Conversation
|
Zabuzard
left a comment
There was a problem hiding this comment.
Can you change the logic to fallback on the previous behavior or similar when loading fails?
I want to prevent that when "loading all users" fails for whatever reason that this means it fails for all categories.
Currently, if I get this right, when there is a problem with one category it will still attempt it for the other categories individually.
You sure that this is needed? |
2133a8e to
cea6143
Compare
|



About
In order to keep helper roles below 250 users, we have a routine that removes a role from a more inactive user.
The old implementation used
Guild#findMembersWithRolesto filter for them, which surprisingly requests allmembers first and does the filtering on the client. This happens for each role (about 17). So we basically query the same
content multiple times. The new implementation requests all members beforehand and then reusing it to prune roles if needed.
This is related to the issue about a
TimeoutExceptionon the member chunk requests when requesting all members. This doesn't mean it is going to fix that issue, but it is still an optimization.