Pydra

- refactoring the error handling by padra: improving raised errors, removing nodes from the workflow graph that can't be run
- refactoring of the `input_spec`: adapting better to the nipype interfaces
- switching from `pkg_resources.declare_namespace` to the `stdlib pkgutil.extend_path`
- moving readme to rst format

* Add ``pydra.tasks`` namespace package to enable separate packages of ``Task``s to be installed into ``pydra.tasks``.
* Raise error when task or workflow name conflicts with names of attributes, methods, or other tasks already added to workflow
* Mention ``requirements.txt`` in README

- removing the tutorial to a [separate repo](https://github.com/nipype/pydra-tutorial)
- adding windows tests to codecov
- accepting `None` as a valid output from a `FunctionTask`, also for function that returns multiple values
- fixing slurm error files
- adding `wf._connection` to `checksum`
- allowing for updates of `wf._connections`
- editing output, so it works with `numpy.arrays`
- removing `to_job` and pickling task instead (workers read the tasks and set the proper input, so the multile copies of the input are not kept in the memory)
- adding standalone function `load_and_run` that can load and run a task from a pickle file
- removing `create_pyscript` and simplifying the slurm worker
- improving error reports in errors flies
- fixing `make_class` so the `Output` is properly formatted

- fixing `hash_dir` function
- adding `get_available_cpus` to get the number of CPUs available to the current process or available on the system
- adding simple implementation for `BoshTask` that uses boutiques descriptor
- adding azure to CI
- fixing code for windows
- etelementry updates
- adding more verbose output for task `result` - returns values or indices for input fields
- adding an experimental implementation of Dask Worker (limited testing with ci)

- reorganization of the `State` class, fixing small issues with the class
- fixing some paths issues on windows os
- adding osx and window sto the travis runs (right now allowing for failures for windows)
- adding `PydraStateError` for exception in the `State` class
- small fixes to the hashing functions, adding more tests
- adding `hash_dir` to calculate hash for `Directory` type

- passing `wf.cache_locations` to the task
- using `rerun` from submitter to all task
- adding `test_rerun` and `propagate_rerun` for workflows
- fixing task with a full combiner
- adding `cont_dim` to specify dimensionality of the input variables (how much the input is nested)