Date-based IDs for superevents

Based on notes from the March 2018 LVC meeting and discussions with Patrick and Deep, I've developed a strategy for implementing date-based IDs for superevents. An example is given below to demonstrate how this will work, but please ask any clarifying questions that are needed since it's a bit confusing to walk through.

Example

1. A superevent is identified for May 13, 2018. (The date is determined from the initial value of t_0 assigned to the superevent)
   • Initial date-based ID assigned: S180513
2. A second superevent is identified for May 13, 2018.
   • Initial date-based ID assigned: S180513b
   • A letter suffix is added to the first superevent: S180513 -> S180513a. Hyperlinks and API calls which use either S180513 or S180513a will both point to this superevent, but S180513a will now be displayed as the ID throughout GraceDB.
3. An alert is sent out about S180513b.
   • Its prefix is upgraded to "GWA", so its official ID is now GWA180513b. The letter suffix is inherited from its previous ID. Hyperlinks and API calls which use either ID (S180513b or GWA180513b) will both point to this superevent, but GWA180513b will be displayed as the ID throughout GraceDB.
4. GWA180513b is confirmed as a GW.
   • Its prefix is upgraded to "GW" and a new letter suffix is determined based on confirmed GWs for this date.
   • Resulting ID: GW180513. As before, hyperlinks and API calls using S180513b or GWA180513b will still work, but GW180513 will be generally displayed as the ID.
5. An alert is sent out about S180513a.
   • ID changed from S180513a -> GWA180513a, hyperlinks still work for both IDs (and the initial, suffixless ID), etc.
6. GWA180513a is confirmed as a GW.
   • ID changed: GWA180513a -> GW180513B. Hyperlinks and API calls still work for all past IDs, etc.
   • GW180513 has a letter suffix added: GW180513 -> GW180513A. Hyperlinks and API calls still work for all past IDs, etc.

Questions

1. Should the intermediate state (GWA prefix) exist or is a two-state arrangement sufficient?
2. If so, should the GWA state have its own letter prefix and not just inherit it from the S state?
3. Is the current strategy for handling the letter suffix for the first superevent on a given date acceptable (i.e., first superevent has no letter suffix until a second superevent is found for that date)? Or should the first superevent always have a letter suffix? Access to the superevent in GraceDB will be unaffected, but there could be confusion in certain cases. Example: offline searches find a second GW on a given date, which adds a letter suffix to the first GW's ID after a paper is already published about the first)
4. Same question as above for the first superevent confirmed as a GW on a given date.
5. Capital or lowercase letter suffixes for confirmed GWs?