Camelia, the Perl 6 bug

IRC log for #metacpan, 2013-07-10

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

All times shown according to UTC.

Time Nick Message
00:18 dsteinbrunner joined #metacpan
00:19 cooper joined #metacpan
00:41 dsteinbrunner_ joined #metacpan
01:02 grantm joined #metacpan
01:12 shmuel joined #metacpan
02:02 Topic for #metacpan is now Have you installed your MetaCPAN VM? https://github.com/CPAN-API/metacpan-developer | Chat logs available at http://irclog.perlgeek.de/metacpan/
03:28 shmuel joined #metacpan
03:55 preflex_ joined #metacpan
04:29 metacpan joined #metacpan
04:29 metacpan [cpan-api01] rwstauner pushed 2 new commits to master: http://git.io/SzRlig
04:29 metacpan cpan-api/master 00b0019 Randy Stauner: Test basic pause login email
04:29 metacpan cpan-api/master 80136e3 Randy Stauner: Use built-in Email::Sender method for testing emails
04:29 metacpan left #metacpan
04:29 dipsy [ Comparing 987be26e0862...80136e3a2ebe · CPAN-API/cpan-api · GitHub ]
04:37 cooper joined #metacpan
04:49 shmuel joined #metacpan
05:07 jayallen joined #metacpan
06:04 dpetrov_ joined #metacpan
06:36 kentnl joined #metacpan
08:05 shmuel joined #metacpan
08:48 thaljef joined #metacpan
08:52 alnewkirk joined #metacpan
09:20 bowtie joined #metacpan
09:23 dolmen joined #metacpan
09:34 bowtie joined #metacpan
11:21 shmuel joined #metacpan
14:12 sawyer_ joined #metacpan
16:05 mst am I right in thinking that barring reindexes due to bugs, the metacpan api data per dist can be considered immutable?
16:05 mst also, if we -do- have to reindex due to a bug, is there anything that'll let me tell that it's changed?
16:57 jayallen joined #metacpan
17:36 thaljef joined #metacpan
18:41 oalders mst: it's pretty stable, but new fields can get added here and there
18:43 oalders i suppose we could add a modification time for each document
18:43 mst or a generation or something
18:43 oalders generation?
18:44 mst er, version number
18:44 oalders right
18:45 mst basically I'm trying to work out how to safely cache
18:45 mst well, in my case, memoize, since the cache will initially be unbounded
18:51 bowtie_ joined #metacpan
18:53 mst I guess for the moment I can do 'cache forever and refetch occasionally'
18:55 oalders in practice, the documents don't change all that much, so that's probably not a bad way to go for now
18:57 mst and knowing my luck, if you need to re-index anything that matters to me it'll be because I've just blown one of my feet off
18:58 trs what about serving an http etag that's a sha1 or something of the json representation of the doc?
18:58 trs does ES have such an internal identifier already that we can repurpose?
18:58 haarg doesn't the status change after PAUSE indexes the dist?
19:22 oalders the etag might work
19:57 ranguard etag would be good
19:58 mst is there an indexer log?
19:58 mst being able to read a log for changes would probably be sufficient for me
20:03 oalders mst: sadly, no log
20:05 mst that seems like it should be relatively easy to add
20:05 mst is it just a question of "nobody got around to it" ?
20:05 oalders and "nobody thought of it"
20:12 theory joined #metacpan
21:20 neilb joined #metacpan
21:47 gshank joined #metacpan
22:14 sivoais joined #metacpan
22:16 rwstauner ES documents have a version number
22:17 rwstauner and i think i heard that the id is predictable so an update should increase the version i think
22:18 rwstauner looks like the ES _version isn't included in the output of /release/Blah though
22:19 rwstauner don't mind me, _version doesn't even show up in search results
22:19 rwstauner i wonder if it's disabled in the mapping or something
22:20 rwstauner but if that's the case it's not obvious (from the output of mapping, anyway)
22:30 trs rwstauner: you're probably thinking of this: http://api.metacpan.org/release/_search?​q=Moose&fields=name&version=true
22:31 neilb joined #metacpan
22:31 rwstauner oh, look at that
22:31 trs there's a special _timestamp field which can be enabled in the mapping and will be autoupdated to the processing date. I'm not sure that it will be updated when the doc is updated though.
22:31 rwstauner didn't realize it didn't output by default
22:31 rwstauner or could be configured
22:32 rwstauner hmm
22:59 metacpan joined #metacpan
22:59 metacpan [cpan-api01] rwstauner pushed 2 new commits to master: http://git.io/eLofoA
22:59 metacpan cpan-api/master 9662dee Randy Stauner: Test pause email for nonexistent pause id
22:59 metacpan cpan-api/master da79c3d Randy Stauner: Test pause login email for a non-ascii name
22:59 metacpan left #metacpan
22:59 dipsy [ Comparing 80136e3a2ebe...da79c3d05cd7 · CPAN-API/cpan-api · GitHub ]
23:44 metacpan joined #metacpan
23:44 metacpan [cpan-api01] rwstauner created travis (+2 new commits): http://git.io/Si-sSw
23:44 metacpan cpan-api/travis ef61042 Randy Stauner: Fix env var after re-reading travis docs
23:44 metacpan cpan-api/travis a9e2d93 Randy Stauner: Add perl-pie so travis tests use standard ES port
23:44 metacpan left #metacpan
23:44 dipsy [ Comparing ef6104248365^...a9e2d93432dc · CPAN-API/cpan-api · GitHub ]

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