Redmine

This patch also fixes issues #7353 and #2761 (which are duplicates of other issues, but nobody flagged them as such, and I don't seem to be allowed to flag issues as duplicates).

Well I would like to test this issue. But currently I'm at the hospital. Just in case of the description, this sounds to be worth to look at it. :-)

No worries. Get back to good health and return home from hospital, redmine is no nearly as important as that!

I just hope this patch won't share the fate of those others and still be here, unresolved (i.e. neither accepted nor rejected) in four years g.

Thanks!

Well I think this one will find a faster way. If there is the next minor release available this could be set to this release or at least the next major.

Sadly after 1 month, it seems nobody had a chance to look at this. Too bad :-(. It would be nice to get at least some feedback on these things... as it is, development of Redmine seems to happen mostly behind closed doors and is not very welcome to outside contributions :-(.

Perhaps somebody did look at this patch and found an issue with that -- but without any feedback, that's impossible to tell for me :-(.

Because this patch does not have tests, I don't know it is correct or not.
I think it is better that Graph logic is moved from controller to helper in order to test.
See: source:tags/2.3.0/test/unit/lib/redmine/helpers/gantt_test.rb

OK, that makes sense. I am not yet very familiar with the testing system (of Redmine, or any other Ruby / RoR app in general -- I am fairly new to Ruby). I will investigate the gantt_test you referred me to, thanks for the point. I will try and see if I can do something. Likewise, moving the logic from controller to helper is something I can do, if you prefer that.

However, currently I am swamped with other work, so it's easily possible that another month or two will pass before I get to address either point. If somebody else beats me to that with a better patch, I won't complain ;-) but in any case, the ball is now in my court, as I have some clear feedback. Thank you very much for that, and for your work on Redmine in general.

Hi,
it's working great.
Thanks

I would love to see this patch in next version,
thanks.

Ruby 1.8.7 PostgreSQL test fails.
https://travis-ci.org/marutosi/redmine-bb/jobs/33601888

Toshi MARUYAMA wrote:

Ruby 1.8.7 PostgreSQL test fails.
https://travis-ci.org/marutosi/redmine-bb/jobs/33601888

Thanks! That should be fixed in r13352. There's no check on Change#action length as it's not user-controlled but it seems SQL drivers don't react the same way so we may be better adding some...

I added committers/users mapping in r13353, but I did not add the other features discussed by Max in the original issue (sort committer names, merge homonyms). In case tests pass, and if JPL and Toshi are OK with that, this feature could be released with Redmine 2.6.0.

Note that this can have a bit of performance impact because with this implementation multiple queries are run to find the correct user (it relies on Repository#find_committer_user while it could use Changeset#user_id directly). I'll improve that in the next few days.

Also it made me realize there are missing tests for Repository#find_committer_user and I think some edge cases are not handled correctly. I'll dig into that later.

Toshi MARUYAMA wrote:

Ruby 1.8.7 PostgreSQL test fails.
https://travis-ci.org/marutosi/redmine-bb/jobs/33601888

Because I restarted, log disappeared.
Ruby 1.9.3 PostgreSQL test fails due to same reason.
https://travis-ci.org/marutosi/redmine-bb/jobs/33601897

  1) Error:
test_stats_by_author_reflect_changesets_and_changes(RepositoryTest):
ActiveRecord::StatementInvalid: PG::StringDataRightTruncation: ERROR:  value too long for type character varying(1)
: INSERT INTO "changes" ("action", "branch", "changeset_id", "from_path", "from_revision", "path", "revision") VALUES ($1, $2, $3, $4, $5, $6, $7) RETURNING "id"