»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.
skink Hey timotimo, is there a list of circumstances in which Buf can or can't substitute CArray? 00:00
timotimo when you get something back from C 00:01
a Buf knows how big it is, a CArray doesn't
sortiz skink, you can't receive from C a Buf
timotimo C also doesn't know how big arrays are, at least not in a unified way
sortiz But if the C API expect the allocated space and the size, a Buf is safe. 00:03
00:04 pierre_ left, jamesnek1 joined
skink Right. c_fill/read_buffer(Buf|CArray, size) is fine, but otherwise you need to use CArray? 00:04
(Assuming you've allocated Buf[size-1]) 00:05
sortiz In particular if the C function will allocate the space, and returns it address, you need CArray.
And you got an "unmanaged" CArray. 00:06
00:07 jamesneko left
BenGoldberg Is there any way to distinguish an unmanaged carray from a managed one? 00:08
00:09 pierre_ joined
sortiz BenGoldberg, yep, see: github.com/salortiz/NativeHelpers-...01-basic.t 00:09
BenGoldberg Also, is there any way to tell an unmanaged carray, "you are $length elements long," so that read accesses outside that will die or return failures? 00:10
timotimo nothing built-in
grondilu m: sub infix:<§>($,$) is tighter(&[+]) {0}; say 1 + 2 § 3; 00:12
camelia rakudo-moar d7698f: OUTPUT«1␤»
grondilu m: multi infix:<§>($,$) is tighter(&[+]) {0}; say 1 + 2 § 3;
camelia rakudo-moar d7698f: OUTPUT«1␤»
arnsholt CArrays are a bit wonky, in general (in part my fault, I'm afraid =)
BenGoldberg m: multi infix:<§>($,$) is looser(&[+]) {0}; say 1 + 2 § 3;
camelia rakudo-moar d7698f: OUTPUT«0␤»
grondilu it did not work last time I tried 00:13
BenGoldberg m: multi infix:<§>($,$) is looser(&infix:<+>) {0}; say 1 + 2 § 3;
camelia rakudo-moar d7698f: OUTPUT«0␤»
grondilu I mean there was a different behavior between sub and multi
BenGoldberg ahh
grondilu that's ok I'm not going to complain about rakudo not failing :P 00:14
BenGoldberg m: multi infix:<§>(@_) { [+] @_ }; say [§], 1..3;
camelia rakudo-moar d7698f: OUTPUT«Cannot call infix:<§>(...); none of these signatures match:␤ (@_)␤ in block <unit> at /tmp/ux8ThIdkh8 line 1␤␤»
BenGoldberg m: sub infix:<§>(@_) { [+] @_ }; say [§], 1..3;
camelia rakudo-moar d7698f: OUTPUT«Too few positionals passed; expected 1 argument but got 0␤ in sub infix:<§> at /tmp/4f9gluDuTu line 1␤ in block <unit> at /tmp/4f9gluDuTu line 1␤␤»
BenGoldberg Well, that's different.
m: sub infix:<§>(@_) { [+] @_ }; say [§] 1..3;
camelia rakudo-moar d7698f: OUTPUT«Too many positionals passed; expected 1 argument but got 2␤ in sub infix:<§> at /tmp/nhf0uG4eQl line 1␤ in block <unit> at /tmp/nhf0uG4eQl line 1␤␤» 00:15
BenGoldberg m: multi infix:<§>(@_) { [+] @_ }; say [§] 1..3;
camelia rakudo-moar d7698f: OUTPUT«Cannot call infix:<§>(Int, Int); none of these signatures match:␤ (@_)␤ in block <unit> at /tmp/36WQDSNVM3 line 1␤␤»
grondilu oh I remember now
m: module A { multi infix:<§>($,$) is tighter(&[+]) is export {0}; } import A; say 1 + 2 § 3; 00:16
camelia rakudo-moar d7698f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/rcXmuV5Wkb␤Strange text after block (missing semicolon or comma?)␤at /tmp/rcXmuV5Wkb:1␤------> 3>($,$) is tighter(&[+]) is export {0}; }7⏏5 import A; say 1 + 2 § 3;␤ expecting any of:␤ …»
grondilu m: module A { multi infix:<§>($,$) is tighter(&[+]) is export {0}; }; import A; say 1 + 2 § 3;
camelia rakudo-moar d7698f: OUTPUT«0␤»
grondilu m: module A { sub infix:<§>($,$) is tighter(&[+]) is export {0}; }; import A; say 1 + 2 § 3;
camelia rakudo-moar d7698f: OUTPUT«1␤»
grondilu ^see?
it was when exporting. 00:17
ZoffixLappy May be worth reporting as a bug 00:18
zwu which module has defined LinearArray?
grondilu reports it
BenGoldberg Yeah, that's rakudobugabble ;)
sortiz zwu, NativeHelpers::CStruct 00:19
zwu sortiz: thank you!
Could not find NativeHelpers::CStruct. And Could not find NativeHelpers::Blob. Is there an option I should set? 00:21
ZoffixLappy zwu, did you install the needed module? 00:24
sortiz zwu, a simple 'zef/panda install NativeHelpers::blob' should suffice.
ZoffixLappy zef/panda install NativeHelpers::Blob 00:25
zwu oh, I have not installed any modules 00:30
Is there a list about important modules? 00:31
timotimo there's only a list of all modules
zwu timotimo:)
sortiz hesitating about the "pros"/"cons" on keep NH::* in one distro or separate them 00:36
00:36 AlexDaniel left 00:38 yasnothil joined 00:46 grondilu left
zwu Can we set memory alignment for NativeHelpers::CStruct too? 00:50
which is important for future vectorization support? 00:51
00:57 grondilu joined
sortiz zwu, not explicitly: LinearArray depends on calloc(..., sizeof(Foo)) semantics, and Rakudo's CStruct try to use the proper alignment. 00:58
dalek osystem: 405185e | (Zoffix Znet)++ | META.list:
Add Bailador::Template::Mojo::Extended

Extra features on top of the default ::Mojo renderer for Bailador: github.com/zoffixznet/perl6-Bailad...o-Extended
01:04
ZoffixLappy NOW, 4 modules later, I think, I can actually do work on the original web app I planned last Thursday :P 01:05
heh... npm gets 447 "new" modules a day: www.modulecounts.com/ 01:08
Either the counter system is broken or npm stands for "New Wheel Reinvention Factory"
01:10 pierre_ left
ZoffixLappy 1111th commit to ecosystem 01:11
01:11 molaf left
ZoffixLappy feels a pleasant feeling 01:12
01:24 molaf joined 01:29 itcharlie left 01:31 BenGoldberg left 01:33 BenGoldberg joined
Xliff_out m: my $b = Blob.new(^100); my $c = $b[10..20]; $c.WHAT.say 01:40
camelia rakudo-moar d7698f: OUTPUT«(List)␤»
01:41 zwu left
Xliff_out say $_ for ^5; 01:45
m: say $_ for ^5;
camelia rakudo-moar d7698f: OUTPUT«0␤1␤2␤3␤4␤»
01:47 ilbot3 joined, zacts joined 01:53 BenGoldberg left 01:54 BenGoldberg joined 02:10 kid511 joined, pierre_ joined 02:11 kid51 left 02:15 pierre_ left 02:17 ZoffixLappy_ joined 02:18 ZoffixLappy_ is now known as sexycodergirl, yqt left 02:19 sexycodergirl is now known as Zoffix 02:22 noganex joined 02:25 noganex_ left 02:34 kid511 left 02:39 skink left 02:48 hotel left, hotel joined 02:55 emdashcomma_ left 02:56 emdashcomma joined 02:58 loren joined 03:07 ZoffixLappy left, Zoffix left 03:10 grondilu left 03:12 pierre_ joined 03:13 kaare_ joined
Xliff_out m: my $b = Buf[int16].new(0 xx 100); 03:14
camelia ( no output )
03:14 skids joined
Xliff_out m: my $b = Buf[int16].new(0 xx 100); say $b.elems 03:14
camelia rakudo-moar d7698f: OUTPUT«100␤»
Xliff_out m: my $b = Buf[int16].new(0 xx 100); $b.map({$_ = 0}); 03:17
camelia ( no output )
03:17 pierre_ left 03:21 kurahaupo joined 03:28 kurahaupo left
Xliff_out m: my $b = Buf[int16].new(0 xx 100); $b.map({$_ = 0}); say $b is Buf 03:37
camelia rakudo-moar d7698f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/F8T6ZcXFG3␤Two terms in a row␤at /tmp/F8T6ZcXFG3:1␤------> 3.new(0 xx 100); $b.map({$_ = 0}); say $b7⏏5 is Buf␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ …»
Xliff_out m: my $b = Buf[int16].new(0 xx 100); $b.map({$_ = 0}); say $b ~~ Buf
camelia rakudo-moar d7698f: OUTPUT«True␤»
richi238 Hi 03:45
Is there a reason why the type num is not documented in the official doc?
doc.perl6.org/type.html
geekosaur because it's a work in progress? 03:47
things that can't be found there yet are often in the older spec docs at design.perl6.org
richi238 ah okay, thanks
04:06 Cabanossi left 04:08 Cabanossi joined
ugexe seems perl6 has totally lost flussence now. a shame... 04:16
04:18 khw left, loren left 04:25 loren joined 04:40 wamba joined 04:54 BenGoldberg left 04:55 dwarring left
MadcapJake ugexe: did something happen? 05:34
05:36 sortiz left
MadcapJake ugexe: gitlab.com/u/flussence 05:37
05:46 loren left 05:54 CIAvash joined 06:13 wamba left
stmuk_ he seems to have been trying to use the JVM backend 06:30
06:44 hotel left, hotel_california joined 06:56 sue joined 07:02 skids left 07:08 firstdayonthejob joined 07:13 firstdayonthejob left 07:24 rindolf joined 07:30 sue left 07:32 domidumont joined 07:34 CIAvash left 07:36 domidumont left 07:37 rurban joined, domidumont joined 07:38 mnick joined 07:39 pierre_ joined 07:43 sue joined 07:44 loren joined 07:48 rurban left 07:49 wamba joined 07:53 lizmat joined 07:54 rurban joined 07:56 loren left, rurban left 08:01 RabidGravy joined 08:10 pierre_ left 08:14 darutoko joined 08:19 firstdayonthejob joined, pierre_ joined 08:31 Ven joined
dogbert2 o/ #perl6, is there anybody out there ? 08:32
08:32 CIAvash joined
dogbert2 hmm 08:33
timotimo o/
_nadim dogbert2: Sunday morning, which is probably after a long "night to Sunday morning"
dogbert2 hello guys
_nadim morning
dogbert2 saw that the doc.perl6.org/language/variables#Dy..._variables had been cleaned up a bit 08:34
still, a lot of vars are still missing and some explanations seem incorrect
_nadim timotimo: I havea, crappy, piece of code with a grammar that core dumps when it is slightly modified. Do you think anyone would be interested in it? it's short.
timotimo i'll take it 08:35
_nadim 2 mn
dogbert2 $*DISTRO Which OS distribution am I compiling under? # this should be 'Which OS distribution am I running under' ?? 08:36
if think I'll write up some changes additions to the variable part and present it here for your perusal. How about that? 08:38
_nadim timotimo: nopaste.linux-dev.org/?1097300 08:41
dogbert2: "that" will probably be welcome but a patch even more. 08:42
dogbert2 _nadim: I can make changes in the repo :) but in a case like this it's nice to have someone take a look first 08:43
RabidGravy make the PR, people will look at it
dogbert2 RabidGravy; can do 08:44
timotimo _nadim: you didn't say anything about including a module :)
_nadim: note that using something like "push" in your action methods is unreliable in the face of backtracking 08:46
08:50 n0xff left 08:51 pierre_ left 08:52 pierre_ joined 09:01 n0xff joined
_nadim timotimo: just ignore the module it's just for display purpose. 09:06
timotimo: yes I am aware of that even though in this case it will work I believe. the whole closure thingy was because of frustration ;)
09:09 AlexDaniel joined
AlexDaniel richi238: if you can't find something then please open a ticket here github.com/perl6/doc/issues/new 09:11
richi238: in this case it was already opened: github.com/perl6/doc/issues/372
09:14 sue left 09:15 pierre_ left, pierre_ joined
dalek c/update-dynamic-vars: e827294 | (Jan-Olof Hendig)++ | doc/Language/variables.pod:
Added TMPDIR and fixed some errors (according to S28)
09:23
dogbert2 ok, PR done, did I get everything wrong? 09:25
09:27 Emeric joined 09:29 pierre_ left
RabidGravy no they look good, $*FOO should always be runtime, but an easy way to check is find the definition - if it's PROCESS::<$FOO> then it's definitely run time 09:29
dalek c: e827294 | (Jan-Olof Hendig)++ | doc/Language/variables.pod:
Added TMPDIR and fixed some errors (according to S28)
09:30
c: f928194 | RabidGravy++ | doc/Language/variables.pod:
Merge pull request #482 from perl6/update-dynamic-vars

Added TMPDIR and fixed some errors (according to S28)
AlexDaniel dogbert2: it looks right, but let's see what precompilation aficionados say :)
dogbert2 AlexDaniel: uh oh :) 09:31
RabidGravy the $*FOO should always be initialised when they are first used with the appropriate run-time values, there may be some hacks in there that make this not the case, but they should 09:32
dogbert2 m: say $. # what will the error message be .. 09:33
camelia rakudo-moar d7698f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/8vFxz5IbCw␤Unsupported use of $. variable; in Perl 6 please use the filehandle's .ins method␤at /tmp/8vFxz5IbCw:1␤------> 3say $.7⏏5 # what will the error message be ..␤»
AlexDaniel what is .ins 09:34
dogbert2 a no longer existing method, removed around Christmas
AlexDaniel irclog.perlgeek.de/perl6/2016-04-30#i_12414356
yeah
timotimo "number of input seperators encountered"
psch m: say $*IN.ins
camelia rakudo-moar d7698f: OUTPUT«Method 'ins' not found for invocant of class 'IO::Handle'␤ in block <unit> at /tmp/t2NZ9VVVql line 1␤␤»
AlexDaniel submits rakudobug
timotimo roughly translates to "current line in file"
what's the rakudobug here? 09:35
dogbert2 the error msg suggesting a method that no longer exists ?
AlexDaniel timotimo: well, you may fix it right away so that I don't have to submit a ticket :) 09:36
github.com/rakudo/rakudo/blob/8d69...n.pm#L1359
m: say $*IN.nl-in 09:37
camelia rakudo-moar d7698f: OUTPUT«[␤
AlexDaniel m: say $*IN.nl-out
camelia rakudo-moar d7698f: OUTPUT«␤␤»
09:38 travis-ci joined
travis-ci Doc build passed. Jan-Olof Hendig 'Added TMPDIR and fixed some errors (according to S28)' 09:38
travis-ci.org/perl6/doc/builds/127000170 github.com/perl6/doc/commit/e827294879e5
09:38 travis-ci left
timotimo ah, *that* is the bug 09:38
ok
09:40 TEttinger left
_nadim masak: going to Cluj? 09:41
09:46 travis-ci joined
travis-ci Doc build passed. Jonathan Stowe 'Merge pull request #482 from perl6/update-dynamic-vars 09:46
travis-ci.org/perl6/doc/builds/127000625 github.com/perl6/doc/compare/05bb1...281940d953
09:46 travis-ci left
dalek c: bceb198 | (Tom Browder)++ | doc/Type/Pod/Block.pod:
add table class
09:50
c: 5c517fe | RabidGravy++ | doc/Type/Pod/Block.pod:
Merge pull request #483 from tbrowder/master

add table class
09:50 wamba left
lizmat AlexDaniel: RT #128045 fixed 10:02
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128045
dogbert2 wonders if unicode contains a bikeshed character :) 10:04
10:05 Actualeyes1 left 10:06 sena_kun joined
tbrowder doc: how does an item become searchable in the search box? 10:07
dogbert2 lizmat: I like the new message :) 10:11
timotimo tbrowder: i think it's either an X<...> tag or some other one 10:12
10:12 tbrowder left, tbrowder joined
AlexDaniel dogbert2: there you go: 🚳 10:12
dogbert2 ok, one possible bikeshed then ... 10:13
AlexDaniel .u 🚳
yoleaux U+1F6B3 NO BICYCLES [So] (🚳)
dogbert2 how would you translate this p5 snippet with .kv
AlexDaniel dogbert2: which snippet? 10:14
dogbert2 perl -n -e 'print $., $_' /etc/passwd
AlexDaniel m: say .[0],.[1] for $*IN.lines.kv; 10:15
camelia rakudo-moar 1e54c5: OUTPUT«Index out of range. Is: 1, should be in 0..0␤ in block <unit> at /tmp/EHSctNz3eu line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/EHSctNz3eu line 1␤␤»
AlexDaniel m: say .k,.v for $*IN.lines.kv;
camelia rakudo-moar 1e54c5: OUTPUT«Method 'k' not found for invocant of class 'Int'␤ in block <unit> at /tmp/DhkGzc8MK9 line 1␤␤»
AlexDaniel oh geez
psch m: .say for lines().kv
camelia rakudo-moar 1e54c5: OUTPUT«0␤Céad slán ag sléibhte maorga Chontae Dhún na nGall␤1␤Agus dhá chéad slán ag an Eireagal ard ina stua os cionn caor is coll;␤2␤Nuair a ghluais mise thart le Loch Dhún Lúich’ go ciúin sa ghleann ina luí␤3␤I mo dhiaidh bhí gleann…»
psch m: .print for lines().kv 10:16
camelia rakudo-moar 1e54c5: OUTPUT«0Céad slán ag sléibhte maorga Chontae Dhún na nGall1Agus dhá chéad slán ag an Eireagal ard ina stua os cionn caor is coll;2Nuair a ghluais mise thart le Loch Dhún Lúich’ go ciúin sa ghleann ina luí3I mo dhiaidh bhí gleanntáin ghlas’ Ghaoth…»
psch that's probably the same
oh, no -n adds new-lines
lizmat for $*IN.lines.kv -> $index, $line { }
psch m: for lines().pairs { put .key, .value }
camelia rakudo-moar 1e54c5: OUTPUT«0Céad slán ag sléibhte maorga Chontae Dhún na nGall␤1Agus dhá chéad slán ag an Eireagal ard ina stua os cionn caor is coll;␤2Nuair a ghluais mise thart le Loch Dhún Lúich’ go ciúin sa ghleann ina luí␤3I mo dhiaidh bhí gleanntáin ghlas…»
AlexDaniel m: for $*IN.lines.kv { say $^a,$^b } 10:17
camelia rakudo-moar 1e54c5: OUTPUT«0Céad slán ag sléibhte maorga Chontae Dhún na nGall␤1Agus dhá chéad slán ag an Eireagal ard ina stua os cionn caor is coll;␤2Nuair a ghluais mise thart le Loch Dhún Lúich’ go ciúin sa ghleann ina luí␤3I mo dhiaidh bhí gleanntáin ghlas…»
lizmat m: for $*IN.lines.kv -> $index, $line { say "$index: $line" }
camelia rakudo-moar 1e54c5: OUTPUT«0: Céad slán ag sléibhte maorga Chontae Dhún na nGall␤1: Agus dhá chéad slán ag an Eireagal ard ina stua os cionn caor is coll;␤2: Nuair a ghluais mise thart le Loch Dhún Lúich’ go ciúin sa ghleann ina luí␤3: I mo dhiaidh bhí gleanntái…»
psch m: for lines() { put $++, $_ } # :P
camelia rakudo-moar 1e54c5: OUTPUT«0Céad slán ag sléibhte maorga Chontae Dhún na nGall␤1Agus dhá chéad slán ag an Eireagal ard ina stua os cionn caor is coll;␤2Nuair a ghluais mise thart le Loch Dhún Lúich’ go ciúin sa ghleann ina luí␤3I mo dhiaidh bhí gleanntáin ghlas…»
AlexDaniel m: for $*IN.lines { say $++, $_ }
camelia rakudo-moar 1e54c5: OUTPUT«0Céad slán ag sléibhte maorga Chontae Dhún na nGall␤1Agus dhá chéad slán ag an Eireagal ard ina stua os cionn caor is coll;␤2Nuair a ghluais mise thart le Loch Dhún Lúich’ go ciúin sa ghleann ina luí␤3I mo dhiaidh bhí gleanntáin ghlas…»
AlexDaniel ah
yeah
psch: right! Forgot that you can use just lines() 10:18
dogbert2 the following hangs: perl6 -n -e 'for $*IN.lines.kv -> $index, $line { say "$index: $line" }' /etc/passwd 10:20
timotimo doesn't -n run the code for every line?
psch dogbert2: i don't think you need -n if you read the lines yourself
AlexDaniel it's just perl6 -ne 'say $++,$_' /etc/passwd 10:21
psch hanging is kind of suboptimal though...
dogbert2 yeah, it's strange
AlexDaniel “readline requires an object with REPR MVMOSHandle” 10:22
and yes, it hangs!
dogbert2 AlexDaniel: your example worked 10:24
AlexDaniel interestingly both get() and lines() kinda work
but $*IN.get and $*IN.lines hang
dalek c: 41bbe3b | Altai-man++ | doc/Language/regexes.pod:
Typo in example fixed

With an order like "0-1-2" substitution makes no sense, since it doesn't change the variable. According to the comment line, the right order should probably be "1-2-0".
10:30
RabidGravy me-- # top tip - if you have an attribute "perl" on an object don't be surprised if things don't appear to behave as expected 10:39
psch RabidGravy: that's a neat braino :) 10:40
10:40 Ven left
timotimo %) 10:41
RabidGravy spent several hours and instrumented a couple of modules trying to debug that :-\
lizmat some lessons learned are expensive :-( 10:42
timotimo uh oh :(
10:42 CIAvash left
RabidGravy anyway off to my 30+ year old stomping ground to meet my brother and family 10:43
have fun
toodles
timotimo :)
i haven't a clue what "stomping ground" refers to :|
psch guesses pub 10:44
timotimo oh
RabidGravy the area I lived until I was 18
but yes there may be a pub involved ;0)
10:49 RabidGravy left 10:52 diakopter left
dalek c: 897f15e | (Tom Browder)++ | doc/Language/5to6-nutshell.pod:
make searchable
10:54
c: 5fa9347 | (Tom Browder)++ | doc/Language/about.pod:
add note about how to make an item searchable
c: 155e9da | (Tom Browder)++ | doc/Language/ (2 files):
Merge pull request #485 from tbrowder/master

make PERL6LIB searchable
10:58 kid51 joined 11:02 kid511 joined 11:03 kid51 left 11:06 CIAvash joined 11:07 diakopter joined
AlexDaniel dogbert2: RT #128046 RT #128047 11:15
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128046
Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128047
11:15 spider-mario joined
dogbert2 AlexDaniel++ 11:20
tbrowder doc: the X<> doesn't seem to make a term searchable, any other ideas? 11:26
11:26 zakharyas joined
tbrowder doc: ah, it looks like it's in htmlify.pl6... 11:28
11:29 Ven joined 11:30 Ven left 11:31 wamba joined 11:41 BenGoldberg joined
kid511 What is the appropriate forum in which to ask questions about libraries that are bundled with Rakudo Star? 11:49
Specifically, I have questions about DBIish
timotimo this here is a good place, but perl6-users@, the mailing list, is also a good place 11:50
and reddit is also all right, though i personally don't read it all that often 11:51
like, reddit.com/r/perl6
11:51 pdcawley joined
kid511 Well, let me ask one question here since it may be more general 11:52
At github.com/perl6/DBIish, the following instruction is given for reading the documentation:
perl6 --doc <filename>
But when I searched the rakudo-star tree, I could not find any file with 'DBIish' in its name. 11:53
Hence, I could not read the documentation. What is wrong?
timotimo oh, that's a good point
because the filenames now have sha1 hashes in them
psch yeah, i've tripped over that a few times too 11:54
kid511 So, doesn't that render 'perl6 --doc <filename>' useless, in practical terms?
lizmat is there a ticket for it ?
psch it probably should check if we have a module named <filename> in the CURI
lizmat is nine_ aware of this ?
11:55 Util joined, pdcawley left
kid511 How, in general, does one read the documentation in Perl6 via an invocation at the command-line? 11:55
timotimo there's the p6doc tool, but it's not for installed modules, is it?
kid511 i.e., is there a working equivalent to 'perldoc'? 11:56
psch well, --doc *should* work
timotimo if you can find the file, it definitely will work
psch i'll have a look in a bit, see if i can hook that up to look into the installation repo
lizmat I guess we need to move up handling of --doc to Perl 6 level as well
timotimo we ought to ship a few nice little tools for inspecting CUR's. like CURls that'll ls all the files that are part of a distribution and their installed paths (if there are any), and also CURcat to cat files that are installed even if there aren't any "files" so to speak 11:57
11:58 pierre_ joined
timotimo CURgrep/CURack, ... :) 11:58
lizmat timotimo: and those tools would simply use the CUR.methods available :-) 11:59
so that people can build their own tools :-)
timotimo yes, of course
those tools all ought to be one to five lines long
but there's still no reason to force users to write them out themselves :) 12:00
12:01 brrt joined
tbrowder doc revert pull #485: can someone please revert doc pull #485? it doesn't work as expected, thanks. 12:02
timotimo hm, no button up on github to do the revert in-place on the website 12:03
12:03 pierre_ left, ZoffixWin left
psch actually, i don't see an easy way to not require e.g. --docee=Module::Name... 12:05
lizmat tbrowder: you mean 897f15e39b3ea78cd ? 12:06
kid511 It appear that if, e.g., I want to read the POD found at github.com/perl6/DBIish/blob/maste...DBIish.pm6 from the command-line, I have to say:
perl6 --doc share/perl6/site/sources/8B5538E93290C550228635C3FE5A7111890C4339
... which meant that I first had to search the source tree for files with the string 'DBIish', read a lot of output, and then make an educated guess. 12:07
... which, IMHO, falls into the category of "barriers to adoption" 12:08
AlexDaniel kid511: you are right
lizmat kid511: completely agree :-)
kid511 Is there work under way to rectify this?
lizmat tries to invoke the nine_ 12:09
AlexDaniel since you've just brought it up, I don't think so
psch gets warnocked again
kid511 Also, here I at least had the advantage of knowing that DBIish was bundled with Rakudo-Star
AlexDaniel and there's also no tickte as far as I can see
ticket*
kid511 I would have been completely clueless if this were "installed from CPAN or the cloud"
I will open a ticket based on my remarks ^^ 12:10
lizmat kid511++
AlexDaniel kid511++
12:13 wamba left
jdv79 kid511: good point 12:14
12:15 zakharyas left 12:17 CIAvash left
AlexDaniel .seen Zoffix 12:17
yoleaux I saw Zoffix 26 Jan 2016 15:53Z in #perl6: <Zoffix> m: class Bar does Associative { has $.foo = 42; method AT-KEY($key) { self."$key"() } }; Bar.new<foo>.say
tadzik wouldn't 'p6doc DBIish' DTRT?
AlexDaniel .seen ZoffixWin
yoleaux I saw ZoffixWin 30 Apr 2016 19:55Z in #perl6: <ZoffixWin> \o
AlexDaniel .tell ZoffixWin where is huggable? :)
yoleaux AlexDaniel: I'll pass your message to ZoffixWin.
12:19 brrt left
tadzik kid511: yes, 'p6doc DBIish' is what you're after :) 12:22
similarly to how you'd do 'perldoc DBI' rather than 'perl -Mpod2text-or-whatever `find -name DBI.pm | head -n1` or something :) 12:23
psch tadzik: in that case i'd almost argue to toss the --doc switch 12:24
tadzik: 'case as-is it's just misleading and somewhat useless too
*cause
tadzik it's quite useful for developing and locally testing your own written doc
12:25 kurahaupo joined
tadzik I think the synopses 12:25
timotimo we could deprecate --doc and call the switch --renderpod instead
tadzik ' idea was to make --doc smart enough
so perl6 --doc DBIish would actually work
psch yeah, i'm looking how we could do that
hm, what's the CUR invocation to look for the path of an installed module..? :) 12:26
12:27 Ben_Goldberg joined, BenGoldberg left, Ben_Goldberg is now known as BenGoldberg
kid511 rt.perl.org/Ticket/Display.html?id=128048 12:28
tadzik I'll respond to the ticket :) 12:29
AlexDaniel what is p6doc
tadzik it's a perldoc for perl6,pretty much :)
AlexDaniel where can I get it? 12:30
tadzik same place where everywhere else :)
'redpadna p6doc', or your other favourite equivalent
lizmat redpadna ?
tadzik frankly, I think a reasonable way to resolve the situation is to make --doc load p6doc-the-module and delegate stuff ot it
reppadna, modulo my pytos :)
12:31 grondilu joined
tadzik github.com/tadzik/App-redpanda this 12:31
AlexDaniel oh, so panda install p6doc
tadzik I mean, the precedent to it this way is --doc=HTML that tries to load Pod::To::HTML
AlexDaniel is it included in rakudo star? 12:32
tadzik yeah, sorry for being deliberately confusing :)
I think it is, yes
ISTR there's at least one other module that does that, being the Debugger::UI::Commandline or so
AlexDaniel and then, how do I use it? 12:33
tadzik I don't see why not to make p6doc '2nd degree core' and have --doc rely on it, at least if --doc cannot find the source for the Pod by itself
as you would perldoc
p6doc DBIish
p6doc -f sort
AlexDaniel “panda install p6doc” does not create a p6doc binary on my system 12:34
tadzik (the latter won't work, as I've found, I'm sure moritz will be happy for patches :))
are you using rakudobrew?
in that case, 'rakudobrew rehashh
why can't I type today :|
AlexDaniel yeah, now I have it
tadzik awesome 12:35
AlexDaniel what if I wasn't using rakudobrew?
tadzik then it should Just Work
as long as the default panda install path is in your PATH
12:35 _nadim left
AlexDaniel hmmm… I wonder why would it be… in case of e.g. debian 12:35
ah, there's no panda in debian repos so perhaps it does not matter right now 12:37
tadzik: anyway, thanks! 12:38
tadzik (nvm) 12:44
12:45 loren joined
tbrowder lizmat: I think the whole commit is: 155e9daf87ce9e38e22ef040987a1b24f3f3dfe6 12:50
dalek c: 32d8fe2 | lizmat++ | doc/Language/5to6-nutshell.pod:
Revert "make searchable"

This reverts commit 897f15e39b3ea78cdd11a704c371b860ad124c5b.
12:53
tadzik lizmat: what is the reason for this? 12:55
dalek c: 232d81e | (Zoffix Znet)++ | doc/Language/5to6-nutshell.pod:
make searchable in a way that works
tadzik I assume "it makes it render badly while not giving us any value at this point".... forget I asked :)
lizmat tadzik: basically doing what tbrowder asked me to do ?
tadzik right, I didn't follow the entire thing, sorry 12:56
12:57 ZoffixWin joined 12:58 Actualeyes joined
ZoffixWin AlexDaniel, she keeps dropping because I never finished IRC::Client and I didn't really see people using her, so I stopped restarting her :) 12:58
yoleaux 12:17Z <AlexDaniel> ZoffixWin: where is huggable? :)
12:59 huggable joined
ZoffixWin huggable, help 13:00
huggable ZoffixWin, I'm a factoid bot. See me my source at github.com/zoffixznet/perl6-IRC-Cl...in-Factoid
13:00 pierre_ joined
ZoffixWin huggable, Int 13:00
huggable ZoffixWin, class Int [Integer (arbitrary-precision)]: doc.perl6.org/type/Int
moritz I'd be very interested in an IRC client that can reconnect automatically
because then I could use a Perl 6 version of ilbot in production, finally
ZoffixWin I'll put more work into it then, after I launch perl6.party
tadzik moritz: is ilbot running on Bot::BasicBot? 13:01
moritz tadzik: yes
tadzik moritz: did you do anything speciall to make it automatically reconnect? I have a bot I maintain on another channel, and every now and then it will "fall asleep" and stop responding, and I always manually restart it :? 13:02
moritz tadzik: I don't think I did :-) 13:04
github.com/moritz/ilbot/blob/master/ilbot2.pl
ZoffixWin There's this, but I'd be surprised if B::BB did not use it by default anyway: metacpan.org/pod/POE::Component::I...:Connector
13:05 pierre_ left
AlexDaniel ZoffixWin: perhaps no one is using her because she keeps dropping? 13:06
ZoffixWin :D
AlexDaniel ZoffixWin: I've tried to use it a couple of times but was very disappointed to see that the bot was not there when I needed it 13:07
ZoffixWin I'll shove it into a auto-restarter script that will kill and restart every so often :) 13:08
AlexDaniel huggable: no bikeshedding please :is: 🚳
huggable AlexDaniel, Added no bikeshedding please as 🚳
ZoffixWin m: '🚳'.uninames.say 13:09
camelia rakudo-moar 6df053: OUTPUT«(NO BICYCLES)␤»
ZoffixWin :S
AlexDaniel yea, it's a bit off… but good enough :)
ZoffixWin I wonder what it renders as
Ahh: www.fileformat.info/info/unicode/ch.../index.htm 13:10
Makes sense now.
AlexDaniel ZoffixWin: files.progarm.org/2016-05-01-16095..._scrot.png
tadzik so what does huggable do? :)
lizmat huggable: hug tadzik 13:11
huggable lizmat, nothing found
ZoffixWin Not much. She just sits around all day
:)
AlexDaniel huggable: what do you do for a living? :is: I am a factoid bot on #perl6
huggable AlexDaniel, Added what do you do for a living? as I am a factoid bot on #perl6
lizmat huggable: what do you do for a living?
huggable lizmat, I am a factoid bot on #perl6
tadzik oh that's useful :)
huggable: how is babby formed?
huggable tadzik, nothing found
lizmat huggable: what do you do for a living ?
huggable lizmat, nothing found
lizmat huggable is rather strict at what she understands :-) 13:12
tadzik can I teach it that old copypasta? :D
13:12 rba_ left
ZoffixWin What old copypasta? 13:12
tadzik ZoffixWin: knowyourmeme.com/memes/how-is-babby-formed 13:13
there's no full text in there though, eh 13:14
answers.yahoo.com/question/index?q...408AAZkbcG this might be the original
ZoffixWin *sigh* I never thought Yahoo would stoop to such bad design as to show damn popups in your face: i.imgur.com/XS0r6H1.png 13:16
AlexDaniel ZoffixWin: no damn popups here with µBlock 13:17
hmmm maybe it is because I am not using Chrome 13:18
tadzik "no need to put that x on that popup, I already have an x on my tabs"
keen eye spots a Steam icon on your screenshot, may I interest you in #perl6-gaming? :) 13:19
masak heh, X-able is a cute naming scheme for IRC bots. kind of self-documenting. ;) 13:32
we could rename camelia to "evalable"
timotimo rakudobuggable will be your new name, masak :) :) 13:34
masak bobs happily in mid-air :>
masak .oO( masak, the bugger-all bot ) 13:35
13:36 nd3i joined
nd3i given an array of pairs: @a = :one(1), :two(2) 13:37
AlexDaniel okay 13:38
13:38 huggable left
AlexDaniel ZoffixWin: :'( 13:38
ZoffixWin I'm making a re-starter
AlexDaniel ah
:) then 13:39
nd3i do I understand correctly that push(@a, :three(3)) fails because the pair is taken as a named parameter for push?
13:39 huggable joined
timotimo that's accurate 13:39
AlexDaniel yes
timotimo you can write push(@a, (:three(3))) if you watn
AlexDaniel m: my@a = :one(1), :two(2); push @a, (:three(3)); say @a
camelia rakudo-moar 6df053: OUTPUT«[one => 1 two => 2 three => 3]␤»
nd3i ok. Right, I found that (:three(3)) is the answer and that sort of made it clear what was happening. 13:40
AlexDaniel m: my@a = :one(1), :two(2); push @a, :three(3)[]; say @a
camelia rakudo-moar 6df053: OUTPUT«[one => 1 two => 2 three => 3]␤»
nd3i Still, it seems dangerous, no? It would be an easy mistake to make and not much fun to track down
would it be possible to detect and warn about at compile time? 13:41
I guess handling named parms is up to the sub, so there's not much the compiler can do 13:42
timotimo m: my @a; push @a, :foobar(1)
camelia ( no output )
timotimo interesting, i'm a bit surprised push has a *% to it
potentially because it just invokes the method form without doing argument checking itself at all?
AlexDaniel m: my@a = :one(1), :two(2); push @a, Pair.new(‘three’, 3); say @a 13:43
camelia rakudo-moar 6df053: OUTPUT«[one => 1 two => 2 three => 3]␤»
13:46 telex left
dogbert2 so, does anyone know what the method flurb does ? doc.perl6.org/routine/flurb 13:46
AlexDaniel “method description here”. Isn't it obvious?
psch m: say &flurb
camelia rakudo-moar 6df053: OUTPUT«5===SORRY!5=== Error while compiling /tmp/sRdaLXiMiw␤Undeclared routine:␤ flurb used at line 1. Did you mean 'slurp'?␤␤»
dogbert2 looks like a joke to me :) 13:47
ZoffixWin :)
dogbert2 would you all cry if I removed it :)
ZoffixWin Well, I'm searching Rakudo source for 'flurb' and nothing comes up. 13:48
timotimo it's in the doc repo
it comes from a tool that starts a doc page with a template for you
13:48 telex joined
timotimo oh 13:48
and actually it's in Submethod, too
=comment ... TODO
dogbert2 that's where I found it (in Submethod) 13:49
ZoffixWin AlexDaniel, what do you mean by "extra" whitespace?
RE github.com/zoffixznet/perl6-IRC-Cl...d/issues/4
AlexDaniel ZoffixWin: I've attached a piece of IRC log. I think that both should work
ZoffixWin disagrees 13:50
dogbert2 timotimo: thx
AlexDaniel huggable: what do you do for a living?
huggable AlexDaniel, nothing found
13:50 sjn joined
ZoffixWin AlexDaniel, '?' is part of the factoid. So inserting a space between it and the rest of the factoid would have to have some heuristics to find a guess 13:50
dogbert2 i'll remove the submethod one then :)
ZoffixWin huggable, what do you do for a living? 13:51
huggable ZoffixWin, I am a factoid bot on #perl6
ZoffixWin huggable, what do you do for a living?
huggable ZoffixWin, nothing found
ZoffixWin THAT, I agree with :)
13:51 brrt joined
AlexDaniel ZoffixWin: instead of saying “nothing found”, why not find something that matches? 13:51
ZoffixWin AlexDaniel, that's already planned: github.com/zoffixznet/perl6-IRC-Cl...ementation 13:52
psch safe the factoids under (tr/ //).fc maybe
AlexDaniel ZoffixWin: so why do you disagree? :)
ZoffixWin huggable, int
huggable ZoffixWin, class int [Integer (native storage; machine precision)]: doc.perl6.org/type/nativeInt
ZoffixWin huggable, Int
huggable ZoffixWin, class Int [Integer (arbitrary-precision)]: doc.perl6.org/type/Int 13:53
ZoffixWin psch, ^ :) that would be a bit problematic
psch m: say "int".fc eq "Int".fc
camelia rakudo-moar 6df053: OUTPUT«True␤»
psch right
psch misremember what exactly foldcase does
timotimo psch: i would only s/ ' '+/ /, tbh
AlexDaniel well so what? if “int” exists then print it, if it does not then print “Int”
timotimo foldcase is a unicode-prescribed transformation that's supposed to be used for case-insensitive comparison
psch huggable: ^search work 13:54
huggable psch, nothing found
psch huggable: ^search living
huggable psch, nothing found
ZoffixWin NIY
Nothing of that design is implemented except for basic storage-retrieval
psch huggable: NIY :is: NYI 13:55
huggable psch, Added NIY as NYI
psch :P
timotimo m)
dalek c: 1a58d1a | (Jan-Olof Hendig)++ | doc/Type/Submethod.pod:
Removed the flurb method
psch resists the urge to try and overwrite the Int factoid 13:56
AlexDaniel psch: overwrite with what?
ZoffixWin You'll succeed.
psch AlexDaniel: well, the impulse was "you probably wanted documentation, eh? too bad" 13:57
AlexDaniel huggable: int
huggable AlexDaniel, class int [Integer (native storage; machine precision)]: doc.perl6.org/type/nativeInt
AlexDaniel huggable: Submethod
huggable AlexDaniel, class Submethod [Member function that is not inherited by subclasses]: doc.perl6.org/type/Submethod
AlexDaniel cool
huggable: ∘ 13:58
huggable AlexDaniel, nothing found
AlexDaniel huggable: o
huggable AlexDaniel, nothing found
psch huggable: infix:<o>
huggable psch, nothing found
13:58 nd3i left
AlexDaniel huggable: huggable 13:58
huggable AlexDaniel, nothing found
psch huggable: huggable :is: did you mean recursion? 13:59
huggable psch, Added huggable as did you mean recursion?
AlexDaniel huggable: huggable :is: github.com/zoffixznet/perl6-IRC-Cl...in-Factoid
huggable AlexDaniel, Added huggable as github.com/zoffixznet/perl6-IRC-Cl...in-Factoid
AlexDaniel :P
psch blames the r-j compiliation process
AlexDaniel after all, man man, help help and stuff like that works 14:00
14:01 mnick left 14:03 ZoffixWin left 14:06 rindolf left, ZoffixWin joined, ZoffixWin left, ZoffixWin joined
ZoffixWin What's a sane way to create a bunch of variables from a hash? say I have %h = foo => 'bar'; meow => 'moo'; *magic happens*; say "$foo $meow"; # prints "bar moo": 14:07
Actually... I'm already evaling stuff... What's a way to test whether a key can be used as a variable name? :P 14:09
14:09 sjn left 14:10 sjn joined 14:11 rindolf joined
ZoffixWin m: say so 'varₓ' ~~ /^ <:L> [<:L> | <:Nd> | <[_'-]> ]+ $/; 14:13
camelia rakudo-moar 6df053: OUTPUT«True␤»
ZoffixWin will use that
moritz - and ' can only be used between word characters
ZoffixWin Thanks
moritz m: my %h = a => a, b => 2; my (:$a, *%) := %h; say $.perl 14:14
camelia rakudo-moar 6df053: OUTPUT«5===SORRY!5=== Error while compiling /tmp/aYCr4zFn8I␤Variable $.perl used where no 'self' is available␤at /tmp/aYCr4zFn8I:1␤------> 3, b => 2; my (:$a, *%) := %h; say $.perl7⏏5<EOL>␤ expecting any of:␤ argument list␤ …»
moritz m: my %h = a => a, b => 2; my (:$a, *%) := %h; say $a.perl
camelia rakudo-moar 6df053: OUTPUT«5===SORRY!5=== Error while compiling /tmp/9U_lB2fAk4␤Undeclared routine:␤ a used at line 1␤␤»
ZoffixWin m: my %h = a => 'a', b => 2; my (:$a, *%) := %h; say $a.perl
camelia rakudo-moar 6df053: OUTPUT«"a"␤»
ZoffixWin :o
That's insane.
14:14 tharkun left
ZoffixWin moritz++ 14:15
psch note that you already know the key there
ZoffixWin Yeah, I just realized that :P 14:16
14:16 tharkun joined 14:25 zwu joined
zwu for NativeCall, what is a way to attach to C memory? For example, C returns a pointer to an array and a size, how can Perl 6 create CArray to attach on the memory for a view? 14:27
timotimo you can use nativecast for that task 14:29
like nativecast(CArray[int], $returned_value)
zwu Just read that Typed Pointers should work for the memory attachment,
timotimo it might be more wordy, i don't think Pointer has a postcircumfix:<[ ]> that works with it 14:30
it'd be cool if it did, though
then you can write pointer arithmetic like in C. which we all know is totally a good idea!
zwu right, I just found the example for Pointer[int32] and Pointer[MyCstruct] 14:31
timotimo right, Pointer[MyCStruct 14:32
] is probably what you'll have to use if the array returned from C has all the structs in-lined, rather than pointed at
zwu really like the idea of the postcircumfix:<[ ]> to pointer! 14:35
timotimo why not have a stab at implementing it, then? :) :)
you can look at how CArray works, you'll probably have to apply the Positional role and implement AT-POS and BIND-POS 14:36
BBIAB 14:38
14:41 khw joined 14:49 ZoffixWin left 14:54 dalek left 14:55 dalek joined, ChanServ sets mode: +v dalek, ecocode joined 14:56 ZoffixWin joined, ZoffixWin left, ZoffixWin joined 15:03 skids joined
zwu I would love to help though I'm not an expert in perl. 15:07
15:12 Relsak joined
timotimo you might be surprised at how easy it can be to hack on core stuff. i'm willing to give guidance 15:12
but you're free to refuse, of course
15:13 ecocode left, BenGoldberg left, ecocode joined, BenGoldberg joined 15:16 richi238 is now known as richi235
BenGoldberg m: class A { submethod b { 42 }; method c { self!b } }; A.c(); 15:17
camelia rakudo-moar e8fd55: OUTPUT«5===SORRY!5=== Error while compiling /tmp/xE1Lu2VzUE␤No such private method 'b' for invocant of type 'A'␤at /tmp/xE1Lu2VzUE:1␤------> 3A { submethod b { 42 }; method c { self!7⏏5b } }; A.c();␤»
BenGoldberg m: class A { submethod b { 42 }; method c { self.b } }; A.c(); 15:18
camelia ( no output )
BenGoldberg m: class A { submethod b { 42 }; method c { say self.b } }; A.c();
camelia rakudo-moar e8fd55: OUTPUT«42␤»
ZoffixWin Ah \o/ reworked how stash vars are used in template in Bailador::Template::Mojo::Extended . Now I'm happy with it :) 15:21
15:22 wamba joined, vike1 left 15:23 RabidGravy joined 15:29 _nadim joined 15:33 zwu left 15:41 vike joined
RabidGravy beware the goars 15:43
goars?
mst goats?
timotimo groots 15:45
geekosaur groats?
tadzik ooh, ::Extended 15:48
consider me interested :)
nice :) 15:49
timotimo i can't seem to build a fresh rakudo 15:50
Could not find symbol '&ssize_t'
what am i missing?
15:55 espadrine_ joined, espadrine_ is now known as espadrine 15:56 n0xff left
RabidGravy I did see that PR go in 15:57
timotimo well, rm -rf install "fixed" it
so yay
RabidGravy HARR
timotimo harr? 15:58
yo RabidGravy, what's your process for making/coming up with melodies?
i've been told i have to learn about harmonies first before i can melodize 15:59
RabidGravy that's a bit strict to be honest 16:00
16:00 ecocode left
RabidGravy as far as the stuff I typically make it might be a trick question as I may eschew melody altogether 16:01
16:01 shibly joined, espadrine_ joined, espadrine left, shibly left
tbrowder perl6-examples: Travis is failing looking for some modules 16:02
RabidGravy but typically it's select a mode around the key of the thing and doodle until it sounds right
psch usually puts a few notes with different length in a row
often that sounds meh, but sometimes it's ok :P 16:03
timotimo "mode"?
psch that's more of that music theory stuff 16:04
16:04 espadrine_ is now known as espadrine
psch ionian, lydian and so on 16:04
timotimo the what now :)
i have never heard of those terms before
RabidGravy I may spend hours working up something with just "noise" processed in various pleasing ways and then just add a few actual notes to hook in the people that don't get minimal ;-)
timotimo en.wikipedia.org/wiki/Ionian_School_(music) - this the right one? 16:05
it's not telling me much, that's for certain
psch oh, yeah, no, ionian is actually wrong i think :S
ah, no 16:06
i don't really know them by heart...
en.wikipedia.org/wiki/Mode_(music)...an_.28I.29
mode kinda are the same as scales, but not really
*modes
timotimo there's not even a german page on modes in music
psch ...or was it keys that are kind of the same but not really?
i get that confused sometimes 16:07
timotimo and mode is a ridiculously general term
psch de.wikipedia.org/wiki/Modale_Tonleitern
timotimo ah
16:07 ecocode joined
tbrowder is there an easy way to update modules on the travis server? 16:08
timotimo you'd invoke panda for that; what are you missing? 16:09
ugexe travis has nothing to do with modules 16:10
timotimo shouldn't it grab the newest versions of all modules anyway?
because it build from scratch every time ...
16:11 rurban joined
ugexe well its requesting to install a 'Web::Request' which does not exist 16:13
RabidGravy if you look in the examples of Audio::PortMIDI then there is a set of definitions of the modes, and I think psch's module has too 16:14
psch wonders if the repeated music discussion (and the GH org) warrant a dedicated channel 16:16
dalek c: 490d255 | (Jan-Olof Hendig)++ | doc/Language/setbagmix.pod:
Fixed typo in operator name, i.e. (+>) to (>+)
tadzik I'd join #perl6-music :)
psch i was thinking #perl6-noise-gang, cause that the GH org vOv
RabidGravy do it 16:17
mst I like #perl6-noise-gang 16:18
ugexe its because supernovas changed his meta.info to meta6.json (for distribution named 'Web') and did not update th ecosystem
mst give me a shout if you manage to botch the channel registration and I'll help fix it for you
(or if you just have questions about managing an IRC channel)
psch mst: fwiw, ChanServ links to freenode.net/channel_guidelines.shtml which is apparently parked 16:19
ah, no, freenode.net just has a search
freenode.net/changuide is the right url 16:20
RabidGravy still none the wiser :) 16:21
mst psch: thanks, filed as github.com/freenode/web-7.0/issues/212
RabidGravy isn't it just "/join #channel" and it makes the channel? 16:22
psch RabidGravy: yeah, but ChanServ does stuff vOv
16:22 loren left
RabidGravy ah right 16:22
psch i do like the habit of not having ops around for example, like we do here 16:23
mst so long as the name matches /^#perl6-/ I can get staff to fix pretty much anything you can do wrong to it
I would, however, prefer to answer questions in advance so I don't have to do that 16:24
psch yeah, i'm reading chanserv help and changuide right now
mst main thing: make sure you've got more than one person with a full set of flags
psch yup, i was planning on that :)
mst oh, and you probably may as well give me full access, since (a) I *effectively* have that anyway (b) if something goes wrong I can fix it faster that way 16:25
psch mst: well, i gave you +F. not sure if i need to add e.g. +e manually too 16:30
mst that'll be fine 16:31
that's enough for me to bootstrap myself up to whatever else if you need me :)
16:35 BenGoldberg left 16:36 BenGoldberg joined 16:44 itcharlie joined 16:52 Ven joined 16:55 BenGoldberg left
Xliff_out Heya, #perl6 16:55
Xliff What type is best for storing binary data that will grow with time?
16:55 n0xff joined
Xliff Blob/Buf or does it matter? 16:56
grondilu Buf is the mutable version
ZoffixWin Buf. I think Blob is immutable
Xliff OK. So how do you append to a Buf?
m: my $b = Buf.new(1, 2, 3, 4); $b.push(5, 6, 7, 8); say $b 16:57
camelia rakudo-moar e8fd55: OUTPUT«Buf:0x<01 02 03 04 05 06 07 08>␤»
Xliff Nevermind. ;)
ZoffixWin :)
Xliff m: my $b = Buf.new(1, 2, 3, 4); my $c.Buf.new(5, 6, 7, 8); $b.push($c); say $b;
camelia rakudo-moar e8fd55: OUTPUT«Method 'Buf' not found for invocant of class 'Any'␤ in block <unit> at /tmp/bdRm_S03uy line 1␤␤»
Xliff m: my $b = Buf.new(1, 2, 3, 4); my $c = Buf.new(5, 6, 7, 8); $b.push($c); say $b;
camelia rakudo-moar e8fd55: OUTPUT«Buf:0x<01 02 03 04 05 06 07 08>␤»
ugexe m: say Buf.new(1,2,3) ~ Buf.new(4,5,6); my $buf = Buf.new(1,2,3); $buf ~= Buf.new(4,5,6); say $buf 17:01
camelia rakudo-moar e8fd55: OUTPUT«Buf:0x<01 02 03 04 05 06>␤Buf:0x<01 02 03 04 05 06>␤»
17:08 ecocode left 17:09 xnrand is now known as they, ecocode joined 17:10 they is now known as xnrand 17:17 wamba left
Xliff ugexe, is ~ more appropriate than .push? 17:18
masak Xliff: that sounds like the kind of question whose answer is benchmarking -- at least if you care about it being fast 17:19
Xliff Heh.
Fair enough.
Hmmm... Audio::OggVorbis isn't showing on modules.perl6.org. When does that refresh? 17:21
arnsholt ~ and .push have subtly different semantics too, don't they?
~ making a new object, while .push modifies its invocant?
Xliff Which means I'd want .push
17:21 TEttinger joined
psch m: say Buf.new(1,2,3) ~ Buf.new(4,5,6); my $buf = Buf.new(1,2,3); say $buf.WHERE; $buf ~= Buf.new(4,5,6); say $buf.WHERE; $buf.push(Buf.new(7,8,9)); say $buf.WHERE 17:22
camelia rakudo-moar e8fd55: OUTPUT«Buf:0x<01 02 03 04 05 06>␤140323522109888␤140323522118904␤140323522118904␤»
AlexDaniel why can't ~= act like .push? 17:23
isn't it something the compiler has to think about instead of me?
17:24 zuchel joined
Xliff m: use Benchmark 17:24
camelia rakudo-moar e8fd55: OUTPUT«===SORRY!===␤Could not find Benchmark at line 1 in:␤ /home/camelia/.perl6␤ /home/camelia/rakudo-m-inst-2/share/perl6/site␤ /home/camelia/rakudo-m-inst-2/share/perl6/vendor␤ /home/camelia/rakudo-m-inst-2/share/perl6␤ CompUnit::Rep…»
psch AlexDaniel: "stringy ops work with stringy semantics, and Str are immutable" is my reasoning
(for the difference, to be clear) 17:25
masak AlexDaniel: that's not what ~= means, though. it means `$buf = $buf ~ $stuff` -- that is, replace the old one with a new one
AlexDaniel: you get what you ask for, basically
17:25 zuchel left
AlexDaniel I get that, I'm just thinking that maybe Perl 6 has to be a bit smarter than me in this case 17:25
psch but if &[~](Buf, Buf) means Buf.push(Buf) we should just not have it 17:26
AlexDaniel if ~ is not somethink that works reasonably well, why have ~ at all
mst AlexDaniel: IME the computer can be significantly smarter than me and still depressingly stupid :)
psch and ISTR that there was a great need for the ~ candidate
...i don't know how "it creates a new Buf" means "it doesn't work reasonably well" 17:27
masak AlexDaniel: it's not about "smarter", it's about consistency. that's what the ~= op does. 17:28
AlexDaniel: making it do the other thing would be inconsistent and surprising.
17:28 rurban left
AlexDaniel masak: I agree with that, yeah 17:28
17:31 BenGoldberg joined
mst AlexDaniel: also, consider, for example, overloading ~ 17:33
AlexDaniel I am not a fan of overloading
mst that's completely nothing to do with the point I'm making 17:34
17:34 CIAvash joined
AlexDaniel mst: well, you mean that I can change it myself, right? 17:34
17:34 mr-foobar joined
mst no 17:34
I mean that the fact people can write their own ~ implementations is another reason why ~= has to behave the way it does 17:35
masak ~ is already overloaded, since it works both on Str and on Buf ;)
17:36 n0xff left
Xliff Odd. When doing this "my $b = Buf.new(@c); my $t2 = timethis(1000, sub { $b.push(@c); }); dd $t2;" via "perl6 -MBenchmark -e" 17:38
17:39 huggable left
Xliff I get this: WARNING: unhandled Failure detected in DESTROY: 17:39
Type check failed in push to Buf; expected uint8 but got Array ($[0, 1, 2, 3, 4, 5, 6...)
17:39 huggable joined
Xliff Benchmark bug? 17:39
Nope. 17:40
m: my @c = ^100; my $b = Buf.new(@c); $b.push(@c) 17:41
camelia rakudo-moar e8fd55: OUTPUT«Type check failed in push to Buf; expected uint8 but got Array ($[0, 1, 2, 3, 4, 5, 6...)␤ in block <unit> at /tmp/9HflsIVrrO line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/9HflsIVrrO line 1␤␤»
Xliff m: my @c = ^100; my $b = Buf.new(@c); $b.push(^50)
camelia rakudo-moar e8fd55: OUTPUT«Type check failed in push to Buf; expected uint8 but got Range (^50)␤ in block <unit> at /tmp/sVxVSLN0b6 line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/sVxVSLN0b6 line 1␤␤»
Xliff m: my @c = ^100; my $b = Buf.new(@c); $b.push(1,2,3,4,5)
camelia ( no output )
Xliff m: my @c = ^100; my $b = Buf.new(@c); $b.push(^50.flat)
camelia rakudo-moar e8fd55: OUTPUT«Potential difficulties:␤ Precedence of ^ is looser than method call; please parenthesize␤ at /tmp/0ZnSRgXEYi:1␤ ------> 3= ^100; my $b = Buf.new(@c); $b.push(^507⏏5.flat)␤Type check failed in push to Buf; expected uint8 but got Range…»
Xliff So argument to Buf.push needs to be Buf 17:42
mst Xliff: the goal here, I believe, is to make sure that you never confuse data intended for a Buf and data intended for a Str, because being pedantic about that makes it much harder to give yourself encoding nightmares 17:43
Xliff kk 17:44
So if I'm reading Benchmark output correctly, .push() is the way to go. 17:46
Thanks!
17:46 sena_kun left 17:47 zakharyas joined 17:54 mr-foobar left 17:56 brrt left
psch anyone around who can teach dalek to join #perl6-noise-gang? 17:57
diakopter: ping
18:06 skids left 18:07 yasnothil left 18:10 zakharyas left 18:11 zakharyas joined 18:13 mspo left 18:20 zakharyas left 18:21 tokomer joined, tokomer left 18:22 tokomer joined 18:27 ecocode left, zacts left, zacts joined 18:29 tokomer left
moritz psch: I think I can 18:30
psch moritz: that'd be nice :) 18:31
18:31 dalek left, dalek joined, ChanServ sets mode: +v dalek 18:32 dwarring joined
psch moritz: thanks! 18:32
moritz psch: no problem
ZoffixWin m: my @c = ^100; my $b = Buf.new(@c); $b.push(^50.flat) 18:39
camelia rakudo-moar e8fd55: OUTPUT«Potential difficulties:␤ Precedence of ^ is looser than method call; please parenthesize␤ at /tmp/RM3BVce53H:1␤ ------> 3= ^100; my $b = Buf.new(@c); $b.push(^507⏏5.flat)␤Type check failed in push to Buf; expected uint8 but got Range…»
ZoffixWin m: my @c = ^100; my $b = Buf.new(@c); $b.push(| ^50);
camelia ( no output )
18:39 ufobat joined, ecocode joined
ufobat ola :) 18:39
ZoffixWin \o 18:40
Ven o
o/*
ZoffixWin ufobat, I sent you a PR to let my new baby work: github.com/zoffixznet/perl6-Bailad...o-Extended
timotimo is there a clever trick to select colors across the hue range when you don't know up front how many colors you'll need, but you want the result to be as well spread out as you can?
ufobat ZoffixWin, it's just merged in :)
i wanted to ask what you're doing since i am happy about your contributions 18:41
ZoffixWin ufobat, I see your comment 'Thank you', but no merge: github.com/ufobat/Bailador/pull/69
ufobat oups, thanks 18:42
ZoffixWin Yey \o/ ufobat++
ufobat now it is!
timotimo oh, golden ratio, eh?
ZoffixWin ufobat, as for what I'm doing, I'm putting up perl6.party that for a start will just be a place to host my blog posts with browser-runnable examples. 18:43
ufobat cool :-)
ZoffixWin And I figured may as well make it Powered By Perl 6(tm)
ufobat of course!
woudn't it be odd if you would run it on something else? 18:44
my wife just drove me around in the car for an hour so i wrote some more documentation \o/
18:45 TEttinger left 18:46 ecocode left
ZoffixWin I wouldn't say "odd," it's just cooler if it is. 18:47
modules.perl6.org is powered by Perl 5, after all.
18:48 molaf left 18:49 BenGoldberg left, ecocode joined 18:50 BenGoldberg joined
ufobat yeah perl5 is great of course. but if you have a website about how awesome perl6 is and someone suddenly realizes "oh yeah.. but your website is powered by JSF" ;) 18:51
i am just babbling :)
18:52 jjido joined
mst ufobat: I feel like we'd be better off going "yeah, cos use what works" 18:53
ZoffixWin mst++ 18:54
Especially when it's really MultiMarkdown + Perl 6 + SASS + JS + CSS Framework + JS Framework 18:56
Perl 6 isn't the only thing that's making the site possible :)
+ Glot.io :P 18:57
ufobat mst, right from the technical point of view. But if it's about transporting enthusiasm to others perl6 should be on your "list" somehow 18:58
mst ufobat: if it's about transporting enthusiasm to others working on things that show off perl6 rather than half-arsed reimplementations of things that already exist is probably a better use of your time 18:59
(also over a long enough timeline things bitrot and then you have the horrible situation where, politically, making it not-$lang even though nobody has time to maintain the $lang implementation becomes a nightmare) 19:00
19:04 kaare_ left, kaare_ joined
ufobat i am not sure if i get the point. for me it reads as like "don't put anything in the ecosystem that's available and maybe better somewhere else" 19:07
19:08 rindolf left, yqt joined 19:09 domidumont left
nine_ ufobat: the website is powered by JSF....and apache...and Linux and maybe Cisco routers. Why exactly is it important that the one component of the stack is written in Perl 6? 19:11
ufobat it's not important form a technical side
it's like in "eat more vegetables! but i am just eating pizza, pasta and drinking beer" 19:12
it sounds better if you eat pizza, paste and eat vegetables 19:13
mst and zoffix does use perl6. just not for that site's code. 19:14
19:14 espadrine left
mst I don't really see why you're being so dogmatic about it 19:14
this is the opposite of TIMTOWTDI to my mind
masak what mst said 19:15
in 2016, one still has to compromise to be a Perl 6 user
of course more Perl 6 every would be Verra Nice
ugexe i only use artisinal bespoke perl6 crafted tcp packets
ufobat well i don't think i am dogmatic about it. I dont want anyone to eat more vegetables, i am sorry if if it was understood like that
mst and from over a decade of cat herding stuff related to perl5 infrastructure, trying to insist on things being in perl5 has repeatedly slowed us down in ways that, in hindsight, were bloody silly 19:16
19:16 kaare_ left
masak dogfooding can be fantastic 19:16
19:17 damnlie joined
masak but I don't understand the desire to replace everything with Perl (5 or 6) just for the sake of... purity or something 19:17
mst like the perl5 wiki itself essentially dying because it was based on socialtext and there isn't really an actively developed modern perl5 wiki but moving it to mediawiki kept being met with "BUT THAT ISN't PERL!!!1111"
masak Perl has always been a glue language, both technically and socially
ufobat i just think if you want express your enthusiasmn about something you should do more then just talk. ( and everyone here is doing it. so i am not complaing of being dogmatic )
masak ufobat: I understand that you don't mean it that way, but it *sounds* like you're being normative 19:18
ufobat no no no sorry!!
mst and zoffix is writing modules as well, so he's already doing much more than 'just talk', and yet you criticised him for using something that worked for his site
ufobat no sorry 19:19
19:19 buharin joined
buharin hi 19:19
:)
mst I know you mean well, but the attitude/approach has proven time and time again to cause a lot more harm than good so I'm basically going 'aaaaaaaaa no please not again'
ufobat i didn't want to critisice him! i am really sorry if anyone understood it like that
19:20 kaare_ joined
mst yeah, it's ok, but having mentioned the previous problems, I hope you can see why I respond like that :) 19:20
RabidGravy and no-one has written a mailing list manager yet 19:23
ufobat i understand the problems that where mentioned. and it makes sense to use things that work. but .. yeah... if you blog about perl6 and a part of your blog uses perl6 somehow it just emphasizes your ideas 19:24
thats what i mean
19:25 zakharyas joined
ufobat ZoffixWin, sorry if you felt offended. 19:25
ugexe if you write all the parts yourself maybe... otherwise it only really emphasizes not using the right tool for the job
19:26 Herby_ joined
Herby_ o/ 19:26
19:29 mr-foobar joined 19:34 mr-foobar left, mr-foobar joined 19:40 CIAvash left
psch does panda have a "update installed modules"? 19:41
ufobat in the end there is a difference between building something professional, that is designed to be lange or designed to be very robust and should be great right from the start and building something small, maybe for your own, or for others, maybe for just fun. And doing things again but differently might lead to good new ideas or evolve to become "big", "robust" or cool. 19:42
so i feel like "emphasizing that you are not using the right tool" it not always negative 19:43
or should not be negative for all cases 19:44
psch, panda --force install <module> does the job, i think
mst in practice any time you find yourself saying 'should' about somebody else's contributions to the community, you're implicitly being critical of anybody who chooses to contribute in a way you don't approve of
19:45 ecocode left
psch ufobat: i think that's more for "reinstall an already installed module even if it has the same version" 19:45
ufobat psch, but if there is a new version its updated... you mean updateing everthing that is already installed?
19:46 ecocode joined
psch ufobat: yes, with "update installed modules" i mean "updateing everything that is already installed" :) 19:46
ufobat psch, i thought a single installed one
psch ufobat: no problem, i actually have another problem first i noticed :) 19:47
ufobat mst, if you say should to to someone else contributions, yeah i fully agree. but thats not what i mean
mst ufobat: but it's what you *said*
"if it's about transporting enthusiasm to others perl6 should be on your 'list' somehow" 19:48
that is an implicit criticism of anybody who tries to do it another way 19:49
ufobat yeah, but thats what i am sorry for, thats what i didn't want to say like that
i never ment to criticize someone 19:50
grondilu I solve a new task on Rosetta Code. I used my Clifford module. The computation only took an hour ;P ! rosettacode.org/wiki/Multiple_regression#Perl_6 19:51
*solved
masak ufobat: I've been blogging with a Perl 6 static web page generator since 2010. I'm very glad the generator itself is in Perl 6. but the Markdown module is still a Perl 5 one. 19:53
ufobat: it was a pragmatic decision back in 2010, and it still kinda is. maybe this year I'll get to replace it with a Perl 6 CommonMark thing.
ufobat is just like "hey i'd like to blog something about perl. some part of my blogging is in perl because i like to do it that why and it makes fun to me" then everything fit's fine. and then its not a using the wrong tool just you didn't choose to run on a <great software> running on <great hardware> behind that <awesome firewall> and so one
ZoffixWin ufobat, nah, I wasn't offended or felt you were criticizing :) 19:54
ufobat thanks :)
masak ufobat: between Rakudo and nqp, there are still dozens of Perl 5 scripts powering everything from the build process to code generation. 19:56
ufobat: some would say that's horrible, or hypocritical, or weak. I say it's a time-saver, and a good use of a proven tool.
ufobat absolutly!
masak some day in the distant future, it might be a cool project to rid Perl 6 of its Perl 5 dependencies 19:57
but that's hard, for various reasons
ufobat but if someone writs a little not yet cool replacmenet, because he feels like doing so, and in the end it might a nice cool replacment
masak and there won't be a big win coming from it, except some kind of feeling of independence from the elder sibling
s/except/merely/ 19:58
heh. "It outsources Markdown generation to Perl 5. This may or may not be a temporary solution." -- strangelyconsistent.org/blog/dog-fo...rl6-flavor (6 years ago) 19:59
ufobat i think my point is really not being dogmatic or normative.... i am just meaning the opposite of "stating that something is cool and avoiding using it".
masak ufobat: what you wanted to say is "gee, I love it when people dogfood stuff!" :D 20:00
ufobat yeah maybe :) 20:01
masak "you go, dogfooders!" :P
mst ufobat: the trick is to phrase it as "I love it when X" rather than "people should X" 20:02
ufobat: that makes it a positive statement of your thing rather than being normative 20:03
ufobat: you can't claim to not be normative while using the word 'should', basically. I've found over the years that most sentences that contain the word 'should' become better sentences once you rephrase them
ufobat i'll take that for a good advice
psch .oO( people should say "i love it when..." ) 20:04
mst psch: trout.me.uk/seriously.jpg
ufobat i should think more before i write! 20:05
20:05 ZoffixWin left
[Coke]_ ~~ 20:06
20:06 [Coke]_ is now known as [Coke]
masak handwave, [Coke] 20:09
the best thing I've ever heard about the word "should" is Guy Steele's quip that "should" is a lambda expression taking one parameter: the context in which applies
in which it* applies 20:10
20:10 jjido left 20:11 mr-foobar left
[Coke] masak: blank stare. 20:11
masak [Coke]: uncomfortable look back.
20:11 jjido joined 20:12 mr-foobar joined
masak .oO( reassessment of the wisdom to speak only in nouns ) 20:13
mst First they came for the verbs, and I said nothing because verbing weirds language. Then they arrival for the nouns, and I speech nothing because I no verbs 20:14
20:16 skids joined 20:17 pmurias joined
masak is reminded of www.panix.com/~tehom/allnoun/allnoun8.faq 20:17
20:23 dolmen joined
ufobat i feel like needing a rest after this discussion.. and being invited on a party last night :-) good night everyone 20:26
masak 'night, ufobat++
20:27 kaare_ left 20:32 TEttinger joined, ufobat left
grondilu still trying to use Rakudo* on Windows. First it complained about git missing, so I installed it from git-scm.com, but now: gist.github.com/grondilu/f03f96622...41cc739d90 20:33
timotimo it's helpfully not telling you what it tries to run
unless it does, and it's trying to invoke ""
20:39 mspo joined 20:41 sortiz joined, n0xff joined
pmurias masak: re Perl 5 dependency of Perl 6, isn't the part that's hard to get rid of Configure? 20:42
20:44 darutoko left
sortiz \o #perl6 20:44
RabidGravy changes META6 in a non-compatible way - waits for the deafening silence
masak pmurias: most likely, yes. 20:48
pmurias: having a Configure written in nqp ought to work, since there's a stage-0. 20:49
timotimo aye
if we distribute moar binaries, too, we can even have moar's Configure in nqp :P
tbrowder travis: still seeing travis build failure; I notice it's using rakudo 2016.01, why not rakudo 2016.04? 20:50
grondilu m: subset Even of Int where * %% 2; role A { multi method f(Even $) {...} }; class :: does A { multi method f(Even $) {} };
camelia ( no output )
grondilu ^the above did not work on rakudo* on Windows. Not sure if it's because of rakudo* or or windows. 20:51
dalek pan style="color: #395be5">perl6-examples: 4d06b5b | (Tom Browder)++ | Makefile:
make the Makefile friendlier; remove colons from options (to avoid confusion)
pan style="color: #395be5">perl6-examples: d41f572 | paultcochrane++ | Makefile:
Merge pull request #42 from tbrowder/master

make the Makefile friendlier; remove colons from options (to avoid co…
sortiz grondilu, known problem with Readline on windows.
grondilu star: subset Even of Int where * %% 2; role A { multi method f(Even $) {...} }; class :: does A { multi method f(Even $) {} };
camelia star-m 2016.01: OUTPUT«5===SORRY!5=== Error while compiling /tmp/m5kVRKnLTu␤Multi method 'f' with signature :(<anon|81244144> $: Int $ where { ... }, *%_) must be implemented by <anon|81244144> because it is required by a role␤at /tmp/m5kVRKnLTu:1␤»
grondilu star: say $*VERSION 20:52
camelia star-m 2016.01: OUTPUT«Dynamic variable $*VERSION not found␤ in block <unit> at /tmp/UnoeORylvO line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/UnoeORylvO line 1␤␤»
grondilu star: say %*VM<VERSION> 20:53
camelia star-m 2016.01: OUTPUT«Dynamic variable %*VM not found␤ in block <unit> at /tmp/CYOcOOa6jh line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/CYOcOOa6jh line 1␤␤»
20:53 buharin left
grondilu can't ever remember the name of the var containing the version number 20:53
star: say $?VERSION
camelia star-m 2016.01: OUTPUT«5===SORRY!5=== Error while compiling /tmp/PW6cMhs7bU␤Variable '$?VERSION' is not declared. Did you mean 'Version'?␤at /tmp/PW6cMhs7bU:1␤------> 3say 7⏏5$?VERSION␤»
20:54 zakharyas left 20:55 jjido left
grondilu star: say $*PERL.perl 20:55
camelia star-m 2016.01: OUTPUT«Perl.new(compiler => Compiler.new(id => "34C40ED80206941C844106C0BECEBD0BCE0E58E3.1457701770.58478", release => "", codename => "", name => "rakudo", auth => "The Perl Foundation", version => Version.new('2016.1.1'), signature => Blob, desc => Str), name => "P…»
grondilu on rakudo.org/downloads/star/ I thought I would get the april edition. Well not so much apparently. 20:56
20:56 Emeric left
grondilu star: say $*PERL.compiler.version 20:57
camelia star-m 2016.01: OUTPUT«v2016.1.1␤»
grondilu m: say $*PERL.compiler.version
camelia rakudo-moar a16f0a: OUTPUT«v2016.04.74.ga.16.f.0.a.4␤»
grondilu meh
sortiz tbrowder, see github.com/perl6/perl6-examples/bl...vis.yml#L6
tbrowder, keep '2016.01' is important for compat, but 'latest' can be added, thought. 21:00
timotimo keep 2016.01 weird!
sortiz timotimo, ?? 21:03
21:06 wamba joined
timotimo nothing, just a random connection in my brain 21:07
21:09 Ven left 21:18 sue joined 21:24 stmuk joined 21:33 espadrine joined 21:34 pmurias left, dolmen left 21:35 mr-foobar left
masak 'night, #perl6 21:35
timotimo gnite masak
our masaked hero
21:36 mr-foobar joined
RabidGravy toodles 21:36
21:36 john51_ left 21:37 yasnothil joined, john51 joined 21:39 huggable left, huggable joined 21:41 ecocode left, cpage_ joined, ecocode joined 21:52 cpage_ left, wamba left 21:54 Relsak left 22:00 mr-foobar left 22:02 mr-foobar joined 22:05 sufrostico joined
ugexe zef has a prototype 'upgrade' command now to upgrade installed modules (inspired by psch's earlier comment) 22:05
BooK I see the core Version.pm has issues with git version numbers 22:10
22:15 cpage_ joined
timotimo m: say $*VM.version.perl 22:16
camelia rakudo-moar a16f0a: OUTPUT«v2016.04␤»
timotimo oh, er
RabidGravy BooK, that's kind of a trick question as git doesn't have version numbers 22:20
it does encourage the use of v on the release tags which don't work too great 22:21
22:23 ecocode left
RabidGravy but 22:23
m: say Version.new("645c96b72112adf30b785e05fa71db081b1cebed") 22:24
camelia rakudo-moar a16f0a: OUTPUT«v645.c.96.b.72112.adf.30.b.785.e.05.fa.71.db.081.b.1.cebed␤»
RabidGravy so yeah that sucks too
but then the only comparison that makes sense for them is equality anyway 22:27
timotimo that's right
it's not actually a version, or a number :)
AlexDaniel m: say :16<645c96b72112adf30b785e05fa71db081b1cebed> 22:40
camelia rakudo-moar a16f0a: OUTPUT«572963874799173193638282104870834911135227767789␤»
AlexDaniel timotimo: that's a great looking number :D 22:41
timotimo :P 22:42
22:42 sue left 22:43 sufrostico left
AlexDaniel interesting, what are the chances that the “highest” hash will be in the, um, let's say last year 22:43
ugexe for security reasons i would hope 0 22:44
AlexDaniel since no new commits are added in the past, you can safely assume that the “latest” version will be moving forward
ugexe: ok, good point :D 22:45
BooK RabidGravy: actually, the output of git describe is telling something about how far one is from the latest tag
in Git::Version::Compare (perl 5), I'm dropping the gdeadbeef bit 22:46
22:46 sue joined
AlexDaniel okay, that's it: github.com/rakudo/rakudo/commit/ff...4aedf70a20 22:47
RabidGravy well do feel free to make a P6 version that returns a comparable Version
22:47 firstdayonthejob left
AlexDaniel the smallest one seems to be this: github.com/rakudo/rakudo/commit/00...6d44320e05 22:48
.oO( I should probably stop this nonsense and just go to bed )
timotimo you can't compare versions in git. the concept doesn't make any sense except for equality or when shas are strictly in a row 22:49
and even then you'd have to have all sha1s in between to figure that out reliably :) 22:50
AlexDaniel but there are some interesting properties to such versioning: 1) there's no need to decide when we the release has to be made. If you want the release to happen sooner, then just commit more! 2) Over time releases will be less frequent, which kinda makes sense
/we// 22:51
psch but git refs aren't ordered..?
BooK actually, `git describe` gives number of additional commits after the tag 22:53
but because branches, that's not enough to disambiguate, hence the abbrievated sha1 after the 'g'
AlexDaniel psch: not sure if you got my crazy idea. The point is that we take the highest sha out of all commits, and treat that as the latest version of the software
timotimo if you drop the gf00f00 bit, you can just move the tag and have something else with the same version
or even just have a branching point after the tag 22:54
BooK yup
metacpan.org/pod/Git::Version::Com...ON-NUMBERS
AlexDaniel psch: new commits will most probably have sha hash that is “smaller”, but at some point a commit will pop up with a bigger hash. Hooray, we have a new release!
psch AlexDaniel: oh. so it's not supposed to be reasonable? :)
AlexDaniel psch: of course not 22:55
jast that makes it kind of important to enforce passing tests for all commits :)
also, for cheating, there's a script somewhere to bruteforce prefixes for git commit IDs 22:56
BooK although, re-reading thatm ^H^H, if^H^HI don't find it f^Hvery clear
my connection is flaky, so my messages may not make much sense :-)
timotimo i've had a bit of energy for improving the spesh grapher, but not enough to improve spesh's optimizer :\ 22:58
22:59 sufrostico joined 23:00 RabidGravy left
_nadim timotimo: did the broken grammar that core dumped tell you anything? 23:01
timotimo didn't actually even run it at all ;( 23:02
_nadim did the extra module stop you, I can remove those three lines 23:03
timotimo that'd be helpful
_nadim do you have grammar::Tracer or you want that one removed too? 23:04
23:04 kurahaupo left, kurahaupo_ joined, sufrostico left
_nadim meh! doesn't core dump without. 23:07
timotimo i need as few dependencies as possible :)
like, if you can make it crash with purely nqp code, that'd be crazy good
_nadim I removed the call to Data::Dump::Tree and it does't crash anymore. Damit! 23:08
23:14 mr-foobar left 23:15 mr-foobar joined 23:22 SHODAN joined
timotimo i've caught a flu or something ... i need to get to bed \: 23:26
23:38 wamba joined
_nadim timotimo: get well, I'll hunt a golfed version 23:39
23:42 kurahaupo_ left 23:43 TEttinger left