request: per-round review forms & multiple per-reviewer, per-submission reviews about hotcrp

[kohler]

Hotcrp already supports per-round review forms: fields can be present or absent depending on review round.

But it currently supports one review per user per paper. For that reason the workflow you mention may be better implemented by (1) different submission classes, so a single “submission” has multiple submission numbers (#1 extended abstract becomes, say, #139 full submission and #172 major revision), all with the same reviewer set, or (2) a different submission system. Who knows maybe Not-Actually-Open Review.

In hotcrp many subsystem semantics, such as formulas, review visibility, etc. depend on the implicit constraint of at most one review per user. A way to make multiple reviews per user per paper happen sooner would be to contribute specific ideas about how to resolve these semantic constraints. Or even a prototype implementation. HotCRP's code is right here in this repository.

[dants]

Thanks for the suggestions and for pointing out the ability to make fields absent and present per round!

Unfortunately, for me, the drawbacks of your suggestions -- spreading the state of a single submission (1) across multiple submissions within the same hotcrp instance, or (2) across multiple submissions in different hotcrp instances -- outweigh the benefits, creating IMO a bigger problem than the lack of the requested functionality. So I prefer to get by with a single submission in a single instance and pay the cost.

I of course understand that the effort required to implement the proposed feature is significant (understatement), as it is incompatible with the current design, seemingly requiring a full overhaul. Understandably, the associated overheads might make this task unfeasible for the hotcrp maintainer. For this reason, however, I doubt that anyone else can realistically pull it off.

Thanks again,
--Dan