Change Details

In the Hungarian Wikipedia, flagged revisions are theoretically used as a counter-vandalism tool (ie. edits should be marked as sighted if they do not contain obvious vandalism, unargued removal of facts, uncited inclusion of facts, formatting errors etc; specifically, edits should be marked as sighted even if the patroller does not have the means or expertise to verify them). In practice that does work well because of lack of capacity and because patrollers don't always follow that rule, so good edits get stuck in a pending state for a long time. If ORES proves accurate in identifying "suspicious" edits, it would be great to use it as input for flagged revisions; ie. only put edits in a pending state if ORES labels them as damaging.

In the Hungarian Wikipedia, flagged revisions are theoretically used as a counter-vandalism tool (ie. edits should be marked as sighted if they do not contain obvious vandalism, unargued removal of facts, uncited inclusion of facts, formatting errors etc; specifically, edits should be marked as sighted even if the patroller does not have the means or expertise to verify them). In practice that does work well because of lack of capacity and because patrollers don't always follow that rule, so good edits get stuck in a pending state for a long time. If ORES proves accurate in identifying "suspicious" edits, it would be great to use it as input for flagged revisions; ie. only put edits in a pending state if ORES labels them as damaging. Potential poor man's solution: {T57081}

In the Hungarian Wikipedia, flagged revisions are theoretically used as a counter-vandalism tool (ie. edits should be marked as sighted if they do not contain obvious vandalism, unargued removal of facts, uncited inclusion of facts, formatting errors etc; specifically, edits should be marked as sighted even if the patroller does not have the means or expertise to verify them). In practice that does work well because of lack of capacity and because patrollers don't always follow that rule, so good edits get stuck in a pending state for a long time. If ORES proves accurate in identifying "suspicious" edits, it would be great to use it as input for flagged revisions; ie. only put edits in a pending state if ORES labels them as damaging. Potential poor man's solution: {T57081}