Camelia, the Perl 6 bug

IRC log for #metacpan, 2013-08-14

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

All times shown according to UTC.

Time Nick Message
00:17 shmuel joined #metacpan
00:24 dsteinbrunner joined #metacpan
01:25 dsteinbrunner joined #metacpan
01:46 klapperl_ joined #metacpan
02:09 ether_ joined #metacpan
02:42 jordan joined #metacpan
03:30 preflex_ joined #metacpan
04:07 Khisanth joined #metacpan
06:26 dpetrov_ joined #metacpan
06:39 ether_ joined #metacpan
07:00 Khisanth joined #metacpan
07:25 Mike-PerlRecruiter_ joined #metacpan
08:54 alnewkirk joined #metacpan
09:15 ether joined #metacpan
11:34 ranguard talexb: hi - have you been able to get anywhere with the sitemaps :) ?
12:59 neilb joined #metacpan
13:01 dsteinbrunner_ joined #metacpan
13:06 neilb Hi. I'm looking up reverse dependencies, using the query that tobyink posted to blogs.perl.org. Is there a way to say "only return these specific fields", as I'm just interested in {metadata}->{name}?
13:10 dsteinbrunner neilb: link?
13:11 neilb http://blogs.perl.org/users/toby_inkster/201​3/07/whos-relying-on-that-cpan-release.html - query is near the bottom of the code
13:11 dipsy urgh. long url. Try http://tinyurl.com/jwxfg6g
13:11 dipsy [ Who's relying on that CPAN release? | Toby Inkster [blogs.perl.org] ]
13:11 dsteinbrunner btw, big fan of you adoption project
13:11 dsteinbrunner I have thought about doing something similar
13:12 neilb cool. That's what I'm working on :-)
13:12 dsteinbrunner I figured
13:12 neilb I've been using David Cantrell's reverse dependency lookup, but it has some bugs, which result in gaps or very wrong counts
13:13 dsteinbrunner so for the ES you can limit what you want like this: "fields": [ "release.name", "release.resources.repository.web" ],
13:13 dsteinbrunner pulled that from a query I have laying around
13:14 dsteinbrunner I'm trying to get authors to update their profiles and update their repo metadata on modules
13:15 dsteinbrunner also checking for typos in modules and reporting bugs or pull requests
13:16 dsteinbrunner here is the metacpan Reverse dependencies lookup https://metacpan.org/requi​res/distribution/Moo?sort=[[2,1]]
13:16 dipsy [ Distributions which depend on Moo - metacpan.org ]
13:16 neilb I noticed all your typo reports: they resulted in a bunch of dists appearing on the adoption list ;-) Maybe you could set the severity of such reports to 'Unimportant'?
13:16 dsteinbrunner there is a link on each module page for it
13:17 neilb Yeah, I nearly thought about scraping those pages, since they're probably fewer bytes than using the API (at the moment).
13:18 dsteinbrunner or you could find the source for it and get ideas from there
13:18 neilb On updating repo metadata, there's a questhub stencil quest for that (http://questhub.io/realm/perl/st​encil/51e2da6e91198f394500009e) - you could always start a quest and set yourself a bigger goal ... :-)
13:18 dipsy [ Questhub.io ]
13:19 dsteinbrunner I kind of figured that was happening with the bugs I'm submitting … and actually think it helps show modules that are in need of at least help if not adoption
13:20 dsteinbrunner on the quest hub bit… I have thought about it but I seem to be doing a good job of it with out it ;-P
13:21 neilb Ah, but QH is not just about doing it yourself, it's also about encouraging others, and letting them see what you're up to, and why
13:21 dsteinbrunner It would be nice if you looked at the github issues also
13:21 neilb Other people see your shit, er, good deeds, and think "that's a good idea, I'll do that to".
13:22 neilb Yeah, github issues on my list too.
13:22 dsteinbrunner metacpan does that if the bug tracker is set
13:22 dsteinbrunner but that does not cover everything
13:23 dsteinbrunner the bug tracker can be set to rt but there are several pull request sitting in github
13:23 neilb yeah. I figured it's less pressing to get to the github issues, because dists on github are generally more actively maintained. But I will get to it eventually ... :-)
13:24 neilb Assuming Ether doesn't keep giving me ideas that distract me ...
13:24 dsteinbrunner another thing… have you thought of or know of an effort to contact the authors
13:24 dsteinbrunner let them know about the list
13:24 dsteinbrunner let them know how many of their modules are on it
13:25 dsteinbrunner suggest that they set ADOPTME if they what it to be adopted
13:25 dsteinbrunner or ask to have it pulled from the list, or whatever
13:26 neilb I think I should wait until the list is a bit better tuned, so I don't piss people off who reply with "It's not bloody well up for adoption, I've just been busy, take my feckin' modules off the list", sorta thing. And tune my preamble wording some more.
13:27 dsteinbrunner Well sure, there needs to be some explanation etc, etc,
13:27 dsteinbrunner and of course regardless you will have someone take it the wrong way
13:27 ranguard neilb: as your here... how goes the Framework review - that you were doing before the perlhub thingy :) ?
13:27 neilb I've emailed a couple of authors, but so far no response, other than some bounces ;-)
13:28 neilb gosh, is that the time? I really must get back to some real work
13:28 ranguard :)
13:28 * neilb slinks away to the shadows shamefully
13:28 dsteinbrunner yeah, I'm dealing with a few of those today myself
13:29 neilb My problem is that I have ideas faster than my ability to code them. They're still on my list *cough*
13:29 ranguard neilb: know what you mean - my todo list is huge - but that's why I try to put as little on it as possible and JFDI :)
13:29 ranguard neilb: your doing other important stuff - so no worries :)
13:30 ranguard s/your/you are/
13:30 neilb The trouble is, todo lists are often stacks, so you just work on the last thing you put on it
13:31 diegok joined #metacpan
13:33 ranguard true
13:40 dsteinbrunner neilb: on that email suggestion.. it would be nice to have the heads up come from the source or someone associated to the source rather than some well intentioned person trying to be helpful and they get the over reaction response
13:40 dsteinbrunner and then also know which authors have bad emails and try to deal with that
13:41 dsteinbrunner make the adoption process as frictionless as possible
13:42 neilb Have had email with ANDK before on the bad email thing. That's another idea on my list, but at least I haven't publicly committed to that.
13:43 neilb when you say "from the source", not sure what you're referring to? Someone who's using the module?
13:43 dsteinbrunner the source of the list
13:43 dsteinbrunner you
13:43 dsteinbrunner ;-P
13:45 Skunix joined #metacpan
13:45 neilb Ah, right. Doh.
14:09 talexb ranguard, I am working on the sitemaps, but I'm having challenges getting the API search tools to work.
14:10 talexb My plan is to call the API and search for the various bits of data, build the gz files, then create the robot.txt from that.
14:10 talexb Since my API searches aren't working, today's plan is to delete all of the mappings and start again.
14:15 ranguard talexb: cool - thanks :)
14:16 talexb ranguard, Still getting an error with 'bin/metacpan release /usr/share/mirrors/cpan/authors/id/' -- 'Dunno what /usr/share/mirrors/cpan/authors/id/ is'
14:17 ranguard oalders: ^^ HELP :)
14:17 talexb /usr/share/mirrors/ doesn't exist.
14:18 talexb I am able to get authors from the API OK, but modules and relaeses don't work. So I've been assuming my local DB is broken.
14:18 * talexb also turns on his Olaf-signal.
14:19 ranguard talexb: do you need to run the API locally - could you not just query live for this?
14:19 ranguard (Testing running it locally is cool, but not actually needed..? )
14:19 talexb I could query live -- I've just been conditioned not to hit production for development purposes.
14:20 talexb Of course, that makes lots of sense.
14:20 talexb I'm an engineer, not a purist. I just want to get the job done.
14:22 shmuel joined #metacpan
14:23 * talexb will be back in ten.
14:39 oalders talexb: i'd encourage you to pummel the live api for this
14:39 oalders if you're actually working on the api itself, having it running locally is a requirement
14:40 oalders but for the sitemaps, i'd honestly just hit the production machine.  that's what it's there for :)
14:40 oalders talexb: you've put your CPAN in /usr/share/mirrors ?
14:44 talexb oalders, Thanks, I will hit the productino API for my testing.
14:44 talexb No, I haven't put my CPAN (?) in /usr/shar/mirrors. Was there a step that I missed?
14:45 oalders well, i'm just wondering about that snippet you posted
14:45 oalders bin/metacpan release /usr/share/mirrors/cpan/authors/id/
14:45 oalders where did you get that from?
14:45 talexb It's from the bottom of https://github.com/CPAN-API/metacpan-developer
14:45 dipsy [ CPAN-API/metacpan-developer · GitHub ]
14:46 talexb I downloaded a full mirror ..
14:47 talexb .. so the only art I didn't do was the WorePan part, which gets a partial mirror (I guess).
14:47 talexb Er, WorePAN. And 'part', not 'art'.
14:53 oalders so, just substitute that with the path to your actual local CPAN
14:54 oalders but in reality, i know ranguard is itching for that site map to get done :)
14:55 oalders so the CPAN indexing can probably wait
14:56 talexb OK. I have been doing research over the last few days to figure out how sitemaps work ..
14:56 ranguard oalders & talexb++
14:56 talexb \o/
15:06 metacpan joined #metacpan
15:06 metacpan [metacpan-developer] talexb opened pull request #10: Update path for local author ID (master...docUpdate-2013-0814)  http://git.io/MgpHnQ
15:06 metacpan left #metacpan
15:06 dipsy [ Update path for local author ID by talexb · Pull Request #10 · CPAN-API/metacpan-developer · GitHub ]
15:45 mdk 'lo all
15:45 mdk Is moritz about?
15:46 mdk I need to speak with whatever passes as the MetaCPAN leads to discuss how we give Bytemark a bit more recognition for their very kind sponsorship
15:46 mdk Have cc'd Moritz into my discussions, if anyone else needs to be part of that just let me know so i can bring you up to speed
15:47 talexb oalders, This message might be for you ..
15:47 mdk oalders: that would be the first person on my list
15:48 mdk oalders: send me an email address and i will forward on the conversations thus far
15:48 oalders mdk: was going to ask to be included :)
15:48 oalders actually, could you just cc noc@metacpan.org?
15:48 mdk Aye.
15:48 mdk Will forward the first couple of emails to that right now
15:48 oalders that will cover everyone who needs to know.  but this would be nice to get taken care of
15:49 mdk Done
15:51 AdmWiggin joined #metacpan
15:51 oalders ok. that all looks good.
15:56 mdk oalders: ace, well if someone wants to respond to the email group we can move it forwards. Thanks :)
15:56 mdk Keep me in the loop of course
15:56 oalders mdk: of course.  email sent :)
16:03 ether neilb: please keep track of when mail to authors bounce - that's valuable information for the pause admins to know
16:04 ether I don't know if there's a formal way of tracking that data but it would be useful - e.g. "last known contact from this author"
16:04 ether it can speed up later adoption processes if we already know an author is missing
16:04 neilb yeah, I'm adding that to my DB, and have added it onto the fguture ideas list on the blog post
16:05 neilb I'll email ANDK and suggest maybe that could be a flag in the PAUSE db
16:06 dsteinbrunner neilb: I'm also getting my own list so if you get farther along with that hit me up
16:06 neilb dsteinbrunner: roger that
16:11 tianon joined #metacpan
16:14 dsteinbrunner neilb: did you notice a recent release that had no changes in it
16:15 neilb dsteinbrunner: huh?
16:15 dsteinbrunner I'm assuming it had something to do with your list
16:15 neilb oh, someone released something with no changes?
16:15 neilb Thanks to the magic of MetaCPAN, the adoption list now has the total # dependent dists, and scores based on this rather than just the direct dists.
16:15 neilb http://neilb.org/adoption/
16:15 dipsy [ CPAN Adoption Candidates v3 ]
16:17 talexb As a side note, I'm wondering what it means when bin/metacpan release declares that an archive is 'impolite'.
16:17 neilb Have you got an example?
16:18 neilb I'm guessing it's the same as an unauth release on search.cpan.org: someone uploaded without having PAUSE perms for one or more module in the dist
16:19 talexb neilb, 2013/08/14 12:17:52 E release: /home/metacpan/CPAN/authors/id/W/WI​/WILSONPM/OutlineNumber-1.00.tar.gz is being impolite
16:19 dsteinbrunner https://metacpan.org/changes/d​istribution/Compress-Raw-Bzip2
16:19 dipsy [ Changes - metacpan.org ]
16:19 talexb neilb, 2013/08/14 12:17:20 I release: release Carp-Datum-0.1.3 contains unauthorized modules: DB,Carp::Datum::Flags,Carp::Datum::Cfg,DB,Ca​rp::Datum::Strip,Carp::Datum::Assert,Carp::D​atum::MakeMaker,Parser,Carp::Datum::Parser
16:19 dsteinbrunner I think it just had to do with the version number staying up to date with other similar modules now that I look at it again
16:20 talexb I'm running the metacpan release script locally while I work on the site map script.
16:21 neilb oops, need to got pick up my daughter ...
16:24 neilb joined #metacpan
16:26 * talexb is lunching, then away this afternoon till about 4pm.
16:32 bowtie_ joined #metacpan
16:43 ether neilb: wow, are you including multiple levels of downstream deps now?
16:43 ether e.g. "if this module disappeared entirely, how many things on the cpan would be broken as a result"?
16:44 neilb ether: yeah, all indirect deps as well now
16:44 ether the number 20085 comes up a few times - is that the total number of dists in the index? :)
16:45 neilb Not quite: planning to look into those later
16:45 ether Test::Pod::Coverage bothers me... because it's used in *author* tests, no dists should be indicating test-requires on it at all.. :/
16:45 ether recently I put in a warning in TEst::Kwalitee that complains if you're running the test without $ENV{AUTHOR_TESTING} or $ENV{RELEASE_TESTING}
16:45 ether I'd love to do the same to Test::Pod(::Coverage) and see how much of cpan starts warning (it would be horrible, but it really is something that ought to be fixed..)
16:46 ether there was a whole bunch of things that became uninstallable under 5.18/5.19 recently when some underlying pod logic changed and Test::Pod started complaining about dists that previously passed tests
16:46 ether leading to huge amounts of "this should be an author-only test!!" tickets
16:47 neilb yeah, I noticed some of those
16:47 ether neilb: if you're looking for more ideas for reports (I say only partially tongue-in-cheek) :D -- "candidates for removal from cpan" -- dists that haven't been updated in years, don't pass tests under 5.18, have few downstream deps...
16:48 ether xdg was talking about curating various indexes based on various criteria, where the ephemeral "quality" could be one of the deciding factors
16:49 dsteinbrunner is there a does not pass 5.18 tests list?
16:49 neilb ether: I was thinking about just such a list as I ran home after my daughter on her scooter!
16:49 ether dsteinbrunner: you could obtain it by comparing hte main index against what cpxxxan has
16:50 ether which I think generates its lists based on "has at least one passing test on version X"
16:51 kablamo joined #metacpan
16:52 tianon joined #metacpan
16:57 genehack joined #metacpan
16:59 neilb ether: the total # dependents is an interesting number. I've got a wee dist with 8 dependent dists. But I just discovered the total # dependents is 3766. Gulp! That will make me a lot more careful about releases now (turns out Dist::Zilla is a 2nd order dependent!). This has prompted an idea which I'll submit as a MetaCPAN suggestion.
16:59 ether ++
16:59 dipsy Thanks!
17:01 neilb ticket: https://github.com/CPAN-AP​I/metacpan-web/issues/913
17:01 dipsy [ Include the total number of dependent dists · Issue #913 · CPAN-API/metacpan-web · GitHub ]
17:02 * ether pokes through some of the buglists at the top of the list, triages a few bugs
17:04 ether hmm, I'm seeing something else I would find really useful in the adoption list... if I can enter my PAUSE id in a textbox, and then have it tell me if I have comaint on any of the dists in the list... ;)
17:05 ether i.e. "hey you, YOU can fix some of these problems.. get on it, why don't you!!" :)
17:16 neilb Nice idea ... I'll have a think about how I can do that.
17:19 neilb Though if I see ETHER I'll probably just guess you are a co-maint and tell you to get on it ;->
17:20 dsteinbrunner I have had the idea of just looking at discs that have high bug count lists and trying to review the queue
17:20 dsteinbrunner there is often stuff that does not need to be in there
17:20 dsteinbrunner I have done that at time.. just making suggestions to close
17:22 ether or marking things as 'wishlist' that didn't have them
17:23 dsteinbrunner so a variation of the list for that kind of thing would be good ;-P  high deeps, high bug count, but other metrics not being factored in
17:23 dsteinbrunner something like that
17:24 dsteinbrunner the needs help but not adopted list or something
17:25 dsteinbrunner of course there are so many variations and each person has their own idea of what is important and how important it is
17:26 dsteinbrunner so you should just make it so there are a few default lists and then let people make customizations for their own lists ;-P
17:26 oalders talexb: is_impolite "this archive is going to unpack into the current directory rather than create its own"
17:26 oalders https://metacpan.org/module/Archive::Any
17:26 dipsy [ Archive::Any - Single interface to deal with file archives. - metacpan.org ]
17:26 oalders there's also is_naughty
17:26 dsteinbrunner speaking of, that module needs help
17:27 dsteinbrunner there is a patch that is needed
17:27 dsteinbrunner has been there forever
17:27 dsteinbrunner the author email fails
17:27 neilb dsteinbrunner: Gabor asked for that feature as well: the ability to 'make your own list' by deciding which bits to include and how to weight them
17:27 oalders yeah, looks like i asked for co-maint 2 years ago https://rt.cpan.org/Public​/Bug/Display.html?id=67738
17:27 dipsy [ Bug #67738 for Archive-Any: Add bzip2 support to Plugin::Tar ]
17:28 dsteinbrunner I tried another email for him.. no response
17:28 dsteinbrunner I think I found his twitter
17:28 dsteinbrunner but I don't use twitter
17:28 neilb oalders: it's on the list -- adopt it, adopt it!
17:28 oalders :)
17:29 neilb :-D
17:29 oalders dsteinbrunner: what is the author's twitter id?
17:29 dsteinbrunner https://twitter.com/hydo
17:29 dipsy [ Clint Moore (hydo) on Twitter ]
17:32 oalders request sent https://twitter.com/wundercoun​ter/status/367700140951613440
17:32 dipsy [ Twitter / wundercounter: @hydo Any chance I could get ... ]
17:32 dsteinbrunner neilb: I have actually created such a project.. a lists with scoring
17:33 dsteinbrunner there is the default scoring and the customer the talked about allowing anyone to login and create their own scoring and results
17:34 dsteinbrunner of course, I did the project 2 years ago and then they have yet to release the website with the single list
17:36 dsteinbrunner oalders: another route on that module would be to get schwern involved
17:36 dsteinbrunner he already has co maint status
17:37 oalders dsteinbrunner: good call.  i'll try that if i don't hear back
17:39 oalders should we add some TIOBE spam to metacpan?
17:40 oalders Mithaldu?
17:40 dipsy Mithaldu is your dist useful if you're already passing critic tests
17:40 Mithaldu well damn
17:40 Mithaldu oalders: yes please!
17:40 oalders heh, ok :)
17:40 Mithaldu you already have a perl.org link
17:40 Mithaldu probably cool to just modify it a bit :)
17:41 oalders ah, right.  good idea
17:50 dsteinbrunner ether: any thoughts on this: https://github.com/rhandom/perl-net-server/pull/1
17:50 dipsy [ populated repo, fixed bugs, added repo info, etc by dsteinbrunner · Pull Request #1 · rhandom/perl-net-server · GitHub ]
17:51 dsteinbrunner just noticed you bumped the severity here https://rt.cpan.org/Public​/Bug/Display.html?id=86707
17:51 dipsy [ Bug #86707 for Net-Server: [PATCH] SSL_test.t fails with IO::Socket::SSL 1.950 and above ]
18:13 dpetrov_ joined #metacpan
18:15 dpetrov__ joined #metacpan
18:15 dpetrov__ oalders: that's correct
18:15 dpetrov__ sorry for the late response
18:16 oalders dpetrov__: thanks.  ranguard  is handling getting the relevant domain names into one account
18:16 dpetrov__ :)
18:16 oalders ranguard: ^^^ mcpan.org domain name transfer
18:16 dpetrov__ ranguard: yep, ping me around when around
18:16 dpetrov__ I would transfer it to you
18:17 oalders dpetrov__: that's great. i'll be a help to have stuff more centralized
18:17 oalders TPF will manage it
18:19 dpetrov__ if only I could remember my godaddy login now
18:19 dpetrov__ :D
18:27 oalders :D
18:27 dpetrov__ okay had to reset the password
18:47 ranguard dpetrov__: https://github.com/tpf/docs/blob/master/domains.md :)
18:47 dipsy [ docs/domains.md at master · tpf/docs · GitHub ]
18:47 ranguard dpetrov__: you still ok hosting it for now - or should we look to moving that to metacpan's server :) ?
18:51 oalders it's essentially just a matter of hosting the DNS at this point
18:51 ranguard cool - then yea just follow instructions there (feel free to CC me - leo@cuckoo.org) - cheers
18:56 dpetrov__ :)
18:56 dpetrov__ ranguard: thanks
18:57 dpetrov__ yeah there is nothing to host
19:00 dpetrov__ is github down just for me
19:06 oalders looks fine from here
19:21 ether dsteinbrunner: for the META_MERGE changes, you should add the fields { 'meta-spec' => { version => 2 }, dynamic_config => 0, ...} and change resources=>{repository=>$url} to resources=>{repository=>{url=>$url}}
19:21 ether otherwise, ++ :)
19:22 dsteinbrunner well….
19:22 dsteinbrunner I was actually looking for ideas on getting things moving towards a release
19:23 dsteinbrunner but I sent an email
19:23 dsteinbrunner and just got a response
19:23 dsteinbrunner so things might be moving along
19:25 Mike-PerlRecruiter_ joined #metacpan
19:44 metacpan joined #metacpan
19:44 metacpan [metacpan-developer] ranguard closed pull request #10: Update path for local author ID (master...docUpdate-2013-0814)  http://git.io/MgpHnQ
19:44 metacpan left #metacpan
19:44 dipsy [ Update path for local author ID by talexb · Pull Request #10 · CPAN-API/metacpan-developer · GitHub ]
20:16 neilb joined #metacpan
20:28 talexb \o/ .. and I'm back.
20:49 oalders talexb: welcome!

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