Perl 6 - the future is here, just unevenly distributed

IRC log for #metacpan, 2014-09-05

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

All times shown according to UTC.

Time Nick Message
00:07 vroom joined #metacpan
00:19 oiami joined #metacpan
01:14 klapperl joined #metacpan
01:37 vroom joined #metacpan
01:55 vroom joined #metacpan
01:58 FROGGS_ joined #metacpan
02:32 vroom joined #metacpan
06:06 priodev joined #metacpan
06:06 sivoais_ joined #metacpan
06:34 priodev joined #metacpan
06:40 talina_ joined #metacpan
06:54 dp_ joined #metacpan
09:55 rashi joined #metacpan
10:13 talina_ joined #metacpan
11:27 oiami joined #metacpan
11:30 vroom joined #metacpan
12:07 vroom joined #metacpan
12:27 vroom joined #metacpan
13:10 vroom joined #metacpan
14:19 vroom joined #metacpan
14:29 hipyhop joined #metacpan
14:59 vroom1 joined #metacpan
15:44 FROGGS joined #metacpan
16:27 metacpan joined #metacpan
16:27 metacpan [metacpan-web] oalders pushed 2 new commits to master: http://git.io/IlWR-w
16:27 metacpan metacpan-web/master 29a3e09 Philippe Bruhat (BooK): Change gittip to Gratipay...
16:27 metacpan metacpan-web/master f267906 Olaf Alders: Merge pull request #1325 from book/master...
16:27 metacpan left #metacpan
16:27 dipsy [ Comparing b5aad7e8561c...f267906cc123 · CPAN-API/metacpan-web · GitHub ]
17:19 rashi joined #metacpan
18:27 stelim joined #metacpan
18:39 miyagawa joined #metacpan
18:40 miyagawa how does metacpan numify version strings with _ in it (dev releases)?
18:41 miyagawa with some blackbox experiment, it seems to me that it strips _ *first*, then does version->parse->numify
18:41 miyagawa is that correct?
18:41 miyagawa ref: https://github.com/miyagawa/cpanminus/issues/367 https://github.com/miyagawa/cpanminus/issues/377
18:41 dipsy [ Odd module range parsing · Issue #367 · miyagawa/cpanminus · GitHub ] [ `--dev` ignores more recent version 0.1.10 in favour of version 0.1.9_1 · Issue #377 · miyagawa/cpanminus · GitHub ]
18:41 ether do you mean versions in dist tarball names? or $VERSION statements in code?
18:41 miyagawa code
18:43 haarg https://github.com/CPAN-API/cpan-api/blob/master/lib/MetaCPAN/Util.pm#L18 looks pretty broken
18:43 dipsy [ cpan-api/Util.pm at master · CPAN-API/cpan-api · GitHub ]
18:46 miyagawa right, numify + 0 fails on _ with numeric warnings
18:46 miyagawa then this fix_version doesn't seem to do the right thing for $VERSION = '0.0.1_1'
18:47 ether that turns 0.0.9_1 into 0.0009001 which isn't horrible actually
18:47 miyagawa it is horrible
18:48 ether how is it incorrect?
18:48 miyagawa version.pm parses it as 0.000009_001
18:48 miyagawa so does perl interpreter
18:48 ether 0.0.9 == 0.000900
18:48 miyagawa it is not.
18:48 miyagawa 0.0.9 == 0.000009
18:49 miyagawa it is ambiguous because it lacks the prefix v, but that's at least how version.pm and perl interprets
18:49 ether perl 5.10.1 == 5.010100
18:49 xdg joined #metacpan
18:49 ether 0.0.9 = 0.000900
18:49 miyagawa ether: that's crazy
18:50 haarg ether: that isn't correct
18:50 haarg to perl, version.pm, or EUMM
18:50 xdg ether tipped me off that versions are being discussed. Context, please?
18:50 miyagawa ether: perl -Mversion -e 'warn version->new("1.2.3")->numify'
18:51 miyagawa xdg: context = https://github.com/CPAN-API/cpan-api/blob/master/lib/MetaCPAN/Util.pm#L18 is incorrect
18:51 dipsy [ cpan-api/Util.pm at master · CPAN-API/cpan-api · GitHub ]
18:51 miyagawa and causes issues like https://github.com/miyagawa/cpanminus/issues/377 https://github.com/miyagawa/cpanminus/issues/367 on my end
18:51 dipsy [ `--dev` ignores more recent version 0.1.10 in favour of version 0.1.9_1 · Issue #377 · miyagawa/cpanminus · GitHub ] [ Odd module range parsing · Issue #367 · miyagawa/cpanminus · GitHub ]
18:52 haarg it should just be version->new($version)->numify =~ s/_//gr
18:52 miyagawa i just added a workaround on my side to numify as metacpan does before sending a query: https://github.com/miyagawa/cpanminus/commit/eeb2b049bd1170cfa62cc59f78230ec0b5ddf25b
18:52 dipsy [ numify version strings if MetaCPAN does, when sending queries to it. Fix... · eeb2b04 · miyagawa/cpanminus · GitHub ]
18:53 miyagawa xdg: MetaCPAN's numify code handles '0.0.9_1' as 0.000901, where as version.pm handles it 0.000009_001
18:53 miyagawa so that's all
18:54 xdg 0.000901?!?
18:54 miyagawa because it strips _ before parsing it
18:54 miyagawa (eventually)
18:54 miyagawa sorry 0.000091
18:55 xdg ah
18:55 haarg miyagawa: that seems wrong
18:55 haarg it converts 0.1.9_1 to 0.1.9001, which numifies to 0.0019001
18:55 miyagawa oh, ok.
18:55 miyagawa $version =~ s/_/00/g;
18:55 miyagawa yeah, smh.
18:56 miyagawa it seems to me it doesn't need this if $version =~ /^v/ check at all
18:56 * xdg look at git blame
18:56 xdg https://github.com/CPAN-API/cpan-api/commit/7386fad91f9659f3286d8e3890717294e23e3227
18:56 dipsy [ workaround for bogus version numbers · 7386fad · CPAN-API/cpan-api · GitHub ]
18:57 xdg fsvo "bogus"
18:57 miyagawa yeah
18:58 miyagawa ether: perl 5.10.1 is 5.010001, not 5.010100 :)
18:58 ether miyagawa: yes, you're right!
18:58 * ether should have another cup of coffee.
18:59 ether (and if I were right, 0.0.10 would never be higher than 0.0.1.)
18:59 miyagawa right ;)
18:59 ether the bit that concerns me is that there's any custom version mangling code in here at all. what is needed that version.pm doesn't provide itself?
18:59 miyagawa xdg: haarg: care to open an issue for this? I'll ship with cpanm with this workaround for now until metacpan gets fixed
19:00 ether every time something needs to do its own implementation of soemthing, bits get missed or done incorrectly
19:00 miyagawa version.pm's numify returning the string with "_" in it is crazy to me
19:00 haarg ether: $version->numify doesn't return numbers
19:00 miyagawa that i need to eval() on it
19:00 haarg ^
19:00 ether haarg: that's a big problem then :)
19:01 xdg that's because version "numbers" aren't numbers.  they are strings.
19:01 miyagawa numify_kinda
19:01 ether numify should return something that passes Scalar::Util::looks_like_number
19:01 haarg pretty sure it's been reported, but jpeacock
19:01 xdg "numify" is a bad name
19:01 haarg https://rt.cpan.org/Ticket/Display.html?id=58084
19:01 dipsy [ Bug #58084 for version: Numifying dev version (e.g. '1.2.3_4') does not return a number ]
19:01 xdg "decimal_form" & "tuple_form" is really the point
19:03 xdg ether, CPAN::Meta::Requirements|Converter both do a little bit of extra cleanup to satisfy version.pm so I understand *some* preprocessing, but that Util code is ... uh... I can't find polite words
19:04 miyagawa > Why did you want to numify() the object in the first place?
19:04 miyagawa ... Why do you provide numify() method if you think you don't need it!?
19:05 haarg he doesn't want it to exist
19:06 miyagawa it's funny every time i come back to maintain perl software i always have to deal with versions
19:08 ether yeah
19:08 ether we keep going around and around wringing our hands about how awful it is, but we can't seem to fix it
19:08 ether which says to me it's a people problem not a tech problem
19:08 ether because tech problems we can solve much easier, usually :)
19:10 xdg at this point it's a back-compat problem as much as a tech/people problem
19:13 haarg commit message: 'improved logging' adds fix_version to the dist version :(
19:14 stelim joined #metacpan
19:36 metacpan joined #metacpan
19:36 metacpan [metacpan-developer] ranguard created leo/puppet3 (+4 new commits): http://git.io/EZiFSA
19:36 metacpan metacpan-developer/leo/puppet3 2d70542 Leo Lapworth: stop daemon control, carton etc now that puppet does this
19:36 metacpan metacpan-developer/leo/puppet3 f630184 Leo Lapworth: comment out puppet3 enabling
19:36 metacpan metacpan-developer/leo/puppet3 a8ede27 Leo Lapworth: mount repos as service name dir
19:36 metacpan left #metacpan
19:36 dipsy [ Comparing 2d70542815d0^...0a3aafea8549 · CPAN-API/metacpan-developer · GitHub ]
19:38 metacpan joined #metacpan
19:38 metacpan [metacpan-developer] ranguard pushed 1 new commit to leo/puppet3: http://git.io/uLv7eA
19:38 metacpan metacpan-developer/leo/puppet3 9d0e779 Leo Lapworth: this is not a fixme any more as the new base box has puppet3
19:38 metacpan left #metacpan
19:38 dipsy [ this is not a fixme any more as the new base box has puppet3 · 9d0e779 · CPAN-API/metacpan-developer · GitHub ]
19:40 metacpan joined #metacpan
19:40 metacpan [metacpan-developer] ranguard opened pull request #17: Upgrade to the new VM box, which has puppet 3 (master...leo/puppet3) http://git.io/Z_dOsw
19:40 metacpan left #metacpan
19:40 dipsy [ Upgrade to the new VM box, which has puppet 3 by ranguard · Pull Request #17 · CPAN-API/metacpan-developer · GitHub ]
19:43 metacpan joined #metacpan
19:43 metacpan [metacpan-developer] oalders closed pull request #17: Upgrade to the new VM box, which has puppet 3 (master...leo/puppet3) http://git.io/Z_dOsw
19:43 metacpan left #metacpan
19:43 dipsy [ Upgrade to the new VM box, which has puppet 3 by ranguard · Pull Request #17 · CPAN-API/metacpan-developer · GitHub ]
19:43 metacpan joined #metacpan
19:43 metacpan [metacpan-developer] oalders reopened pull request #17: Upgrade to the new VM box, which has puppet 3 (master...leo/puppet3) http://git.io/Z_dOsw
19:43 metacpan left #metacpan
19:43 dipsy [ Upgrade to the new VM box, which has puppet 3 by ranguard · Pull Request #17 · CPAN-API/metacpan-developer · GitHub ]
19:51 metacpan joined #metacpan
19:51 metacpan [metacpan-developer] oalders pushed 1 new commit to master: http://git.io/IBsl3g
19:51 metacpan metacpan-developer/master 6f25f7c Olaf Alders: Merge pull request #17 from CPAN-API/leo/puppet3...
19:51 metacpan left #metacpan
19:51 dipsy [ Merge pull request #17 from CPAN-API/leo/puppet3 · 6f25f7c · CPAN-API/metacpan-developer · GitHub ]
19:51 metacpan joined #metacpan
19:51 metacpan [metacpan-developer] oalders deleted leo/puppet3 at 9d0e779: http://git.io/iZc0Hg
19:51 metacpan left #metacpan
19:51 dipsy [ this is not a fixme any more as the new base box has puppet3 · 9d0e779 · CPAN-API/metacpan-developer · GitHub ]
19:55 metacpan joined #metacpan
19:55 metacpan [metacpan-developer] oalders pushed 1 new commit to master: http://git.io/hBpIxg
19:55 metacpan metacpan-developer/master 05a152d Olaf Alders: Updates path to api checkout.
19:55 metacpan left #metacpan
19:55 dipsy [ Updates path to api checkout. · 05a152d · CPAN-API/metacpan-developer · GitHub ]
20:04 haarg miyagawa, xdg: https://github.com/CPAN-API/cpan-api/issues/337
20:04 dipsy [ version numification is incorrect · Issue #337 · CPAN-API/cpan-api · GitHub ]
20:08 ranguard ribasushi: cpan.metacpan.org now running on new backend server - please let me know if you get any issues
20:18 miyagawa haarg: thanks <3
20:20 metacpan joined #metacpan
20:20 metacpan [metacpan-developer] ranguard pushed 1 new commit to master: http://git.io/BxT1Zg
20:20 metacpan metacpan-developer/master 58b34a0 Leo Lapworth: if we are renaming checkout dir, need to change mount!
20:20 metacpan left #metacpan
20:20 dipsy [ if we are renaming checkout dir, need to change mount! · 58b34a0 · CPAN-API/metacpan-developer · GitHub ]
20:23 ranguard ribasushi: main change... nginx/0.7.67 (2010 ish) -> nginx/1.2.1 (though maybe we should look at using backports to get an even newer one)
21:36 stelim left #metacpan
23:19 harleypig joined #metacpan
23:21 harleypig The _search method using the api is returning "SearchPhaseExecutionException[Failed to execute phase [query], total failure; shardFailures {[... blah blah blah"
23:24 harleypig http://api.metacpan.org/v0/release/_search?size=10&amp;q=status:latest%20AND%20B::
23:24 harleypig I just tried without the :: and it worked fine, but I'm pretty sure I've been able to search before like that ...
23:36 harleypig This response is being returned even if :: is url encoded.
23:41 kentnl harleypig: try using \:\:
23:42 kentnl otherwise it sees ":" as a parameter:value seperator, regardless of URL encoding ( because URL encoding gets un-done as soon as it hits the server side code )
23:43 harleypig Ah. That works. Thank you.

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