»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 What the... 00:06
Apparently I wrote an Argon2 module like five months ago
And just completely forgot about it 00:07
teatime lol
timotimo ZoffixLappy: once to see if things match, another time to get a precise error diagnostic out of the thing 00:15
ZoffixLappy: have you ever tried a say or note inside a Proxy's FETCH method?
ZoffixLappy No
timotimo, but why doesn't that happen during my Subset $x; ? 00:16
timotimo you should. it's an enlightening experience
because it happens every time we do an istype
we consider "istype" to be an extremely cheap operation
ZoffixLappy Is there some sane way to detect which... type... of call is done, just so I could print my message only once? 00:17
timotimo check the backtrace, or have a state variable
or use a once block
ZoffixLappy state and once won't work, since the subset can be used many times. 00:18
ZoffixLappy tries the backtrace
timotimo hm.
ZoffixLappy Not sure how to "check the backtrace". Will leave this for now. 00:21
Actually, state can work: set it on first message, unset it on second message, 00:22
dalek osystem: 9e4b84f | (Zoffix Znet)++ | META.list:
Add Subset::Helper

Easily create subsets that sanely work with optional parameters and provide awesome error messages: github.com/zoffixznet/perl6-Subset-Helper		 00:31
ZoffixLappy m: my $s = 0; my $v := Proxy.new( FETCH => method () { note "Meow"}, STORE => method () {} ); say $v 00:35
camelia rakudo-moar 4b0e5b: OUTPUT«Meow␤Meow␤Meow␤Meow␤Meow␤Meow␤Meow␤Meow␤Meow␤Meow␤Meow␤Meow␤True␤»
ZoffixLappy :o
timotimo, I wasn't expecting this lol :)
This morning I was going to make a Proxy-based LWP::Simple lol. Glad I didn't waste my time. 00:36
timotimo i told you
ZoffixLappy Why does it fetch it so many times?
timotimo every time "decont" is called on it, it fetches
ZoffixLappy Ah
timotimo you could dump a backtrace every time fetch happens 00:37
that ought to be interesting
but decont, like istype, is expected to be trivial in cost; it's usually a pointer dereference or two 00:38
well, it's a c function invocation via the REPR
ZoffixLappy How do you dump a backtrace? I don't have an exception to call it on. 00:39
timotimo m: say callframe(0)
camelia rakudo-moar 4b0e5b: OUTPUT«/tmp/TNLLkCl38B at line 1␤»
timotimo you can just try { die "oh"; CATCH { ... } }
well, with a default clause, ideally
ZoffixLappy That's not too interesting :) gist.github.com/zoffixznet/617a627...2bbce88344 00:40
timotimo well, only callframe(0) won't be interesting 00:43
oh
you'll want --ll-exception
ZoffixLappy Same output 00:44
timotimo hum. so it is.
ah! 00:45
there's .full on the backtrace object
not too much variance
00:47 kid51 joined
timotimo leont: i trust you ran the patch in your pull request through spec tests? 00:47
timotimo also runs spectests to verify 00:48
though i suppose it can hardly impact other spec tests, really 00:49
leont I cheated a little and only ran t/spec/S17-procasync/
ZoffixLappy :)
ZoffixLappy learned two days ago an "unrelated" change can affect other tests :)
(this one github.com/rakudo/rakudo/pull/743#...-210489723 ) That I originally ignored entirely :P 00:50
timotimo heh 00:51
leont I immediately moved on to rebasing my tap harness PR, but now I need to sleep first
timotimo have a good rest!
and thanks for your continued involvement in Tap6 and stuff! 00:52
ZoffixLappy leont++ 00:53
00:56 mr-foobar left, sue joined 00:57 kurahaupo joined 00:58 mr-foobar joined 01:00 Sgeo joined
ZoffixLappy Type panda install Test::Output .... go on a Social Media excursion while the module installs... Come back... Realize you forgot to press ENTER 😂😂 01:02
01:02 kurahaupo left 01:04 leont left 01:09 boegel left 01:13 mr-foobar left
timotimo i went ahead and merged it, leedo_ 01:14
sorry, meant leont
but they left the channel
01:16 mr-foobar joined
skink > $encoded.decode 01:18
$argon2i$v=19$m=65536,t=2,p=1$D6dL87cRUdNBUkW0RlV+bA$wsbEjnb0zmWmxmbPNE1ZxavQfuIbo37mhThG/tfEnrw
Oh
So... that works.
01:27 kurahaupo joined 01:30 yurivish left
skink sortiz, Got a question you'd probably have an opinion for 01:32
01:32 kurahaupo_ joined
skink Default to returning Str for a password hash, or give the encoded Buf which you'd .decode for the Str 01:32
?
01:33 kurahaupo left 01:37 kurahaupo joined 01:38 kurahaupo_ left 01:45 Ben_Goldberg joined, jnap_ left, BenGoldberg left, Ben_Goldberg is now known as BenGoldberg
skink What the... 01:55
sortiz, I just ran into something that apparently you reported as a bug already :D
>Bytecode validation error at offset...
01:56 ZoffixLappy left
skink > my $hash = argon2-hash("password").decode; 01:59
$argon2i$v=19$m=65536,t=2,p=1$PmrqP5tZRGRWc4SiS9mvPA$0R9KvVCjntFHcWJNCYdrWg
> argon2-verify($hash, "password")
True
woo 02:00
02:03 kurahaupo left 02:07 chansen_ left 02:09 chansen_ joined 02:15 telex left 02:16 telex joined 02:17 djbkd left 02:27 adhoc left 02:31 jnap_ joined 02:37 araujo__ joined 02:38 noganex joined, araujo_ left, kid51 left 02:41 noganex_ left
sortiz skink, For the password question: as the password is all ascii, returning a Str is ok. 02:56
s/password/hash/ 02:57
03:10 kipd joined 03:12 p6_nb joined, p6_nb left 03:16 Peter_R left 03:17 clkao joined, araujo__ left 03:21 Peter_R joined 03:23 kaare_ joined 03:29 Herby_ joined
Herby_ Evening, everyone! 03:29
o/ 03:30
03:33 Herby_ left, Herby_ joined 03:34 hankache joined
hankache m: my @a = (1,2,3); my @b = (4,5,6); say @a == @b; 03:35
camelia rakudo-moar c4f04d: OUTPUT«True␤»
hankache :(
was this always the case ? 03:36
sortiz hankache, you are comparing the number of elements. 03:40
\o Herby_ 03:42
hankache sortiz yeah i just remarked, i thought it compared the also the elements inside.
so what would you use to compare the elements? eq eqv or something else?
sortiz Yep, eqv 03:43
hankache thanks
sortiz Or ~~
Smart match is, well, "smart" :-) 03:44
hankache indeed :)
sortiz m: say (1,2,3) eqv (1..3), (1.2.3) ~~ (1..3) # for example 03:51
camelia rakudo-moar c4f04d: OUTPUT«5===SORRY!5=== Error while compiling /tmp/pzwoxVq1X7␤Malformed postfix call␤at /tmp/pzwoxVq1X7:1␤------> 3say (1,2,3) eqv (1..3), (1.2.7⏏053) ~~ (1..3) # for example␤»
sortiz m: say (1,2,3) eqv (1..3), (1.2,3) ~~ (1..3) # for example 03:52
camelia rakudo-moar c4f04d: OUTPUT«FalseTrue␤»
hankache aha 03:53
~~ is "smarter" than eqv
sortiz Yep, 'cus "inside" (1..3) you have a Range. 03:55
hankache sortiz: github.com/hankache/perl6intro/pull/37 03:56
sjn m: say (1..3).Bag eqv (1,3,2).Bag
camelia rakudo-moar c4f04d: OUTPUT«True␤»
hankache hello sjn 03:57
sjn \o
skink Is there an efficient way to trim null bytes off an allocated buffer? 03:58
sortiz skink, if the null bytes are at the end, and you know its start position, a subbuf is ideal. 04:01
It is a common case in NC, when you give an allocated Buf to a function returning something with a max size, and the function returns the size used. 04:03
04:04 kurahaupo joined
skink Well, actually I'm not sure if that's a problem 04:06
m: say Blob.new(65, 66, 67, 68).decode
camelia rakudo-moar c4f04d: OUTPUT«ABCD␤»
skink m: say Blob.new(65, 66, 67, 68, 0, 0).decode
camelia rakudo-moar c4f04d: OUTPUT«ABCD␀␀␤»
skink Okay, maybe that is part of the problem 04:07
sortiz, The Argon2 lib has encodedlen() to determine the size of the char array you need to allocate
But it seems it gives too large a result
04:08 skids joined
skink hey skids 04:08
04:08 kurahaupo left
skids o/ 04:08
skink I'm using the same defaults as the Python module, and can verify hashes created with that module, but that module cannot verify my outputs 04:09
Mine are 3char too long
sortiz m: Blob.new(65, 66, 67, 68, 0, 0).subbuf(0,4).say
camelia rakudo-moar c4f04d: OUTPUT«Blob:0x<41 42 43 44>␤»
skink 4, actually
sortiz, It's variable how many extra bytes it gives, though, depending on your settings 04:10
So I would have to scan for the position of the first null byte or something
But also, having the null bytes isn't adding extra chars 04:11
I'm not sure what's causing that
Oh!
I'm dumb.
My Python module is an old version which doesn't support the new API 04:13
04:13 Cabanossi left
skink That support was only added last week 04:13
04:15 Cabanossi joined 04:17 sue left 04:30 hankache left
sortiz good night o/ 04:32
04:33 sortiz left
Herby_ ZoffinWin: trying out your Twitter module 04:36
having some issues (probably self-induced)
skink ZoffixWin, Submitted PR to add Crypt::Argon2. Already has travis set up and passing :) 04:39
04:49 Ben_Goldberg joined 04:50 BenGoldberg left, Ben_Goldberg is now known as BenGoldberg 04:55 kurahaupo joined, geekosaur left 04:58 CIAvash joined, djbkd joined 04:59 khw left 05:02 geekosaur joined 05:03 djbkd left 05:05 skink left 05:09 molaf joined
dalek ar: c36ceae | (Naoum Hankache)++ | docs/perl6intro.pdf:
Update docs/perl6intro.pdf
 05:11
ar: 6d24982 | (Steve Mynott)++ | docs/perl6intro.pdf:
Merge pull request #69 from hankache/master

Update docs/perl6intro.pdf
05:13 mane joined
Herby_ anyone still awake? 05:52
messing around with this example: github.com/sergot/http-useragent/b...crawler.p6
skids barely
Herby_ getting an error: Method 'for' not found for invocant of class 'List' 05:53
its specifically the: sub get-urls function that is throwing the error 05:54
AlexDaniel hm, what's the perl 5 equivalent for perl 6 Supply?
Herby_ AlexDaniel: o/
AlexDaniel o/
Herby_: what did you mean by this “for”? 05:55
Herby_: oh, you meant .map ? 05:56
geekosaur github.com/sergot/http-useragent/b...ler.p6#L39
skids Herby: I bet that's suposed to be .map not .for
AlexDaniel m: say (2,3,4).for({$_ × 2}) 05:57
camelia rakudo-moar c4f04d: OUTPUT«Method 'for' not found for invocant of class 'List'␤ in block <unit> at /tmp/Ez61aR1shJ line 1␤␤»
AlexDaniel m: say (2,3,4).map({$_ × 2})
camelia rakudo-moar c4f04d: OUTPUT«(4 6 8)␤»
Herby_ trying to play around with HTTP::UserAgent and copied that example, got that error
let me try map
AlexDaniel is it an old thing or just a typo?
Herby_ it works now. thats why they pay you guys the big bucks :)
AlexDaniel is not getting paid
geekosaur nov 2015 according to github... 05:58
skids Assuming it was ever tested :-)
AlexDaniel Herby_: also note that you can still use for, just not as a method
geekosaur I have a vague recollection that .for was an alias that got removed at some point
AlexDaniel m: say ($_ × 2 for (2,3,4))
camelia rakudo-moar c4f04d: OUTPUT«(4 6 8)␤»
AlexDaniel geekosaur: ok, then I'm not hallucinating 05:59
Herby_ AlexDaniel: i'll poke around at it, thanks
AlexDaniel it's interesting that nobody wrote an awesome error message for this case 06:00
“use .map instead!”
I also remember several cases when people did not know about map and were trying to do it with .for
ok not several but maybe two 06:01
geekosaur I'd have to dig in the repo, but if it was there and removed then it may have been long enough ago that it went through a period with such a deprecation message before complete removal
also I think all of those were yanked before 6.c release
AlexDaniel oooooooh
geekosaur because deprecations during prerelease were not considered relevant for the first actual release 06:02
AlexDaniel yeah, that totally makes sense
still, if I stumble upon a newcomer who is trying to use “for” instead of “map” (without knowing that it ever existed) again I'll file a ticket 06:03
Herby_ the other newcomers will probably know better :) 06:07
AlexDaniel still, I'd love to know how to have “Supply” in perl 5…
I'd ask that in #perl5 but I'm afraid that they will execute me if I phrase my question this way 06:08
although maybe they are more friendly than I expect
06:11 kurahaupo left 06:14 djbkd joined 06:20 aries_liuxueyang left 06:21 aries_liuxueyang joined 06:28 BenGoldberg left 06:30 skids left
llfourn AlexDaniel: have you seen metacpan.org/pod/Mojo::EventEmitter ? 06:34
AlexDaniel llfourn: nope!
llfourn it does the job pretty well :)
obviously you need to combine it with the IO::Loop for it to be useful 06:35
AlexDaniel hmm 06:36
06:40 xiaomiao left 06:46 xiaomiao joined 06:47 cpage__ joined 06:48 AlexDani` joined 06:49 Cabanossi left, clkao left 06:50 jameslen_ joined, jameslenz left, kipd left, AlexDaniel left, cpage left, cpage__ is now known as cpage 06:52 Xliff joined 06:53 Cabanossi joined
Xliff 'lo 06:54
06:54 kipd joined 06:57 clkao joined 07:01 dolmen joined 07:03 rindolf joined 07:05 AlexDani` is now known as AlexDaniel
Xliff \o AlexDaniel 07:07
AlexDaniel \o/
Xliff What are you up to, this fine Sunday? 07:08
AlexDaniel programming like crazy for my thesis, haven't slept yet :) So in my mind it is not Sunday yet 07:09
Xliff knows that feeling.
Don't wear yourself out, too much. And good luck!
AlexDaniel thanks 07:10
07:18 dwarring left 07:39 dolmen left
nine AlexDaniel: Perl 5 people are usually way more friendly than you seem to expect :) 07:45
Xliff tries to grok out to get Pointer[Pointer[Pointer[num32]]] into something usable. 07:46
07:47 Xliff_ joined, Herby__ joined, djbkd_ joined 07:48 notostraca joined 07:49 avenj_ joined 07:51 jeek_ joined 07:54 Spot___ joined, rafl_ joined, avalenn_ joined 07:56 MadcapJake_ joined, Herby__ left, Xliff left, djbkd left, Herby_ left, TEttinger left, jeek left, rafl left, b2gills left, infina left, Spot__ left, MadcapJake left, avenj left, tadzik left, avalenn left, chee left, Guest23064 left, eiro left, gabiruh left, rafl_ is now known as rafl 07:57 b2gills joined, MadcapJake_ is now known as MadcapJake 07:58 chee joined, Spot___ is now known as Spot__, M-tadzik joined
AlexDaniel llfourn: I ended up using a shared array of Thread::Queue's 07:59
07:59 gabiruh joined 08:02 noganex_ joined 08:03 djbkd joined 08:04 astj joined 08:05 jameslenz joined, uruwi joined 08:06 kshannon_ joined, Some-body_ joined, richi235 joined, masak_ joined 08:07 ilbelkyr joined, jdv79_ joined, lucs_ joined, perlpilot joined, jferrero_ joined, jantore_ joined, Praise- joined, Praise- left, Praise- joined 08:08 gtodd1 joined, nine_ joined, shmibs_ joined, kmwallio_ joined, pochi_ joined 08:09 kaare__ joined, M-tadzik left 08:10 Upasaka_ joined, Celelibi left, Celelibi_ joined 08:11 edenc_ joined, avalenn joined 08:12 b2gills left, avalenn_ left, jeek_ left, avenj_ left, djbkd_ left, jameslen_ left, AlexDaniel left, kaare_ left, noganex left, M-matthew left, uruwi_ left, nine left, autarch left, raydiak_ left, Woodi left, mls left, gtodd left, aindilis left, jamesneko left, orbus_ left, m0ltar left, ribasushi left, nightfrog left, zostay_ left, Praise left, lucs left, Util left, PerlJam left, jferrero left, shmibs left, ilbelkyr_ left, charsbar_____ left, richi238 left, jdv79 left, agentzh left, jolts left, Some-body_ is now known as DarthGandalf, firstdayonthejob joined, kmwallio_ is now known as kmwallio 08:13 Celelibi_ left, M-matthew joined, cgfbee left, M-matthew left, M-matthew joined, Celelibi joined, xnrand left 08:14 ribasushi joined, Woodi joined, mls joined, Peter_R left, m0ltar joined, domm_ joined, cgfbee joined, aindilis joined 08:15 mephinet left, Spot__ left 08:16 darutoko joined 08:17 shmibs_ is now known as shmibs 08:18 sivoais joined, sivoais left, sivoais joined, orbus_ joined, xnrand joined, mephinet joined 08:19 raydiak_ joined 08:20 charsbar_____ joined, Util joined, b2gills joined, jamesneko joined 08:21 zostay joined, matt_ joined, nightfrog joined, matt_ is now known as Guest55473 08:24 autarch joined 08:25 M-tadzik joined, masak_ is now known as masak 08:26 masak is now known as Guest90489, Guest90489 is now known as masak_grrr, Peter_R joined
masak_grrr alright. 08:26
I'd like to solve this SSE thing. I think it might be easier to solve it on the side, not in my existing webapp. 08:27
so here's my set of requirements: one route `a/` and one route `b/`, both of which respond to a GET request.
08:27 jeek joined
masak_grrr then one route `events/` which streams back server-side events, specifically `a` and `b` respectively whenever someone hits those other two routes. 08:28
that's all.
open questions as I go into this:
(a) can I already do this in Bailador proper? (probably not) 08:29
(b) if I get this to work, which levels do I have to hack? (probably HTTP::Easy or something)
(c) could I publish the eventual solution as an extension module to Bailador? (or will that not even be necessary?)
(d) can I expose my SSE response as a Channel? (I really, really hope so) 08:30
Xliff_ masak: Sorry, bud. I have yet to grok Bailador. 08:31
masak_grrr 's ok
I'm brain-dumping just in case *someone* has some prior art somewhere
08:31 Spot__ joined
masak_grrr actually, the most likely place to find prior art is probably CPAN 08:31
but I don't know if I'll be lucky enough to stumble over it 08:32
Xliff_ Moar NativeCall fun: What's the best way to get Pointer[Pointer[Pointer[num32]]] into something useful?
masak_grrr hits the Dancer docs
Xliff_ my @outer_ap = nativecast(CArray[CArray[num32]], $pcm.deref); 08:33
my @chan_a = nativecast(CArray[num32], @outer_ap[i]);
^^ First attempt
08:35 notostraca is now known as TEttinger
masak_grrr Xliff_: I'm sorry -- I'm not at all 'versed in NativeCall. 08:35
Xliff_ LOL!
No worries.
masak_grrr wow, Dancer is pretty cool. 08:36
08:37 djbkd left
masak_grrr finds no prior art re Dancer + Server-Sent Events 08:38
Xliff_ Server-Side or Server-Sent? Or is the difference smaller than a hair?
masak_grrr I just stopped saying the (wrong) "Server-Side", and hoped no-one would notice... :P 08:40
Xliff_ =)
masak_grrr ok, wow, so I'm on my own, and breaking ground both in Perl 6 and Perl 5
well, let's first get this working in isolation
mojolicious, however: mojolicious.org/perldoc/Mojolicious...eb-service 08:41
of course this works in mojolicious ;)
Xliff_ LOL 08:43
I just too a deeper look into Mojolicious, today.
08:43 jdv79_ is now known as jdv79
Xliff_ Previously I had only explored Mojo::DOM 08:43
Strong stuff!
masak_grrr yeah
Xliff_ masak: lists.preshweb.co.uk/pipermail/danc...02974.html 08:45
This one looks generic SSE 08:48
www.perlmonks.org/?node_id=1140968
masak_grrr no-one replied to that email, back in 2013 08:49
Xliff_ Yeah. Just noticed.
I pasted that more for the code segment in the email.
masak_grrr *nod*
I'm in a little bit over my head here
Xliff_ Aaaaand. THAT's all the relevant looking results I was able to pull.
masak_grrr, you sound like me and NativeCall! LOL! O_O 08:50
masak_grrr but I think github.com/ufobat/Bailador/blob/b1...#L144-L148 in Bailador is a show-stopper
because it presupposes a synchronously computed response content
which is exactly what I won't have ;)
Xliff_ oooo
masak_grrr so, either (a) patch Bailador, or (b) work around Bailador 08:52
seems to me (b) might be easier in the short run
Xliff_ new module contents: use Bailador; class Bailador::AsyncRoute is Bailador { # override dispatch here } 08:54
Maybe ?
08:55 pierrot left
masak_grrr yeah, something like that 08:57
may not go with class inheritance, because it is a little too Highlander-ish ;)
good or bad style: `my @params = $match.list if $match;`
I could write the above as `my @params = $match ?? $match.list !! ();`, but I realized I could just use the if there to do the *assignment* conditionally 08:58
Xliff_ :) 08:59
masak_grrr it's basically turning the assignment into a thunk
(I'm replacing the `if` statement I linked to above in a pull request) 09:00
09:01 xinming left 09:02 wamba joined
masak_grrr ufobat: github.com/ufobat/Bailador/pull/62 :) 09:05
09:05 agentzh joined, jolts joined
Xliff_ masak_grrr, I thought you didn't want to refer to $app.response.content? 09:08
Or are you golfing 2 lines, there? :) 09:09
Xliff_ used "golf" as a verb.
Xliff_ slaps himself silly.
masak_grrr just a stylistic PR, not fixing my actual problem 09:11
I'm nowhere near an action plan yet, except "might skip the Bailador level on the first go"
nine_ masak_grrr: maybe you want to work with Crust instead of Bailador? 09:17
09:17 jervo joined, literal joined 09:18 avar joined
masak_grrr nine_: maybeee...? 09:18
09:18 wtw joined
masak_grrr nine_: I have a Bailador app already up and running on heroku, though. so it'd take some convincing for me to switch. 09:19
nine_ I guess combining them is not possible yet which is a shame. 09:20
masak_grrr it wouldn't be a very big conversion. I have a small app. 09:21
but how would it help me with SSE?
nine_ I don't know :) I haven't seen much of either Bailador or Crust. Just figured, since Bailador is very high level and doesn't support your use case and Crust is much more low level, the latter might be easier to coerce into supporting you. 09:22
masak_grrr *nod* 09:23
09:23 xinming joined
masak_grrr judging by the eg/ directory, Crust also doesn't support SSE 09:30
I'm sorry to say that it's not clear to me from the Crust README, test, and eg/ directory how to get started with Crust. 09:31
I can see that .p6sgi files play a role, but not how to run them in my application 09:33
09:36 aindilis left
nine_ masak_grrr: you run it with bin/crustup 09:36
masak_grrr ok... 09:45
I'd like to point out that that is *not* in the README
m: class D { method foo { say "OH HAI" }; method dispose { say "...and that is that!" } }; { my $d will leave { .dispose } = D.new; $d.foo; die "augh" }; CATCH { default { say .message } }
camelia rakudo-moar c4f04d: OUTPUT«OH HAI␤augh␤...and that is that!␤»
masak_grrr man, that is so cool.
I posit that 95% of the time when people want DESTROY, what they *actually* want is the above 09:46
we would do well to put something like the above code in a FAQ somewhere, to short-circuit the anguish for those people
09:48 infina joined, eiro joined, infina left, infina joined 09:50 spider-mario joined
masak_grrr this is my first use of `will leave` in production. yay! 09:50
09:53 masak_grrr is now known as masak 09:56 mane left 10:00 keix left 10:01 avenj_ joined 10:03 keix joined 10:16 azawawi joined
azawawi hi 10:16
10:16 TEttinger left
azawawi so how does one close an RT perl6 bug via email? :) 10:16
10:22 jack_rabbit left
spider-mario yeah, tbh I find our current bug tracker quite horrible 10:24
it feels like 1997
10:28 araujo joined
masak RT is not perfect, to be sure 10:28
but we're quite invested in it -- lots and lots and lots of information-rich tickets -- and I don't like "feels like 1997" as a reason to switch 10:29
10:30 RabidGravy joined
masak what teams usually do when they switch (whether to Github issues or something else) is declare ticket bankrupcy. I think that would be a shame, and a high price to pay 10:30
vendethiel well, "makes people not want to participate in it because it's so hard to access, search etc" is a good reason
masak yes, that's a better one
vendethiel declaring ticket bankrupcy would be insane, though :).
masak though it *is* very easy to send an email to [email@hidden.address]
the ease of this is a big part of why I've submitted so many
I kinda like RT a little :) I seem to be in a minority, though 10:31
vendethiel and the reason why the UX is so horrible is probably because it's from 1997. But I know I try to avoid RT at all costs just because I can't comprehend it. Why do I get permission refused trying to see the ticket list? How do I filter by this or that? How to search by "tags" or whatever-else
masak yeah. 10:32
again, it's not perfect and I'm not defending RT for its technical shortcomings
but I'm getting a lot of value out of it, personally
one does get into a sort of symbiosis with the tech one uses
azawawi masak: so how does one close an RT perl6 bug via email? :) # backtracking 10:33
masak azawawi: one doesn't. one closes a ticket via the web interface. 10:34
azawawi: email is only for creating the ticket.
azawawi masak: i see
masak (this is just my answer, not an "official" answer. maybe there is a way to close a ticket via email that I don't know about.)
spider-mario well, I can’t seem to reply to anything from the web interface, for example 10:35
(also, what on earth is the difference between “respond” and “comment”?) 10:36
azawawi can someone please close rt.perl.org/Public/Bug/Display.html?id=127883 ? :)
azawawi starts reading raw.githubusercontent.com/perl6/do...tshell.pod
spider-mario another issue I have with only being able to report bugs via e-mail is that of privacy 10:41
10:41 firstdayonthejob left
spider-mario I must make sure that the From: field is properly configured in my e-mail client and that the smtp server won’t change it back 10:41
so, I don’t know how other people feel about RT, but to me, the barrier to interaction simply feels too high to bother 10:42
(especially if, for some reason, I can’t even reply to existing bugs) 10:43
azawawi or close them lol 10:44
it feels like 1990 to be exact :) 10:45
10:46 firstdayonthejob joined
spider-mario yeah :) 10:46
if I click a “Respond” link, all I get is: i.imgur.com/ol4Sr2L.png
that’s it 10:47
azawawi spider-mario: The thing i hate the most about RT is the concept of anonymous user and the ancient login bitcard system. It feels a lot out of age. 10:48
RabidGravy well that's weird, but on each entry on a ticket there is a 'Reply' and 'Comment' in the right top corner, both provide a form me
azawawi RabidGravy: boom :) 10:49
masak spider-mario: I've had the response a couple of times lately too.
10:49 gdo joined
masak spider-mario: I suspect it means you're not logged in. 10:49
spider-mario ah.
that’s not very clear
masak (but yeah, as an error message, it sucks)
azawawi masak: and the cross-site forgery warning. isnt that scary? :) 10:50
masak haven't seen that.
azawawi well i have after submitting a perlbug
masak my stance is, sure, we can leave RT behind. I won't stand in the way. but I want us to have a very deliberate plan about what to move to, and how.
azawawi: via the web interface? 10:51
azawawi masak: yup
masak azawawi: ah, that's why. I always do email.
RabidGravy it seems that the moaning about RT comes up at least once a week, but other than the almost certain non-starter of "use github issues" no-one has actually provided a concrete and doable alternative
10:51 rvchangue joined
masak I like Github's issue system, and that would certainly lower the barrier for many people. but I'm not 100% sure Github's issue system will be a satisfactory replacement for RT. 10:52
azawawi masak: pasteboard.co/fFR2Q8t.png # scary
RabidGravy yeah that's why I describe it as a non-starter
masak another component of this, of course, is that RT is written in Perl. as much as we're complaining about it, as long as we're using it, we're dogfooding.
azawawi: agreed; scary. 10:53
azawawi: also, "Referrer" should be spelled "Referer"
(it's "Referrer" in *English*, but "Referer" in HTTP)
masak really hopes that's not the reason it doesn't find the header :P 10:54
RabidGravy of course if half the people complaining were to make a replacement issue tracking system in Perl 6 , arrange to have it hosted somewhere and migrate all the existing tickets, it would become someone elses problem 10:55
10:56 Ven joined, Ven left
psch afair respond and comment differ by notifying the original submitter or not 10:57
azawawi RabidGravy: If you build it, they will come? :)
10:57 pmurias joined
psch although that's just preselected checkboxes that can be un-/checked anyway 10:57
pmurias masak: using a Perl 5 request tracker isn't really dogfooding Perl 6 10:58
re moving to github issues would it even be possible (someone mentions we couldn't assign ticket ownership properly on github when migrating them) 11:00
azawawi what's the status of the Perl 6 on JVM? Is it passing `make spectest`?
psch azawawi: not at all
although it does look a lot better than i thought... 11:02
well, unless i'm misreading [Coke]++'s summary github.com/coke/perl6-roast-data/b...ummary.out
arnsholt masak: I think "referrer" vs. "referer" might be a British English vs. American English thing
11:04 gdo left
RabidGravy I think it's just a mis-spelling on the part of the original auther of the spec that didn't get noticed until it was too late 11:04
11:04 leont joined 11:05 gdo joined 11:06 gdo left
arnsholt Or that. Being a non-native, I'm more or less incapable of keeping track of americanisms vs. britishisms 11:07
dalek c/rb-nutshell: 27a2f72 | azawawi++ | doc/Language/rb-nutshell.pod:
Fix typos and Improve documentation
 11:08
azawawi RabidGravy: 81 days github contribution streak so far :) 11:09
timotimo cool :) 11:10
masak arnsholt: it's a misspelling which became spec, not a BrE/AmE distinction. 11:11
you're right that there are BrE/AmE distinctions like that, like "canceled"/"cancelled" 11:12
11:12 jnthn left 11:13 khagan left, nowan left, khagan joined
spider-mario wow, I had no idea about that one 11:13
11:13 M-tadzik left, nebuchadnezzar left, gypsydave5 left, azawawi left 11:14 stux|RC-only left, apejens_ left, nebuchadnezzar joined, nowan joined 11:15 gypsydave5 joined, Xliff__ joined 11:17 ocbtec joined 11:18 Xliff_ left 11:19 stux|RC-only joined, Vitrifur left, Timbus left, jnthn joined,