datalad 0.14.0rc1 breaks test_orc_datalad_pair_new_submodule

[yarikoptic]

discovered while testing #567

copy/paste of output for test_orc_datalad_pair_new_submodule from raw log

�[32mINFO    �[0m datalad.gitrepo:log.py:323 Finished writing objects
�[31m�[1m_____________________ test_orc_datalad_pair_new_submodule ______________________�[0m

job_spec = {'_command_array': ["sh -c 'echo a >sub/a'"], '_extra_inputs_array': [[]], '_inputs_array': [[]], '_outputs_array': [[]], ...}
dataset = Dataset('/tmp/pytest-of-travis/pytest-0/dataset33')
shell = Shell(name='localshell', id=None, type='shell', status='available')

    �[94mdef�[39;49;00m �[92mtest_orc_datalad_pair_new_submodule�[39;49;00m(job_spec, dataset, shell):
        �[94mwith�[39;49;00m chpwd(dataset.path):
            orc = orcs.DataladPairOrchestrator(
                shell, submission_type=�[33m"�[39;49;00m�[33mlocal�[39;49;00m�[33m"�[39;49;00m, job_spec=job_spec)
            orc.prepare_remote()
            orc.submit()
            orc.follow()
            orc.fetch()
    
            �[90m# prepare_remote() doesn't fail when a new subdataset is added after�[39;49;00m
            �[90m# the first run.�[39;49;00m
            sub = dataset.create(�[33m"�[39;49;00m�[33msub�[39;49;00m�[33m"�[39;49;00m)
            dataset.save()
    
            job_spec[�[33m"�[39;49;00m�[33m_resolved_command_str�[39;49;00m�[33m"�[39;49;00m] = �[33m"�[39;49;00m�[33msh -c �[39;49;00m�[33m'�[39;49;00m�[33mecho a >sub/a�[39;49;00m�[33m'�[39;49;00m�[33m"�[39;49;00m
            job_spec[�[33m"�[39;49;00m�[33minputs�[39;49;00m�[33m"�[39;49;00m] = []
            job_spec[�[33m"�[39;49;00m�[33moutputs�[39;49;00m�[33m"�[39;49;00m] = []
    
            orc = orcs.DataladPairOrchestrator(
                shell, submission_type=�[33m"�[39;49;00m�[33mlocal�[39;49;00m�[33m"�[39;49;00m, job_spec=job_spec)
>           orc.prepare_remote()

�[1m�[31mreproman/support/jobs/tests/test_orchestrators.py�[0m:867: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
�[1m�[31mreproman/support/jobs/orchestrators.py�[0m:830: in prepare_remote
    �[96mself�[39;49;00m._fix_up_dataset()
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

self = <reproman.support.jobs.orchestrators.DataladPairOrchestrator object at 0x7f2d847dd7f0>

    �[94mdef�[39;49;00m �[92m_fix_up_dataset�[39;49;00m(�[96mself�[39;49;00m):
        �[33m"""Try to get datataset and subdatasets into the correct state.�[39;49;00m
    �[33m    """�[39;49;00m
        �[96mself�[39;49;00m._checkout_target()
        �[90m# fixup 0: 'datalad create-sibling --recursive' leaves the subdataset�[39;49;00m
        �[90m# uninitialized (see DataLad's 78e00dcd2).�[39;49;00m
        �[96mself�[39;49;00m._execute_in_wdir([�[33m"�[39;49;00m�[33mgit�[39;49;00m�[33m"�[39;49;00m, �[33m"�[39;49;00m�[33msubmodule�[39;49;00m�[33m"�[39;49;00m, �[33m"�[39;49;00m�[33mupdate�[39;49;00m�[33m"�[39;49;00m, �[33m"�[39;49;00m�[33m--init�[39;49;00m�[33m"�[39;49;00m])
    
        �[90m# fixup 1: Check out target commit in subdatasets. This should later be�[39;49;00m
        �[90m# replaced by the planned Datalad functionality to get an entire�[39;49;00m
        �[90m# dataset hierarchy to a recorded state.�[39;49;00m
        �[90m#�[39;49;00m
        �[90m# fixup 2: Autoenable remotes:�[39;49;00m
        �[90m# 'datalad publish' does not autoenable remotes, and 'datalad�[39;49;00m
        �[90m# create-sibling' calls 'git annex init' too early to trigger�[39;49;00m
        �[90m# autoenabling. Temporarily work around this issue, though this�[39;49;00m
        �[90m# should very likely be addressed in DataLad. And if this is here�[39;49;00m
        �[90m# to stay, we should avoid this call for non-annex datasets.�[39;49;00m
        lgr.info(�[33m"�[39;49;00m�[33mAdjusting state of remote dataset�[39;49;00m�[33m"�[39;49;00m)
        �[96mself�[39;49;00m._execute_in_wdir([�[33m"�[39;49;00m�[33mgit�[39;49;00m�[33m"�[39;49;00m, �[33m"�[39;49;00m�[33mannex�[39;49;00m�[33m"�[39;49;00m, �[33m"�[39;49;00m�[33minit�[39;49;00m�[33m"�[39;49;00m])
        �[94mfor�[39;49;00m res �[95min�[39;49;00m �[96mself�[39;49;00m._execute_datalad_json_command(
                [�[33m"�[39;49;00m�[33msubdatasets�[39;49;00m�[33m"�[39;49;00m, �[33m"�[39;49;00m�[33m--fulfilled=true�[39;49;00m�[33m"�[39;49;00m, �[33m"�[39;49;00m�[33m--recursive�[39;49;00m�[33m"�[39;49;00m]):
            cwd = res[�[33m"�[39;49;00m�[33mpath�[39;49;00m�[33m"�[39;49;00m]
            �[96mself�[39;49;00m._assert_clean_repo(cwd=cwd)
            lgr.debug(�[33m"�[39;49;00m�[33mAdjusting state of �[39;49;00m�[33m%s�[39;49;00m�[33m"�[39;49;00m, cwd)
>           cmds = [[�[33m"�[39;49;00m�[33mgit�[39;49;00m�[33m"�[39;49;00m, �[33m"�[39;49;00m�[33mcheckout�[39;49;00m�[33m"�[39;49;00m, res[�[33m"�[39;49;00m�[33mrevision�[39;49;00m�[33m"�[39;49;00m]],
                    [�[33m"�[39;49;00m�[33mgit�[39;49;00m�[33m"�[39;49;00m, �[33m"�[39;49;00m�[33mannex�[39;49;00m�[33m"�[39;49;00m, �[33m"�[39;49;00m�[33minit�[39;49;00m�[33m"�[39;49;00m]]
�[1m�[31mE           KeyError: 'revision'�[0m

�[1m�[31mreproman/support/jobs/orchestrators.py�[0m:754: KeyError

we installed rc:

Collecting datalad>=0.13.0rc2
  Downloading datalad-0.14.0rc1.tar.gz (1.6 MB)

It also failed similarly in recent master run but was all green a few days back (whenever there were no datalad rc) https://travis-ci.org/github/ReproNim/reproman/builds/755466822 . So -- must be datalad rc related .