I updated joda-time to 2.3 (low risk), and along with it tzdata version to 2013d, on master. I noticed that localrepo/org/jruby/joda-timezones/2013c/joda-timezones-2013c.pom contains references to joda-time and joda-timezones versions.
As it is unclear how this is intended to be maintained, I left the file untouched, but it may need to be updated as well.
Update to joda-time 2.3 (and tzdata 2013d)
$ jruby -ve 'require "rbconfig";
jruby 1.7.5.dev (1.9.3p392) 2013-09-02 ba28553 on Java HotSpot(TM) 64-Bit Server VM 1.7.0_25-b15 [darwin-x86_64]
More precisely, please explain how the files committed in c70c359 were generated.
the localrepo and its content needs to go away COMPLETELY for the release
next jruby version.
we have already a project https://github.com/jruby/joda-timezones
there is another idea about a timezone provider around:
JodaOrg/joda-time#37 which could solve the 'using
new timezone data' as well.
so the purpose is a temporary hack to get away the build_lib directory.
headius already started
which just needs a 'release'.
the joda-timezone still needs some attention on "how" we are going to do it
but there is no need it in the localrepo anymore.
requireTest is unclear for me - i.e.from where it comes from - could not
Pushed a6af6c1 so Travis can move forward with up-to-date tzdata.
joda-timezones has not moved forward because I got wrangled into writing a more comprehensive solution for joda-time. jsr292-mock has not moved forward because the author updated it to use ASM 5, which is not release is probably will not be released until after Java 8.
So I think both of these need to just get pushed under jruby groupId, with joda-timezones just supporting the one we want to ship with for 1.7.5 and jsr292-mock being just a binary release for now. Hopefully for 1.7.6 we can get them released as proper artifacts under the right orgs.
localrepo is now gone, replaced by real artifacts.