cluster accept group now but our APIs still require user to pass name rather than key...
ObjectKey's json parser is about to accept following json representation.
"key": {
"group":"g",
"name":"n"
}
```json
"key": {
"name":"n"
}
```json
"key": "n"
Hence, the "default group" to all objects should be identical to make a global json parser for key type.
The sub issues are listed below.
related to #2769
cluster accept creation without group specified. If we use key instead of group and name, how could we create an object with default group?
If we use key instead of group and name, how could we create an object with default group?
The Scala APIs accept the key and the json APIs take both key and name.
The former is for consistent in writing pure scala code. Converting key to name is a weird and ugly style across our project.
The later take “key” but accept the key lacking of “group”. For example: the following requests are identical.
bkKey: {
group:default,
name:”a”
}
bkkey: a
The Scala APIs accept the key and the json APIs take both key and name.
Since scala APIs are part of "client", this conclusion is very important for later developers/users.
Could you add this conclusion to the key() comment? :)
Could you add this conclusion to the key() comment? :)
sure, boss
/cc @oharastream/frontend
The latest changes of cluster APIs are traced here, I will remove all the stale fields in this release (0.8).
If you had any problem, please let me know :)
The latest changes of cluster APIs are traced here, I will remove all the stale fields in this release (0.8).
If you had any problem, please let me know :)

@saivirtue Brace for the impact...
@oharastream/frontend
All done! You can test the APIs of cluster to see the stale fields should not available now :)
Most helpful comment
/cc @oharastream/frontend
The latest changes of cluster APIs are traced here, I will remove all the stale fields in this release (0.8).
If you had any problem, please let me know :)