»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.
tbrowder Zoffix: good change! 00:02
Geth modules.perl6.org: c795817a3b | (Zoffix Znet)++ | 2 files
Reword/Restructure/Restyle lead para

  - Use less text
  - Make it smaller		 00:04
Geth modules.perl6.org/new-site: 9bbdce3b9c | (Zoffix Znet)++ | 9 files
Nuke all the things
 00:11
modules.perl6.org/new-site: d6429c1f17 | (Zoffix Znet)++ | lib/ModulesPerl6.pm
Refuse to start without secrets in production
 00:14
zengargoyle tbrowder: i would probably do file, but would want to do named pipe but i think there's a bit of "try to not be POSIX-y". 00:17
timotimo some libraries let you give them a struct of callbacks so you can have them write to or read from whatever you like; SDL2 with its rwops is one example 00:19
ijneb Is there a way to declare multiple constants (or other types) at the same time like this: 00:33
const (a = 1, b = 2, c = 3);
BenGoldberg m: my constant a = 1, b = 2; 00:34
camelia 5===SORRY!5=== Error while compiling <tmp>
Preceding context expects a term, but found infix = instead
at <tmp>:1
------> 3my constant a = 1, b =7⏏5 2;
BenGoldberg m: my constant (a = 1, b = 2);
camelia 5===SORRY!5=== Error while compiling <tmp>
Missing initializer on constant declaration
at <tmp>:1
------> 3my constant7⏏5 (a = 1, b = 2);
jnthn Not for constants. Variables can be declared as my ($a, $b, $c). 00:35
timotimo oh hey jnthn
did it get nice and cool yet?
ijneb I see, thanks
jnthn But in that case we have list assignment semantics to work out what goes into where
Whereas constants are bound to at compile time, not assigned
And there's no list binding 00:36
ijneb I see
That makes sense
jnthn timotimo: Not so much as hoped; the rain/storm didn't happen. But it's cooling :)
Tolerable enough that I even managed to finish writing a blog post :) 00:37
timotimo oooooh 00:37
there it is! 00:38
jnthn 6guts.wordpress.com/2017/08/06/moa...ring-data/ fwiw :)
zengargoyle lols at the-eater 'shinit' because reasons. 00:45
ijneb Are `constant`s in perl6 the same as preprocessor `#define`s in C 00:46
Zoffix No 00:47
ijneb How do they work differently? 00:48
both compile time
Zoffix ijneb: In Rakudo, they're "variables" that are created at compile time and don't have a container and arrays/hashes's elements get their containers stripped out (you can still stick a container in them and mutate the variable or mutate the object in it). While with `#define` you can stick any chunk of code in it and it'll get replaced 00:49
ijneb ahh
geekosaur #define creates a textual replacement, so it can do things at compile time that code can't easily do
ijneb is there something like that in perl6? 00:50
geekosaur like paste symbols together
there's an experimental macro infrastructure, it's fairly limited and expected to change
but most of the work on it has been taking place outside of perl 6, see for example masak's 007
Zoffix ijneb: macros are planned for eventual implementation. There's an experimental implementation of them in 007
buggable: eco 007
buggable Zoffix, 007 'Small experimental language with a license to macro': github.com/masak/007
ijneb I see, thanks guys 00:51
Zoffix man &nbsp; in Perl&nbsp;6 on my linux VM gets rendered like twice wider than normal space... Such a small thing but is driving me nuts 00:54
i.imgur.com/CVn01D1.jpg
timotimo good post, jnthn++
Zoffix Just another reason to use a name without any space in it :P
timotimo Zoffix: that is really strange! 00:55
Zoffix yeah. Looks fine on Windows
zengargoyle Zoffix: is that modules.perl6.org? 00:57
does not look strange for me. your VM default font doesn't have &nbsp; glyph and is falling back to some double-wide font. 00:58
you just need a better font for matching whatever default-ish serif/sans/monospace or whatever that the browser will use. 01:01
guess it's a different page... 01:04
Zoffix yeah mp6o 01:11
Zoffix doesn't know how to get fonts
zengargoyle no package manager? i'm sorta assuming VM and some extra-minimal like distro install.... for size or speed or whatever. 01:13
Zoffix Yeah, I got package manager. It's just a standard Bodhi Linux install (minimalistic fork of Ubuntu) 01:14
ijneb Zoffix: Truetype or bitmap fonts 01:18
Zoffix shrugs
ijneb *.tff -> /usr/share/fonts/truetype
zengargoyle i'd check like: apt list '*font*' and apt list '*ttf*' and see what's installed and maybe install somthing else like dejavu or bitstream or such.. unless that's what's alread there. 01:19
ijneb fc-list to display installed fonts
Once you move the font files to '/usr/share/fonts', you can use fc-cache to load them 01:20
zengargoyle usually debian (and i'd expect ubuntu) manage fontconfig stuff decently. and installing some other font just werks. 01:22
ijneb There is probably some gui font manager if you'd like to see all of your fonts in a manageable way
fontmanager.github.io seems nice
zengargoyle unless you want to be like me and have ~/.fonts with a plethora of custom fonts.
ijneb zengargoyle: Do you like bitmap fonts? lol 01:23
The smallness of them isn't great for readability but something attracts me to them
zengargoyle ijneb: no, can't use them. HiDPI monitor. bitmapped fonts are like ant poop on my screen.
:)
ijneb lol 01:24
I like being able to display 50 billion characters on my screen though
zengargoyle i do miss them ... like fixed 6x13 or whatnot was great for most things.
ijneb font.gohu.org/ 01:25
zengargoyle ijneb: that's why i have like symbola for heiroglyps. and half a dozen japanese fonts for calligraphy. 01:26
ijneb zengargoyle: Do you have a japanese keyboard?
How does it feel to type japanese 01:27
I've never done that
zengargoyle it's a mess. but browser for the most part use fontconfig to find or fallback to 'serif' or 'sans' or 'monospace' and fontconfig/packages installed *usually* picks the best combination.
unless you go into your browser and select default fonts for languages yourself. 01:28
zengargoyle thinking VM and purposefuly small/limited install vs desktop. 01:29
ijneb: mostly easy. you use an Input Method Editor (IME) like 'anthy' and there are various ways.... i'm not native speaker so i do like: naruhodo and it shows up in ime as なるほを 01:31
ijneb I think Japanese looks cool
I'd much prefer learning Japanese over any sort of Chinese too lol 01:32
seems a lot more concise
zengargoyle and you hit space bar to turn it into kanji (sorta semi-smart)...　元気ですか
that's 'genkidesuka' + <SPC> + <RET>
ijneb Do you ever put Japanese characters into your code lol 01:33
zengargoyle ijneb: not really. i've totally thought about it, but dont. :) 01:34
i do want to make someday a monitoring application of some sort that is like a grid of kanji (sorta matrix-y) because there are a few thousand single character things to put in an (x,y) box vs ascii. and they mean something. 01:35
that's very high density information transmission. 01:36
Zoffix Well, 500MB of fonts later, still space difference, so screw it. don't care
zengargoyle ijneb: japanese is as not-concise in usage that it's crazy. 01:37
ijneb zengargoyle: Yeah seems like a mess to "digitize"
Zoffix: What do you mean?
zengargoyle Zoffix: :(
zengargoyle fontconfig is a beast that haunts me all the time not doing the thing i want. but i'm also to lazy to try and configure it the hard way. :) 01:38
ijneb: no, the 'digitizing' bit is pretty easy, it's the greater context that is hard to keep track of. 01:40
zengargoyle japanese will just not say 'words' that you're supposed to understand are there because you've been paying attention. 01:40
and will huffman encode phrases down to 'oosu' => 'ohahou goazaimasu' 01:41
ijneb: you should give it a shot if you're interested. i was a moderator on a japanese language learning website for a while so i can definately point you in the right directions. zengargoyle at gmail.com 01:44
zengargoyle and i can go on forever about encodings and fonts and phonological mutations. :P 01:47
ijneb My girlfriend knows a good amount of Japanese, so maybe it would be cool to know .. hm. 01:48
zengargoyle ijneb: give it a shot. the sounds aren't that hard, the speaking has a few quirks, but it is syllabic, so nowhere as strange as english. 01:55
the grammr is postfix.
and there's a lot of leaving out things you don't absolutely need to say that's the hard part 01:56
and a complex politeness scale where there are differnt words for something sepending on your 'rank' whith whom you're talking about. 01:57
and a social expectation that you can pick up on the ebb and flow and pick the appropriate words for the appropriate time. 01:59
Geth modules.perl6.org/new-site: 2f92f6a9c7 | (Zoffix Znet)++ | lib/ModulesPerl6.pm
Use faked out secret in dev mode
 02:16
modules.perl6.org/new-site: 20a10cbe9c | (Zoffix Znet)++ | 5 files
New index page
Zoffix Is testers.perl6.org still an active website? 02:31
ijneb Seems like it, never seen it before though 02:32
Zoffix and smoke.perl6.org/report 02:34
if they ain't; may as well kill all the cronjobs and servers... 02:35
geekosaur sounds like infrastructure intended for cpan(6)?
ijneb Zoffix: There's also cpandatesters.perl6.org 02:38
Zoffix yeah
ugexe zef works with testers.perl6.org heh 02:39
zengargoyle <expletive deleted> Amazon says my moritz book was delivered and it's not here. GRAAAAR 02:44
GRAAAR rescinded, GRAAAAR reinstated (the folded it up and stuffed it into my tiny mailbox. that's no way to treat a bit bubble wrapped envelope with a book in it. :/ 02:51
geekosaur ow 02:53
zengargoyle oooooo, pretty colors and graphs. me likey. 02:54
zengargoyle flipping through... it looks awesome to the eye. 02:55
ijneb what are you looking at zengargoyle 02:57
Zoffix ijneb: moritz++'s freshly-released book: www.amazon.com/dp/1484228987 02:59
zengargoyle moritz's book 'Perl 6 Fundamentals' -- A Primer with Examples, Projects, and Case Studies 03:00
Geth_ perl6.org: 5b26af2750 | (Zoffix Znet)++ (committed using GitHub Web editor) | source/resources/index.html
P6F is now published
 03:01
Geth perl6.org: 5b26af2750 | (Zoffix Znet)++ (committed using GitHub Web editor) | source/resources/index.html
P6F is now published
 03:02
zengargoyle it's pretty nice looking for a dead-tree (it's been a while since i've actually seen a dead-tree technical thing) 03:03
nice flow between text and code (with basic keyword bold highlighting) that just looks really nice. 03:07
zengargoyle still has to actually read it later.. :) 03:13
ijneb in GD::Raw, R, G, and B of colors are Int32's 03:33
How can I allocate a color with using the 1-255 format for each color 03:34
ijneb with gdImageColorAllocate() 03:34
are they already formatted as 1-255?
oh, they are, nevermind lol 03:35
Geth modules.perl6.org/new-site: 0d46495323 | (Zoffix Znet)++ | 6 files
Start sketching out new search page
 03:45
Zoffix purty \o/ i.imgur.com/XerZBEa.jpg 03:46
araraloren good noon! 03:50
Zoffix \o 03:52
Geth modules.perl6.org/new-site: 1a85a79a67 | (Zoffix Znet)++ | 2 files
Direct to github repo for github dists
 04:35
modules.perl6.org/new-site: 9be33b216a | (Zoffix Znet)++ | 2 files
Add search box to search page
modules.perl6.org/new-site: 34272fa942 | (Zoffix Znet)++ | 4 files
Add top nav
Geth modules.perl6.org/new-site: 3b69ed0859 | (Zoffix Znet)++ | lib/ModulesPerl6.pm
Update BS/jQuery CSS/JS
 04:40
Geth modules.perl6.org/new-site: 53ceea1878 | (Zoffix Znet)++ | 4 files
Implement "I'm feeling Lucky" feature

Fixes #52		 04:52
modules.perl6.org/new-site: 307c13f3e9 | (Zoffix Znet)++ | site-tips.txt
List Feelin' luck endpoint in site tips
 04:54
modules.perl6.org/new-site: 134a5fb5cc | (Zoffix Znet)++ | 3 files
Display site tips on home
 04:57
Zoffix huggable: swear jar
huggable Zoffix, Help Perl 6. Donate: donate.perlfoundation.org/
Geth ecosystem: 81e6da3454 | (brian d foy)++ (committed using GitHub Web editor) | META.list
Added Chemistry::Elements
 04:59
modules.perl6.org/new-site: e60f825f78 | (Zoffix Znet)++ | 2 files
Add "Donate to TPF" message
 05:01
Geth ecosystem: 6b3224b975 | (brian d foy)++ (committed using GitHub Web editor) | META.list
Update meta filename for PrettyDump
 05:07
Geth modules.perl6.org/new-site: 6a75605206 | (Zoffix Znet)++ | 8 files
Implement tags panel/route
 05:54
Geth modules.perl6.org/new-site: 55dc1fefbf | (Zoffix Znet)++ | 4 files
List dists' tags on search page
 06:04
Zoffix damn. That took no time at all. Perl++ 06:05
Just gonna streighten up JSON API and then merge already
Geth modules.perl6.org/new-site: 35a9915b42 | (Zoffix Znet)++ | 3 files
Polish to be more like old site
 06:18
modules.perl6.org: zoffixznet++ created pull request #80:
New site		 06:21
modules.perl6.org/master: 17 commits pushed by (Zoffix Znet)++
review: github.com/perl6/modules.perl6.org...2f35e8d756		 06:22
Zoffix buggable: eco test 06:23
buggable Zoffix, Found 20 results: Testing, Test::Mock, Test::Builder, TestML, Test::Junkie. See modules.perl6.org/#q=test
Zoffix buggable: eco test 06:24
buggable Zoffix, Found 20 results: Testing, Test::Mock, Test::Builder, TestML, Test::Junkie. See modules.perl6.org/s/test
Zoffix seems to work
Geth modules.perl6.org: f9edddfc42 | (Zoffix Znet)++ | lib/ModulesPerl6/Model/Dists/Schema/Result/AppveyorStatus.pm
Toss largely wrong doc
 06:32
Geth modules.perl6.org: a07aa3c190 | (Zoffix Znet)++ | 2 files
Allow empty term for search

As a means to show all dists (handy during dev)		 06:40
Geth modules.perl6.org/dist-source: c242b1b2ee | (Zoffix Znet)++ | 7 files
Intermediary; debug dist source DBIC issue
 06:58
Geth modules.perl6.org: 4358632ae0 | (Zoffix Znet)++ | 7 files
Store dist source in db

Already have `github` and `gitlab` dist sources. CPAN will be 3rd. Storing this datapoint in DB will let us cater search results differently based on source.		 07:18
Zoffix starts db rebuild for ^ that 07:19
can probably automagicated: a commit flag that tells it to build from scratch into a different db file and then move it to the normal place 07:20
ya. I'll just do that instead
Geth modules.perl6.org: e2c705a7b3 | (Zoffix Znet)++ | 2 files
Implement NEWDB commit trigger to re-deploy DB from scratch

Allows easy changes to structure of the database, without having to manually rebuild the DB and copy it over.		 07:51
Zoffix W00t! 1000th commit to that repo :D
ryu0 slackers. come back when it's 10,000 =p 07:52
Zoffix :)
Geth modules.perl6.org: 6a1b04ca74 | (Zoffix Znet)++ | README.md
[NEWDB] Fix typo
 07:54
Zoffix oh shit 07:55
Zoffix--
Zoffix nuked the db by accident. Rebuilding 08:00
Geth modules.perl6.org: dded43b884 | (Zoffix Znet)++ | update-modules.perl6.org
In new db build, if building db fails, exit early

Otherwise, we end up copying a failed, empty db file over our shiny working database.		 08:07
Geth modules.perl6.org: 6831e0f66b | (Zoffix Znet)++ | templates/root/search.html.ep
Display count of found results
 08:12
Geth modules.perl6.org: ee44a2780f | (Zoffix Znet)++ | templates/layouts/default.html.ep
Behave more sanely when last build time not available
 08:18
Zoffix :| 08:31
db rebuilt but now it's giving me a 500 -_-
Zoffix k; it's back in business... 08:34
Time for attempt #2, this time with a backup in place :P 08:35
Geth modules.perl6.org: 30d95e5736 | (Zoffix Znet)++ | 2 files
[NEWDB] Make search box wider on search page
 08:36
Zoffix k; db remains un-nuked, but rebuild failed for some reason 08:37
Geth modules.perl6.org: 82248d41fd | (Zoffix Znet)++ | update-modules.perl6.org
Do not let mktemp create temp db file

Otherwise db builder thinks it's an already-deployed db		 08:40
Zoffix (sloppy... prolly should instead try to query db to figure out whether to deploy)
Geth modules.perl6.org: 2e5aae2180 | (Zoffix Znet)++ | templates/layouts/default.html.ep
[NEWDB] Simplify IRC contact link
 08:42
Zoffix third time's the charm...
Geth modules.perl6.org: 4e42840278 | (Zoffix Znet)++ | 2 files
[NEWDB] Remove uwanted padding when dist has no tags
 08:47
Geth modules.perl6.org: 0988751590 | (Zoffix Znet)++ | lib/ModulesPerl6/Controller/Root.pm
Sort results by stars

Probably makes sense for now, at least until we add a search indexer.
Also fix lucky results giving lowest-starred result instead of highest.		 08:54
modules.perl6.org: 9ce6808dee | (Zoffix Znet)++ | templates/layouts/default.html.ep
Add language docs to nav / hide non-essentials...

  ...on smaller screens to maintain purtiness (though fully-visible
on -xs devices who get the toggle nav).		 09:11
Geth modules.perl6.org: ccbffd8dea | (Zoffix Znet)++ | 2 files
Add gitlab icon to GitLab source dists
 09:18
modules.perl6.org: 396dcfa715 | (Zoffix Znet)++ | templates/root/search.html.ep
Show N/A for GitLab issues/stargazers

Instead of 404 urls.
Probably can hide them and the update date altogether; will see what the stuff looks like once CPAN source is added.		 09:28
Zoffix .ask ugexe how is this file generated? I see source URLs are still pointing to GitHub; how come all of them aren't pointing to CPAN? github.com/ugexe/Perl6-ecosystems/.../cpan.json 09:31
yoleaux Zoffix: I'll pass your message to ugexe.
Geth modules.perl6.org: 6931493d64 | (Zoffix Znet)++ | lib/ModulesPerl6/Controller/Root.pm
Prefer exact name match in feelin-lucky search
 09:44
Geth modules.perl6.org: 88066562b1 | (Zoffix Znet)++ | 2 files
Fix "ALL" tag

  - Make it actually show all dists
  - Make both search box and tag well show up on search page when
   neither search term nor a tag was specified		 09:55
Geth modules.perl6.org: ef52f58cb7 | (Zoffix Znet)++ | update-modules.perl6.org
Restart the app after copying over new DB
 10:11
Zoffix d'oh the modules site had a bug where non-JS search (e.g. API) was searching only by name and not also description 10:18
my @a = 1, 2; in Perl saves only 1 in the array -_-
eater[sha2017] can I turn off precompilation in perl6?
kind of breaks on a read-only filesystem :')
Zoffix You can use `no precompilation` in a module... dunno if there's a global switch 10:19
AlexDaniel eater[sha2017]: no precompilation;
timotimo eco: test 10:24
araraloren buggable: eco test
buggable araraloren, Found 20 results: Test::Mock, Test::Fuzz, Testing, Test::Base, Test::META. See modules.perl6.org/s/test
timotimo oh, doh 10:25
that works now, cool
i only saw the link up above with #q=test followed by a "seems to work" and i thought "no it doesn't" :P
zengargoyle eater[sha2017]: still here? 10:33
araraloren buggable: perl 10:34
buggable: eco perl
buggable araraloren, Found 15 results: Inline::Perl5, BioPerl6, Perl6::Parser, Perl6::Maven, Perl6::Literate. See modules.perl6.org/s/perl
eater[sha2017] zengargoyle: yep 10:35
zengargoyle eater[sha2017]: i just wanted to share my appreciation.... Shinji, "I'm Sorry", i loled hard.
Geth modules.perl6.org: 54a715cea0 | (Zoffix Znet)++ | 2 files
Add `author:foo` feature to search

Filters search results to only by that author
  (author is what `author_id` is in db, which is 1st author from
meta file or repo user)		 10:36
Zoffix buggable: eco WWW author:zoffix 10:37
buggable Zoffix, WWW 'No-nonsense, simple HTTPS client with JSON decoder': github.com/zoffixznet/perl6-WWW
Zoffix hell ye
buggable: eco author:zoffix
buggable Zoffix, Found 35 results: Acme::Anguish, IRC::Client, WWW, GlotIO, CoreHackers::Sourcery. See modules.perl6.org/s/author%3Azoffix
Zoffix Alright... I've done enough damage for the night 10:38
Zoffix goes to relax
araraloren OH, modules.perl6.org/ looks pretty now
timotimo really pretty
eater[sha2017] zengargoyle: haha thanks :') feel free to help :p 10:42
nadim who's behind the new look? 10:49
Zoffix Some random guy off the Internet. 10:50
timotimo zoffix is
i mean zoffix is some random guy off the Internet.
nadim Zoffix: nice job! 10:51
I think it is important to have nice looking things to draw in more people. The hard core crew never has problems compiling stuff but the rest needs bling. 10:52
zengargoyle i'm still wondering how CPAN integration is going.... and if like MetaCPAN is eventually going to be a pretty Perl 6 module search/browse, or if it's just going to stay a repository for .tar.gz blobs. 10:55
Zoffix CPAN integration is done, pretty much (the only question I have is why the METAs have a non-CPAN URLs in them) 10:59
Zoffix Nothing's gonna happen with MetaCPAN.org; it's a Perl tool 10:59
zengargoyle it would be pretty awesome if MetaCPAN did perl6 modules with interface like perl5 modules *and* understood github like stuff like current ecosystem in addition to.
Zoffix And whatever forkage of it was being worked on looks to be ded to me, so I'm just gonna add the listing to modules.perl6.org 11:00
zengargoyle oh, that's sorta sad.
i've considered them the same for so long i didn't think that way.
Zoffix MetaCPAN doesn't support Perl 6's model of "many authors can have a dist with same name"
zengargoyle *nods* 11:01
i was totally thinging MetaCPAN pretty. :)
i had CPAN question about the MANIFEST file. which i've thought is mandatory for CPAN for indexing things. 11:03
maybe more historical as META spec changed. do perl 6 .tar.gz need MANIFEST of contents?
zengargoyle a bit hazy on MANIFEST vs META and the indexing on CPAN. 11:05
Zoffix don't see any MANIFESTs in a few dists I opened on www.cpan.org/authors/id/P/PS/PSIXDI...6/?C=M;O=A 11:06
seems to generate a META.info if it's not present and META6.json if it's not present, so there are two META files :/ Yet both have source pointing to github instead of CPAN 11:07
zengargoyle yes, it works whith no MANIFEST for perl 6, every perl 5 authoring tool i've ever used creates MANIFEST and i *thought* it was related to what is displayed and searchable vs what is actually in the .tar.gz. like CPAN doesn't extract the .tar.gz fully to get contents but relies on MANIFES or META to tell it what to put in search results. 11:09
so a module with 1000 tests doesn't have 1000 hits for the test names.... like that.
i'm not sure if MANIFEST is purely from CPAN time befor META or not... 11:10
guess we'll find out soon enough. :} 11:11
zengargoyle mostly does simple things and lets authoring tools decide. 11:12
Zoffix but all files go into MANIFEST. It's the list of all files that are supposed to be in the archive 11:13
eater[sha2017] Zoffix: zengargoyle: no precompilation worked :)) now can finally reboot w/ my init system :')))
zengargoyle but App::Mi6 doesn't create manifest and current CPAN-ized perl6 modules work without manifest... i'm wondering if it is needed.... and App::Mi6 and other tools need to make manifest as part of .tar.gz 11:15
zengargoyle and i *think* mst mentioned that perl6 META spec was more like old perl5 META spec and is missing some of the things that have been added/changed since p6 META spec was spec'd. 11:16
Zoffix p6 META spec isn't specced tho 11:17
Zoffix adds that as a TODO to 6.d prep
zengargoyle so CPAN is a shot in the dark of hosting .tar.gz, but not searching or displaying.?
MANIFEST and then META (to my memory anyway) evolved to avoid things like trying to display a random binary blob on search.cpan.ort. 11:19
Zoffix "shot in the dark"?
zengargoyle or to avoid indexing the names of lots of files that aren't really searchable in the first place.
CPAN, even the old search.cpan.org is much more than hosting .tar.gz 11:20
is it indexing and searching and displaying contents of dist. 11:21
ryu0 \o/. Packaging of the core components of perl6 is done. Next I'll need to work on some module packages.
launchpad.net/~ryu0/+archive/ubuntu/perl6
zengargoyle i don't see how without syncing perl 6 MANIFEST/META with perl 5 MANIFEST/META that the site will work well....
El_Che ryu0++ 11:22
ryu0 if this works out i may extend it to other ubuntu releases.
zengargoyle unless perl 6 CPAN is just a search for module and get .tar.gz and it doesn't do all the other stuff... 11:23
zengargoyle thinks probably a question to wait for or to go to toolchain or something. 11:24
ryu0 Wow. make spectest takes a long time.
Zoffix zengargoyle: modules.perl6.org will show all the things and CPAN will just host the tars 11:25
ryu0: try make stresstest :)
ryu0: did you set TEST_JOBS?
ryu0 Zoffix: heh. No. It defaults to single core.
I was watching it take its time in the build log. 11:26
mostly just to ensure i run the testsuite of these packages.
seemed like a good idea. :o
Hm. What shall I package first? zef seems to be a wise choice. 11:27
Zoffix Is www.cpan.org/authors/id/P/PS/PSIXDISTS/Perl6/ still the place where all the P6 modules go into? They ain't got an author name so it still has the issue of "can't have same-dist-multiple-authors" 11:28
Zoffix Ahh... That's just for the PSIXDISTS project. Each PAUSE user has a Perl6 dir. e.g. www.cpan.org/authors/id/A/AS/ASTJ/Perl6/ 11:35
Zoffix Oh.. yeeey.. these have proper META source URL. 11:39
zengargoyle Zoffix: i never heard back from requests about modules.perl6.org/todo/zengargoyle i went to great pains to try and 'pass' the checks (an they all do have a MANIFEST) but it might be a bug or something and a new commit might fix it... that's part of why i'm trying to figure out the MANIFEST thing. 11:43
Zoffix zengargoyle: dunno, I wasn't the one who coded that features 11:45
I saw some glitches with it too
zengargoyle yeah, i notified somebody but didn't try make another commit just to see if that worked in case somebody wanted to look at it. 11:46
Zoffix k, lemme look at it right now 11:48
zengargoyle i sorta want to patch Mi6 to make MANIFEST or something but i'm totally in a quandry about CPAN+MANIFEST+META and what's wanted and what's needed and what is just my old cargo-culting. 11:49
Zoffix Let's not add anything until there's clear reason to add it :) 11:50
zengargoyle doesn't really want to bother if if's not Metacpan and nice interface...
ryu0 Zoffix: do perl6 modules have to be recompiled each time the moarvm/nqp/rakudo packages are? 11:54
Zoffix ryu0: I think so. nine++ or ugexe++ would know more. 11:55
zengargoyle: you can make whatever commits you wanted to make. I repro-ed the issue with one of my modules.
zengargoyle ryu0: in theory no. the source for modules is in the repo and it *should* just notice and recompile the source if needed. 11:56
i asked this once and that's what i was told.... if it is fully bullet-proof and always works... maybe.
nadim ryu0: you have to do nothing, that was your question, you may notice a short delay the first time they are used after a change 11:57
ryu0 ok.
Teaa You're correct, Zen. I've been told the same answer before.
zengargoyle yeah, if it doesn't *work* it's probably a bug that somebody would be interested in. (in practice.... sometimes you nuke from orbit) 11:58
zengargoyle but i noticed i didn't have to re-install things anymore and asked and yeah, it's suppesed to work that way. 11:59
i think some package maintainer type people are pondering the "force recompilation now" thing so that users don't get the random pause when they use something that needs recompilation. 12:01
Zoffix k, 99% sure the /todo issue is 'cause the writer of the code did not know that there's a short-circuiting code path 12:02
And same happened with appveyor code. Wonder how to make that we have that mode more obvious :/
zengargoyle i have a cut-n-paste from nine about forcing recomp of installed modules, but haven't tested, but sorta trust nine :) 12:04
perlawhirl Can I specify something to happen at a programs INIT... *from* a module. ie. not at the modules INIT time but at the INIT time of the script that `use`d that module 12:20
araraloren m: module A { INIT { say "123"; }; }; INIT { say "456"; }; 12:30
camelia 123
456
araraloren perlawhirl, have you do some test ? 12:31
Skarsnik Hello 12:58
timotimo greetings
Skarsnik no 2017.08 yet? ^^ 13:03
timotimo releases are usually after the middle of the month 13:09
Skarsnik AlexDaniel, I am working on L:p:statm, what do you think of use Linux::Proc::Statm; say get-statm-human.perl => {:data("232.404 kB"), :dirty("0 kB"), :lib("0 kB"), :resident("112.580 kB"), :share("24.104 kB"), :size("259.832 kB"), :text("8 kB")} 13:34
AlexDaniel Skarsnik: I don't think I'll ever need this, but it could be useful 13:39
Skarsnik I also changed it so you can do get-statm(:unit("b")) if you want in byte and not kbytes 13:39
AlexDaniel usually I'd want numbers with the same range, so I'll just sprintf it and slap a kB suffix if needed
Skarsnik I found it hard to read without some small formating like that ^^ 13:40
moritz m: my $*a = 42; { my $*a = CALLER::<$*a>; say $*a 13:48
camelia 5===SORRY!5=== Error while compiling <tmp>
Missing block
at <tmp>:1
------> 3 = 42; { my $*a = CALLER::<$*a>; say $*a7⏏5<EOL>
expecting any of:
postfix
statement end
statement modifier
stateme…
moritz m: my $*a = 42; { my $*a = CALLER::<$*a>; say $*a }
camelia 42
moritz is there a way to achieve the same thing without resorting to reference-by-name?
$*CALLER::a doesn't work 13:49
moritz m: my %*a = :a, :b; say %*a; { my %*a = CALLER:<%*a>; say %*a }; 13:58
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared name:
CALLER:<%*a> used at line 1
moritz m: my %*a = :a, :b; say %*a; { my %*a = CALLER::<%*a>; say %*a }; 13:59
camelia {a => True, b => True}
{a => True, b => True}
moritz huh
I'm having a problem where something similar doesn't seem to work in a grammar
moritz m: gist.github.com/moritz/944e304f0d6...98e5b969a2 14:01
camelia Odd number of elements found where hash initializer expected:
Only saw 1 element
in regex block at <tmp> line 33
in regex statement at <tmp> line 23
in regex TOP at <tmp> line 18
in block <unit> at <tmp> line 51
AlexDaniel what's NYFI?
( github.com/rakudo/rakudo/blob/acae....java#L698 )
moritz AlexDaniel: "not yet fully implemented"?
AlexDaniel oh… ok!
moritz anybody see the problem in my gist above?
it seems to work in regular code
moritz m: grammar A { token TOP { :my %*VAR; <add> <recurse> }; token add { a { %*VAR<a> = True } }; token recurse { :my %*VAR = CALLER::<%*VAR>; b }}; say A.parse('ab'); 14:03
camelia ｢ab｣
add => ｢a｣
recurse => ｢b｣
moritz huh, works there
m: grammar A { token TOP { :my %*VAR; <add> <recurse> }; token add { a { %*VAR<a> = True } }; token recurse { :my %*VAR = CALLER::<%*VAR>; b <add> { say %*VAR.perl } }}; say A.parse('aba');
camelia {:a}
｢aba｣
add => ｢a｣
recurse => ｢ba｣
add => ｢a｣
moritz m: grammar A { token TOP { :my %*VAR; <add> <recurse> }; token add { (.) { %*VAR{$0} = True } }; token recurse { :my %*VAR = CALLER::<%*VAR>; b <add> { say %*VAR.perl } }}; say A.parse('abc'); 14:04
camelia {:a, :c}
｢abc｣
add => ｢a｣
0 => ｢a｣
recurse => ｢bc｣
add => ｢c｣
0 => ｢c｣
patience I am still learning perl6 idioms - can someone refactor the compound stmt beginning on line 3 of the following? 14:05
my $proc = run '/bin/ls', '-l', '/tmp', :out;
say "/bin/ls -l /tmp";
evalable6 /bin/ls -l /tmp
patience with $proc.out { 14:06
moritz patience: please use a pastebin
patience no strict;
repeat {
$return = .get; 14:07
moritz patience: stop pasting multi-line code directly in IRC
patience: use a pastebin/nopaste site
patience how to do that?
} 14:08
while $return.Bool;
}
ENDS
can I paste <pre> .. </pre> bloc directly and have it be formatted (< 10 lines) automatically ? 14:09
timotimo there's thousands if not millions of nopaste sites on the 'net
no, irc doesn't have anything like that
patience will check logs of this channel - back later 14:10
moritz patience: you something like pastebin.com/ or gist.github.com/ 14:10
nadim nopaste.linux-dev.org 14:12
timotimo: You were right, displaying type for Int is just noisy, Ints, as long as they do nothing else, will be displayed with a type that is blank. in columnar as in vertical mode. imgur.com/qfhepwy 14:14
should be in the next version, with less noise for IntStr too. Maybe I get myself to do the same for Str too.
timotimo hm, it'd perhaps be good to be able to tell the difference between Int and IntStr 14:17
maybe just output them with < > aronud 'em
Skarsnik how does it show stuff like int32?
nadim IntStr still get a type but when the int part and the str part are identical then only the int part is shown, there's an example in the image above 14:18
timotimo an int32 can't really be inside a container like an array or hash 14:19
nadim Skarsnik: good question, I guess .int32
let me try
timotimo so it'll have a Buf[int32] or similar on the next level outside
Skarsnik :m my int32 @foo; say @foo.WHAT;
m: my int32 @foo; say @foo.WHAT;
camelia (array[int32]) 14:20
nadim I must admit that I have dumped no Buf not int32
yet
moritz m: grammar A { token TOP { :my %*VAR; <add> <recurse> }; token add { (.) { %*VAR{$0} = True } }; token recurse { 'b' :my %*VAR = CALLER::<%*VAR>; <add> { say %*VAR.perl } }}; say A.parse('abc');
camelia {:a, :c}
｢abc｣
add => ｢a｣
0 => ｢a｣
recurse => ｢bc｣
add => ｢c｣
0 => ｢c｣
moritz can't figure out why it works in this simple example, but not in gist.github.com/moritz/944e304f0d6...98e5b969a2 14:21
nadim It's show like this but nothing of what's in the array, bleah!,└ 4 = .array[int32] @18 14:24
Skarsnik what is the @18 14:25
nadim it's a reference, any array and other containre gets them, if they are fered to somewhere elese, the data is not shown again but the reference is 14:27
Skarsnik m: my uint32 $foo; $foo = -42; say $foo;
camelia 4294967254
nadim Skarsnik: it would make much more sense if the example I gave had elements in the array 14:28
Skarsnik m: my uint64 $foo; $foo = -42; say $foo;
camelia -42
Skarsnik hm, these unsigned int stuff is still not fixed it seems
patience How do I insert a clickable link in my message, e.g. as nadim did above at 10:14 ? 14:29
Skarsnik just copy paste the link
your irc client will make it clickable
ilmari irc itself is just plain text
any fancy display is up to the client
moritz right 14:30
patience I am still learning perl6 idioms - how can I tighten up the following short program: pastebin.com/F4sRYY0y 14:32
AlexDaniel patience: maybe you need $proc.out.slurp-rest ? 14:33
patience What I am trying to do is test .get so I can terminate the script when it returns Nil 14:34
IE I need line by line input 14:35
Any way it works .. 14:36
moritz patience: you're never assigning to $return
AlexDaniel patience: what about $proc.out.lines then?
nadim m: my int32 $x ; $x.^name.say
camelia Int
nadim so int 32 is not a type? is it a hint? or what? 14:37
moritz patience: also don't use "no strict;". Strict mode caught an error for you, and you ignored it
moritz patience: you can replace your whole loop with for $proc.out.lines -> $line { say $line } 14:38
AlexDaniel patience: also, there is a non-perl6 problem with your code: mywiki.wooledge.org/ParsingLs
patience: but maybe it's just an example, I don't know
patience OK thanks, will try - back later 14:40
AlexDaniel patience: if you want to go through files/directories in a directory, then use ‘dir’. For example: for dir ‘/tmp’ -> $smth { … }
nadim So, what's a int32, obviously not a type and what is int32 @array, since it is not an array either, hand how can one dig out it's values (this suddenly made my sunday exciting) 14:51
timotimo int32 is a native type, those are special 14:52
when you call a method or something on an object like that you'll get it boxed into an Int
nadim ok
timotimo they are slippery like that
nadim is it possible to know if an Int is an int32? 14:53
ryu0 say 3/5;
evalable6 0.6
ryu0 say 3/5.WHAT;
evalable6 (exit code 1) Parameter 'de' of routine 'DIVIDE_NUMBERS' must be an object instance of type 'Int', not a type object of type 'Int'. Did you forget a '.new'?
in block <unit> at /tmp/VGZ6tIQ7cL line 1
ryu0 say (3/5).WHAT;
evalable6 (Rat)
timotimo that's more a property of a variable, really
m: my int32 $foo = 99; say $foo.WHAT; say $foo.VAR.WHAT;
camelia (Int)
(IntLexRef)
timotimo IntLexRef is a kind of object that forwards all accesses to a lexical slot in some frame 14:54
this way the int32 can stay an int32 but you can still pass it around with read-write access
nadim ok, is it possible to know what it is forwarded to?
perlawhirl bisectable6: say (1,2,3).combinations(2..*)
bisectable6 perlawhirl, Bisecting by exit code (old=2015.12 new=aca4b94). Old exit code: 0
timotimo i don't think so
bisectable6 perlawhirl, bisect log: gist.github.com/b3a3d6a44a4bb7081c...2b7d695e4a
perlawhirl, (2017-01-14) github.com/rakudo/rakudo/commit/50...dd10f3fbe9
nadim timotimo: ok, what about int32 @array? 14:55
timotimo there you'll get a IntPosRef
which is like an IntLexRef but instead of a lexical pad it refers to a native array
m: my int32 @foo = 1, 2, 3; say @foo[1].WHAT; say @foo[1].VAR.WHAT
camelia (Int)
(IntPosRef)
nadim but it is not an Array, how do I get it's elements? 14:56
timotimo how do you mean?
m: my int32 @foo; say @foo.WHAT
camelia (array[int32])
nadim hmm, maybe have to handle IntPosRef
timotimo in what circumstance would you have to? 14:57
araraloren m: my int32 @foo = 1, 2, 3; say @foo[1];
nadim m: my int32 @foo = 3, 4 ; dd @foo
camelia 2
array[int32].new(3, 4)
AlexDaniel commit: 502fc77a6^,502fc77a6 say (1,2,3).combinations(2..*)
committable6 AlexDaniel, ¦502fc77a6^: «((1 2) (1 3) (2 3) (1 2 3))» ¦502fc77: «This type cannot unbox to a native integer: P6opaque, Num␤ in block <unit> at /tmp/_5iwIZuPqX line 1␤ «exit code = 1»»
AlexDaniel commit: HEAD say (1,2,3).combinations(2..*)
committable6 AlexDaniel, ¦HEAD(aca4b94): «Cannot determine integer bounds␤ in block <unit> at /tmp/P75ThMhAwg line 1␤ «exit code = 1»»
timotimo oh is that because it treats * as Inf? 14:58
AlexDaniel sure, and so? 14:59
both * and ∞ should work I think
perlawhirl: you are going to submit a ticket, right? 15:00
perlawhirl yes, they did. at least, there should be a way to ask for combinations of n or greater, without having to specify a concrete upper bound
yeap... will log it to rt 15:01
quick question... is rt still only accepting emailed submissions 15:02
AlexDaniel perlawhirl: there is a hidden door for some of the club members. rt.perl.org/m/ 15:04
perlawhirl ta 15:05
Skarsnik_ was a mail for bug?
like [email@hidden.address]
timotimo right, it should definitely work, the code just has to be changed to make it work (again)
AlexDaniel my face was slapped several times now for advertising this as an option, but the whole situation is a joke to be honest
moritz Skarsnik_: [email@hidden.address]
AlexDaniel commit: all say (1,2,3).combinations(2..*) 15:08
committable6 AlexDaniel, gist.github.com/74f64a1ef4bf80453f...a570964d89
nadim Skarsnik: something like this in vertical mode, some time this week when I have understood better how the int32 style types are used imgur.com/0REbrsM 15:09
moritz ok, I figured out my grammar conundrum; turns out I need to use CALLERS:: instead of CALLER:: if there is a chance of there being an intermediate rule 15:10
AlexDaniel Skarsnik: hey. I want Gumbo to be fixed ASAP, how can I help with the bug?
Skarsnik Well, look like a MoarVM bug ~~ 15:10
perlawhirl AlexDaniel: I'm not a master with these bots... So it looks like it broke between 2017.02 and .03... am i interpreting the commitable output correctly? 15:11
AlexDaniel Skarsnik: yeah, but maybe I can golf it down to something managable?
perlawhirl: bisectable gave you the right answer, which is 502fc77a68924a68115e739ffae64fdd10f3fbe9
perlawhirl k, thanks
AlexDaniel perlawhirl: we can double check like this
Skarsnik hm, I wanted to try writing a small C program that does lot of allocation/free to try to replicate this thing 15:12
AlexDaniel commit: 502fc77a6^ say (1,2,3).combinations(2..*)
committable6 AlexDaniel, ¦502fc77a6^: «((1 2) (1 3) (2 3) (1 2 3))»
AlexDaniel so it works before 502fc77a6
commit: 502fc77a6 say (1,2,3).combinations(2..*)
committable6 AlexDaniel, ¦502fc77: «This type cannot unbox to a native integer: P6opaque, Num␤ in block <unit> at /tmp/cK9C5OFdUt line 1␤ «exit code = 1»»
AlexDaniel but not on 502fc77a6
ryu0 Is it normal for perl6 to be slow to start up?
Skarsnik The issue, it does not crash with the minigumbo thingy if I remember correctly
Skarsnik ryu0, yeah first start can be slow with lot of modules 15:12
moritz ryu0: depends on what you do
AlexDaniel ryu0: how slow?
moritz a perl6 -e '' should be around 0.1s 15:13
ryu0 i'm getting ~ 0.25s of startup time each time I run it.
AlexDaniel ryu0: what rakudo version?
ryu0 2017.07
moritz is your hardware really old? 15:13
ryu0 then again this PC has a slower processor. N3150.
No, but it's a low power model.
oddly perl5 starts very fast. 15:14
moritz yes, perl 5 is highly optimized for low startup times
Skarsnik perl6 is still quite slow
timotimo yeah, perl6 is mildly optimized for low startup times
moritz on my machine, I get around 5ms startup for perl5, 100ms startup for perl6, 40ms for python2.7
timotimo was perl6 already faster than perl5 if you only add moose or do you have to add more modules? 15:15
moritz ruby around 60ms
timotimo: on my machine, perl -MMoose -e '' is slower than perl6 -e ''
timotimo ok, cool 15:16
Moo is likely a lot faster, right?
Skarsnik AlexDaniel, the issue, if the minigumbo code I wrote does not make it crash, etheir you need to tweak the minixml code to make Moar do more stuff with memory and figure what trigger the heap corruption, and dig into Moar with the actual crash ~~ 15:17
Skarsnik s/and/or 15:21
Skarsnik AlexDaniel, I wonder how we can create leak test/crash 15:29
AlexDaniel ryu0: there's one more thing. I don't have a startup time graph yet, but due to latest changes the startup time dropped by tens of milliseconds I think 15:30
timotimo dropped? don't you mean increased?
AlexDaniel timotimo: huh? 15:31
timotimo wait, what i remember were only like 6 miliseconds
AlexDaniel hm, maybes so, but it felt like more than 10 at least 15:32
timotimo ooh
that was spesh enabled vs disabled
AlexDaniel ryu0: in any case, the startup time is constantly improving, but there's still room for improvement… like, a lot of it 15:33
I was surprised to know that python3 has significantly slower startup than python2
it's not that surprising now that I think about it, but it was a bit unexpected for sure 15:34
timotimo oh? what makes it unsurprising? 15:35
they've had a few years of making semantics right and since then have been working on performance, too, no?
AlexDaniel timotimo: imo it's easy to make a language that has very low startup time but very few useful features. And once you're below 50ms, who cares? So I'm totally fine with things getting a bit slower, if they're fast anyway 15:39
does not apply to rakudo though :P
timotimo yeah, 10ms startup would be nice 15:53
but i'd settle for 40ms 15:54
jdv79 nope, not using XML via replacing it with a run call out to p5 doesn't seem to impact the new moarvm mem hogging 16:00
jdv79 a little but seems more systemic than i hoped 16:00
AlexDaniel jdv79: do you have a runnable example to reproduce the issue? 16:16
perlawhirl: hmmm I think “Anon state var doesn't work in regex” is not a bug actually 16:19
perlawhirl: lemme find it…
or at least is a duplicate 16:20
perlawhirl: RT #131102
synopsebot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=131102
jdv79 looks like its run that causes thread starvation or something like that 16:22
working on it
AlexDaniel perlawhirl: yea, ok, I merged these tickets 16:23
jdv79 moving from HTTP::Tinyish in sync to async mode cleared that portion up 16:23
AlexDaniel jdv79: irclog.perlgeek.de/moarvm/2017-08-05#i_14973437 ? 16:25
jdv79 maybe 16:26
jdv79 hmm. removing all Proc seems to help. 16:38
rakudo really doesn't like proc not-async i guess
mst guessing the more recent rakudos' ability to not tie up threads during an await makes that less likely to cause starvation? 16:41
timotimo proc async was already not using up a thread, though 16:44
timotimo at least i think so 16:50
moritz do we have LEAVE phasers for regexes? 16:55
jdv79 yes, the new await works as advertized 16:57
but for some reason if you use Proc somewhere it'll seemingly deadlock the scheduler if your task count is > the max thread count 16:58
so just converting it all to async and its fine again
gfldex jnthn: are you testing moar with long running programs? I'm asking because I got a segfault with a really simple program. 17:44
after about 3 days
jnthn gfldex: I don't have anything set up for that, no. Though have been thinking about it a bit. 17:49
gfldex jnthn: i got coredumps active now. I will yell when I get another hit. 17:50
jnthn gfldex: Alright, thanks. 17:51
El_Che I don't mind setting a VM with this type of a smoker, but I have no clue what a representative program could be 17:52
gfldex El_Che: that's what crashed for me: github.com/gfldex/perl6-bt-auto-co...connect.p6 18:03
Geth modules.perl6.org: badc389ecd | (Zoffix Znet)++ | 2 files
Don't pre-set appveyor status

The Dist model already handles the not-set-up case and the appveyor postprocessor handles the 'unknown'		 18:04
Zoffix m: my @foo = qw/foo bar ber foo/; sub what { $_ //= shift @_; /meows/ }; if (my ($foo) = grep &what, @foo) { say "I am a Rakudo program" } else { say "I am a Perl program" } 18:35
camelia I am a Rakudo program
Zoffix :)
Petit_Dejeuner Neat! 18:38
Zoffix m: say "0" && "I am a Rakudo program" || "I am a Perl program" 18:41
camelia I am a Rakudo program
Zoffix Shorter version :)
Zoffix See also: github.com/mauke/poly.poly 18:45
jdv79 oh wow. XML and HTML::Parser::XML timed out after a minute on a reddit parse whereas p5 and libxml does it is < 2s 18:52
Skarsnik use Gumbo
Zoffix use DOM::Tiny
buggable: eco dom
buggable Zoffix, Found 9 results: DOM::Tiny, Digest::PSHA1, Haikunator, Math::Random, Crypt::Random. See modules.perl6.org/s/dom
Skarsnik :modules Gumbo
jdv79 i'd prefer if the grammars were better
but its fine
does gumbo do both xml and html5? 18:53
Zoffix jdv79: well, you know where the source is :) help yourself
jdv79 or any of those
Skarsnik It's not a good idea to waste time trying to write a html5 parser IMO
too much work
buggable: eco Gumbo 18:54
buggable Skarsnik, Gumbo 'Binding of the gumbo C library, a html parser lib': github.com/Skarsnik/perl6-gumbo
Skarsnik Too bad I did not manage to make the XML parsing work with gumbo
could have be nice in GPTrixie to not wait to XML module to take 30sec+ on sizeable header 18:55
Zoffix jdv79: Perl's version of DOM::Tiny does, and it's meant to be a clone, so if it was a faithful one, then it should do too in Rakudo 18:56
s/clone/port/;
jdv79 ok 18:58
nadim guys, is .array only used with native types?
nadim Skarsnik: Native type support is certainly possible, had to change 2 lines in the core code and a few handlers have to be added. I' am prety sure it can be made better. I'd like to show that things are used for native calls but once instantiated it becomes difficult 19:26
first shot is not too bad imgur.com/43p37nf
I'd need to read the docs more and write tests for all the supported types 19:27
this is what is being dumped nopaste.linux-dev.org/?1160473 19:28
moritz do we have a list supported character classes anywhere? samcv, halp! :-) 19:42
Skarsnik nadim, not, maybe you could add the CStruct info to the Point class 20:02
nadim, this can maybe help figure stuff about native type github.com/rakudo/rakudo/blob/nom/...l.pm6#L233 20:04
nadim It would certainly be helpful to display what the classes are in native land, I'll have a look at it 20:08
Skarsnik and be careful, int32 are "Perl 6" native, a class with is repr('CStruct') is "C" native 20:17
nadim want to give it a try ? ;) I'll provide a template, examples and support 20:34
Geth modules.perl6.org: 5db8ada712 | (Zoffix Znet)++ | 5 files
De-spaghettify DistSource <-> PostProcessor / Fix TODO issue finder

  - Move all TODO issue checks to METAChecker postprocessor
   - Lets us avoid duplicating the checks in each dist source
   - Fixes false positives MANIFEST issue
  - Make MANIFEST TODO check only applicable to `cpan` dist source as it's a bit
   silly to tell users they need to create useless files in their repos
  - Add TODO check for version being set to `*` which we now deem invalid
  - Do not use author from GitHub URL if missing from META; hosting != authorship
  - Move all Travis and AppVeyor logic entirely into corresponding postprocessors		 20:48
Geth modules.perl6.org: ddea979132 | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder/Dist/Source.pm
Pass a `time` query arg when fetching META file

GitHub overly eagerly servers cached version of the META file, so that even after a commit is done, it serves the old META for some time.
This interferes with testing during dev, and might also cause inconguency between the META data we have and latest commit data, since we fetch META off the Web, not the API call.		 20:49
Zoffix .tell zengargoyle this fixes the MANIFEST in TODO issues you were having (fixes it as well as moves it to be only for `cpan` dist source): github.com/perl6/modules.perl6.org...5db8ada712 20:50
yoleaux Zoffix: I'll pass your message to zengargoyle.
Zoffix I hope I didn't mess anything up in that commit. Ain't no unbittrotted tests for any of that code :'( 20:51
Geth modules.perl6.org: 88874dedfc | (Zoffix Znet)++ | use-me-for-commit-triggers
[REBUILD] Add special file to use for commit triggers

So we don't edit random files just for triggers		 20:54
Geth modules.perl6.org: 6244527199 | (Zoffix Znet)++ | 3 files
Add commit trigger for faster app restart
 20:59
Zoffix
.oO( [MAKEMEASAMMICH] ... )
 21:00
Geth modules.perl6.org: 881fbe6740 | (Zoffix Znet)++ | 6 files
Rename PostProcessors to order them

In preparation to add interacting postprocessors, e.g. README fetcher to fetch the README just once and re-use that data in both METAChecker and AppVeyor postprocessor.		 21:06
jdv79 if i call exit(), shouldn't all threads be halted? 21:30
once they can - not in a syscall, etc... 21:31
Geth modules.perl6.org: 8998676fd6 | (Zoffix Znet)++ | 2 files
Implement DataFetcher postprocessor

And fetch README content in it		 21:32
modules.perl6.org: 259adf66f5 | (Zoffix Znet)++ | templates/layouts/default.html.ep
[REAPP] Fix broken last-update date in footer
 21:35
Geth modules.perl6.org: 444c586e58 | (Zoffix Znet)++ | 5 files
Implement heuristic for AppVeyor project URL

  - Try to fetch status from API using GitHub's user/repo
   for AppVeyor user/slug; if success, that's the AppVeyor url
  - If that fails, fetch status from badge and rake repo's README
   for the URL to AppVeyor project and use that
  - If that fails, just direct AppVeyor URL to dist repo
Fixes github.com/perl6/modules.perl6.org/issues/76		 22:27
modules.perl6.org: 19b4da489d | (Zoffix Znet)++ | use-me-for-commit-triggers
[NEWDB] Rebuild db for AppVeyour URL db entry
 22:29
Geth modules.perl6.org: 1b7771a686 | (Zoffix Znet)++ | update-modules.perl6.org
Do not bail out if re-app commit trigger fails to start app
 22:35
modules.perl6.org: de43f672ad | (Zoffix Znet)++ | templates/root/search.html.ep
In tag view, suggest keyword search for current tag
 22:48
Zoffix chicks Geth 22:49
Oh it did report it... never mind :)
I think that tells me I should call it a day :D
nadim if we lived in house without windpws we would't know if it was day or night anymore 22:51
Zoffix: just before you leave, giveen that I have an object, I want to get an undefined object out of it. the meta class. how do I do that? 22:52
Zoffix nadim: I've no idea what you mean... Do you mean the type object or the meta class? 22:54
m: say WHAT 42
camelia (Int)
Zoffix m: say 42.HOW
camelia Perl6::Metamodel::ClassHOW.new
nadim does WHAT return a string or the meta object? 22:55
Zoffix Neither. It returns the type object 22:56
andreoss m: 1.WHAT.new
camelia ( no output )
andreoss m: say 1.WHAT.new
camelia 0
nadim thanks, let's see what I can do with tht :) 22:57
Zoffix s/chicks/kicks/; # heh 22:59
Geth modules.perl6.org: 6d00aeef99 | (Zoffix Znet)++ | assets/sass/main.scss
Get rid of page shift "bug"
 23:02
andreoss modules.perl6.org is 500 23:10
rackstrawe In the statement "while $++ > 5 { .say }", what does the construction "$++" do, and how is it parsed? 23:20
jnthn Parsed ($)++ which is short for (state $)++ 23:21
That is, declaring an anonymous state variable 23:22
rackstrawe thanks!
sjn Hey, it seems modules.perl6.org is down (500 server error) 23:26
MasterDuke Zoffix: ^^^ 23:28
sjn build log says Received fatal error while building raw.githubusercontent.com/perl6/fo...ETA6.json: Use of uninitialized value in string eq at lib/ModulesPerl6/DbBuilder/Dist/PostProcessor/p05DataFetcher.pm line 56.
sjn starting from dist number 110 of 870 23:30
MasterDuke i guess the content has no encoding set 23:36
zengargoyle darn. 23:44
yoleaux 20:50Z <Zoffix> zengargoyle: this fixes the MANIFEST in TODO issues you were having (fixes it as well as moves it to be only for `cpan` dist source): github.com/perl6/modules.perl6.org...5db8ada712
zengargoyle .tell rackstrawe and in 'say $++ while $++ < 5;' those are two different $ state varibles. to me, they are *very* local and a bit tricky to undersand. 23:46
yoleaux zengargoyle: I'll pass your message to rackstrawe.
zengargoyle doesn't get `cpan` dist source but it can wait. 23:48
.tell Zoffix thanks for looking and i don't quite get what you mean. but i think i'll just wait a while before i worry about CPAN integration stuff. 23:51
yoleaux zengargoyle: I'll pass your message to Zoffix.
zengargoyle it's not like i ever put p5 modules on CPAN either. :) (yay DarkPAN) 23:52
Zoffix is back
yoleaux 23:51Z <zengargoyle> Zoffix: thanks for looking and i don't quite get what you mean. but i think i'll just wait a while before i worry about CPAN integration stuff.
Zoffix sjn: looking into it
zengargoyle remembers my CPAN account/name is totally not ZENGARGOYLE. 23:54
Zoffix sjn: it's back up 23:56
zengargoyle: I meant that the MANIFEST check no longer gets run for dists on github/gitlab; only for those from CPAN 23:57
And the glitch where it said there were no manifest when there was a manifest is gone too
But... looks like the NEWDB rebuild ain't working yet :\ wonder wtf
zengargoyle oh, so it's like going to check CPAN for .tar.gz dists separate from git-ish links in the p6 ecosystem list-o-urls....? 23:59
Zoffix CPAN urls won't be in list-o-urls
You just upload to CPAN

Powered by Cro and the Raku® Programming Language.

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

Thank you!