Permalink
Commits on May 27, 2016
  1. @timpokorny
  2. @timpokorny

    Minor formatting updates to previous PR and removal of unused imports.

    No functionality changes.
    timpokorny committed May 27, 2016
  3. @DoctorRuss @timpokorny

    Fix for NPE in ModelMerger (#182)

    Fixed NPE in ModelMerger that can occur when FOM fragments do not contain all base scaffolding.
    
    hla1516e.fomparser.FOM now ensures that icRoot and ocRoot are not null by creating scaffolding HLAinteractionRoot or HLAobjectRoot metadata respectively.
    
    Reference: #181
    DoctorRuss committed with timpokorny May 27, 2016
Commits on May 19, 2016
  1. @timpokorny
  2. @anthonycramp @timpokorny

    Implementation of HLA1516eVariantRecord (#178)

    * Added HLAvariantRecord encoding and decoding
    
    Added IEEE 1516e HLA Variant Record encoding and decoding. Unit tests
    (inspired by the tests for HLA1516eFixedRecord) are included and pass.
    
    Required implementation of hashCode and equals on HLA1516eOctet. It
    will be necessary to implement hashCode and equals on any type that
    will be used as a variant record discriminant.
    
    * Added and removed some commentary
    
    * Commentary and small code changes
    
    Mainly a commentary commit. Identified logic errors in getOctetBoundary
    (discriminant octet boundary is not considered anymore) and decode
    (wasn’t handling the case where there was no variant for the decoded
    discriminant).
    
    * Deleted errant import auto inserted by my IDE.
    anthonycramp committed with timpokorny May 19, 2016
Commits on May 15, 2016
  1. @timpokorny

    Fix problem with HLAprivilegeToDelete not getting added to FOM.

    Due to recent change in MOM processing, the MIM is no longer loaded from
    disk, but rather built in code. This lets us give specific handles,
    which allows us to look up attributes quickly without using their name
    (and in turn, allows us to do things in a HLA v1.3 & 1516e compliant
    way).
    
    This commit fixes the missing attribute problem by adding it on
    validation if it is missing.
    
    Working: #175
    timpokorny committed May 15, 2016
Commits on May 4, 2016
  1. @timpokorny
Commits on May 2, 2016
  1. @timpokorny
Commits on May 1, 2016
  1. @timpokorny
  2. @timpokorny
Commits on Apr 30, 2016
  1. @timpokorny

    Updates to the various README files.

    Fix: #50
    timpokorny committed Apr 30, 2016
  2. @timpokorny
Commits on Apr 29, 2016
  1. @timpokorny
Commits on Apr 25, 2016
  1. @timpokorny

    Fixing (i hope) problem with hanging test

    testConnectWhenAlreadyConnected() appears to be hanging intermittently
    on Travis builds. This never happens for me locally and getting logs out
    of that environment can be hard, especially as the problem is
    intermittent (maybe 1/10 builds). Shuffling some things around to
    hopefully cause the condition this test is confirming to execute
    earlier, before anything has the chance to hang. If the tests starts
    failing on local builds a better investigation is needed.
    timpokorny committed Apr 25, 2016
Commits on Apr 24, 2016
  1. @timpokorny
  2. @timpokorny

    Re-enable federate crash detection after move to in-app GMS.

    Ever since we moved to in-application group management services (rather
    than relying on JGroups), the failure/crash detection for federates has
    been broken. The messages were coming through JGroups, but we could
    never action them since we were not using those services and had an
    additional GMS layer on top. This has now been fixed, and JGroups
    crashes are passed up to our custom GMS layer.
    
    Fix: #162
    timpokorny committed Apr 24, 2016
Commits on Apr 23, 2016
  1. @timpokorny

    Fix problem with rapid exchange of attribute ownership.

    Previously, if a federate release attributes and then requested them
    back before they received an acquisition notification for some other
    federate, the request wouldn't store properly inside the Ownership
    Manager. This has now been fixed.
    
    Fix: #166
    timpokorny committed Mar 2, 2016
Commits on Feb 9, 2016
  1. @timpokorny
  2. @timpokorny

    Fix example C++ 1.3 & 1516e federates on Mac OSX and ensure a full

    sandbox (including JRE) is built any time we specify the generic
    "sandbox" target as a build argument
    timpokorny committed Feb 9, 2016
Commits on Feb 8, 2016
  1. @timpokorny

    Fix compilation of C++ core and example federates on the Mac and some

    potentially crashy code relating to the fetching of env vars that may
    not exist
    timpokorny committed Feb 8, 2016
Commits on Feb 6, 2016
  1. @timpokorny
Commits on Jan 26, 2016
  1. @timpokorny
  2. @timpokorny
  3. @timpokorny
Commits on Jan 24, 2016
  1. @timpokorny

    Add initial support for MOM to 1516e

    Many of the values are not currently supported, but the basics are now
    complete for 1516e MOM support, including the proper serialization of
    those attribtutes of `HLAfederate` and `HLAfederation` that are
    supported.
    
    New test additions have been included, as have refactorings of the major
    MOM data providers to deliver values in format that is properly encoded
    depending on the version of the interface being used. This has also
    triggered some changes through the 1516-2000 tests to match this.
    
    Fix: #55
    timpokorny committed Jan 24, 2016
  2. @timpokorny

    Add partial MOM implementation for 1516e.

    Name resolution now works, and updates are generated in response to
    `RequestAttributeUpdate` and `RequestClassAttributeUpdate` requests.
    
    Basic MOM support added for 1516e. The MOM hierarchy is now manually
    added to a federation so that handles can be controlled, and names are
    resolved properly. Federates can request updates and they will receive
    the responses, but currently the formatting of the attribute reflections
    is not correct (just follows the "everything as a string" approach).
    
    Updates to the tests and implementation needed to ensure that the
    encoding is correct.
    
    Working: #55
    timpokorny committed Jan 19, 2016
Commits on Jan 17, 2016
  1. @timpokorny
  2. @timpokorny
Commits on Jan 16, 2016
  1. @timpokorny

    Accept order & transport types specified in attributes for 1516e FOMs

    Previously <interactionClass> and <attributeClass> elements required
    <order> and <transportation> sub-elements to be considered valid. Now
    these can be:
    
      - Specified as sub-elements
      - Specified as attributes on the actual element
      - Omitted entirely, falling back to defaults
    
    Defaults are `TimeStamp` (for order) and `Reliable` for transport.
    
    Fix: #163
    timpokorny committed Jan 16, 2016
Commits on Dec 30, 2015
  1. @timpokorny
  2. @timpokorny
  3. @timpokorny
  4. @timpokorny
Commits on Dec 23, 2015
  1. @timpokorny

    Let user specify WanRouter IP addr with symbolic names

    In addition to an IP Address or DNS name, a user can now specify one of
    the following symbols on the command line when defining the address the
    WAN Router should communicate on:
    
      - LOOPBACK   : 127.0.0.1
      - LINK_LOCAL : 169.254.1.0 through 169.254.254.255
      - SITE_LOCAL : 10.*.*.*, 192.168.*.*, 172.16.*.* through 172.31.*.*
      - GLOBAL     : Any globally routable address (none of the above)
    timpokorny committed Dec 23, 2015
Commits on Dec 13, 2015
  1. @timpokorny

    Update the version number to 2.1.0.2 (still beta) and some random Ecl…

    …ipse C project setting update
    timpokorny committed Dec 13, 2015