-
-
Notifications
You must be signed in to change notification settings - Fork 886
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Changes to get_random_community_id: add filters for nsfw and private, use algorthm that doesn't scan the entire table #5267
base: main
Are you sure you want to change the base?
Conversation
…thm that doesn't scan the entire table
CREATE INDEX idx_community_random_number ON community (random_number) INCLUDE (local, nsfw) | ||
WHERE | ||
NOT (deleted OR removed OR visibility = 'Private'); | ||
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
why smallint? due to birthday paradox you have a high collision probability already with 2^(n/2) communities = 2^8 = 256 communities
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It uses less space, and collisions are fine because the random number is updated after the community is picked
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
why would you update the number after picking?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Its described here. I'm sure there are others ways of doing it, allowing collisions and updating the random number after picking, being one way to do it.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
right, i missed the last paragraph there. Still seems a bit awkward to me, especially having a read-only innocuous request cause a table write.
Why don't you just do select * from community where id = 1+(random() * (select max(id) from community))::int
in a loop? The id of communities is already unique and uniformly distributed, with a fill rate of like 99% (so needing more than one/two queries is exceedingly unlikely) - no new column or modifying query or function needed.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I guess that would make filtered (nswf/local) more difficult, mh.
CREATE INDEX idx_community_random_number ON community (random_number) INCLUDE (local, nsfw) | ||
WHERE | ||
NOT (deleted OR removed OR visibility = 'Private'); | ||
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Its described here. I'm sure there are others ways of doing it, allowing collisions and updating the random number after picking, being one way to do it.
No description provided.