Perl 6 - the future is here, just unevenly distributed

IRC log for #metacpan, 2015-08-28

| Channels | #metacpan index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
01:03 vroom joined #metacpan
01:12 metacpan joined #metacpan
01:12 metacpan [metacpan-web] oalders pushed 1 new commit to master: http://git.io/vGI7K
01:12 metacpan metacpan-web/master fb28c90 Olaf Alders: Merge pull request #1587 from CPAN-API/profile-extra-json...
01:12 metacpan left #metacpan
02:06 vroom joined #metacpan
04:24 SineSwiper joined #metacpan
05:39 jwang joined #metacpan
06:56 oiami joined #metacpan
07:28 pjfl joined #metacpan
07:53 jdv79 so, vagrant or something is cranky
07:54 jdv79 https://gist.github.com/anonymous/e69cd38eb2d90556e12d
07:54 jdv79 anyone?
07:54 jdv79 oh, nm
07:54 jdv79 i read it wrong
07:54 jdv79 local issue
07:56 jdv79 downloading... cool.
08:05 jdv79 https://gist.github.com/anonymous/e52da9e948934c14862f - is that an error worthy of help?
08:17 dolmen joined #metacpan
08:23 jdv79 ok, i'm kinda stuck:  https://gist.github.com/anonymous/6c942efb6a1dcf5776f3
08:40 jdv79 oh...  its up
08:40 jdv79 had to enable virt support in the bios
08:40 jdv79 nice
08:42 jdv79 oh, there's red text
09:18 margeas joined #metacpan
09:55 melo joined #metacpan
10:52 ranguard jdv79: i will be around at 7:30pm BST if you need help
10:53 ranguard you do get some red text... sometimes
10:53 * ranguard wave bbl
10:58 margeas joined #metacpan
11:25 jdv79 ok
11:25 jdv79 so, one thing i already don't like about working in random containers is not having my shell and editor config
11:25 jdv79 meh, not a big deal - seem to work overall - the vm i mean
11:26 nicomen just import your homedir?
11:49 oiami joined #metacpan
12:17 vroom joined #metacpan
12:41 oalders jdv79: you shouldn't need to work in the VM for the most part
12:42 oalders you can edit all of the files on your host machine.
12:42 oalders you'd just need to get in there to restart services
12:43 oalders you've got all of the repositories cloned on the host, so it's easiest to work on them there rather than via the VM
12:52 jdv79 i haven't worked with containers much so i'm not sure what i'm doing
12:56 jdv79 the docs for metacpan-developer say to ssh into the vm and do stuff there, right?
12:58 ranguard jdv79: umm, the 'src' dir (in the mc-developer dir should hvae the repos cloned that you can edit directly
12:59 ranguard they are then mounted into the VM, so you can run cmds in the VM, but edit files on the host
12:59 jdv79 so i tried to run the tests in the api checkout and got module missing failures
13:00 ranguard in the VM, run puppet, that will install everything
13:01 jdv79 is that in the README?
13:04 * ranguard goes to check
13:07 jdv79 how do i call puppet
13:07 ranguard cd /etc/puppet/
13:07 ranguard sudo sh run.sh
13:08 * ranguard boots up his VM to check
13:09 jdv79 is it ok if i get red output from puppet?
13:09 ranguard probably, though depends what.. just running it my self
13:10 jdv79 nope.  tests still fail the same.
13:10 ranguard "Warning: Non-string values..." is ok to ignore
13:10 jdv79 Missing stuff like Digest::SHA1, Perl::Critic, etc...
13:10 ranguard what other red stuff did you get ?
13:12 jdv79 https://gist.github.com/anonymous/1744939baedda3ffcce6
13:14 ranguard jdv79: hmm, do a 'sudo apt-get update'
13:14 ranguard then run puppet again
13:15 vroom joined #metacpan
13:17 jdv79 https://gist.github.com/anonymous/533b5898b65d1668a1a9
13:17 jdv79 i haven't run deb type systems
13:18 jdv79 that output looks bad though
13:18 vroom joined #metacpan
13:19 ranguard yea, http://debian.yorku.ca seems to be fubared
13:19 jdv79 apt-get doesn't have a way to choose another?
13:19 jdv79 or do i have to import/edit something?
13:20 ranguard just checking how it's setup
13:22 ranguard please change /etc/apt/sources and change to: https://gist.github.com/ranguard/adc5ff92ed8126f99009
13:22 ranguard then sudo apt-get update
13:23 * ranguard is just doing the same to test
13:23 jdv79 it seems to be making progress
13:25 jdv79 feels like less puppet blood
13:25 jdv79 maybe...
13:26 jdv79 tests still fail 
13:26 ranguard try the web ones... https://github.com/CPAN-API/metacpan-developer#run-the-tests
13:27 jdv79 i don't get how those module could not be installed
13:27 jdv79 same thing - a bunch bomb
13:29 ranguard hmm, wfm...
13:29 jdv79 good times - looks like Perl::Critic is in ~/perl5...
13:30 ranguard oalders: ^^ over to you I'm afraid, $child now demanding attention
13:30 jdv79 that's cpanm or locallib or something, right?
13:30 jdv79 i guess i'll figure a way to install modules in the system perl since taht seems to be the one being used
13:30 ranguard jdv79: it should all be ending up in ~/carton/..
13:30 ranguard it shouldn't be usning system perl afaik
13:31 ranguard you re doing bin/prove t (e.g. the one in the DIR, not the system one?)
13:31 ranguard sorry have to go *wave*
13:31 jdv79 perl -v is hitting all stuff under /opt/perl-5.18.2
13:31 jdv79 no
13:31 jdv79 systemm
13:31 jdv79 i'll try that tehn
13:31 ranguard ahh, that's the problem!
13:31 jdv79 who has their own prove?  honestly:)
13:31 jdv79 ok, maybe that will work better
13:32 jdv79 specialization-- :(
13:33 jdv79 anyway, thanks!  the tests are chgging along
13:33 jdv79 *chugging
14:02 Relequestual joined #metacpan
14:02 oalders jdv79: what happens if you "vagrant provision"?
14:02 oalders that should basically do everything, including module installs
14:03 oalders bin/prove is just a wrapper to make sure the right libs etc are being used
14:03 oalders i had the same issues with apt-get recently, so i guess it's something we should look at
14:10 vroom joined #metacpan
14:17 jdv79 it seems to all be working much better now
14:18 oalders jdv79: did you get past the apt sources problems?
14:19 jdv79 i even got a p6 dist imported, broken, but partially there already!
14:19 oalders nice!
14:19 jdv79 i'm sure it'll take a lot of work to fix it up that was fast rogress
14:20 jdv79 basically just imported it...
14:20 jdv79 nice
14:20 oalders ok, that's great to hear
14:20 jdv79 it seems th ancilary issues have abated for now - on to real work.  thanks.
15:37 vroom joined #metacpan
17:21 vroom joined #metacpan
17:31 vroom joined #metacpan
17:35 vroom joined #metacpan
18:06 metacpan joined #metacpan
18:06 metacpan [metacpan-developer] ranguard pushed 1 new commit to master: http://git.io/vGOqH
18:06 metacpan metacpan-developer/master 77f705b Leo Lapworth: Make it clearer that one should use ./bin/prove
18:06 metacpan left #metacpan
18:10 ranguard oalders: do you run `vagrant provision` to run puppet on the VM each time, or only the first (using /etc/puppet/run.sh otherwise) ?
18:10 oalders ranguard: i mostly use "vagrant provision" because i'm lazy
18:11 ranguard oalders: I guess the question is which should I put in the docs?
18:11 ranguard as currently if someone sees errors they don't know they can just re-run it after fixing the issue
18:12 oalders i like "vagrant provision" because it's easier to remember and you don't have to log in to the VM
18:12 ranguard ok
18:12 oalders and it can run some additional cleanup that puppet can't handle
18:12 oalders or doesn't handle right now
18:12 oalders we might want to add something to fix up the apt sources
18:13 oalders i'm guessing the fact that it picked a mirror close to me is because i set up the box
18:13 ranguard oalders: yea, it's all your fault ;)
18:13 oalders :)
18:14 ranguard we have a puppet module for managing apt/sources but it's not configured at the moment other than for panatopia
18:15 oalders ah, maybe we could tweak that then
18:15 oalders i tried a provision today and got the same flood of messages
18:16 metacpan joined #metacpan
18:16 metacpan [metacpan-developer] ranguard pushed 1 new commit to master: http://git.io/vGO3I
18:16 metacpan metacpan-developer/master 948fa69 Leo Lapworth: Mention that vagrant provision can be run several times, and relation to puppet
18:16 metacpan left #metacpan
18:16 oalders ranguard++
18:17 ranguard hmm, production sources should probably be left as is ( using the hosting companies mirrors )
18:18 oalders sure. if we can just have it on the vagrant node, that's fine
18:18 ranguard created a ticket for myself for YAPC::EU
18:22 oalders :)
18:22 oalders when do you leave?
18:27 ranguard Tuesday, got several hours between planes so maybe hacking then
18:41 oalders vagrant box is good for that :)
18:45 Relequestual joined #metacpan
19:01 vroom joined #metacpan
20:04 jdv79 so, if i print to STDERR shouldn't that end up in the startman error log?
20:19 ranguard jdv79: no, because it's Catalyst via plack... best to use `plackup` instead of the service if you want that level of debug
20:21 ranguard oalders: ^^ what's the invocation from within the VM?
20:22 ranguard jdv79: on the VM:
20:22 oalders ranguard, jdv79 : i think it's ./bin/run plackup in the appropriate folder
20:22 ranguard oalders: cheers
20:23 oalders yep, that works
20:23 ranguard ahh, yea, I'd just found the long winded way (which is what `run` bascially does for you)
20:23 oalders :)
20:24 jdv79 its been as while since i used "modern" p5 web stuff.  sorry.
20:24 ranguard jdv79: you'll regret using that new fangled 6 stuff ;)
20:25 jdv79 i would prefer it not print logging to the console.  is there a conf i can set to log to a normal file?
20:25 ranguard oalders: ^^ ?
20:25 jdv79 actually, for the past 8 or so years i was stuck on mod_perl(mostly 1.3)/mason/that stack...
20:25 ranguard jdv79: I do puppet/sys stuff here, but avoid most of the actual code
20:25 * ranguard comforts jdv79
20:26 jdv79 i was around when sri gave birth to catalyst but after about a year after i got sucked into legacy
20:27 oalders i only got away from mod_perl last december...
20:27 ranguard ick, 8 years is a LONG time for legacy not to get upgraded
20:27 oalders not sure about redirecting the logging
20:28 jdv79 well, i said mostly because ?2? years til the end it was upgraded to plack that wrapped the legacy app so for all intents it was teh same
20:28 oalders if you deploy via Starman you get --error-log
20:29 oalders plackup gets you --access-log
20:29 jdv79 ok, i'll try some things
20:29 ranguard arrg, battery going... good night
20:29 oalders ranguard: good night!
21:00 margeas joined #metacpan
21:55 vroom1 joined #metacpan
22:46 jdv79 so i've loaded the same dist a few times and the api is giving me back different answers
22:47 jdv79 is there a cache or something else i need to do to fix hat
22:47 jdv79 or, better yet, what is a good way to refresh/reload a dist?
23:06 jdv79 whoa.  is the syntax highligher written in js?
23:42 oalders jdv79: syntax highlighting is in js
23:42 oalders if you're hitting the local api then there should be no caching
23:42 oalders just make sure you are actually hitting your local api and not going over the network
23:43 oalders ie check the config in metacpan-web.  by default it goes to the live api
23:43 oalders also, just do proper searches on metacpan-web.  the search suggestions are from a 3rd party and they won't take your local api into account
23:45 metacpan joined #metacpan
23:45 metacpan [metacpan-web] oalders deleted profile-extra-json at 15e06bd: http://git.io/vGsfc
23:45 metacpan left #metacpan
23:52 jnbek joined #metacpan

| Channels | #metacpan index | Today | | Search | Google Search | Plain-Text | summary