Tags: * link meta/reviews to acceptance * against 8 week cycle * make conference PC honorary role
If an RR system continues to be used, reviews and acceptance/rejection decisions should **not** be decoupled.
Actually RR should make decisions. If the decision is acceptance then the *ACL conference **must** accept
them for presentation first-come, first-served based on tracks. For example, if the RR MT track has
100 papers accepted in its back-log the next conference, say ACL fill its quota of 70 MT papers from
the RR MT accepted pool and the next 30 go to the next conference (e.g. EACL) and so one. This pretty
much eliminates program chairs but it's more fair. PCs could still exist for honorary purposes. In case
of RR rejection, authors could resubmit after the next cycle by keeping or changing reviewers. 8-week
cycles is still too much and introduces big admin burden. There should be a maximum 4 deadlines a year
for a RR system to work well.