There were following score actions:
NAMES_MAP = { NEW: 'TA', EDITED: 'TE', EDITED_OWN: 'TX', DELETED: 'TD', REVIEWED: 'R', EDIT_PENALTY: 'XE', REVIEW_PENALTY: 'XR', MARKED_FUZZY: 'TF', SUGG_ADDED: 'S', SUGG_ACCEPTED: 'SA', SUGG_REJECTED: 'SR', SUGG_REVIEWED_ACCEPTED: 'RA', SUGG_REVIEWED_REJECTED: 'RR', }
I wonder what wordcount stats we need to store separately to allow scoring to be clear and understandable.
TA - translated wordcount (state change 0 -> 200)
TA
TE - edited wordcount (target change without state)
TE
S - suggestion added wordcount
S
R
There was an error while loading. Please reload this page.