Skip to content
Commits on Jul 9, 2010
  1. @hpoydar @defunkt

    Work with updated Campfire API

    hpoydar committed with defunkt Jun 5, 2010
  2. @shingara @defunkt
  3. @shingara @defunkt
  4. @shingara @defunkt
Commits on Apr 7, 2010
  1. @defunkt

    rake publish

    defunkt committed Apr 7, 2010
  2. @defunkt

    v0.4.0

    defunkt committed Apr 7, 2010
  3. @fritzek @defunkt

    Add build queueing

    fritzek committed with defunkt Apr 7, 2010
Commits on Apr 4, 2010
  1. @dcrec1 @defunkt

    Added Signal to the bad boys :)

    dcrec1 committed with defunkt Apr 2, 2010
Commits on Apr 1, 2010
  1. @defunkt

    Link to Passenger blog post

    defunkt committed Mar 31, 2010
Commits on Mar 25, 2010
  1. @vjt @defunkt
  2. @tmm1 @defunkt

    set environment variables using ENV

    tmm1 committed with defunkt Feb 13, 2010
Commits on Feb 28, 2010
  1. @defunkt

    0.3.0

    defunkt committed Feb 28, 2010
  2. @defunkt
  3. @antono @defunkt
  4. @antono @defunkt

    Fixed tests broken by ec1677a

    antono committed with defunkt Feb 21, 2010
     * project_path initialization in Config
       take global (cmdline) or use self as default
  5. @rsanheim @defunkt

    Wider layout for wide monitors (40em -> 80%)

    rsanheim committed with defunkt Feb 4, 2010
     Cherry picked from Rob's Sanheim master:
    
       We have big monitors - lets use a little more space for the output
  6. @rsanheim @defunkt

    cherry-picked: Fix a very annoying bug that was breaking HTTP basic a…

    rsanheim committed with defunkt Feb 3, 2010
    …uth...
    
      - The load order was such that the config class was being loaded
        before the project_path was being chdir'ed to.  This meant the first
        time the http auth user/pass was being retrieved (via `git config`)
        it was being done in the current dir, instead of the project dir --
        and the current dir was often not a git repo.
    
        So that meant that the http user/pass was empty strings the first
        time, which was the only time the server would check it - so http auth
        was effectively never working.  At least for me.
    
        Not sure how this worked for anyone before -- maybe everyone who has used
        http auth starts cijoe from within the repo dir?
    
        Anyways, this is a fairly ugly hack to ensure that the git commands in Config
        are always run in the context of the project path.
Commits on Feb 13, 2010
  1. @adonaldson @defunkt
Commits on Jan 21, 2010
  1. @jbarnette @defunkt
  2. @jbarnette @defunkt
Commits on Jan 20, 2010
  1. @tmm1 @defunkt

    prefer globals to constants

    tmm1 committed with defunkt Jan 20, 2010
  2. @defunkt

    prefer globals to constants

    defunkt committed Jan 20, 2010
  3. @tmm1 @defunkt
Commits on Jan 3, 2010
  1. @defunkt

    Always show some SHA1

    Evan Phoenix committed with defunkt Dec 31, 2009
  2. @defunkt

    Fix ugly thread bug

    Evan Phoenix committed with defunkt Dec 31, 2009
  3. @defunkt

    Fix build.pid and show build status on STDOUT

    Evan Phoenix committed with defunkt Dec 31, 2009
Commits on Dec 24, 2009
  1. @defunkt

    bump version

    defunkt committed Dec 24, 2009
  2. @defunkt

    style

    defunkt committed Dec 24, 2009
  3. @phuesler @defunkt

    adding some tests to check for a valid git config values.

    phuesler committed with defunkt Nov 9, 2009
     This fixes a bug with earlier versions of git which do not
     have a config command. With older versions, cijoe would start
     up in auth mode and take gits error message as username
     and password.
  4. @threedaymonk @defunkt
Commits on Nov 11, 2009
  1. @rtomayko

    we don't need the diff, thanks

    rtomayko committed Nov 11, 2009
  2. @rtomayko

    fix issues parsing raw commit info when a commit is a merge

    rtomayko committed Nov 11, 2009
    There's an additional "Merge: SHA1 SHA2" header line so we can't
    assume the line positions mean anything.
  3. @rtomayko

    write current build immediately after obtaining SHA

    rtomayko committed Nov 11, 2009
    Prevents issues with the last_build not having a SHA when
    something fails before the command is executed.
  4. @rtomayko
  5. @rtomayko
Something went wrong with that request. Please try again.