Perl 6 - the future is here, just unevenly distributed

IRC log for #metacpan, 2014-07-02

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

All times shown according to UTC.

Time Nick Message
01:02 klapperl joined #metacpan
01:12 oiami joined #metacpan
01:12 grantm joined #metacpan
01:12 cooper joined #metacpan
01:12 alnewkirk joined #metacpan
01:12 FROGGS joined #metacpan
01:12 mr-foobar joined #metacpan
01:12 Khisanth joined #metacpan
01:12 berekuk joined #metacpan
01:12 rwstauner joined #metacpan
01:12 ribasushi joined #metacpan
01:12 lifeofguenter joined #metacpan
01:12 burnersk joined #metacpan
01:12 mtj joined #metacpan
01:12 tianon joined #metacpan
01:12 genehack joined #metacpan
01:12 Seveas joined #metacpan
01:12 priodev joined #metacpan
01:12 ether joined #metacpan
01:12 alh____ joined #metacpan
01:12 haarg joined #metacpan
01:12 lestrrat joined #metacpan
01:12 charsbar___ joined #metacpan
01:12 shibayu36 joined #metacpan
01:12 kentnl joined #metacpan
01:12 Ralesk joined #metacpan
01:12 mattp joined #metacpan
01:12 diegok joined #metacpan
01:56 FROGGS_ joined #metacpan
01:59 klapperl_ joined #metacpan
05:10 talina_ joined #metacpan
05:34 oiami joined #metacpan
06:25 neilb joined #metacpan
06:47 neilb joined #metacpan
06:47 FROGGS_ joined #metacpan
07:13 neilb joined #metacpan
07:43 oiami1 joined #metacpan
07:44 ingy joined #metacpan
08:19 neilb joined #metacpan
08:20 dpetrov_ joined #metacpan
08:36 DerAlex joined #metacpan
09:24 neilb Are other people feeling that MetaCPAN is slower recently? When typing a URL in, or hitting return in the search bar, there's a noticeable latency. I'm assuming this is down to the CDN.
09:26 * ranguard was just thinking it felt sluggish
09:26 oiami joined #metacpan
09:27 ranguard we've been a little more loaded than previously: http://munin.metacpan.org/metacpan.org/bm-n2.metacpan.org/load.html
09:27 dipsy [ Munin :: metacpan.org :: bm-n2.metacpan.org :: load ]
09:29 ranguard we've got some new hardware being setup in the next couple of weeks (though not sure on switch over time) but we'll have a LOT more umpf
09:31 rashi joined #metacpan
09:49 neilb Ooh, munin graphs!
09:51 neilb interesting -- what job kicked off around 06:30 this morning?
09:51 ranguard https://github.com/CPAN-API/metacpan-puppet <- everything is in there somewhere
09:51 dipsy [ CPAN-API/metacpan-puppet · GitHub ]
09:51 ranguard https://github.com/CPAN-API/metacpan-puppet/search?q=cron&amp;ref=cmdform searches is
09:51 dipsy [ Search Results · GitHub ]
11:01 heytitle_ hi guys
11:02 heytitle_ Can I do anything with this link? https://www.dropbox.com/s/dw9gqhacyxlgil6/Screenshot%202014-07-02%2017.59.51.png
11:02 dipsy [ Dropbox - Screenshot 2014-07-02 17.59.51.png ]
11:02 heytitle_ It leads to 404 page.
11:38 ranguard heytitle_: contact the author of the module (see the issues link on the left hand side of that screen shot) and let them know
11:39 ranguard oh, sorry it's a perl built in
11:40 ranguard oalders: ^^ ? https://metacpan.org/pod/File::Flock::Tiny
11:40 dipsy [ File::Flock::Tiny - yet another flock package - metacpan.org ]
12:42 oalders ranguard: so the author is linking to a built-in function L<flock>. is that valid Pod?
12:43 ranguard This syntax is also occasionally used for references to UNIX man pages, as in "L<crontab(5)>"
12:44 ranguard so maybe not ( reading perldoc perlpod )
12:45 ranguard also doesn't work on search.cpan.org (because you need the '-f' on perldoc for 'flock')
12:46 ranguard heytitle_: ok, so back to my original suggestion - contact the author :)
13:01 ilmari oalders: L<perlfunc/flock FILEHANDLE,OPERATION>
13:02 ilmari the thing after / has to match the content of a =headX or =item in the target document
13:02 oalders so, the author does need to fix this
13:03 oalders i hadn't seen that before
13:07 ilmari https://metacpan.org/pod/perlpod#Lname----a-hyperlink
13:07 dipsy [ perlpod - the Plain Old Documentation format - metacpan.org ]
14:29 FROGGS joined #metacpan
17:19 neilb joined #metacpan
17:32 rashi joined #metacpan
18:05 oalders anyone care to review the bootstrap v2 => v3 branch before we go live with it?
18:25 haarg have the contributors always only visible on the release page and not the module page?
19:46 oalders haarg: not that i remember
20:22 DerAlex joined #metacpan
20:57 mattp joined #metacpan
20:59 mattp joined #metacpan
21:09 ether huh, why is https://metacpan.org/release/MSCHWERN/base-2.11 showing up in schwern's list of releases, when there doesn't seem to be an indexable module in there?
21:09 dipsy [ base-2.11 - Establish an ISA relationship with base classes at compile time - metacpan.org ]
21:11 ether is the author page now showing all uploads, rather than just those that have an indexed module in them?
21:11 ether looks like it's also showing the last upload of a dist name made by that author, even when it's no longer the latest release (e.g. schwern's page also has ExtUtils-MakeMaker-6.64 listed)
21:12 ether it would be really really helpful to mark those in some way as "not latest"
21:12 * ether doesn't see mention of thise in /news
21:12 ether -e
21:12 dipsy -2.71828182845905
21:16 * ether files https://github.com/CPAN-API/metacpan-web/issues/1240
21:54 grantm joined #metacpan
22:19 ether another weird occurance of an unauthorized module: https://metacpan.org/source/INGY/Acme-1.1111111/lib/Acme.pm
22:19 dipsy [ lib/Acme.pm - metacpan.org ]
22:20 ether a second package declaration inside a module is a common way of monkeypatching some other module, so listing it as being in that dist is pretty weird
22:21 ether I wonder if metacpan should only list the first package found in each module (or use 'provides' as canonical, if present)
22:22 grantm left #metacpan
22:23 haarg afaik PAUSE will attempt to index UNIVERSAL for that module as well
22:23 haarg so it makes sense for metacpan to follow that
22:25 ether so ingy should get an indexing failure for that dist? I don't think that will happen in this case
22:25 ether ingy: can you let us know what you get from PAUSE? ^^
22:25 grantm joined #metacpan
22:26 haarg he would get an indexing failure on UNIVERSAL and a success on Acme
22:26 haarg indexing is per package, not per dist
22:27 trs ether: there are many, many valid examples of more than one package in a file.
22:27 trs where you want to indexing all of em
22:30 ether you generally don't, because it's impossible to 'use' that other package name directly
22:30 ether so there's not much point in it being indexed
22:30 ether but I agree it is done a lot :)
22:31 trs the value of indexing does not solely correspond to being able to 'use' a package.
22:32 trs indexing merely describes what is in a distribution and provides a way to claim namespaces.
22:33 ether I see the main point of a module being indexed that you can install it via a cpan client, but if it's not a module itself, you can't use it on its own anyway
22:34 ether plus, if ingy happened to have comaint on the UNIVERSAL dist, the fact that he's monkeypatching it there would mean the real UNIVERSAL just got replaced in the index - not so good!
22:35 ether so maybe PAUSE itself shouldn't be trying to index anything but the first package declaration :)
22:35 ether but backcompat; the end :)
22:39 trs we come to different conclusions. my conclusion is that if you're monkey-patching something you have comaint on, it's your own fault for screwing up and you should split the package line or better yet explicitly declare no_index.
22:39 ether oh I'd agree with that too :)
22:39 ether but your behaviour shouldn't be any different just because you *don't* have comaint
22:40 trs using the first package would be so arbitrary.  what if the monkey patch *was* the first package?
22:41 trs I agree it shouldn't be any different, except that you don't cause any problems by it.
22:41 trs (if you don't have comaint, that is)
22:42 ether yeah, it would have to be s/not first/not matching the filename it's in/
22:43 ether this is what we get for being so lax - we have to support all the crazy permutations
22:43 ether and it really sucks sometimes!
22:43 trs it's not clear what's sucking in this case. :)
22:44 trs what's the problem?
22:44 ether remember perigrin's mishap earlier today
22:44 ether that's another, because .pm files can be anywhere in a dist
22:44 ether if they were only under lib/, extra cruft does no harm.
22:45 trs that has nothing to do with multiple packages in a file...
22:45 trs it's not clear what's sucking *in this case*
22:49 ether that PAUSE attempts to index that module, which results in ingy getting a failure response from PAUSE, and also that metacpan displays "UNAUTHORIZED" against that module
22:52 trs why is that bad?
22:53 trs it tells ingy, "you should probably hide this from pause" or "hey, someone is using that namespace already"
22:53 trs while metacpan is telling you, the casual onlooker, that ingy is doing something funky with UNIVERSAL
22:54 trs if you're just ranting about the state of indexing, OK :)  but if you actually want something in metacpan to change, I'm not sure why.
22:55 ether it did seem odd to me that metacpan would list UNIVERSAL there.
22:56 trs right, but _why_?
22:56 ether I don't know how I can rephrase again.
22:57 ether there's not likely going to be any documentation there, so what is the point?
22:57 trs it's under Provides, not Documentation for a reason.
22:57 ether the code is always browsable, so the list of modules in the main release page is (IMHO) documentation-centric
22:57 trs the point isn't just to show pod.
22:57 trs the release page is already doc centric, but it also lists other stuff about the release.
22:59 trs i.e. documentation is first, then modules with doc, then provides (modules without doc), then examples, then other build files, tests, etc.
22:59 trs s/modules/packages/ really
23:00 ether yes exactly :) as modules vs. packages is really what I'm niggling
23:03 trs a module's just a file, though.
23:03 trs what I'm asking is why do you want to exclude from the release page packages which aren't the primary package of a file, given that documentation is already the most prominent?
23:04 trs what use is there in excluding them?  or alternatively, what problem do they cause by being included?
23:04 trs these are sincere questions. I don't understand why they're bothersome or surprising.
23:05 trs ("they're" meaning the packages, not my questions...)
23:16 neilb When did MetaCPAN start displaying things from BackPAN on the author pages? I can't see anything in /news about it. I'm assuming that's what it's doing
23:19 trs neilb: see scrollback and the gh issue :)
23:23 neilb ah, just found the github issue, so not intentional :-)  thanks
23:39 Khisanth joined #metacpan
23:54 neilb joined #metacpan

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