1. 01 Jun, 2018 1 commit
  2. 31 May, 2018 6 commits
  3. 27 May, 2018 1 commit
  4. 23 May, 2018 2 commits
    • Ben North's avatar
      Avoid error if gitk not found on windows-nt · 4c6e9ed0
      Ben North authored
      On windows-nt, if gitk is not in the path, then the 'which' produces
      nothing to stdout (although it does produce an error message to
      stderr).  The magit-git-string call then returns nil, which is not a
      valid input to file-executable-p, causing an error when trying to
      autoload anything defined in magit-extras.el.
      
      Include 'exe' as a first conjunct to the 'and', short-circuiting the
      file-executable-p test in this case.
      4c6e9ed0
    • Jonas Bernoulli's avatar
      magit-insert-remote-branches: Guard against empty message · 22747048
      Jonas Bernoulli authored
      Fixes #3454, like 44f8e45d fixed #3382.
      22747048
  5. 22 May, 2018 3 commits
    • Kyle Meyer's avatar
      dc687d59
    • Kyle Meyer's avatar
      relnotes: Update for #3443 · b07ef8f8
      Kyle Meyer authored
      b07ef8f8
    • Li-Yun Chang's avatar
      magit-blame--format-time-string: Correct timezone handling · 9d6e9c84
      Li-Yun Chang authored
      Because the time variable is an Unix time which is not related to time zone,
      we don't have to update time according to time zone.
      
      To deal with the different behavior of `format-time-string`
      
      In emacs version > 25 :
      
      If time-format contains the time zone symbol ( %z or %Z ),
      use tz-in-second to format the time in committer time zone or author time zone.
      
      If time-format doesn't contain the time zone symbol ( %z or %Z ),
      assign tz-in-second to nil to format the time in local time zone.
      
      In emacs version < 25 :
      
      Just format the time string in local time zone
      9d6e9c84
  6. 19 May, 2018 1 commit
    • Jonas Bernoulli's avatar
      magit-diff-wash-diffstat: Handle renames · 66d9558f
      Jonas Bernoulli authored
      But ignore the special-case of file-names that contain " => ".
      Handling that properly would require the use of "-z", which would
      require changes all over the place and this code is going to be
      refactored soon anyway.
      
      Fixes #3440.
      66d9558f
  7. 17 May, 2018 7 commits
  8. 15 May, 2018 3 commits
    • Jonas Bernoulli's avatar
      magit-blame-process-filter: Uglify for the benefit of Emacs 24 · adbf6962
      Jonas Bernoulli authored
      * Consult match-data outside of `with-slots' before that messes
        with match-data on Emacs 24.
      * Call the `magit-blame-chunk' with an empty string as the value
        of the positional and optional NAME argument. That argument is
        obsolete as of Emacs 25, but in Emacs 24 it isn't optional yet.
      * Don't call `magit-blame-chunk' directly, instead let-bind a
        variable and then `funcall' that.  This leaves the byte-compiler
        on Emacs 25 and 26 in the dark, which keeps it from complaining.
      
      Fixes #3446.
      adbf6962
    • Jonas Bernoulli's avatar
      magit-log-buffer-file: Fix checking if file is regular · c271ddeb
      Jonas Bernoulli authored
      Fix the issue reported and fixed in #3418: the filename is relative to
      the top-level but `default-directory' may be a sub-directory, and also
      fix the regression introduced by that attempt and reported in #3448
      and #3449: `buffer-file-name' is nil in blob-visiting and Dired
      buffers, where this command is expected to work too.
      c271ddeb
    • Jonas Bernoulli's avatar
      Revert "magit-log-buffer-file: Use absolute filename in call to file-regular-p" · fb35b4af
      Jonas Bernoulli authored
      This reverts commit c234bf59.
      
      Reverts #3418.
      Fixes #3449.
      See next commit.
      fb35b4af
  9. 14 May, 2018 2 commits
    • Jonas Bernoulli's avatar
      Silence byte-compiler · 36de4e39
      Jonas Bernoulli authored
      36de4e39
    • Jonas Bernoulli's avatar
      Fix remembering whether user confirmed editing published commits · a2e1139c
      Jonas Bernoulli authored
      We now use a variable to remember whether the user already confirmed
      that s/he wants to edit published commits when initiating the rebase
      so that s/he won't be asked again when there are staged changes when
      continuing the rebase.  Previously we tried to use a file in the
      control directory used by Git, but that directory does not exist at
      the time we need to remember and if we create it ourselves, then Git
      trips over it.
      
      Fixes #3447.
      a2e1139c
  10. 12 May, 2018 8 commits
  11. 11 May, 2018 3 commits
  12. 10 May, 2018 3 commits