»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.
00:00 tardisx left 00:08 tardisx joined 00:25 sufrostico left 00:27 sufrostico joined 00:35 sufrostico left 00:36 sufrostico joined 00:37 vytas joined 00:38 dfcarpenterak left 00:44 Sgeo_ joined 00:46 Sgeo left 00:48 Sgeo__ joined 00:49 dfcarpenterak joined 00:50 Sgeo_ left 00:52 dfcarpenterak left 00:59 abaugher left 01:09 itaipu joined 01:10 telex left 01:13 tardisx left 01:15 telex joined
perlawhirl who's awake 01:18
01:20 sufrostico left 01:21 tardisx joined 01:23 itaipu left 01:24 Actualeyes joined
sortiz perlawhirl, I'm o/ 01:25
geekosaur fsvo "awake" 01:26
perlawhirl sortiz: this is not explicitly a perl question, but git. specifically making pull request to the ecosystem
do i need to fork the ecosystem repo to make a PR
I'm relatively green to this git stuff 01:27
geekosaur yes. you fork it, make your change(s), and make the pull request from that
perlawhirl righto 01:28
sortiz Not necessarily an explicit fork. If you change only a file, you can edit it on github, that make the dance. 01:29
Util for example when you are changing META.list 01:30
perlawhirl actually, it looks like the last commit is wrong, there's a leading "i" on the last added module in META.list 01:32
or is that some hackery for newly added module?
sortiz Let me see...
01:33 labster left
sortiz Indeed, seems a bug, changing that... 01:33
Xliff_ Hotkeys, if I didn't hate the fscking Metro look so much, I would be right there with you. 01:35
dalek osystem: 9334cbe | (Salvador Ortiz)++ | META.list:
Update META.list

Remove trailing 'i' in P6-Text-Template entry.
Xliff_ Read somewhere that some dude thought that Templates were just another form of spaghetti code.
While yes, some Templates can look god awful (depending on the complexity of the intended output), I find MVC a better design pattern than anything else. 01:36
I don't care if it "compiles" to something unreadable. I just don't want to maintain it.
sortiz perlawhirl++ # For the report. 01:37
Xliff_ www.workingsoftware.com.au/page/You...de_yes_you
sortiz Xliff_, To me the main issue with *some* template engines is that pretend to do too much, and the template ceases to be declarative, to become procedural. :) 01:39
01:39 tardisx left
Xliff_ sortiz, perl5 Template Toolkit, maybe? 01:39
01:40 kid51 left
Xliff_ In the case where the template engine offers too much. Remember: Just because you can doesn't mean you have to. 01:40
perlawhirl sortiz: leading, not trailing... but who cares :D 01:41
Xliff_ perlawhirl, o/
prl a wrl
perlawhirl haha... about to add a pr to the ecosystem
it lives: github.com/0racle/p6-wig 01:42
Xliff_ W00 h00~!
perlawhirl++
sortiz Yes, but the if an abstraction barrier can be trivially bypassed is bad design.
leading, yep :D 01:43
imo a good template engine should offer you a good *macro* system, not a new programming language. 01:45
01:45 ilbot3 left 01:47 ilbot3 joined
dalek osystem: 3795dba | 0racle++ | META.list:
Add wig to ecosystem

See github.com/0racle/p6-wig
01:51
osystem: 99aea7f | (Salvador Ortiz)++ | META.list:
Merge pull request #183 from 0racle/patch-1

Add wig to ecosystem
perlawhirl sortiz: there was an error in the url of my PR :( 02:01
02:02 BenGoldberg left
sortiz perlawhirl, what is the correct one? 02:03
perlawhirl raw.githubusercontent.com/0racle/p...META6.json 02:04
02:04 labster joined
perlawhirl there was a '.git' after the module name before 02:04
so the url was invalid
dalek osystem: fbb91a5 | 0racle++ | META.list:
Add wig to ecosystem

See github.com/0racle/p6-wig
02:05
osystem: 9db9800 | (Salvador Ortiz)++ | META.list:
Merge pull request #184 from 0racle/patch-2

Fix wig URL
perlawhirl sortiz++ 02:06
sortiz Now you can remove your branches.
02:07 labster left 02:10 BenGoldberg joined 02:14 tardisx joined
perlawhirl oh, hrm... 02:16
oh, you mean the forked repos i now have?
02:17 yqt left
sortiz Yep. 02:21
perlawhirl ok done 02:25
sortiz All ready :) 02:26
ZoffixWin m: subset Foo of Int where { $_ > 10_000 or warn "You need a number more than 10,000" };␤␤␤␤my Foo $x = 42; 02:45
camelia rakudo-moar ae3a79: OUTPUT«You need a number more than 10,000 in block at /tmp/7G_HLCRktW line 1␤Type check failed in assignment to $x; expected Foo but got Int (42)␤ in block <unit> at /tmp/7G_HLCRktW line 5␤␤»
02:45 noganex joined
ZoffixWin Is there a way to make this a better error message? I could live with at least the "more than 10,000" line number being line 5; but if that Type check failed message could also be removed (and changed into the "more than 10,000") that would be great. 02:46
02:48 noganex_ left
ZoffixWin m: subset Foo of Int where { $_ > 10_000 or fail "You need a number more than 10,000 on line {(callframe 4).line}" };␤␤␤␤my Foo $x = 42; 02:52
camelia rakudo-moar ae3a79: OUTPUT«You need a number more than 10,000 on line 5␤ in block <unit> at /tmp/NdbPQyk3Bi line 1␤␤»
ZoffixWin ¯\_(ツ)_/¯
AlexDaniel m: subset Foo of Int where { $_ > 10_000 or fail "You need a number more than 10,000 but you passed $_ on line {(callframe 4).line}" }; my Foo $x = 42; 02:54
camelia rakudo-moar ae3a79: OUTPUT«You need a number more than 10,000 but you passed 42 on line 1␤ in block <unit> at /tmp/uPvYWlvgr3 line 1␤␤»
AlexDaniel ZoffixWin: that's a very important bit :)
ZoffixWin AlexDaniel++ 02:55
m: subset Foo of Int where * > 10_000; try { my Foo $x = 10; CATCH { fail "That value does not match the subset" } }; say "That value did match the subset" 03:01
camelia rakudo-moar ae3a79: OUTPUT«That value does not match the subset␤ in block at /tmp/DcnZUWk_xA line 1␤ in block <unit> at /tmp/DcnZUWk_xA line 1␤␤»
ZoffixWin This feels a bit like a hack... is there a better way to check whether the value can be used for a type (a subset in particular)?
03:02 RandalSchwartz joined
RandalSchwartz tommy is giving a perl6 talk 03:02
a question he couldn't answer: 03:03
are instance methods and class methods sharing the same namespace?
or is there a way to say "this foo method is for instances" vs "this foo method is for class"?
or if they share, what can I test inside the method to go either way? 03:04
ZoffixWin m: class Foo { method bar { self.DEFINITE ?? 'instance' !! 'class' } }; say Foo.bar; say Foo.new.bar
camelia rakudo-moar ae3a79: OUTPUT«class␤instance␤»
ZoffixWin This is possible. Unsure about whether it's shared.
RandalSchwartz ahh ok 03:05
passed that along to tommy 03:07
thanks
Xliff_ Wonder how difficult it would be to add perl 6 support to HexChat.... Hmmm.... 03:08
ZoffixWin There's Inline::Perl6 in P5 :)
Xliff_ ZoffixWin, well, I'd have to be sure that the P5 module in HexChat could handle that. 03:09
03:09 nadim left 03:10 Ben_Goldberg joined
Xliff_ ZoffixWin, so does Inline::Perl6 convert P6 code to P5, then? 03:12
Juerd RandalSchwartz: The same method is used as a class method an as an instance method. The signature can specify that the invocant must be undefined (type object => class method) or defined (not a type object => instance method)
03:13 BenGoldberg left
ZoffixWin Xliff_, I've never used it, but from the docs, I see "This module embeds a MoarVM based Rakudo Perl 6 and allows you to run Perl 6 code, load Perl 6 modules, use methods of Perl 6 objects and much more.": metacpan.org/pod/Inline::Perl6 03:13
Xliff_ Ah. OK... so it includes XS code.
That might be tricky on a Windows system. 03:14
Although I -could- run it in a VM and export to the desktop via X-Ming.... hmmmmm....,
Juerd RandalSchwartz: class Example { method cm (Example:U: $arg1, $arg2) { ... }; method im (Example:D: $arg1, $arg2) { ... } } 03:15
Xliff_ ZoffixWin, thanks for the input. 03:16
Juerd RandalSchwartz: If you want a class method and instance method to have the same name, 'multi' can be used as such:
m: class Example { multi method m (Example:U:) { say "class method" }; multi method m (Example:D:) { say "instance method" } }; Example.m; Example.new.m; 03:17
camelia rakudo-moar ae3a79: OUTPUT«class method␤instance method␤»
Juerd RandalSchwartz: If you're not using a multi method, you can determine whether it was called as a class or instance method, simply using 'defined', like: 03:18
m: class Example { method m { say .defined ?? "instance method" !! "class method" } } 03:19
camelia ( no output )
Juerd m: class Example { method m { say .defined ?? "instance method" !! "class method" } }; Example.m; Example.new.m
camelia rakudo-moar ae3a79: OUTPUT«class method␤class method␤»
Juerd Hm.
Ouch, bitten by $_ 03:20
m: class Example { method m { say self.defined ?? "instance method" !! "class method" } }; Example.m; Example.new.m
camelia rakudo-moar ae3a79: OUTPUT«class method␤instance method␤»
Juerd There :)
03:21 tardisx left 03:26 cosarara left 03:28 vytas left 03:31 Fleurety left 03:33 _notbenh joined
ZoffixWin Someone mentioned during my talks there are these for 100s of languages and we should have one for Perl 6 too: rubykoans.com/ 03:33
03:33 clkao_ joined
diakopter ZoffixWin: TimToady and others spent many hundreds of hours on the rosettacode examples 03:33
03:33 cosarara_ joined
ZoffixWin diakopter, the guy said they were hard to navigate 03:33
sortiz .tell abraxxa I finally have a working oracle test environment!
yoleaux sortiz: I'll pass your message to abraxxa.
03:33 MadcapJake_ joined, orbus_ joined
ZoffixWin (not the same guy, but someone brought up they were not ideal_ 03:33
perlawhirl ZoffixWin: did you just volunteer yourself :D
03:33 Vitrifurr joined, AiHasBeenSolved joined
ZoffixWin nope :) 03:33
03:33 nadim joined
diakopter ZoffixWin: there are per-language lists 03:33
AiHasBeenSolved wiki.opencog.org/wikihome/index.php/Ghost
AiHasBeenSolved Ghost AI coder doing his bit to promote Perl6 and Perl AI. Bye
q
quit
diakopter rosettacode.org/wiki/Category:Perl_6
03:33 AiHasBeenSolved left, retupmoc_ joined, M-tadzik joined, vytas` joined
diakopter -_- lol 03:33
ZoffixWin *shrug* I'm unsure what the exact issue was. I suggested opening an Issue on the User experience repo so it could be addressed.
03:33 andrewalker_ joined, sQuEE` joined, roguelazer_ joined, daxim_ joined, sunnavy__ joined, jnap_ joined, kipd_ joined
saaki rubykoans is really geared for learning the language in an active fashion. rosettacode is about examples of common patterns. kind of different. 03:33
03:33 Spot___ joined, MadcapJake_ is now known as MadcapJake, roguelazer_ is now known as roguelazer, clkao_ is now known as clkao, jnap_ is now known as jnap, kipd_ is now known as kipd 03:34 Spot___ is now known as Spot__ 03:35 jack_rabbit joined
skids wonders if oeis.org would welcome Perl6 worked code or not. 03:35
03:35 ChanServ sets mode: +o diakopter, diakopter sets mode: -o diakopter
diakopter oops 03:35
03:35 Fleurety_ joined 03:36 musca joined
Xliff_ **SPLIT** !!! 03:37
03:43 ecojud joined 03:47 Ben_Goldberg left 03:50 Fleurety_ is now known as Fleurety
Xliff_ W00 h00! supernovus merged my PR 03:51
ecojud well done, I'm jealous 03:52
03:52 tardisx joined 03:55 skids left 03:58 pierrot left, pierrot joined 04:04 tardisx left, labster joined 04:06 tardisx joined 04:07 tardisx left 04:08 khw left 04:09 tardisx joined 04:12 tardisx left, ecojud_ joined 04:13 tardisx joined 04:15 tardisx left, ecojud left 04:20 tardisx joined 04:25 jamesalbert joined 04:29 AlexDaniel left
jamesalbert Hey, how would I extend a module's class from a different directory. Doing Preexisting::Module::NewClass doesn't allow Preexisting::Module to load NewClass dynamically with `require ::Preexisting::Module::("NewClass")` 04:29
and by "Doing Preexisting::Module::NewClass" I mean creating a class by that name, in another directory 04:30
04:33 RandalSchwartz left 04:38 quester joined 04:43 jamesalbert left 04:44 Cabanossi left 04:46 Cabanossi joined 04:51 molaf joined, tardisx left 05:14 tardisx joined 05:28 jack_rabbit left 05:35 tardisx left 05:36 ecojud_ left 05:39 jack_rabbit joined
Xliff_ jamesalbert: Read up on "augment". 05:39
perlawhirl he's gone... 05:40
i was gonna reply, then some work stuff came up, then when i tured back he had quit
Xliff_ Oh. LOL. I didn't even see the quit message.
perlawhirl Xliff: my modules is now on the modules directory. 05:41
pretty happy about it, am inspured to do start working on some other unfinished modules i have lying around 05:42
Xliff_ Yeah. I saw. 05:43
05:43 Xliff_ is now known as Xliff
perlawhirl yes, but did you see my cool mascot for wig {[: ) 05:44
it's wearing a wig!
Xliff *snerk*
You should make it so that appears next to the project name in the modules list. 05:45
modules.perl6.org/#q=wig
perlawhirl i will
Xliff whips up a quick SVG version
05:46 tardisx joined 05:47 znpy joined 05:56 jack_rabbit left
Xliff {[: )wiki.shadowplayinternational.org/te...20logo.svg 06:01
06:01 znpy left
Xliff Er. 06:01
wiki.shadowplayinternational.org/te...20logo.svg
It's an SVG so you can download and do whatever.
The second one has character. 06:02
perlawhirl: Logo goes in logotype/logo_32x32.png and the site should automatically pick it up. 06:03
(I am assuming logo then needs to be sized 32 x 32....) 06:04
perlawhirl, note error when trying to install from panda: 06:06
fetch stage failed for wig: Failed cloning git repository 'git://github.com/0rac1e/wig.git'
This is after "panda update"
perlawhirl dagnabbit! 06:13
the source-url is wrong :(
fixed, but it will be another hour before modules directory updates 06:15
in the meantime, you could just panda install github.com/0racle/p6-wig.git
06:17 rindolf joined 06:18 RabidGravy joined
perlawhirl Xliff, i remember now. when i first created the repo it was just called wig. i renamed it to p6-wig... becuase i dunno. forgot to update the meta 06:21
this is my first github repo, first perl module... training wheel are getting a workout
06:27 araujo left, araujo joined
Xliff perlawhirl, LOL! 06:35
06:36 domidumont joined, domidumont left
Xliff Hrm. 06:36
Using the URL doesn't work, either. 06:37
perlawhirl, tried github.com/0racle/p6-wig.git
and github.com/0racle/wig.git
06:37 domidumont joined
Xliff Now why did I read that name as nomadonut? 06:37
Xliff just got his eyes checked. 06:38
06:38 wamba joined 06:39 darutoko joined
Xliff wants an infix:<∞> operator. 06:40
It should do something NIFTY! 06:41
perlawhirl hmm, panda must look at the repo's META if you give it a url
i guess the only option until the moduless dir rebuilds is to git clone, cd pg-wig, panda install . 06:42
but just wait under ~30 mins :D
06:42 domidumont left 06:47 teatime joined 06:48 sue_ joined 06:52 domidumont joined 06:54 domidumont left 06:55 CIAvash joined 06:56 domidumont joined, nadim left, tardisx left
moritz ecosystem-api.p6c.org/projects.json now has wig 06:58
06:59 nadim joined
perlawhirl thanks moritz... though i made a booboo and messed up the source url. i've fixed it... now sitting idly waiting for the build to run again :D 06:59
not long now
moritz 8,28,48 * * * * bash update.sh > update.log 2>&1 07:01
that's the update cron job for ecosystem-api.p6c.org
moritz kicks it off manually
... done 07:03
RabidGravy boom
07:05 Actualeyes left, perlawhirl left 07:06 perlawhirl joined
perlawhirl moritz++ 07:09
07:11 perlawhirl left 07:13 zakharyas joined 07:14 perlawhirl joined
perlawhirl Xliff: it works now {[: ) 07:15
07:17 yeahnoob joined, perlawhirl left 07:18 yeahnoob left, perlawhirl joined 07:20 nadim left 07:26 sunnavy__ is now known as sunnavy 07:29 Actualeyes joined, fireartist joined 07:33 anaeem1 joined 07:34 anaeem1 left 07:44 abraxxa joined 07:45 perlawhirl left
sortiz abraxxa, ping 07:46
abraxxa good morning!
yoleaux 30 Mar 2016 23:52Z <sortiz> abraxxa: Changes that only touch DBDish/Oracle* can be pushed directly to master
03:26Z <sortiz> abraxxa: I finally have a working oracle test environment!
abraxxa great!
sortiz abraxxa, I have only 5 tests failing now.
abraxxa wow, great!
that means I've wasted lots of time yesterday 07:47
sortiz :( Why?
07:48 TEttinger left
sortiz The 'is rw' is now working for handlers, the problem was the base datatype. 07:48
07:49 dakkar joined 07:51 perlawhirl joined 07:52 sue_ left 07:59 leont joined
abraxxa i'll update my checkout 08:02
sortiz: have you not pushed it yet? 08:03
sortiz I haven't pushed yet.
abraxxa so what was required to get the Oracle docker container working on your Fedora?
08:03 telex left 08:04 telex joined
sortiz As commented yesterday, my problem was with the client library vs selinux, but googling found a solution. 08:04
Pushing my WIP... 08:05
dalek Iish: b9ea19b | (Salvador Ortiz)++ | lib/DBDish/Oracle (4 files):
Oracle: WIP, Updated some bindings
08:07
sortiz abraxxa, My remaining problem is that somehow the returned strings don't match the reference ones, but seems identical. :-( 08:12
08:12 sue joined
sortiz Aha! 08:13
The returned ones are marked, for example: Str+{NativeCall::ExplicitlyManagedString} <element> = "BEOM" 08:14
08:14 sjoshi joined 08:15 sjoshi left
abraxxa sortiz: yes 08:15
sortiz: don't know where this comes from
sortiz: you wrote the OO code a bit different than I did yesterday, for example I moved OCIEnvNlsCreate into OCIEnv.new 08:16
sortiz Yes, I don't want to change the overall approach, yet :) 08:17
08:20 azawawi joined
sortiz And there are lots of calls not converted yet to avoid the CArray wrapper. 08:20
azawawi hi
RabidGravy: ping
.seen RabidGravy
yoleaux I saw RabidGravy 07:03Z in #perl6: <RabidGravy> boom
sortiz azawawi, o/ 08:21
azawawi sortiz: hi
.tell RabidGravy github.com/azawawi/perl6-file-which/issues/4 # whence is now done as per your earlier request :)
yoleaux azawawi: I'll pass your message to RabidGravy.
08:22 |Tux| joined 08:23 perlawhirl left 08:29 perlawhirl joined
abraxxa sortiz: FYI the OCI naming convention is: postfixed with p => pointer, pp => pointer to pointer 08:34
08:34 azawawi left
abraxxa so you might want to rename $envhpp to $envhp now that you're using is rw 08:34
sortiz++ # that refactor lead to really nice code! 08:35
I wonder why the is rw didn't work for me yesterday, did the same thing you did
sortiz: you seem to have tab/space problems in your editor 08:36
08:36 nadim joined
sortiz For the 'is rw' see Native.pm6#L19 08:37
08:40 g4 joined
abraxxa that does the double pointer trick? 08:40
sortiz That plus 'is rw' 08:41
Xliff perlawhirl, panda is still failing on the git clone. 08:43
==> Fetching wig
fatal: remote error:
Repository not found.
fetch stage failed for wig: Failed cloning git repository 'git://github.com/0rac1e/wig.git'
sortiz And now we can convert all subs that take an OCIEnv as first parameter to methods in that class, for example.
abraxxa I defined it as class OCIEnv is repr('CPointer') is export {}; and is rw on OCIEnvNlsCreate and it didn't work
perlawhirl try umm --force or --reinstall or something
maybe sortiz knows, but theres a dist folder, in ~/.perl6/dist on linux 08:44
that has a file in it
that still says 'wig.git' in it, it hasn't updated
Xliff "git clone github.com/0racle/wig.git; panda install ." <-- worked
perlawhirl that's weird, because the repo is now called p6-git... 08:45
maybe github remembers i renamed it and still lets you clone it via the old name ??
sortiz abraxxa, then I don't know what happened, but now is working.
Xliff perlawhirl, did you see the SVG?
abraxxa sortiz: please check your tab/space editor settings 08:46
sortiz Xliff, try 'panda update'
Xliff Oh! I see a logo. Was it one of mine or did you refont from the SVG?
perlawhirl i did, but i didn't have an editer to crop it to 32x32
nah i just knocked seomthing up in one of those icon maker sites
sortiz abraxxa, sure, don't worry about that, we have some more serious problems ;-) 08:47
08:47 perlawhirl2 joined
abraxxa sortiz: it makes reviewing commits harder when lines only have whitespace changes 08:48
Xliff *gasp* 08:49
perlawhirl has a client that ignores CTCP VERSION!!!
Philistine!
sortiz abraxxa, Yes, i know, promise fix that.
abraxxa sortiz: thanks!
Xliff Now how can I know what OS you have so that I can taunt you mercilessly? ^_^
perlawhirl2 Xliff: i just got off the train. Switched from irssi on my laptop to some app on my phone 08:50
Xliff perlawhirl, on a serious note. www.google.com/url?sa=t&rct=j&...HKKMi7uwDQ
arnsholt I used to have a client that replied to CTCP VERSION with a claim that it was running on an ENIAC =D
Xliff arnsholt++
perlawhirl2 Any recommendation for irc client on Android
Xliff perlawhirl2, Inkscape is multiplatform SVG editor that can export to PNG. 08:51
perlawhirl2 Something lite, i don't need a million protocols
arnsholt I think I had something weird for CTCP TIME as well. Can't remember what, though
Xliff perlawhirl2, I used to have one but that was WAY back in the day.,
play.google.com/store/apps/details...&hl=en
08:52 perlawhirl left
Xliff That's the default option. 08:52
Works better with a tablet. IRC demands width-based real estate.
perlawhirl2, heading to work or heading home?! 08:53
perlawhirl2 Heading home, albeit a bit late. 19:53 here 08:54
nadim Morning! 08:57
RabidGravy erp 09:02
yoleaux 08:21Z <azawawi> RabidGravy: github.com/azawawi/perl6-file-which/issues/4 # whence is now done as per your earlier request :)
RabidGravy azawawi :)
.tell azawawi RARRRR! 09:03
yoleaux RabidGravy: I'll pass your message to azawawi.
09:03 ab6tract joined
nine perlawhirl2: I just use vxconnectbot to ssh to my server where irssi is running 09:05
Xliff Rebuilding rakudo to check my Match infix:<eqv> implementation. 09:07
To my knowledge there are no formal test cases for Match... or are there? 09:08
09:10 leont left, donaldh joined 09:19 CIAvash left 09:23 xinming_ is now known as xinming
Xliff *sigh* 09:23
Now I've found a failed test case.... just when I was getting ready to push the PR.
ARGH!
I need sleep. I'll figure this out tomorrow.
'night (or 'morn for some o ya) #perl6
09:24 Xliff is now known as Xliff_z_OUT 09:25 perlawhirl joined 09:27 quester left 09:28 perlawhirl2 left 09:31 labster left 09:32 espadrine joined 09:36 sue left
ab6tract lizmat: updated the PR again :) 09:42
09:55 perlawhirl left
dalek Iish: ce20a74 | (Salvador Ortiz)++ | lib/DB (3 files):
Oracle: Passing the standard test suite

Not yet "in an ideal world", TODO #50
09:55
10:00 kid51 joined 10:04 perlawhirl joined, cibs_ joined
nadim can someone make a review on Text::Template that is in the ecosystem? it's pretty short. 10:09
10:11 cibs joined 10:14 cibs_ left
dalek Iish: 91594d7 | (Alexander Hartmaier)++ | lib/DBIish/CommonTesting.pm6:
fix typo
10:17
ab6tract jnthn: not sure if you have any objections to the PR? 10:38
lizmat ab6tract: right in the middle of something else, will try in a mo 10:52
jnthn ab6tract: Didn't have time to look yet, sorry :(
ab6tract lizmat, jnthn: no problem! sorry, just a little excited about my first NQP patch (however trivial :) ) 10:53
Ulti how much of a big deal would it to be to get some helper methods added to Duration to convert to minutes/hours/days 11:03
11:04 dvinciguerra__ joined 11:07 M-tadzik is now known as tadzik
ab6tract Ulti: sounds useful to me... a bit surprised that functionality is not already there 11:12
11:13 woodruffw joined, woodruffw left 11:14 woodruffw joined
lizmat Ulti: isn't that what polymod is ? 11:20
11:20 nemo left 11:21 nemo joined
moritz well, days is dependent on which day you're talking about 11:21
11:21 fireartist left, Some-body_ joined
moritz even minutes (there are minutes, hours, days with leap seconds) 11:21
timotimo polymod isn't enough when you have to deal with leap seconds/days/...
11:21 nemo is now known as Guest27595 11:22 cosarara_ left, pierrot left, DarthGandalf left, Upasaka_ joined, Some-body_ is now known as DarthGandalf, cosarara joined 11:23 ZoffixWin joined, ZoffixWin left, ZoffixWin joined, wtw joined, jervo joined, Timbus joined, skarn joined, nadim left
lizmat myabe I'm understanding Duration wrong then ? 11:23
11:24 Guest27595 is now known as nemo
ab6tract m: sub f() { True }; say f ?? 'yay!' !! 'nay!'; 11:28
camelia rakudo-moar ae3a79: OUTPUT«5===SORRY!5=== Error while compiling /tmp/Nvl4jDwA5l␤Expected a term, but found either infix ?? or redundant prefix ?␤ (to suppress this message, please use a space like ? ?)␤at /tmp/Nvl4jDwA5l:1␤------> 3sub f() { True }; say f ??7⏏5 'ya…»
moritz or f() 11:29
ab6tract I would have hoped that specifying no args would provide the same behavior as an empty prototype in p5
moritz: aye, that works. but it feels weird that p6 is less capable than p5 here
moritz your hopes have been squashed
11:30 tony-o joined
ab6tract ok. well usually i like my tears to flow with an explanation 11:30
moritz m: sub term:f { True }; say f ?? 'yay' !! 'nay'
camelia rakudo-moar ae3a79: OUTPUT«5===SORRY!5=== Error while compiling /tmp/u9x4khtkwZ␤Expected a term, but found either infix ?? or redundant prefix ?␤ (to suppress this message, please use a space like ? ?)␤at /tmp/u9x4khtkwZ:1␤------> 3sub term:f { True }; say f ??7⏏5 …»
moritz m: sub term:<f> { True }; say f ?? 'yay' !! 'nay'
camelia rakudo-moar ae3a79: OUTPUT«yay␤»
moritz ab6tract: we allow post-declaration of subroutines 11:31
ab6tract: so we can't have the declaration influence the parsing
also, things get tricky once you're in multi territory
ab6tract gotcha. i had forgotten about the term category for ops 11:32
thanks for clarifying
moritz you're welcome
ZoffixWin If anyone's interested, the recording of my last night's "Wow, Perl 6!" talk has been posted and I've written up answers to some of the questions I didn't answer during the talk: blogs.perl.org/users/zoffix_znet/20...tions.html 11:36
jnthn ZoffixWin: "Can you check whether or not a value fits the subset?" - just smartmatch against the subset type 11:47
ZoffixWin :o
timotimo you, too, can be smart 11:48
jnthn m: subset Even where * %% 2; say 3 ~~ Even; say 42 ~~ Even
camelia rakudo-moar ae3a79: OUTPUT«False␤True␤»
ZoffixWin jnthn++
Ulti moritz as a unit though a duration in days would be 86400 seconds right? 11:49
like after a day you can't really talk about a unit month or year 11:50
11:50 woodruffw left
Ulti but a lot of epoch times are numbers of days since X as a real value with a decimal day 11:50
specifically MS Excel dates are :) which is what I'm playing with
11:54 kaare_ joined
timotimo ZoffixWin: sadly the noise cancellation or something else has done bad things with the audience questions in your recording :( 11:55
moritz Ulti: not all days are 86400 seconds
Ulti: which is why day isn't really a unit that can be carelessly converted from/to seconds 11:56
ZoffixWin I hear next time there'll be a better setup.
moritz has a dream that one, all our days are created equal 11:57
moritz has a dream
Ulti moritz which days arent? leap second is a specific date not a day and if you really care you can average the leap second over the calculation from knowing the full Duration 12:01
moritz Ulti: days with leap seconds, days on which DST is switched on or off
Ulti sure those are specific dates not a day
moritz days on which a country changes what timezone it considers itself to be in
Ulti you dont change the duration of a second in phyiscs because there is a leap nanosecond 12:02
moritz: those are all specific dates not the length of any day 12:03
moritz Ulti: well, if you make the distiction between a date and a day, you can't add durations based on days to datetimes anymore
Ulti even a year you can define fairly well
moritz well, a year is rougly made of 365.25 days; but if you use that definition in a datetime library, you make everybody unhappy 12:04
Ulti they arent based on days at the moment it only holds seconds, I'm suggesting if you want to know how many days those seconds add up to for no particular date you can avoid magic numbers
delta_days in Perl 5 DateTime for example
though I think that does do the full calculation
but DateTime::Duration in Perl 5 also has delta_days which is more like what I'm talking about 12:05
12:05 kid51 left
timotimo ZoffixWin: your "custom terms and operators" slide has an image, so i can't copy-paste all those values in order to try writing that program :| 12:05
Ulti DateTime delta_days tells you the actual number of days between dates but the Duration just treats a day as a unit
ZoffixWin timotimo, go nuts :) unicodeemoticons.com/
timotimo oh, fantastic 12:06
but what the hell should these actually do?
you can just turn each of them into a term and an infix alternatingly
ZoffixWin Cure cancer? That'd look good in history books :)
timotimo i was hoping i could get a way with short one-liners for the implementations 12:07
.u ❨ 12:09
yoleaux U+2768 MEDIUM LEFT PARENTHESIS ORNAMENT [Ps] (❨)
timotimo fantastic
12:10 andreoss joined
andreoss perl6-j dies with Unhandled exception: java.lang.StackOverflowError 12:13
in repl
12:14 xinming_ joined
ab6tract andreoss: not sure if the jvm backend has ever gotten on its legs after the GLR.. 12:16
timotimo only partially 12:17
could very well be that the REPL is just "too much" for it right now
12:18 xinming left 12:22 xinming joined
andreoss will be rakudo-jvm released? or it's not a priority 12:23
12:24 xinming_ left
timotimo if i understand correctly, it's currently a single of our devs who's investing time into that 12:24
12:26 Sgeo__ left
ab6tract andreoss: indeed, the moar backend gets the vast majority of work at the moment. i believe that the reasoning is more or less "one full featured and fast backends is better than multiple slow, ever so slightly broken ones" 12:28
timotimo well, a lot of the work really is cross-backend, or backend-agnostic 12:30
12:41 M-matthew left, nemo left 12:43 dvinciguerra__ left
timotimo ZoffixWin: about aborting/killing promises; nope, not possible. if you need that kind of thing, you'll have to use processes, or you'll have to build the code so that it exposes some kind of flag "should i continue working?" 12:43
12:44 nemo joined, nemo is now known as Guest52450 12:46 Guest52450 is now known as nemo 12:47 M-matthew joined 12:49 sufrostico joined 12:51 rindolf left 12:53 xinming left
timotimo oh, i just now remember, ZoffixWin, yo ucould have also shown off the ANTLR4 thingie for turning ANTLR grammars into a perl6 grammar 12:54
ZoffixWin k
timotimo in general, "kill a thread" is something you're not really getting anywhere
donaldh ab6tract: the JVM backend did just fine after the GLR 12:55
12:55 rindolf joined
donaldh ab6tract: there were some issues on the JVM backend with the precompilation changes. One those manifested as a stack overflow error. 12:55
andreoss: ^ 12:56
13:02 sufrostico left 13:13 xinming joined 13:17 sufrostico joined 13:21 cdg joined
timotimo ZoffixWin: your security tip thingie about RESTRICTED.setting is moot since you're allowed to "use nqp" and use anything moar offers 13:21
13:21 xinming left
donaldh The jvm repl blows up here: github.com/rakudo/rakudo/blob/7e8c...PL.pm#L243 13:31
hoelzro hmm...maybe we should disable REPL.pm on the JVM for now 13:32
donaldh hoelzro: Why is it: has $.compiler 13:33