My main goal in trying to include L1 DiaObjects is to handle the case where an alert goes out in year N and someone wants to find out what the updated light curve is in DRX some years later. I don't really have a feel for how we map from a VOEvent alert containing a DiaSource ID and a DiaObject ID to a DiaObject ID from a later data release.

An Alert packet has a L1 DiaSource ID, DiaObject ID and (if you are lucky) and up to 6 Object IDs from the previous data release. The DiaObject has attached to it a light curve that is hopefully due to one of those 6 Object IDs but might be a light curve from multiple objects that are close to each other or none of them (a supernova from an object that was too faint before). So how do I go about finding the light curve from the current release? Do I just do a spatial search in the DiaObject table for the current release? Is the assumption that L1 cross-matching is a moving target and we should stay away from it? Is it possible that the L1 DiaObject no longer exists (say we decided that it was a composite and there were two distinct objects with different light curves)?

If we are ignoring L1 cross-matching, is my draft requirement text okay?

If we are ignoring L1 cross-matching, is my draft requirement text okay?

Fine with me, but I don't really have much experience with the language of requirement specification.

I have added this as DMS-REQ-0350. Colin Slater would you have a quick look at the text? Since I assume you don't have EA access the new text is:

DMS-REQ-0350: Associating Objects across data releases

Specification: It shall be possible to associate an Object in one data release to the most likely match in the Object table from another data release. This shall be possible without the previous data releases being online.

Discussion: This could be done with a database table mapping every Object in one data release to the matched Object in every other data release.

Looks fine.

(Apologies. Hit the wrong button by mistake)