Permalink
Commits on Mar 24, 2010
  1. Merge remote branch 'upstream/master'

    Conflicts:
    	generators/formtastic/templates/formtastic.css
    committed Mar 24, 2010
Commits on Mar 23, 2010
  1. don't render hints hidden inputs

    spec coverage for commit 694bffa
    committed Mar 23, 2010
Commits on Mar 20, 2010
Commits on Mar 19, 2010
  1. don't show hints on hidden inputs

    committed Mar 19, 2010
Commits on Mar 16, 2010
Commits on Mar 13, 2010
  1. change input widths to max-width, this allows :size to work, also mak…

    …e password input shorter
    twar59 committed with justinfrench Jan 15, 2010
  2. Make the legend width slightly smaller.

    Set the legend width to ~95%, this will add a little padding between the legend>label and the form elements to the right.
    
    This will only effect webkit and IEs
    twar59 committed with justinfrench Jan 10, 2010
  3. tiny README tweak

    justinfrench committed Mar 13, 2010
  4. resolves #219, failing spec when running time_input_spec by itself, a…

    …lthough in both cases, it should be for a different select altogether, looking into it
    justinfrench committed Mar 13, 2010
  5. Since we're not planning on support :default going forward, it doesn'…

    …t make sense to temporarily push people to migrate over to :default. Instead, the next release will just have a different (better) implementation of :selected.
    justinfrench committed Mar 1, 2010
  6. added a deprecation warning for :selected/:checked/:default -- they'l…

    …l be gone in 1.0 in favor of after_initialize or controller set-ups
    justinfrench committed Mar 1, 2010
Commits on Mar 10, 2010
  1. Fixes issue #209, which hidden fields were not always hidden (Chrome …

    …for example). The dual declarations are required because of our clearfix display hack on the LIs, which is more
    
    specific than the more general rule below.
    
    Thanks to adurity for testing.
    justinfrench committed Mar 10, 2010
Commits on Mar 1, 2010
  1. make eval'ed code's line-numbers line up

    It just makes debugging easier this way.
    timcharper committed with justinfrench Jan 30, 2010
  2. Made CSS rules for fieldset lists more specific

    This will make conflicts with other rules less likely. Tested
    in Chromium/Mac TRUNK, Safari 4.0.4 Mac, Firefox 3.5 Mac, IE8,
    IE8 compat. view without any noticable differences.
    Mange committed with justinfrench Jan 18, 2010
  3. Better workaround for ActiveRecord's flawed Base.human_name:

    * affecting commit buttons
    * in 0.9.6 and earlier we called @object.class.human_name, but this resulted in things like "Userpost" instead of "User post"
    * in 0.9.7 we called @object.class.name.underscore.humanize
    * 0.9.6 worked for people with their own i18n for model names (human_name falls back to "Userpost" if there's no i18n)
    * 0.9.7 worked for those without i18n ("User post" instead of "Userpost")
    * the crappy fallback in human_name is the issue we're routing around (@object.class.name.humanize)
    * so we now call @object.class.name.underscore.humanize only if it looks like human_name has fallen back to the crappy @object.class.name.humanize
    justinfrench committed Feb 16, 2010
  4. Radio and checkbox inputs no longer have the legend label associated …

    …with the first choice's input, resolving #101.
    justinfrench committed Feb 14, 2010
  5. shifting things around

    justinfrench committed Feb 7, 2010
  6. renamed a spec

    justinfrench committed Feb 7, 2010
  7. smart_require is tripping up the CI build, and also tripped me up a f…

    …ew times, so let's try explicit lines that do what we need to do... specs pass with what I have here, the only real change is we're explicitly requiring Rails 2.3.5 components to explude Rails 3
    justinfrench committed Feb 7, 2010
  8. Revert "generated config file is now called formtastic_config.rb rath…

    …er than just formtastic.rb, although it really doesn't matter what it's called"
    
    This reverts commit b872f39.
    justinfrench committed Feb 7, 2010
Commits on Jan 18, 2010
  1. quick fix for a spec that assumes Time.now.hour is a two digit number…

    …, better fix is to use Timecop or similar
    justinfrench committed Jan 18, 2010