- compatible with all recent Linux kernels
- no need for kernel rebuilding/patching
- verified (to what degree?) compatibility with Python (what versions?)
- ...
So, the topics/examples/use cases we could give to Python programmers
- Generic Condor integration (conditioned on availability)
** + jobs migration * "take home" or postponed computation * reversible debugging ** dmtcp-snapshot your Python debugger prior critical point to resurrect/share debugging section if/when needed
Specific for neuroscientists somehow?
- Package/expose Python bindings
- Verify operation on snapshotting common analysis pipelines (pymvpa, nipype (+fsl, +matlab/spm, etc))