»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
timotimo masak: when i pass repeatedly, the AI doesn't win, it just keeps making two immediately neighbouring towers higher and higher :D 00:00
00:02 kurahaupo left 00:04 labster left 00:07 mcmillhj joined
timotimo Use of uninitialized value $github-user of type Any in string context. / in code at /home/timo/perl6/install/share/perl6/site/sources/08A003E9585BF4F4FEBD6559AE70DDC410B29DDC (META6::bin) 00:08
i'm so good at this %)
00:09 gdonald left 00:11 mcmillhj left 00:19 grondilu left 00:25 mcmillhj joined
timotimo Sub object coerced to string (please use .gist or .perl to do that) 00:28
in any dump_children at /home/timo/perl6/install/share/nqp/lib/QASTNode.moarvm line 1
:\
00:30 mcmillhj left
timotimo No such source rod 'left to right left to middle right to middle left to right middle to left middle to right left to right left to middle right to middle right to left middle to left right to middle left to right left to middle right to middle' 00:32
:D
00:35 Actualeyes left
geekosaur hanoying? 00:36
00:36 skids joined
timotimo yup 00:37
List $source = $(((("left", "to", "right"), ("left", "to", "middle"), ("right", "to", "middle")), ("left", "to", "right"), (("middle", "to", "left"), ("middle", "to", "right"), ("left", "to", "right"))), ("left", "to", "middle"), ((("right", "to", "middle"), ("right", "to", "left"), ("middle", "to", "left")), ("right", "to", "middle"), (("left", "to", "right"), ("left", "to", "middle"), ("right", "to",
"middle"))))
List $target = $(((("middle", "to", "left"), ("middle", "to", "right"), ("left", "to", "right")), ("middle", "to", "left"), (("right", "to", "middle"), ("right", "to", "left"), ("middle", "to", "left"))), ("middle", "to", "right"), ((("left", "to", "right"), ("left", "to", "middle"), ("right", "to", "middle")), ("left", "to", "right"), (("middle", "to", "left"), ("middle", "to", "right"), ("left", "to",
"right"))))
that is not a rod that is on the stack!
i should go to bed 00:40
i can continue this tomorrow
00:46 perlawhirl left, perlawhirl joined 00:56 mcmillhj joined, mr-fooba_ left
samcv what am i seeing here 00:57
Geth specs: 63686a0042 | (Samantha McVey)++ | S22-package-format.pod
Update META info for the license section

Use the SPDX license identifier under license, which is the most widely used license tags meant for interoperability.
If the license does not have an identifier, then it is recommended to add a URL to the license under the support key.
01:01
specs: 07a57f3cae | (Samantha McVey)++ (committed using GitHub Web editor) | S22-package-format.pod
Merge pull request #120 from samcv/package

Update META info for the license section
01:02 Cabanossi left 01:04 Cabanossi joined 01:06 gdonald joined 01:32 awwaiid joined
Geth specs: dce35d18d5 | (Samantha McVey)++ | S22-package-format.pod
META6.json describe how to denote no license or no assertion
01:37
samcv very good. covered all the likely cases now :) 01:38
01:46 Cabanossi left 01:49 Cabanossi joined 02:08 baest left, noganex_ joined
samcv hehe 02:09
bag(emulates(10), test-depends(301), version(813), name(813), api(4), description(813), excludes(9), resources(145), build-depends(195), supersedes(9), production(10), support(211), source-uri, licence(4), author(285), source-url(779), superseded-by(9), resource(13), tags(149), license(108), auth(155), perl(739), provides(803), meta6(54), history, repo-type(31), authors(411), owner(9), authority(8), depends(771), source-type(69)) 02:10
so we have 4 modules with misspelled license (they spelled it as licence)
and 108 of our modules have licenses plus uh. whatever those 4 that misspelled them..
02:11 noganex left
TimToady well, that's the british spelling, they spell 'defence' differently as well 02:16
and a couple other three things...
Geth ecosystem: e6a2236f61 | (Zoffix Znet)++ (committed using GitHub Web editor) | META.list
s/META.info/META6.json/
02:22
02:24 baest joined
samcv ok my json is fixed 02:25
02:47 baest left 02:51 baest joined
samcv yay i can't use the meta6 bin because 03:01
===SORRY!===
Failed to parse „~/.gitconfig“.
03:01 Cabanossi left
samcv :( 03:01
won't even start.... ugh
03:04 Cabanossi joined, rindolf joined 03:14 rindolf left 03:15 mcmillhj left 03:27 AlexDaniel left 03:32 rindolf joined 03:46 Cabanossi left 03:47 Cabanossi joined 03:49 baest left 03:54 khw left 03:58 baest joined
u-ou How do you do a runtime-mixin of an already created object? Is it just $s = $s but role? 04:05
TimToady that will clone $s 04:06
04:06 g0d355__ joined
TimToady use '$s does role' for modification in place 04:06
u-ou ahhh
thanks
04:15 rindolf left, TimToady left 04:16 wamba joined 04:17 TimToady joined 04:21 mcmillhj joined 04:25 mcmillhj left 04:32 skids left 04:45 Actualeyes joined 04:57 kurahaupo joined 05:01 Cabanossi left 05:02 Cabanossi joined 05:03 mcmillhj joined 05:11 mcmillhj left 05:18 BenGoldberg left 05:25 gdonald left 05:26 gdonald joined 05:30 mcmillhj joined 05:31 bolangi left 05:32 jroth joined 05:33 jroth is now known as bolangi 05:35 mcmillhj left 05:37 CIAvash joined 05:38 xtreak joined 05:40 lowbro joined, lowbro left, lowbro joined 05:45 wamba left 05:46 Cabanossi left, nadim joined 05:47 Cabanossi joined 05:51 RabidGravy joined 06:02 mcmillhj joined 06:06 mcmillhj left
Woodi hallo #perl6 :) 06:09
some days ago I read that Prolog requires first class backtracking. we have it in v6 ? also I learned logic programming is superset of functional programming... but not saying I understand all of this :) 06:11
06:13 gdonald left, gdonald joined
TimToady backtracking is pretty easy; it's unification that's hard 06:15
06:16 Cabanossi left 06:18 Cabanossi joined
TimToady the hardest part of logical programming, though, is managing your lexicon such that words always mean the same thing everywhere, which they usually don't in real life :) 06:21
06:22 MasterDuke_ left
TimToady funny how real life actually means "complex"... 06:23
but the opposite of "imaginary"...
06:24 wamba joined 06:32 darutoko joined 06:43 domidumont joined 06:47 mr-foobar joined, domidumont left 06:48 domidumont joined
El_Che TimToady: hehe 06:48
yoleaux 18 Apr 2017 19:52Z <Zoffix> El_Che: you're good for packages. 2017.04.2 is out
El_Che will start
06:48 nadim left 06:50 domidumont left 06:51 domidumont joined
mst TimToady: without continuations it's non-trivial to push backtracking down into the runtime though 06:52
of course the WAM doesn't really do it that way anyway
Woodi: first class backtracking at the VM level is pretty rare, and not at all a requirement to implement prolog 06:56
Woodi: hence why there's more than one implementation of prolog in perl5
Woodi: also, please call perl6 perl6, not 'v6'
06:56 sunnavy left 06:57 sunnavy joined, llfourn left
u-ou yay my perl6 thing works 06:58
samcv whew. just finished PR to META6 module and the META6::Test modules 06:59
once accepted META6::Test will be able to check license names against the full list of standardized names, and if it's not correct, tell the user to fix it, or if license isn't on the list, if you set a 07:00
URL for the license text then it will also be able to pass
and if you put NONE or NOASSERTION it will give a little warning to remind you that's nonideal 07:01
github.com/jonathanstowe/Test-META/pull/21 can we call it the Licensining? or uh. idk something catchier 07:02
07:04 llfourn joined 07:05 labster joined
gfldex samcv: it's JSON, no license should be null, not "NONE" 07:06
samcv what?
null does not convey any actual afirmative information thuogh
whether there is none or if it's just an empty field 07:07
though we could depart from SPDX on that i guess? hm
or take NONE or null as both being no license?
gfldex null indicates the absence of a value. "NONE" may very well be an acronym. 07:08
samcv abcense of a value is different from no license in the affirmative though 07:09
just how NOASSERTION means that there could be a license but the creator of the meta file does not know and makes no assertion
because no license and no assertion are distinct things
if anything License being null would be closer to NOASSERTION 07:10
07:10 nadim joined 07:12 Actualeyes left
gfldex samcv: I disagree, NOASSERTION would be indicated by not having the field in the JSON file 07:13
not sure if META6.pm6 is handleing that properly tho
samcv i we could define NOASSERTION is no license key and that license: null means NONE and they are equivilant. that would be okay with me i guess. 07:14
as long as there is a way to distinguish between those two. that is the important part. and that they are defined
i still don't like license: null.... but regardless. it still is fine to have it mean SOMETHING which is better than meaning nothing i think 07:15
gfldex aye
samcv thanks for bringing that up gfldex
gfldex besides, if there is no license it's rather pointless to put it into the ecosystem to start with 07:16
samcv well. most modules have no license....
gfldex as nobody knows if the module can be used at all
07:16 parv joined
samcv so 07:16
hopefully we can make issues with all of those modules to add a license. but it's a big undertaking
when only about 10% of them have one in the META 07:17
gfldex they don't got a license because it was never enforced
samcv well yes
gfldex I may be able to automate that problem away
samcv with travis?
gfldex githubs api is rather nice
samcv oh? to notify?
that would be nice
gfldex but first, $dayjob 07:18
samcv though it's useful now that i have META6::Test being able to check validity of license type things. may eventually want to just cause an error when no license is there. but it doesn't have to happen overnigtht. we're working toward licenses and that's the important part
nice talking to you gfldex
07:18 jian joined
gfldex yw 07:19
jian HELP
gfldex WHAT?
no wait, I can answer that myself
m: "HELP".WHAT.say 07:20
camelia (Str)
samcv hi jian
jian hi.
hmm..
samcv please let me try and help
ZofBot, HELP
jian I don't actually know what it is 07:21
samcv oh :(
gfldex can you provide some details of your undescribable problem please? 07:22
jian that is human or what ? 07:23
gfldex humans are an undescribable problem indeed :) 07:24
samcv humans are homo sapiens and are sentient beings who populate this channel in addition to the IRC bots here
see en.wikipedia.org/wiki/Human
07:24 jeek left
moritz humans are meat bags. There, I described the problem for you. 07:24
samcv en.wikipedia.org/wiki/IRC_bot
AFAIK those are the only two things in this channel 07:25
gfldex the IRC bots are smarter tho
u-ou I'm not sure what the difference is
but I think I'm a bot
just sayin'
mst moritz: this is where I turn up with an emu, right? 07:26
07:26 Actualeyes joined, jeek joined
moritz mst: an emu_lator? :-) 07:28
mst I was going for a flightless biped joke, but you do you
jian maybe I should go =-= 07:29
samcv no!
don't go
07:30 zakharyas joined
jian but I don't know how 07:30
close the tab is ok? 07:31
samcv no
i want to help!
jian help what 07:32
samcv <jian> HELP 07:33
help you
you need help
gfldex samcv: don't be so selflessly selfish!
samcv i will help
XD
jian ok
just tell me
waht is it 07:34
samcv ask your question again please
jian or what's this
moritz an IRC channel!
jian IRC? 07:35
moritz that was easy. Next customer!
samcv this in an IRC channel. this is the #perl6 room where we talk about the perl6 programming languages
moritz jian: en.wikipedia.org/wiki/Internet_Relay_Chat
gfldex I wonder if I should use subparse in Git::Config. 07:36
jian maybe I know something :-) 07:37
gfldex samcv: PR welcome (read: I'm not good with grammars.) github.com/gfldex/perl6-git-config...Config.pm6
07:38 dakkar joined 07:39 abraxxa joined
samcv what do i do with this 07:40
oh
yeah it doesn't even work if i delete my .gitconfig 07:41
so idk how to make this thing work
not sure why META bin needs it tho :\ 07:42
why does it care about my git config
gfldex samcv: try meta6 --create-cfg-dir
samcv: it needs your github username and your e-mail to generate README.md 07:43
and a few other things
parv gfldex: why is there a specific need for anything GitHub when the module name is more generic? 07:44
gfldex samcv: at some point it will ask nicely for your github PW and do anything for you. :) 07:45
parv: you can skip the github part but if you don't any new module will be pushed automatically.
parv "github" part should be easily switchable 07:46
gfldex i wanted to add automatic travis support too but failed due to travis wonky api
and so we read: "meta6 --new-module --name=<Module::Name::Here> --force --skip-git --skip-github"
meta6.cfg needs to learn a new trick it seams 07:47
gfldex takes notes
07:48 wamba left, ufobat joined 07:49 jian left
samcv gfldex, anything? :-D 07:50
gfldex the goal is to telepathically sense the general intend of the programmer and do anything automatically indeed 07:51
the same is true for Perl 6 I believe
07:55 wamba joined 07:56 xtreak left
gfldex self.dayjob() 07:56
07:56 xtreak joined 07:59 xtreak left 08:00 xtreak joined 08:01 g4 joined 08:12 eyck left 08:13 eyck joined 08:14 robertle left, xtreak left 08:15 xtreak joined, xtreak left, xtreak joined
samcv u: end 08:17
unicodable6 samcv, U+02FA MODIFIER LETTER END HIGH TONE [Sk] (˺)
samcv, U+02FC MODIFIER LETTER END LOW TONE [Sk] (˼)
samcv, 363 characters in total: gist.github.com/a2197e703ec67db1f7...1251c2aabe
08:18 eyck left 08:19 eyck joined 08:23 eyck left 08:27 Actualeyes left 08:29 rindolf joined 08:31 jonas1 joined
u-ou roles! 08:36
08:36 eyck joined 08:43 xtreak left, xtreak joined 08:47 cpage_ left 08:48 xtreak left
u-ou hi 08:50
08:52 Actualeyes joined 08:58 astj joined, astj_ left 09:02 Cabanossi left 09:03 Cabanossi joined 09:12 ChoHag joined 09:25 mcmillhj joined 09:29 mcmillhj left 09:31 labster left 09:37 kurahaupo left 09:41 mcmillhj joined 09:42 bjz joined 09:45 eiro joined, mcmillhj left 09:47 Cabanossi left 09:48 Cabanossi joined, cpage_ joined 09:50 xtreak joined 09:57 mcmillhj joined 10:05 mcmillhj left 10:12 astj left 10:13 astj joined, astj left, astj joined 10:14 notbenh left 10:16 Cabanossi left 10:17 notbenh joined, mcmillhj joined, TEttinger left 10:18 Cabanossi joined 10:19 parv left 10:21 mcmillhj left 10:27 xtreak left 10:30 xtreak joined, zakharyas left 10:32 mcmillhj joined 10:33 bjz_ joined 10:34 bjz left 10:37 mcmillhj left 10:41 astj left, astj joined 10:46 gdonald left, gdonald joined 10:48 astj left 10:50 astj joined
Geth ecosystem: 72de400aaa | (Zoffix Znet)++ (committed using GitHub Web editor) | META.list
.info -> .json
10:52
10:53 CIAvash left, mcmillhj joined 10:56 xtreak left, xtreak joined 10:58 mcmillhj left 11:01 Cabanossi left 11:02 xtreak left 11:03 Cabanossi joined 11:08 kurahaupo joined 11:22 tadzik left 11:27 noganex joined 11:29 mcmillhj joined 11:30 noganex_ left, tadzik joined 11:34 mcmillhj left 11:41 wamba left 11:58 TimToady left 12:00 TimToady joined
timotimo masak: fixing Adventure::Engine and Crypt is a fascinating experience in "how have lists changed in the last years" 12:01
also, if you %foo.push(%foo.invert), be ready to see some duplicates, because invert is lazily being iterated and is updating as the hash gets stuff pushed into it, lol.
Geth ecosystem: scmorrison++ created pull request #322:
AWS::Pricing META.info > META6.json
12:08
12:09 xtreak joined 12:10 xtreak left 12:11 kurahaupo left 12:12 zakharyas joined
Geth ecosystem: c143a14304 | (Sam Morrison)++ (committed by Zoffix Znet) | META.list
AWS::Pricing META.info > META6.json (#322)
12:14
12:15 domidumont left, Cabanossi left 12:17 domidumont joined, Cabanossi joined 12:23 japh__ joined 12:32 mcmillhj joined 12:33 MasterDuke_ joined
japh__ I'm getting "Cannot find method 'setlang' on object of type Perl6::HookGrammar" when I run perl6-debug-m (Rakudo 2017.04.2, MoarVM version 2017.04). I saw some references to this #perl6-dev on 2/28 and #perl6 on 3/21, but nothing that's a clear solution 12:35
12:36 mcmillhj left
japh__ I was getting the same set of errors under Rakudo 2017.03, so upgraded MoarVM, NQP and Rakudo to the latest tagged releases, but still no luck 12:39
jnthn I think timotimo++ was taking a look at it quite recently 12:40
nadim hi all, do we have something like Devel::Size? 12:44
12:45 salva left 12:50 japh____ joined 12:54 japh__ left 13:00 mcmillhj joined
timotimo japh__ there is a trivial fix for that error but then you get a debugger that fails at everything anyway 13:04
13:04 wamba joined
timotimo nadim, would that give you thelayout of your classes in memor or something? 13:04
i wanted something like pahole forp6opaque in the past 13:05
oh gotta go
jnthn I think the closest we have is the heap analyzer 13:07
13:23 cdg joined 13:29 salva joined 13:37 skids joined 13:40 beginner joined 13:52 gdonald left 13:53 gdonald joined 13:55 lowbro left 13:57 mcmillhj left 13:59 mcmillhj joined
beginner which is the best IDE for perl6 13:59
14:01 lizmat left, eroux joined
El_Che beginner: atom, vim, emac (non of which are IDE's, I know) 14:03
14:06 AlexDaniel joined
moritz isn't it ironic that, being immaterial, atom doesn't actually contain atoms? 14:13
while most things that contain atoms don't have atom in the name
14:15 Cabanossi left 14:18 Cabanossi joined 14:29 committable6 left 14:30 committable6 joined, ChanServ sets mode: +v committable6 14:34 khw joined
timotimo moritz: atomoritz? 14:46
SmokeMachine m: my @a = :1a, :2b, :3a, :4a; say [⊎] @a 14:51
camelia bag(b(2), a(8))
SmokeMachine m: my @a = :1a, :2b, :3a, :4a; say Bag.new ⊎ @a 14:52
camelia bag(b(2), a(8))
14:55 bjz_ left 14:56 eroux left
nadim timotimo: Data::Dump::Tree can give the layout, it can show the "perl address" but it can't show how much memory each element uses. In the P5 version I used Devel::Size to get that information. I wonder if there is some equivalent functionality in P6. 14:59
timotimo it can not give the kind of layout i'm talking about 15:00
El_Che aaaagh. Now that I build 9 different rakudo packages I hit travis' 50 minute limit
15:04 cdg left 15:06 iyra joined
iyra does anyone have knowledge of the SCGI module? I'm trying to use the example in the readme here (first under 'Usage') github.com/supernovus/SCGI 15:07
however I'm getting "Nothing given for new socket to connect or bind to" when I try to use it, seems to be caused by connect method, here: github.com/supernovus/SCGI/blob/ma...GI.pm6#L22 15:08
skids m: say Bag.new ⊎ :e_with_oatmea
camelia bag(e_with_oatmea)
15:10 nadim left
timotimo iyra: how does your code look? did you SCGI.new with a :port? 15:13
iyra yes, I used my $scgi = SCGI.new( :port(8118) ); 15:14
15:14 kurahaupo joined
iyra however I tried with my $scgi = SCGI.new( :host<localhost>, :port(8118) ); and I get the same error 15:14
15:17 Cabanossi left, Cabanossi joined
timotimo interesting 15:17
ugh, scgi doesn't have any tests 15:18
iyra using IO::Socket::INET.new in general seems to work, though (using the first example in the docs for it) docs.perl6.org/type/IO::Socket::INET 15:20
i can't think why it's not working with scgi though 15:21
timotimo it could just have bitrotted 15:22
iyra hmmm 15:23
timotimo just running one of the examples gives me errors 15:24
15:25 eroux joined
timotimo oh 15:25
iyra here's the code and full error that i got: fh.tl/CO 15:26
timotimo i know what's wrong 15:27
it's passing :listen(1), which upsets it
it must be just :listen or :listen(True)
iyra what would be the best way to make that modification locally on my computer? 15:28
I'm not familiar with perl modules :)
timotimo hold on
i'd like to know why the code had \ in these weird places to begin with 15:29
ah, it uses psgi, let's see. 15:30
ok, i think my fix is ifne 15:31
fine*
you have zef, right? 15:33
iyra yes
timotimo zef can install from git links, so try this: 15:34
zef install github.com/timo/SCGI.git
might need --force because i didn't change the version, actually
iyra do I have to remove the original scgi?
titsuki hi #perl6, I couldn't send a bug report to the RT system with email. Is there something trouble with the RT today ? 15:35
timotimo now i've bumped thev ersion
titsuki: it's supposed to work. only the ticket submission thing on the rt website has been disabled long ago because of spam 15:36
iyra: you do not, it should replace it
also, if you do it now, i actually did bump the version
titsuki timotimo: thx. I'll try it again. 15:37
iyra timotimo, that's strange, I'm still getting the same error, even after changing :listen(1) to just :listen 15:41
timotimo did zef install --force give any errors or something? 15:42
iyra no, it installed fine 15:47
I cloned the repo and made the change, then tried again with --force and zef didn't give any errors
15:48 sufrostico left
timotimo huh 15:49
can you put some very obvious thing in there, like a "say 'hello'" near the problematic code?
iyra i'll give it a go
timotimo if you want to make sure that code definitely gets used, just -I ../path/to/SCGI/lib/ 15:51
15:52 aindilis joined
iyra so I've found that even after doing zef install with my new git repo, it still uses the old sources 15:54
and using -I ../path/to/SCGI/lib/ doesn't seem to get it done either, it still references the old sources in my .rakudobrew directory
i checked by looking at the source file it was using
15:57 astj left
timotimo oh interesting 15:58
it could be it just grabs the META.info which has source-url pointing at the original repo
iyra oh, good point, i'll change that
in my fork, i changed the version number to 2.4, but zef install is still reporting that it's installing 2.3 github.com/iyra/SCGI/blob/master/META6.json 15:59
timotimo are you doing "zef install --force ."? 16:01
iyra no, I was using the my github URL; doing the command you just posted installed my new version, though 16:02
16:02 mcmillhj left
iyra oh, nice! it works 16:03
timotimo ah
iyra the :listen thing was the problem, not sure why it wasn't getting the new version from github though
thanks a lot for your help
timotimo you're quite welcome
i opened a pullrequest on SCGI, too
iyra thanks, would be nice to see it fixed 16:04
timotimo turns out there was an issue posted about this back in march
but it didn't look like anybody looked for the solution
i just now also fixed two other modules :)
Adventure::Engine and Game::Crypt
TimToady is there an easy way to do a pullrequest from the comfort of my command line, without going off to github and forking and all that?
timotimo TimToady: yup, get the "hub" program
masak: kindly look at your pull requests for Adventure::Engine and Game::Crypt :) 16:05
and yesterday i made Druid work again, too
since we're about to have a Rakudo Star release, it's a good time to be fixing ecosystem modules 16:06
i can't find the huge log samcv made where it shows what modules failed to install? 16:10
oh, is this it
hack.p6c.org/~samcv/install_log.txt
masak: you think it might be time to throw the GGE out of the ecosystem? 16:11
16:11 robertle joined 16:13 mcmillhj joined 16:16 Cabanossi left 16:18 Cabanossi joined, mcmillhj left 16:26 astj joined 16:28 domidumont left
El_Che if someone here bumps agains the travis timeout my solution is here: github.com/nxadm/rakudo-pkg/blob/m...travis.yml 16:30
16:31 mcmillhj joined, astj left 16:32 dakkar left, titsuki left 16:34 AlexDaniel left 16:35 mcmillhj left 16:45 mcmillhj joined, domidumont joined
TreyHarris TimToady: if you run into any issues just holler; I've set up hub and dealt with many issues on many different platforms for my companies' devs over the years 16:49
16:50 gregf_ left, mcmillhj left
timotimo m: say (1..10).map({ { :a($_) } }); # what is preventing the inner block from being parsed as a block? it has $_ in it … 16:51
camelia (a => 1 a => 2 a => 3 a => 4 a => 5 a => 6 a => 7 a => 8 a => 9 a => 10)
TreyHarris Erm, can't run mi6 anymore: gist.github.com/treyharris/bc343fe...98acc2da40
timotimo TreyHarris: huh, what happened to that file? it's got some powershell stuff in it? 16:52
TreyHarris I'm on ubuntu
16:52 abraxxa left
timotimo yeah but %~dpn0 is powershell 16:52
TreyHarris I installed it with zef, maybe retry w/panda? People have been telling me to use zef 16:53
timotimo no zef is right
that's literally all there is in the file 16:54
um
it's installing a windows launcher
TreyHarris the newest commits to zef are windows-related 16:55
timotimo 2 bin/ scripts [mi6 mi6.bat] installed to:
/home/timo/perl6/install/share/perl6/site/bin
but only one of the two actually exist 16:56
my bet is that it accidentally installed mi6.bat over mi6 for some reason
TreyHarris nope, no they aren't, I don't know where I saw recent win-related commits
timotimo for the time being, just clone mi6 locall, and rm bin/mi6.bat, then zef install --force .
16:56 beginner left
TreyHarris timotimo++ thanks 16:58
16:59 japh__ joined
timotimo i don't think you should ever have a bat in your bin/ 17:00
rakudo or zef or whoever makes a launcher for you if you're on windows
17:03 japh____ left, mcmillhj joined
TreyHarris does the bat file let an author on windows test directly before install time? I know nothing of windows except the see-through kind and the orbital kind 17:04
timotimo nah, you'd just perl6 -Ilib bin/mi6 in that case 17:05
oh man, having fought cmake today left me in a surpremely foul mood 17:06
TreyHarris timotimo: try dealing with an airline and aviation lawyers, I bet after that cmake would seem like a soothing hottub 17:07
timotimo i know i could always have it worse 17:08
trying to develop krita is a hundred times more annoying than contributing to pypy. and pypy has a one-hour compile cycle.
17:08 mcmillhj left 17:12 mcmillhj joined 17:15 bwisti joined, bwisti left 17:16 sufrostico joined 17:23 sufrostico left 17:27 khw left, zakharyas left 17:31 Cabanossi left 17:33 Cabanossi joined
skids bisect: role A { }; sub A ($) { 42 }; A(2).say # I thought this used to DWIM. 17:34
bisectable6 skids, Bisecting by output (old=2015.12 new=d272667) because on both starting points the exit code is 1
skids, bisect log: gist.github.com/123c5b03ba5697fccf...aa416b1ad4
skids, (2016-05-12) github.com/rakudo/rakudo/commit/33...d4efe221de
17:36 titsuki joined
skids commit: role A { }; sub A ($) { 42 }; A(2).say # I thought this used to DWIM. 17:41
committable6 skids, ¦role: «Cannot find this revision (did you mean “all”?)»
17:41 ChoHag left, jonas1 left
jnthn Probably not since the SomeTypeName(...) syntax was taken for coersions 17:42
*coercions
17:42 g0d355__ left
skids Isnn't the way to implement coercions to define a sub though? 17:42
Oh, coercive types. 17:43
timotimo nah, subs for coercions is no longer a thing
skids Wait it was postcirumfix, then subs, now what?
17:53 sufrostico joined
timotimo postcircumfix is a sub 17:57
or .. was it built on top of a postcircumfix method?
skids I think it was a method.
timotimo now coercions are just methods named the target type
which has different problems than the multi sub had
but everything that was attempted was problematic
skids Yeah, noticing that. Somewhat LTA when you are trying to emulate an enum. 17:58
18:00 xzhao left 18:02 gdonald left, gdonald joined 18:05 domidumont left 18:08 iyra left
TreyHarris Every time I see somebody use something like bisectable6 or committable6 I want to shout "this is sorcery, burn the witch!" 18:09
Though at Google I taught git to nooglers and when I got to the bisect section, I saw lots of mouths hanging open. So I guess it's just one of those things you either are used to or seem like pure magic. 18:12
moritz I'm kinda used to it, and still looks like pure magic to me :-) 18:13
speaking of like magic: I'm trying github.com/dxa4481/truffleHog right now, and I'm quite impressed so far 18:14
18:15 eroux left
moritz it found faked key material that we use in unit tests (for code that validates the format of the key material) 18:15
and remarkably few false positives
18:17 AlexDaniel joined
tony-o anyone familiar with a p6 module connecting to dynamodb? 18:18
yoleaux 23 Nov 2016 15:21Z <jnthn> tony-o: Rakudo 6bb8823 nails, best I can tell, the odd crashes I've been seeing in things using Proc::Async in chars mode.
tony-o jnthn++ 18:19
though i think we've chatted since nov :-)
timotimo perhaps in the other channel
[Coke] /win 6 18:22
18:22 ufobat left
timotimo we should have a bot that collects lines of /win N and draws a winning number at the end of each week 18:22
geekosaur .oO ( and the winner is: bal ) 18:26
El_Che new rakudo linux packages: github.com/nxadm/rakudo-pkg/releas.../2017.04.2
timotimo cool 18:29
18:41 sufrostico left, ufobat joined 18:43 sufrostico joined 18:44 iyra joined
RabidGravy Boom! 18:44
18:45 iyra left, iyra joined 18:51 lizmat joined 18:52 iyra left, iyra joined 19:00 Cabanossi left 19:02 Cabanossi joined 19:03 AlexDaniel left, AlexDaniel joined, xinming joined 19:06 xinming_ left 19:08 zakharyas joined 19:17 espadrine joined 19:25 sufrostico left 19:27 darutoko left 19:32 Cabanossi left, Cabanossi joined
timotimo how's it going, gravy sempai? 19:32
19:33 gdonald left
samcv good ** 19:33
timotimo good HyperWhatever to you, too
19:33 gdonald joined
RabidGravy Harr, samcv - I'm just testing all of the ecosystem with your licence test 19:35
timotimo samcv: i sent out three pullrequests to fix three modules in the ecosystem today :) 19:36
samcv oh nice
RabidGravy also, rather than "meta-version" there is an already specced "meta6" 19:37
samcv RabidGravy, tell me what you find. probably gonna have almost 108 failures due to licenses. since some had "Artistic" when their project had Artistic 2.0, or put Artistic2 or Artistic 2 or they have the url to the license
yeah i saw that RabidGravy
thank-
interestigly thre are 3 different artistic 1.0 licenses on the spdx website |X confusing. at least there's only one Artistic-2.0 19:41
timotimo clearly Artistic-2.0 was made because there were too many different Artistic-1.0s 19:42
samcv was that why? i thought it was other reasons
timotimo i have no idea :) 19:43
samcv :)
timotimo i thought it was a funny reference
MasterDuke_ which xkcd is that? the standards one
timotimo that one, yes 19:44
wait, there are too many xkcds about standards
we should make a better xkcd
samcv not sure why there are multiple artistic 1.0 i mean. it is 1.0 it should be the same X|
lol!
xkcd.com/927/
19:44 labster joined
samcv also seems you can view the spdx spec in at least two formats that are the names of animals spdx.org/specifications 19:45
what's up with that
RabidGravy 106 fail 19:47
samcv k. do you have that list? 19:48
or log or something, or were they all due to license failures and not other things?
19:49 japh____ joined 19:52 japh__ left
RabidGravy samcv, would have failed the licence test gist.github.com/jonathanstowe/e8b5...ab016b69d9 19:54
rindolf samcv: hi! I've collected more links like that xkcd strip here - shlomifishswiki.branchable.com/How_...e/?updated
19:54 perigrin left, perigrin joined 19:56 perigrin left 19:57 perigrin joined 19:58 lizmat left 19:59 geekosaur left
samcv making lots of PR for projects that have a LICENSE file in the repo, but no field in the META file. progress 20:00
thx for the list RabidGravy
20:00 geekosaur joined
samcv time to PR away! 20:03
RabidGravy do not send any for my modules 20:04
and it is spelled LICENCE in all of mine
samcv in the meta? or the filenames 20:06
timotimo LICHENS 20:07
samcv hehehe
RabidGravy the filename, being the correct spelling and all
samcv oh that's fine. but the meta field should be license since that's what the spec says
RabidGravy using a verb there is obtuse 20:09
samcv in the meta or in the filename or both 20:10
or *anywhere* heh
google says licence is a verb and license is a noun 20:11
www.google.com/search?q=licence&am...p;ie=UTF-8
with their dictionary thing
idk though. 20:12
timotimo samcv: license AE / licence BE
samcv i don't know what that means
this ones' license is "The Artistic License 2.0"
so many variations
MasterDuke_ it's another british/american thing 20:13
samcv the artistic, artistic, Artistic 2, Artistic-2 so many
RabidGravy license is a verb, licence a noun yes
samcv yeah but the json should probably be the standardized thing yes?
since it's processed by computers
i don't care what your license file is named, it could be licence.doc but 20:14
as long as the meta file is fine
RabidGravy and atrribute which isn't boolean should be a vnoun
maybe I should delete all my modules
samcv :\
please don't 20:15
20:16 iyra left
mst given "atrribute" and "vnoun" I'm pretty sure he'll end up typing 'mr' instead of 'rm' anyway 20:16
yoleaux 12:26Z <Zoffix> mst: would you know whether it's possible to debug Travis's IRC bot connection timeouts to Freenode? We've not seen their bot for a while and they responded to me: twitter.com/travisci/status/855034311103455232
mst so we're probably safe
RabidGravy pfft hippy 20:17
samcv vnoun sounds like a futuristic nounn
i'm going to start with all packages that depend on Test::META though RabidGravy
RabidGravy as long as you don't send me any, feel free :) 20:18
samcv :( 20:20
mst if you're going to send RabidGravy a package it generally needs to contain alcohol
RabidGravy I'm going to make it not fail unless meta6 is 1 or greater anyway
samcv you don't want any license field in your meta file RabidGravy even if you have a licence for the project? 20:21
like this github.com/jonathanstowe/JSON-Infe...META6.json you have a license github.com/jonathanstowe/JSON-Infe...er/LICENCE
RabidGravy it's an optional attribute in the spec 20:23
samcv so you don't want to do any option attributes? 20:24
jnthn Do we have an immediate use for the data?
samcv well it would be nice if 100% of packages in the ecosystem had license fields 20:25
ranguard setting standards for people to follow?
samcv and eventually it should probably be required to hav a license to get in the ecosystem
if you have a non ecosystem module fine have no license field... but if it's in the ecosystem it should have one
RabidGravy then the spec shouldn't have it as optional 20:26
samcv the meta spec? or ecosystem requirements?
or either?
RabidGravy either
samcv i mean ecosystem is different than the meta spec itself
mst is still boggling at travis
jnthn I'd think the first place to update would be the documentation that we link to from modules.perl6.org: docs.perl6.org/language/modules.ht...ng_Modules 20:27
Which at present doesn't include license in the template that, in reality, 90%+ of people will just copy/paste/fill out :)
samcv yeah i need to update the docs there
so the example has a license field. people need all the reminders they can get. cause some may copy paste the meta6.json on that page and edit it
and if it starts with no licese field, they're unlikely to add one 20:28
jnthn It's exactly what I'd do if I was new to a community and looking to submit modules
samcv yeah
RabidGravy so you are proposing removing modules without a liicense field in the meta?
samcv i am not RabidGravy
i am proposing not accepting new additions to the ecosystem without a license field. at some future point in time 20:29
starting out with a gentle nudge to do so when submitting a PR to the ecosystem: github.com/perl6/ecosystem/blob/ma...EMPLATE.md
but a module without a license, you won't know if you can use it before installing it 20:30
which is non-ideal
samcv goes to edit that docs page 20:32
omg. navbar plus got a filtering capability so i can filter routines and methods in the bar! woo 20:33
that's pretty awesome
20:34 robertle left
samcv also if you have two licenses which at least one thing i saw uses Artistic-1.0-Perl and GPL, it would be `Artistic-1.0-Perl OR GPL-1.0` 20:43
and yes RabidGravy i still need to program in that situation into the meta tester X| don't shoot me!
TreyHarris samcv: non-ideal, and in several corp environments I've worked at, impossible b/c you need pre-approval for even downloading.
samcv yes.
which is why the meta files themselves are under permissible licensing, by submitting to ecosystem, so you can dl the meta files. but dling the module you can't see until you dl... 20:44
and without standardized names, having a whitelist in your package manager (zef supports this) becames useless
[Coke] I find the licensing of the meta files outrageous, I hope it's actually required. 20:46
but absolutely appreciate the effort to get things licensed so I know if I can use them or not. 20:47
jnthn is, personally, rather doubtful that the META6.json files are copyrightable
But yeah, I've been rotten about making sure my modules have a license 20:48
RabidGravy I'm going with removing my modules from the ecosystem until this has some proper discussion
jnthn So a reminder to do so upon submitting is good, as is the documentation reminding me. 20:49
samcv heh jnthn well i gave you two new issues :)
i had no liceensed modules either so. i am guilty too
sjn jnthn: any written work is copyrightable. if you can write it, you can copyright it
samcv hence my reasoning for github.com/perl6/ecosystem/issues/321 [Coke] and TreyHarris 20:50
to have a metadata license that covers the metadata itself. to be more explicit. cause the metadata can be hosted anywhere, it's not like cpan where submitting is controlled, they could replace it at any time with something different that may or may not be copyrighted differently 20:51
(regarding the meta file itself)
so i take freedesktop.org's recommendations for allowed licenses for metadata and add in artistic 2.0 in addition to that, since i'm mostly certain it fullfillls the requirements 20:52
jnthn sjn: cyber.harvard.edu/property00/alter...table.html
gfldex sjn: not all countries got the same laws, here in Germany we got the concept of "Erfindungshöhe" what makes it questionable if porn is copyrightable.
what also means we are better safe then sorry by having some sort of ruling on the meta file 20:53
samcv jnthn, united states is not the whole world :P and there are so many different countries
but that's somewhat encouraging at least. but
sjn gfldex: isn't that more related to patents? (we have something similar in Norway) 20:54
samcv also what if the person writes like pages in the description of the metadata file?
like writes a book. and pastes it in the description of the metadata
gfldex sjn: it applies to any creation (Werk is not easy to translate to english)
samcv that could be copyrightable easily 20:55
sjn gfldex: the translation is "a Work" (e.g. "a Work of art")
20:55 ufobat left
jnthn samcv: Yes, I know, was just pointing out it's not so clear cut 20:55
gfldex samcv: a poem would do. And given that we are talking about perl folk …
samcv yep
hehe
sjn but sure, I get jnthn's and gfldex' points. there *might* be some expectations of effort before one can claim a © 20:56
gfldex perl -e 'do not $fear and do not die'
sjn one way to get around it is by making it easy to auto-generate the META6.json file :) 20:57
gfldex already done 20:58
samcv “ Four of these were fictitious listings that Rural had inserted into its directory to detect copying.” 21:00
hehehe
putting fake phone numbers and names in. sneaky
gfldex they do that with maps too
samcv yea 21:01
does google maps do that though?
and how many poeple have been in an accident due to inaccurate listings caused by copy protection and has anybody sued heh
en.wikipedia.org/wiki/Trap_street
gfldex may become dangerouse with self driving cars
21:02 rindolf left
samcv trap streets aren't copyrightable aparently. but i guess they are evidence that the other sections are copied 21:03
There, the court stated: "[t]o treat 'false' facts interspersed among actual facts and represented as actual facts as fiction would mean that no one could ever reproduce or copy actual facts without risk of reproducing a false fact and thereby violating a copyright ... If such were the law, information could never be reproduced or widely disseminated." (Id. at 733)
i like that decision!
21:04 skids left, dct joined 21:07 mcmillhj left, TEttinger joined 21:09 kyan joined, RabidGravy left 21:12 itaipu joined 21:17 bjz joined 21:20 japh__ joined 21:22 sufrostico joined 21:23 japh____ left 21:27 wamba left 21:29 zakharyas left 21:34 lizmat joined 21:40 dct left 21:45 Cabanossi left 21:48 Cabanossi joined, kyan left 21:50 bjz left 22:04 itaipu left
samcv on reading the do what the fuck you want to license again. it's really bad. it doesn't even specify that the license relates to the softwares use or otherwise 22:06
oh. i guess version 2 does sorta
DO WHAT THE FUCK YOU WANT TO PUBLIC LICENSE 22:07
TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
0. You just DO WHAT THE FUCK YOU WANT TO.
still not very clear even in version 2.0. sure you can always do WTF you want to. i mean. but legally. it gives you nothing
i would argue that you *can* always do wtf you want to, and people already do. but it only says what you can do, not that the copyright holder is actually granting you those privledges 22:08
the CC-zero actually dedicates to public domain with plenty of clauses that take effect if you cannot actually reliquish copyright in your or some other country and in that case grants licenses to all persons, if the other sections are not fully valid 22:20
pretty nice
timotimo yeah the CC people know their stuff 22:21
22:22 itaipu joined
samcv gfldex, i made a zef issue open for comments on the NONE vs NOASSERTION thing and mentioned what you suggested github.com/ugexe/zef/issues/154 22:23
if anybody wants to comment on it
22:28 iyra joined
iyra what's the best way to modify a hash/array in perl 6 from another function? I want to add values to a hash for example from the function I am passing it to 22:29
timotimo you can just assign to subscripts, or use the push method or sub
samcv iyra, hashes are able to be written to by default
timotimo if you want a read-only hash, you'd use Map 22:30
samcv m: sub thing ( %hash) { %hash<blah> = 'foo' }; my %hash = 'abc' => '123; thing(%hash); say %hash
camelia 5===SORRY!5=== Error while compiling <tmp>
Unable to parse expression in single quotes; couldn't find final "'"
at <tmp>:1
------> 3= 'abc' => '123; thing(%hash); say %hash7⏏5<EOL>
expecting any of:
single quotes
samcv m: sub thing ( %hash) { %hash<blah> = 'foo' }; my %hash = 'abc' => '123'; thing(%hash); say %hash
camelia {abc => 123, blah => foo}
iyra what about arrays? (the things with the @ sigil I think)
samcv but also push works too
timotimo same deal with @ sigiled things
samcv yeah arrays and hashes i believe are writable, sigil things are not unless you use `is rw` timotimo that's correct yes?
timotimo you mean $ sigil things 22:31
iyra ah, thanks okay
timotimo "is rw" is one way, "is raw" is another
samcv yeah that's what i said. sigil things are different than @ or %
timotimo and things returned from a sub are also made read-only, unless you make the sub itself "is rw" and use either return-rw or last-statement-of-block return
@ is a sigil and % is also a sigil :) 22:32
samcv: are you going to re-generate install_log.txt regularly? 22:33
samcv uh
i shuold
i forget what happened at the end that caused it to stall
timotimo do we have a big checklist like the io checklist zoffix made for the IO grant? where there were checkboxes for "pr submitted" and such?
samcv yes
github.com/perl6/ecosystem/issues/318 incomplete list 22:34
timotimo aha!
samcv feel free to add more or edit it
timotimo wait wtf 22:35
where did my clone of druid go? with all the fixes?
oh!
i had push access to druid and pushed them without forking
samcv i'll regen install_log_live.txt in a bit probably or something. i'll probably implement randomizing install order of modules too
i should look back in logs and see why it stalled. i have timeouts but. something weird. brb 15 mins 22:36
timotimo what makes the list incomplete? you had to do it by hand?
samcv uhm
it's not updated to everything in install_log_live.txt
it was based on an older log that didn't have the ability to end zef when it timed out 22:37
timotimo OK
samcv so it was partial
timotimo that explains it
samcv also. note that there is an issue. where it'll say OK before instaling certain modules
22:37 Actualeyes left
samcv so look for the FAIL's not OK cause there's an issue with my script sometimes saying OK before it installs it 22:37
brb 22:38
timotimo OK, looking for FAIL only is easy 22:39
22:55 itaipu left
timotimo github.com/MoarVM/MoarVM/releases/...-use-these - perfect 22:55
sammers samcv, looks like Template::Mustache was reported here github.com/softmoth/p6-Template-Mu.../issues/16
22:58 astj joined 22:59 BenGoldberg joined 23:01 mcmillhj joined 23:03 astj left 23:06 mcmillhj left 23:08 kaare_ left, kaare_ joined
samcv i wonder how many PR i can get done today gist.github.com/samcv/b7809f404778...f08efeb6bd 23:10
at least trying to do the more popular ones 23:11
timotimo whoa, you wanna *fix* all of these?
samcv uh
timotimo oh
samcv idk :(
timotimo no, that's for licenses, right?
samcv they have licenses, but it's not standard identifier
the list of no license is bigger than that
timotimo for a moment i thought this was about the fail-to-install log
samcv idk can maybe automate PR?
haha 23:12
dunno
timotimo you can get the "hub" program and shell-script it
samcv i've been copy pasting the same commit message and such
i have hub
i'm worried about it making a mistake though :X
plus i have to review all the different ones
timotimo hmm
fair enough
samcv and connect module names with repos etc. idk.
timotimo well, for module names and repos you can definitely use modules.perl6.org 23:13
samcv but we can automate opening issues with everybody that has NO license in meta
yeah
timotimo i.e. modules.perl6.org/repo/Foo::Bar
samcv ah
timotimo will http redirect you
samcv yeah
timotimo no* parsing required
23:15 zapwai joined
samcv and many people have "Artistic" as their license in the meta, but the project is obviously Artistic 2.0, well the license file in the project says so 23:17
so another reason i don't want to automate. but could automate all the ones that use the artistic license url
but i have a text file that i have variations of issues or of commit title/descriptions. so 23:20
will see about automating
23:20 iyra left
timotimo right 23:20
i'll go to bed now :)
23:20 iyra joined
samcv night! 23:21
23:21 mcmillhj joined
timotimo night! 23:22
23:22 skids joined
samcv gonna start on some automated thing now. first gonna compile all the different fields of those modules then decide which i can automate and which i cannot 23:23
AlexDaniel TreyHarris: well, 「git bisect」 is just a binary search 23:27
TreyHarris: so I don't know… what's so magic about binary search? :) 23:28
23:28 mcmillhj left
TimToady samcv: you can probably derive an "auth" from "source-url" 23:31
23:31 Cabanossi left
TimToady looks like a lot of people are using the "github:myname" form of auth 23:32
AlexDaniel TreyHarris: bisectable6 does a bit more sorcery though. Normally 「git bisect」 will take your word on whether the starting points are good or bad, but bisectable6 actually *runs* the code on starting points to figure out what strategy should be used (exit code / signal / output)
23:33 Cabanossi joined
TimToady samcv: some people are using "author" to put a name in, but that's kind of different than "auth", so we might leave "author" as is (or change to "creator"?) and go for something unique for auth, github:foo or cpan:bar or an email or something that is unlikely to have collisions, compared to names 23:35
samcv yeah. i think that makes sense
idk maybe it should create the auth i mean the package manager. cause package manager knows where it got them... otherwise auth is meaningless 23:36
iyra what's a nice way in perl to check if all members of a list return true on a certain predicate? for example, to know if all IO::Paths in an array are directories?
samcv unless you can verify identity
TimToady it would also be nice if there were some way to force a new version when a module is patched, but I dunno how to do that...
samcv if it didn't really come from that github user... then could be malicious
TimToady, uhm
AlexDaniel iyra: what about @smth.all ~~ :d
samcv it's possible. we collect all the META into one json file
iyra AlexDaniel, thanks, i'll check it out 23:37
samcv apparently. that the zef uses
but. that. would be hard to do? without scraping
argh
AlexDaniel iyra: see docs.perl6.org/type/Junction
samcv TimToady, make so to reinstall ALL of your modules every time!
TimToady like I sent a pr to jnthn to fix a bug, but forgot to increment the patchlevel, so zef doesn't see it 23:38
(apparently)
the S11 specs require that any given name/ver/auth correspond to a unique text if installed into an official library 23:39
23:40 espadrine left
TimToady or we should include a hash as part of the identity, if that's not practical 23:40
23:40 cpage_ left
samcv hmm 23:40
TimToady, OR we can have uh git revision
TimToady but it's gonna be a nightmare if you can't name the "correct" version you're looking for uniquely 23:41
samcv in the metadata, like we add it when we collate all of the metadata
and have it know it has changed
though the version number will be the same. but realistically. that version number is technically the newest git commit that has that version number in the json...
so it changes... but. yeah
TimToady would be nice to have a patchlevel sequence one way or another, whether you believe in semver or not 23:42
samcv since we clone the git when installing a module we could add an extra field or data that denotes the commit hash that the json came from
yeah
TimToady but we really need to enforce the invariant somehow that, if you know the identity, you know the exact contents, or we won't be able to have a distribution system that doesn't care where things come from 23:45
when it comes to source control, we're still kind of in the world of Usenet, where an email address had to know the route it took to get to you 23:46
we haven't quite reached the goal of Internet email where you have a unique identity that is independent of route
that's why "auth" is "auth" and not "author", because it's really supposed to be a naming authority 23:48
samcv yeah
ugexe the auth has no relation to where the package manager gets the package from 23:56
and verifying identity is something a recommendation manager / content storage (CPAN, MetaCPAN) could enforce if they wished 23:57
they could simply refuse to index anything with an auth they couldnt verify
TimToady right, from the standpoint of Perl 6, there is no requirement that an "auth" of "github:TimToady" be fetched from github; we're only using it to establish a unique id 23:58
some well-known place where Someone Else is managing the name collisions 23:59