, I'm asking about the reasoning behind the () | 19:17 | |
s/etc/ect/ | |||
viki | Oh around sort() and unique() and stuff | ||
lizmat | El_Che: we're not guilt shamed that easily :-) | ||
viki | m: sort | ||
camelia | rakudo-moar 7f3224: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Calling sort() will never work with any of these multi signatures: (@values)  ($cmp, + is raw)at <tmp>:1------> 3<BOL>7⏏5sort» | ||
viki | kalkin-_: possibly to disambiguate whether you want the feed stuff to be used as an argument to the subs? | 19:18 | |
viki knows nothing about feeds :( | |||
19:19
darutoko left
|
|||
kalkin-_ | m: 3..1 ===> sort() | 19:19 | |
camelia | rakudo-moar 7f3224: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Preceding context expects a term, but found infix > insteadat <tmp>:1------> 033..1 ===>7⏏5 sort()» | ||
kalkin-_ | m: (3...1) ===> sort() | ||
camelia | rakudo-moar 7f3224: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Preceding context expects a term, but found infix > insteadat <tmp>:1------> 3(3...1) ===>7⏏5 sort()» | ||
viki | m: 3..1 ==> sort() | ||
camelia | ( no output ) | ||
viki | m: say 3..1 ==> sort() | ||
camelia | rakudo-moar 7f3224: OUTPUT«3..1» | ||
viki | m: say 3..1 ==> sort | ||
camelia | rakudo-moar 7f3224: OUTPUT«3..1» | ||
psch | m: say (unique() <== sort() <== [1,1,3,2,5,2,1,6]) | 19:20 | |
camelia | rakudo-moar 7f3224: OUTPUT«(1 2 3 5 6)» | ||
kalkin-_ | argle one = to many | ||
ahh i understand I didn't know that i can call a method like this: sort 1...3 | 19:21 | ||
I always thought I need to use sort: 1...3 | |||
psch | kalkin-_: that's because you can't. that 'sort' isn't a method | ||
m: say &sort eqv Any.^can('sort') | |||
camelia | rakudo-moar 7f3224: OUTPUT«False» | ||
psch | m: say &sort; say Any.^can('sort') | ||
camelia | rakudo-moar 7f3224: OUTPUT«sub sort (| is raw) { #`(Sub|69319472) ... }(Method+{<anon|73532432>}.new)» | ||
kalkin-_ | When do I want to use ':' after a function? | 19:22 | |
viki | kalkin-_: it's just a sub. Perl 6 is a multiparadigm language :) | ||
kalkin-_ | I.e what is the difference between sort 1..3 and sort_:_ 1...3? | ||
psch | anyway, the feed operators have the second lowest precedence, just above ; (i.e. statement end) and { } (i.e. block curlies) | 19:23 | |
viki | One works and another doesn't? | ||
m: say sort: 1...3 | |||
camelia | rakudo-moar 7f3224: OUTPUT«5===SORRY!5===Calling sort() will never work with any of these multi signatures: (@values)  ($cmp, + is raw)at <tmp>:1------> 3say 7⏏5sort: 1...3Variable '&infix:<:>' is not declaredat <tmp>:1------> 3say sort7⏏…» | ||
viki | m: say sort 1..3: | ||
camelia | rakudo-moar 7f3224: OUTPUT«(1 2 3)» | ||
viki | ^ in that case it's a method call | ||
psch | uh wow, what about that &infix:<:> there..? | ||
oh, right, colon in the wrong spot does that | 19:24 | ||
kalkin-_ | viki: ohh, hmm i think i misunderstood before | ||
[Coke] | m: <1 3 2> ==> sort() ==> say | ||
camelia | rakudo-moar 7f3224: OUTPUT«5===SORRY!5===Argument to "say" seems to be malformedat <tmp>:1------> 3<1 3 2> ==> sort() ==> say7⏏5<EOL>Other potential difficulties: Unsupported use of bare "say"; in Perl 6 please use .say if you meant $_, or use an explicit…» | ||
kalkin-_ | psch: thanks | ||
[Coke] | m: <1 3 2> ==> sort() ==> say() | ||
camelia | rakudo-moar 7f3224: OUTPUT«(1 2 3)» | ||
[Coke] | m: <1 3 2> ==> { sort } ==> { say } | ||
viki | kalkin-_: to be fair, you're not the first one I see write colons after a sub. Maybe it worked some time in the past?> | ||
camelia | rakudo-moar 7f3224: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Unsupported use of bare "say"; in Perl 6 please use .say if you meant $_, or use an explicit invocant or argument, or use &say to refer to the function as a nounat <tmp>:1------> 3<1 3 2> ==> { sort }…» | ||
[Coke] | m: <1 3 2> ==> { sort } ==> { .say } | ||
camelia | rakudo-moar 7f3224: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Sorry, do not know how to handle this case of a feed operator yet.at <tmp>:1------> 3<1 3 2> ==> { sort } ==> { .say }7⏏5<EOL> expecting any of: postfix» | ||
psch | m: <1 3 2> ==> &sort # is this supposed to work or is the error just an artifact? | 19:25 | |
camelia | rakudo-moar 7f3224: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Sorry, do not know how to handle this case of a feed operator yet.at <tmp>:1------> 3o work or is the error just an artifact?7⏏5<EOL>» | ||
kalkin-_ | m: sort: 1...3 | ||
camelia | ( no output ) | ||
kalkin-_ | m: say sort: 1...3 | ||
camelia | rakudo-moar 7f3224: OUTPUT«5===SORRY!5===Calling sort() will never work with any of these multi signatures: (@values)  ($cmp, + is raw)at <tmp>:1------> 3say 7⏏5sort: 1...3Variable '&infix:<:>' is not declaredat <tmp>:1------> 3say sort7⏏…» | ||
19:25
RabidGravy joined
|
|||
kalkin-_ | m: sort: 1...3 # this works in repl | 19:26 | |
camelia | ( no output ) | ||
viki | m: dd (sort: 1...3) | ||
camelia | rakudo-moar 7f3224: OUTPUT«(1, 2, 3).Seq» | ||
psch | m: sort: 1..3; say sort.WHAT | ||
camelia | rakudo-moar 7f3224: OUTPUT«WARNINGS for <tmp>:Useless use of ".." in expression "1..3" in sink context (line 1)(Label)» | ||
psch | m: quietly sort: 1..3; say sort.WHAT # for clarity | ||
camelia | rakudo-moar 7f3224: OUTPUT«(Label)» | ||
kalkin-_ | m: sort 1...3; say sort.WHAT | ||
camelia | rakudo-moar 7f3224: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Calling sort() will never work with any of these multi signatures: (@values)  ($cmp, + is raw)at <tmp>:1------> 3sort 1...3; say 7⏏5sort.WHAT» | ||
viki | ahh | ||
psch++ | |||
kalkin-_ | m: sort: 1...3; say sort.WHAT | 19:27 | |
camelia | rakudo-moar 7f3224: OUTPUT«(Label)» | ||
psch | m: sort: 1..3; redo sort unless $++ | ||
camelia | rakudo-moar 7f3224: OUTPUT«WARNINGS for <tmp>:Useless use of ".." in expression "1..3" in sink context (line 1)labeled redo without loop construct in block <unit> at <tmp> line 1» | ||
psch | :P | ||
19:27
andrzejku joined
|
|||
psch | kalkin-_: try 'sort: 3,2,1' in the repl instead maybe? :) | 19:27 | |
kalkin-_ | I have no idea where i picked up the colons form | ||
FROGGS | huh, someone broke Label.gist it seems | 19:28 | |
psch | kalkin-_: probably misplaced intuition. "sort: 1,2,3" seems more natural, but we demand the invocant marker *after* the invocant, not before | ||
kalkin-_ | psch: sort: 3,2,1 & sort 3...1 works in repl | ||
psch | kalkin-_: does it print 3,2,1? :) | ||
kalkin-_ | aehm | ||
the second one is sort_:_ 3...1 | |||
psch | right, but it doesn't sort | 19:29 | |
is my point | |||
i don't recall if your repl actually prints non-printy stuff though... | |||
kalkin-_ | ahh | ||
viki | m: sort: 1...3; say sort | ||
camelia | rakudo-moar 7f3224: OUTPUT«Cannot find method 'Stringy': no method cache and no .^find_method in block <unit> at <tmp> line 1» | ||
kalkin-_ | I haven't noticed that it doesn't sort | ||
viki | m: sort: 1...3; say sort.gist | ||
camelia | rakudo-moar 7f3224: OUTPUT«Cannot find method 'Stringy': no method cache and no .^find_method in block <unit> at <tmp> line 1» | ||
dalek | osystem: ebc1542 | (Paweł Szulc)++ | META.list: Added Statistics::LinearRegression See github.com/hipek8/p6-Statistics-Li...Regression |
||
osystem: e31bda4 | RabidGravy++ | META.list: Merge pull request #266 from hipek8/master Added Statistics::LinearRegression |
|||
psch | point is, [ \w+ ':' .+ ] gets parsed as a Label followed by a statement | ||
viki | bisect: sort: 1...3; say sort.gist | 19:30 | |
bisectable6 | viki, Bisecting by exit code (old=2015.12 new=7f32243). Old exit code: 0 | ||
psch | err, plus anchors :S | ||
FROGGS | viki: multi method gist(Label:D:) { | ||
bisectable6 | viki, bisect log: gist.github.com/f53d92a6264535b1bd...7f86bcf1c7 | ||
viki, (2016-07-19) github.com/rakudo/rakudo/commit/34...c39db42ab8 | |||
psch | so [ ^ \w+ ':' .+ ;? $] # or something like that | ||
AlexDaniel | that being said: | ||
m: say sort 1..3: | |||
camelia | rakudo-moar 7f3224: OUTPUT«(1 2 3)» | ||
FROGGS | :S | ||
psch | m: say unique sort 3,1,1,5,2,5: : | 19:31 | |
camelia | rakudo-moar 7f3224: OUTPUT«(1 2 3 5)» | ||
AlexDaniel | which should not be that surprising | ||
kalkin-_ | thanks for the explanation. Any hints where i can read up on labels? | ||
19:31
atweiden left
|
|||
psch | docs.perl6.org/type/Label probably? | 19:31 | |
AlexDaniel | psch: although I am surprised that you can nest it like that… | ||
psch | AlexDaniel: it gets wonky though | 19:32 | |
m: say unique sort 3,1,1,5,2,5: : : # | |||
camelia | rakudo-moar 7f3224: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Confusedat <tmp>:1------> 3say unique sort 3,1,1,5,2,5: : :7⏏5 # expecting any of: colon pair» | ||
19:32
cdg left
|
|||
psch | AlexDaniel: although that's probably &say being somewhat special :) | 19:32 | |
AlexDaniel | m: say 25 : | ||
camelia | rakudo-moar 7f3224: OUTPUT«25» | ||
AlexDaniel | not really? | ||
m: say 25,40 : | |||
camelia | rakudo-moar 7f3224: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Confusedat <tmp>:1------> 3say 25,40 :7⏏5<EOL> expecting any of: colon pair» | ||
AlexDaniel | m: say (25,40) : | ||
camelia | rakudo-moar 7f3224: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Confusedat <tmp>:1------> 3say (25,40) :7⏏5<EOL> expecting any of: colon pair» | ||
FROGGS | viki: let me try to fix it | 19:33 | |
AlexDaniel | m: say sort (25,40) : | ||
camelia | rakudo-moar 7f3224: OUTPUT«(25 40)» | ||
psch | m: sort unique (2, 12, 1, 2): : | ||
camelia | rakudo-moar 7f3224: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Confusedat <tmp>:1------> 3sort unique (2, 12, 1, 2): :7⏏5<EOL> expecting any of: colon pair» | ||
psch | hm, maybe just the outer most doesn't parse | ||
although i can't imagine why | |||
might be we commit to a sub call somewhere along the way..? dunno, haven't looked at the grammar in a few weeks :/ | 19:34 | ||
bisect: sort unique (2, 12, 1, 2): : | |||
bisectable6 | psch, On both starting points (old=2015.12 new=7f32243) the exit code is 1 and the output is identical as well | ||
psch, Output on both points: ===SORRY!=== Error while compiling /tmp/r2wO3Ni78NConfusedat /tmp/r2wO3Ni78N:1------> sort unique (2, 12, 1, 2): :<HERE>  expecting any of: colon pair | |||
psch | well, apparently it's been like that for a while | ||
AlexDaniel | commit: all sort unique (2, 12, 1, 2): : | ||
19:34
kalkin-_ left
19:35
kalkin-_ joined
|
|||
psch | m: $a = sort unique (2, 12, 1, 2): : | 19:35 | |
camelia | rakudo-moar 7f3224: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Variable '$a' is not declaredat <tmp>:1------> 3<BOL>7⏏5$a = sort unique (2, 12, 1, 2): :» | ||
psch | m: my $a = sort unique (2, 12, 1, 2): : | ||
camelia | rakudo-moar 7f3224: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Confusedat <tmp>:1------> 3my $a = sort unique (2, 12, 1, 2): :7⏏5<EOL> expecting any of: colon pair» | ||
committable6 | AlexDaniel, gist.github.com/f110e1039926477f4c...59c4f694ae | ||
psch | aha, 2015.03 and earlier explicitly forbade it | ||
then came the confusion | |||
honestly, it seems weird to forbid it there | 19:36 | ||
i don't see the difference between allowing nesting inside and argument list and allowing nesting up to the statement level | |||
so either nested invocant markers should be all the way gone or they should be fine up to statement level | |||
FROGGS | $ perl6 -e 'say: say(say)' | ||
Label<say>(at -e:1, '⏏say: say(say)') | |||
psch | and considering a single invocant marker on statement level is fine, it should probably be fine? | ||
FROGGS++ | |||
kalkin-_ | Why can I do say 1…3 but not sort 1…3 ? | 19:38 | |
has sort some kind of special precedence? | |||
viki | kalkin-_: you can do both | ||
m: say sort 1…3 | |||
camelia | rakudo-moar 7f3224: OUTPUT«(1 2 3)» | ||
viki | m: say sort 3…1 | ||
camelia | rakudo-moar 7f3224: OUTPUT«(1 2 3)» | ||
kalkin-_ | m: sort 1…3 | 19:39 | |
camelia | ( no output ) | ||
kalkin-_ | Repl throws me an error on that | ||
Error while compliing: Two terms in a row | |||
psch | kalkin-_: is your sort maybe still a Label? | ||
viki | kalkin-_: what sort of error? | ||
kalkin-_ | psch: good point | ||
psch | m: sort: 1; sort 1..3 | ||
camelia | rakudo-moar 7f3224: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Two terms in a rowat <tmp>:1------> 3sort: 1; sort7⏏5 1..3 expecting any of: infix infix stopper statement end statement modifier statement…» | ||
kalkin-_ | yes it was still a label, thank you psch | ||
psch | m: X: 1; X::AdHoc.new.throw: msg => "phooey" # /o\ | 19:40 | |
camelia | rakudo-moar 7f3224: OUTPUT«WARNINGS for <tmp>:Useless use of constant integer 1 in sink context (line 1)Could not find symbol '&AdHoc' in block <unit> at <tmp> line 1Actually thrown at: in block <unit> at <tmp> line 1» | ||
19:40
nicq20 joined
|
|||
psch | we probably still have a ticket about 'class X' sitting somewhere... | 19:41 | |
m: sub f(Int(Cool) $x) {$x}; Int: quietly 1; f 2.0 | 19:42 | ||
camelia | rakudo-moar 7f3224: OUTPUT«===SORRY!===P6opaque: get_boxed_ref could not unbox for the representation '20' of type Label» | ||
psch | snrk | ||
AlexDaniel | m: say (sort (2, 12, 1, 2):): | ||
camelia | rakudo-moar 7f3224: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Confusedat <tmp>:1------> 3say (sort (2, 12, 1, 2):):7⏏5<EOL> expecting any of: colon pair» | ||
AlexDaniel | ah wait | 19:44 | |
m: say sort unique (2…3…1): : | 19:45 | ||
camelia | rakudo-moar 7f3224: OUTPUT«(1 2 3)» | ||
AlexDaniel | psch: it kinda works | 19:46 | |
m: say sort unique [3, 5, 3, 1]: : | |||
camelia | rakudo-moar 7f3224: OUTPUT«(1 3 5)» | ||
19:47
dextertzu left
|
|||
AlexDaniel | ah, you mean without say… | 19:49 | |
m: sort unique (2…3…1): : | |||
camelia | ( no output ) | ||
AlexDaniel | m: sort unique [3, 5, 3, 1]: : | ||
camelia | rakudo-moar 7f3224: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Confusedat <tmp>:1------> 3sort unique [3, 5, 3, 1]: :7⏏5<EOL> expecting any of: colon pair» | ||
AlexDaniel | I have no idea | ||
19:50
andrzejku left
|
|||
psch | yeah, there's definitely some weird state in the parser making things... weird | 19:50 | |
dalek | c: 3fd8f9d | (Zoffix Znet)++ | doc/Type/Real.pod6: Direct to parse-base as reverse of base Since the :16("") notation has some magiks in it and it can't parse negatives |
||
synopsebot6 | Link: doc.perl6.org/type/Real | ||
psch | s:2nd/weird/inconsistent/ # probably | ||
from what i remember it might just come down to the non-parens'd branch not expecting an invocant marker anymore for some reason | 19:51 | ||
but yeah, been some time since i spend time with the grammar | |||
19:53
mohae_ joined
|
|||
dalek | c: 3dd93ed | (Zoffix Znet)++ | doc/Type/Real.pod6: Document behaviour of Rational.base(Int:D, Whatever) |
19:54 | |
synopsebot6 | Link: doc.perl6.org/type/Real | ||
19:56
mohae left
|
|||
viki | hm | 19:57 | |
psch | r: say sort 1: : | ||
camelia | rakudo-moar 2432c4, rakudo-jvm 76b061: OUTPUT«(1)» | ||
psch | r: say sort 2,1: : | ||
camelia | rakudo-jvm 76b061: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Confusedat <tmp>:1------> 3say sort 2,1: :7⏏5<EOL> expecting any of: colon pair» | ||
..rakudo-moar 2432c4: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Confusedat <tmp>:1------> 3say sort 2,1: :7⏏5<EOL> expecting any of: colon pair» | |||
psch | wicked | ||
r: say (sort 2,1:) : | 19:58 | ||
camelia | rakudo-moar 2432c4: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Confusedat <tmp>:1------> 3say (sort 2,1:) :7⏏5<EOL> expecting any of: colon pair» | ||
..rakudo-jvm 76b061: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Confusedat <tmp>:1------> 3say (sort 2,1:) :7⏏5<EOL> expecting any of: colon pair» | |||
psch blames the comma | |||
viki | r: say sort (2,1): : | ||
camelia | rakudo-moar 2432c4: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Confusedat <tmp>:1------> 3say sort (2,1): :7⏏5<EOL> expecting any of: colon pair» | ||
..rakudo-jvm 76b061: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Confusedat <tmp>:1------> 3say sort (2,1): :7⏏5<EOL> expecting any of: colon pair» | |||
19:59
ufobat left
|
|||
psch | r: say (sort 1:) : | 19:59 | |
camelia | rakudo-moar 2432c4, rakudo-jvm 76b061: OUTPUT«(1)» | ||
20:00
dextertzu joined
20:04
brillenfux left
20:06
Tonik joined
|
|||
viki | What are the 10 extra second agbout? | 20:10 | |
*about | |||
m: say Instant.from-posix: 100 | |||
camelia | rakudo-moar 2432c4: OUTPUT«Instant:110» | ||
psch | j: say Instant.from-posix: 100 | ||
camelia | ( no output ) | ||
psch | j: say Instant.from-posix: 100 | ||
camelia | rakudo-jvm 76b061: OUTPUT«Instant:110» | ||
psch | leap seconds? | 20:11 | |
viki doesn't follow how those fit in there | |||
psch | j: say Instant.from-posix: 100, :!prefer-leap-second | ||
camelia | rakudo-jvm 76b061: OUTPUT«Instant:110» | ||
psch | yeah, it was a completely blind guess | 20:12 | |
"Nil while ($i = $i + 1) < $elems && $p > nqp::atpos($posixes,$i);" | 20:13 | ||
seems kind of not-that-self-documenting, that line | |||
in Rakudo/Internals.pm:1546, .tai-from-posix | |||
which is what's behind Instant.from-posix | |||
20:13
sufrostico left
|
|||
viki | Hm github.com/rakudo/rakudo/blob/7f32...s.pm#L1517 | 20:14 | |
"# TAI - UTC at the Unix epoch (1970-01-01T00:00:00Z)." | |||
psch | ah, so it *is* leap seconds | ||
and we had 10 leap seconds before epoch start | |||
viki | Oh, "TAI is exactly 36 seconds ahead of UTC. The 36 seconds results from the initial difference of 10 seconds at the start of 1972, plus 26 leap seconds in UTC since 1972." | 20:15 | |
psch | j: say Instant.from-posix: 0, :!prefer-leap-second | ||
camelia | rakudo-jvm 76b061: OUTPUT«Instant:10» | ||
psch | m: say Instant.from-posix: 0, :prefer-leap-second | ||
camelia | rakudo-moar 2432c4: OUTPUT«Instant:10» | ||
psch | m: say Instant.from-posix: 0, :!prefer-leap-second | ||
20:15
sufrostico joined
|
|||
camelia | rakudo-moar 2432c4: OUTPUT«Instant:10» | 20:15 | |
psch is completely unclear about what that named does | |||
oh wait | 20:16 | ||
that's the named for tai-from-posix... | |||
ah, but it's called the same in .from-posix | 20:17 | ||
so, yeah, it might have been intended to allow "ignore leap seconds"..? | |||
geekosaur | but that might be buggy if you can't ignore the fixed initial difference | ||
psch | ohh | ||
geekosaur | which might well be a thinko somewhere | ||
psch | prefer-leap-second means "interpret this as a leap second" | 20:18 | |
Xliff | *SNARL* | ||
I went and upgraded libssl on my linux box. | |||
viki | # If $prefer-leap-second is true, 915148800 is interpreted to | ||
# mean 1998-12-31T23:59:60Z rather than 1999-01-01T00:00:00Z | |||
Xliff | OpenSSL 1.1.0 changed the initialization and went and broke the OpenSSL module. | ||
AlexDaniel | I don't want to sound annoying, but… | ||
20:18
domidumont left
|
|||
psch | m: say Instant.from-posix(78796800, $_) for True, False | 20:18 | |
AlexDaniel | what about negative values? | ||
camelia | rakudo-moar 2432c4: OUTPUT«Instant:78796810Instant:78796811» | ||
psch | m: say Instant.from-posix(78796800, $_).Date for True, False | ||
camelia | rakudo-moar 2432c4: OUTPUT«1972-06-301972-07-01» | ||
psch | there it is | ||
Xliff | So OpenSSL support for perl6 will BREAK if users upgrade to OpenSSL 1.0.2 | 20:19 | |
psch | well, assuming there's a positional candidate :P | ||
Xliff | Er... | ||
psch | m: say Instant.from-posix(78796800, prefer-leap-second => $_).Date for True, False | ||
camelia | rakudo-moar 2432c4: OUTPUT«1972-07-011972-07-01» | ||
psch | m: say Instant.from-posix(78796800, prefer-leap-second => $_).DateTime for True, False | ||
camelia | rakudo-moar 2432c4: OUTPUT«1972-07-01T00:00:00Z1972-07-01T00:00:00Z» | ||
Xliff | Let's try that again. -- "So OpenSSL support for perl6 will BREAK if users upgrade to OpenSSL 1.1.0" | ||
viki | Isn't 1.1.0 the buggy one with the DoS bug? | 20:20 | |
RabidGravy | Xliff, I'm sure your patch to mitigate that would be welcomed :) | 20:21 | |
AlexDaniel | m: say Instant.from-posix: -9999999999999999999999999999999999999999999 | ||
camelia | rakudo-moar 2432c4: OUTPUT«Instant:-9999999999999999999999999999999999999999989» | ||
Xliff | RabidGravy: Already looked. It is non-trivial. | 20:22 | |
viki | m: say Instant.from-posix: -10 | ||
camelia | rakudo-moar 2432c4: OUTPUT«Instant:0» | ||
20:23
ocbtec left
|
|||
viki | AlexDaniel: to me it does sound like the 10 extra seconds do not apply prior to start of 1972 | 20:23 | |
m: say Instant.from-posix(-10).DateTime | 20:24 | ||
camelia | rakudo-moar 2432c4: OUTPUT«1969-12-31T23:59:50Z» | ||
viki | hm | ||
m: say Instant.from-posix(10).DateTime | |||
camelia | rakudo-moar 2432c4: OUTPUT«1970-01-01T00:00:10Z» | ||
viki | So we basically always sub them? | ||
From what I'm reading at en.wikipedia.org/wiki/International_Atomic_Time that sounds like a bug | 20:25 | ||
geekosaur | sounds like we're back to the old days of openssl | ||
RabidGravy | yeah | 20:26 | |
20:27
acrussell_ joined
|
|||
Xliff | GRRR.... and there is no graceful way to downgrade without breaking everything. | 20:28 | |
psch | ahh, it's *not* named in .from-posix | ||
so yeah, my first eval there was correct | |||
Xliff | Sooo... | ||
psch | m: say Instant.from-posix(78796800, $_).Date for True, False | ||
camelia | rakudo-moar 2432c4: OUTPUT«1972-06-301972-07-01» | ||
psch | first is leap-second ignorant, second is *-aware | ||
20:28
sufrostico left
|
|||
psch | m: say Instant.from-posix(-10, $_).Date for True, False | 20:29 | |
camelia | rakudo-moar 2432c4: OUTPUT«1969-12-311969-12-31» | ||
psch | m: say Instant.from-posix(-0, $_).Date for True, False | ||
Xliff | Compile openssl 1.0.2 in /usr/local and have a wrapper for my script that forces its use. | ||
camelia | rakudo-moar 2432c4: OUTPUT«1970-01-011970-01-01» | ||
Xliff | yay. | ||
psch | m: say Instant.from-posix(0, $_).Date for True, False | ||
camelia | rakudo-moar 2432c4: OUTPUT«1970-01-011970-01-01» | ||
psch | m: say Instant.from-posix(0, $_).DateTime for True, False | ||
camelia | rakudo-moar 2432c4: OUTPUT«1970-01-01T00:00:00Z1970-01-01T00:00:00Z» | ||
RabidGravy | I've always tended to treat openssl specially in a build, it went through so long being totally bizarre in its changes | ||
psch | ...the first 10 seem to always be there though, viki++ | 20:30 | |
20:31
sufrostico joined
|
|||
viki | AlexDaniel: FWIW, I didn't mean that you open too many tickets or anything here: irclog.perlgeek.de/perl6/2016-11-15#i_13573265 it was just general frustration at dead-end tickets :) | 20:32 | |
AlexDaniel | viki: I never thought about it this way, don't worry | ||
psch . o O( don't dead-end tickets just mean that people with notable insight don't write enough patches..? ) | |||
viki | OK. Just making sure :) | ||
AlexDaniel | thanks :) | 20:33 | |
psch | in any case, AlexDaniel++ for digging into things | ||
AlexDaniel | viki: that being said, will you open a ticket for 10 second issue? | 20:34 | |
I don't understand it as much as you do | |||
viki | Yeah | ||
AlexDaniel | viki++ | ||
psch | m: say Instant.from-posix(100, $_) for True, False | ||
camelia | rakudo-moar 2432c4: OUTPUT«Instant:110Instant:110» | ||
psch | one of those should say "100", probably the first one | ||
so, yeah, +1 on the ticket and all | 20:35 | ||
although, honestly, 'prefer-lead-second' sounds like False is the wrong default..? | |||
...leap | 20:36 | ||
20:36
troys joined
|
|||
RabidGravy | Xliff, my reading of the openssl changes would indicate that you have a choice when it is compilled whether to be backward compatible or not | 20:37 | |
seems strange to package one that isn't | |||
Xliff | Backwards compatible (from what I've read) involves a macro in a header. | 20:38 | |
I'll download the 1.1 source in a few and see. | |||
But every time I seem to get somewhere with this latest project, something always come back to bite me in the ass. | |||
It's frustrating. | |||
20:41
FROGGS left
|
|||
moritz | github.com/supernovus/perl6-http-client/pull/13 preparation for the perl6.org HTTPS redirect. AlexDaniel++ for digging it up | 20:42 | |
dalek | c: 523f08a | (Zoffix Znet)++ | doc/Type/Real.pod6: Document .base(Int:D, Whatever) for all Reals The Whatever was made acceptable by all Reals and not just the Rationals in github.com/rakudo/rakudo/commit/c1...c94acda6a6 |
20:44 | |
synopsebot6 | Link: doc.perl6.org/type/Real | ||
20:45
mohae joined
20:46
TEttinger joined,
wamba left
20:47
mohae_ left
|
|||
Xliff | \o/ | 20:48 | |
Holy shinkies... I fixed it! | |||
Xliff looks at RabidGravy++ | |||
RabidGravy | top man! | ||
20:51
kyclark left
|
|||
viki | Instant.from-posix ticket: rt.perl.org/Ticket/Display.html?id=130123 | 20:52 | |
viki & | |||
lizmat | viki: from the WP article: "Since this system of correction was implemented in 1972, 26 leap seconds have been inserted" | 20:54 | |
viki: "By then, the UTC clock was already 10 seconds behind TAI" | 20:56 | ||
geekosaur | yes, viki quoted a different statement of that earlier | ||
[17 20:15:01] <viki> Oh, "TAI is exactly 36 seconds ahead of UTC. The 36 seconds results from the initial difference of 10 seconds at the start of 1972, plus 26 leap seconds in UTC since 1972." | 20:57 | ||
lizmat | so, anything before 1972 has 10 | ||
since this only goes back to 1970 anyway, the current behaviour feels correct to me | |||
20:57
Tonik left
21:01
kyclark joined
|
|||
geekosaur | I think viki's intepretation is that it was ignored before 1972 so should be 0, with the 10 seconds abruptly appearing at 00:00:00 1 Jan 1972 | 21:01 | |
if nobody was actually keeping track officially before that point, the interpretation makes sense | 21:02 | ||
21:02
cognominal left
21:04
khw left
|
|||
psch | so the 11th leap second was at 78796800, and nobody cared about them before? | 21:05 | |
lizmat | en.wikipedia.org/wiki/Leap_second | ||
"which had been synchronized with UT1 in 1958, " | 21:06 | ||
21:06
cognominal joined
|
|||
lizmat | so, I one could argue that 10 seconds got added in 1958, and between then and 1972 the difference got corrected at 78796800 | 21:06 | |
psch | m: say Instant.from-posix(78796800).DateTime | 21:07 | |
camelia | rakudo-moar 53d7b7: OUTPUT«1972-07-01T00:00:00Z» | ||
21:07
cdg joined
|
|||
psch | ...this comes down to metrology doesn't it | 21:08 | |
i mean, it's about solar seconds vs SI seconds | 21:09 | ||
21:09
sufrostico left
|
|||
psch | the question is, are the precise instants for the 10 pre-1972 leap seconds | 21:09 | |
because if not we can just do whatever, where one possibility is "we already are doing the right thing" | 21:10 | ||
*are there | |||
21:11
ydhf joined,
sufrostico joined,
itaipu left
|
|||
ydhf | sometimes I ask myself what would have been Perl 6 if it would have fully embraced POSIX apis at the nativecall level throwing away multiplatform and fancy things like multithreading (a part for the huge amount of stuff POSIX already gives you) but keeping the syntax and features | 21:14 | |
lizmat | feels like an ecosystem module to me :-) | 21:16 | |
ydhf | an huge one | 21:17 | |
stmuk_ | there is one | ||
21:18
firstdayonthejob joined
|
|||
moritz | but it's tiny at the moment, no? | 21:18 | |
stmuk_ | yes but its just case a case of adding nativecall | 21:19 | |
github.com/cspencer/perl6-posix/bl.../POSIX.pm6 | |||
21:20
espadrine joined
|
|||
psch | dunno, integrating POSIX on a language level seems way more work than what POSIX.pm does currently | 21:20 | |
i mean, start {} would start posix thread, wouldn't it? | |||
+a | |||
lizmat | psch: but "start" doesn't start a thread | ||
start merely schedules code to be executed at a later time by another thread | 21:21 | ||
ydhf | pthread_create would be ok | ||
$ | |||
would it work? | |||
psch | lizmat: that's kind of my point, start would have to schedule a pthread, instead of a vm thread | ||
lizmat | it does not necessarily start a tread | ||
psch | lizmat: sorry if i wasn't particularly clear in my wording | ||
21:21
Vynce joined
|
|||
moritz | ydhf: try it and see! | 21:22 | |
psch | m: use NativeCall; sub pthread_create($, $, &, $) is native('pthread') {*} | 21:23 | |
camelia | rakudo-moar 53d7b7: OUTPUT«Potential difficulties: In 'pthread_create' routine declaration - Not an accepted NativeCall type for parameter [1] : Any --> For Numerical type, use the appropriate int32/int64/num64... at <tmp>:1 ------> 3ate($, $, &, $) is nat…» | ||
psch | well... | ||
psch isn't gonna dig up the corresponding defines... | |||
m: use NativeCall; sub pthread_create(Pointer $, Pointer $, &, Pointer $) is native('pthread') {*} | |||
camelia | ( no output ) | ||
psch | that probably almost works :P | ||
21:25
sufrostico left
|
|||
stmuk_ | isnt the whole point of perl6 that you dont have to do that? | 21:26 | |
RabidGravy | quite | ||
moritz | you really don't have to | ||
psch | right, the hypothetical is "what if Perl 6 was a POSIX extension" | ||
21:30
nadim_ left
21:31
nadim_ joined
21:33
atweiden-air joined
|
|||
atweiden-air | in a regex, is it possible to specify that a general quantifier (**) is an even number? | 21:35 | |
21:36
itaipu joined
21:38
firstdayonthejob left
21:40
pmurias left,
nicq20 left,
firstdayonthejob joined
|
|||
AlexDaniel | atweiden-air: well kinda | 21:42 | |
geekosaur | ** {0, {$_ + 2} ... *} # something like? | ||
AlexDaniel | 0, 2 ... * should work too | ||
m: .say for 0, 2 ... * | |||
oops | |||
m: say (for 0, 2 ... *) | |||
camelia | rakudo-moar 53d7b7: OUTPUT«(timeout)0246810121416182022242628303234363840424446485052545658606264666870727476788082848688909294969810…» | 21:43 | |
rakudo-moar 53d7b7: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Missing blockat <tmp>:1------> 3say (for 0, 2 ... *7⏏5) expecting any of: block or pointy block» | |||
AlexDaniel | the problem is that you probably want longest token matching | ||
21:43
troys is now known as troys_
|
|||
AlexDaniel | but it seems to prefer shorter ones, if I'm not mistaken | 21:43 | |
atweiden-air | m: say so 'ww' ~~ / w ** {0, 2 ... *}/ | 21:44 | |
camelia | rakudo-moar 53d7b7: OUTPUT«Cannot .elems a lazy list in block <unit> at <tmp> line 1Actually thrown at: in block <unit> at <tmp> line 1» | ||
AlexDaniel | so just add a <?{}> check | 21:45 | |
m: say ‘abcdef’ ~~ /(.*) <?{$0.chars %% 2}> / | |||
camelia | rakudo-moar 53d7b7: OUTPUT«「abcdef」 0 => 「abcdef」» | ||
AlexDaniel | m: say ‘abcdefg’ ~~ /(.*) <?{$0.chars %% 2}> / | ||
camelia | rakudo-moar 53d7b7: OUTPUT«「abcdef」 0 => 「abcdef」» | ||
AlexDaniel | m: say ‘abcdefg’ ~~ /^ (.*) <?{$0.chars %% 2}> $/ | 21:46 | |
camelia | rakudo-moar 53d7b7: OUTPUT«Nil» | ||
AlexDaniel | m: say ‘abcdef’ ~~ /^ (.*) <?{$0.chars %% 2}> $/ | ||
camelia | rakudo-moar 53d7b7: OUTPUT«「abcdef」 0 => 「abcdef」» | ||
AlexDaniel | atweiden-air: any reason why that ↑ is not good enough? | ||
perlpilot | That's probably closer to what's wanted anyway. (using <?{}> like that) | ||
atweiden-air | thanks, will use that | ||
AlexDaniel | though you will have to rely on backtracking in this case… | 21:47 | |
perlpilot | btw, I'm pretty sure any solution like 'ww' ~~ / w ** {0, 2 ... *} / isn't going to do the right thing anyway. For example ... | 21:50 | |
m: "aaaa" ~~ / a ** {0,2,4,6} /; # works? | |||
camelia | ( no output ) | ||
perlpilot | m: "aaaa" ~~ / a ** {0,2,4,6} /; say $/; # works? | ||
camelia | rakudo-moar 53d7b7: OUTPUT«「aaaa」» | ||
perlpilot | m: "aaa" ~~ / a ** {0,2,4,6} /; say $/; # nope | ||
camelia | rakudo-moar 53d7b7: OUTPUT«Nil» | ||
AlexDaniel | number of elems, yeah | ||
it does .elems on it | 21:51 | ||
atweiden-air | m: my token item { \h* <?{$0.chars %% 2}> w+ } ; say so 'w' ~~ &item | 21:52 | |
camelia | rakudo-moar 53d7b7: OUTPUT«Use of Nil in string context in block at <tmp> line 1True» | ||
perlpilot | It should die with an NYI error IMHO as it could easily confuse someone | ||
AlexDaniel | m: say ‘abcdefgh’ ~~ /.**{2|5}/ # this is sad too | ||
camelia | rakudo-moar 53d7b7: OUTPUT«This type (Junction) does not support positional operations in block <unit> at <tmp> line 1» | ||
AlexDaniel | atweiden-air: you didn't capture it with () | ||
atweiden-air | right | 21:53 | |
AlexDaniel | m: say ‘abcdefgh’ ~~ /.**{Inf}/ # this is sad too | ||
camelia | rakudo-moar 53d7b7: OUTPUT«This type (Failure) does not support positional operations in block <unit> at <tmp> line 1» | ||
AlexDaniel | m: say ‘abcdefgh’ ~~ /.**{NaN}/ # this is sad too | ||
camelia | rakudo-moar 53d7b7: OUTPUT«This type cannot unbox to a native integer: P6opaque, Failure in block <unit> at <tmp> line 1» | ||
AlexDaniel | m: say ‘abcdefgh’ ~~ /.**{3.4}/ # this is sad too | ||
camelia | rakudo-moar 53d7b7: OUTPUT«「abc」» | ||
21:54
sufrostico joined
|
|||
AlexDaniel | m: say ‘abcdefgh’ ~~ /.** 3..5/ | 21:55 | |
camelia | rakudo-moar 53d7b7: OUTPUT«「abcde」» | ||
AlexDaniel | and this is also not a normal range syntax | ||
m: say ‘abcdefgh7’ ~~ /.** 3..5..8/ | |||
camelia | rakudo-moar 53d7b7: OUTPUT«Nil» | ||
AlexDaniel | m: say ‘abcdefgh7’ ~~ /.** 3..5..✓/ | ||
camelia | rakudo-moar 53d7b7: OUTPUT«5===SORRY!5=== Error while compiling <tmp>Unrecognized regex metacharacter ✓ (must be quoted to match literally)at <tmp>:1------> 3say ‘abcdefgh7’ ~~ /.** 3..5..7⏏5✓/ expecting any of: infix stopper» | ||
AlexDaniel | oops | ||
m: say ‘abcdefgh7’ ~~ /.** 3..5..7/ | |||
camelia | rakudo-moar 53d7b7: OUTPUT«「bcdefgh7」» | ||
21:58
lizmat left
|
|||
AlexDaniel | meh | 22:00 | |
m: say ‘abcdefgh’ ~~ /.** {0..Inf}/ | |||
camelia | rakudo-moar 53d7b7: OUTPUT«「abcdefgh」» | ||
AlexDaniel | m: say ‘abcdefgh’ ~~ /.** {0..-Inf}/ | ||
camelia | rakudo-moar 53d7b7: OUTPUT«Range maximum in quantifier (**) cannot be -Inf in block <unit> at <tmp> line 1» | ||
AlexDaniel | cool, somebody special-cased -Ing, right? | ||
m: say ‘abcdefgh’ ~~ /.** {0..NaN}/ | |||
camelia | rakudo-moar 53d7b7: OUTPUT«This type cannot unbox to a native integer: P6opaque, Failure in block <unit> at <tmp> line 1» | ||
AlexDaniel | but not NaN! | 22:01 | |
22:02
acrussell_ left
|
|||
AlexDaniel | m: say ‘abc’ ~~ /^(.** {-10..-5}) {say $0} <!>/ | 22:03 | |
camelia | rakudo-moar 53d7b7: OUTPUT«「abc」「ab」「a」「」Nil» | ||
AlexDaniel | and negative numbers are bugged | ||
22:05
Dunearhp left
22:06
Dunearhp joined
|
|||
AlexDaniel | ah dammit | 22:17 | |
m: say ‘abcefghij’ ~~ / .**{2..^3} / | |||
camelia | rakudo-moar 53d7b7: OUTPUT«「abc」» | ||
AlexDaniel | m: say (2^..^2).elems | ||
camelia | rakudo-moar 53d7b7: OUTPUT«0» | ||
AlexDaniel | m: say ‘abcefghij’ ~~ / .**{2..^2} / | ||
camelia | rakudo-moar 53d7b7: OUTPUT«「ab」» | ||
AlexDaniel | m: say ‘abcefghij’ ~~ / .**{2^..^2} / | ||
camelia | rakudo-moar 53d7b7: OUTPUT«「ab」» | ||
22:22
wamba joined
|
|||
RabidGravy | heads up for anyone interested, there's something leaky around Capture literals - will track down and golf | 22:26 | |
anyone nighty night | |||
AlexDaniel | m: say ‘abcefghij’ ~~ / .**{2..1} / | 22:27 | |
camelia | rakudo-moar 53d7b7: OUTPUT«Nil» | ||
AlexDaniel | so this doesn't match anything, makes sense? | ||
22:27
Vynce left
|
|||
AlexDaniel | yeah | 22:27 | |
m: say ‘abcefghij’ ~~ / .**{2..0} / | |||
camelia | rakudo-moar 53d7b7: OUTPUT«「abcefghij」» | ||
AlexDaniel | /o\ | ||
22:31
RabidGravy left
|
|||
[Coke] | viki: 3fd8f9d0f4a1 in perl6/doc breaks with: L«C<parse-base>/routine/parse-base» . getting a failure in 'make xtest' | 22:32 | |
er, file is doc/Type/Real.pod6 | 22:33 | ||
perl6 xt/aspell.t doc/Type/Real.pod6 shows the error. | |||
er, warning. | |||
samcv | is there a perl 6 module to find all url's in a text or should I make one? | 22:36 | |
El_Che | samcv: make one, I use Perl5 URI::Find on some perl 5 scripts :) | 22:37 | |
samcv | kk :) | ||
22:38
kurahaupo left,
kurahaupo__ joined
22:43
troys_ is now known as troys
22:45
cpage_ left,
wamba left
22:52
Vynce joined,
Vynce left
23:01
canopus left
23:06
canopus joined
|
|||
dalek | c: 8b85121 | (Zoffix Znet)++ | doc/Type/Real.pod6: Fix broken POD directive [Coke]++ |
23:06 | |
synopsebot6 | Link: doc.perl6.org/type/Real | ||
23:09
seatek joined
23:13
cpage_ joined
|
|||
atweiden-air | is it possible to have more than one C<:my $such-and-such = 2> code block inside a grammar token? | 23:17 | |
timotimo | should be possible, yeah | 23:20 | |
23:22
rindolf left
|
|||
MasterDuke | atweiden-air: github.com/rakudo/rakudo/blob/nom/...r.nqp#L199 | 23:23 | |
23:24
ydhf is now known as trollolollo
23:28
atweiden-air left
23:30
atweiden-air joined
23:32
espadrine left
23:36
trollolollo left
|
|||
atweiden-air | m: gist.github.com/anonymous/86f67673...7ec6927efc | 23:36 | |
camelia | rakudo-moar 53d7b7: OUTPUT«Use of Nil in string context in regex abc at <tmp> line 8[DEBUG] $leading-whitespace: 0[DEBUG] real $leading-whitespace: 2False» | ||
timotimo | you may have to cause the captures to be reified with a {} | 23:37 | |
literally just {} after \h* | |||
atweiden-air | timotimo: what does reified mean in this instance? | 23:38 | |
timotimo | ah, just "turned into a real thing" | 23:40 | |
23:44
kyclark left
23:48
lizmat joined
23:50
troys is now known as troys_
23:52
firstdayonthejob left
23:59
sufrostico left
|