Closed GeorgDangl closed 2 years ago
I suggest adding a UTC timestamp query parameter since
that would request the server to only return versions that were created on or after that timestamp. This should be easy for clients since all they need to do is provide the last polling time and would, I believe, reduce implementation complexity for the servers.
Did you deliberately choose a flat list of versions rather than grouping the versions of each document together?
Did you deliberately choose a flat list of versions rather than grouping the versions of each document together?
Yeah, that was on purpose. My reasoning was to just return a flat list, containing the latest version for each queried document. We could also nest it deeper by document_id
, but I thought the flat list fits well here.
I suggest adding a UTC timestamp query parameter
since
that would request the server to only return versions that were created on or after that timestamp. This should be easy for clients since all they need to do is provide the last polling time and would, I believe, reduce implementation complexity for the servers.
Sounds good, added with 2c7c3c71326c493cf84a37d858426414321b7588
We've made some more changes during the call today, but we're not finished yet. We will look at it again next Monday😀
Hey guys,
as discussed, here's a first idea how the bulk query endpoint could be added to the API. Looking forwards to your input😀
CC @pasi-paasiala @ykulbak @jasollien