- include input file name (and chunk? and CDX?) in sentry context
- play with test image on older releases (eg, trusty)

- how to get argument (like --hbase-table) into mrjob.conf, or similar?
- fix pig gitlab-ci tests (JAVA_HOME). also make fetch_deps *way* more quiet
- sentry: https://github.com/getsentry/raven-python

potential helpers:
- https://github.com/martinblech/xmltodict
- https://github.com/trananhkma/fucking-awesome-python#text-processing
- https://github.com/blaze/blaze (for catalog/analytics)
- validation: https://github.com/pyeve/cerberus
- testing (to replace nose):
    - https://github.com/CleanCut/green
    - pytest
    - mamba ("behavior driven")