»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_log/perl6 | UTF-8 is our friend! 🦋
Set by Zoffix on 25 July 2018.
timotimo but this also means that if i have the hash from the filename, i can't differentiate between released versions of a module and unreleased local modifications, right? since the file contents don't enter into it? 00:00
though i guess for many installed ones i get :auth<cpan:...>
whereas the auth for non-released ones tends to be :auth<github:...> 00:01
nadim_: i don't know what that is
nadim_ pack everything in an executable archive, perl 6, scripts, modules, data, ...
ugexe to calculate the sha1 you do nqp::sha1('lib/Relative/Path/Name.pm6' ~ $foo), where $foo is $.id from github.com/rakudo/rakudo/blob/89f8...m6#L56-L65 00:02
nadim_ metacpan.org/pod/distribution/PAR/...torial.pod
00:02 mowcat joined
timotimo ugexe: hum. i'd have liked there to be a separator between the path name and the $.id, or perhaps sha1 them recursively 00:03
i.e. instead of sha1($foo ~ $bar) we'd do sha1(sha1($foo) ~ $bar)
ugexe Goofus installs his unreleased local modifications without bumping the version. Galant uses -I. or bumps the version.
yeah that'd be great, except it wasn't done like that to begin with. so you'd have o write something to upgrade all existing sha1 using stuff 00:04
timotimo right, that problem ...
00:05 moony left
ugexe also its already sha1($foo ~ sha1($bar)) 00:06
timotimo oh, it is?
sorry about that, then
ugexe sha1("lib/Foo.pm6" ~ sha1('Foo:ver<...>:auth<...>:api<...>')) 00:07
timotimo m) 00:08
i'm too tired
i should go to bed
00:20 farcas82 joined
nadim_ I have the weirdes peoblem, one that has been bugging me all afternoon, I have a script that uses DDT, and it shows some dd output, but there is no dd call! 00:21
I did have that yesterday but removed it.
I got tired so I removed the whole DDT module ... and the script still works! where does it find the librry?!!! 00:22
ugexe: any idea? I have removed all .precompiled I know of
ugexe what do you mean "removed the whole DDT module". be more specific
timotimo you can use "RAKUDO_MODULE_DEBUG" in the environment to find out where it's looking
nadim_ ugexe: i move DDT in another file so the script would fail, and it doesn't 00:23
ugexe you have to be more specific still 00:24
what file exactly did you move
nadim_ lib/Data/Dump/Tree.pm to somewhere else 00:25
ugexe what is the exact command you are invoking your script that uses lib/Data/Dump/Tree.pm ?
nadim_ perl6 -Ilib,t example/script.pl 00:26
00:27 epony joined
nadim_ itseems to me that thescript load from the cache (/home/nadim/.perl6/precom) even the module has changed, or is not there at all 00:28
ugexe the precomp from that file will be in lib/.precomp
otherwise it may be loading some old installed version, which would be apparent with `zef locate Data::Dump::Tree` 00:29
nadim_ there is no lib/.precomp 00:30
ugexe there will be if you run the command you pasted
timotimo oh i didn't know you can put a , there 00:31
bedtime! 00:33
ugexe otherwise follow timotimo advice and run your script with RAKUDO_MODULE_DEBUG, then grep through the wall of text to find the file being loaded
00:33 epony left
nadim_ ugexe: you're righ, there's a shebang inthe scrip I didn't always run it the same way 00:34
ugexe timotimo: put your repos in `C:\Users\Last, First` 00:35
let the space and the comma duke it out 00:36
00:39 epony joined 00:44 Manifest0 left 00:50 Manifest0 joined 00:55 Manifest0 left 01:00 MasterDuke joined, MasterDuke left, MasterDuke joined, Manifest0 joined 01:06 sena_kun left
Kaiepi um... design.perl6.org/S05.html#Predefined_Subrules 01:08
01:11 farcas82 left
Xliff Is sourcebaby still around? 01:12
01:26 Manifest0 left 01:32 Manifest0 joined 01:47 hythm joined
hythm .tell moritz any suggestion regarding this Grammar that works intermittently github.com/rakudo/rakudo/issues/2933 01:52
yoleaux hythm: I'll pass your message to moritz.
MasterDuke hythm: you might be able to use the code coverage log env variable and compare what different code gets run between the two different outcomes, could help with debugging 01:56
01:57 molaf left
ugexe .tell timotimo run this from perl6-all-modules/ 01:57
yoleaux ugexe: I'll pass your message to timotimo.
ugexe -- gist.github.com/ugexe/e00ade523d15...311d07e420 01:58
doh
.tell timotimo run this from perl6-all-modules/ -- gist.github.com/ugexe/e00ade523d15...311d07e420
yoleaux ugexe: I'll pass your message to timotimo.
hythm Thanks MasterDuke , checking logs for related debug variables 01:59
docs* 02:00
02:01 avalenn left
MasterDuke hythm: don't think the info is in the docs. but `moar --help` has some info 02:02
hythm noted
02:03 avalenn joined
MasterDuke MVM_COVERAGE_LOG=filename is the simple method 02:03
02:04 Manifest0 left, mowcat left
MasterDuke using MVM_COVERAGE_CONTROL and then nqp::coveragecontrol in your code allows for a little more control 02:05
02:06 Manifest0 joined
hythm thanks, trying that 02:06
02:10 molaf joined 02:12 Manifest0 left 02:14 avalenn left 02:18 Manifest0 joined 02:25 Manifest0 left 02:30 Manifest0 joined 02:31 Xliff left 02:35 Manifest0 left 02:41 Manifest0 joined 02:48 vrurg left 02:52 ejd left 02:53 Manifest0 left 02:58 adu joined 02:59 Manifest0 joined 03:04 Manifest0 left 03:12 Manifest0 joined 03:18 Manifest0 left 03:23 Manifest0 joined 03:26 araraloren joined 03:28 Manifest0 left 03:33 Manifest0 joined 03:34 kurahaupo_ joined 03:40 Manifest0 left 03:41 kaare__ left, kaare__ joined 03:42 Kaiepi left 03:46 Kaiepi joined, Manifest0 joined 03:53 Manifest0 left 03:58 Kaiepi left 04:00 Manifest0 joined 04:04 Kaiepi joined 04:08 Manifest0 left 04:14 hythm left 04:21 Manifest0 joined 04:27 mtg joined
Kaiepi i'm going to start maintaining perl 6 on openbsd ports 04:33
freebsd too potentially
04:38 Woodi joined 04:42 curan joined 04:57 Manifest0 left 04:58 Kaiepi left 04:59 Kaiepi joined 05:01 Manifest0 joined
bartolin Kaiepi++ 05:05
05:11 Kaiepi left 05:12 skids left, Manifest0 left 05:15 mtg left 05:17 Kaiepi joined 05:18 Manifest0 joined, kurahaupo_ is now known as kurahaupo 05:22 sauvin joined 05:23 Kaiepi left, Kaiepi joined 05:26 molaf left 05:36 Manifest0 left 05:41 Manifest0 joined 05:49 Manifest0 left 05:57 Manifest0 joined 06:00 aborazmeh joined, aborazmeh left, aborazmeh joined 06:10 Manifest0 left 06:11 Kaiepi left 06:15 Manifest0 joined 06:19 ChoppedBacon left 06:20 Manifest0 left 06:21 Kaiepi joined 06:22 domidumont joined 06:24 ChoppedBacon joined, Manifest0 joined 06:54 Manifest0 left 06:57 aborazmeh left 06:59 Manifest0 joined 07:03 Manifest0 left 07:07 patrickb joined 07:08 Manifest0 joined 07:13 Manifest0 left
holyghost Kaiepi, I want to start the perl6 pkg for netbsd, I've not come around to doing anything due to work but if you can send me a tarball with pkgsrc Makefiles, patches etc, I might put it in 07:14
I'm on their mailing list, openbsd ports should be the same as netbsd AFAIK, so it'd be done in a jiffy 07:16
Kaiepi aight
07:18 Manifest0 joined
holyghost Problem is, I don't have a x86_64, I only have i386 netbsd 07:18
patrickb o/
holyghost 'lo patrickb
Then I have a amd64 openbsd 5.7 on a macbook 07:19
and a i386 freebsd on a dell
I might only be interested in the netbsd though, as I don't have the time or the courage to develop much just for maintenance 07:20
07:20 kurahaupo left, kurahaupo_ joined
holyghost I don't have the money for new machines, 64 bit I mean 07:21
I sold my SGI O2 and Indy, if I can find another machine I'd be interested to port perl6 to IRIX (or SGI's linux e.g. debian) 07:22
IRIX is a freebsd or 4.4 bsd derived
IRIX has a gnu c compiler (gcc) 07:24
I tried to port perl6 to MacOS X PPC but that didn't go well
PPC = IBM's powerpc 07:26
used in a GameCube, Nintendo Wii, and WiiU
Then later on I'll try to load an OS on my G5 Mac and try again to work on it for the perlmongers 07:27
It's a small niche though
since G5 is 64 bit, perl6 might work better although it might need headers from something else 07:28
I know you just walk into another OS with perl6 but people should keep it portable C so you can load it on other machines 07:29
The gcc master holds the key to it I presume 07:30
*I know you don't just walk 07:32
Kaiepi m: say $*PERL.compiler.release 07:37
camelia
07:46 Manifest0 left 07:47 veesh left 07:48 veesh joined
holyghost In some worlds, e.g. the openssh, there might exist something like a perl6-portable version for porting 07:49
This is far-fetched and extremely difficult if we don't have the time or folks or ports to brew that 07:50
07:51 Manifest0 joined, arkaitz joined 07:52 dolmen joined 07:57 Manifest0 left, zakharyas joined 07:58 arkaitz left, Kaiepi left 08:00 noisegul joined 08:01 ravenousmoose_ joined, Kaiepi joined 08:02 ravenousmoose left 08:03 Manifest0 joined 08:05 aborazmeh joined, aborazmeh left, aborazmeh joined 08:08 sena_kun joined 08:11 dolmen left 08:13 dolmen joined
timotimo greppable6: trusts 08:14
yoleaux 01:57Z <ugexe> timotimo: run this from perl6-all-modules/
01:58Z <ugexe> timotimo: run this from perl6-all-modules/ -- gist.github.com/ugexe/e00ade523d15...311d07e420
greppable6 timotimo, 47 lines, 20 modules: gist.github.com/2dacdbdc3e636bc8ea...d118792f7e
timotimo ugexe: gasp, that's so cool 08:16
ooh, perl6/doc is already in what greppable uses so you can already "doccable", though of course it links to the source, not the rendered result 08:18
08:21 ravenousmoose_ left 08:27 arkaitz joined, Kaiepi left 08:28 Kaiepi joined, [particle]1 joined 08:29 [particle] left 08:35 Manifest0 left 08:40 Manifest0 joined 08:43 farcas82 joined 08:53 Manifest0 left 08:56 ravenousmoose_ joined 08:58 Manifest0 joined 09:01 zakharyas left 09:02 Manifest0 left 09:07 natrys joined, Manifest0 joined 09:11 rindolf joined 09:13 adu left
nadim_ morning all 09:16
noisegul nadim_: good morning
holyghost gm
nadim_ how do I pass a named arg to a sub conditionaly? eg: some_sub(:named_arg. :other_named_arg unless $some_bool, :third_named_arg) ; 09:18
moritz you need to build a hash and interpolate it with | 09:19
yoleaux 01:52Z <hythm> moritz: any suggestion regarding this Grammar that works intermittently github.com/rakudo/rakudo/issues/2933
nadim_ moritz: good morning, so I can't do something like this nopaste.linux-dev.org/?1216576 09:20
not even with putting parenthesis around? 09:21
jnthn m: sub foo(:$x = 42) { say $x }; foo(|%(x => 1 if 1)) 09:22
camelia 1
jnthn m: sub foo(:$x = 42) { say $x }; foo(|%(x => 1 if 0))
camelia 42
09:33 Manifest0 left 09:38 Manifest0 joined
nadim_ timotimo: latest with github.com/nkh/P6-Data-Dump-Tree#c...etup-roles 09:40
timotimo: it was a good idea, it also gives me a better "framework" to try rendering idea 09:42
09:54 Manifest0 left, domidumont left
Kaiepi m: role A { has Int $.a; submethod BUILD(:$!a = 1) { } }; class B does A { submethod BUILD(:$!a = 2) { } }; my B $b .= new; $b.a 09:58
camelia ( no output )
Kaiepi m: role A { has Int $.a; submethod BUILD(:$!a = 1) { } }; class B does A { submethod BUILD(:$!a = 2) { } }; my B $b .= new; say $b.a
camelia 2
Kaiepi i thought that didn't work?
or maybe i'm mixing up roles with classes
specifically :$!foo = 'bar'
09:59 Manifest0 joined 10:11 noisegul left 10:16 Kaiepi left 10:23 Manifest0 left 10:24 aborazmeh left, arkaitz left 10:29 Manifest0 joined 10:30 lgandras joined 10:34 dolmen left 10:40 Kaiepi joined 10:41 curan left 10:43 noisegul joined 10:47 noisegul left
holyghost I remember I descended into Ultima 6 dragon dungeons with "Heirate Mich" from Rammstein on my headphones 10:48
I'll make some good dungeons for my game systems 10:49
dungeon == Master Tower in Hollandaise
of a castle I presume 10:50
Kaiepi i remember grinding for weeks in ffii for the best magic 10:51
i learned nothing
besides that i'm a masochist
holyghost "Heirate mich" is a song from the movie "Lost Highhway", film album 10:52
libffi ?
Kaiepi no, final fantasy 2
it's by far the hardest in the series
holyghost I have it on PSP 10:53
It's a very cool game
Kaiepi funny you do the same thing i do though, i associate certain songs with what i was doing at the time
holyghost sure
You have to dungeoneer with music, I guess
Kaiepi the later versions aren't as hard
i played the famicom version
holyghost I gave my PS3 FF13 to my girls 10:54
Kaiepi luckily there's a bug where if you select a spell and cancel 100 times in a row in a battle you instantly level up that spell
holyghost I don;t know the SNES/Super Nintendo/Famicom versiob
Kaiepi it's mostly the same, just missing some extra items
holyghost sure, NES and SNES have those easter eggs
Kaiepi and bug fixes
holyghost I am preparing my combat systems with one of FF2 10:55
And, for you elder gamers, Quest for Glory (also Hero's Quest) 10:56
I need to sub/super system modularize it
There's also Darkside Chronicles on Windows Store, it's the same system more or less 10:57
It's a free game from M$
or it's named Heroes of Larkwood, I don't remember 11:00
I write games, I don't play them
As a solo designer/gfx/programmer you cannot just work out current games, e.g. I write them mostly in 2D 11:02
Kaiepi the only game i've written is a trivia plugin for a chat website 11:03
holyghost I hope Xliff can help out with the system, he's sleeping now, so don't wake the dragon :-)
Kaiepi i find writing games tedious
holyghost tedious ?
Kaiepi well, a lot of work
holyghost I know my CSS3 games too
Kaiepi the trivia plugin wasn't but...
holyghost I see 11:04
It's indeed sped up with programming them
You just don't program them fully featured without hard work
You miss stuff, although there's perl sotware for easily programming game scenarios 11:05
Then again, I almost always start from scratch with a system I found in 2003
*I found out myself in 2002 11:06
It's a simple program but you can do lots of stuff fast
Anyway, start dungeoneering with Rock music or something on your headphones of your CDROM/computer 11:11
You have to go into Dark Wave for that, for example :-) 11:12
"Save Me" from Suicide Commando (a Belgian artist) might do the trick
I also like "Feuer Frei" from Rammstein, the dungeoneering music is legio these days 11:20
Blood, violence and Perl :-) 11:23
I'm going to do some good dungeoneering in PaganVisions2 (it does _not_ compile right now), with FF and QFG systems, I hope to incorporate several game subsystems and mix them for game dev 11:28
11:28 dolmen joined
holyghost is a bit drunk on fridays 11:28
If you look at github.com/theholyghost2/PaganVisions2, there's a MILESTONES.txt in it 11:29
I hope to extend it to 3D sometime
The game subsystems need to become driven by the AI subsystems 11:30
11:30 noisegul joined
holyghost There's also a TODO.txt 11:30
Most simple game systems cover 2D as well as 3D, e.g. by collision detection 11:31
It's lots of work but it's not difficult if you simplify several
It held onto Linux and Cairo(1-2) for dev with Xliff of that gfx subsystem 11:32
he's offline now, maybe next week can debug the gfx subsystems 11:33
*we can
Then I'm going to use rpgrevolution.com and opengameart.org gfx 11:34
and my own
11:34 noisegul left
holyghost It needs things like a priest/wizard spellbook, widget tree, you design game widgets in a graphical way such as smalltalk morphs 11:34
not just rectangled widgets such as buttons 11:35
It's better IMO to use a graphics system such as Cairo or SDL2 to develop widgets
Then they have lots more functionality 11:36
11:36 Kaiepi left 11:37 arkaitz joined, Kaiepi joined
holyghost Anyway, these things need to be sorted out before developing games 11:38
It's more or less just a heavyweight game kit
It should depend on GL and extentions later on 11:39
*OpenGL
11:39 lgandras_ joined, domidumont joined
holyghost I hope everyone likes to make games of it, as I'm going to cover the broad range of systems 11:40
holyghost is a bit too drunk on fridays, I'll explain later on
11:42 lgandras left 11:43 lgandras_ left
lizmat
.oO( and there was me hoping this would only happen on a good friday )
 11:44
holyghost neither :-)
It's good friday, lizmat 11:45
although for me
11:45 Kaiepi left
holyghost Anyway, I'm going to put the word dungeon into Perl6, if I may :-) 11:46
*say so
using nethack, ncurses, SDL2 and Cairo 11:47
angband, thangorodrim, zangband, moria, rogue and angmar 11:48
11:49 Guest6771 left
holyghost *lol* 11:49
11:49 success joined 11:50 success is now known as Guest89556
holyghost I'm a bit afk, I'll explain some more dungeon/gaming software later on 11:50
11:51 Kaiepi joined
nadim_ holyghost: I am sure you can put all this in a nice blog and just past a link. 11:51
holyghost I had a blog once, it's on thediaryofelvishhealer.blogspot.com 11:53
I speak through my code, I'll make some nice docs as you say, for it
11:54 lgandras joined
holyghost I hope to incorporate all major game systems in my PaganVisions2, with explanation in test dir and pod6 11:55
the tests will be somewhat heavyweight by all means
11:55 lgandras_ joined
holyghost Larry is probably going to call me onto making extra modules for PaganVisions2 11:56
PaganVisions2 is monolithic now 11:57
although it will be a modular system, meaning it has game subsystems which are worked out on their own, compile on their own and test on their own
People can always contact me for further development 11:58
11:58 lgandras left
holyghost As I said before, I want Xliff to debug the Cairo backend then move on with the SDL2 backend and then code game subsystems 11:59
11:59 Kaiepi left
holyghost ok, I'm going to try to shut up now 12:00
12:00 Kaiepi joined
holyghost Kaiepi, if you have a tarball of your openbsd port, do tell me 12:00
then I'll make a netbsd version
Kaiepi i haven't started it yet 12:01
holyghost If I find the time
ok
Kaiepi i'll get started on it once i finish working on this small project
holyghost Kaiepi : in debian there's modules as packages
so p6-JSON-Fast for example 12:02
I hope o/n/bsd can figure that out for themselves with zef
It's enormous work 12:03
*an enormous
ok, follow me up on it then 12:04
most p6/python packages become old really soon
12:04 kurahaupo joined
holyghost Then they delete them from the ports tree 12:04
I guess, we should just start out with perl6 on its own 12:05
12:07 kurahaupo_ left 12:09 molaf joined
nadim_ a perl 5 question but I'll need to know for Perl 6 too, when setting the record separator to "", is it possible to know what was read in the recod separator, ie: how many lines? 12:12
12:13 lgandras_ left
timotimo nadim_: i'm not sure i understand what you mean; does setting the record separator to "" give you one byte per record or something? 12:15
holyghost Kaiepi : do you know already if a perl5 ./Configure would work in the distro of a ports/pkgsrc of perl6 ? 12:17
nadim_ timotimo: given a RS = "", $_ will contain the paragraph but I want toknow how much of thetext was eaten by the record separator
timotimo: not sure you've seen it but you :does[MyCustomSetup] is on github 12:18
timotimo i saw it, but i didn't look at it in detail
i don't think perl6 supports a record separator of ""
tyil the link for " 12:19
holyghost Kaiepi : that's one of the troubles I ran into, by compiling perl5 for a perl6 package
tyil compiler only installation" is still bork on rakudo.org/
(rakudo.org/files/rakudo is what it points it)
12:19 cpan-p6 left
holyghost Kaiepi : you can of course make a predefined pkgsrc 12:20
12:20 cpan-p6 joined, cpan-p6 left, cpan-p6 joined 12:25 ravenousmoose joined 12:27 ravenousmoose_ left, robinsmidsrod left 12:28 BinGOs left
Kaiepi holyghost, i don't know if it would work 12:28
holyghost I understand
Kaiepi not that i doubt it, i just don't know
i have no experience working with ports
12:28 BinGOs joined
holyghost maybe, it's shell code in a Makefile config 12:29
12:29 robinsmidsrod joined
holyghost We would need it as you understand, for a clean perl6 installation 12:29
12:30 epony left
holyghost I'm also looking in to that later on 12:30
*into
12:37 Kaiepi left 12:38 Kaypie joined 12:39 arkaitz left, Kaypie left 12:40 Kaypie joined 12:45 gfldex left 12:47 gfldex joined 12:50 kurahaupo left, kurahaupo joined
holyghost is a bit afk, I'm going to watch LOTR - "The fellowship of the ring" 12:53
12:59 molaf left 13:03 vrurg joined
holyghost hi, vrurg 13:13
Kaiepi is doing openbsd pkgsrc of perl6, I'll try to do the netbsd one 13:14
vrurg Hi
Hope it doesn't have that limitations of OpenBSD. 13:15
holyghost I wouldn't know, it's my first package for netbsd 13:17
we need to figure out the perl5 ./Configure
Kaypie m: sub to-id(Str $str) { $str.samemark(' ').lc.subst(/<-[a..z0..9\-]>+/, '', :g) }; $/ = 'POAoaijodPJfiAPdfa' ~~ /$<command>=.+/; say do ~$<command> R[//] $<command> $[//] Nil ==> to-id()
camelia 5===SORRY!5=== Error while compiling <tmp>
Unsupported use of $/ variable; in Perl 6 please use the filehandle's .nl-in attribute
at <tmp>:1
------> 3.subst(/<-[a..z0..9\-]>+/, '', :g) }; $/7⏏5 = 'POAoaijodPJfiAPdfa' ~~ /$<command>=.
Kaypie m: sub to-id(Str $str) { $str.samemark(' ').lc.subst(/<-[a..z0..9\-]>+/, '', :g) }; my $/ = 'POAoaijodPJfiAPdfa' ~~ /$<command>=.+/; say do ~$<command> R[//] $<command> $[//] Nil ==> to-id()
camelia 5===SORRY!5=== Error while compiling <tmp>
Unsupported use of $/ variable; in Perl 6 please use the filehandle's .nl-in attribute
at <tmp>:1
------> 3bst(/<-[a..z0..9\-]>+/, '', :g) }; my $/7⏏5 = 'POAoaijodPJfiAPdfa' ~~ /$<command>=.
Kaypie m: sub to-id(Str $str) { $str.samemark(' ').lc.subst(/<-[a..z0..9\-]>+/, '', :g) }; 'POAoaijodPJfiAPdfa' ~~ /$<command>=.+/; say do ~$<command> R[//] $<command> $[//] Nil ==> to-id() 13:18
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 3.+/; say do ~$<command> R[//] $<command>7⏏5 $[//] Nil ==> to-id()
expecting any of:
infix
infix stopper
postfix
…
Kaypie m: sub to-id(Str $str) { $str.samemark(' ').lc.subst(/<-[a..z0..9\-]>+/, '', :g) }; 'POAoaijodPJfiAPdfa' ~~ /$<command>=.+/; say do ~$<command> R[//] $<command> R[//] Nil ==> to-id()
camelia Type check failed in binding to parameter '$str'; expected Str but got Match (Match.new(hash => Map.new(...)
in sub to-id at <tmp> line 1
in block <unit> at <tmp> line 1
Kaypie m: sub to-id(Str $str) { $str.samemark(' ').lc.subst(/<-[a..z0..9\-]>+/, '', :g) }; 'POAoaijodPJfiAPdfa' ~~ /$<command>=.+/; say do ~$<command> R[//] Nil ==> to-id()
camelia poaoaijodpjfiapdfa
Kaypie m: sub to-id(Str $str) { $str.samemark(' ').lc.subst(/<-[a..z0..9\-]>+/, '', :g) }; 'POAoaijodPJfiAPdfa' ~~ /$<command>=\d+/; say do ~$<command> R[//] Nil ==> to-id()
camelia Use of Nil in string context

in code at <tmp> line 1
Kaypie is there a less dumb way to write this?
m: sub to-id(Str $str) { $str.samemark(' ').lc.subst(/<-[a..z0..9\-]>+/, '', :g) }; 'POAoaijodPJfiAPdfa' ~~ /$<command>=.+/; say do ~$<command> R[//] $<command>.ast R[//] Nil ==> to-id() 13:19
camelia poaoaijodpjfiapdfa
timotimo instead of the subst, a trans might be good
alternatively a comb and join perhaps
Kaypie the gibberish is kind of a bad example 13:20
the text can be any unicode 13:21
vrurg holyghost: good luck then! 13:22
13:29 Manifest0 left 13:31 noisegul joined 13:35 Manifest0 joined
Kaypie is there a way to speed up precompilation? on one of my modules it can take 5-10 minutes to finish 13:37
13:39 Manifest0 left
timotimo you can start by checking what RAKUDO_MODULE_DEBUG says what it does, is it rebuilding more than you think it should? 13:40
13:43 AlexDaniel joined 13:44 Manifest0 joined
ugexe just reduce the amount of code that needs to be precompiled 13:51
i dont think it matter if its rebuilding more than one thinks it should... whats the alternative 13:54
lets say you find out it is rebuilding more than you think. you could try and make it smarter, but i'm not sure how you do that since you must run the code to pull in further dependencies. additionally the order is now sorted so that reproducible builds can happen 13:55
Kaypie it's recompiling modules that have never been changed 13:57
timotimo "no precompilation;"?
ugexe if the repo-id changes then they must be reprecompiled
the repo-id off a -I. or -Ilib will if the source of any files change
s/will/will change/ 13:58
s/off/of/
same as if you bumped the version of a distribution and installed it (it doesn't not precompile stuff that didnt change from the last version) 13:59
Kaypie ohhhhhh
so it's -Ilib that's the problem
timotimo there's a simple solution
split every file into a separate folder structure and add lots of -I
hold on, the repo id also contains the next-in-chain's repo id, no?
ugexe yes, so you would need to use the same order of lib folders 14:00
every time
timotimo well, that's the easy part
but whenever one of them has a file in it that changes, everything "above" that will rebuild? 14:01
ugexe yeah
Kaypie in lib i have a plugins and games directory 14:02
i'm planning on moving more code over to them so i can have more pure code
timotimo well, if you've got 20 files and are planning to only work on a single one of them for the next few days, it could help to split 1 vs 19
14:26 Manifest0 left 14:29 epony joined 14:32 Manifest0 joined
patrickb Is it true that the need to recompile is to some extent related to the hard linking of compilation units that moarvm does? I.e. if we'd manage to decouple the compiled files via more intelligent linking we'd need to recompile less? 14:40
I remember a discussion of mst and jnthn a while ago...
ugexe i think the killer is how to handle e.g. `augment Test { ... }` when Test has already been used somewhere else earlier 14:43
jnthn patrickb: It only does that because there isn't any absolute naming, which would be needed for dynamic linking. 14:45
patrickb Is dynamic linking an achievable goal? Or architecturally unfeasable? 14:47
Kind of sounds like a potential GSoC project...
jnthn I've no idea how it'd be achieved.
Or a potential PhD project :P 14:48
14:53 Manifest0 left
ugexe isn't alexdaniel almost done with their masters? :P 14:58
14:58 Manifest0 joined
timotimo isn't today the defense? 15:00
patrickb I know near to nothing about linking. But intuitively I'd say it shouldn't be that hard. I guess an uninformed intuition isn't worth much. :-P 15:01
15:02 antoniogamiz joined 15:14 Manifest0 left
antoniogamiz mmm do you know how much time get an CPAN ID takes? 15:15
I asked one 2 or 3 days but I've got no answer 15:16
15:19 Manifest0 joined
timotimo yeah, it takes a couple of days; i guess whenever the person or people who are in charge notice and have a spare moment 15:21
antoniogamiz ah okay, ty timotimo :) 15:23
timotimo surely there's an address you can write an email to to remind them 15:24
15:24 ravenousmoose left
Geth ecosystem: f9b2b51777 | Antonio++ (committed using GitHub Web editor) | META.list
public Perl6::TypeGraph
 15:28
15:34 antoniogamiz left 15:45 gc29180 joined
AlexDaniel timotimo: yes, and it's over! 15:46
👌
15:50 mowcat joined
holyghost NOTE : I've moved PaganVisions2 project name to just "pagan" on github, it does not compile yet 15:52
PaganVisions2 will be deprecated to the fullyness 15:53
15:56 aleck joined
sena_kun AlexDaniel, my congratulations! 15:56
15:56 lgandras joined, aleck left
holyghost AlexDaniel : congrats 15:59
16:04 gc29180 left 16:06 dolmen left
moritz AlexDaniel++ 16:08
16:09 dolmen joined 16:12 mtg joined
holyghost AlexDaniel : it was a Ph. D defense, I guess ? 16:12
AlexDaniel holyghost: no, just Master's
holyghost what topic ? 16:13
AlexDaniel you can find a pdf here: github.com/RGVID-EU/RoboRG-Docs/releases 16:14
holyghost ok
I remember my Master's thesis, it was called "Development on NK Fitness landscapes", it has a paper in a Springer-Verlag book 16:16
anyway, glad you're finished with it :-) 16:17
jnthn AlexDaniel: Congrats! :)
AlexDaniel thanks everyone :)
16:18 Kaypie is now known as Kaiepi
Kaiepi i keep struggling to remember how to write proxies so that they don't throw when you try to use them 16:19
idk why
they're nowhere near as bad as js proxies, i don't get why i can never remember not to use methods for FETCH/STORE 16:20
AlexDaniel Kaiepi: why does it sound like you're using prxies a bit too often? :)
proxies*
Kaiepi i've only used them a couple times
but i've wasted a lot of time on them
holyghost design patterns in perl6 I guess, cool 16:22
We're going totally java :-)
TMTOWTDI 16:26
I like tizen.org as a git server for Samsung linux, it uses C/C++/C++17 AFAIK 16:27
as a devel language, I mean, Windows devel env though, the rest of git.tizen.org casts to linux 16:29
Samsing linux == Android os 16:31
16:31 patrickb left
holyghost *Samsung 16:31
16:42 farcas82 left
Kaiepi `is rw` ended up being a better fit than proxies 16:43
16:48 arkaitz joined 16:51 araraloren left 17:01 dolmen left 17:15 Manifest0 left 17:20 Manifest0 joined 17:21 ravenousmoose joined
vrurg AlexDaniel: Wow! +1 to the congrats chorus! 👍🏻 17:21
nadim_ AlexDaniel: Congratulations and what a fun idea! 17:25
AlexDaniel thank you :)
nadim_ I also wantto see the video of you doing rythmic dance to test your ideas.
AlexDaniel heh. Just went to a real competition to test it :) 17:26
MasterDuke AlexDaniel++ 17:27
17:29 robertle joined 17:31 patrickb joined 17:48 voldemortensen joined 17:49 voldemortensen1 joined, voldemortensen left
voldemortensen1 Not sure who to contact, but the cert for examples.perl6.org/ doesn't include examples.perl6.org in the name list. 17:50
17:50 arkaitz left 17:51 voldemortensen1 left
sena_kun rba, ^? 17:51
lizmat timotimo ^^ 17:53
17:58 Manifest0 left 18:02 arkaitz joined 18:03 Manifest0 joined 18:04 rindolf left 18:05 rindolf joined
moritz the site hasn't been resurrected yet 18:08
MasterDuke moritz: perl6-all-modules looks like it hasn't had an update since April 19 18:15
18:16 Manifest0 left 18:18 Kaiepi left
moritz MasterDuke: yes, it had been updated on hack 18:18
18:18 Kaiepi joined
rba sena_kun: timotimo did this temporary setup. we need to ask him. 18:18
MasterDuke moritz: ah, ok
moritz I'll need to replicate the setup somewhere else 18:19
rba we are working on giving these websites new homes. hopefully soon.
18:20 lgandras_ joined, domidumont left
sena_kun rba, I see. glad to know it. 18:21
18:21 Manifest0 joined 18:22 lgandras left 18:24 sauvin left, MasterDuke left 18:25 noisegul left 18:27 chakli joined 18:29 MasterDuke joined, MasterDuke left, MasterDuke joined 18:31 arkaitz left
chakli Is there a way to profile Grammar and action? I have a grammar that runs over a file thats around 30 mb and it usually take a minute and half to parse. I 18:43
timotimo yeah, the profiler can do that 18:44
chakli perl6 --profile ? I did try that but it generated a 150 MB html file! Any hints what I should look out for? 18:47
timotimo oof. yeah, in that case you'll have to use the sql output and explore that with moarperf 18:48
18:52 ravenousmoose left
chakli Ok, ill try that 18:54
I also notices that GC took ~38% time
is that normal?
timotimo it can be, depends on a lot of factors 18:55
roughly, the more objects survive, the longer it takes and the more inter-object references there are, the longer it takes
chakli RES memory was around 160MB on top but VIRT went upto 20GB 18:57
timotimo i'm not sure what kinds of stuff moar does that brings virt memory up a lot without using res
but normally big amounts of virt are no problem 18:58
chakli Also i read the file like 'file'.IO.lines.map({ ...parsng code... })
not sure if that might be the bottleneck
18:58 ravenousmoose joined
timotimo the shorter the lines, the higher the overhead from mapping, but it shouldn't be terribly slow 18:59
oh, grammars also have quite a bit of overhead for startup and finishing, which is also not great for very short lines
chakli ah, they are short lines around 160 characters/line 19:00
timotimo ah i was thinking more like 5 characters
chakli but even with ~200/lines its around 150,000 lines 19:01
If probably Grammar is the reason i can probably use grep to filter the lines first 19:08
19:08 squashable6 left
cpan-p6 New module released to CPAN! JSON::Fast (0.9.13) by 03TIMOTIMO 19:09
timotimo functions like .starts-with .ends-with and .contains with only strings can be a great performance boost
chakli instead of grep? cool, ill try that then. 19:11
19:11 squashable6 joined
timotimo yeah, you can use "next" inside of mapped blocks 19:11
that should be good for skipping things faster
19:19 lgandras_ left, Manifest0 left 19:23 chakli left 19:24 Manifest0 joined
holyghost timotimo++ 19:26
timotimo lizmat's responsible for this newest release fwiw 19:28
19:32 rindolf left 19:46 johnjohn101 joined 19:47 Manifest0 left 19:52 Manifest0 joined 19:54 xinming joined
sena_kun greppable6, fopen 19:55
greppable6 sena_kun, 131 lines, 30 modules: gist.github.com/13904ea3167fa692f8...f3321de3d9
19:57 xinming_ left
holyghost I'm going to go offline for the weekend, I'll work on pagan next week and talk some more about it 20:01
then
CUL
20:01 holyghost left
nadim_ I gett error: This type cannot unbox to a native string: P6opaque, Failure, yes that's all I get. that's not very helpful. Now I know what the problem is, I think, becaue I can run the code successfully when I use -Ilib. 20:08
I think I have another version ofthe module installed and that versions is broken
1/ I have no idea where that versiionis installed 2/ the error message is still not heplful 20:09
ugexe zef uninstall Module::Name -- will uninstall all versions
nadim_ I'd like to remove the installed version not install over,
20:10 Manifest0 left
ugexe zef uninstall "Module::Name:ver<...>" 20:10
nadim_ ugexe: yes, thank you but I would have liked to know where it is installed
20:10 Rocco joined
nadim_ for science 20:10
ugexe zef locate "Some::Module:ver<...>" i would guess
nadim_ we discussed this yesterday, I though I had dd somewhere in my code but it is in the installed version not my dev one 20:11
20:11 robertle left
nadim_ hmm, how do I know what version of the module gets loaded? I have no idea which one wasinstalled 20:11
ugexe zef locate doesn't work with versions yet it seems
20:11 Rocco left 20:12 dolmen joined
ugexe run your script with RAKUDO_MODULE_DEBUG=1 20:12
it will output a wall of text
but you can use that to determine exactly which files were loaded
perl6 -e 'say $*REPO.resolve(CompUnit::DependencySpecification.new(:short-name<Zef>)).distribution' 20:13
that will show you which version of zef gets loaded for just `use Zef;` 20:14
perl6 -e 'say $*REPO.resolve(CompUnit::DependencySpecification.new(:short-name<Zef>)).distribution.meta'
nadim_ zef locate says which version! cool, From Distribution: Data::Dump::Tree:ver<1.8.0>:auth<github:nkh>:api<>
ugexe zef list --installed will show you all installed modules and all their versions as well
nadim_ good to know too 20:15
20:15 Manifest0 joined
nadim_ is there a completion file for zef somewhere? 20:16
20:16 rindolf joined
ugexe no. ingy said he was going to look at doing that sometime 20:18
20:26 Manifest0 left 20:27 dolmen left
sena_kun greppable6, help 20:30
greppable6 sena_kun, Like this: greppable6: password # See wiki for more examples: github.com/perl6/whateverable/wiki/Greppable
sena_kun greppable6, crt 20:31
greppable6 sena_kun, 128 lines, 20 modules: gist.github.com/191d1e3c27e625c39f...158e63e8cc
20:31 Manifest0 joined 20:34 ravenousmoose left 20:36 Manifest0 left, ijneb joined 20:42 Manifest0 joined 21:05 dolmen joined 21:07 dolmen left 21:12 ToddAndMargo joined
ToddAndMargo Anyone on newbie duty? 21:12
Question: is there a way to tripper parent directory creation with `mkdir` as does bash `mkdir --parents xxx`? docs.perl6.org/routine/mkdir is beyond confusing 21:14
trigger not tripper
alexghacker That doc says: "Also creates parent directories, as needed (similar to *nix utility mkdir with -p option); that is, mkdir "foo/bar/ber/meow" will create foo, foo/bar, and foo/bar/ber directories if they do not exist, as well as foo/bar/ber/meow." 21:15
If it's failing then there's a bug in either the doc or the implementation 21:17
(I have no idea which)
if, on the other hand, you're just not seeing the directories you expect then maybe your code isn't executing where you think it is?
ToddAndMargo I did a search in page an could not fid that verbage, but it did show on Google short discription. And I found that I had a file called "a" which was what I used as a parent. So "short between the headsets" 21:19
# p6 'mkdir "a/b/c, 0o766";' Failed to create directory '/home/linuxutil/a/b/c, 0o766' with mode '0o777': Failed to mkdir: Not a directory in block <unit> at -e line 1
alexghacker are we looking at the same docs.perl6.org/routine/mkdir ?
ToddAndMargo could be a little more explanitory
ugexe well, its telling you what is happening already 21:20
alexghacker perhaps the error message was just LTA
ugexe you are trying to create a path '/home/linuxutil/a/b/c, 0o766', including command and space
ToddAndMargo no, googel popped me over to docs.perl6.org/type/IO.html and I did not realize it. Showed correctly o the hits page, which was weird. 21:21
ugexe mkdir "a/b/c, 0o766"; -- i mean how the hell can the compiler know that you didnt mean that nonsensical path
mkdir "a/b/c, 0o766"; is NOT the same as mkdir "a/b/c", 0o766; 21:22
and the error is telling you everything you need to know to figure that out already 21:23
ToddAndMargo not sure what yo mean. My mistake was that I had an "a" file (not directory) alrady existing
ugexe read what you pasted 21:24
the code
21:24 dolmen joined
ToddAndMargo bash does a nice error report: `touch a; mkdir -p a/b/c` 21:24
mkdir: cannot create directory ‘a’: Not a directory
ugexe you are trying to create a file called "a/b/c, 0o766". The error explicitly tells you this, and also that it is using mode 0o777 which is not what you were trying to do
ToddAndMargo Could we have something like that in p6? 21:25
Kaiepi moarvm's openbsd port is just about ready! just two more to go
ToddAndMargo I was trying to create parents directories a, a/b, then a/b/c. It crashed because I already had an "a" file. I was not trying to create a file. I am confused as to your question 21:26
ugexe i dont have any questions. i was pointing out what is wrong with everything you are doing. 21:27
ToddAndMargo I do realize what I did wrong, now. Did I do somethign else wrong as well? 21:28
ugexe ToddAndMargo># p6 'mkdir "a/b/c, 0o766";' Failed to create directory '/home/linuxutil/a/b/c, 0o766' with mode '0o777': 21:29
you are passing ONE argument to mkdir
you should be passing TWO arguments to mkdir (if you want to use mode 0o766)
the error message explains everything. look at the name of the directory it is trying to create
ToddAndMargo As soon as I deleted the file called "a", this worked perfectly `p6 'mkdir "a/b/c, 0o766";'` I am confused by your statement. 21:30
timotimo that's a very strange name for a folder 21:31
ToddAndMargo oh I see. I created a fiel called 0o766
21:31 natrys left
timotimo no, you created a folder named "c, 0o766" 21:31
ToddAndMargo I was trying to set attributes
how would you modfiy my line to get the attributes (mode)? 21:32
ugexe ugexe>mkdir "a/b/c, 0o766"; is NOT the same as mkdir "a/b/c", 0o766;
21:33 AlexDaniel left
ugexe i gave you the answer 10 minutes ago 21:33
ToddAndMargo Please understand that I can be dense as a rock at times. `p6 'mkdir ("a/b/c", 0o766);'` does look like two paramters to me. Would you mind modifying my line so I can see exactly what I am doing wrong? 21:35
ugexe ...that is not the code you posted before
ToddAndMargo># p6 'mkdir "a/b/c, 0o766";
ToddAndMargo give the same error.
ugexe that is ONE argument
no it doesnt 21:36
ToddAndMargo you are correct. I though I saw what I saw. Thank you!
I kept seeing `"a/b/c", 0o766` when it was `"a/b/c, 0o766"`; the 0o766 is a Samba specifiy attribute 21:39
timotimo could it be you have very mild dislexia and could benefit from a font specifically made to help dislexic people read stuff more easily? 21:43
21:44 AlexDaniel joined
AlexDaniel MasterDuke: thank you especially for taking a look ♥ 21:45
ToddAndMargo Hi Timo, I am not dislexic. I am "weird". There is a difference. Dislexic folks don't reach for the salt when they want the pepper 21:46
AlexDaniel :) 21:47
Altreus Is there a variable containing the current package name? 21:53
AlexDaniel m: say $*PACKAGE 21:58
camelia Dynamic variable $*PACKAGE not found
in block <unit> at <tmp> line 1
AlexDaniel m: say $?PACKAGE
camelia (GLOBAL)
AlexDaniel this?
nadim_ ugexe: a simple comletion for zef in the meantime gist.github.com/nkh/68c8462f170fad...578447a3af 22:02
I haven't found how to pass --/option to bash yet 22:04
22:08 dolmen left 22:10 johnjohn101 left
ugexe if you wanted to get extra fancy it'd look at the zef config and add options for every field called `short-name` such as github.com/ugexe/zef/blob/e7389737...g.json#L33 22:10
--<short-name> to enable any of them, and --/<short-name> to disable any of them
so some shmuck might have a custom config with or without everything in the base config 22:11
perl6 -e 'use Zef::Config; say Zef::Config::parse-file(Zef::Config::guess-path).values.grep(* ~~ Array).map(*.Slip).grep({.<short-name>:exists}).map({.<short-name>})' 22:15
that will get you a list of all the options from the config
22:15 rindolf left
nadim_ arent't those in the help enough? 22:15
I very rarely use zef, except for install,like 3 times the last days and never before 22:16
ugexe dunno, i dont use tab completion 22:17
nadim_ are those short names commands or options?
ugexe options
nadim_ I'll add them, update the gist and then you and I can forget about them :) 22:18
Altreus AlexDaniel: ah that's it, thanks
SmokeMachine: what have I done wrong? A model cannot have attributes, but you tried to declare '$!student-email' 22:19
has Str $!student-email is referencing{ :model<Student>, :column<email> }
22:19 rindolf joined
Altreus gist.github.com/Altreus/1482449a60...b0a1061281 22:21
22:21 AlexDaniel left
Altreus gonna go to bed and pick it up later 22:21
22:21 cpan-p6 left 22:23 cpan-p6 joined, cpan-p6 left, cpan-p6 joined
ingy nadim_: I'll have completion for zef ready for yapc :) 22:25
ToddAndMargo this stopped working on me after upgrading to Fedora 30 from 28. "Malformed UTF-8" 22:28
@LogFiles = @LogFiles.sort: {my ($month, $day, $year, $hour, $minute, $second) = .comb(/\d+/);($year // 0, $month // 0, $day // 0, $hour // 0, $minute // 0,$second // 0, $_);}
And it only bombs if there are more than one entry in @LogFiles 22:29
entries look like this: /home/tony/Documents/CimTrak/logs/cimtrak.log.06-07-2019_15:21:12.zip
nadim_ ingy: good work. the one I uploaded is more than enough for me, except install I have never used anything before yesterday 22:32
22:39 patrickb left, rindolf left
Geth ecosystem: fc70c02798 | thundergnat++ (committed using GitHub Web editor) | META.list
Rename module

Rename Lingua::EN::Numbers::Cardinal to Lingua::EN::Numbers. Has expanded to a more general purpose module.		 22:40
vrurg zef completion for fish somebody?? 22:41
22:41 rindolf joined
ugexe i think ingys will work on all shells 22:41
22:44 mowcat left 22:55 farcas82 joined 23:06 mtg left 23:08 rindolf left
ToddAndMargo came up with a workaround for my utf-8 issue and sort. I threw a `@LogFiles[$I] ~~ s:global|''||;` on each element; What the heck? 23:20
Kaiepi moarvm on openbsd ports is a no go 23:29
23:29 aindilis` joined
Kaiepi the version of libuv it has is too old 23:29
maybe i should upgrade it
ToddAndMargo This did the trick. I added .Str to $F `for dir $ZipCimLogPath -> $F { push @Files, $F.Str; }`; Some weirdness with Fedora 390 upgrade 23:30
30 not 390
vrurg Kaiepi: it won't hurt anyone, I think.
23:32 aindilis left
Kaiepi actually i think i'll hold off on updating moar/nqp/rakudo until the next release 23:34
mainly because there are still issues like github.com/rakudo/rakudo/pull/2963 23:35
23:45 ToddAndMargo left, aborazmeh joined, aborazmeh left, aborazmeh joined 23:47 nadim_ left 23:57 sena_kun left

Powered by Cro and the Raku® Programming Language.

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

Thank you!