Permalink
Commits on Sep 27, 2016
  1. Merge tag 'v2.6.7'

    Tagging v2.6.7.
    segiddins committed Sep 27, 2016
  2. Version 2.6.7 with updated history

    segiddins committed Sep 26, 2016
Commits on Sep 26, 2016
  1. Auto merge of #1699 - rubygems:bugfix/gem_sources_load_gemrc, r=bronzdoc

    Load config in Gem.sources
    
    # Description:
    When calling `Gem.sources` load sources from configuration if present, else use default sources.
    
    closes #1613
    # Tasks:
    
    - [x] Describe the problem / feature
    - [x] Write tests
    - [x] Write code to solve the problem
    - [ ] Get code review from coworkers / friends
    - [ ] [Squash commits](http://gitready.com/advanced/2009/02/10/squashing-commits-with-rebase.html)
    
    I will abide by the [code of conduct](https://github.com/rubygems/rubygems/blob/master/CODE_OF_CONDUCT.md).
    homu committed with segiddins Aug 26, 2016
  2. Auto merge of #1683 - coderanger:nativeuser, r=segiddins

    Set loaded_from before using extension_dir to ensure it is correct for --user or other alternate install path modes
    
    Fixes #1680
    
    Not 100% sure the best way to write a test for this.
    
    I will abide by the [code of conduct](https://github.com/rubygems/rubygems/blob/master/CODE_OF_CONDUCT.md).
    homu committed with segiddins Aug 11, 2016
  3. Auto merge of #1711 - rubygems:seg-remote-fetcher-no-location-redirec…

    …t, r=indirect
    
    [RemoteFetcher] Fail gracefully when attempting to redirect without a Location
    
    # Description:
    
    Fixes http://help.rubygems.org/discussions/problems/22729
    
    # Tasks:
    
    - [ ] Describe the problem / feature
    - [ ] Write tests
    - [ ] Write code to solve the problem
    - [ ] Get code review from coworkers / friends
    - [ ] [Squash commits](http://gitready.com/advanced/2009/02/10/squashing-commits-with-rebase.html)
    
    I will abide by the [code of conduct](https://github.com/rubygems/rubygems/blob/master/CODE_OF_CONDUCT.md).
    homu committed with segiddins Sep 21, 2016
  4. Auto merge of #1714 - rubygems:seg-molinillo-0.5.1, r=indirect

    Update vendored Molinillo to 0.5.1
    
    # Description:
    
    See https://github.com/CocoaPods/Molinillo/releases/0.5.1
    
    # Tasks:
    
    - [ ] Describe the problem / feature
    - [ ] Write tests
    - [ ] Write code to solve the problem
    - [ ] Get code review from coworkers / friends
    - [ ] [Squash commits](http://gitready.com/advanced/2009/02/10/squashing-commits-with-rebase.html)
    
    I will abide by the [code of conduct](https://github.com/rubygems/rubygems/blob/master/CODE_OF_CONDUCT.md).
    homu committed with segiddins Sep 14, 2016
Commits on Sep 24, 2016
  1. Auto merge of #1723 - rubygems:drbrain/no-squashed-commits, r=bronzdoc

    Do not squash commits
    
    # Description:
    
    Squashing commits may be helpful for small pull requests where the total
    change is around 10—20 lines or two or three hunks.  With such a small
    patch it's easy to infer intent in the future.
    
    For larger patches it becomes increasingly difficult to determine why
    the change was made when a single commit contains dozens of hunks where
    none of them are related.
    
    This is especially important after either:
    * Enough time has passed that you don't remember why you made the patch
    * The patch was made by a maintainer that is no longer active
    * The patch was made by a contributor that is no longer interested
    
    In these situations, when an issue is found related to the patch the
    person working in the area has to infer the purpose of the patch from
    the context they have.  The new work won't necessarily be a bug in the
    original patch.  It may add functionality that incidentally became an
    important feature for some of our users, and the new work wishes to
    change something related to that change.
    
    The new work may be a bug fix that has a subtle interaction, so
    determining if the first patch was a buggy (even if unintentionally so)
    change, or more importantly how the change is buggy, will be difficult
    to determine the less context we have.
    
    By aggressively squashing commits we remove the commit messages that
    contain that very important context.  I see no reason to forbid
    squashing commits in a way that makes the intent of the patch more
    clear.  On the other hand, encouraging it via a check-box on our PR form
    may cause the loss of this important context when we need it.
    
    ______________
    
    # Tasks:
    
    - [x] Describe the problem / feature
    - [ ] Get code review from coworkers / friends
    
    I will abide by the [code of conduct](https://github.com/rubygems/rubygems/blob/master/CODE_OF_CONDUCT.md).
    homu committed Sep 24, 2016
  2. Do not squash commits

    Squashing commits may be helpful for small pull requests where the total
    change is around 10—20 lines or two or three hunks.  With such a small
    patch it's easy to infer intent in the future.
    
    For larger patches it becomes increasingly difficult to determine why
    the change was made when a single commit contains dozens of hunks where
    none of them are related.
    
    This is especially important after either:
    * Enough time has passed that you don't remember why you made the patch
    * The patch was made by a maintainer that is no longer active
    * The patch was made by a contributor that is no longer interested
    
    In these situations, when an issue is found related to the patch the
    person working in the area has to infer the purpose of the patch from
    the context they have.  The new work won't necessarily be a bug in the
    original patch.  It may add functionality that incidentally became an
    important feature for some of our users, and the new work wishes to
    change something related to that change.
    
    The new work may be a bug fix that has a subtle interaction, so
    determining if the first patch was a buggy (even if unintentionally so)
    change, or more importantly how the change is buggy, will be difficult
    to determine the less context we have.
    
    By aggressively squashing commits we remove the commit messages that
    contain that very important context.  I see no reason to forbid
    squashing commits in a way that makes the intent of the patch more
    clear.  On the other hand, encouraging it via a check-box on our PR form
    may cause the loss of this important context when we need it.
    drbrain committed Sep 24, 2016
  3. Auto merge of #1722 - koic:fix_typo_an_integer, r=hsbt

    Fix a typo
    
    # Description:
    
    Just modified `a Integer` to `an Integer` .
    ______________
    
    # Tasks:
    
    - [x] Describe the problem / feature
    - [ ] Write tests
    - [ ] Write code to solve the problem
    - [ ] Get code review from coworkers / friends
    - [ ] [Squash commits](http://gitready.com/advanced/2009/02/10/squashing-commits-with-rebase.html)
    
    I will abide by the [code of conduct](https://github.com/rubygems/rubygems/blob/master/CODE_OF_CONDUCT.md).
    homu committed Sep 24, 2016
  4. Fix a typo

    koic committed Sep 24, 2016
Commits on Sep 21, 2016
  1. Auto merge of #1697 - rubygems:seg-runtime-metadata-failure-suggestio…

    …n, r=indirect
    
    Add a helpful suggestion when `gem install` fails due to required_rub…
    
    # Description:
    
    This should help make it clearer what users (especially those not familiar with rubygems) can do to make installation succeed on an unsupported ruby/rubygems.
    
    I can add tests if others think this feature is worthwhile.
    
    Examples from 1.8.7:
    
    ```
    ERROR:  Error installing cocoapods:
    	There are no versions of cocoapods-core (= 1.0.1) compatible with your ruby & rubygems. Maybe try installing an older version of the gem you're looking for?
    	cocoapods-core requires Ruby version >= 2.0.0.
    ```
    
    ```
    ERROR:  Error installing cocoapods-downloader:
    	The last version of cocoapods-downloader (>= 0) to support your ruby & rubygems was 0.7.2. Try installing it with `gem install cocoapods-downloader -v 0.7.2`
    	cocoapods-downloader requires Ruby version >= 2.0.0.
    ```
    
    # Tasks:
    
    - [ ] Describe the problem / feature
    - [ ] Write tests
    - [ ] Write code to solve the problem
    - [ ] Get code review from coworkers / friends
    - [ ] [Squash commits](http://gitready.com/advanced/2009/02/10/squashing-commits-with-rebase.html)
    
    I will abide by the [code of conduct](https://github.com/rubygems/rubygems/blob/master/CODE_OF_CONDUCT.md).
    
    …y{gems,}_version
    homu committed Sep 21, 2016
  2. Auto merge of #1711 - rubygems:seg-remote-fetcher-no-location-redirec…

    …t, r=indirect
    
    [RemoteFetcher] Fail gracefully when attempting to redirect without a Location
    
    # Description:
    
    Fixes http://help.rubygems.org/discussions/problems/22729
    
    # Tasks:
    
    - [ ] Describe the problem / feature
    - [ ] Write tests
    - [ ] Write code to solve the problem
    - [ ] Get code review from coworkers / friends
    - [ ] [Squash commits](http://gitready.com/advanced/2009/02/10/squashing-commits-with-rebase.html)
    
    I will abide by the [code of conduct](https://github.com/rubygems/rubygems/blob/master/CODE_OF_CONDUCT.md).
    homu committed Sep 21, 2016
  3. Auto merge of #1716 - rubygems:update-contributing, r=segiddins

    Update Contributing.rdoc with new label usage
    
    ---
    
    I will abide by the [code of conduct](https://github.com/rubygems/rubygems/blob/master/CODE_OF_CONDUCT.md).
    
    ---
    
    reference issue: #1715
    
    Some labels were changed in https://github.com/rubygems/rubygems/labels to match this document
    
    - Add the `closed: discussion` label
    - Workflow labels are denoted by `status: *`
    - All labels in the same broad category now share a color
    - Adding a space to "bug fix" allows it to be sorted next to "bug report"
    - Add type: documentation
    - Minor syntax and wording changes
    - A list of other changes are mentioned in the commit descriptions
    homu committed Sep 21, 2016
  4. Auto merge of #1721 - rubygems:seg-update-bundler, r=segiddins

    Update bundler to handle new rails 3 release in specs
    
    # Description:
    
    Fix CI
    
    # Tasks:
    
    - [ ] Describe the problem / feature
    - [ ] Write tests
    - [ ] Write code to solve the problem
    - [ ] Get code review from coworkers / friends
    - [ ] [Squash commits](http://gitready.com/advanced/2009/02/10/squashing-commits-with-rebase.html)
    
    I will abide by the [code of conduct](https://github.com/rubygems/rubygems/blob/master/CODE_OF_CONDUCT.md).
    homu committed Sep 21, 2016
  5. Update bundler to handle new rails 3 release in specs

    segiddins committed Sep 21, 2016
  6. Split up working / ready

    lynnco committed on GitHub Sep 21, 2016
Commits on Sep 15, 2016
  1. Update CONTRIBUTING.rdoc

    Platforms do not need descriptions, so they do not need a list
    lynnco committed on GitHub Sep 15, 2016
  2. Update CONTRIBUTING.rdoc

    move color descriptions to the top
    lynnco committed on GitHub Sep 15, 2016
  3. Update CONTRIBUTING.rdoc

    Add a new "Contribution" label type
    lynnco committed on GitHub Sep 15, 2016
  4. Update CONTRIBUTING.rdoc

    fix rdoc code block syntax for multiple words
    lynnco committed on GitHub Sep 15, 2016
  5. Update CONTRIBUTING.rdoc

    Fix rdoc code block and bolding syntax
    lynnco committed on GitHub Sep 15, 2016
  6. Update CONTRIBUTING.rdoc

    - Add `status: user feedback required`
    - Change `category: *` syntax to match the other `type: subtype` labels
    - Standardize the bolding syntax
    - Change `accepted` to `confirmed`, shift its use slightly
    - Change `blocked` to `blocked / backlog`
    - Describe `user feedback required`
    - Change `ready for work` to `working / ready`, don't remove once claimed
    - Change `claimed` to `unclaimed`, as unclaimed issues are what people search for
    - Remove `feedback` label pending a re-assessment of its use
    - Describe the workflow... flow
    - Fix spacing issue
    lynnco committed on GitHub Sep 15, 2016
  7. Update Contributing.rdoc with new label usage

    ref: #1715
    
    Some labels were changed in https://github.com/rubygems/rubygems/labels to match this document
    
    Add the `closed: discussion` label
    Workflow labels are denoted by `status: *`
    All labels in the same broad category now share a color
    Adding a space to "bug fix" allows it to be sorted next to "bug report"
    Add type: documentation
    Minor syntax and wording changes
    Likely other smalls changes that I forgot to mention
    lynnco committed on GitHub Sep 15, 2016
Commits on Sep 14, 2016
  1. Auto merge of #1714 - rubygems:seg-molinillo-0.5.1, r=indirect

    Update vendored Molinillo to 0.5.1
    
    # Description:
    
    See https://github.com/CocoaPods/Molinillo/releases/0.5.1
    
    # Tasks:
    
    - [ ] Describe the problem / feature
    - [ ] Write tests
    - [ ] Write code to solve the problem
    - [ ] Get code review from coworkers / friends
    - [ ] [Squash commits](http://gitready.com/advanced/2009/02/10/squashing-commits-with-rebase.html)
    
    I will abide by the [code of conduct](https://github.com/rubygems/rubygems/blob/master/CODE_OF_CONDUCT.md).
    homu committed Sep 14, 2016
Commits on Sep 12, 2016
  1. Update vendored Molinillo to 0.5.1

    segiddins committed Sep 12, 2016
Commits on Sep 8, 2016
  1. [RemoteFetcher] Fail gracefully when attempting to redirect without a…

    … Location
    segiddins committed Sep 8, 2016
Commits on Sep 7, 2016
  1. Auto merge of #1709 - aycabta:add-host-sample-to-help, r=segiddins

    Add --host sample to help
    
    # Description:
    
    `push`, `yank` and `owner` subcommands have `--host` option. This needs URI scheme. For example, `--host rubygems.org` is not valid, `--host https://rubygems.org` is valid. So I added sample to help.
    
    before:
    
            --host HOST                  Use another gemcutter-compatible host
    
    after:
    
            --host HOST                  Use another gemcutter-compatible host
                                           (e.g. https://rubygems.org)
    
    # Tasks:
    
    - [x] Describe the problem / feature
    - [x] Write tests **(no need)**
    - [x] Write code to solve the problem
    - [x] Get code review from coworkers / friends
    - [x] [Squash commits](http://gitready.com/advanced/2009/02/10/squashing-commits-with-rebase.html)
    
    I will abide by the [code of conduct](https://github.com/rubygems/rubygems/blob/master/CODE_OF_CONDUCT.md).
    homu committed Sep 7, 2016
  2. Update Manifest.txt for bundler/feature_flag.rb

    segiddins committed Sep 7, 2016
  3. Update to a bundler that has spec support for compact_index 0.11

    segiddins committed Sep 7, 2016
  4. Add --host sample to help

    aycabta committed Sep 7, 2016
Commits on Aug 26, 2016
  1. Auto merge of #1699 - rubygems:bugfix/gem_sources_load_gemrc, r=bronzdoc

    Load config in Gem.sources
    
    # Description:
    When calling `Gem.sources` load sources from configuration if present, else use default sources.
    
    closes #1613
    # Tasks:
    
    - [x] Describe the problem / feature
    - [x] Write tests
    - [x] Write code to solve the problem
    - [ ] Get code review from coworkers / friends
    - [ ] [Squash commits](http://gitready.com/advanced/2009/02/10/squashing-commits-with-rebase.html)
    
    I will abide by the [code of conduct](https://github.com/rubygems/rubygems/blob/master/CODE_OF_CONDUCT.md).
    homu committed Aug 26, 2016
  2. Use config sources if available else use default source

    bronzdoc committed Aug 26, 2016
  3. Add accessor for sources

    bronzdoc committed Aug 26, 2016
  4. Auto merge of #1633 - ntl:master, r=indirect

    Sort ordering of sources is preserved
    
    Resolves #1609
    
    I will abide by the [code of conduct](https://github.com/rubygems/rubygems/blob/master/CODE_OF_CONDUCT.md).
    homu committed Aug 26, 2016
  5. Auto merge of #1696 - sonalkr132:suggest_prerelase, r=indirect

    Retry with :prerelease when no suggestions are found
    
    # Description:
    closes: #116
    When no matches are found with latest specs, `suggest_gems_from_name` will retry with prerelease specs.
    # Tasks:
    
    - [x] Describe the problem / feature
    - [x] Write tests
    - [x] Write code to solve the problem
    - [ ] Get code review from coworkers / friends
    - [ ] [Squash commits](http://gitready.com/advanced/2009/02/10/squashing-commits-with-rebase.html)
    
    I will abide by the [code of conduct](https://github.com/rubygems/rubygems/blob/master/CODE_OF_CONDUCT.md).
    homu committed Aug 26, 2016