»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.
AlexDaniel any ideas on github.com/rakudo/rakudo/wiki/Mont...mendations ? 00:02
Aaronepower raschipi: How do I add zef Apps to my path? 00:02
[Coke] are you using rakudobrew or no? 00:05
gfldex Aaronepower: are you running zef as a non-root user?
Aaronepower gfldex: Yes. 00:06
I installed perl6 with homebrew
gfldex add $HOME/.perl6/bin to $PATH
jferrero Hi! I have problems installing Rakudo Star 2017.07 00:13
Aaronepower gfldex: there is not .perl6/bin 00:16
jferrero zef-m show: No such method 'run-script' for invocant of type 'CompUnit::RepositoryRegistry' 00:21
jferrero Huummm... I suspecting that modules have not installed with make install-modules 00:35
samcv nice on this test of nqp::index, improves the speed for haystacks that are strands from 14.9s to 8.2495808s 02:15
should be commiting that very shortly
hopefully eventually we can get to the point where it takes more time to flatten the haystack than to keep it as strands for our regex 02:16
someday
raschipi I have faith in you, samcv 02:23
moritz jdv79: please give me an SSH public key 05:14
yoleaux 3 Sep 2017 23:14Z <jdv79> moritz: surely it can be fixed. ill try.
ipatrol Will this do what I want it to do? gist.github.com/anonymous/fc6f6a3e...23549b360a 05:37
yoleaux 30 Aug 2017 20:55Z <gfldex> ipatrol: not quite what you where looking for: irclog.perlgeek.de/perl6/2017-08-30#i_15093185
ipatrol ...ok 05:38
ipatrol It's supposed to represent a table of verb conjugations 05:39
ipatrol memo me or whatever yoleaux just did if you can 06:12
geekosaur ".tell gfldex your message here" 06:21
oh they left already
daxim i.imgur.com/pBW3mJn.png hey, the docs layout recently changed. can you undo the style worsening? horizontal scrolling sucks 08:37
lizmat daxim: fwiw, I don't see such horizontal scrolling on MacOS / Safari 08:41
so maybe it's a browser dependent / css issue 08:42
daxim .tell zoffix i.imgur.com/pBW3mJn.png it's docs commit d7ed308 08:57
yoleaux daxim: I'll pass your message to zoffix.
pmurias daxim: it seems to work fine here on chrome docs.perl6.org 09:00
pmurias hi 09:19
Skarsnik hello 09:23
yoleaux 03:14Z <AlexDaniel> Skarsnik: I have tested your proposed MoarVM fix. Are you sure that it affects anything on your machine? Have you tried the snippet from this bug report? RT #131003
synopsebot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=131003
yoleaux 03:15Z <AlexDaniel> Skarsnik: because it makes no difference at all here. :S
Geth doc: zakame++ created pull request #1533:
Link to non-raw GitHub page, with edit link		 09:32
Geth doc: 145005a2bd | (Elizabeth Mattijsen)++ | doc/Type/Thread.pod6
Initial doc for Thread.is-initial-thread
 10:06
sproctor Morning all. 10:23
teatime \o
sproctor So I got back from holiday and decided to give Cro a poke. I'm getting an error about peer-host not existing on an IO::Async::Socket. Would this be related to running and requesting locally? 10:26
jnthn sproctor: No, just means that it depends on a feature in Rakudo 2017.08 10:27
sproctor Ahhhhhh...
See, the other thing I did this morning was move from rakudobrew to star. 10:28
I'll swap back :)
jnthn 2017.08 also contains an important memory leak fix, and you'd really not want to run Cro without that :) 10:29
El_Che sproctor: do you run it on linux? 10:30
sproctor Yup
El_Che github.com/nxadm/rakudo-pkg/releases/
sproctor Ooooo. 10:31
jnthn: It looks very impressive. :) 10:33
pmurias is rakudo.js running on top of electron interesting to people? (I had a cobbled together demo of that for TPF::EU but I could package it up in a more sane manner) 11:13
Zoffix .ask daxim what browser is that on do you have any user styles enabled? I can't repro your issue 12:02
yoleaux 08:57Z <daxim> Zoffix: i.imgur.com/pBW3mJn.png it's docs commit d7ed308
Zoffix: I'll pass your message to daxim.
Zoffix It's freakin' ems for content sizing. 12:07
Geth doc: 23fb79c83b | (Zoffix Znet)++ | assets/sass/style.scss
Revert "Reduce width when vertical TOC shows up"

This reverts commit d7ed308f3c23debc72e1030189b6821898c62b44.
The site's styles are broken and use em sizes all over the place, which blows up on UAs that override our font size. Reverting this, with the original issue to be fixed with BS conversion.		 12:10
Zoffix .tell daxim nevermind, your browser just has custom font size set. I reverted the commit 12:11
yoleaux Zoffix: I'll pass your message to daxim.
census I'm getting an error I can't understand at all: error:00000000:lib(0):func(0):reason(0) Internal Error: 'server returned no data' in sub get at C:\rakudo\share\perl6\site\sources\A43871137AF18FF267579BE6084BD32415A09002 (WWW) 12:12
Zoffix .ask [Coke] how's the conversion of docs site to bootstrap going? Are you still progressing or is anyone welcome to jump in with it?
yoleaux Zoffix: I'll pass your message to [Coke].
census pastebin.com/3bVPV46K 12:13
pmurias short rakudo.js update: blogs.perl.org/users/pawel_murias/2...tests.html 12:14
lizmat pmurias++ 12:15
Skarsnik :)
Zoffix \o/
pmurias++ that's pretty awesome
Zoffix census: the error says the server returned no data. You're probably banned from google so it cuts the connection short. Or your sock proxy isn't working right 12:18
census Zoffix I'm able to get to google on my web browswer though 12:18
Zoffix census: WWW.pm6 doesn't use your browser to connect. For one, it doesn't execute any of the JavaScript, nor does it replicate any of potential quirks the UA you claim to be has. 12:20
census Zoffix I understand -- I just meant i'm not IP-banned 12:21
Zoffix I thought it at least relied on my IP
Zoffix census: it might be just refusing requests. I know it refuses to talk with Perl's LWP::UserAgent module with its default UA string. 12:21
census Zoffix I didn't have that problem a few days ago 12:22
raschipi Zoffix: How is rendering Pod6 on github going? 12:23
Zoffix I'm not aware of any progress.
There's an open issue somewhere.
raschipi I'm aware. 12:25
Skarsnik the issue is not just .pod6 or not for pod file? 12:26
raschipi Sorry Skarsnik, can you rephrase it? 12:28
Skarsnik like should we name pod file as pod6 so github have a clue it's perl6 pod and not p5?
Zoffix The issue is there's no (current) way to safely render pod6 as you can inject arbitrary code and IIRC GitHub never responded to whether they have a problem with that; and if they don't how we go about implementing that. IIRC they asked for a PR but no idea what the PR should do or where it should go. 12:29
Skarsnik: no, the issue is actually rendering .pod6 so they could display it on GitHub, instead of just raw file
lizmat starts on the P6W (it's going to be a big one this week)
Zoffix with .pod abiguity with Perl's pod... I'd say it's a non-issue TBH. Rakudo users should name their pod .pod6 and the issue disappears. 12:30
census Zoffix maybe they changed the website? i tried to make a simpler version of the code and got this erro: Could not parse URI: patents.google.com/xhr/query?url=q...before%3D~ 2001 ~0101%26after%3D~ 2001 ~1231&exp=&download=true in sub get at C:\rakudo\share\perl6\site\sources\A43871137AF18FF267579BE6084BD32415A09002 (WWW) line 9 12:32
census Zoffix Indeed, I went to the website, and it does look a little different: patents.google.com/?q=Babesiosis&a...r=20010101 12:32
For example, there's no more /xhr/query?url
Zoffix xhr => JS stuff 12:33
census Zoffix okay are you saying I need that then?
Zoffix You probably shouldn't have it in your code and instead use that different one.
census Zoffix I don't follow what you mean. Are you saying to make the change I was thinking? 12:34
Zoffix census: I mean the URL you're accessing in your code might be wrong. Try using this format: patents.google.com/?q=Babesiosis&a...r=20010101
census: also that "could not parse URI" error suggest the code in your pastebin differs from the code you're actually using. 12:35
'cause the latter looks like you stuck ~ chars in an attempt to concatenate, but you stuck them inside the string itself
census Zoffix i wasn't clear. i just rewrote the code. so that's why it's 2 different errors
Zoffix ok
Well, try the patents.google.com/?q=Babesiosis&a...r=20010101 lin 12:36
k
census Zoffix that seems to work better: pastebin.com/yUReR4Yn however i got an error now about the my $results-num variable. i understand the error, but i'm not sure why it's an i issue. is it the csv part is not good? 12:37
Variable '$results-num' is not declared at C:\Users\owner\Desktop/testgoogle.pl6:12 ------> my <HERE>$results-num = csv(:in<temp>).elems - 2; 12:38
Zoffix census: you forgot the quote on the previous line, so it's trying to interpolate $results-num in the string
census Zoffix okay that code worked 12:39
Zoffix \o/
census now what i want to do is abstract. go back to a form where i loop over years and loop over search-terms 12:40
raschipi census: you need syntax highliting
census raschipi what do you mean by that exactly? like colors visually
raschipi yes 12:45
census Zoffix the code doesn't work, however. It outputs `3' no matter what 12:46
wait hold on. i might have figured it out. give me a second
Zoffix .tell [Coke] FWIW, I'm not yet jump-in-ready. Was just curious. Though, I think I'll have time to give the SQL-based doc site a try in 4-5 weeks; it'll naturally have a new web app powering the site, so I'd be able to do BootStrap conversion then, if needed. The new system is a lot of work tho, so I'm hesitant to say "I'm gonna do it" :) 12:50
yoleaux Zoffix: I'll pass your message to [Coke].
census There's an issue with the .csv line I presume: pastebin.com/gFbJeBPF to indicate it's 3 no matter what the search is 12:51
Zoffix Don 12:58
Don't presume. Print it and inspect it for yourself.
census Zoffix I just checked the .csv and I still want to delete the first 2 rows. This one should give about 700, not 3
Zoffix I say I presume because I can't figure out where else the error is coming from. I did inspect it 13:00
census in perl5 there was a way to do something like a print screen (i forgot what it was called) to confirm that the html was indeed the correct screen 13:02
Zoffix census: how did you inspect it? 13:06
census Zoffix inspect what? i went to the link that corresponds with what i'm typing into perl6. then i downloaded the .csv and examined how many elements there are in it 13:07
census by went, i mean I went to the link in my browser 13:08
Zoffix census: so you only verified that the link you open in the browser works. It tells you nothing about the issue in your program. 13:09
census Zoffix so that means either perl6 isn't going to the link the way i think it is or it's not grabbing the .csv the way i think it is.
Zoffix census: print the data your script is downloading and inspect it before you parse it with CSV module.
census Zoffix yes that's what i was saying before. there was a way to do so in perl5. what is the way to do so in perl6?
i used to test my code out like that all the time in perl5 13:10
Zoffix census: simple `put` will suffice. There's also `dd` routine
census are you sayign the `put' command will save it as an .html file
Zoffix I'm looking at: docs.perl6.org/routine/put 13:11
and docs.perl6.org/programs/01-debugging 13:13
araraloren_fake evening 13:21
sproctor So I'm trying to see if I can use Cro::HTTP::Client standalone to get around the libSSL and threading issues. And I'm getting some errors. Does anyone know if there's anyway to get the details of the request made? the response object in the exception has a request in it but that seems to be empty. Thoughts? (And it's not done yet is a fine answer I know this is all bleeding edge). 13:27
timotimo cro has the tracing environment variable 13:29
you could see if that does anything for the standalone client module, too
sena_kun >the response object in the exception has a request in it but that seems to be empt 13:32
sproctor, can you file an issue about ^?
sena_kun with your code samples. 13:32
raschipi sena_kun: You guys have a plan to use the new linux TLS accelarator? 13:37
jnthn raschipi: Link? 13:38
sena_kun raschipi, never heard about this one, so at least me - no. :) And it's going to be cross-platform(at least for windows/macos), so I can feel if's.
*ifs
jnthn Also, is that not something OpenSSL could use, and thus we'd get the benefit via that? 13:39
raschipi link: lwn.net/Articles/666509/ It landed in 4.13 13:42
sena_kun >Linux miskatonic 4.12.10_1 #1 SMP PREEMPT Wed Aug 30 09:44:02 UTC 2017 x86_64 GNU/Linux 13:43
I was close.
raschipi OpenSSL needs to be involved in supporting it, but it can't be transparent.
sproctor sena_kun, I'll try and get some more info first. 13:44
sena_kun sproctor, thanks~ 13:45
census Zoffix I'm not following the exact syntax
sena_kun >he said the in-kernel TLS showed 2-7% better performance than the equivalent done in user space 13:46
census i tried dd "temp";
Zoffix i had also tried "temp".put; 13:47
raschipi sena_kun: It's not certain that will help, they are going for the usecase where it can be offloaded to the NIC. 13:49
jnthn I don't think this is going to help much without significant changes to the way IO::Socket::Async::SSL works fwiw 13:51
jnthn Since the socket is a normal libuv socket 13:51
And so we are getting buffers and then using the BIO interface to give them to the SSH lib to decrypt 13:52
Maybe I'm missing some detail, though 13:53
geekosaur this support would need to live in libssl 13:55
sproctor sena_kun running with CRO_TRACE=1 it looks like the query string is being stripped off the URL. I'll try and golf it a bit and get something more useful.
sena_kun sproctor, sounds scary, so waiting for the details. 13:56
araraloren_fake codepaste.net/532dnp I wrote a task runner example. Any idea? I stolen some idea from p6-app-installermaker-wix@jnthn. 13:58
yuyuyu Zoffix, Timotimo What about introducing a type "Expo", which is a combination of two Rats, a base and an exponent. This way you solve those issues and you also get a precise representation of Complex numbers 14:05
sproctor sena_kun: I've raised the issue. I *think* I might know that's causing it from a casual poke through the code. 14:07
Zoffix yuyuyu: what? I think you're confusing me with someone 14:09
Zoffix ... 14:11
.oO( the hell was that about )
timotimo it was about your log10 function
if we carry on information about how such a number was made, i.e. by exponentiation, we could have exact results always 14:12
except of course if we parse an int from a string, or calculate a gigantic sum, or stuff like that
Zoffix Wouldn't a mathemical calculation be a lot more efficient than hauling every Int made outta a power op as an Expo type that has two Rats? 14:15
Zoffix And what do you do with Ints made outta other operations? Fail to calculate the log? I think if I could type log 10**2000, I could type the result of it too. The goal is to get the answer for arbitrary numbers 14:16
timotimo it seems a bit fragile
Zoffix m: gist.github.com/zoffixznet/ba9a873...765f249fae 14:34
camelia 1
100
99
1000
10000
Zoffix This one actually shows up as 3.35x faster than core version :P
census what's the syntax for dumper or for put? i tried dd "temp"; from docs.perl6.org/programs/01-debugging
Zoffix census: what's "temp"? 14:35
Dump the output of get() routine
census "temp".IO.spurt: get
Zoffix census: temp is the filename, dump "get"
census "temp".IO.spurt: get "http:// ...... "
Zoffix without quotes
m: sub get { <a b c> }; dd get
camelia ("a", "b", "c")
census Zoffix do i need to initialize use Debugger::UI::CommandLine ? i forget the syntax for that in the cmd 14:36
i got Dynamic variable $*DEBUG_HOOKS not found 14:37
Zoffix ummmm
To install the module, you run zef install Debugger::UI::CommandLine in the cmd, but `dd` doesn't require that module 14:38
census did i do something wrong? pastebin.com/TU7dzwuU 14:39
i tried to install the module but i still get the same error
Zoffix Ah, it's faster 'cause &log10 delegates to methods and stuff 14:40
census Zoffix what do you mean? 14:41
Zoffix I'm talking about log10 14:42
m: gist.github.com/zoffixznet/2539e8b...43e173fe8f 14:43
camelia 1
100
99
1000
10000
Zoffix Though even this version comes out as 2.41x faster AND it can handle huge Ints :/
census yes i love using logs in programming. there's a lot of neat tricks it can exploit 14:44
Zoffix Gonna stash it for a later day 14:45
huggable: log
huggable Zoffix, nothing found
Zoffix huggable: log :is: Comes out 2.41x faster than core .log10 and can handle Ints larger than 1.8e308: gist.github.com/zoffixznet/2539e8b...43e173fe8f
huggable Zoffix, Added log as Comes out 2.41x faster than core .log10 and can handle Ints larger than 1.8e308: gist.github.com/zoffixznet/2539e8b...43e173fe8f
Geth doc: 950ad926d6 | (Jonathan Worthington)++ | doc/Type/IO/Handle.pod6
Document output buffering.
 14:55
sjn o/ 15:29
sjn takes another stab at cro today o/ 15:30
El_Che sjn: it's not that bad. Don't stab it! 15:32
sjn El_Che: the docs are a bit confusing :-\ 15:35
census Zoffix you have a second to clarify how i'm getting the dumper still wrong? otherwise how is the put syntax? i tried how i understood it from the perl6 docs 15:37
rindolf Hi all! I am having some problems bootstrapping zef - paste.debian.net/984408/ 15:42
jnthn sjn: What, in particular, is confusing? 15:44
sjn jnthn: I'm looking at the "cro stub" stuff 15:46
jnthn Hm, that'd perhaps be nicer if it more prominently mentioned that if you don't specify any options then you'll be prompted for them interactively 15:48
Any other tips for improvement are welcome
sjn some of the terminology is a bit confusing 15:49
e.g. is a "service name" the same as "id"?
sjn also, I'm having a little trouble getting the ":link" option to work 15:52
jnthn No, name is human-friendly and can contain anything, id is what it's referred to in other instructions 15:53
Though name is set to id by default
We're still largely to do :link stuff (it's pretty high on the roadmap) 15:55
The upside of writing docs before code is that everything is doc'd. The downside of writing docs before code is sometimes you didn't write the code yet. :)
sjn ...until n00bs like be come rushing head-first in and then hitting that wall. :-) 15:56
sjn *smacks head into wall*
timotimo put "die 'NYI'" everywhere :)
rindolf Can anyone help me? Is it a bug?
sjn or just a "..." :)
jnthn Yeah, sorry about that. Once we're a release or two down the line (and certainly once we hit 1.0) we'll make sure to have versioned docs. 15:57
sjn here's another doc-confusion: under "HTTP Services" in the cro-tool document, it says ":link<service>", but the terms before this talk about "service-type" or "id" 15:58
jnthn Yeah, :link<service-id> would be clearer.
sjn then the first example for using "cro stub" should also say "<service-id>" 15:59
jnthn Please file an issue, if you've a moment 16:01
sjn sure 16:01
Geth ecosystem: 6dcc590d7b | (Tom Browder)++ | META.list
make maintenance easier
 16:02
sjn btw, how do I get the interactive version of cro stub?
jnthn Just don't specify any options 16:03
There's also a web UI planned for the future
jnthn m: say 1.11 / 1.29 16:07
camelia 0.860465
jnthn oops, wrong window
sjn sjn: I get a usage message if I do that 16:11
jnthn: ^ 16:12
jnthn sjn: huh, odd... 16:13
Ulti sooo I am having a really unhappy time in Python... I think Perl 6 has ruined all languages for me
jnthn Wait, by options I mean the last ['options'] bit
Not all the other options
Ulti www.attrs.org/ <--- the problems this fixes are in many ways worse problems than Perl 5 OO and Moose is at least one relatively standardised one way to get nicer OO meta programming 16:15
sjn jnthn: ah
rindolf OK, I reported it here - github.com/ugexe/zef/issues/202 16:19
sjn jnthn: github.com/croservices/cro/compare...jn:patch-1 16:25
tried to fix a bunch of issues, but I'm not sure about the best way to write about the relationship between service-id and service-name 16:26
timotimo "is by default set to this" sounds strange; "is this by default" instead perhaps? 16:27
jnthn "to be used with other `cro` commands. `service-name` is by default" - that . wants to be a ; I think
Also what timotimo noted
sjn Allright; github.com/croservices/cro/compare...jn:patch-2 16:30
sjn hm. the cro stub http foo example needs still some love 16:32
jnthn link<service-name>` should be link<service-id>` 16:32
To run a specific service, write its `service-name`
That should also be service-id
And appears as `id` in the .cro.yml
sjn mm 16:33
I've sent a pull-request already, but if you want I can make another one
tbrowder hi, #perl6, i'm having trouble with my modules failing on travis and i can't figure out why. one of the modules is at github.com/tbrowder/Net-IP-Lite-Perl6; branch "release". i would appreciate any help. thanks.
jnthn sjn: I can fix it up later on today if you don't have chance to fix it. About time to go for dinner here :) 16:35
sjn I'll give it a shot before they throw me out of the cafe here :)
Ulti tbrowder: your json isnt valid 16:37
Error: Parse error on line 10:
...ite-Perl6/issues",},"provides": {
----------------------^
Expecting 'STRING', got '}'
excuse the paste
Zoffix rindolf: at a guess, zef is having trouble creating its config file 16:38
timotimo because of the trailing , inside the { }
Ulti tbrowder: github.com/tbrowder/Net-IP-Lite-Perl6/pull/3 16:39
tbrowder: jsonlint.com/ to the rescure on that, your master branch doesn't have the same issue
Zoffix census: well, the code you pastebinned last has an error. You're meant to dump the output of *your* get call, not a random get(). You should consider reading some of the introductory rourses: perl6.org/resources/ 16:40
sjn jnthn: allright, final attempt done. :) 16:43
github.com/croservices/cro/pull/15
rindolf Zoffix: thanks! 16:47
jnthn sjn: Thanks, will review/merge after dinner :) bbl 16:51
rindolf tbrowder: it says this isn't valid - github.com/tbrowder/Net-IP-Lite-Pe...META6.json 16:57
moritz \o 16:58
rindolf moritz: hey 17:01
moritz ho
raschipi Let's go. 17:03
tbrowder okay, my bad, i modified the file and missed the comma, oh for hjsan! thanks, Ulti! 17:07
tbrowder *hjson* 17:07
census Zoffix yes i agree it looked off. i was trying to read the perldocs docs.perl6.org/programs/01-debugging and docs.perl6.org/routine/put for sntax help 17:14
Zoffix census: it's just a regular routine that takes arguments. 17:17
census Zoffix i thought the get from "temp".IO.spurt: get "https: ... " is stored in "temp" so i had tried "temp".put 17:21
i'm not going to do data dump -- i do remember that from perl5, but i'm still having trouble with that package
that package in perl6. i'm just not figuring out how to set it up on this computer for perl6 17:22
geekosaur census: "temp" is not a handle. "temp".IO is a file path, to which you can spurt (open, write an entire thing, close) 17:23
census can i do something like dd "temp.IO ; 17:24
geekosaur you can but it'll just show you an IO::Path object, not file contents
"temp" is a Str. "temp".IO is an IO::Path object --- a pathname. 17:25
census may i ask how to see the file contents? saved as an .html file. i've done it before in perl5
geekosaur "temp".IO.slurp
census i can't remember in perl5 geekosaur. is there a way to save data dump to an .html. i feel like i used to do a print command with it so i'd save as an .html and then i can open up the file 17:28
geekosaur I don't recall an html dumper but there's this bit where web programmers always assume everyone is a web programmer and eat and shits html and json and no other formats ever exist 17:29
not to mention that "ibfrstructure" can only mena web infrastructure, "api" can only mean web api, etc. 17:30
census geekosaur i don't follow you but i know in perl5 i was able to see the .html i was doing
geekosaur basically my whole career has been erased because web programmers know there only is and only ever has been the web
I do not know what html magic you used in perl 5
census i'll have to read my old codes again 17:31
anyhow, here's my current code geekosaur: pastebin.com/qp0fLmx3
geekosaur I know you expect that all perl 5 programmers know that everything is, always has been, and always will be html, and therefore everyone knows only of those things
in perl 5 I used Data::Dumper. which does nto create html
I do not know what you used
yes, that's right, there are dumpers that do not produce html 17:32
and some programmers even use them
census you're correct now that I think about it. but I did use Data::Dumper I believe with an html print sort of statement
geekosaur in my current code, the "say" statement prints the wrong number. i don't know it he problem is in how the $results variable is being created or with the get statement
tbrowder hm, Test::Deeply::Relaxed looks like it may be the culprit, I filed a bug report there on 22Jul, i guess i'll have to change my tests--bummer! 17:35
sjn jnthn: I'm also trying to test cro-zeromq, but it looks like I'm getting random segfaults here (!) 17:39
timotimo might have the wrong zmq version perhaps 17:41
sena_kun wrong version.
github.com/arnsholt/Net-ZMQ#supported-version
4.2 will give you segfaults, errors, failed asserts, occasional Ctugha awakening, etc. 17:42
sjn, ^^^
sjn hm 17:46
rindolf Hmmm... seems like moar consumes 10% of my 3gb of ram 17:50
leont I'm not sure what «Type check failed in binding to parameter '<anon>'; expected Any but got NQPMu» means, probably .made related 18:01
moritz leont: try running your program with --ll-exception, then you should get the source location even if it's within rakudo's source code 18:02
leont It's triggered in Test.pm's _is_deeply 18:03
[Coke] .tell Zoffix anyone is welcome to jump in. if I had progress to show, it'd be in a branch.
yoleaux 12:12Z <Zoffix> [Coke]: how's the conversion of docs site to bootstrap going? Are you still progressing or is anyone welcome to jump in with it?
[Coke]: I'll pass your message to Zoffix.
12:50Z <Zoffix> [Coke]: FWIW, I'm not yet jump-in-ready. Was just curious. Though, I think I'll have time to give the SQL-based doc site a try in 4-5 weeks; it'll naturally have a new web app powering the site, so I'd be able to do BootStrap conversion then, if needed. The new system is a lot of work tho, so I'm hesitant to say "I'm gonna do it" :)
leont moritz: paste.scsys.co.uk/565029. Can't see anything refering to my code in it, except the line calling is-deeply 18:04
[Coke] getting a doc failure: $ perl6 t/meta.t 18:06
===SORRY!===
Could not find META6:ver<0.0.4..*> at line 66 in:
moritz m: Any eqv Mu 18:06
camelia WARNINGS for <tmp>:
Useless use of "eqv" in expression "Any eqv Mu" in sink context (line 1)
Type check failed in binding to parameter '<anon>'; expected Any but got Mu (Mu)
in block <unit> at <tmp> line 1
moritz leont: you're passing a Mu to is-deeply, which passes it on to eqv, which doesn't like that 18:07
Zoffix And can't be made to like it; there are roast tests testing that it autothreads with junctions
yoleaux 18:03Z <[Coke]> Zoffix: anyone is welcome to jump in. if I had progress to show, it'd be in a branch.
Zoffix But the error says NQPMu, not regular Mu; perhaps there's another place where it leaks out, like it did with Match.made 18:08
leont Zoffix: Match.made is involved, how recently was that fixed? 18:22
Zoffix On my box it's still present with 2017.07-209-g26287a1 18:24
bisect: say 42 eqv Match.new.made 18:25
bisectable6 Zoffix, On both starting points (old=2015.12 new=4b02b8a) the exit code is 0 and the output is identical as well
Zoffix, Output on both points: «False»
Zoffix bisect: old=2017.07-209-g26287a1,new=HEAD say 42 eqv Match.new.made
bisectable6 Zoffix, Bisecting by exit code (old=2017.07-209-g26287a1 new=4b02b8a). Old exit code: 1
Zoffix, bisect log: gist.github.com/1eb7c70bb85340b147...8158debf56
Zoffix, (2017-08-21) github.com/rakudo/rakudo/commit/5d...72387e9f75
Zoffix ^ then
Skarsnik_ we should call bisectable6 to "whotoblame" 18:26
leont Yeah, not observing the behavior with my git checkout perl6, guess I should upgrade my rakudobrew perl6
Zoffix rather call it "whofixedit" :) 18:26
leont This isn't isn't even the first time that commit affected my code!
Zoffix :P 18:27
[Coke] is doc's t/meta.t failing for anyone else?
Zoffix m: 2017.05,2017.06,2017.07,2017.08 say 42 eqv Match.new.made 18:29
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 032017.05,2017.06,2017.07,2017.087⏏5 say 42 eqv Match.new.made
expecting any of:
infix
infix stopper
statement end
…
Zoffix c: 2017.05,2017.06,2017.07,2017.08 say 42 eqv Match.new.made
Zoffix committable6: c'mon, bruh 18:30
c: 2017.07,2017.08 say 42 eqv Match.new.made
Zoffix Dammit, now I don't have evidence to blame leont for using dev commits instead of proper releases :P 18:30
Zoffix hm, what happened to *able ./start-all script? 18:33
MasterDuke there's a Sakefile now 18:34
Zoffix How do I use it?
Trying to kill and restart commitable
MasterDuke i think just `sake start-all`
Zoffix And even after I killed its process it's still on IRC :/
Zoffix c: 2017.05,2017.06,2017.07,2017.08 say 42 eqv Match.new.made 18:35
committable6_ Zoffix, ¦2017.05,2017.06,2017.07,2017.08: «Type check failed in binding to parameter '<anon>'; expected Any but got NQPMu (?)␤ in block <unit> at /tmp/RyG9b1WkOr line 1␤ «exit code = 1»»
MasterDuke it has become more powerful than you can possibly imagine!
Zoffix Heh, OK, looks like that bug existed every since Cursor/Match merge.
c: 2017.05,2017.06,2017.07,2017.08 Match.new.made.say 18:36
committable6_ Zoffix, ¦2017.05,2017.06,2017.07,2017.08: «No such method 'say' for invocant of type 'NQPMu'␤ in block <unit> at /tmp/OalhCLwHdo line 1␤ «exit code = 1»»
Zoffix hm, well, YAMLish shows up as a SUCC on last toast result: toast.perl6.party/module?module=YA...it=2017.07 18:37
I guess leont++ was writing new tests that showed the Any vs. Nil issue and the eqv NQPMu thing 18:38
Zoffix feels less guilty about breaking stuff :) 18:38
rindolf Zoffix: hi 18:44
rindolf Zoffix: the zef problem was fixed 18:44
[Coke] (doc t/meta) rakudobrew+installed module hell. nevermind. 18:45
rindolf Zoffix: it should have been -Ilib instad of -I. 18:45
Zoffix Ah 18:47
rindolf Zoffix: :-) 18:49
Zoffix: thanks for your help, though
[Coke] m: my $at = "this@that"; say $at.gist; dd $at 18:59
camelia this@that
Str $at = "this\@that"
[Coke] ^^ why does dd escape the @?
[Coke] (that seems like a p5-ism) 18:59
lizmat m: my $at = "this@that"; say $at.perl
camelia "this\@that"
lizmat it does because Str.perl apparently does
geekosaur I think it's a holdover 19:00
we used to expand @ (possibly pre-6.c) but it was decided that literal use is more common 19:01
geekosaur er, I meant pre-glr there 19:02
lizmat [Coke]: the reason is probably that .perl output is intended to be EVALlable 19:03
Zoffix m: dd ">>»[]@foo" 19:04
camelia ">>»[]\@foo"
Zoffix Probably because it's easier to escape the `@` instead of all the possible following chars that could make it interpolatable
lizmat m: my @a = <a b c>; my $s = '@a[]'; dd $s.perl
camelia "\"\\\@a[]\""
Zoffix m: my @a = <a b c>; say "@a».uc()" 19:05
camelia A B C
Zoffix m: my @a = <a b c>; say "@a».uc()".perl
camelia "A B C"
geekosaur oh right, only interpolates with Stuff
Zoffix m: my @a = <a b c>; say '@a».uc()'.perl
camelia "\@a».uc()"
lizmat m: my @a = <a b c>; my $s = '@a[]'; dd $s 19:06
camelia Str $s = "\@a[]"
lizmat m: my @a = <a b c>; my $s = '@a[]'; dd $s.perl.EVAL
camelia "\@a[]"
lizmat that's why :-)
Zoffix :)
ipatrol Amazingly, both Perl 6 and Japanese grammar are starting to make sense to me 19:31
tyil .hug samcv 19:34
huggable hugs samcv
sena_kun ipatrol, Japanese grammar is (mostly) easy, isn't it? Very ambiguous though. And hard to produce correctly. And sonkeigo is horrifying. And numerical system is a disaster. Well, ok, don't listen to me.
samcv yay hugs
.hug tyil 19:35
huggable hugs tyil
tyil :3
ipatrol sena_kun: the syntactical grammar is easy, what's hard are the semantics
sena_kun ipatrol, true.
ipatrol Right now I'm trying to write a toy program designed to automatically conjugate Japanese verbs given their hiragana form and the desired mood/tense 19:36
sena_kun oh, sounds cool.
ipatrol Which probably already exists, but I'm writing it mostly for study/reinforcement
sena_kun but why do you need kana? you have already the ending of the verb. 19:37
ipatrol sena_kun: do I? I'm not far enough along in the course to know if that's so 19:37
I just figured that conjugations are a thing, looked them up, and indeed they are 19:38
sena_kun I know about 2k kanji as for now and I've never seen an exception from what I'm going to say... For example, if you have "使う"(or (perhaps, almost) any other verb), you can derive conjugation just from "う" already. You don't need to feed "つかう"(i.e. hiragana of the verb) to your program. 19:40
Or I didn't understand properly how you're doing it.
ipatrol sena_kun: I'm just using a simple tail regex to separate the verb endings. Doesn't matter either way, actually 19:41
sena_kun true.
ipatrol I started with this: gist.github.com/anonymous/fc6f6a3e...23549b360a but I'm actually rearranging it to take advantage of the vowel table 19:42
Geth ecosystem: gabrielash++ created pull request #363:
Update META.list		 19:43
ecosystem: 74bcef6d66 | gabrielash++ (committed using GitHub Web editor) | META.list
Update META.list
 19:44
ecosystem: 22a1a22ea7 | lizmat++ (committed using GitHub Web editor) | META.list
Merge pull request #363 from gabrielash/patch-1

Update META.list
sena_kun ipatrol, seems good.
>(given 'pol'
>λ
ipatrol polite form
sena_kun I feel some Lisp presence. :)
ipatrol no, I've never so much as glanced at Lisp, actually
tbrowder ref: Test::Deeply::Relaxed, i can't figure out why it's a dependency. any way to get zef to tell me the actual dependency chain? 19:45
ipatrol And as for the lambda, that's a bit of a personal pun on "lemma" 19:46
sena_kun Quite a fun concidence here, since I need to review kanji right now. :) So won't be interfering into your fun, ipatrol. Have luck! 19:48
ipatrol sena_kun: am I correct that when conjugating a verb as polite in addition to another tense, you just conjugate it as polite first and then do the other conjugation on the 〜す tail? 19:54
sena_kun ipatrol, it depends. Do you mind if I pm you? Offtopic a bit. 19:57
ipatrol sena_kun: dozo
tbrowder trying the --serial option to zef to see if that sheds some light... 20:06
sena_kun tbrowder, you can file an issue at zef repo(github.com/ugexe/zef), maybe. Or browse through the sources to make a little patch, maybe. 20:07
tbrowder sena_kun: good idea on the feature request issue (a patch would be nice, but zef to me is in the stratosphere!) 20:19
sena_kun tbrowder, yeah, the code seems complex and it depends on mostly undocumented stuff. 20:20
ipatrol black magic 20:21
sena_kun I wanted to do something with panda(when zef wasn't so stable), but in the end couldn't do anything useful.
"To do something" in a good meaning. 20:22
ugexe tbrowder: according to `zef rdepends Test::Deeply::Relaxed` only one module depends on Test::Deeply::Relaxed - Getopt::Std 20:44
tbrowder ugexe: thanks! 20:45
ugexe wonder what the best way to display the entire dependency graph is 20:51
i.e. is a simple \t per level going to take up too much space to be useful for any non-trivial dependency chain? 20:52
i had this feature implemented before and it just labeled things 1.0.0.0, 1.0.0.1, 1.0.1.0, 1.0.1.1, 1.0.1.2 etc, so it was machine readable but to a human not as much 20:53
sena_kun good night, #perl6! o/ 20:54
lizmat and another Perl 6 Weekly hits the Net: p6weekly.wordpress.com/2017/09/04/...a-reading/ 20:59
Geth ecosystem: 8e2181b067 | (Tom Browder)++ | META.list
make maintenance easier
 21:07
lizmat notices at least one new ecosystem entry in next week's P6W 21:15
lizmat gets an early night before a very early morning
Zoffix lizmat++ # good weekly; lizmat++ # pluggin my blog; lizmat++ nice presentation for Weeklies :) (I looked through slides)
Crazy how recent some of the functionality in rakudo is... Makes me feel anything's possible :) 21:16
timotimo :) :) 21:18
rindolf github.com/nkh/P6-Data-Dump-Tree/pull/32/files - yay! Removed 80 lines 21:22
Zoffix \o/ 21:29
rindolf Zoffix: :) 21:50
nadim_ rindolf: those lines where there for compatibility with multiple browsers, I tested those a few years ago, like 10, on different browsers, are the ones left enough for most modern browsers? 22:29
rindolf nadim_: practically all of them 22:30
nadim_: see my commit msg
nadim_: and link
nadim_ rindolf++ a good start 22:31
rindolf nadim_: thanks 22:39
Geth doc: d607c78644 | (Zak B. Elep)++ | 2 files
Link to non-raw GitHub page, with edit link

Let the pod6 docs link to their non-raw GitHub pages, so visitors can see and maybe help out on improving some more. And for convenience, also provide an edit link.
There's probably some more improvement to do on the footer as well; will further work on separate commits/PRs.		 23:33
doc: f7c92249cb | (Aleks-Daniel Jakimenko-Aleksejev)++ (committed using GitHub Web editor) | 2 files
Merge pull request #1533 from zakame/link-to-non-raw-pages-on-github

Link to non-raw GitHub page, with edit link
[Coke] lizmat: what does HLL mean in the context of doc commit 145005a2bde389848cb5a12bfc60790915d33696 ? (HLL is tripping the spell check; I was going to add it, but I realize that I don't know what you're trying to say there. 23:43
AlexDaniel where are the bots :o
stuck trying to reconnect
MasterDuke [Coke]: HLL usually means high-level language (i.e., Perl 6 compared to C (or maybe NQP)) 23:49

Powered by Cro and the Raku® Programming Language.

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

Thank you!