ID	Type	Status	Priority	Milestone	Owner		Summary + Labels	...
	8	Enhancement	Accepted	Medium	----	fonnesbeck		Add multi-processor support
	18	Enhancement	Accepted	Medium	----	fonnesbeck		Prior updating
	32	Enhancement	Accepted	Medium	Post2.0	anand.prabhakar.patil		Gibbs samplers
	34	Enhancement	Started	High	Release2.0	fonnesbeck		Update docs for PyMC2
	55	Enhancement	Accepted	Medium	Release2.0	fonnesbeck		Implement GOF methods
	56	Enhancement	New	Low	Post2.0	anand.prabhakar.patil		Derivatives
	61	Enhancement	Started	Medium	Post2.0	anand.prabhakar.patil		Dirichlet processes
	62	Other	New	Medium	Release2.0	anand.prabhakar.patil		Release target?
	70	Enhancement	Accepted	Medium	Post2.0	anand.prabhakar.patil		Arbitrary objects in traces
	73	Enhancement	Accepted	Medium	----	fonnesbeck		Updating parameters in-place
	76	Defect	Started	Medium	----	anand.prabhakar.patil		sqlite database bug
	79	Enhancement	Accepted	Medium	----	fonnesbeck		raising exceptions on reserved words
	82	Enhancement	Accepted	Low	Release2.0	fonnesbeck		Summary statistics
	96	Defect	Accepted	Low	Release2.0	anand.prabhakar.patil		Huge numbers of warnings from gibbsit.f and math.f
	108	Other	Accepted	Low	Post2.0	anand.prabhakar.patil		Ravelling groups of stochastics
	112	Defect	Accepted	Low	----	fonnesbeck		Parameterization of autoregressive normal
	115	Enhancement	Accepted	Medium	Post2.0	anand.prabhakar.patil		Proposals from the prior for dataless submodels
	118	Enhancement	New	Low	Post2.0	anand.prabhakar.patil		On-the-fly deterministic subclassing
	122	Enhancement	New	Medium	----	fonnesbeck		Recursive value initialization
	127	Enhancement	Accepted	Medium	Post2.0	anand.prabhakar.patil		making 'p' an infinite generator in Categorical and Multinomial
	128	Defect	Accepted	Medium	Post2.0	anand.prabhakar.patil		Infinite IID sequences
	134	Defect	Accepted	Medium	----	fonnesbeck		Acceptance rate reporting error in binary variables
	137	Defect	Started	Medium	Release2.0	anand.prabhakar.patil		numpy 1.2 breaks testsuite
	138	Task	New	Low	Release2.0	anand.prabhakar.patil		PyPI
	141	Defect	Accepted	Medium	Release2.0	david.huard		Naming confusion
	145	Defect	Accepted	Medium	Release2.0	fonnesbeck		Valid initial values rejected from Poisson stochastic
	147	Enhancement	New	Low	Post2.0	anand.prabhakar.patil		'Time' flag in variables for profiling
	148	Other	New	Low	----	anand.prabhakar.patil		Should categorical be a discrete distribution?
	150	Enhancement	Accepted	----	Post2.0	fonnesbeck		Data cloning
	152	Enhancement	New	Low	Post2.0	anand.prabhakar.patil		'debug' option in isample
	154	Defect	Accepted	Medium	----	anand.prabhakar.patil		Attempting to import pymc from inside source tree should raise error
	155	Defect	Accepted	Medium	----	anand.prabhakar.patil		Sampling with thin > iter causes error with ram backend
	156	Defect	Accepted	Low	----	fonnesbeck		isampler does not indicate when it is finished
	159	Defect	Accepted	Medium	----	anand.prabhakar.patil		Fortran likelihoods need to check shapes of input arguments
	160	Defect	Assigned	Medium	Post2.0	fonnesbeck		Sampler should not re-initialize if trace is not empty
	161	Enhancement	New	Medium	Post2.0	anand.prabhakar.patil		Metropolis should be variable-length
	162	Enhancement	New	Low	Post2.0	anand.prabhakar.patil		Parsing R model formula syntax
	163	Defect	Accepted	Medium	Release2.0	anand.prabhakar.patil		rtruncnorm bug with integer arguments
	164	Enhancement	Accepted	Low	Post2.0	anand.prabhakar.patil		Log-sums