Perl 6 - the future is here, just unevenly distributed

IRC log for #metacpan, 2014-12-17

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

All times shown according to UTC.

Time Nick Message
00:05 haarg cpanmetadb.plackperl.org is miyagawa's
00:19 neilb_ joined #metacpan
01:20 metacpan joined #metacpan
01:20 metacpan [metacpan-developer] rose opened pull request #28: Docs (master...docs) http://git.io/Ueazbg
01:20 metacpan left #metacpan
01:20 dipsy [ Docs by rose · Pull Request #28 · CPAN-API/metacpan-developer · GitHub ]
01:38 riveter what version of ubuntu is mc-dev based on?
01:54 metacpan joined #metacpan
01:54 metacpan [metacpan-developer] oalders closed pull request #28: Docs (master...docs) http://git.io/Ueazbg
01:54 metacpan left #metacpan
01:54 dipsy [ Docs by rose · Pull Request #28 · CPAN-API/metacpan-developer · GitHub ]
01:54 oalders riveter: it's debian
01:55 oalders should be wheezy
01:55 oalders https://github.com/CPAN-API/metacpan-vagrant
01:55 dipsy [ CPAN-API/metacpan-vagrant · GitHub ]
02:02 riveter whoa, how did I not see that before?
02:03 riveter thanks
02:08 klapperl_ joined #metacpan
02:38 oalders :)
04:51 khisanth_ joined #metacpan
07:45 neilb joined #metacpan
08:23 neilb joined #metacpan
08:55 neilb joined #metacpan
09:17 neilb joined #metacpan
12:23 punter joined #metacpan
13:23 oiami joined #metacpan
13:55 vroom joined #metacpan
14:43 Ptolemarch joined #metacpan
15:14 alnewkirk joined #metacpan
15:16 Ptolemarch joined #metacpan
15:18 Ptolemar_ joined #metacpan
15:43 Ptolemarch joined #metacpan
16:18 andreeap joined #metacpan
16:20 andreeap when I set up the vm , I chose a full CPAN mirror, but now this is happening http://i.imgur.com/vukKLMk.png?1
16:42 Ptolemarch joined #metacpan
17:16 Ptolemarch joined #metacpan
17:38 Ptolemarch joined #metacpan
17:52 punter joined #metacpan
18:22 SineSwiper joined #metacpan
18:30 riveter after reprovisioning my mc-dev machine, ssh authentication fails.  I can log in manually as the vagrant user.  Anyone have any idea why that would happen?
18:31 riveter ranguard rwstauner ^
18:38 riveter vagrant ssh-config says it's using .vagrant/machines/default/virtualbox/private_key as the identity file
18:38 riveter it doesn't match the private one on https://github.com/mitchellh/vagrant/tree/master/keys
18:38 dipsy [ vagrant/keys at master · mitchellh/vagrant · GitHub ]
18:38 riveter but the public one in /home/vagrant/.ssh is the same
18:39 riveter (as the public one in that repo)
18:45 rwstauner riveter: what OS is your host, what version is your vagrant?
18:45 rwstauner i don't have that private_key file that you do
18:45 riveter oO
18:45 rwstauner and i haven't had the issue you're describing
18:45 riveter interesting
18:46 riveter well, copying the private key from that repo fixed it
18:46 rwstauner that's good
18:46 rwstauner it sounded like it should be that easy
18:46 rwstauner but i don't know why it was out of sync in the first place
18:46 riveter I don't see how provisioning could've changed that though
18:46 riveter yeah
18:46 rwstauner no, it shouldn't
18:46 riveter what version of vagrant are you using?
18:47 rwstauner unless it was a weird vagrant thing (like you upgraded it)
18:47 rwstauner 1.6.5
18:47 riveter I'm on 1.7.1
18:47 riveter and I'm using arch so I do upgrade things regularly, I guess that could be it
18:47 riveter can you tell me what your vagrant ssh-config looks like?
18:48 rwstauner "If the default insecure keypair is used, Vagrant will automatically replace it with a randomly generated keypair on first vagrant up. [GH-2608]"
18:48 rwstauner that's in the vagrant changelog for 1.7.0
18:48 riveter derp
18:48 rwstauner https://github.com/mitchellh/vagrant/blob/master/CHANGELOG.md
18:49 dipsy [ vagrant/CHANGELOG.md at master · mitchellh/vagrant · GitHub ]
18:49 riveter so, but shouldn't it have replaced the public key too?
18:49 rwstauner i would think so
18:49 riveter I just spent more time than I want to admit on this :P
18:49 riveter ah well, thanks for your help
18:50 riveter nice to know I'm not just crazy
18:50 rwstauner unless metacpan's puppet config overwrites is
18:50 rwstauner s/is/it
18:50 rwstauner ranguard: ^
18:50 * rwstauner says that b/c the file exists
18:50 * riveter rubs hands together eagerly and runs provision again
18:52 riveter this will take a while
18:53 ranguard don't think it does
18:53 * ranguard can't be 100% sure though as could be a knock on from wanting the dev machines to run as 'vagrant' rather than 'metacpan' as we have in production
18:55 neilb joined #metacpan
18:55 ranguard vagrant doesn't seem that great at backwards compat I must say, we've all wasted a lot of time on things like that
18:56 riveter I wonder if docker would be better for this use case
18:56 riveter we want to be able to send people an up to date image
18:56 ranguard probably not to be honest, you still need all the setup / users / accounts / firewalls etc
18:57 ranguard though someone has started https://github.com/CPAN-API/metacpan-docker as a play project
18:57 dipsy [ CPAN-API/metacpan-docker · GitHub ]
18:57 riveter oh neat
18:57 ranguard we are very much NOT supporting it, unless someone wants to convert EVERYTHING over to it :)
18:57 riveter heh
18:58 riveter wise
19:01 rwstauner i don't see why we need the ssh file in puppet for the vagrant user, unless it's just a consistency thing
19:01 rwstauner in which case i say we make it configurable
19:02 riveter is puppet setting it?
19:02 riveter that would explain a lot
19:02 rwstauner yeah, that's what i meant
19:03 riveter mystery solved, good job gang
19:03 rwstauner https://github.com/CPAN-API/metacpan-puppet/tree/master/modules/metacpan/files/default/home/vagrant/ssh
19:03 dipsy [ metacpan-puppet/modules/metacpan/files/default/home/vagrant/ssh at master · CPAN-API/metacpan-puppet · GitHub ]
19:03 riveter yeah, that's going to be painful once people start upgrading to 1.7
19:03 rwstauner it makes sense to have the ssh files like that for all the other users
19:03 rwstauner but not vagrant
19:04 rwstauner however it may make sense to have puppet know about the vagrant user for other things (like repo paths)
19:04 rwstauner so we probably need to configure the user type to not do the ssh file for vagrant
19:04 riveter what if we just deleted that file?
19:04 riveter would it handle that ok?
19:05 rwstauner probably, but i need to check the puppet conf
19:05 ranguard riveter: try it and see :)
19:05 riveter xp
19:05 * riveter shakes fist impotently at the general concept of provisioning tools
19:06 ranguard https://github.com/CPAN-API/metacpan-puppet/blob/master/modules/metacpan/manifests/user.pp#L83 is the bit that's going to get broken by it I think
19:06 dipsy [ metacpan-puppet/user.pp at master · CPAN-API/metacpan-puppet · GitHub ]
19:06 rwstauner yeah, it's specifically trying to copy that file
19:06 riveter ftp was good enough for my grandaddy and durn it it's good enough for me
19:06 riveter hm
19:06 rwstauner so we might need to make that section conditional
19:07 riveter I'm behind on my stuff for today but I can take a look at that tomorrow if you want
19:07 ranguard riveter: next you'll be saying who needs a dev/stage box or version controll for that matter :)
19:07 ranguard riveter: all yours - thanks
19:07 riveter yeah, and don't get me started on these high level languages
19:07 ranguard please create yourself a issue so we know why
19:07 rwstauner an assembly fan, are you?
19:08 riveter darn tootin'
19:08 riveter can't wait until this whole 'compiler' fad fades out
19:08 * riveter mutters ominously about c-writing hipsters
19:09 riveter ok, I have to actually go pretend to do something useful
19:27 oalders andreeap: did you index the modules in your CPAN?
19:31 andreeap yes
19:36 neilb_ joined #metacpan
19:37 oalders andreeap: which command did you run to index them?
19:41 andreeap I followed the instructions for the full cpan mirror in the  README_API.md
19:50 oalders andreeap: what happens if you click on "all releases by this author" on that page
19:52 andreeap this http://i.imgur.com/Q4Gh4MS.png
19:55 oalders andreeap: so either the releases were never indexed, or the indexer found nothing in your CPAN
19:56 oalders can you re-run the indexer and check the output in your terminal?
19:56 andreeap yes, I will try that
20:00 ranguard rebooting bm-mc-03
20:03 Ptolemarch joined #metacpan
20:04 andreeap I can't even re-run the indexer, I get errors
20:05 oalders andreeap: gist?
20:06 andreeap I just wanted to do that, just a sec
20:08 andreeap https://gist.github.com/andreeap/c98c3cf07cd79842dcd3
20:08 dipsy [ cpan index ]
20:12 * ranguard is upgrading nginx and various bits - please yell if stuff breaks
20:13 oalders andreeap: which command are you running there?
20:15 andreeap ./bin/run bin/metacpan release $MINICPAN/authors/id/
20:16 oalders andreeap: what do you see when you run "./bin/run carton install"
20:17 andreeap https://gist.github.com/andreeap/c78487d8beda1b8df6e6
20:17 dipsy [ carton install command result ]
20:19 oalders andreeap: what do you see for "which perl"
20:19 andreeap sorry for that
20:20 andreeap https://gist.github.com/andreeap/d5f933de95530fe148cf
20:20 dipsy [ which perl ]
20:20 oalders there's your problem
20:20 oalders try "source ~/.metacpanrc && which perl"
20:21 andreeap it's the same
20:22 oalders what do you see with " cat ~/.metacpanrc"
20:23 andreeap https://gist.github.com/andreeap/807be5df8733a25edb65
20:23 dipsy [ gist:807be5df8733a25edb65 ]
20:23 oalders that's really old
20:23 oalders andreeap: 1) make sure you've got the latest commits on all of your repositories 2) run "vagrant provision"
20:24 andreeap ok, thanks
20:30 ranguard all servers upgraded (bm-mc-01 could do with more debian packages upgraded, but don't want to do that until we have a hot fail over because I'm a chicken :)
20:39 oalders :)
20:39 oalders you don't want to start moving indices between servers?
20:42 ranguard I have no clue about clustering 0.20.0 - only the 1.4 on liquidweb is setup as cluster
20:42 ranguard once we can do a full reindex in 0.20 I'll be more comfortable
20:56 jnbek joined #metacpan
21:50 gtodd does metacpan.org (cpan.org) indicate whether a module is: in CORE; when @ which version it became a CORE module; ... i.e. if a module works as a "dual life" module what dependencies will it need to install etc.
21:50 gtodd errm I guess I mean "Metacpan API"
21:52 ranguard gtodd: not at the moment afaik
21:52 gtodd hmm I guess can just grab stuff with MetaCPAN::Client and make comparisons
21:52 ether gtodd: that's all in Module::CoreList and the corelist executable
21:52 gtodd with CoreList
21:52 gtodd yeah that
21:53 ranguard See: ... yea
21:53 gtodd :)
21:53 ranguard or submit a patch to the api to do it ;)
21:54 neilb gtodd: also see: https://github.com/CPAN-API/metacpan-web/issues/403
21:54 dipsy [ "Core Module" Indicator in search results · Issue #403 · CPAN-API/metacpan-web · GitHub ]
21:56 gtodd ranguard:  heheh ...  it will have that feature one day ... since perl is a living being in an ectoplasm of cpan and human thought
21:56 jnbek joined #metacpan
21:58 gtodd I mean ectoplasm's second definition ... and only metaphorically
21:59 gtodd neilb: thanks ... something to think about
22:01 ether if the API had that information, we could present it on the webpage... e.g. omitting prereqs from being listed if they are core as of <some perl version>
22:01 ether dunno how useful that would be though.
22:01 ether there's nothing really wrong with some redundant prereqs being listed
22:02 alh Is it a good idea to not list the prereqs? What if the person looking has an old perl that doesn't yet have the modules that are now in core?
22:03 ranguard you could let people set the min perl version - or a specific version they care about and update results accordingly
22:09 gtodd I'm starting to think I just want to know lots about my own perl and have metacpan API provide "research resources" for my scripts (that probably copy and/or use cpanm) ... and the best way to know my perl is to stick to system/local separation; perlbrew/local::lib etc   sometimes having to work backwards to get a carton/pinto description for an installation that already exists is necessary.
22:10 gtodd but the tools are there to describe a perl/+cpan installation pretty thoroughly
22:14 oalders gtodd: have you looked at https://metacpan.org/pod/Dist::Surveyor ?
22:14 dipsy [ Dist::Surveyor - Survey installed modules and determine the specific distribution versions they came from - metacpan.org ]
22:17 gtodd ok that will be perfect
22:20 gtodd not sure how it missed it ... I think I mixed it up with something I'd used in the past
22:20 gtodd thanks all, cheers
22:22 Ptolemarch joined #metacpan
22:36 andreeap ok, thanks89/
23:22 neilb joined #metacpan
23:31 neilb_ joined #metacpan

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