Camelia, the Perl 6 bug

IRC log for #metacpan, 2013-07-29

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

All times shown according to UTC.

Time Nick Message
00:44 shmuel joined #metacpan
01:07 klapperl joined #metacpan
01:11 ether_ joined #metacpan
01:34 rwstauner dipsy: forget all users
01:35 dipsy rwstauner: I forgot about all users.
02:17 ether joined #metacpan
03:07 thaljef joined #metacpan
04:27 metacpan joined #metacpan
04:27 metacpan [metacpan-web01] oalders pushed 1 new commit to master: http://git.io/NU8aYA
04:27 metacpan metacpan-web/master a3d56a4 Olaf Alders: Updates 'About MetaCPAN' page.
04:27 metacpan left #metacpan
04:27 dipsy [ Updates 'About MetaCPAN' page. · a3d56a4 · CPAN-API/metacpan-web · GitHub ]
04:35 shmuel joined #metacpan
06:26 Mike-PerlRecruiter_ joined #metacpan
06:39 daxim joined #metacpan
08:24 plu joined #metacpan
08:29 tobyink joined #metacpan
09:01 dolmen joined #metacpan
09:54 bowtie_ joined #metacpan
10:50 tobyink joined #metacpan
11:32 ether_ joined #metacpan
11:49 dsteinbrunner joined #metacpan
12:41 tobyink joined #metacpan
13:43 Skunix joined #metacpan
13:57 Skunix joined #metacpan
15:02 oalders http://blogs.perl.org/users/toby_inkster/201​3/07/whos-relying-on-that-cpan-release.html
15:02 dipsy urgh. long url. Try http://tinyurl.com/jwxfg6g
15:02 dipsy [ Who's relying on that CPAN release? | Toby Inkster [blogs.perl.org] ]
15:17 talexb joined #metacpan
16:16 tobyink joined #metacpan
16:29 jayallen joined #metacpan
16:31 ether a little chart for recursive upstream would be cool, like there are for recursive downstream
16:31 ether although the presentation would have to be different
16:32 ether iirc someone did one of the ecosystem depending on Moose
16:34 BinGOs I would have my usual rant about dynamic_config, but it's too hot and I can't be arsed.
16:57 ether should the "...and possibly others" string in the right sidebar be conditional on the dynamic_config flag?
16:57 ether perhaps "..and *probably* others" if it is set, at least
16:58 * ether notices that the optional_feature dzil plugin isn't setting dynamic_config - doh! will fix tonight.
17:55 talexb Just trying to run the API test suite .. it seems that 'sudo /opt/elasticsearch-0.20.2/bin/elasticsearch -f -Des.http.port=9900 -Des.cluster.name=testing' takes a long time ..
17:56 talexb The VM appears to be quiescent, but no command prompt .. and it's been six minutes.
17:57 rwstauner talexb: -f means "run in the foreground" so it's never going to return
17:57 talexb Unless that's supposed to be left along, while the test suite is run from another ssh session.
17:57 rwstauner it should be printing some basic output though
17:57 rwstauner in the vm i think ES is already running
17:58 talexb rwstauner, OK -- that makes sense. But the instructions don't indicate that.
17:58 * rwstauner hasn't reviewed the instructions
17:58 talexb :)
17:58 talexb No problem -- I'll suggest some updates.
17:58 rwstauner please do
18:03 ranguard talexb++ # feedback
18:04 ranguard ES will likely also not habe any data in it As far as i know (this mayhave been covered)
18:04 talexb :) Issue #3 posted.
18:04 talexb I'm still brand new to this project, so my plan is to learn and issue bugs as I go.
18:05 talexb Hopefully the Stupid Bug ratio will remain low.
18:05 talexb My next issue is that when I ran the test script, 'Test::Aggregate::Nested version 0.366 required--this is only version 0.364' .. is that an issue that I should report, or just install a newer module and move on?
18:06 * talexb wonders if that's something that should be in the provisioning script.
18:07 rwstauner just install the newer module
18:07 talexb Okey doke. Thanks.
18:07 rwstauner the api repo declares it, so that 'cd $dir; install_modules' thing should work
18:07 rwstauner though i did consider adding known missing deps to the provisioner temporarily
18:08 rwstauner it may be worth doing
18:08 rwstauner ('temporarily' meaning until a newer vm is built)
18:08 talexb OK -- I freshened all of the repos before I started .. maybe I should do the install_modules step as well.
18:09 rwstauner yeah, do that for the web and api repos (/home/metacpan/metacpan.org and /home/metacpan/api.metacpan.org)
18:12 talexb Done .. the basic command 'sudo /home/metacpan/bin/install_modules --installdeps .' said Nothing to install, so then I explicitly told it to install Test::Aggregate::Nested and all was well.
18:14 rwstauner ok
18:14 rwstauner i still don't know why that doesn't work for you
18:14 rwstauner it should
18:14 rwstauner but installing them by name is fine for now
18:55 sawyer_ joined #metacpan
18:57 talexb OK, got it upgraded after manually typing 'sudo /usr/local/perlbrew/perls/perl-5.16.2/bin/cpanm Test::Aggregate::Nested'
19:24 rwstauner i thought you said doing as 'install_modules Test::Aggregate::Nested' worked
19:32 talexb I did -- then I realizes that it looked like it installed on the system Perl, and not the perlbrew location.
19:32 talexb That's to say, it showed that it installed correctly, but the prove command still failed with the same error.
19:33 oalders talexb: did you source ~/.metacpanrc first?
19:33 talexb Yes.
19:34 oalders what does "which perl" give you?
19:34 talexb I think the call to cpanm to find the dependencies didn't work properly.
19:34 talexb which perl gives me /usr/local/perlbrew/perls/perl-5.16.2/bin/perl
19:35 talexb I added 'set -x' to the install_modules script, and the cpanm command line looks wrong.
19:37 rwstauner something is wrong about how that's working but i haven't figured out what it is yet
19:37 talexb http://nopaste.info/a6daed8790.html
19:37 dipsy [ Nopaste - powered by project-mindstorm IT Services ]
19:40 rwstauner when i do that i get totally different output
19:41 rwstauner are you sure the puppet run finishes successfully?
19:43 shadowpaste joined #metacpan
19:43 shadowpaste "rwstauner" at 217.168.150.38 pasted "metacpan@debian:~/api.metacpan" (5 lines) at http://paste.scsys.co.uk/264268
19:43 dipsy [ magnet_web paste from "rwstauner" at 217.168.150.38... ]
19:43 talexb Puppet did complain about differences in /etc/resolv.conf .. but it seemed to finish OK.
19:44 rwstauner puppet will show a diff when it changes some of the files
19:44 rwstauner i clipped the above output to just the first two modules... the list is dozen of lines long
19:44 rwstauner but as the same user in the same dir running the same command i'm getting totally different output
19:45 rwstauner talexb: cat /home/metacpan/.metacpanrc
19:45 talexb export PATH=/usr/local/perlbrew/p​erls/perl-5.16.2/bin:$PATH
19:46 rwstauner yeah, that's right
19:46 rwstauner hmm
19:46 talexb And I have /usr/local/perlbrew/perls/perl-5.16.​2/lib/site_perl/5.16.2/CPAN/Faker.pm and /usr/local/perlbrew/perls/perl-5.16.2/li​b/site_perl/5.16.2/Perl/PrereqScanner.pm installed ..
19:46 rwstauner ok
19:47 rwstauner what does 'git describe' say in /home/metacpan/api.metacpan.org ?
19:48 talexb v0.2.0-656-g5c16692
19:48 rwstauner yep
19:48 talexb (Phew)
19:48 rwstauner hmm
19:50 talexb I'm not too worried about it -- it's a loose end I'd like to sort out later.
19:50 rwstauner ok
19:51 talexb In the meantime, I'd like to make some updates to the metacpan-developer README .. and I'm a nebie about this ..
19:51 rwstauner /usr/local/perlbrew/perls/perl-5.16.2/bin/cpanm --version
19:51 rwstauner i have 1.6922
19:52 talexb http://nopaste.info/3e28c110d2.html
19:52 dipsy [ Nopaste - powered by project-mindstorm IT Services ]
19:52 talexb Likewise.
19:53 rwstauner looks the same
19:53 rwstauner so weird
19:53 rwstauner so do you have questions about editing the README and submitting a pull request?
19:54 talexb So do I make a branch on the repo that I cloned, and push that branch?
19:54 rwstauner yes
19:54 talexb Or do I push my version of the repo to github and point that to you?
19:54 rwstauner did you fork the repo into your github account?
19:54 talexb OK, I'll do that -- and I'll name the branch with the issue number I guess.
19:54 rwstauner or did you clone directly from the cpan-api repo?
19:54 talexb Oh -- no, I didn't.
19:54 talexb I cloned directly.
19:55 rwstauner start with that
19:55 talexb So, blow away the previous version?
19:55 rwstauner no, you don't need to do that
19:55 rwstauner if you click the "Fork" button on github
19:55 rwstauner it will take you to your fork
19:55 rwstauner and you can just edit your git config
19:55 talexb Change the remote, I guess.
19:55 rwstauner git remote rename origin upstream
19:56 rwstauner git remote add origin $YOUR_GITHUB_URL
19:56 talexb Right. OK, I'll get onto that.
19:56 rwstauner then you can make a branch, and commit your changes to it
19:56 rwstauner push the branch to github
19:57 rwstauner and then on your page on the github site it will show you little buttons to compare your branch to the main repo and make a pull req
19:57 talexb Ah. OK. Sorry this is so basic. I've used git for 3-4 years, but haven't really used github with a team before.
19:57 rwstauner sure, np
19:58 rwstauner github tries to make it easy
19:58 oalders and i did not know about "git remote rename" :)
19:58 rwstauner :-)
19:58 oalders talexb: yeah, it gets a bit more interesting with a team
19:58 oalders but if you send pull requests for the doc changes we can integrate them right away
19:59 talexb Right.
19:59 talexb OK, I've forked the repo, done the re-naming and I'm going to start updating the documentation as per today's adventures. Thanks all! :)
19:59 rwstauner talexb++
19:59 talexb rwstauner++
20:06 rwstauner ok, i figured it out
20:06 rwstauner it's another vagrant/perm issue
20:06 talexb Really?
20:06 dipsy no, not really
20:07 talexb LOL
20:07 rwstauner and the problem is that i've used my api repo outside the vm
20:07 rwstauner so my EUMM files are already present
20:07 talexb Right.
20:07 rwstauner but on your vm they don't get built
20:07 rwstauner honestly i thought there was a cpanfile there
20:07 talexb Amazing what gets found when some green guy arrives.
20:07 rwstauner :-)
20:07 oalders i think it was just cpanfile on metacpan-web
20:08 rwstauner probably
20:08 oalders should we just switch everything to cpanfile?
20:08 rwstauner idk
20:08 rwstauner but i can fix this perm issue without that
20:09 oalders right
20:12 rwstauner we probably could
20:13 rwstauner i think the only thing we'd lose is that 'make test' works right
20:13 rwstauner but i don't know if anybody does that
20:13 rwstauner they aren't if they're reading the READMEs
20:13 rwstauner and there's the bin/unlisted_prereqs helper but i'm not sure how useful that is either
20:19 tobyink joined #metacpan
20:25 oalders i don't know if anyone is using 'make test'. i think we could assume that's not needed
20:28 mst I don't understand why using cpanfile for deps would affect make test
20:33 oalders i think the idea is to remove the Makefile.PL
20:33 oalders rather than list deps in 2 places
20:38 mst ah, yes, "I don't know how to do something so let's just start deleting files"
20:38 mst I'll go chase you down a non-stupid answer.
20:40 oalders i'm not sure how this got stupid
20:40 mst ok, it appears the normal approach if using cpanfile
20:41 mst is to use cpanfile as part of the generation of META.json for building a release
20:41 oalders well, in this case we're just installing deps for the API using cpanm
20:41 mst right, but 'cpanm --installdeps .' can use the cpanfile
20:41 mst so I don't see why you need to delete the Makefile.PL
20:42 oalders well, what other purpose does the Makefile.PL serve in our case? we're not releasing this as a tarball
20:42 oalders we just need to deploy the code via git checkout
20:43 oalders and travis deals with cpanfile just fine
20:43 oalders because it defaults to cpanm
20:44 mst I thought the metacpan code was released to cpan
20:44 oalders there are a few pieces on cpan, but it's mostly just github
20:45 mst ok, if you're not actually doing proper releases then you're right, it doesn't matter
20:45 rwstauner it used to be a dist.ini for the sake of listing prereqs
20:46 rwstauner and we switched to Makefile.PL to remove the dzil dependency
20:46 rwstauner for contributors
20:46 rwstauner there was a time when cpanfile wasn't properly bootstrapped for cpanm, but i'm sure that's no longer an issue
20:49 tobyink left #metacpan
21:32 oalders rwstauner: did you see https://github.com/CPAN-API/cpan-api/issues/277 ? is that related to something you were working on?
21:32 dipsy [ API test suite fails / CPAN::Faker source directory not writable · Issue #277 · CPAN-API/cpan-api · GitHub ]
21:32 oalders the t/var stuff
21:32 rwstauner no, but i can fix it
21:35 oalders just wondering how it happened. did i mess up the VM?
21:36 mst oalders: aha!
21:36 rwstauner no
21:36 mst oalders: see #toolchain, it is as easy as I thought.
21:36 rwstauner the vm fixes that issue for web, but not for api
21:36 oalders mst: thanks. checking logs
21:38 grantm joined #metacpan
21:42 oalders mst: thanks for pointing that out
21:43 mst oalders: not that it's -that- important but it'll help people to not be confused :)
21:45 oalders yeah, i don't think we need both files in our case, but i can already think of at least one other place i can make use of that
21:50 mst I'm now tempted to figure out how to integrate it with Distar

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