davisp commented on issue #1085: Return update_seq and offset when update_seq 
is true and keys is set
URL: https://github.com/apache/couchdb/pull/1085#issuecomment-357793594
 
 
   Well, offset certainly doesn't make any sense. Though to match things we 
could make that return offset:null or similar.
   
   Secondly, this is gating the inclusion of these keys on whether 
update_seq=true is passed, it seems like if we're going to add them they should 
be included regardless.
   
   For the network overhead I'm not sure it'd be a huge deal? I guess we could 
add a more specific endpoint rather than shipping entire db info blobs but I'm 
not super concerned about it.
   
   Although, this update_seq is kind of a lie because we open each document 
individually so a) the update_seq may or may not reflect the shards used to 
open documents and b) it could change for each invidual doc open call depending 
on the combination of shards that responded.
   
   I read the PouchDB issue and I'm not seeing the motivation for adding this 
other than its not the same as the other behavior. I wonder if we shouldn't 
either a) update the documentation, or b) just set offset and update_seq to 
null if they're requested. However without knowing the motivation for wanting 
these keys I'm not sure if that breaks the downstream use or not.

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
[email protected]


With regards,
Apache Git Services

Reply via email to