Tags: * link all versions in openreview * for sticky meta/reviews (keep old reviews even when reviewers are changed) * for partial tracks * track ethical breaches in the infrastructure (for both authors and reviewers)
"While I agree that asking new reviewers should be allowed, previous reviews should not be forgotten.
Otherwise, it could happen that the same authors send the same paper multiple times, without changing
anything, until they find a favorable reviewer. I think that having a ""history"" of resubmission would
filter out this kind of behavior. At some point during the decision process, all the reviews ever received
should be shown, not only the most recent ones. I don't know when this would be better, maybe they should
be visible to the meta-reviewers, maybe only to the area chair, I don't know. For what concern the tracks,
I think they should be flexible. If a work belongs to multiple tracks, it should be allowed to indicate
both. If authors feel their work does not belong to any track, they should be allowed to choose none,
and rather specify keywords or similar. I would recommend the introduction (if it does not already exist)
of ""warnings"" or some other type of measure, that would be visible to the AC or senior AC in case
an author or a reviewer commits a SERIOUS violation of research ethics. For example, meta-reviewers
that write reviews based only on their own opinion rather than on the opinion of the reviewers. Such
a warning would remain associated with the person for a reasonable amount of time, and it will not be
used as a sort of ban or exclusion. But it should be used to ask for additional attention in the review
process."