Switch branches/tags
Nothing to show
Commits on May 15, 2015
  1. Properly wait for mtouch to exit and for stdout/stderr to complete

    Touch.Server was not waiting for mtouch to exit or for the last
    logging statements to come from mtouch before it exited. Now it
    will! My testsuites were hitting this case and we were losing
    logging statements.
    alanmcgovern committed May 15, 2015
Commits on May 7, 2015
Commits on Feb 27, 2015
  1. [Touch.Server] Add a -startup-timeout, to specify how long to wait fo…

    …r the app to connect back.
    The iOS 8+ simulators will kill the app if it takes longer than ~20
    to launch, but the simulator will simply kill the app, and not produce
    any other diagnostic.
    This allows us to produce crash reports when the app doesn't start
    up in an orderly manner.
    rolfbjarne committed Feb 27, 2015
  2. [Touch.Server] Print output from mtouch as soon as we get it.

    Just a simple trick to make the test run look faster.
    rolfbjarne committed Feb 27, 2015
  3. [Touch.Server] Refactor app killing to a separate method, and store t…

    …imeout using a TimeSpan.
    rolfbjarne committed Feb 27, 2015
Commits on Dec 15, 2014
Commits on Nov 5, 2014
Commits on Jun 24, 2014
Commits on Jun 13, 2014
  1. Always cancel the listener when mtouch exits for simulator runs.

    mtouch will only return a failure code for simulator runs if the
    app couldn't be launched in the simulator, in particular mtouch
    will always return 0 for any exit code from the app.
    This means that once mtouch exits (for simulator runs), we know
    the app isn't executing anymore, and we can cancel the tcp
    Previously we could wait forever if the app just exits from the
    Main method (we'd wait forever for a connection), or if the app
    crashes during startup.
    Just in case it was a clean and expected exit from the app, and
    for some reason the tcp listener hasn't finished yet, we wait
    a second before cancelling the server. This should flush everything.
    rolfbjarne committed Jun 13, 2014
  2. Add a timestamp to the mtouch output.

    Makes it easy to see if a test run finished immediately (indicator
    of a startup crash for instance), or something else happened later.
    rolfbjarne committed Jun 13, 2014
Commits on Jun 12, 2014
Commits on May 16, 2014
Commits on May 15, 2014
Commits on Mar 14, 2014
Commits on Mar 11, 2014
  1. [Runner] Log the GC (Boehm or sgen) and options (e.g. NewRefCount) be…

    …ing used while executing the tests
    committed Mar 5, 2014
Commits on Dec 10, 2013
  1. Mark test with [Timout] attribute as it will make it execute on a sep…

    …arate thread (not the main UI thread favored by Touch.Unit)
    committed Dec 10, 2013
Commits on Nov 29, 2013
  1. Make sure we don't call 'uniqueIdentifier' unless checking if the obj…

    …ect in question actually supports that selector.
    rolfbjarne committed with Nov 22, 2013
  2. Fixlet for 64bits support

    committed Nov 26, 2013
Commits on Nov 25, 2013
Commits on Nov 11, 2013
  1. Fix 64 bits build

    committed Nov 11, 2013
Commits on Oct 6, 2013
  1. Update runner for NUnitLite 1.0.0. Main run loop is now done from the…

    … ThreadPool and tests are executed in the main UI thread (like before). That makes less likely to 'freeze' in async testing [#15104]
    committed Oct 6, 2013
Commits on Oct 5, 2013
  1. Ensure [TestFixture[Setup|TearDown]] are called when running an indiv…

    …idual test case [#13390]
    committed Oct 5, 2013
Commits on Oct 4, 2013
  1. Merge pull request #15 from garuma/master

    Fix previous commit
    committed Oct 4, 2013
Commits on Oct 3, 2013
  1. Fix previous commit

    garuma committed Oct 3, 2013
  2. Merge pull request #14 from garuma/master

    Capture and log output from mtouch invocation.
    committed Oct 3, 2013
Commits on Sep 1, 2013
  1. Make it a one line change (in AppDelegate) to support NUnit[2-3] XML …

    …reporting (or any custom one based on OutputWriter), either local (on console) or remote (socket server). Similar in result to pull-request #13 by Clancey
    committed Sep 1, 2013
Commits on Aug 31, 2013
  1. Fix the defines to use CLR_4_0 (no just on Release) and NET_4_5 (on b…

    …oth Debug and Release). Fix running async tests [#14312] (based on pull request from Clancey)
    committed Aug 31, 2013
Commits on Jul 9, 2013
  1. Fix duration scale. There's no more need to multiply to 1000 since we…

    …'re using TotalMilliseconds, not a double value
    committed Jul 9, 2013