You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Segment Index that is stored in FossilDB should include additionalCoordinates. We should consider adding these to the key as well (just like the mag) or design a more elaborate lookup structure, depending on the queries we might get.
My guess is that the typical query is for a single spot along the additionalCoordinates axes.
The text was updated successfully, but these errors were encountered:
@frcroth fyi, we added you as assignee, read this as a “whoever of us is out of issues first”. The other issues, especially #6740, #7243 and #7275 are higher priority (and should hopefully require less pending discussion)
Currently the segment index maps from a segment id to a list of buckets(xyz).
To adapt this for ND we could make the value a list of buckets (xyz+additionalCoordinates). However, I’m not sure that can answer all relevant queries.
We could also include the additional coordinates in the key instead, so it would be (segment id + additional coordinates) → list of buckets (xyz)
That would support getting the full list for a particular additional-coordinates point (e.g. time-point).
However, on top of that, we might also implement an index for segment id to list of additional coordinates, where this segment exists at all.
This way, we could also answer queries about in which time-points the segment exists.
Segment Index that is stored in FossilDB should include additionalCoordinates. We should consider adding these to the key as well (just like the mag) or design a more elaborate lookup structure, depending on the queries we might get.
My guess is that the typical query is for a single spot along the additionalCoordinates axes.
The text was updated successfully, but these errors were encountered: