99xxx is what we have been assigning to a new satellite in SatNOGS before it has a proper NORAD ID or we otherwise identify it as one of the tracked OBJECTs after launch. It will be changed to one of the 473xx IDs in SatNOGS once identified.

Its a hack for a problem not quite anticipated when the initial apps were designed to use NORAD ID as the primary identifier: being that there's a lag between launch and identification whereas we want to use the information and schedule observations right away. It is also a hack that does not scale as we will trip over ourselves when NORAD IDs hit the 99000's.

From a data perspective there are some other challenges in normalizing on NORAD as a primary identifier in SatNOGS as well, as the association is with a singular object in orbit where we'd like to broaden the information to include pre-launch objects, and organizing data around individual components or satellites that start off as one only becomes feasible once they've separated (ie: prox-1/lightsail-2) and we risk certain items of historical data being lost when combined (all things ISS). So with that we've tossed around the idea of Yet Another Identifier that we'll likely implement early this year. Forum thread with the details, ideas, and gitlab issue is here (and we welcome input from others): https://community.libre.space/t/unique-satellite-identifier-definition-in-satnogs/2970/1  The intent is not to create another identifier to confuse the world with, but internally we need to shift away from NORAD as the primary identifier. If it is of any benefit to use externally, that's welcome, but not the driver.

:)