Permalink
Commits on Nov 5, 2010
  1. Update docs for 1.0b release

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Nov 5, 2010
Commits on Nov 4, 2010
  1. Fall back to name rather than giving up when selecting

    See:
    
      https://wincent.com/issues/1617
    
    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Nov 4, 2010
  2. Check one more place where Vim's always-0-buffer-number bug can crop up

    See:
    
      https://wincent.com/issues/1617
    
    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Nov 4, 2010
  3. Better fix for issue #1617

    Even though I can't reproduce the Vim "largefile" bug, I did some
    testing by stubbing $curbuf.number and forcing it to always return 0. It
    turns out that my fallback in this case wasn't working because if you
    do:
    
      :bunload! GoToFile
    
    Vim will fail to close the window because the GoToFile buffer is hidden
    and it can't find it.
    
      :bwipeout! GoToFile
    
    However, does work, so go with that in the fallback case.
    
    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Nov 4, 2010
  4. Swap conditional order

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Nov 4, 2010
  5. Fix a comment typo

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Nov 4, 2010
  6. Fall back to buffer name on platforms affected by buffer number bug

    I was reluctant to use the buffer name instead of the buffer number, but
    on platforms affected by the bug we have two choices:
    
      - allow the bug to break Command-T, causing support requests and
        requiring people to either patch the source or rebuild Vim
    
      - fall back to the buffer name, which may not be entirely robust but
        at least is better than being totally broken
    
    Seeing as we can maintain the correct, robust behavior on platforms not
    affected by the bug, and only use the fallback behavior when required,
    adding this workaround is probably acceptable until upstream Vim gets
    fixed.
    
    See:
    
      https://wincent.com/issues/1617
    
    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Nov 4, 2010
  7. Dispose of match listing using :bunload! rather than :bwipeout!

    Previously we created a new buffer every time we showed the match
    listing and then disposed of it each time with :bwipeout!. This caused
    the buffer numbers to steadily grow.
    
    Now, we dispose of the match listing using :bunload!; this frees the
    contents of the buffer and closes the window, but the buffer sticks
    around and can be re-used next time we show the match listing. In this
    way we always have a stable buffer number for the match listing and
    repeatedly using Command-T doesn't drive the buffer numbers up any more.
    
    For more info see:
    
      https://wincent.com/issues/1724
    
    One concern with this change is that it increases our reliance on the
    accuracy of the buffer number, and there is a known bug in Vim 7.3 when
    built with --enable-largefile; see this ticket for full details:
    
      https://wincent.com/issues/1617
    
    If you are affected by this bug the solution is to build Vim using
    --disable-largefile. Until the issue gets fixed upstream I am thinking
    of adding a workaround in Command-T (falling back to buffer name rather
    than number for people with a broken Vim).
    
    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Nov 4, 2010
Commits on Oct 8, 2010
  1. Update docs for 0.9 release

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Oct 8, 2010
  2. Use a ternary to make a more compact conditional

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Oct 8, 2010
  3. Code comment fix (concordance)

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Oct 8, 2010
  4. Add license header to recently-added file

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Oct 8, 2010
  5. Use expect/to rather than lambdas for readability

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Oct 8, 2010
  6. Remove unneeded "private" in anonymous module

    No external caller is going to try calling these internal methods, so
    no need to mark them as private.
    
    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Oct 8, 2010
  7. Reword comment that was looking at me funny

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Oct 8, 2010
  8. Use a single "before" block in specs rather than two

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Oct 8, 2010
  9. Spec style fixes: reword to avoid unnecessary "should"

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Oct 8, 2010
  10. Require spec helper using relative path

    RSpec adds the spec directory to the load path, so there's no need to
    specify an absolute path.
    
    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Oct 8, 2010
  11. Bump minimum RSpec version to 2.0.0.rc

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Oct 8, 2010
  12. Open relative paths under the working directory

    This reduces absolute path noise in the buffer list, status line, tab
    line, etc., for the usual no-arg case of opening files in the working
    directory.
    
    It falls back to the previous absolute path behavior for files outside
    the working directory.
    
    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    matthewtodd committed with wincent Oct 8, 2010
  13. Add a source to the Gemfile

    `bundle install` dies otherwise.
    
    (At least for me; I'm using Bundler 1.0.2.)
    
    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    matthewtodd committed with wincent Oct 8, 2010
Commits on Sep 14, 2010
  1. Teach multi-spec.sh about Bundler

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Sep 14, 2010
  2. Rakefile: remove reference to long-gone readme:check task

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Sep 14, 2010
  3. Update docs for 0.8.1 release

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Sep 14, 2010
Commits on Sep 13, 2010
  1. `set timeout' on match window creation and restore on close

    See:
    
      https://wincent.com/forums/command-t/topics/432
    
    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    guns committed with wincent Sep 13, 2010
Commits on Sep 2, 2010
  1. doc: update with info for known-working versions on Windows

    See:
    
      https://wincent.com/issues/1673
    
    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Sep 2, 2010
Commits on Aug 23, 2010
  1. doc: update links and info for installing on Windows

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Aug 23, 2010
Commits on Aug 19, 2010
  1. Doc: update HISTORY section for 0.8 release

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Aug 19, 2010
  2. Doc: updates for Vim 7.3 release

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Aug 19, 2010
  3. Doc: update MacVim download link

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Aug 19, 2010
Commits on Aug 16, 2010
  1. Remove "releases" subdirectory

    Releases are now archived in the separate "releases" branch so as to
    stop them from polluting "grep -R", "git grep" and "ack" output with
    probably-irrelevant noise.
    
    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Aug 16, 2010
  2. Remove now-redundant 'readme' tasks

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Aug 16, 2010
  3. Make doc/command-t.txt a symlink to README.txt

    Originally README.txt was a symlink to doc/command-t.txt, added for the
    benefit of GitHub, but GitHub doesn't follow such symlinks so it didn't
    work.
    
    Instead, I set up Rake tasks for keeping the two files in sync.
    
    Evidently this wasn't enough because I've received patches from people
    editing one file or the other.
    
    So, invert the original approach and make doc/command-t.txt a symlink to
    README.txt. The full text gets included when making the Vimball, and
    Pathogen users will also get access to the documentation too because
    ":call pathogen#helptags()" does the right thing with symbolic links.
    
    Of course, in hindsight, I should have done it this way all along.
    
    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Aug 16, 2010
  4. Update README.txt for GitHub

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Aug 16, 2010
  5. doc: update contributor list and history

    Signed-off-by: Wincent Colaiuta <win@wincent.com>
    wincent committed Aug 16, 2010