WIP: update rugged to 0.25.1.1, #17325, #14972, #17019

@smcgivern seems gollum-rugged_adapter new version needs to be released first:

➔ bundle update rugged
Fetching source index from http://rubygems.org/
Resolving dependencies.............................................................................................................................................................
Bundler could not find compatible versions for gem "rugged":
  In Gemfile:
    gollum-rugged_adapter (~> 0.4.2) ruby depends on
      rugged (>= 0.21.3, ~> 0.24.0) ruby

    rugged (0.25.1)

mentioned in issue #17325 (closed)

Related: https://gitlab.com/gitlab-org/gitlab-ce/merge_requests/2999 / https://github.com/gollum/rugged_adapter/pull/17

It seems like a PR like the latter is needed here too, then

added ~164274 ~112478 labels

mentioned in issue #17019 (moved)

changed title from WIP: update rugged to 0.25.1, #17325 (closed), #14972 (closed) to WIP: update rugged to 0.25.1, #17325 (closed), #14972 (closed), #17019

waits for https://github.com/gollum/rugged_adapter/pull/22

added 207 commits

5c7d0ed6...bf073583 - 206 commits from branch gitlab-org:master
6dc5ba11 - update rugged to 0.25.1, #17325 (closed), #14972 (closed), #17019 (moved)

Compare with previous version

mentioned in issue #26179 (closed)

mentioned in issue #14765 (closed)

added 1 commit

ad7f2878 - use :git ref for rugged_adapter

Compare with previous version

i pushed to use git ref. perhaps that's the way to go?

because looks like dead loop:

rugged_adapter maintainer(s) don't want to merge because no testing
gitlab can't be tested because no new version of rugged_adapter

rugged_adapter maintainer(s) don't want to merge because no testing

Did they say that? I don't see that in the linked PR

In general, we'd rather fork and push a gem rather than use a git ref, I think.

i just assumed so: https://github.com/gollum/rugged_adapter/pull/22#issuecomment-272261491

at least with current MR state people can do testing if they manage to merge this MR into their source based install :)

The spec failures look legitimate - did something in libgit2 change that would affect comparisons?

i retried one rspec job which timeouted; but indeed the other rspec/spinach failures seem to be related to libgit2/rugged update

fixing those rspec/spinach failures is over my leaguage/available free time.

also, seems this needs fixing:

Gemfile:63:58: C: Style/HashSyntax: Use the new Ruby 1.9 hash syntax.
gem 'gollum-rugged_adapter', '~> 0.4.3', require: false, :git => 'https://github.com/glensc/rugged_adapter', :ref => '954c6a313'
                                                         ^^^^^^^

altho this was from official bundle documentation syntax

added 89 commits

ad7f2878...1028b111 - 87 commits from branch gitlab-org:master
284280de - update rugged to 0.25.1, #17325 (closed), #14972 (closed), #17019 (moved)
10e51099 - use :git ref for rugged_adapter

Compare with previous version

That RuboCop error just means to write git: 'https://...' instead of :git => 'https://...'

Thanks for working on this, @glensc! We will try to have someone pick https://gitlab.com/gitlab-org/gitlab-ce/issues/17325 up in 8.17.

@smcgivern yeah, they can base their branch on !8538 (closed)

direct links for failed builds (for easy access):

WIP: update rugged to 0.25.1.1, #17325, #14972, #17019

What does this MR do?

Are there points in the code the reviewer needs to double check?

Why was this MR needed?

Screenshots (if relevant)

Does this MR meet the acceptance criteria?

What are the relevant issue numbers?

Activity

WIP: update rugged to 0.25.1.1, #17325, #14972, #17019

What does this MR do?

Are there points in the code the reviewer needs to double check?

Why was this MR needed?

Screenshots (if relevant)

Does this MR meet the acceptance criteria?

What are the relevant issue numbers?

Merge request reports

Activity