Perl 6 - the future is here, just unevenly distributed

IRC log for #pr-challenge, 2015-02-02

| Channels | #pr-challenge index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:13 gansh joined #pr-challenge
00:15 neilb ooh, just produced first graph to see if there’s a visible impact of the PRC …
00:20 Peter The suspense
00:31 neilb http://neilb.org/2015/02/02/monthly-cpan-pr-graph.html
00:32 Peter Nice
00:33 Peter I can certainly say I've got more involved with non-perl github projects too
00:43 manchicken joined #pr-challenge
01:20 frederico :)
01:20 frederico it seems that travis is build the module now
01:39 tinypig joined #pr-challenge
02:25 vairav joined #pr-challenge
02:52 jamessan joined #pr-challenge
02:53 oalders joined #pr-challenge
02:57 Su-Shee_ joined #pr-challenge
03:20 oalders joined #pr-challenge
03:26 ether neilb (asleep): would you be able to take manual dist name -> repo address additions, for considering contendors?
03:26 ether if you start a text file in the PRC repo I'll happy send PRs to add entries
03:26 ether then the dist can be handed out to someone, and their first PR can be to add that repo metadata :)
03:27 ether I just discovered that I have a few (adopted) dists myself that are missing that data
03:27 ether and it would be a *real shame* for them to not get assigned :D
03:29 oalders joined #pr-challenge
03:31 ether also - github itself might be interested in hearing about the PR challenge
03:31 ether could be even more good publicity
03:39 boreas joined #pr-challenge
03:39 raj4perl joined #pr-challenge
05:26 BRAD_ joined #pr-challenge
05:30 raj4perl joined #pr-challenge
05:36 Peter joined #pr-challenge
05:53 oalders joined #pr-challenge
07:29 reneeb joined #pr-challenge
07:49 boreas joined #pr-challenge
08:01 neilb joined #pr-challenge
08:01 reneeb joined #pr-challenge
08:15 Mike__B neilb: graph is beautiful!
08:16 neilb yay us!
08:53 raj4perl joined #pr-challenge
08:54 neilb joined #pr-challenge
09:05 Relequestual joined #pr-challenge
09:07 neilb We’re now 47.1% complete for January. Hoping we hit 50%… (that’s 146/310)
09:09 neilb 208 people have an assignment for February
09:13 Relequestual I should really do my assignment for Jan...
09:16 Su-Shee neilb: _very_ nice!
09:20 Su-Shee neilb: lovely graph :)
09:21 Su-Shee neilb: I'm slowly thinking 2016 as "writing perl applications challenge" might be have impact too :)
09:31 neilb joined #pr-challenge
09:33 reneeb Su-Shee++
09:36 veryrusty joined #pr-challenge
09:37 Su-Shee I'm noticing I'm still a bit fuzzy about the entire toolchain involved. did anybody ever write a modern tutorial "from module via github onto cpan, dipping into travis ci"?
09:37 Su-Shee I'm finding bits and pieces, but not the entire workflow explained once.
09:50 fuzzix That would probably amount to "Put these lines into dist.ini"
09:59 BooK neilb: since metacpan has info about PAUSE id -> github account, I should probably be able to produce some /once-a/*/PR boards :-)
10:00 BooK restricting the PR targets to cpan dists, I guess
10:00 neilb yeah, similar to the code I wrote last night to get the stats for numbers of PRs per month
10:01 neilb combined with the MetaCPAN query that was in my perladvent, for about 15th Dec I think
10:01 BooK given that cpan.io is made of code forked off yours, I'll ask for links to the code
10:01 marderh joined #pr-challenge
10:13 neilb Su-Shee: I think a “write an application” would sound a lot more like work to people, so might have fewer signups :-)
10:14 Su-Shee neilb: lemme sit on the idea a bit longer and warm it.. :)
10:14 Su-Shee neilb: I'm assuming two things: people LOVE to help. people would like to know what to do exactly.
10:15 Su-Shee neilb: I know a lot of open source developers which are basically just waiting that somebody throws work at them.
10:15 neilb yup. And things that sound small are good too
10:15 neilb break it down into 12 standard tasks, one per month?
10:16 Su-Shee neilb: so imagine you'd have something like 5 applications to write over the course of a year and 5 people managing it mostly and then people can sign up for whatever app they like.
10:16 Su-Shee neilb: for example.
10:16 BooK month 1: come up with a good idea
10:16 BooK month 2: agonize over the name
10:16 Su-Shee exactly NOT how to do it.
10:16 neilb 3: design the icon
10:16 BooK worked fine for me
10:16 neilb 4: pick the font
10:16 Su-Shee 5: make prototype
10:16 BooK except for s/month/year/ obviously (in my case)
10:16 Su-Shee 6: pick up fight over libraries and tools
10:17 Su-Shee 7: abandon project
10:17 neilb 8: check if domain name is available
10:17 Su-Shee 9: squat on domain. just in case
10:18 Su-Shee neilb: jokes aside, I think most missing apps in perl are actually quite doable in a year when somebody prevents the typical grand idea problem that I see a lot in perl..
10:18 BooK https://metacpan.org/pod/wallflower#ACKNOWLEDGEMENTS  https://metacpan.org/changes/distribution/App-Wallflower # 2 years from idea to fix release
10:18 neilb Su-Shee: not just in perl
10:18 neilb but yes. Maybe team challenge!
10:19 Su-Shee BooK: awesome. even if I read the POD I don't know what it actually DOES?
10:19 BooK Su-Shee: I'll take your PR!
10:20 BooK did you try https://metacpan.org/pod/wallflower#DESCRIPTION ?
10:20 Su-Shee BooK: move the first sentence of "DESCRIPTION" up ABOVE "NAME"
10:20 Su-Shee BooK: yes. I have to read down a lot until somebody finally tells me what it does :)
10:20 BooK and lose the joke?
10:20 Su-Shee BooK: yes.
10:20 BooK INCONCEIVABLE!
10:20 Su-Shee BooK: I don't get it anyways, but I absolutely _hate_ "cute" docs and "funny" manuals.
10:21 Su-Shee I want to read and parse them as fast and precisely as possible.
10:21 BooK this project was basically born from the joke
10:21 Su-Shee that's lovely. put it at the bottom under "FOLKLORE".
10:21 Su-Shee BooK: no, sarcasm aside, it really takes me an entire page until I can even guess what wallflower does.
10:23 BooK the tutorial has all the information in the right order, I think
10:24 Su-Shee neilb: also, the app project is of course purely self-serving to motivate me to finally finish two I have in the making
10:24 BooK the problem with this distro is really: which one is the entry point?
10:24 BooK Su-Shee: how dare you scratch you own itch?
10:24 Su-Shee BooK: are you seriously trying to convince me that a joke on top of documentation is more important than simply stating what the application really does? ;)
10:26 BooK I'm not trying to convice anyone, just explaining why things are like they are
10:27 Su-Shee BooK: then leave them as they are, it's your module.
10:28 BooK as I said, PR welcome!
10:29 Su-Shee na, too much work. still don't really get what I would use the module for, would have to try it first, too many open questions I have to find out first.
10:29 BooK and if you want to get the joke, it's explained somewhere in http://blogs.perl.org/users/book/2012/09/sorry-i-cant-dance-im-holding-on-to-my-friends-purse.html
10:29 Su-Shee and I have a module to do PRs for. :)
10:30 BooK neilb: if you really plan to assigne Archive::Tar to me in March, you *have* to assign App::Wallflower to Su-Shee that month :-)
10:30 Su-Shee BooK: no, I don't want to get the joke, really :)
10:32 neilb BooK: it’s all random™!
10:33 BooK neilb: sure thing
10:33 rnewsham joined #pr-challenge
10:43 Mike__B BooK: sorry for asking the stupid question but you work for 'that big hotel site' right?
10:44 Mike__B BooK: like the one with the name which starts with your IRC nick
10:45 Mike__B BooK: I attended sawyers talk on FOSDEM saturday and he mentioned booking.com uses Dancer; so do you happen to use wallflower for the extreme caching, or is dancer not used for anything huge like that?
10:48 dolmen neilb: I haven't received my assignment for February...
10:50 Mike__B KPN also uses Dancer or at least I made some APIs based on dancer but it's nothing very big so far.
10:50 dolmen Is it because I 'haven't told you so by e-mail (but instead just via Github in my first PR, via IRC, and 5 times via Twitter)?
10:51 BooK Mike__B: we use dancer, but not for the main site as far as I know
10:51 BooK dolmen: that's actually because he hates your guts
10:52 BooK my own current project is using dancer2! but it's probably as far as one can be from the main site :-)
10:56 Mike__B ah ok, I see!
10:57 Mike__B so wallflower was not meant at least for use within booking.com directly?
10:57 BooK nope
10:57 neilb dolmen: have you emailed me saying “I’m done” for January?
10:57 BooK read the blog post above and Wallflower::Tutorial for the historical details
10:58 neilb if not, are you done? My DB says not officially
10:58 BooK actually, the tutorial does not have any historical info
10:59 BooK Mike__B: wallflower is one of the building blocks of my jekyll replacement, whenever I actually start working on that :-)
10:59 Mike__B nice
11:00 BooK deven_: just like Git::Repository is a building block for Git::CPAN::Hook
11:00 BooK when I die, I'll have baked a few bricks, but not even started building that wall I dream about ;-)
11:02 neilb dolmen: coming your way!
11:18 dolmen neilb: if you are still relying on e-mail to verify achievements, it means continuing my script that uses the Github API is worthful...
11:19 neilb doing a PR doesn’t mean someone’s done. They’re done when they say they’re done…
11:19 dolmen neilb: so what is the criteria besides 'at least one PR'?
11:20 neilb at least one PR and the punter says they’re done
11:21 * dolmen still doesn't understand why the second point. But the rules are the rules.
11:32 fuzzix You have to use the string "I'm done"?
11:35 fuzzix What about "I’m done"? :)
11:37 neilb fuzzix: :-p
11:37 neilb dolmen: I don’t ever want to give the same dist to two people in parallel, hence the “I’m done”.
11:37 neilb I don’t plan to give out a dist more than oince anyway, but we might need to later in the year
11:48 Mike__B neilb: can you upload the February assignments?
11:49 neilb nope. They get published after 5 days, to give people time to request a different dist before it becomes public knowledge
11:49 Mike__B ok, I see
11:49 neilb Allowd for “I don’t want that bozo’s dist”, or “They’re a competitor” (which has actually happened more than once)
11:50 Mike__B I might also have one or two people on my list which I would not look forward to making PRs.
11:51 neilb (yeah, people might say “oh, not BOOK, he’ll just try and get me to use Wallflower”, or similar :-)
11:52 neilb or “not NEILB, he’s so tardy merging PRs” :-(
12:18 tinypig joined #pr-challenge
12:35 kid51 joined #pr-challenge
12:36 cakirke joined #pr-challenge
13:35 shorty_mu joined #pr-challenge
13:36 itcharlie joined #pr-challenge
13:36 itcharlie left #pr-challenge
13:51 tinypig joined #pr-challenge
13:58 manchicken_ joined #pr-challenge
14:02 StylusEater joined #pr-challenge
14:04 vroom joined #pr-challenge
14:10 vroom joined #pr-challenge
14:16 deven BooK: What's the big issue for Git::CPAN::Hook?
14:21 tinypig joined #pr-challenge
14:30 gryphon joined #pr-challenge
14:35 neilb joined #pr-challenge
15:18 kanashiro joined #pr-challenge
15:24 manchicken_ joined #pr-challenge
15:32 jrhunt neilb, are all the emails sent out for Feb?
15:33 neilb jrhunt: everyone who claimed “done” for January got an assignment for February
15:33 BooK deven: so the idea being Git::CPAN::Hook is to save the result of installing a module in its own little git tree, each pointed to by an orphan commit, tagged with the distribution name
15:33 PerlJam neilb: How hard would it be to add a link to the RT queue to the email?
15:34 BooK so we can easily install/uninstall/resinstall various dists
15:34 neilb PerlJam: not at all
15:34 BooK my idea behind this is to bisect dependencies, for example
15:34 * neilb adds to his todo list
15:34 PerlJam neilb: the first thing I do it look at github issues and the RT queue, then I look at the module itself :)
15:35 BooK you install all versions of Foo-Bar, create a branch with each install as successive commits, and bisest with your test suite to find the oldest version with which your test suite still passes
15:35 jrhunt neilb, I claimed done on Jan 22 for Crypt::OpenSSL::X509 (via the cpan-prc.org mailing list), but have yet to receive a Feb assignment
15:35 neilb github username pkease?
15:35 jrhunt filefrog
15:35 neilb ah, to the mailing list and not me?
15:35 BooK this would be all fine, except when you install v1.01 on top of v1.00 and some files have not changed, git does not include them in the commit
15:36 BooK neilb: this is soon to become bigger than you :-)
15:36 jrhunt yes, to the list.  was I supposed to email you directly?
15:36 jrhunt (apologies, I'll try to do better in Feb :)[1~[3~
15:37 neilb Book, well if everyone actually read the emails from start to finish! :-)
15:37 neilb jrhunt: recorded. You take us to 47.7% for Jan
15:37 BooK neilb: you're not talking about the first email for february that some of got, are you?
15:37 BooK deven: so the solution I have in mind for that is simply to do the last step of the install on top of an *empty* directory, so all files will be new to git, and the orphan commit has all needed files
15:38 neilb BooK: :-p
15:38 BooK then we have finally accomplished the first step, which is to isolate all the installed files for a given dist
15:38 PerlJam neilb: how long before whatever software you use to manage the PRCs becomes something that can participate in the PRC?  ;)
15:39 BooK it turns out this is still a little bit optimistic, because some module installations might actually depend on what's already installed or not. after all, installation is a program, which can do different things depending on the environment
15:39 BooK but I decided to ignore this for the moment
15:42 BooK also, Git::CPAN::Hook is currently tied to CPAN.pm, but this should work with any cpan client where one can hook itself at some of the critical steps
15:43 deven BooK: Bisecting the modules to find the oldest working version is a cool idea.
15:43 BooK yes
15:44 BooK I actually had it in 2009, I'm just so damn slow
15:44 deven BooK: Because I expect that many modules claim to "require" whatever version of a module it happened to be developed against.
15:44 BooK indeed
15:45 BooK it would also speed up the creation of specific dev environments
15:45 jb360 neilb: Hi Neil.  I emailed the owner of my Feb. cpanpr assignment.  If I don't hear anything from him about working on his package do we ping you for another assignment?  Just curious what the protocol is?
15:45 BooK basically a description could be a simple text file with the list of dists you want, and you could easily swap between envs
15:45 punter joined #pr-challenge
15:46 BooK install once, use whenever
15:46 neilb depends. Plenty of people have gone ahead and done a PR anyway. But you might choose to limit the scope for a first one, so you don’t sink too much effort in that might not result in a merge.
15:46 deven BooK: Here's where it gets particularly tricky -- you can't really just bisect a single module dependency because there could be dependencies amongst the dependent module that only certain combinations work.
15:46 neilb Submitting the PR is still useful, as if someone adopts the module, they can benefit from your work.
15:47 neilb It’s a good idea to log the fact in the RT queue for the dist.
15:47 neilb PerlJam: hopefully after the QA hackathon, but no promises :-)
15:47 BooK deven: the trick is one could programmatically create branches with anything installed
15:47 raj4perl joined #pr-challenge
15:47 PerlJam jb360: or, if the module author has moved on to other things ... maybe you could adopt it :)
15:47 deven BooK: I'm not sure what to think of the orphan commit idea.  Why orphaned?
15:47 jb360 PerlJam: Sure, but he's active.
15:48 BooK deven: just so that it's not in the way
15:48 jkg jb360: to be fair, if you've not had a response on the 2nd of the month I wouldn't worry just yet :-)
15:48 deven jb360: Which hopefully translates into responsive. :)
15:48 dolmen BooK: make ExtUtils::Install pluggable...
15:48 PerlJam what jkg said :)
15:48 jb360 Recent work on other modules he owns.  I'm just trying to be prepared if he does not respond in a week.
15:48 BooK dolmen: is that used by all clients?
15:48 jkg yeah, a week (especially if there's other activity in that week) is probably indicative of ... well, something
15:49 jkg but equally, people take holidays :-)
15:49 BooK as ribasushi noted, doesn't need to be commits, tagged trees work fine too
15:49 dolmen BooK: you mean all installers?
15:49 deven BooK: Seems like it would clutter the repository. :/
15:49 BooK who cares?
15:50 BooK dolmen: yes
15:50 BooK the point is that this specific repository should be throwaway, anyway
15:53 BooK deven: right now, the blocker (and the test for it) is: install a dist, record the commit/tree, install another version where at least one install file is unchanged on top of it, record the commit again, and make sure the tree contains the unchanged file
15:53 BooK my current use of diff-tree doesn't detect changed files
15:54 deven BooK: If these are orphan commits, what are the parent commits?
15:54 BooK mmmm, actually, I think I've found the trick
15:54 BooK deven: they are just building blocks
15:54 dolmen deven: undef
15:54 deven BooK: Seems like doing a hard reset to the same parent and making a new commit ought to include all files.
15:54 BooK they contains all the files installing dist Foo-Bar adds to your system
15:54 manchicken_ joined #pr-challenge
15:54 BooK deven: exactly, that's a problem. I want only the files for this install
15:54 BooK so anyway, I found the trick
15:55 deven BooK: git won't store two copies of the same object, but the shared object ought to still be referenced from both commits.
15:55 BooK I know all that
15:55 deven BooK: From what you described, I don't follow why it isn't already working then.
15:55 dolmen deven: BooK is the author of Git::Repository
15:56 deven dolmen: I wasn't questioning his knowledge, just trying to confirm my understanding.
15:56 BooK because if Foo-1.00 contains Bar.pm (hash=1234) and  Foo-1.01 contains Bar.pm (hash=1234, the same), if I install Foo-1.01 on top of a cpan directory with Foo-1.00 already install, then it wont see that Bar.pm was overwritten by the same content
15:56 deven BooK: It seems it should already work now -- what am I missing?
15:56 BooK so the diff won't list it
15:57 BooK it's an issue with the way I'm doing things
15:57 BooK turns out the trick is horribly simple
15:57 deven BooK: Why not wipe the old copy completely first?
15:57 BooK well, installation will require some dependencies to be there
15:57 deven BooK: just always install into an empty dir?
15:58 BooK dependencies!
15:58 BooK you can never have an empty dir
15:58 BooK actually, the trick it to lie to the installers
15:59 BooK pick a temporary directory $tmp, and do $ENV{PERL_MB_OPT}="--install_base $tmp"; $ENV{PERL_MM_OPT}="INSTALL_BASE=$tmp";
15:59 BooK and they'll install this one in a perfectly clean environement
15:59 BooK and I can just add .  and commit all
16:00 BooK instead of playing tricks with the output of diff-tree
16:00 dolmen BooK: local::lib
16:00 BooK dolmen: exactly!
16:01 BooK and when in a chain of install, you can always cherry-pick this commit you just did back in the actual cpan repo
16:01 BooK probably needs to play tricks with GIT_WORKTREE too, but that's fine
16:02 dolmen BooK: see ensure_dir_structure_for, setup_env_hash_for methods of local::lib
16:02 BooK dolmen: aren't those two env variable enough?
16:04 dolmen BooK: PATH and PERL5LIB must be adapted too. So to do things right, trust local::lib.
16:05 BooK right. but I can't simply "use local::lib"
16:05 dolmen BooK: use local::lib ();
16:05 BooK dolmen: you know I don't need that level of hand-holding...
16:06 dolmen BooK: you know I know you know I know. So why can't you "simply" use local::lib?
16:07 * Su-Shee counts the knows.
16:07 BooK because I need to "use local::lib ()" :-)
16:07 BooK I also need to be able to make that temporary
16:08 BooK local::lib->setup_env_hash_for($path) seems to be able to deactivate a path
16:08 sue joined #pr-challenge
16:08 BooK local %ENV = %ENV would also do the trick
16:10 BooK so basically, run the install in a temporary local lib, record the commit, and then cherry-pick it back in your normal cpan tree, and move on to the next distro to install
16:10 dolmen BooK: build_environment_vars_for
16:11 BooK dolmen: I also have an editor open on local/lib.pm :-)
16:11 dolmen BooK: start with local::lib->new
16:13 BooK the version I have here does not seem to have a sub new, not a parent module
16:13 dolmen BooK: if you don't want to rely on an installed local::lib, just fatpack it
16:13 BooK yeah, I was wondering about that. about the minimum amount of modules that should be required
16:14 dolmen BooK: the _for methods are shortuts. If you have to call multiple ones, it's better to build an object and call the lower level API
16:15 BooK dolmen: I think there are no objects, just the local::lib $class
16:16 BooK but not necessarily: modules used to run the git-cpan tool are not necessarily in the way when playing around with the worktree
16:16 BooK anyway, we've found a way to get past the roadblock I had, so all is fine
16:17 BooK will look at the next steps when ready to take them
16:17 * BooK &
16:28 BooK dolmen: what you said about installers made me think a little more
16:29 BooK there probably something I can do to hook around the install step, no matter which client is doing the install
16:31 BooK I just don't know what yet :-)
16:33 BooK I really just want to hook around 'make install' or 'build install'
16:36 ptc joined #pr-challenge
16:49 shorty_mu1 joined #pr-challenge
17:03 BooK for 'make install' I can probably muck around with $PATH (on unix at least)
17:03 rvandam joined #pr-challenge
17:15 ptc silly question time: is a META.yml required in a git repository?
17:15 ptc or can it simply be generated via ExtUtils::MM?
17:16 ptc in other words, should it be removed from the repo since it's generated anyway?
17:16 ribasushi yes, and gitignored
17:16 BooK deven: actually, the manual for Git::CPAN::Hook has lots of explanations
17:17 ptc ribasushi: do you mean, yes: required, or yes: delete?
17:17 deven BooK: I haven't looked through it yet. :)
17:17 ribasushi <ptc> or can it simply be generated via ExtUtils::MM? <--- not only can, but should be generated in virtually all cases
17:17 BooK some people do keep a copy of the generated file, though
17:17 ribasushi <ptc> in other words, should it be removed from the repo since it's generated anyway? <--- yes, and gitignored
17:18 ptc ribasushi: cool!  thanks for the explanation
17:18 ptc ribasushi: I'd asked too many questions and wasn't so sure which one you were answering :-)
17:18 ribasushi ptc: note that as BooK said, opinions differ, so checking with the author would be best
17:18 ptc ribasushi: will do, thanks!
17:22 thrig this month's low hanging fruit includes a README that says perl Makefile.PL for a Module::Build module
17:22 Su-Shee thrig: which module do you have?
17:23 thrig String::Template
17:24 Su-Shee I have something Google::Something OpenID something.
17:24 Su-Shee which I think has just a failing test on a weird dependency.
17:45 Peter joined #pr-challenge
17:48 shorty_mu joined #pr-challenge
17:56 vairav joined #pr-challenge
18:10 gansh joined #pr-challenge
18:11 reneeb joined #pr-challenge
18:11 nawglan joined #pr-challenge
18:11 nawglan did anyone get a YAML distro this month?
18:14 athos joined #pr-challenge
18:21 neilb joined #pr-challenge
18:47 quicoju joined #pr-challenge
18:51 quicoju Hi!, I'm running test cover on the module I'm working on, and I'm getting 50% branch coverage in lines using the conditional ternary operator.
18:51 quicoju i.e: $BACKEND = $opt{backend} eq 'AC' ? 'GD' : $opt{backend};
18:52 quicoju How can I get around this ?
19:00 shorty_mu joined #pr-challenge
19:20 veryrusty joined #pr-challenge
19:34 veryrusty joined #pr-challenge
20:06 mvuets joined #pr-challenge
20:26 choroba joined #pr-challenge
20:28 Peter joined #pr-challenge
20:30 neilb_ joined #pr-challenge
20:41 neilb_ joined #pr-challenge
20:50 balban quit
20:53 veryrusty joined #pr-challenge
20:56 Peter joined #pr-challenge
21:40 athos joined #pr-challenge
21:47 choroba I got https://metacpan.org/pod/Dist::Zilla::MintingProfile::SYP for February
21:47 choroba Do I understand correctly that it's basically a personal boilerplate?
21:48 ambs choroba: yeah, in that case I think a new module would be better. neilb?
21:49 neilb yeah, that’s not a good choice. I skip various namespaces, that should be one of them.
21:49 neilb choroba: remind me please, what’s your github username?
21:49 choroba choroba :)
21:49 neilb :-)
21:50 ambs (coherent people)++
21:51 choroba The documentation might include a link to https://metacpan.org/pod/Dist::Zilla::Role::MintingProfile, perhaps, because a random CPAN wanderer might wonder
21:56 thrig just one wafer thin minting profile...
21:57 pink_mist perhaps add +1 to the score of modules with more than 300 LoC? :P
21:59 pink_mist or subtract from modules with less than 50? :P
22:02 * choroba is checking...
22:02 choroba ok, 50 sounds like a good threshold
22:03 thrig or first look at a histogram of LoC?
22:04 pink_mist doing research instead of throwing a random number out there? whatever for? :P
22:05 neilb Yeah, I’ve wondered about some kind of size measure.
22:05 neilb tiny not good, but small and bijou +1, very huge maybe -1
22:05 neilb I also thought people could flag themselves as “ok to take monsters”
22:06 neilb and if they don’t have the flag, then monster dists would either get a -1 or -2 for that user, or I’d just not consider for that user.
22:08 neilb what other namespaces should we skip, like /^Dist-Zilla-MintingProfile/
22:08 pink_mist /^Task/
22:11 neilb choroba: new assignment on the way
22:11 neilb I notice that it needs the abstract fixing, as a minimum :-)
22:17 BooK nice to provide some hints about things to do in the email, btw
22:19 neilb I’m hoping to build up new ones each month, so we can gradually identify things that we all agree are worth doing.
22:20 PerlJam pink_mist: /^Task/  or /^Task::BeLike/  ?
22:21 pink_mist PerlJam: well all of the Task namespace seems like the wrong thing to focus on in my opinion
22:22 neilb I think definitely Task-Belike-* should be skipped, and possibly Task-*
22:23 pink_mist also, if it still even exists? Bundle::
22:23 BooK I have some Task modules that non trivial
22:25 veryrusty joined #pr-challenge
22:28 choroba got it
22:28 choroba thanks
22:29 choroba seems much more suitable
22:30 athos joined #pr-challenge
22:47 rvandam joined #pr-challenge
22:57 athos joined #pr-challenge
22:58 vroom joined #pr-challenge
23:01 tinypig joined #pr-challenge
23:34 rjbs joined #pr-challenge
23:34 athos joined #pr-challenge
23:38 punter joined #pr-challenge
23:49 boreas joined #pr-challenge
23:53 Peter joined #pr-challenge
23:56 vairav joined #pr-challenge
23:58 deven joined #pr-challenge

| Channels | #pr-challenge index | Today | | Search | Google Search | Plain-Text | summary