»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.
kalkin-_ ohh nevermind 00:00
skids kalkin-_: maybe you are just looking for a dynamically scoped variable?
kalkin-_ skids: you mean $*FOO? 00:01
skids right. Does not have to be uppercase. uppercase is usually for builtin stuff.
00:05 pierre_ joined 00:07 IRCBot1465147004 left 00:09 kurahaupo left 00:12 dalek joined, ChanServ sets mode: +v dalek 00:17 kurahaupo joined 00:20 kid51 left 00:22 kid51 joined 00:23 gorgonzola left 00:29 kurahaupo left 00:30 labster joined 00:34 raiph joined 00:37 mr-foobar joined, retupmoca joined 00:38 Herby__ joined
Herby__ Evening, everyone! 00:38
emdashcomma Hi__ 00:39
Herby__ o/
00:41 kurahaupo joined 00:42 bisectable left, bisectable joined 00:44 kurahaupo left, Zoffix joined
AlexDaniel .tell Zoffix try your .wut snippet again. I thought that only I would throw in a code that produces non-zero exit code on a good commit, but since you've done it too I decided to cover it as well 00:44
yoleaux AlexDaniel: I'll pass your message to Zoffix.
Zoffix Any idea why this program works with `echo` but not with `perl`, even though if I start the perl script alone it produces output? gist.github.com/zoffixznet/0ec1c2f...a7804d4103 00:45
yoleaux 00:44Z <AlexDaniel> Zoffix: try your .wut snippet again. I thought that only I would throw in a code that produces non-zero exit code on a good commit, but since you've done it too I decided to cover it as well
AlexDaniel mmm…
yoleaux: great job!
Zoffix AlexDaniel, no idea what .wut snippet you're talking about
AlexDaniel bisect: class A { has $.wut = [] }; my $a = A.new; $a.wut = [1,2,3]; $a.wut.say
bisectable AlexDaniel: exit code on a “good” revision is 1 (which is bad), bisecting with inverted logic
AlexDaniel: (2016-03-02) github.com/rakudo/rakudo/commit/fdd37a9
Zoffix It wasn't mine. Don't care about :/ 00:46
retupmoca .botsnack
synopsebot6 om nom nom
yoleaux :D
Zoffix Pissed off that I'm constantly fighting with things that aren't primary code writing. Like this proc stuff
AlexDaniel bisect: for ‘q b c d’.words -> $a, $b { }
bisectable AlexDaniel: exit code on a “good” revision is 1 (which is bad), bisecting with inverted logic
AlexDaniel: (2016-03-01) github.com/rakudo/rakudo/commit/1b6c901
AlexDaniel but look how great it is now! 00:47
Zoffix The docs don't say how the .tap'ed stdout is chunked. Is it by new lines? Does it buffer and waits? *sigh*
00:49 Herby__ left
AlexDaniel Zoffix: it may be the buffering in 01-basic.pl itself 00:49
Zoffix It prints fine when I run it from command line. 00:50
I tried now with a small perl script that prints a few lines and sleeps and I'm getting nothing from the tap
AlexDaniel Zoffix: what if you run it piped to | cat ?
geekosaur remember that programs buffer output not to terminals, usually in 4k or 8k chunks
on linux you can use stdbuf to defeat this 00:51
AlexDaniel or just turn on autoflush
in 01-basic.pl
geekosaur oh, yeh, probably doesn't defeat perl's since it's doing its own buffering (p5 via the perlio stuff, rakudo via libuv) 00:52
whereas stdbuf uses a backchannel into glibc's buffering code
AlexDaniel somehow all buffering problems I experienced were associated with perl5… hmm… 00:53
sometimes it feels like it is overdoing it
Zoffix AlexDaniel++ thanks, adding $|++ fixed it 00:54
Perl5--
AlexDaniel a bunch of code is not commited… :P gonna get some sleep 00:58
01:05 AlexDaniel left 01:09 ssotka left 01:11 aborazmeh joined, aborazmeh left, aborazmeh joined
BenGoldberg m: say &x 01:13
camelia rakudo-moar 85682a: OUTPUT«5===SORRY!5=== Error while compiling /tmp/hwvFiF6jOt␤Undeclared routine:␤ x used at line 1␤␤»
BenGoldberg supposes that the x operator is a syntactically wossname, not a subroutine. 01:14
m: (say 2) xx 1;
camelia rakudo-moar 85682a: OUTPUT«2␤» 01:15
BenGoldberg m: (say 2) xx 0;
camelia ( no output )
geekosaur m: say &infix:<xx>
camelia rakudo-moar 85682a: OUTPUT«sub infix:<xx> (Mu $, $, *%) { #`(Sub+{Precedence}|43087568) ... }␤»
01:15 finanalyst joined
BenGoldberg m: sub postfix:<twice> (Mu $lhs) { lhs xx 2 }; (say 'a') twice; 01:17
camelia rakudo-moar 85682a: OUTPUT«5===SORRY!5=== Error while compiling /tmp/IpmyXCbgE8␤Two terms in a row␤at /tmp/IpmyXCbgE8:1␤------> 3twice> (Mu $lhs) { lhs xx 2 }; (say 'a')7⏏5 twice;␤ expecting any of:␤ infix␤ infix stopper␤ statement end…»
BenGoldberg m: sub postfix:<twice> (Mu $lhs) { $lhs xx 2 }; (say 'a') twice;
camelia rakudo-moar 85682a: OUTPUT«5===SORRY!5=== Error while compiling /tmp/4fVDA6M5WM␤Two terms in a row␤at /tmp/4fVDA6M5WM:1␤------> 3wice> (Mu $lhs) { $lhs xx 2 }; (say 'a')7⏏5 twice;␤ expecting any of:␤ infix␤ infix stopper␤ statement end…»
geekosaur m: sub postfix:<twice> (Mu $lhs) { $lhs xx 2 }; (say 'a')twice;
camelia rakudo-moar 85682a: OUTPUT«a␤»
geekosaur m: sub postfix:<twice> (Mu $lhs) { $lhs xx 2 }; (say 'a')\ twice;
camelia rakudo-moar 85682a: OUTPUT«a␤»
geekosaur it's a bit peculiar that way 01:18
Zoffix You're not using the result of of `say`
BenGoldberg Ok, but now how do I make it actually run it's argument twice, the way xx does?
Zoffix It doesn't 01:19
timotimo not possible without a slang (or macros) at the moment, as far as i know
Zoffix m: (say 42) xx 1;
camelia rakudo-moar 85682a: OUTPUT«42␤»
Zoffix m: (say 42) xx 2;
camelia rakudo-moar 85682a: OUTPUT«42␤42␤»
Zoffix never miknd
geekosaur xx likely gets to cheat. user subs don't get to do that, they always receive something already evaluated 01:20
BenGoldberg m: macro postfix:<twice> (Mu $lhs) { quasi { {{{ $lhs }}} x 2 } }; (say 'a')twice;
camelia rakudo-moar 85682a: OUTPUT«5===SORRY!5=== Error while compiling /tmp/sVLyzPmmGU␤Use of macros is experimental; please 'use experimental :macros'␤at /tmp/sVLyzPmmGU:1␤------> 3macro7⏏5 postfix:<twice> (Mu $lhs) { quasi { {{{␤»
gfldex is infix:<xx> meant to be eager? 01:21
BenGoldberg m: use experimental :macros; macro postfix:<twice> (Mu $lhs) { quasi { {{{ $lhs }}} x 2 } }; (say 'a')twice;
camelia rakudo-moar 85682a: OUTPUT«Weird node visited: QAST::BVal␤a␤»
BenGoldberg m: use experimental :macros; macro postfix:<twice> (Mu $lhs) { quasi { {{{ $lhs }}} xx 2 } }; (say 'a')twice;
camelia rakudo-moar 85682a: OUTPUT«5===SORRY!5=== Error while compiling /tmp/mtWb0FEkTA␤Unknown QAST node type QAST::Unquote␤at /tmp/mtWb0FEkTA:1␤»
gfldex m: my @a := (say 42) xx 2; dd @a;
camelia rakudo-moar 85682a: OUTPUT«42␤42␤(Bool::True, Bool::True)␤»
gfldex m: my @a := (say 42) xx 2; say 'ello'; dd @a; 01:22
camelia rakudo-moar 85682a: OUTPUT«42␤42␤ello␤(Bool::True, Bool::True)␤»
01:22 araujo__ joined, kalkin- joined 01:25 molaf left
BenGoldberg does not yet sufficiently grok macros. 01:25
gfldex neither does Rakudo :)
01:25 araujo_ left 01:30 araujo_ joined 01:35 araujo__ left, colomon left
timotimo that.s right 01:35
01:36 araujo__ joined 01:37 molaf joined 01:39 araujo_ left 01:43 aborazmeh left, Actualeyes joined 01:47 ilbot3 joined, colomon joined 01:49 colomon left 01:53 kid51 left, FROGGS_ joined 01:56 FROGGS left
Zoffix Well, I'm gonna disappear for a week+. Burning out.... I do remember that I'm meant to release Rakudo on 18th and I will. So, don't worry, I remember :) 02:16
\o
02:16 Zoffix left 02:23 colomon joined, mr-foobar left 02:25 mr-foobar joined 02:26 pierre_ left 02:30 user9 joined 02:34 pierre_ joined 02:37 Zoffix joined 02:38 colomon left, colomon joined, noganex joined
Zoffix This will sound weird, but... My brain just won't shut up that my .seen message would be the wave and not the actual last message, so... 02:38
Well, I'm gonna disappear for a week+. Burning out.... I do remember that I'm meant to release Rakudo on 18th and I will. So, don't worry, I remember :)
02:38 Zoffix left 02:42 gorgonzola joined, colomon left, colomon joined 02:49 user9 left 02:54 huggable left 03:01 gorgonzola left 03:06 finanalyst left 03:12 khw left 03:13 yqt left 03:37 TEttinger joined 03:39 khw joined 03:47 finanalyst joined 04:07 ssotka joined 04:14 xinming joined 04:22 khw left 04:25 Cabanossi joined 04:45 pierre_ left 04:48 pierre_ joined 04:51 pierre_ left 04:52 pierre_ joined 05:01 sno joined 05:02 BenGoldberg left 05:03 araujo__ left 05:09 _nadim left 05:10 _nadim joined 05:18 user9 joined 05:22 user9 left 05:43 sno left 05:52 FROGGS_ left 05:57 mr-foobar left 06:01 salv0 left 06:18 girafe joined 06:19 pierre_ left 06:24 jrusso joined, firstdayonthejob joined, jrusso is now known as MadcapJake 06:26 pierre_ joined 06:28 pierre_ left, pierre_ joined 06:37 mtj_ joined 06:39 pierre_ left 06:40 firstdayonthejob left 06:41 pierre_ joined 06:48 aries_liuxueyang left 06:50 aries_liuxueyang joined 06:54 huggable joined 06:55 sno joined 06:58 RabidGravy joined 07:01 darutoko joined
RabidGravy BOOM! 07:08
07:11 ssotka left 07:21 pierre_ left 07:22 girafe left, rurban joined 07:24 zakharyas joined 07:41 bjz joined 07:45 bjz left 07:49 g4 joined 07:50 rindolf joined 07:51 Humdai joined
RabidGravy I just have to say that among all the cool and fabulous things in Perl 6, the combination multi of dispatch and the richness of signature declaration is racing ahead as my favourite thing 07:53
s/multi of/of multi/
moritz that is indeed a very powerful combination 07:54
and it nudges you in the direction of writing small routines
07:55 bjz joined
RabidGravy yeah, for a long while now I have aspired to making routines with the minimum possible complexity and removing all the logic surrounding determining what the arguments are and branching accordingly to a langiage feature leaps forward into that 07:59
08:01 ocbtec joined, user9 joined 08:08 dakkar joined 08:10 azawawi joined
azawawi hi 08:10
RabidGravy: github.com/azawawi/gtk-simple/blob...ar.pl6#L32 # Feedback please :)
RabidGravy: github.com/azawawi/gtk-simple/blob...ox.pm6#L13 08:11
pasteboard.co/1tyObr6M.png # Screenshot of the above example after GTK::Simple::Box.new change 08:12
RabidGravy looks good matey 08:14
if it passes the tests and the rest of the examples continue to work as before then it's good to go :) 08:15
08:16 cosimo joined 08:24 wamba joined 08:25 sdsdfsd joined
azawawi installs homebrew on macbook pro :) 08:25
DrForr RabidGravy: I might have a working wiki to show @ YAPC, I did most of the heavy lifting last night. 08:29
RabidGravy cool, WordPress replacement next ;-) 08:31
DrForr++ # making stuff
DrForr Oh, I've got *other* plans as well, they might earn a LT mention if I can get the REPL module to behave the way I want it.
RabidGravy I've been sidetracked for days trying to make something that does what SQL::Abstract does 08:32
08:33 araujo joined, sdsdfsd left
DrForr I almost let myself get distracted by factoring out a MediaWiki grammar before realizing what a headache that actually was. 08:34
08:36 zakharyas left
gregf_ morning Perl6! 08:37
DrForr Mornin'.
08:38 zakharyas joined
RabidGravy MARNIN 08:39
right, I'm going to use this window of fine weather opportunity to do some stuff in the garden 08:40
literal m: time 08:44
camelia ( no output )
08:44 araujo_ joined
literal m: say time 08:44
camelia rakudo-moar 85682a: OUTPUT«1465202659␤»
literal how do I get the epoch time in milliseconds?
moritz m: say now 08:45
camelia rakudo-moar 85682a: OUTPUT«Instant:1465202791.008717␤»
moritz m: say now.tai 08:46
camelia rakudo-moar 85682a: OUTPUT«1465202818.009479␤»
08:47 ufobat joined, araujo left
literal I see 08:51
Woodi Zoffix: gl with resting! don't worry, be truly yourself :) 08:54
ufobat morning :)
Woodi hallo :) 08:56
08:59 azawawi left, araujo_ left
arnsholt DrForr: IIRC, MediaWiki's language is implemented as a cascade of regex search-replace routines, which results a context-*sensitive* language 09:12
So implementing a grammar for it is likely quite hard 09:13
09:13 finanalyst left
arnsholt Besides, you want MarkDown ;) 09:13
09:13 xinming left, iH2O joined 09:14 xinming joined
dogbert17 o/ #perl6 09:16
iH2O hi Dagobert
dogbert17 ashello iH20
s/ashello/hello/
iH2O is the final / really needed? 09:17
dogbert17 iH20, I always put it there for some reason :) 09:18
jast in perl? yes. in sed? yes. in IRC? depends on who is parsing your expression ;)
dogbert17 :)
does anyone have a minute to review my attempt to document List.categorize at gist.github.com/dogbert17/c0b286f9...f431f22d30 ? 09:20
arnsholt It's optional in vi, as long as you don't need any modifiers =)
09:21 bjz_ joined 09:22 bjz left
jast dogbert17: I found the first paragraph a little hard to follow, here's my attempt to simplify: gist.github.com/jast/45e6533056692...a0e83b62df 09:30
might be a matter of taste, I don't know 09:31
dogbert17 jast: I like it an will change :)
*and
running it through htmlify.p6 atm 09:32
09:39 rindolf left 09:44 g4 left
dalek c: 0937200 | (Jan-Olof Hendig)++ | doc/Type/List.pod:
Added docs for List.categorize. jast++ for making the text easier to understand
 09:45
09:46 rindolf joined 09:48 iH2O left 10:11 Hor|zon joined 10:23 kaare_ joined 10:25 labster left 10:27 jack_rabbit joined, kid51 joined 10:45 mr-foobar joined 10:46 DHAD05 joined, DHAD05 left 10:47 TEttinger left 10:50 araujo joined 10:54 huggable left, huggable joined
literal m: say "abc".split("", :skip-empty) 11:11
camelia rakudo-moar c4b3fa: OUTPUT«(a b c)␤»
literal is there a shorter way of splitting a string into characters?
lizmat m: dd "abc".comb 11:12
camelia rakudo-moar c4b3fa: OUTPUT«("a", "b", "c").Seq␤»
literal ah, nice 11:13
lizmat literal: in Perl 6, you usually want to use comb instead of split
literal alright 11:14
11:22 rurban1 joined 11:29 rurban left
jast I used comb in one of my very first Perl 6 programs: say [Z](Q%`sd`Kbmsotj7!nrd!Quusidi!%.comb.map({chr(ord($_)+^1)}).rotor(5)).reverse>>.join.join 11:31
11:31 rurban joined
moritz m: say [Z](Q%`sd`Kbmsotj7!nrd!Quusidi!%.comb.map({chr(ord($_)+^1)}).rotor(5)).reverse>>.join.join 11:32
camelia rakudo-moar a64607: OUTPUT«Just another Perl6 hacker␤»
moritz starting with an obfu, you fit right in here :-) 11:33
jast I wanted to play with the fancy operators as soon as possible :}
moritz m: say [Z](Q%`sd`Kbmsotj7!nrd!Quusidi!%.comb.map({chr(ord($_)+^1)}).rotor(5)).reverse>>.join.join
camelia rakudo-moar a64607: OUTPUT«Just another Perl6 hacker␤»
moritz m: say [Z](Q%`sd`Kbmsotj7!nrd!Quusidi!%.comb.map({chr(.ord+^1)}).rotor(5)).reverse>>.join.join
camelia rakudo-moar a64607: OUTPUT«Just another Perl6 hacker␤»
moritz m: say [Z](Q%`sd`Kbmsotj7!nrd!Quusidi!%.comb.map((*.ord+^1).chr).rotor(5)).reverse>>.join.join 11:34
camelia rakudo-moar a64607: OUTPUT«Just another Perl6 hacker␤»
jast dang :)
moritz m: say [Z](Q%`sd`Kbmsotj7!nrd!Quusidi!%.comb.map((*.ord+^1).chr).rotor(5)).reverse.flat.join
camelia rakudo-moar a64607: OUTPUT«Just another Perl6 hacker␤»
moritz m: say [Z](Q%`sd`Kbmsotj7!nrd!Quusidi!%.ords.map((*+^1).chr).rotor(5)).reverse.flat.join 11:35
camelia rakudo-moar a64607: OUTPUT«Just another Perl6 hacker␤»
jast no fair, I didn't know about that function ;)
moritz jast: it's not meant as competition, just as playing (and potentially education)
jast (btw I used Q% instead of ' because it looks stranger, even though it's slightly longer) 11:36
yeah, sure, it's all good, clean fun :)
moritz m: say [Z](<`sd`Kbmsotj7!nrd!Quusidi!>.ords.map((*+^1).chr).rotor(5)).reverse.flat.join 11:37
camelia rakudo-moar a64607: OUTPUT«Just another Perl6 hacker␤»
moritz 'nother option
jast right. I still prefer Q%...% because fewer people would think of it as matching pairs
obfuscation is applied psychology, right? :) 11:38
moritz aye 11:40
I've never been good at that :-)
11:41 kid51 left 11:48 dontaki joined 11:53 samb1 joined 11:54 rurban1 left, rurban1 joined, cessna joined 12:01 rurban1 left 12:02 dontaki left 12:08 zakharyas left
gregf_ m: ("!".ord-1).chr 12:11
camelia ( no output )
gregf_ m: ("`".ord-1).chr 12:12
camelia ( no output )
gregf_ m: say ("`".ord-1).chr
camelia rakudo-moar a64607: OUTPUT«_␤»
dogbert17 m: my $i = 1; say ($i %% 2 ?? "even" !! "odd", "prime" if $i.is-prime) # stupid question, why doesn't this return a one element list? 12:13
camelia rakudo-moar a64607: OUTPUT«()␤»
gregf_ that post if messes things up i guess 12:17
or rather, gets the parser confused :|
m: my $i = 1; say $i.is-prime *smiles* 12:19
camelia rakudo-moar a64607: OUTPUT«5===SORRY!5=== Error while compiling /tmp/3MLiAeGcOW␤Missing required term after infix␤at /tmp/3MLiAeGcOW:1␤------> 3my $i = 1; say $i.is-prime *smiles*7⏏5<EOL>␤ expecting any of:␤ prefix␤ term␤»
gregf_ m: my $i = 1; say $i.is-prime #*smiles*
camelia rakudo-moar a64607: OUTPUT«False␤»
psch m: my $i = 1; say ($i %% 2 ?? "even" !! "odd", ("prime" if $i.is-prime)) #
camelia rakudo-moar a64607: OUTPUT«(odd)␤»
psch m: my $i = 3; say ($i %% 2 ?? "even" !! "odd", ("prime" if $i.is-prime)) #
camelia rakudo-moar a64607: OUTPUT«(odd prime)␤»
gregf_ m: (^10).map(*.is-prime.say) 12:20
camelia rakudo-moar a64607: OUTPUT«False␤False␤True␤True␤False␤True␤False␤True␤False␤False␤»
dogbert17 psch: and why does that work?
psch dogbert17: parenthesis change precedence
dogbert17: statement_mod if modifies the whole statement
dogbert17 so gregf_ was right the post if messes things up
psch dogbert17: infix:<,> is part of that statement
dogbert17: it doesn't "mess anything up", it does exactly what it's suppose to do :)
gregf_ dogbert17: no, prime says 1 is not a prime. my $i = 3; say ($i %% 2 ?? "even" !! "odd", "prime" if $i.is-prime) # run this 12:21
dogbert17 psch: ah, thanks for the explanation, it sure messed with my brain :) 12:22
psch m: my $i = 1; say (do if $i.is-prime { $i %% 2 ?? "even" !! "odd", "prime" }) # without parens, it's equivalent to this 12:23
camelia rakudo-moar a64607: OUTPUT«()␤»
psch m: my $i = 1; say ($i %% 2 ?? "even" !! "odd", do if $i.is-prime { "prime" }) # with parens, it's equivalent to this
camelia rakudo-moar a64607: OUTPUT«(odd)␤»
dogbert17 psch: that definitely makes things clearer 12:24
gregf_: indeed this 'odd' problem showed itself when is-prime returned false
12:25 perlpilot joined
dogbert17 psch: do you remember discussing an async bug a few days ago, an example running 'hexdump'? 12:25
pscg: and a missing error message 12:26
psch dogbert17: i remember talking about it, yeah. not sure what - if any - came of it though
dogbert17 psch: nothing came of it, should we report it, perhaps it interests jnthn?
psch ...i thought that was already ticketed? 12:27
dogbert17 psch: not by me, but perhaps it had been reported earlier
psch in any case, yes, if there is no ticket there definitely should be 12:28
dogbert17 psch: how would I search for it in RT
psch dogbert17: i usually use google, with "site:rt.perl.org" and well, keywords i hope that match :) 12:29
the rt search isn't bad, but a bit unwieldy 12:30
dogbert17 psch: thx, I should have figured that out myself :(
12:32 rindolf left
dogbert17 can't find anything on RT so either it isn't there or my keyword(s) were bad 12:33
12:35 Possum joined 12:39 sammers left
dogbert17 ok, I'll try to rakudobug it then 12:39
12:47 sufrostico joined 12:54 brrt joined, Sgeo_ left 12:56 zakharyas joined 13:09 Emeric joined 13:10 yqt joined 13:16 rurban2 joined, azawawi joined
azawawi hi 13:16
masak \o 13:18
brrt good hi masak azawawi
masak hìíìí 13:19
azawawi github.com/perl6/gtk-simple/pull/46
مرحبا # hi in Arabic :) 13:20
masak irssi shows it all backward :) 13:23
masak .oO( íìíìh )
azawawi m: say "مرحبا".reverse; 13:24
camelia rakudo-moar a64607: OUTPUT«(مرحبا)␤»
psch looks the same in irssi as on the clog for me
...and reverse *still* looks the same
masak m: say "مرحبا".flip 13:27
camelia rakudo-moar a64607: OUTPUT«ابحرم␤»
azawawi i was searching for that .flip :)
masak :)
psch i find it really hard to tell where one letter begins and another ends 13:28
masak .oO( always useful to be flippant on channel )
psch but i suppose that'd be the same for cursive for anyone who never learned the latin alphabet
masak psch: that's how you know you're not seeing it reversed ;)
psch masak: well, it's both - flipped and non-flipped - that look very... connected, for lack of a better word :) 13:29
azawawi if you need to specify unconnected arabic letters, you can use en.wikipedia.org/wiki/Arabic_Prese...on_Forms-B btw 13:30
13:30 cdg joined
azawawi connected = line on the left/right 13:30
en.wikipedia.org/wiki/Arabic_script_in_Unicode # to be exact all blocks 13:31
13:31 skids joined 13:36 rurban left, stevieb joined
azawawi pasteboard.co/1tUpg8C8.png # GTK::Scintilla Editor Demo :) 13:42
azawawi starts wrapping developer.gnome.org/gtk3/stable/Gt...utton.html
13:44 brrt left 13:45 brrt joined
dalek osystem: f923b67 | titsuki++ | META.list:
Add Algorithm::NaiveBayes to ecosystem

See github.com/titsuki/p6-Algorithm-NaiveBayes		 13:55
osystem: 03ddc2c | titsuki++ | META.list:
Merge pull request #216 from titsuki/add-naive

Add Algorithm::NaiveBayes to ecosystem
13:59 girafe joined
azawawi any feedback on github.com/perl6/gtk-simple/pull/46 ? 14:01
timotimo doesn't look bad at first glance 14:06
azawawi going to merge it then since it provides the ability to do this pasteboard.co/1tUpg8C8.png :) 14:10
dalek k-simple: 611a180 | azawawi++ | lib/GTK/Simple/Box.pm6:
Introduce the idea of an array of hashes or objects to Box.new(@packees)
 14:12
k-simple: f9cc80a | azawawi++ | / (2 files):
Use the new Box expand attribute
timotimo yeah
k-simple: e17ffa9 | azawawi++ | / (2 files):
More typing
k-simple: ea78db1 | azawawi++ | / (2 files):
Merge pull request #46 from azawawi/master

Expose more Box API to fine tune layout
14:17 khw joined 14:26 pullphinger joined 14:28 rindolf joined
azawawi github.com/perl6/gtk-simple/pull/47 # File Chooser Button :) 14:46
masak m: my ($x1, $x2) = 5, 3; $x1 = .min, $x2 = .max given minmax $x1, $x2; say ($x1, $x2).perl
camelia rakudo-moar bf099b: OUTPUT«(3, 5)␤»
azawawi pasteboard.co/1tYxawu0.png # screenshot
masak is there a nicer way to do the above?
14:46 AlexDaniel joined 14:47 sherwool joined
timotimo masak: yes! 14:47
m: my ($x1, $x2) = 5, 3; ($x1, $x2) .= sort
camelia ( no output )
arnsholt .=sort perhaps?
timotimo m: my ($x1, $x2) = 5, 3; ($x1, $x2) .= sort; say ($x1, $x2)
camelia rakudo-moar bf099b: OUTPUT«(3 5)␤»
arnsholt Yeah, that =) 14:48
timotimo o/⁵o arnsholt
brrt oh, wow, that is nothing if not awesome
and evil
masak ooh, .sort
Juerd .= on a list of lvalues... oh my. 14:49
brrt indeed
masak yeah, that's magical
Perl 6 is fairly alone in doing that kind of thing
gregf_ m;say 5,4,100 .=min 14:50
m: say 5,4,100 .=min
camelia rakudo-moar bf099b: OUTPUT«Cannot modify an immutable Int␤ in block <unit> at /tmp/dZoFm_Ca7o line 1␤␤»
jnthn First class l-values are both awesome and such a PITA as an implementer I'm not sure I'd ever include them in a language I designed :P
yoleaux 5 Jun 2016 21:44Z <AlexDaniel> jnthn: Here is the ticket: rt.perl.org/Ticket/Display.html?id=128325 (gnull++)
brrt i suppose an array doesn't do that
masak gregf_: need to store it in variables first
gregf_ hmm, lists are immutable
AlexDaniel .
masak jnthn: I wonder if the same effect could be had through just-sweet-enough syntactic sugar
arnsholt I hadn't thought of doing it that way, actually
jnthn masak: Not sure :)
Juerd m: my $foo = "foo"; my $bar = "bar"; ($foo.substr-rw(1), $bar) .= sort; say $foo; say $bar
camelia rakudo-moar bf099b: OUTPUT«fbar␤oo␤»
timotimo haha 14:51
masak m: my ($a, $b, $c) = 4, 100, 5; ($a, $b, $c) .= sort; say ($a, $b, $c).perl
camelia rakudo-moar bf099b: OUTPUT«(4, 5, 100)␤»
brrt m: my ($x1, $x2) = 5,3; [$x1,$x2] .= sort; say $x1, $x2;
camelia rakudo-moar bf099b: OUTPUT«53␤»
brrt thought as much
masak brrt: no, because the array has already de-<mumble>d the lvalues
docont'd 14:52
decont'd*
Juerd Are there any more interesting lvalue methods like substr-rw?
timotimo yeah, array is a list of scalar containers. unless you bind the originals, you'll end up with new ones
Juerd: just grep the core setting for Proxy.new :)
literal m: class Foo { has $.val handles <AT-POS EXISTS-POS elems end iterator>; }; my $f = Foo.new(val => [5,6]); say $f[1], $f[*-1] 14:53
camelia rakudo-moar bf099b: OUTPUT«65␤»
literal am I missing some method in 'handles', since it's getting the wrong size of the array for '*' ? 14:54
arnsholt No, it's just that *-1 is the last but one element
brrt m: my @a = (1,2,3); @a .= sort; say @a;
camelia rakudo-moar bf099b: OUTPUT«[1 2 3]␤»
brrt eh...
literal m: my $f = [1,2]; say $f[1], $f[*-1]
camelia rakudo-moar bf099b: OUTPUT«22␤»
14:54 huggable left
literal arnsholt: see --^ 14:54
brrt m: my @a = (3,2,1); @a .= sort; say @a;
camelia rakudo-moar bf099b: OUTPUT«[1 2 3]␤»
14:54 huggable joined
brrt m: my @a = (3,2,1); say @a.WHAT; 14:55
camelia rakudo-moar bf099b: OUTPUT«(Array)␤»
arnsholt Or wait, am I confused again? Could be
I thought @arr[*] was the last element
brrt m: my @a := (3,2,1); say @a.WHAT;
camelia rakudo-moar bf099b: OUTPUT«(List)␤»
timotimo no, @arr[*] is the last-plus-one element 14:56
arnsholt Right, I was confused
In that case it's a weird thing 14:57
perlpilot sounds like everyone is confused to me.
m: my @foo = 1,2,3; say @foo[*];
camelia rakudo-moar bf099b: OUTPUT«(1 2 3)␤»
arnsholt m: class Foo { has $.val handles *; }; my $f = Foo.new(val => [5,6]); say $f[1], $f[*-1] # I wonder
camelia rakudo-moar bf099b: OUTPUT«Index out of range. Is: 1, should be in 0..0␤ in block <unit> at /tmp/Qycq4lTWvE line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/Qycq4lTWvE line 1␤␤»
arnsholt Oh, no that won't work 14:58
That'll eat .new too
timotimo hehehe 14:59
moritz and .val, for that matter
psch m: class A { has $.val handles * }; say A.new(:2val).perl
camelia rakudo-moar bf099b: OUTPUT«A.new(val => 2)␤»
timotimo hm, interesting. 15:00
psch afaik it doesn't do anything currently
m: say *()
camelia rakudo-moar bf099b: OUTPUT«No such method 'CALL-ME' for invocant of type 'Whatever'␤ in block <unit> at /tmp/V6oqL2AqBL line 1␤␤»
psch $target.set_handles($thunk()); # $thunk is Whatever, in this case
15:01 get_drunk joined
literal doc.perl6.org/language/subscripts ==> "In other words, @alphabet[*-1] becomes @alphabet[@alphabet.elems - 1]." 15:01
get_drunk good morning smart people
perlpilot get_drunk: don't start asking too many random questions here :)
literal since "elems" is in "handles", why doesn't it work?
get_drunk that is a premature assumption
psch so i'd think that the handles call just gets Whatever-curried and not actually applied..? 15:02
literal but it is applied. $f.elems returns the correct value 15:03
psch m: class A { has $.val handles * }; say A.new(:2val).^attributes[0].handles
camelia rakudo-moar bf099b: OUTPUT«*␤»
literal psch: I wasn't using "handles *", that's just something arnsholt mentioned
psch m: class A { has $.val handles * }; say A.new(:2val).succ 15:05
camelia rakudo-moar bf099b: OUTPUT«3␤»
literal m: class Foo { has $.val handles <AT-POS EXISTS-POS elems end iterator>; }; my $f = Foo.new(val => [5,6]); say "elems: {$f.elems}, second: $f[1], last: $f[*-1]"
camelia rakudo-moar bf099b: OUTPUT«elems: 2, second: 6, last: 5␤»
15:05 ssotka joined
literal second and last should be the same there 15:05
llfourn I think it's because it's [*-1] is calling SELF[^SELF.cache.elems]; 15:07
not .elems
see array_slice.pm
psch m: class Foo { has $.val handles <cache AT-POS EXISTS-POS elems end iterator>; }; my $f = Foo.new(val => [5,6]); say "elems: {$f.elems}, second: $f[1], last: $f[*-1]" 15:08
camelia rakudo-moar bf099b: OUTPUT«elems: 2, second: 6, last: 6␤»
llfourn or is that only for slices?
psch well, anyway, 'handles *' (and 'handles **') does a lot currently actually
llfourn nvm that's just for slicing with *, but it might be the same problem 15:09
psch i'll blame my statements about that yesterday on being too tired to read the source properly then vOv
same for my statement just now, which was informed by my insufficient reading yesterday :S
llfourn m: class Foo { has $.val handles <cache AT-POS EXISTS-POS elems end iterator>; }; my $f = Foo.new(val => [5,6]); say $f[1], $f[*-1] # yup it's because it calls .cache.elems internally 15:11
camelia rakudo-moar bf099b: OUTPUT«66␤»
llfourn kinda annoying
psch it probably would be nice to have something convenient hanging around that makes delegating all Positional (or Associative, for that matter) methods to an Attribute 15:15
llfourn I agree.
psch not sure i have a sane way to do that though - someone did something with .^methods and grep a few days back, but i don't recall the specifics
llfourn then we could have spectests to make sure that at least works
that was me I think 15:16
but it wouldn't have worked I rkn
m: say ?Array.^methods.first(*.name eq "cache") 15:17
camelia rakudo-moar bf099b: OUTPUT«False␤»
llfourn m: say ?Array.^methods(:all).first(*.name eq "cache")
camelia rakudo-moar bf099b: OUTPUT«True␤»
llfourn maybe if you did that :S
but you are starting to get into evil territory with that
psch m: say Array.^methods(:all).grep(* !(elem) Mu.^methods)>>.name # not quite it... 15:18
camelia rakudo-moar bf099b: OUTPUT«(iterator from-iterator new is-lazy STORE reification-target shape pop shift splice name of default dynamic flat List AT-POS ASSIGN-POS BIND-POS DELETE-POS push append unshift prepend perl gist WHICH iterator STORE pick Slip perl fmt unshift item EXISTS-PO…»
llfourn I'm going to RT this because it would probably be easy to implement and would be a great feature.
b2gills perlpilot: Your last comment could easily be taken the wrong way by people new to #perl6. ( In fact it seems as if it was )
psch well, the easiest thing i could think of is just something like Rakudo::Internals::ListOfPositionalMethods 15:19
although it'd probably be in the wrong spot in Rakudo::Internals
llfourn there should simply be a role which is a coathanger for all positionaly methods 15:20
perlpilot looks to see what his last comment was ...
llfourn then you can do handles ThatRole;
psch llfourn: i don't think that's particularly sensible, because then you'd have to stuff Iterator in there too
llfourn: or you don't get cache
wait, Iterable i think actually? 15:21
perlpilot b2gills: oh, I was just giving get_drunk a hard time because he was asking questions about P6 in #perl before he came here.
llfourn I think the implementors will have to worry about that :)
what about $.a handles @;
get_drunk it was a joke b2gills
about chat in another channel
so it sorta seems bad but it is not in the context that it was meant in
llfourn but that probably won't be ok because handles is just a trait hmmm 15:22
psch llfourn: as an (on and off...) implementor, i don't think compromising internal interface sanity for a feature is a good idea :)
FSVO "internal" - Iterator and Iterable can be implemented outside of CORE...
oh, but maybe i'm thinking bad again
psch steps away from hard programming thoughts 15:23
perlpilot get_drunk: though ... you were strangely silent here after all that activity in #perl ...
psch it's definitely too warm for that o/
llfourn the role would be purely superficial just there a way of tagging methods that make postcircumfixes work as intended
get_drunk i am observing
llfourn I don't know how it could make things more complex interanally?
perlpilot get_drunk: you didn't even re-try your "wildly successful" question, which might have gotten some different answers here (though I doubt it :) 15:24
get_drunk this is my wildly cautious side
i feel like I should learn both though
15:25 ChanServ sets mode: +o mst
llfourn anyway I'll RT the general problem and this convo can be continued there if need be :) 15:25
15:26 mst sets mode: -q *!*@90-156-117-46.internetia.net.pl
mst whyyyyy do we have a quiet for lambdabot2 15:26
when there's not been one in here for ... aaages 15:27
15:27 mst sets mode: -q lambdabot2!*@*, mst sets mode: -o mst
mst I'm going to assume that wasn't helping anything 15:27
perlpilot mst: because no one has been keeping up with channel maintenance in that regard?
mst (also that first -q was buharin; apparently I did remember)
dalek k-simple: 66bf155 | azawawi++ | / (4 files):
Add File Chooser Button
k-simple: 063388c | azawawi++ | / (4 files):
Merge pull request #47 from azawawi/master

Add File Chooser Button
mst perlpilot: I was just looking while I was there 15:28
argh, and all our bans are timed by netsplat 15:29
b2gills perlpilot get_drunk: Fair enough. Just trying to keep this an inclusive community, I see that it was a false positive ( or is that false negative, whatever ) 15:32
get_drunk ok 15:33
mst b2gills: he did ask a metric fucktonne of random questions in #perl which I think triggered the comment
b2gills At least it wasn't an Imperial fuckton 15:34
get_drunk i only ever asked two questions 15:35
gregf_ m: 2 == 'metric ****tonne' 15:37
camelia rakudo-moar bf099b: OUTPUT«WARNINGS for /tmp/oGbr3bcjD1:␤Useless use of "==" in expression "2 == 'metric ****tonne'" in sink context (line 1)␤Cannot convert string to number: base-10 number must begin with valid digits or '.' in '3⏏5metric ****tonne' (indicated by ⏏)␤ …»
gregf_ get_drunk: what were your questions *if i may ask* that did cause an uproar? *sorry, dont want to flood this channel* 15:41
get_drunk they were dumb
gregf_ you'll proly get 2 to-the-point answers on here ;)
or more.. lol 15:42
mst there wasn't an uproar, they were just random 15:44
wtf is it with everybody assuming everything was more drama than it aactually was
literal m: my $f = Q{b\ar}; $f .= trans(Q{\} => Q{\\}); say $f 15:45
camelia rakudo-moar bf099b: OUTPUT«b\ar␤»
mst b2gills, gregf, please calm down and stop imagining drama where none exists so we can get back to our regularly scheduled scaring newbies with awesome language features
gregf_ quotemeta?
jast mst: but drama is so much fun :(((
get_drunk agreed
literal m: my $f = Q{b\ar}; $f .= trans(Q{\} => Q{_}); say $f
camelia rakudo-moar bf099b: OUTPUT«b_ar␤»
literal how come \\ doesn't work as a replacement? 15:46
mst jast: go join a local theatre group then, they're usually awesome :D
get_drunk blegh
jast that's the boring kind of drama
get_drunk actually
if it was a bunch of perl hackers I'd do it
jast ideally nobody actually enjoys it. there's a residual risk of enjoyment in theatre
mst jast: you've clearly never been backstage during a performance of wyrd sisters that was going wrong
15:46 Humdai left
get_drunk interdisciplinary perling could be a thing 15:46
jast you've got me there 15:47
literal or is trans() only for single-char replacements?
jast literal: AFAICT, yes. just like most transliterate features in various languages. 15:48
get_drunk depends on their gender
jast but I'm wrong \o/
get_drunk same
jast based on skimming the docs, maybe try this: ... .trans([Q{\}] => [Q{\\}]) 15:49
(I always forget about the bot and so don't try it myself :))
sufrostico /clear 15:50
jast now I want to know 15:52
m: my $f = Q{b\ar}; say $f.trans([Q{\}] => [Q{\\}])
camelia rakudo-moar bf099b: OUTPUT«b\\ar␤»
15:52 Humdai joined 15:59 _nadim left 16:01 _nadim joined 16:06 azawawi left 16:07 Emeric left 16:08 Emeric joined
llfourn psch: RT #128330 16:10
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128330
16:12 setty1 joined 16:14 nadim joined 16:17 _nadim left 16:18 _mg_ left 16:19 silug left, vike joined 16:21 sno left 16:27 kurahaupo joined 16:29 notbenh joined 16:34 silug joined 16:44 yqt left 16:50 brrt left 16:55 dha joined 16:56 zakharyas left 16:59 Humdai left 17:06 TreyHarris joined, telex joined 17:12 kurahaupo left 17:13 firstdayonthejob joined, dakkar left 17:22 zakharyas joined, grondilu joined 17:25 domidumont joined 17:26 mr-foobar left 17:44 rurban2 left 17:54 girafe left
masak m: class A::B {}; class C { has A::B $.ab = A::B.new }; say "alive" 18:10
camelia rakudo-moar 3156a9: OUTPUT«alive␤»
masak m: class A::B {}; class C { has A::B $.ab .= new }
camelia rakudo-moar 3156a9: OUTPUT«===SORRY!===␤Could not locate compile-time value for symbol A::B␤»
perlpilot that one seems familiar. I bet there's already a ticket for it. 18:11
masak yes, that's from RT #126975 18:18
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=126975
18:18 spider-mario joined
masak just found it in some 007 code, and wanted to check if it's still alive and kicking :) 18:19
dha James Bond does Perl?
masak oh, he's at the cutting edge of Perl 18:20
perlpilot oh
masak dha: it's the Q department, providing Mr Bond with all the latest futuristic tech
finest secret Britain can produce
dha Rule Brittania! 18:21
Wait, is it the *real* Q giving him this stuff or that new 12 year old?
masak that's... a surprisingly apt slogan for the Bond movie franchise
TimToady why do you think MI-6 and such are illegal identifiers?
masak dha: it's the real Q
dha whew.
TimToady++
masak yeah, just becase Perl 6 is a 15-year-old doesn't mean it'd trust a 12-year-old to produce its stuff 18:22
arnsholt once read a spy thriller where MI6 had been mangled into M16
TimToady no, we want the 7-year-olds, not the 12-year-olds
arnsholt Not sure if it was copy-editing or the translator's fault
TimToady by 12 they start to develop a conscience
18:23 rurban joined
TimToady realizes his recent utterances could be taken the wrong way, and shuts up... 18:24
timotimo what, monkeys?
masak TimToady: haha
timotimo .o( it's not even called cont, it's called context! )
masak TimToady: it's already too late -- that one is already in *someone's* quote file :P
TimToady blames camelia 18:25
18:25 jack_rabbit left, sno joined
literal is there a shorthand to perform all substitutions according to some %hash? 18:25
on a string
TimToady with which semantics? 18:26
18:26 rurban left
literal build up an alternation (LTM I suppose) of all the keys, do a global subst on the string with the values as replacements 18:26
TimToady with or without overlap after substitution? 18:27
probably without, most people would expect
literal without
basically the equivalent of this in Ruby: string.gsub!(Regexp.union(myhash.keys), myhash)
TimToady if trans were up to original spec, it would work for this
since trans was specced to to LTM on left-sides longer than a character 18:28
AlexDaniel m: ‘a123b123c’.trans(‘23’ => ‘’, :delete);
camelia ( no output )
AlexDaniel m: say ‘a123b123c’.trans(‘23’ => ‘’, :delete);
camelia rakudo-moar 3156a9: OUTPUT«a1b1c␤»
TimToady oohie
maybe it'll work
perlpilot doubts 18:29
literal I tried trans but it wouldn't insert replacements that were longer than 1 character
AlexDaniel m: say "a123b123c".trans(‘23’ => ‘X’, ‘a’ => ‘Z’, :delete);
camelia rakudo-moar 3156a9: OUTPUT«Z1Xb1Xc␤»
AlexDaniel m: say "a123b123c".trans(‘23’ => ‘LONGER’, ‘a’ => ‘Z’, :delete);
camelia rakudo-moar 3156a9: OUTPUT«Z1LOb1LOc␤»
AlexDaniel WTF?
TimToady bug, I say
perlpilot trans needs some love looks like
18:30 zakharyas left
TimToady literal: so yeah, trans does that, it just doesn't do it right :) 18:30
18:30 yqt joined
literal alright 18:30
18:31 FROGGS joined
AlexDaniel m: say "abCDEe".trans(‘CDE’ => ‘LONGER’); 18:31
camelia rakudo-moar f0f1d4: OUTPUT«abLONe␤»
AlexDaniel :DD 18:32
okay
literal I guess it just replaces the character positions it matched and no more 18:33
FROGGS o/
18:33 girafe joined
AlexDaniel oh 18:34
look
m: say "abCDEe".trans(/CDE/ => ‘LONGER’);
camelia rakudo-moar f0f1d4: OUTPUT«abLONGERe␤»
literal huh
FROGGS O.o
AlexDaniel wait…
m: say "DabCDEe".trans(/CDE/ => ‘LONGER’);
camelia rakudo-moar f0f1d4: OUTPUT«DabLONGERe␤»
AlexDaniel m: say "DabCDEe".trans(‘CDE’ => ‘LONGER’);
camelia rakudo-moar f0f1d4: OUTPUT«OabLONe␤»
masak why am I getting this error: ?
Missing or wrong version of dependency '/home/masak/ours/007/lib/_007/Parser/Syntax.pm (_007::Parser::Syntax)' (from '/home/masak/ours/007/lib/_007/Parser/Actions.pm (_007::Parser::Actions)')
FROGGS one of these must be wron, no? 18:35
masak (I just changed Syntax.pm)
FROGGS masak: whipe /home/masak/ours/007/lib/.precomp
AlexDaniel FROGGS: uh, not really… I'm not sure
masak FROGGS: just did, reflexively
FROGGS: but I don't think the error message should ever happen
AlexDaniel ‘CDE’ => ‘LONGER’ means ‘C’ => ‘L’, ‘D’ => ‘O’ and so on 18:36
FROGGS masak: something uses X which uses Syntax.pm, and that isnt handled right it seems
masak: aye
masak: from my weekend experience it does not seem to happen if you only have one level of dependency
AlexDaniel m: say "CabCDEe".trans( ‘CDE’.comb => [‘LONGER’, ‘heh’, ‘test’]); 18:37
camelia rakudo-moar f0f1d4: OUTPUT«LONGERabLONGERhehteste␤»
18:37 sno left
AlexDaniel so it does work 18:37
it's just that it anti-dwim a little bit 18:38
is*
masak FROGGS: not sure how to count my levels of dependency, but 007 is a pretty multi-layered project
AlexDaniel literal: I believe that you can use it today
FROGGS I guess this is not enough: rakudo/src/core/CompUnit/PrecompilationRepository.pm:121: return False if $modified > $since; 18:39
AlexDaniel so there's no bug :) 18:41
literal AlexDaniel: yeah
18:41 labster joined
AlexDaniel I find the current behavior a little bit weird but… it does the job if you play by its rules 18:41
m: say ‘abc’.trans(‘b’ => ‘hello’) 18:43
camelia rakudo-moar f0f1d4: OUTPUT«ahc␤»
AlexDaniel ok, so, why no warning? 18:44
m: say ‘abc’.trans(‘bc’ => ‘h’)
camelia rakudo-moar f0f1d4: OUTPUT«ahh␤»
masak g'ah, had to blow lib/.precomp again... :(
AlexDaniel ok, that I can understand
FROGGS masak: every time :/
masak: let me try to actually golf it 18:45
masak FROGGS: much appreciated
18:45 zakharyas joined
dalek osystem/MARTIMM-patch-1: f712c39 | (Marcel Timmerman)++ | META.list:
Add Config::DataLang::Refine to ecosystem

See github.com/MARTIMM/config-datalang-refine but pod file is better		 18:53
18:54 huggable left 18:57 _nadim joined
dalek osystem: f712c39 | (Marcel Timmerman)++ | META.list:
Add Config::DataLang::Refine to ecosystem

See github.com/MARTIMM/config-datalang-refine but pod file is better		 18:59
osystem: d9aebe1 | (Zoffix Znet)++ | META.list:
Merge pull request #217 from perl6/MARTIMM-patch-1

Add Config::DataLang::Refine to ecosystem
18:59 nadim left
dalek osystem/MARTIMM-patch-1: 9cd1e6f | (Zoffix Znet)++ | META.list:
Remove unwanted empty line
 18:59
19:01 nadim joined 19:03 _nadim left 19:06 revdiablo joined 19:08 andreoss joined 19:15 tbrowder joined
tbrowder m: say 'hi' 19:15
camelia rakudo-moar f0f1d4: OUTPUT«hi␤»
yoleaux 30 May 2016 10:50Z <Zoffix> tbrowder: any word on merging this? I recall you received positive feedback on adding a new tab: github.com/perl6/doc/compare/maste...perl6-docs
FROGGS masak: it is not just a A-uses-B-uses-C problem
masak FROGGS: ok...
tbrowder Zoffix: I'm still working on it, but advice helpful; issues for me before any merge: (1) be able to sort input pod docs by file name, not internal title 19:17
(2) figure out a way to work around the "table row with leading hyphen issue (#128221) 19:18
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128221
FROGGS huh 19:19
when C was changed, it tries to precompile A and fails, then B and fails, then C, continues, and then fails that D has an outdated dependency on C 19:20
so it is more about a diamond dependency relationship
tbrowder in the meantime, I may have found another bug: by my reading of the docs, this should be a valid scalar: "$issue-1234-fixed" but I'll try it:
m: my $issue-1234-fixed = False;
camelia rakudo-moar f0f1d4: OUTPUT«5===SORRY!5=== Error while compiling /tmp/DXfR9HBMMD␤Preceding context expects a term, but found infix = instead␤at /tmp/DXfR9HBMMD:1␤------> 3my $issue-1234-fixed =7⏏5 False;␤» 19:21
19:21 araujo_ joined
tbrowder m: my $issue-N1234-fixed = False; 19:21
camelia ( no output )
tbrowder note adding a letter in front of the numbers quells the problem
m: my $issue-_1234-fixed = False;
camelia ( no output )
perlpilot tbrowder: pretty that's how it's supposed to be. 19:22
tbrowder m: my $issue-1234_fixed = False;
camelia rakudo-moar f0f1d4: OUTPUT«5===SORRY!5=== Error while compiling /tmp/aZXEhWC44D␤Confused␤at /tmp/aZXEhWC44D:1␤------> 3my $issue-12347⏏5_fixed = False;␤»
perlpilot i.e. the thing immediately after a - must match <.alpha>
[Coke] tbrowder: where in the docs? (I can't even find "identifier" in the index) 19:23
tbrowder Coke: that's another doc bug; the def is found here: <doc.perl6.org/language/syntax#Identifiers>
AlexDaniel m: sub test { 42 }; my $hello = 69; say $hello-10-test
camelia rakudo-moar f0f1d4: OUTPUT«17␤»
AlexDaniel tbrowder: ↑ I think that this is why it is not valid 19:24
perlpilot yep
19:24 araujo left
perlpilot (at least that seems right to me) 19:24
19:24 domidumont left 19:25 zakharyas left, darutoko left
tbrowder well the def in the docs doesn't agree (but I haven't checked the synopses) 19:25
19:26 andreoss left
AlexDaniel I see no tests at all: github.com/perl6/roast/blob/ab7476...ntifiers.t 19:27
19:27 telex left
AlexDaniel ah no, that's S15 19:27
dalek k-simple: baec00c | azawawi++ | examples/12-frame.pl6:
Add a seperate frame demo
k-simple: 99cc705 | azawawi++ | /:
Merge remote-tracking branch 'upstream/master'
k-simple: 81d43c7 | azawawi++ | / (5 files):
Add initial Places sidebar example
k-simple: 46499b1 | azawawi++ | / (3 files):
More documentation and name fixes
k-simple: 27e0315 | azawawi++ | / (3 files):
More fixes
k-simple: 7522fd8 | RabidGravy++ | / (6 files):
Merge pull request #48 from azawawi/master

Separate Frame widget example + Places Sidebar + Documentation fixes
AlexDaniel here github.com/perl6/roast/blob/ab7476...entifier.t 19:28
dha Actually, the description of identifiers says that that should be valid, but I think the examples don't necessarily agree. the first example of a NON-valid identifier is "with-numbers1234-5" which seems to agree with the description.
19:28 sno joined, telex joined
avalenn_ the def and the examples in the doc don't agree with each other 19:28
design.perl6.org/S02.html#Names_and_Variables 19:29
An identifier is composed of an alphabetic character followed by any sequence of alphanumeric characters. The definitions of alphabetic and numeric include appropriate Unicode characters. Underscore is always considered alphabetic. An identifier may also contain isolated apostrophes or hyphens provided the next character is alphabetic.
dha Well, the design docs are pretty much considered historical at this point.
perlpilot avalenn_++
grondilu m: say my $apostroph'attempt 19:30
camelia rakudo-moar f0f1d4: OUTPUT«(Any)␤»
grondilu TIL
perlpilot m: my $foo't-gun;
camelia ( no output )
dha Ah, so the design document has that right, and the user docs have it wrong.
(apparently)
avalenn_ examples in the language documentation agree with design doc
AlexDaniel dha: what's right about it?
ah 19:31
didn't read the last few words…
dha the last bit about what follows the hyphen or quote needing to be alphabetic.
tbrowder But the S02 definition clarifies that each portion delimited with a leading hyphen or comma must start with a alphabetic character--that's what's missing in the docs 19:32
AlexDaniel but also note that there's no test for that
dha If we're agreed that that's the problem, I'm happy to add that.
oop. Then there should be a test for that. :-)
tbrowder dha: and change the docs
AlexDaniel dha: didn't you just say that you'd be happy to add a test?
perlpilot AlexDaniel: I think he was talking about the docs 19:33
AlexDaniel perlpilot: shhh
dha Nope. I'd be happy to add docs. :-) But I'm willing to take a whack at the test as well.
[Coke] assuming that's the desired behavior, aye. A pull request on the docs requesting someone verify that's the desired behavior (and/or on the spec test suite) appreciated.
dha If someone would just tell me where said test should be put, I'll see what I can do.
AlexDaniel [Coke]: well, keep in mind that the only thing that is out of sync is the docs 19:34
[Coke] so we don't run afoul of pmichaud's concerns about the defacto implementation becoming spec.
AlexDaniel [Coke]: both the speculations and the implementation does it “right”
[Coke] AlexDaniel: no, if there's no tests, then there's no spec.
AlexDaniel [Coke]: ok, indeed
[Coke] the speculations help drive it, but post-christmas, we shouldn't assume they are right.
AlexDaniel dha: S02-names/identifier.t
perlpilot [Coke]++ for keeping things in perspective. 19:35
[Coke] that said, I think they are, and pre-emptively vote +1 on this. :)
dha ok. Will take a shot at it.
AlexDaniel ah wait
“foob-4 is not a valid identifier (not alphabetic after hyphen)";”
how did I miss that
dha Where's that?
AlexDaniel github.com/perl6/roast/blob/ab7476...fier.t#L25
so yeah, *everything* is right except the docs 19:36
perlpilot so there *are* tests
one less thing for dha to do :)
AlexDaniel I'm probably blind, not sure how I missed it
dha perlpilot - but not for the hyphen.
AlexDaniel dha: github.com/perl6/roast/blob/ab7476...fier.t#L28
dha oh. how did I miss that. 19:37
bleah.
ok, so the doc just needs to change
AlexDaniel yep
as simple as that
OH COME ON. Who is “as”??
lizmat m: my $foo = 666; say $foo-42 # we probably don't want to lose this functionality 19:38
camelia rakudo-moar 055ca1: OUTPUT«624␤»
dha Ok. I'll add the aphabetic thing.
lizmat - being able to subtract? Yeah, I'm sure *someone* needs that. 19:39
:-)
lizmat subtract without whitespace, was my point :-)
tbrowder of course (slaps forehead hard)!
dha Yes. 19:40
perlpilot yeah, there are lots of examples I think that do that.
lizmat I think that's the only reason... if we would force whitespace for an infix:<->, then we could have $foo-42
tbrowder I've just been enjoying kebab case too much...
lizmat but I think that's too alienating for people coming from p5
tbrowder but us non-typists love it, even coming from p5! 19:41
(no shift key!!)
19:42 rindolf left
perlpilot just make sure you don't have variables named $A, $A-B and a constant named B that you want to subtract ;) 19:43
dha How's this for wording: You can also embed dashes C<-> or single quotes C<'> in the middle, but not two in a row, and only if followed by an alphabetic character.
perlpilot dha: maybe add the word "immediately" for emphasis 19:44
dha ok.
lizmat m: my $a = 42; my \b = 666; say $a-b # perlpilot: I think LTM will take care of that
dha Anything else before I submit?
camelia rakudo-moar 055ca1: OUTPUT«5===SORRY!5=== Error while compiling /tmp/0xaTKAlml2␤Variable '$a-b' is not declared␤at /tmp/0xaTKAlml2:1␤------> 3my $a = 42; my \b = 666; say 7⏏5$a-b # perlpilot: I think LTM will ta␤»
19:46 _nadim joined 19:49 nadim left, _nadim left, _nadim joined
perlpilot m: my $a-b = 42; my $a = 39; my \b = 7; say $a-b; # oops, I wanted 32 19:52
camelia rakudo-moar 055ca1: OUTPUT«42␤»
dha Grr... I really need to figure out how to create a pull request for JUST one commit instead of EVERY commit I've done since I last made a pull request.
perlpilot lizmat: what I was saying is that (under those contrived circumstances), you wanted to subtract, but the compiler thought otherwise. You've just made a new footgun for yourself. 19:53
dha: you can do that by making a branch for just the one commit.
lizmat true, hence perhaps we should emphasize longest token match at this place in the documentation ?
FROGGS masak: do you ex-/import tags in the packages the packages that fails with "Missing or wrong version of dependency"? 19:54
dha perlpilot - that sounds like a HORRIBLE solution... :-/
FROGGS masak: do you have*
perlpilot dha: eh, branches are cheap. I make them all the time.
lizmat: I haven't the words, but certainly a word or two of caution could be put in the docs there. 19:55
lizmat is currently busy with words for the P6W 19:56
dha Ok, I can't remember the last time I did a branch. Will probably have to look that up now...
19:56 Zero_Dogg left, Zero_Dogg joined, Zero_Dogg left, Zero_Dogg joined 20:01 sno left 20:02 TEttinger joined 20:06 sufrostico left 20:08 sufrostico joined
geekosaur branches are very cheap in git, and they're how it keeps track of this stuff 20:09
it's not like svn or cvs where a branch is effectively a repo copy
mst dha: ideally, you'd use a branch for each unit you'll later PR 20:10
dha It's times like this I wish I were an actual developer. :-| 20:11
mst because sungo and I are so more happy go lucky, yeah :P 20:12
dha Well, it's not like I'm so happy go lucky to begin with, and were I a developer, I'd probably actually understand git better. So, result. 20:13
20:13 rurban joined 20:14 _mg_ joined
RabidGravy nobody understands git,they just use it 20:15
20:15 rurban_ joined
dogbert17 git happens ... 20:16
RabidGravy :)
mst dha: most important thing I ever learned about git 20:17
(cd ..; cp -a MyRepo MyRepo.bak73)
before I do anything clever
RabidGravy I've lost count of the version control systems I've used, but they all do the same thing
20:17 [particle] joined
dha I... wouldn't have thought that trying to do a pull request on a single commit would be considered "clever" 20:18
Hm. Looks like I need to figure out cherry-pick to make this work. 20:19
b2gills Considering you can force push a branch to change a PR, how would doing a PR on a single commit work? 20:20
20:22 rurban_ left
dha Uh... I'm sure that means something, but apparently I'm too dim to know what that is. :-( 20:22
The great irony here is that I'm just adding a few words, and I actually have commit rights on the actual doc repository, so I *could* do it directly. But experience tells me that that way lies unpleasantness... 20:23
[Coke] dha: once you great the branch and are on it, "git cherry-pick sha-of-one-commit-you-care-about"
s/great/create/ # huh 20:24
dha :-)
yeah, I'm seeing that's what I should do. I'm just not clear on what cherry-pick actually *does*. but that doesn't mean it wouldn't work, even if I have no idea what's going on. 20:25
lizmat PSA: does anybody have something in the Perl 6 world that "defined" the past week ? 20:26
stmuk_ lizmat: I think there are new videos at www.youtube.com/channel/UCMDEevWdp...JId-bPyReg since the last P6W 20:27
dha eek. I cherry-pick'ed and now it's... saying things.
stevieb dha: here's how I create PRs with an individual commit (or numerous commits related to one thing): gist.github.com/stevieb9/8b40e992e...0e9a387735
TEttinger lizmat: something something jon snow 20:28
I don't watch game of thrones, but apparently it's popular recently
TEttinger is unhelpful
dha m: say 'hodor'
camelia rakudo-moar 055ca1: OUTPUT«hodor␤» 20:29
stmuk_ actually maybe those videos did go in the last p6w (just)
lizmat stmuk_: yeah, they did :-)
I think
stmuk_ the last day's video aren't out yet 20:30
FROGGS masak: I filed it as RT #128332 20:32
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128332
AlexDaniel lizmat: bisectable bot is now more user-friendly 20:33
bisect: class A { has $.wut = [] }; my $a = A.new; $a.wut = [1,2,3]; $a.wut.say
bisectable AlexDaniel: exit code on a “good” revision is 1 (which is bad), bisecting with inverted logic
AlexDaniel: (2016-03-02) github.com/rakudo/rakudo/commit/fdd37a9
20:34 ocbtec left
AlexDaniel bisect: good=2016.03 bad 2016.02 say (^∞).grep({ last })[5] 20:34
bisectable AlexDaniel: exit code is 0 on both starting points, bisecting by using the output
AlexDaniel: “bisect run” failure
AlexDaniel bisect: say (^∞).grep({ last })[5]
bisectable AlexDaniel: exit code is 0 on both starting points, bisecting by using the output
AlexDaniel: (2016-03-18) github.com/rakudo/rakudo/commit/6d120ca
AlexDaniel and so on
FROGGS gnight 20:35
20:35 FROGGS left
lizmat gnight freeze 20:35
grrr
AlexDaniel yeah, most importantly this:
bisect: exit 42
bisectable AlexDaniel: on both starting points the exit code is 42 and the output is identical as well
20:36 sufrostico left 20:38 jonadab joined
dha Ok, cherry-pick'ing did *not* remove the unwanted commits from my pull request. I think I need to murder github. 20:44
jnthn bisect: say $*VM.version
20:44 tbrowder left
bisectable jnthn: exit code is 0 on both starting points, bisecting by using the output 20:44
jnthn: “bisect run” failure
jnthn o.o :)
20:44 kaare_ left
AlexDaniel that's very interesting… 20:46
jnthn I didn't expect it to do anything sensible, fwiw. I was mostly curious what it would do (and hoping "nothing too bad" :))
[Coke] dha: you need to make sure you have a clean place to branch from; branch; cherry-pick; push; then PR that branch. 20:47
AlexDaniel jnthn: it shouldn't've failed
dalek sectbot: 72d52f5 | (Aleks-Daniel Jakimenko-Aleksejev)++ | / (3 files):
Garbage in, rainbows out

This commit includes a lot of changes to how problematic cases are handled. The basic idea is that first the code snippet is tested on good and bad revisions (git is not doing that by default), and further behavior is adapted according to that.
Also, 2015.12 is now used instead of v6.c.		 20:48
20:50 inokenty joined
lizmat and another Perl 6 Weekly hits the net: p6weekly.wordpress.com/2016/06/06/...bisecting/ 20:51
20:51 patrickz joined
jnthn AlexDaniel: btw, bisectbot is a really great idea :) 20:51
AlexDaniel jnthn: yea, thanks :) 20:52
vendethiel lizmat++ :D
20:52 get_drunk left 20:53 dha left
[Coke] AlexDaniel++ for sure 20:53
20:54 dha joined
masak FROGGS++ # rt.perl.org/Public/Bug/Display.html?id=128332 20:56
AlexDaniel jnthn: gist.github.com/AlexDaniel/0f90c29...1b4b13435f
20:56 _28_ria joined
dha [Coke] - define "clean place to branch from" 20:56
AlexDaniel jnthn: seems like you was just a bit unlucky
dha Also, unless I've done something wrong, git cherry-pick seems to have actually worsened the issue at hand. 20:57
mst yes, for pulling a single commit
AlexDaniel jnthn: there are like 15 revisions where the build process fails, so there's no perl6 executable. Such commits are skipped
mst once it gets merged you'll want to rebase your stack o' commits branch onto master
and it should just re-order itself
dha Well, the branch and the master didn't actually differ in the first place, although cherry-picking probably has changed that. 21:00
21:01 cpage_ joined
mst dha: right, I said *once it gets merged* 21:01
i.e. after the PR is accepted into master
[Coke] dha: are you working in a fork? 21:02
dha ah. But in the meantime, the problem I was trying to solve by making a branch (i. e. only have one commit be used in the pull request) persists.
[Coke] - yes.
[Coke] if so, then rebase your master so it include's the upstream's master commits.
21:03 wamba left
[Coke] then branch off your clean master that doesn't have any of your other commits. 21:03
21:03 Emeric left
[Coke] *includes 21:03
AlexDaniel what's the idea behind MONKEY-WRENCH and MONKEY-BARS ?
mst [Coke]: I think you've omitted about the first six steps from dha's point of view there
dha: so, if you keep a stack of commits in a wip branch atop master 21:04
[Coke] dha, mst: yes, well, not knowing how off-kilter his fork is, hard to tell. And feel free to jump in with any missing steps.
mst dha: you can PR one via 'git checkout master; git checkout -b some_name; git cherry-pick <the commit from wip>; git push origin some_name'
and then PR some_name
if currently your local master *is* your wip 21:05
then 'git checkout master; git checkout -b wip; git checkout master; git reset --hard origin/master'
dha define "wip"
work in progress?
mst (remember that I said about 'cp -a' first before being clever though)
yeah. I mean, if you're more comfortable generating a pile of commits in a single branch
then teasing them out one at a time to PR
you can use a wip branch like that
if they're completely independent commits, I'd give 'em a branch each up front
git branches are basically a symlink to a commit, they're dirt cheap 21:06
mst resists urge to try and write Yet Another Git Tutorial
dha To be clear, what's going on on this end is: I forked the perl6/doc repo some time back. As of the beginning of this whole mess, it was up to date with the upstream repo.
mst ... wait, I think I might have a really easy way to get you a chance to 'redo from start' without losing any work 21:07
[Coke] note that if you have commit bits on the original repo, you can skip the fork entirely and put your branches to be merged in the original repo and PR from those branches. 21:08
dha But, there have been merges along the way from the upstream repo, so as to keep my fork in sync. Those commits are *all* showing up when I make a PR.
[Coke] (to cut this knot going forward)
21:08 pullphinger left
[Coke] dha: you shouldn't be working in master in your fork if you're doing PRs 21:08
mst so, 'git format-patch origin/master' will turn each of your local commits into its own .patch file
dha Ah. That's probably what I missed.
[Coke] which doesn't help you clean up what you have today.
21:08 skids left
mst at which point, I would expect you can start again with a completely fresh fork 21:09
dha No. no it doesn't. I think at this point I should just shoot myself and avoid the christmas rush.
mst dha: yeah, GH workflow is fork, clone, 'git checkout -b mything', add comits, submit mything branch as PR, git checkout master, repeat
[Coke] how many commits are we talking about? if you made one huge PR and someone reviewed it...
21:09 geraud joined
dha Yep. I missed the branching part of that. 21:09
mst yes but (1) that would be a waste of a perfectly good bullet (2) we need your doc patches 21:10
(3) after bothering to try and explain it to you, clearly you owe me finishing the job, because the 10 minutes I spent totally give me the right to control your volunteer efforts, honest
dha [Coke] - I imagine that, if I put the pull request through, someone could certainly weed out the crap. But I was going to *try* to avoid that.
perlpilot lizmat++
stmuk_ github.com/perl6/doc/compare/maste...dha:master ? 21:11
mst dha: the other key problem is that you merged rather than rebasing, which makes for some untangling
though not that much
honestly, cherry picking each of your real commits into its own branch
then blowing away your current master
dha mst - please. I've been dealing with comic book fans lately. your claims to privilege over me pales in comparison. :-)
mst seems like it'd work fine to me
RabidGravy well "the crap" just makes it difficult to review, it won't mess up the target
dha stmuk_ - yes, that would be my fork
mst dha: bah. I'll try and up my game next time. 21:12
dha goes to read up on rebasing 21:13
21:14 TeTpaAka joined
stmuk_ what is the actual problem? 21:14
[Coke] dha: stackoverflow.com/questions/724432...repository
mst stmuk_: basically he wants to take his single master with assorted patches
and turn it into one PR per patch
without the mess
I think
dha: correct?
[Coke] top answer on how to do the rebase from upstream. (you might need a clean master to do that.) 21:15
dha Well, most of the commits I don't want to PR at all, but, basically, yes.
[Coke] - thanks. Do they also have a "how do I clean my master" entry? :-)
[Coke] you could probably programmatically find the commits you have that upstream doesn't have and do a per-branch commit for each one with a shell script.
21:16 cpage_ left
[Coke] if it's 100 commits, that might be worth it. If it's 10... eh. 21:16
heading out. g'luck.
dha thanks
21:17 cpage_ joined, cpage_ left
dha I'm actually trying to PR *one* commit. I just can't figure out how to not PR all the other crap. 21:17
21:18 cpage_ joined
stmuk_ just delete the branch and change the few lines again? 21:18
and in future "git pull --rebase" if you care (and most people don't bother but then most would only git pull once before committing) 21:19
mst dha: I thought you managed that with cherry-pick
hm. do you have a remote for upstream?
dha Nope. cherry-pick just added the relevant commit *again* 21:20
I believe so.
mst dha: er, yes, you need to do the cherry pick on a fresh branch that contains only the perl6 repo's master
dha: do you have a remote for upstream? 21:21
dha I believe so. Yes. Alternately, I am totally misunderstanding you.
mst ok, what's it called?
21:22 _mg_ left
mst (origin is presumably your fork, not the perl6/doc repo itself) 21:22
dha yes. 21:23
mst so, what's the remote for perl6/doc called?
dha git tells me I have origin, perl6 and upstream
mst AHA
dha aha?
mst except that now I dunno whether perl6 or upstream is the real one
which one have you been merging from?
perl6/master or upstream/master ? 21:24
21:24 TeTpaAka left
mst dha: ? 21:25
dha I should note that the urls for perl6 and upstream are the same.
mst ok
so it probably doesn't matter
RIGHT, let's see if this works then
dha Oh, I'm at the point where *nothing* matters anymore...
mst git checkout perl6/master; git checkout -b this_is_msts_fault; git cherry-pick <commit goes here>; git push origin this_is_msts_fault;
and then see if the PR button on origin looks like it's going to PR the commit you wanted it to 21:26
dha it doesn't like "git checkout perl6/master"
masak 'night, #perl6 21:27
dha night masak
mst dha: "doesn't like" ?
dha error: pathspec 'perl6/master' did not match any file(s) known to git.
mst it'll do some complaining about 'detached head'
ok. try 'upstream/master'
there must be one somewhere in here
dha oh. wait. i did something wrong.
mst the idea was that appart from <commit goes here> 21:28
you could actually copy-paste what I said into IRC
thereby reducing ti to 'mst did something wrong'
dha ok, now I'm getting the detached head thing.
mst that's fine 21:29
git checkout -b this_is_msts_fault
dha sorry about that. I tried to checkout in the wrong window before.
mst # now you're in a new branch from that
git cherry-pick <commit> # cherry pick it
21:29 stevieb9 joined
mst git push origin this_is_msts_fault # push just this new branch to origin 21:29
21:29 patrickz left
dha ok.... let me cherry-pick... 21:29
RabidGravy I'm your ch-ch-ch-cherry bomb! 21:30
literal is there a way to do an exact type check against a typename, analogous to «$foo.^name eq 'Hash'»? comparing against .WHAT seems to cause warnings, and smartmatching is too lax 21:31
dha ok, picking of cherry achieved.
dha pushes
Now which branch am I trying to do the PR from? 21:32
mst this_is_msts_fault
stevieb9 the one you just pushed
21:32 setty1 left
mst to perl6/doc/master 21:32
dha \o/ 21:33
It's now just the one commit.
mst++
stevieb9 after you've submitted the PR, don't forget to switch back to master, or you'll add more commits to the branch (and the PR) 21:34
mst right, at this point you can do 'git checkout master' to get back to your own master-with-stuff
then do 'git checkout -b this_went_horribly_wrong' to make a copy of it
then 'git checkout master' to go back to your master 21:35
then 'git reset --hard perl6/master' to force it to look like the upstream master
and then you can start from fresh
and cherry pick out anything else you want from this_went_horribly_wrong later
dha: oh, and before you do any of this, do a 'cp -a' in case it goes even more horribly wrong - I did those all the time while learning git 21:36
dha ok, done. 21:37
Thanks!
mst when it goes well, git is amazing. once you get into a mess, it's about as fun to escape as "I'm not sure what shape my undo tree in vim is but I am sure I can't find the state I wanted anymore" 21:38
hence the 'cp -a' every so often while you learn
dha *nod*
mst++
mst grins 21:39
dha is there any point to putting this_went_horribly_wrong up to github? 21:40
stevieb9 when we switched over to git from svn at work last year, there wasn't enough training, and those new to git caused all sorts of mayhem. Not fun
mst dha: backup, and being able to link it to other people later, and being able to clone it on another machine to steal more bits later 21:41
and if anybody judges you negatively on the contents of a branch with *that* name, I think they can safely be ignored
dha ok. :-)
AlexDaniel literal: what do you mean by smartmatching is too lax?
dha now I just have to remember how to do that.
stmuk_ there is actually one commit anyway .. github PR would have worked
mst dha: wait
dha waits
mst stmuk_: no, we spent a long time unfucking to get to having only one commit. 21:42
stmuk_: that was the entire conversation
dha: to get this_went up to GH:
stmuk_ the rest are just merges
mst dha: git checkout this_went_horribly_wrong; git push origin this_went_horribly_wrong;
dha: then to get your rest master back up to GH
dha: 'git checkout master; git push --force origin master'
stmuk_ so there is no point in saving anything
dha ok. 21:43
mst oh, the only commit with any contents at all is the one already PRed now?
dha pretty much.
mst then, yeah, who cares
git branch -D this_went_horribly_wrong; # and don't bother pushing it to origin 21:44
mst was doing belt-and-braces lets-not-lose-anything
dha eh. already pushed
mst without ever actually looking at the repo, because I was trying to achieve 'dha at least can pretend he knows what's going on' as well as things being sorted
stevieb9 having issues installing Inline::Python today. Started on FreeBSD, so when I got home I tried Ubuntu in hopes it was my box. Travis is not enabled, so I can't see if this is a global problem. Can someone please have a look to see if I'm missing something? gist.github.com/stevieb9/6e8afa08c...5d4bfac03b
mst dha: I mean, I can't magically teach you to understand all of git
dha: but, like, a delusional sense of warm fuzziness? that's a doabla goal :D 21:45
stevieb9 I'm installing the module per the CONTRIBUTING doc
for perl6/doc
stmuk_ this is clear from the url I gave at 2111 above
dha Hm. I checked out master and it's telling me I'm behind by 382 commits, and it can be fast-forwarded. It also says I can use git pull to update my local branch.
Should I do that, or is there something else I should do that will not screw up everything again? 21:46
Also, I'm happy to be deluded in a fuzzy manner.
mst dha: yeah, if it says 'it can be fast-forwarded' then a pull shouldn't break anything 21:47
dha ok. things have been pulled. I assume I should now push? 21:48
mst I tend to always specify a branch to push
so 'git push origin master'
dha ah. ok.
mst (this means that, have I screwed up another branch, I don't accidentally propagate that without noticing)
stevieb9 very good habit to get into for sure... 21:49
mst "if git was a physical tool, it would have black and yellow stripes"
dha Apparently everything is up to date.
...so it would be a bumblebee?
mst I don't know, it's rjbs' excuse for dzil having sharp edges
stmuk_ you should have made the whitespace changes in another commit really :) 21:50
dha Oh. sorry.
dalek c: f97fc88 | (David H. Adler)++ | doc/Language/syntax.pod:
Added alphabetic-after-hyphen-or-quote restriction to Identifier
c: 7bb0eff | (Steve Mynott)++ | doc/Language/syntax.pod:
Merge pull request #563 from dha/this_is_msts_fault

Added alphabetic-after-hyphen-or-quote restriction to Identifier
mst dha: later, we can talk about "how to hate yourself while turning one commit into two" 21:51
but right now it's 22:51 UK time
and i'm thinking it's time to start drinking
dha I don't actually need lessons on how to hate myself.
ooh. drinking.
I don't think I'm doing that until wednesday.
mst dha: no, but doing it *and* getting two working commits afterwards takes a little extra knowledge 21:52
you provide the generalised self hatred, i'll provide the git specific self hatred 21:53
SYNERGY!
dha deal.
stmuk_ | gist.github.com/stevieb9/6e8afa08c...5d4bfac03b 21:55
oops
AlexDaniel mst: just nuke this commit, and then add -p ?
mst: assuming that it's the last commit 21:56
stmuk_ stevieb9: are you using python from ports/pkg? if so you need to recompile your own with -fPIC
mst AlexDaniel: read through the past hour of backlog, then tell me again that you think dha's at a level where 'just' is a word that's even remotely sensible to use for explaining that 21:57
AlexDaniel I think that dha was doing great :)
dha Aw... thanks.
mst yes. but I wouldn't describe unwinding and re-assemblin a commit as 'just' for ... oh never mind
RabidGravy just drink beer and it will be fine 21:58
dha FWIW, I think I understand what you're suggesting, but not certain I would be able to actually implement it.
RabidGravy - that's your answer to everything.
mst dha: which is precisely why I was offering to talk you through it rather than the offhand 'just'
that's because it's a GOOD answer 21:59
dha And I thank you for it.
22:00 cpage_ left
stmuk_ stevieb9: BTW Inline::Python is optimal for perl6/doc (athough building HTML is a lot slower under pygments) 22:01
s/optimal/optional 22:02
optimal as well
22:02 cpage_ joined, _28_ria left
stevieb9 stmuk_: I did compile with --enable-shared, but I'm doing a new build now as perhaps the new python wasn't being used. 22:03
lizmat good night, #perl6! 22:05
dha night lizmat 22:06
stmuk_ stevieb9: its probable noone has tried this under clang either 22:07
stevieb9 stmuk_: I'll rebuild on my fbsd box tomorrow if this works here on my ubuntu laptop now. If it still won't work on fbsd, I'll open a ticket 22:09
stmuk_ stevieb9: some stuff assumes gmake too
maybe not perl6/doc I forget 22:10
22:10 kurahaupo joined 22:11 cpage_ left, cpage_ joined 22:12 tbrowder joined 22:14 kid51 joined 22:27 sftp joined, skids joined 22:28 labster left 22:29 tomboy64 joined 22:30 Sgeo_ joined 22:32 bjz_ left 22:34 tbrowder left 22:36 _nadim left 22:41 sufrostico joined 22:47 mr-foobar joined, _28_ria joined 22:48 cdg left 22:53 bjz joined 22:54 huggable joined 22:55 dha left, dha joined 22:57 firstdayonthejob left 22:59 firstdayonthejob joined 23:03 mohae_ joined 23:04 rurban left, bjz left 23:06 mohae left 23:11 firstdayonthejob left, labster joined, kurahaupo left, kurahaupo joined 23:13 kurahaupo_ joined 23:16 captain-adequate joined, kurahaupo left 23:18 girafe left 23:20 RabidGravy left 23:25 jack_rabbit joined
Hotkeys_ is there a way to bind to a class attribute 23:27
say I have some class Foo { has $.attr } and I want to bind some $bar = "bar"
to some $foo.attr 23:28
but I want $foo.attr to change if I change $bar
how would I do that?
Juerd I don't know how, or if it's even possible. 23:29
Why do you want this?
Hotkeys_ I'm just curious really
since you can bind to regular variables
awwaiid m: class Foo { has $.attr; method bindit($v is rw) { $!attr := $v } } ; my $x = 5; my $f = Foo.new; $f.bindit($x); say $f.attr; $x = 42; say $f.attr 23:35
camelia rakudo-moar f80cc7: OUTPUT«5␤42␤»
yoleaux 5 Jun 2016 20:35Z <DrForr> awwaiid: It seems that lrepl only lets you invoke one middleware layer at a time, is this by design?
raydiak w
wrong terminal 23:36
though while I'm here...hi #perl6 \o 23:37
awwaiid DrForr: The intent, iirc, is for middleware to work very much like PSGI (rack, ring, wsgi,...) middleware -- each layer calls filters the data and invokes the next, and then can filter again on the way back out
23:37 spider-mario left
Hotkeys_ oh 23:38
thanks awwaiid
idk why i didn't think of that
awwaiid Hotkeys_: ya. It told me the left-hand-side couldn't be bound, so I figured wrapping it into a method so that the left-hand-side was more boring would work
23:39 wamba joined
awwaiid I don't know WHY you'd want to do that though :) 23:45
Juerd Binding at a distance sounds like a bad idea. 23:49
23:51 _28_ria left 23:53 _28_ria 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!