Switch branches/tags
Commits on Feb 21, 2013
  1. Bump for 1.7.3

    enebo committed Feb 21, 2013
Commits on Feb 20, 2013
Commits on Feb 19, 2013
  1. Prevent failure in MRI 1.9 to run this and also to actually extract p…

    enebo committed Feb 19, 2013
    …aragraph (formatting must have broken this at some point) -- I think JRuby does not barf because our default encoding is UTF-8 and not US-ASCII
  2. * Updated jars for Eclipse

    donv committed Feb 19, 2013
Commits on Feb 18, 2013
  1. Support marshaling and unmarshaling strings with repeated encodings

    shepmaster authored and enebo committed Jan 20, 2013
    When an encoding other than UTF-8 or US-ASCII is used, the encoding is
    marshaled as a String, which means that the value can be reused via
    the `LINK` type.
  2. Merge pull request #537 from bruceadams/add-epl-text

    headius committed Feb 18, 2013
    Add the complete text of the Eclipse Public License
Commits on Feb 17, 2013
Commits on Feb 15, 2013
  1. Hack jruby-rake-plugin to work on a windows machine

    Syver Enstad authored and BanzaiMan committed Nov 21, 2012
    classpath.erb: escape backslashes for classpath entries
    dump_classpath.erb: Workaround bug in Pathname#relative_path_from
    Signed-off-by: Hiro Asari <>
  2. Merge pull request #536 from sluukkonen/fix-failing-spec

    BanzaiMan committed Feb 15, 2013
    Fix occasionally failing spec.
Commits on Feb 14, 2013
  1. Fix occasionally failing spec.

    sluukkonen committed Feb 14, 2013 measures time with millisecond precision, so the spec will 
    fail if - start is less than 0.501 seconds.
Commits on Feb 13, 2013
  1. Merge pull request #533 from bruceadams/j9-patch3

    BanzaiMan committed Feb 13, 2013
    Default indy off for early IBM Java 7 releases #477
  2. Update stdlib to 1.9.3p385.

    headius committed Feb 13, 2013
  3. Default indy off for early IBM Java 7 releases #477

    bruceadams committed Feb 13, 2013
    Use the date stamp embedded in IBM Java's java.runtime.version
    for deciding if a known crash issue is present.
  4. Update json to 1.7.7.

    headius committed Feb 13, 2013
Commits on Feb 12, 2013
  1. Switch JRuby's CPL license to EPL.

    headius committed Feb 12, 2013
    Governance of the CPL license was moved to the Eclipse Foundation
    at some point in the past, and Eclipse blessed the EPL (Eclipse
    Public License v1.0) as the successor to CPL. Because the CPL
    contains a clause allowing transparent upgrades to future blessed
    successor licenses, we can make the move to EPL without
    re-contacting past contributors.
    The change in going from CPL to EPL largely just removes some
    punitive language that apparently made many companies nervous. All
    other language in the license remains identical.
    Several documents about Eclipse's move from CPL to EPL are here:
    Note that Eclipse's move came before the blessing of EPL as the
    upgrade path for CPL, so they had to do a bit more legal juggling.
    With Eclipse now the administrator of CPL and with their blessing
    of EPL as the upgrade path for CPL, all we need to do is make the
Commits on Feb 11, 2013
  1. Fixes for JRUBY-7058

    yokolet committed Feb 11, 2013
Commits on Feb 8, 2013
  1. Use a PhantomReference for releasing native FFI memory, to ensure the…

    Wayne Meissner
    Wayne Meissner committed Feb 8, 2013
    …y don't get run until after the object has been finalized and is guaranteed to be no longer reachable.
  2. Fixed up whitespace

    Bob Beaty authored and BanzaiMan committed Nov 19, 2012
    I just noticed that it's all spaces, and no tabs. My bad.
  3. Made changes as per Charles' request on exceptions

    Bob Beaty authored and BanzaiMan committed Nov 19, 2012
    Charles asked me to change the code to NOT generate a second BindException, but
    to change the newErrnoEADDRFromBindException() method, so I did that and just
    make a second, more general form of that method that took a context message
    string that would be put at the end of the excpetion's message. The old version
    doesn't have to worry about this, but in these two cases, we can put a lot more
    information at the user's fingertips.
  4. Added detailed host:port to Bind exceptions

    Bob Beaty authored and BanzaiMan committed Nov 8, 2012
    We really need to know what's failing, and the exceptions just aren't telling
    us - but in the context of the call we know EXACTLY what it's trying to do.
  5. Forced all sockets to use SO_REUSEADDR

    Bob Beaty authored and BanzaiMan committed Nov 8, 2012
    If you have to have a default for ALL sockets, the safer assumption is that
    there's going to be a lot more activity than less, so let's make ALL the
    sockets set-up with the SO_REUSEADDR set to 'true'. Ideally, we will make
    this a system define check, and maybe even an option, but those can come later
    after we've proven this helps things.
Commits on Feb 7, 2013
  1. Fix #529.

    BanzaiMan committed Feb 7, 2013
    Here, elt() needs an offset for the effective array we are considering. This was not a problem since 'begin' is often 0.
Commits on Feb 6, 2013
  1. Merge pull request #528 from igorbonadio/mvm

    BanzaiMan committed Feb 6, 2013
    fix samples/mvm.rb
Commits on Feb 5, 2013
  1. fix samples/mvm.rb

    igorbonadio committed Feb 5, 2013
  2. Unused imports.

    BanzaiMan committed Feb 5, 2013
Commits on Feb 4, 2013
  1. isARG() noticed

    enebo committed Jan 29, 2013
  2. That didn't quite work, so we'll comment out these examples to assess…

    BanzaiMan committed Feb 4, 2013
    … the fitness of other specs.
Commits on Feb 3, 2013
  1. Revert "Putting this set to pending to examine what other JI specs fa…

    BanzaiMan committed Feb 3, 2013
    …il on Travis"
    This reverts commit a85f52b.