Final Pre-AMP Release

This is the final "2018 configuration" release. The next v0.5.x series of releases will introduce several new features regarding on-manifold products and multisession solving (in combination with Caesar.jl). This release has several bug fixes and minor feature modifications. See related milestone information here.

Maintenance Release

Several maintenance issues have been addressed in this release. See milestones for more details. Main purpose is:

• Release constraints to allow newest NLsolve v3.x,
  • Help resolve Pkg errors from recent JuliaRegistries/General implementation (likely automation mistake) towards stricter versioning policies.
• Mutable types for serialization of some factor types,

Bug Fix Release

A previous release introduced a hidden bug in the Julia 1.0 compliant solver code. This release corrects the bug as well many documentation additions.

Rapid Release for Upstream Breaking Changes

A rapid release to contain an NLsolve upstream problem.

See other milestone chances and enhancements here.

Also see discussion here:
#183

Experimental ZMQ interface Support

This release introduces experimental functions needed downstream for multi-language support to the mm-iSAM solver via Caesar.jl.

Experimental Fixed Lag Smoothing

This pre-release fixes a bug (needed downstream) that allows a first draft of fixed-lag smoothing operation.

Maintenance Release, Bug / Deprecation Fixes

Fixes bugs and improvements needed downstream -- see related work at milestones here.

Maintenance Release, Julia 0.7/1.0

Support for Julia 0.7 and 1.0. See relevant work here.

Final Julia 0.6 Release, many new features

The next release v0.4.0 will not be backwards compatible with Julia 0.6 ecosystem. This release captures the last of the work done on the 0.6 work cycle, which includes:

• fixed lag smoothing,
• improved jld file handing,
• label name handling,
• refinement of ThreadModel for single or multithreaded use,
• arbitrary number of multihypotesis associations (beyond original bimodality),
• defining SamplableBelief, LinearConditional, etc.,
• Introduction of the AliasingScalarSampler as a usable probability belief method,
• various maintenance and test improvements and bug fixes,
• addition of new fields in VariableNodeData and GenericFunctionNodeData, breaking jld formats relative to v0.3.8,
• access to more features of graphviz for visualization of graphs,
• added the fastnorm function,

Note at this point that PR #111 for Julia 0.7 is already passing Travis tests. Please see Milestones v0.3.9 for work related to this release.

Mutlithreaded Convolutions

This pre-release has a relatively small code change with few PRs, but represents a major change in the computations performed during approximate convolution. Each particle is now projected through a shared memory multithreaded loop -- and applies to both FunctorPairwise and FunctorPairwiseMinimize. This release also introduces a new constraint on factors, such that multiple threads do not conflict on any memory shared during residual computations. Please use Threads.nthreads() and Threads.threadid() for allocating separate memory space in the functor objects. The new keyword addFactor!(..., threadmodel=MultiThreaded / SingleThreaded) can be used for better error stack trace in SingleThreaded mode. Please set the number of shared memory threads in the .bashrc file with export JULIA_NUM_THREADS=4, or which ever amount you prefer. Lastly, also note that the shared memory threads are an addition on top of the existing multiprocess computations already used with the Bayes tree formulation.