List ingress pagination #1383
List ingress pagination #1383
Conversation
|
milos-lk
force-pushed
the
list-ingress-pagination
branch
from
fe047a0 to
54c5988
Compare
protobufs/livekit_models.proto
Outdated
| string token = 1; | ||
| } | ||
|
|
||
| message CursorPagination { |
There was a problem hiding this comment.
essentially same to TokenPagination wrt state it holds - but wanted to add it as a separate type to have clean separation on methods for encoding/decoding
There was a problem hiding this comment.
Considering we already have an anonymous "TokenPagination" type we can decode to anything we want depending on context, do we really need to expose this mew type at the API level?
There was a problem hiding this comment.
existing token pagination isn't fully anonymous -
Line 228 in edf9ff2
my intention with having a separate type was to prevent the misuse of the api, but maybe we don't need to go that far - still not sure - maybe we could use that TokenPagination and have the new set of encoder/decoder APIs with it..
There was a problem hiding this comment.
maybe reusing the TokenPagination + new codecs is actually neater - I think I will update it
There was a problem hiding this comment.
Not sure what's best: overloading the protocol with 3 pagination schemes or requiring implementers (ie LiveKit engineers) to select the right decoding format. Curious to see what other folks think.
There was a problem hiding this comment.
we should standardize on this encoding... sip implements this pattern incorrectly
Adding CursorPagination - to allow for cursor based list APIs where ordering key could be set to an arbitrary column value (e.g created_at) and tie breaker column (unique) specified to allow for stable result sets.
Adding utilities to encode/decode the token.