»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.
timotimo the code was trying to be careful here, but didn't have in mind one important thing 00:00
vsnprintf will only write up to n bytes, and it returns how many bytes it has written ... except!
except when it *would have* written more than n bytes 00:01
in which case it returns *that* number, i.e. a number bigger than n
00:03 bjz left 00:05 Cabanossi left 00:08 Cabanossi joined
timotimo bedtime! o/ 00:24
00:25 isBEKaml joined
sammers m: Str.^add_method('こんにちは', method () {"{self}ちゃん、こんにちは。"}); "Camelia".こんにちは.say; 00:32
camelia Cameliaちゃん、こんにちは。
sammers so this will work: "Camelia"\n.こんにちは\n.say 00:33
even with non alpha method names 00:34
AlexDaniel sammers: docs.perl6.org/language/unicode_te...Characters 00:35
say ‘こんにちは’.uniprops
evalable6 (Lo Lo Lo Lo Lo)
sammers ah 00:36
ok, that is cool
00:53 gdonald left, bjz joined, gdonald joined 01:03 bjz left 01:05 tomboy64 joined, Cabanossi left 01:06 azawawi joined
azawawi hi #perl6 :) 01:07
sammers ola
01:08 Cabanossi joined, isBEKaml left
azawawi :) 01:09
azawawi begins removing the dust off his old perl 6 modules 01:10
01:26 bjz joined 01:33 cyphase left 01:37 Cabanossi left 01:38 Cabanossi joined 01:39 AlexDaniel left, cyphase joined 01:46 ilbot3 left 01:48 ilbot3 joined, ChanServ sets mode: +v ilbot3 02:05 Cabanossi left 02:08 Cabanossi joined 02:21 noganex joined
labster m: say Ⅹ×ⅩxⅩ 02:22
camelia 100100100100100100100100100100
02:24 noganex_ left 02:28 TimToady left 02:30 TimToady joined 02:34 cpage___ joined, cpage_ left, cpage___ is now known as cpage_ 02:37 bjz left 02:45 ChoHag left 02:47 BenGoldberg joined
TEttinger labster: wat 02:50
m: say Ⅹ
camelia 10
TEttinger m: say x 02:51
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared routine:
x used at line 1
TEttinger so confused
geraud TEttinger: just guessing, but X (roman numeral) is changed to 10. And depending on the font you use, there are 2 kinds of 'x', the first one looks weird so I would assume it's some unicode character for multiplication, and the second one is for string duplication. 02:57
TEttinger oh, it's duplicating "100" ten times 02:58
geekosaur 3 roman-numeral-Ⅹ-s, one multiplication symbol ×, one string replication ordinary x) 02:59
where'd that ) come from... 03:00
03:06 BenGoldberg left 03:10 tyil left 03:11 tyil joined 03:15 bjz joined 03:18 BenGoldberg joined 03:19 Ben_Goldberg joined, BenGoldberg left, Ben_Goldberg is now known as BenGoldberg 03:33 nimajneb joined
nimajneb How would I play sounds through its frequency in hz 03:33
w/ perl6 or anything that I can program 03:34
03:34 Cabanossi left
sammers nimajneb, have to taken a look at any of these? modules.perl6.org/t/AUDIO 03:35
nimajneb Yeah, but do any of those allow for outputting via frequency 03:36
Not sure about how I would accomplish that
03:37 Cabanossi joined
nimajneb I'd be fine with using the portaudio module 03:37
still not sure about how I would accomplish this though
BenGoldberg Any sound player is going to require, not a frequency, but some data in a valid sound format. So, your first step is to say, "How do I create a sound file whose content is nothing more than a tone at X frequency and Y volume for Z duration." 03:43
Then, only after you've created the data, do you worry about playing it.
nimajneb Is somehow generating .wmv files fairly simple? 03:49
would .au be simpler?
BenGoldberg Honestly, I'm not sure. 03:54
azawawi pasteboard.co/4359L3M3R.png # Hello world using ncurses color github.com/azawawi/perl6-ncurses/b...colors.pl6 :) 04:16
04:21 zapwai left 04:35 skids left 04:37 nimajneb left 04:39 dj_goku left
azawawi pasteboard.co/43xgAJ0hR.png # Move windows using NCurses github.com/azawawi/perl6-ncurses/b...indows.pl6 :) 04:40
04:46 dj_goku joined 04:50 Cabanossi left 04:52 Cabanossi joined
sammers azawawi, I have a sudden urge to play solitaire 04:52
TimToady I'm afraid you'll have to play it by yourself... 04:53
sammers ha 04:54
azawawi :)
samcv ah. i think i have a pdf not in unicode maybe. cause i copy the letters. and. it copies something different
or maybe something else weird idk
oh. unless... they're using a non unicode font
latin letters are normal but the special characters here all copy as different other characters 04:59
sammers is there any way to utilize hyper operators / similar when using ==> ? 05:00
samcv i think they must be using some non-unicode font because some things copy as multiple characters, so it's like they used ligature hacks with this font 05:02
05:11 Sgeo_ joined
samcv not sure sammers 05:13
05:13 Sgeo left 05:16 wamba joined 05:17 BenGoldberg left 05:22 stmuk_ joined, azawawi left, stmuk left 05:32 domidumont joined
moritz feed operators are meant to provide some level of parallelity, they just don't yet 05:33
the idea is that they are like pipes, running separate threads for each block that's connected by feeds/pipes 05:34
sammers moritz, is the idea that the parallelity will auto-apply when applicable or will there be a way to influence the behavior? 05:36
like, ==> should imply >>. whenever it is possible? 05:37
moritz sammers: I don't know yet
sammers ok, just curious
05:38 domidumont left 05:41 domidumont joined 05:43 domidumont left 05:50 Cabanossi left 05:52 Cabanossi joined 05:55 RabidGravy joined 05:58 domidumont joined 06:20 Cabanossi left 06:22 Cabanossi joined 06:32 darutoko joined 06:56 labster left 07:02 [particle]1 left 07:03 [particle] joined, ChoHag joined 07:05 labster joined
samcv perl 6 is a very ____ project. i'm trying to think of the word i'm looking for. which means optimistic, and an effort which has high breadth and larger than normal scope 07:06
but my mind is not thinking of the word. can anybody help 07:07
labster ambitious? grippy?
samcv ambitous
labster is just pulling words from the opening of the Colbert Report. 07:08
samcv thank you 07:09
i hate when i know exactly the word i mean but i can't think of it 07:10
07:11 rindolf joined 07:12 mr-foobar left, mr-foobar joined
labster As a creative writer, I feel this pain. 07:13
samcv feels like your brain is broken 07:14
07:14 TreyHarris left
samcv *why* won't it thinkkkk of the word i already know i know! 07:14
07:14 tomboy64 left
RabidGravy I have that problem with whole concepts and categories of ideas 07:19
samcv please tell me how to fix this 07:20
07:21 Cabanossi left
RabidGravy I usually trick my brain by thinking about something completely different 07:22
07:22 Cabanossi joined
samcv need an industrial strength therarus of words 07:22
a word web or something
with deep learning we can find words that are created when you add two words together. it's very interesting 07:23
also 07:25
m: "ā".NFD.say
camelia NFD:0x<0061 0304>
samcv when is this gonna be a thing
0x<0061 0304> to make an array of hex digits or something
that would be neat 07:26
07:29 abraxxa joined 07:34 abraxxa left 07:35 astj_ joined 07:36 astj left
moritz m: say Blob[uint16].new: <0061 0304> 07:46
camelia Blob[uint16]:0x<3d 130>
TEttinger samcv: perl 6 is a ____ project <-- quixotic, grandiose, over-achieving, moon shot of a, mind-blowing, unicodatious 07:47
napoleonic
samcv well. they did go to the moon ;) so we have a chance
TEttinger and napoleon did conquer most of europe 07:48
and don quixote did joust windmills
samcv yes
TEttinger and I liek vocabularistery 07:49
samcv that's not a work but ok :-)
TEttinger neither is :-) and neither is the american flag emoji
but you can't deny that they have meanings! 07:50
samcv they do have meanings
TEttinger I wonder when you can start assigning unicode gender identifiers to flags
samcv what...
i mean
TEttinger female north korea
samcv unlikely
geekosaur expansive, audacious 07:51
TEttinger the audacity of wat
(the main practical use of gendered flags would be representations of officials from that country, using the gendered modifier and the flag modifier on a person emoji) 07:52
such as female + germany + leader for merkl
or however it's spelled
samcv that's very specific
TEttinger unicode is very specific
samcv i don't think that is used in practice though. 07:53
there is not a huge desire for this to be implemented
TEttinger that could be said for many parts of unicode!
samcv like which
07:54 abraxxa joined
TEttinger the counterexample is actual stuff that needs special PUA stuff to work and is non-standard because of Unicode consortium decisions, like klingon having books written in it that cannot be transcribed to unicode 07:54
similarly with tengwar, an elven script from LotR 07:55
samcv oh i thought you were going to argue the oppostie that there's things in unicode that have no purpose
TEttinger oh there totally are
samcv well.
maybe when they have put in all the languages they can add klingon
TEttinger the phaistos disk they don't know if it's a language or not, same with linear a and b 07:56
samcv there are argubly more books written in ancient languages than klingon
TEttinger there are no books written in linear b
samcv idk what that is
TEttinger it's a unicode block 07:57
samcv i don't think it's an ancient language though
TEttinger it's an ancient <unclear if language or symbolic system>
samcv not true
idk.
en.wikipedia.org/wiki/Linear_B_Syllabary
TEttinger hm?
samcv there are things in it
there are more books in ancient languages than klingon though. probably i don't have any statistisc though
TEttinger then fine en.wikipedia.org/wiki/Phaistos_Dis...ode_block) 07:58
this is one that is and may forever be undeciphered
samcv lol wat
TEttinger they have no idea what it means, it's in unicode
samcv reads the proposal to add it
huf that's just those animal shaped crackers
TEttinger yes, we need animal crackers in unicode 07:59
samcv The fundamental issue, of course, is should this be encoded at all? There is exactly one extant
document in this script, and a short one at that. That could change at any minute, but given the
failure of any other document written with this script to surface in nearly a century, it doesn't seem
likely.
07:59 astj_ left
huf i still think we need quake3 in unicode 07:59
samcv :|
TEttinger yep 08:00
08:00 astj joined
samcv m: "𐇐".uniname.say 08:00
camelia PHAISTOS DISC SIGN PEDESTRIAN
samcv glad i have a font with these characters in it
otherwise how could i live with my life
08:01 baest left
huf but seriously why not add whatever to unicode? we're not going to run out of numbers... 08:01
samcv well they should add klingon probably if it's used
TEttinger there is more text in the voynich manuscript than in phaistos disc objects
(one copy of each)
samcv but. isn't that not an actual language
or. is that the one we still don't know 08:02
TEttinger also, undeciphered
just as actual as phaistos
samcv i forget there was one that we deciphered there are several ones undeciphered though
several manuscripts i mean that are mysterious
TEttinger there's an xkcd comic about it www.xkcd.com/593/
08:03 gdonald left
samcv hah 08:03
08:03 gdonald joined 08:04 baest joined
samcv TEttinger, we might know 10 words from the book www.livescience.com/43542-voynich-m...acked.html 08:11
TEttinger nice
samcv which seems likely since like there are drawings 08:12
and it's a long enough book
TEttinger "the", "a", "foxglove", "warhammer"
"+1 Vorpal" 08:13
samcv what?
TEttinger it has to be a gaming text! 08:14
and dungeons and dragons has +1 Vorpal Longswords and such
samcv nope
vorpal is a made up word though
created in modern times
TEttinger the confusing thing is that there are what appear to be alchemical diagrams using the metaphor of a bath house 08:15
lots of nude fairies or women or something in a system of pipes, kinda suggesting an apparatus to move stuff through
that would sync up with the plant diagrams from earlier in the text 08:17
but it could also signify star movement or internal medicine symbolism
geekosaur contemplates someone in the future trying to decipher "Jabberwocky" 08:19
08:20 Cabanossi left 08:22 Cabanossi joined
samcv we have a written record though. so 08:23
seems pretty likely
TEttinger, i'm watching the video www.youtube.com/watch?time_continu...pZD_3D8_WQ 08:26
08:28 astj_ joined, astj left, baest left 08:51 cyphase left 08:53 cyphase joined 09:06 synopsebot6 joined, ChanServ sets mode: +v synopsebot6 09:21 baest joined 09:25 astj joined, astj_ left 09:29 baest left
RabidGravy Hmm, that was simpler to fix than I thought 09:48
09:51 Cabanossi left 09:52 Cabanossi joined
RabidGravy timotimo, if you're around be a love and merge github.com/arnsholt/Net-ZMQ/pull/11 - it fixes the last problem with Net::ZMQ and makes properly usable again 09:56
09:58 baest joined 09:59 wamba left 10:00 wamba joined 10:21 Cabanossi left 10:22 Cabanossi joined 10:34 mr-foobar left 10:36 mr-foobar joined 10:43 baest left
Geth ecosystem: e6e0583979 | (Julien Simonet)++ | META.list
[Log::Any] Fix META6 URL.

s/META6.info/META6.json/		 10:45
10:58 labster left 11:06 baest joined 11:16 bjz_ joined 11:18 bjz left 11:24 baest left 11:29 baest joined 11:31 Possum left 11:34 mxco86 joined 11:49 noganex left 11:50 mr-foobar left 11:51 baest left 12:01 baest joined
dalek t-ZMQ: f499171 | RabidGravy++ | lib/Net/ZMQ/Message.pm:
Fix the wrong sized Message struct.

This fixes the memory access faults and runs clean with valgrind		 12:09
t-ZMQ: 15da7f5 | azawawi++ | lib/Net/ZMQ/Message.pm:
Merge pull request #11 from jonathanstowe/opaque-message

Fix the wrong sized Message struct.
RabidGravy BOOM! 12:10
12:10 azawawi joined
azawawi hi 12:10
RabidGravy my plans for world domination are coming along nicely
yo azawawi 12:11
azawawi our $plans; # you mean :)
moritz that's what he wants you to think! 12:13
speaking of world domination 12:14
I've signed a contract with a major tech publisher, who will publish "Perl 6 by Example"
though with a different title, likely
jnthn moritz++ # cool! :)
azawawi grats
moritz thanks
RabidGravy fabulous! 12:15
12:21 noganex joined, nadim joined
nadim Yo! I get the feeling that there is a nicer way to write this: 12:21
if %options<display_info>.defined && %options<display_info> == False
azawawi uses the NCurses card to begin his world domination :)
tbrowder ugexe: when a META6.json has a line pointing to a github release zip archive, is that the exact source for an ecosystem install? 12:22
with zef, anyway? 12:23
moritz nadim: if %options<display_info> === False { } =
s/\= $//
nadim can I also haz an explication? 12:24
moritz === is the identity comparison operator
12:24 lucas-1 left
moritz so it's only true if the other side is also False 12:24
nadim ah, I get the {} now, it's the block 12:25
moritz if you want a 0 be true too, you could write unless %options<display_info> // True { ... }
nadim would it be possible to do it with a junction (in case one has more tests) something like $options<xxx> ~~ .defined && True && > 3 12:27
12:27 mr-foobar joined
jnthn if (.defined && .so given %options<display_info>) { } 12:28
moritz junctions use single &
m: say so 2 ~~ (.defined & * > 3) 12:29
camelia False
moritz m: say so 2 ~~ (.defined & (* > 3))
camelia False
moritz m: say so 2 ~~ (*.defined & (* > 3))
camelia Too few positionals passed; expected 2 arguments but got 1
in block <unit> at <tmp> line 1
moritz oh, I guess & participates in whatever-currying
m: say so 2 ~~ (.defined & { $_ > 3})
camelia False
moritz confused 12:30
m: say 2 ~~ (.defined & { $_ > 3})
camelia False
nadim 2 > 3 false?
moritz erm, yes
nadim hehe
moritz m: say 2 ~~ (.defined & { $_ 3})
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 3say 2 ~~ (.defined & { $_7⏏5 3})
expecting any of:
infix
infix stopper
statement end
statement modifier
…
moritz m: say 2 ~~ (.defined & { $_ < 3})
camelia True
moritz too tired
lizmat wonders if "say "foo" if $_ === False with ⏏%h<a>" should be legal syntax
(minus the eject symbol" 12:31
moritz lizmat: s/with/given/ ?
lizmat no, with, so you get your defined check and localization 12:32
moritz m: say "foo" if $_ === False with False;
camelia 5===SORRY!5=== Error while compiling <tmp>
Confused
at <tmp>:1
------> 3say "foo" if $_ === False with 7⏏5False;
nadim given a Bool, is the comparison to True(False better written $b == True or $b === True?
lizmat m: dd $_ with False
camelia Bool::False
moritz nadim: I'd simply write it as "$b" and "!$b" 12:33
if you already know it's a Bool
lizmat nadim: are you really checking for existence of the key, rather than definedness of the value ?
12:34 Util_ is now known as Util
nadim for $b === True, you are right, better use it as "$b only", don't know why I compared it altogether! 12:35
as for the first %h<xxx>.defined && $h<xxx> == False; I wanted to know if it was set to false by the user, it is set to True by default 12:37
moritz %h<xxx> //= True; 12:38
nadim I don't want to set it if it is undefined
just know it it is True
lizmat %h<xxx> // True
if it's not defined, it is supposed to be True, no? 12:39
nadim is something undefined always False?
lizmat yep
well....
nadim but False is defined, no?
lizmat m: class A { method Bool { True } }; say ?A
camelia True
lizmat nadim: yes 12:40
moritz m: say False.defined
camelia True
lizmat m: dd $_ with False
camelia Bool::False
nadim m: my $x ; say so $x ~~ .defined 12:42
camelia False
lizmat nadim: without given, you're comparing $_, which is probably not defined 12:43
m: say 42 ~~ False
camelia Potential difficulties:
Smartmatch against False always fails; if you mean to test the topic for truthiness, use :!so or *.not or !* instead
at <tmp>:1
------> 3say 42 ~~ 7⏏5False
False
tbrowder hi, #perl6 12:46
ref bailador: does anyone have a systemd or sysv init script to manage baile? 12:47
12:56 eroux joined 12:57 lucas-1 joined 13:01 mcmillhj joined 13:11 mcmillhj left
moritz tbrowder: systemd service files are pretty easy to write; here is one for a mojolicious application: github.com/moritz/package-info/blo...fo.service 13:12
should be pretty easy to adapt to a bailador app 13:13
and if you leave out ExecStop command, systemd just sends it a KILL
13:20 mcmillhj joined 13:21 skids joined
azawawi Can NativeCall handle slurpy (varargs) C functions (e.g. printf) ? 13:23
13:28 ugjka left
jnthn azawawi: No 13:28
azawawi Noooo :)
13:29 ugjka joined
azawawi jnthn: any workarounds? 13:30
jnthn Implement support for them? :) 13:32
But seriously, I think the calling conv is different so you can't just write the signatures you need
13:33 MilkmanDan left
jnthn oh, maybe you can 13:33
perl6-m -e 'use NativeCall; sub printf(Str, int32) is native {*}; printf("answer is %d\n", 42)'
azawawi I see. Thanks. Anyway for NCurses it is basically formatting functions so i can pass to sprintf and then pass the final string to the native function
jnthn answer is 42
azawawi :) 13:34
13:35 MilkmanDan joined
jnthn You can use is symbol('printf') to get different overloads too 13:39
Only thing that worries me is if this cheat will actually be...reliable... :)
13:40 MilkmanDan left 13:42 MilkmanDan joined
azawawi the actual problem is with scanw* window functions 13:43
13:49 Cabanossi left 13:52 Cabanossi joined 13:56 st_elmo joined 13:59 baest left
azawawi RabidGravy: ping 14:00
RabidGravy erp
azawawi RabidGravy: you wrote the starfields demo right in ncurses?
RabidGravy yeah
worked quite nicely 14:01
azawawi RabidGravy: well now you can resize it :)
RabidGravy kewl
azawawi RabidGravy: epic with the transparency
14:03 baest joined 14:04 bjz_ left, bjz joined
RabidGravy Hmm programming is tiring, I need beer 14:08
azawawi Take your time.... muhaahaahhaa :) 14:11
azawawi enables Atom Perl 6 linter and begins conjuring more n curses examples :) 14:15
14:18 Possum joined 14:21 abraxxa left 14:24 baest left 14:25 rindolf left 14:26 khw joined 14:28 robertle joined 14:32 ponpon is now known as ponbiki 14:35 baest joined 14:42 rindolf joined 14:50 baest left 15:01 baest joined 15:05 AlexDaniel joined 15:12 baest left 15:16 MilkmanDan left, baest joined 15:18 domidumont left 15:19 cdg joined 15:20 Cabanossi left 15:21 MilkmanDan joined 15:22 Cabanossi joined 15:33 kurahaupo_ joined 15:38 w4and0er96 left, w4and0er96 joined 15:41 eady left, eady joined 15:43 baest left 15:44 abraxxa joined 15:45 abraxxa left 15:53 bjz left, baest joined 15:59 zapwai joined
azawawi mouse events with ncurses... done :) 15:59
github.com/azawawi/perl6-ncurses/b...-mouse.pl6
Geth ecosystem: 397598895b | (Jonathan Stowe)++ | META.list
Add MessagePack::Class

See github.com/jonathanstowe/MessagePack-Class		 16:33
RabidGravy azawawi, nice one 16:34
azawawi Thanks 16:35
azawawi throws some salt around him and begins preparing the next curse-d example :) 16:37
tbrowder ugexe: i'm trying to automate creating a module release on github, and then checking it before adding it to the ecosystem. which of the paths in the META6 is used by zef? i think you told me the zip file, but i can't find my notes about 16:41
moritz: thanks, that's helpful. BTW, can you say who will publish the paper copy of yr book, and maybe what its ne name will be? 16:43
moritz tbrowder: not yet, but I'll keep everybody informed when the facts are settled and public 16:47
16:48 baest left
timotimo tbrowder: isn't that what Meta6::bin does? 16:48
tbrowder i'll check, thanks! 16:49
16:54 baest joined 16:55 domidumont joined 16:56 cdg left
tbrowder timotimo: that may have the answer, but i'm have to look a bit deeper... 16:56
timotimo META6 is a module that you can use in your own tool, too 16:57
rindolf azawawi: hi, long time
azawawi: what's new?
timotimo it ought to give you a bit of a head-start
azawawi rindolf: hi. Currently dusting old modules. How are you? 16:58
17:00 cdg joined
rindolf azawawi: i'm fine - working on freecell solver 17:02
azawawi: I attended an interesting conversation at the dog owners' gathering
azawawi: about europe and stuff 17:03
tbrowder timotimo: it's definitely a good head start, thanks for pointing to it! 17:05
RabidGravy timotimo, tbrowder I have a little script that uses META6 to generate the meta file
I never bothered putting it in with META6 because well, it's tiny 17:06
azawawi META6::Teensy32 ? :) 17:08
RabidGravy I have a tab open in my browser that is to remind me to do an experiment with an ATTiny85 17:10
I'm pretty sure Perl 6 will never run on one of those 17:16
17:18 domidumont left
azawawi RabidGravy: power consumption is like 5 mA at 8 MHz @ 5V. Awesome. 17:19
Enough distractions :)
17:21 kent\n left
[Coke] tadzik: readme on rakudobrew has some bad md - look for "run something like" on the front page of the project on github 17:34
17:37 wamba left 17:42 bjz joined 17:49 bjz left 17:53 espadrine joined 17:57 ChoHag left 18:00 cdg left, sftp left, sftp joined 18:03 ChoHag joined 18:06 Cabanossi left 18:07 Cabanossi joined 18:15 domidumont joined 18:17 baest left 18:22 baest joined 18:23 eroux left 18:31 benji__ joined, benji__ left 18:36 baest left 18:40 baest joined
Geth ecosystem: blippy++ created pull request #317:
Remove NCurses6 from ecosystem		 18:49
18:51 darutoko left, rodarmor left
Geth ecosystem: 233002ea56 | blippy++ (committed by Zoffix Znet) | META.list
Remove NCurses6 from ecosystem (#317)

My NCurses6 module is obviated by recent changes to upstream:
  github.com/azawawi/perl6-ncurses
Therefore NCurses6 should be removed as a redundant package		 18:51
18:52 rodarmor joined 18:56 TreyHarris joined 18:57 kurahaupo_ left, kurahaupo_ joined 18:59 stmuk joined 19:01 baest left, stmuk_ left
azawawi github.com/azawawi/perl6-ncurses/b...07-acs.pl6 19:02
19:04 Xliff joined, Xliff left 19:10 Xliff joined
Xliff Hello! 19:11
rakudobrew is not working ATM: Configuration FAIL. You can try to salvage the generated Makefile.
Command failed (status 256): /usr/bin/perl Configure.pl --optimize --prefix=/home/cbwood/.rakudobrew/moar-nom/install --make-install
19:11 baest joined
Geth doc: fff866f52b | (Zoffix Znet)++ | doc/Type/IO/Path.pod6
[io grant] Fix docs for symlink/link routines

  - Fix issues with argument naming
  - List new type constraints for args
  - Mention symlink caveat on Windows
Rakudo impl: github.com/rakudo/rakudo/commit/8c09c84d64
   github.com/rakudo/rakudo/commit/da1dea2ea0
Tests: github.com/perl6/roast/commit/d4353b6d00
   github.com/perl6/roast/commit/8fa49e10a4		 19:15
19:16 ttkp6 joined
Xliff rakudobrew is currently broken. Does anyone know if this is known and if there is a fix avail? 19:17
Actually, I think it might not be rakudobrew specific but a Configuration error. 19:18
timotimo is that the only output you get? 19:20
because that output doesn't give any information at all
Xliff You want everything? 19:21
timotimo well, a bit more would be nice at least
geekosaur rakudobrew is prone to oddnesses in my experience. every so often I have to nuke and reinstall the whole thing 19:23
and there's one build path that hasn't worked for me for months 19:24
timotimo yeah like when i did that thing with libtommath >_<
geekosaur I'm thinking of the "can't find a working make" bit 19:25
timotimo i haven't heard of that
Xliff Not a rakudobrew thing. Configure.pl 19:31
Generating build/mk-moar-pc.pl ......................... OK
Configuration FAIL. You can try to salvage the generated Makefile.
Command failed (status 256): /usr/bin/perl Configure.pl --optimize --prefix=/home/cbwood/.rakudobrew/moar-nom/install --make-install
Command failed (status 7424): /usr/bin/perl Configure.pl --prefix=/home/cbwood/.rakudobrew/moar-nom/install --backends=moar --make-install --git-protocol=https --git-reference=/home/cbwood/.rakudobrew/bin/../git_reference --gen-moar
19:31 domidumont left 19:32 pochi_ left
moritz try deleting the whole moarvm repo, and try again 19:32
19:32 labster joined
moritz a subdir changed to a git submodule or so, and a regular 'git pull' can't handle that 19:33
timotimo "rakudobrew nuke"
yeah, it was basically the dumbest idea and worst mistake i've done in my whole development life
moritz wow, then I guess you never dropped a column from a big table in a production database
(or similar fun things) 19:34
Xliff Same error after nukage
moritz: Been (somewhere near) there. Glad for backups. Egg on face.
19:34 pochi joined
Xliff But then... company didn't really have a place to do alternative. 19:34
Customer didn't have another environment. 19:35
19:35 st_elmo left
moritz who was it recently who dropped their prod database by accidentally running test scripts against it? 19:36
that was like, last week or so
19:37 st_elmo joined
geekosaur was that the aws outage? 19:37
moritz no, much more recently 19:38
geekosaur only other one I can think of recently was gitlab
moritz even more recently :-)
geekosaur uh, the aws outage I mean was a month ago 19:39
19:39 baest left
gfldex El_Che: i'm trying to teach travis to use your ubutu packages and am getting odd errors. Did you have success with travis in that regard? 19:39
geekosaur bit it wasn't a test script against a production db, it was a typo 19:40
gfldex El_Che: see line 215 in travis-ci.org/gfldex/gfldex-meta-z.../222185215
timotimo moritz: yup, never destroyed a production database like that 19:41
in fact, nobody has yet let me near a production database :)
moritz oh, it was Digital Ocean: www.digitalocean.com/company/blog/...17-outage/
19:41 st_elmo left
timotimo they just wanted to be hip and cool and in the news just like amazon 19:42
moritz " A process performing automated testing was misconfigured using production credentials."
geekosaur missed that one 19:46
moritz it was much smaller than AWS 19:47
fewer users, and only the control thingy down
timotimo Xliff: how about you go into the folders and manually go through the steps? 19:48
geekosaur "do'h"
Xliff timotimo: That was manually using Confiugre from top level rakudo dir. 19:49
timotimo yeah, but it's clearly trying to build moarvm 19:51
so go into the moarvm folder and run Configure in there
and hopefully it'll give you some more output
Xliff Updating submodules .................................... FAIL 19:52
git error: fatal: destination path '/home/cbwood/.rakudobrew/moar-nom/nqp/MoarVM/3rdparty/libtommath' already exists and is not an empty directory.
timotimo yup
that's exactly the problem
Xliff Wasn't libtommath already mentioned>?
timotimo yeah
Xliff Well piss!
geekosaur yes, that's your issue
timotimo you need to rm -rf that folder
geekosaur or git submodule update?
timotimo alternatively, write some code for us that does that automatically
geekosaur: no, it won't let you go into a state where the submodule even exists
because there's already stuff in the way
Xliff \0/ 19:53
timotimo: I would first need to know how things SHOULD be done. 19:54
So that's less me and more .... someone.
19:54 lostinfog joined
Xliff So what has life been in rakudoworld since... Feb? 19:55
I'd read bland changenotes, but why do that when I can give timotimo high blood pressure!? ^_^
timotimo oh 19:56
hm, can we even put code in that'll rm -rf that folder? 19:57
can we checkout a checkout where the submodule is configured?
potentially we can
19:57 baest joined
timotimo we need to somehow figure out the problem is "already exists and is not an empty directory", but we should also make sure any user changes don't erroneously get rm -rf'd there 19:59
i suppose it'd be enough to just mv it out of the way
Xliff When running "./Configure" can you check the output? If you see that git error then mv it out of the way and retry. 20:02
Say in big red letters "CHECK LIBTOMMATH DIR FOR LOCAL CHANGES. IF NONE EXIST YOU CAN SAFELY REMOVE" 20:03
timotimo *shrug*, it's a perl5 script 20:05
i don't do perl5
Xliff timotimo: I sense an upturned nose. ^_^ 20:08
And system command output redirection in perl5 is a bit of a chore. 20:10
I know there are modules for it, but...
timotimo no, i just don't want to write some perl5 code that goes against every single best practice out there 20:11
because i am truly 100% ignorant of how to write good perl5 code
20:20 Cabanossi left 20:21 baest left 20:22 Cabanossi joined 20:23 st_elmo joined 20:25 lostinfog left
samcv a thought i had today. if most ENV vars are all in caps. then wouldn't that make lowercase env vars the least likely to have collisions? 20:29
for personal scripts and such
20:32 baest joined
geekosaur that's the general idea, yes 20:33
except that there's an edge case: Bourne-derived shells conflate env vars with local shell vars, so you generally use lowercase for the latter and uppercase for the former 20:34
ugexe tbrowder: it uses whatever is listed for `source-url` 20:37
20:42 ChoHag left
tbrowder ugexe: great! thanks. 20:43
20:44 ChoHag joined
samcv geekosaur, what about posix shells like dash? 20:51
geekosaur that is Bourne-derived
samcv ok
geekosaur fish, rc, and csh/tcsh are shells that are not Bourne-derived
20:52 mcmillhj left, Sound joined 20:53 Sound is now known as Guest85297 20:54 Guest85297 left, Sound_ joined
gfldex lolibloggedalittle: gfldex.wordpress.com/2017/04/14/sp...up-travis/ 20:55
SmokeMachine m: say {try 5 == 5}()
camelia WARNINGS for <tmp>:
Useless use of "==" in expression "5 == 5" in sink context (line 1)
True
21:01 skids left 21:03 bjz joined 21:04 Cabanossi left 21:05 Sound_ left 21:06 baest left 21:07 Cabanossi joined 21:09 RabidGravy left 21:10 baest joined
TreyHarris samcv: another wrinkle to know about before you intentionally pick the lowercase form of an env variable you know is i use somewhere: zsh at least has the concept of autobound scalar/array pairs, so that you can do "typeset -TU PERL6LIB perl6lib ,", then $perl6lib can be manipulated with array ops while PERL6LIB will remain in its comma-delimited scalar form. (The -T does that; the -U just forces uniqueness as 21:10
you usually want with search path variables.) PATH/path, FPATH/fpath--in general, zsh uses the caps/lowercase convention for this.
I set up a -TU binding for pretty much every list variable because it's so useful. (With the uniqueness in particular, so you can just do "perl6lib+=mylib" and know it'll be added to the end iff it isn't already there.) 21:11
I don't know if other shells have the same as a general concept; most have it as a special case for PATH/path and other core library env vars. 21:12
geekosaur it's a csh-ism 21:13
$path vs. $PATH is the canonical one, hardwired into csh/tcsh; zsh generalized the mechanism 21:14
and I think otherwise only fish does it
Geth doc: aeeec94478 | (Zoffix Znet)++ | doc/Type/IO/Path.pod6
[io grant] Straighten up copy, move, rename

  - Shorten and unify prose
  - List new type constraints as implemented in
   github.com/rakudo/rakudo/commit/ff97083419
TreyHarris geekosaur: that was my guess but I wasn't going to delve into the arcane world of unfamiliar shells' manpages to try to figure it out :-) 21:16
Geth doc: 6992c483e3 | (Zoffix Znet)++ | doc/Type/IO/Path.pod6
For the love of copy-paste is the root of all evil

1 Timothy 6:10		 21:17
TreyHarris when people complain about Perl's un-googleability, Perl (5 or 6) have nothing on shells
timotimo agreed 21:26
21:27 st_elmo left 21:39 nadim left 21:41 baest left 21:44 baest joined 21:49 cpage joined 22:04 espadrine left 22:06 Cabanossi left, wamba joined 22:07 Cabanossi joined 22:11 kerframil joined
Geth doc: 923ea058b6 | (Zoffix Znet)++ | doc/Type/IO/Path.pod6
[io grant] Straighten up mkdir docs

  - Remove removed multi-dir multi[^1]
  - Direct to `mode` docs for mode info
  [1] github.com/rakudo/rakudo/commit/0d9ecae6fe		 22:12
TreyHarris I love everything about jnthn's "auto-new" trait suggestion except the name. "method table() is auto-new" just reads wrong. But I can't come up with a better one. "autovivified"? It's not quite the same as subscript autovivification directly translated into attribute autovivication, but it's not that far off.... 22:15
22:24 lucas-1 left, robertle left
gfldex TreyHarris: autoinstance 22:27
AlexDaniel TreyHarris: wait, auto-new? Where can I read about it? 22:28
jnthn Names are hard, not to mention bikesheddable :) 22:29
(And no, I don't consider auto-new a particularly nice name either :)) 22:30
It's hard to think of something relatively short but descriptive
Maybe `is creative` (as in, creates an instance, which plays OK with the Perl 6 .CREATE primitive) 22:31
timotimo we have .CREATE?
oh, that's the really low-level one?
jnthn Yeah
It's why the low-level object creation op got named nqp::create
timotimo i completely missed the suggestion for this auto-new thing
22:32 raiph joined
raiph m: sub foo-Int (Int $a) { my Int $b; $b++ for ^$a }; sub foo-int (int $a) { my int $b; $b++ for ^$a }; my $time; $time = now; foo-Int my Int $ = 1e6.Int; say now - $time; $time = now; foo-int my int $ = 1e6.Int; say now - $time; # 0.7627174 22:32
camelia 0.73760479
0.479592
timotimo gfldex: it'd be cool if you could figure out short instructions for using samcv's appimages; the star ones come with zef if i understand correctly 22:33
samcv they do come with zef but you can't really get to it any way
even though they are inside the image
timotimo hm, maybe through perl6 -MZef::Something?
samcv could even add a --zef flag to start zef instead lol 22:34
ugexe perl6 -MZef::CLI
samcv that might be broken though
timotimo fantastic! ugexe to the rescue
samcv i think module preloading is broken on relative paths
ugexe ah
samcv i can try it though
timotimo samcv: wow, all those html files are just one single long line? 22:35
samcv which?
timotimo github.com/MoarVM/MoarVM/blob/106a...uct.c.html
samcv oh hey... that worked
magic!
ugexe $ perl6 -MZef::CLI -e '' install CSV::Parser
22:35 skids joined
samcv i guess installed modules are fine to preload. i think non installed modules from $ANY_PATH_HERE don't work 22:36
since rakudo distro is relative path, somehow it confuses the precompiler
timotimo you're not running a lot of testy stuff there yet?
samcv for appimages?
it does run tests before pushing the updated appimages
timotimo no the coverage report
samcv oh
timotimo i was switching topics :P 22:37
samcv it does the nqp test
timotimo oh facepalm
samcv though you can't do commands after -MZef::CLI because it tries to resolve the paths in the perl6 script
timotimo i was looking at moar, not libmoar
samcv oh
oh hey 22:38
it worked! \o/ \o/
nice
22:38 rindolf left
samcv moarvm.github.io/MoarVM/libmoar/ fancy 22:38
timotimo very fancy indeed 22:39
22:51 Cabanossi left 22:52 Cabanossi joined 22:54 wamba left 22:55 nowan left
azawawi off to sleep q { github.com/azawawi/perl6-ncurses/t...r/examples :) } 22:57
22:58 nowan joined 23:01 azawawi left
Xliff Er.... 23:07
NativeHelpers::Blob fails, which blows DBIish install. 23:08
==> Testing NativeHelpers::Blob
t/00-trivial.t .. ok
t/01-basic.t .... ok
# Failed test 'Indeed'
# at t/02-cstruct.t line 112
# expected: '-2080500184'
# got: '2214467112'
# Looks like you failed 1 test of 35
t/02-cstruct.t ..
Dubious, test returned 1 (wstat 256, 0x100)
Failed 1/35 subtests
t/03-pointer.t .. ok
t/99-my-meta.t .. ok
gfldex timotimo: quite frankly I'm not happy with that appimage stuff. We would do much better in providing proper Rakudo Star distro packages on a regular basis, including a repo that could be added to sources.list. 23:10
mst gfldex: ... 23:11
Geth doc: 47b05262e8 | (Zoffix Znet)++ | doc/Type/IO/Path.pod6
[io grant] Explicitly spell out caveats of IO::Path.Str
mst gfldex: yes. it would be nice to have those *as well*
Xliff Nevermind. I filed a ticket with the proper repo.
23:12 labster left
mst gfldex: but that's no reason to be mad at somebody for spending their free time building something else 23:12
timotimo gfldex: well, we got the appimages now.
it serves as a package for about one gazillion different distros
gfldex mst: what makes you think I'm mad at anyone?
mst and is relocatebale and on some systems user-installable and self-contained
timotimo it only took like 4 days to build
that's about 0.001 seconds per distro 23:13
i'd say that's hard to beat
mst gfldex: "I'm not happy with X, we would do much better doing Y" is an implicit criticism of X and the work that went into it.
23:13 ChoHag left
mst gfldex: that might not have been your intention but is how such comments are likely to be read 23:13
gfldex mst: please stop
23:13 kent\n joined
mst gfldex: you asked a question. I was answering it. 23:14
23:14 riatre_ is now known as riatre
mst normally open source contributors care when they accidentally looked like they were shitting on somebody else's work 23:14
gfldex so let me rephrase then, I personally don't like nor have I ever used appimages. As a consequence of my personal preference I have to decline timotimos kind request to write about stuff I never had plans to write about. 23:16
mst: happy now?
timotimo that's fine, it just came across a bit more general than that 23:19
23:20 Cabanossi left 23:21 geekosaur left 23:22 Cabanossi joined 23:23 geekosaur joined
Xliff Huh! 23:24
gfldex zef got failing tests with bleeding rakudo. No time to hunt it down, I got $dayjob tomorrow. 23:25
TimToady something to do with link or symlink maybe 23:30
Xliff Wow! "sprintf("%p", pointer)" is returning a negative value. WTF? 23:31
TimToady that often means pointer is on the stack 23:33
many machines put the stack in "negative" memory 23:34
I suppose this is a pointer from C code calling into some kind of callback? 23:35
if not, you may have nested runloop problems... 23:36
mst gfldex: yep, that's completely reasonable, and since I agree that vendor packages would also be useful, I'd be perfectly happy to test a set if you ever come up with them
Xliff It's a NativeCall generated pointer array passed back into C code.
LinerArray[]
So the pointer is created by rakudo and passed back to a C function. 23:37
Which means I can't test the C side.
TimToady I don't see how rakudo can allocate a (valid) pointer on the C stack unless it was called into by some C code that is still live 23:38
Xliff This one test is the only thing that fails in NativeHelper::Blob
TimToady: That's what I was thinking.
gfldex mst: I strongly doubt I will do so as Perl 6 is becoming less and less enjoyable to me.
TimToady I suppose some OSes *could* be putting heap values that high, but it's not normal in my experience 23:39
Xliff Everything else works, though. Passing that same pointer to C to change values and print them out works fine.
It's just checking the "%p" value to the perl6 one that failes. 23:40
TimToady gfldex: is that because we intentionally torment the implementors a bit more than the users?
Xliff I almost think that test is no longer "valid"
TimToady gfldex: or is there something else (that is fixable)? 23:41
Xliff TimToady: See lines 33, 41 and 112, here: github.com/salortiz/NativeHelpers-...-cstruct.t 23:42
myaddr() is a C routine that simply does sprintf("%p"...) to the passed parameter. 23:43
And returns the output
ugexe m: say mkdir("foo") # this is why zef is broke 23:44
camelia mkdir is disallowed in restricted setting
in sub restricted at src/RESTRICTED.setting line 1
in sub mkdir at src/RESTRICTED.setting line 8
in block <unit> at <tmp> line 1
ugexe well that returns True now instead of the directory it created
23:45 BenGoldberg joined
TimToady there was something about that over in -dev 23:46
23:51 Cabanossi left 23:52 Cabanossi joined 23:53 baest left 23:56 labster joined 23:57 baest joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!