So it sounds like compression before encryption should only be done in specific circumstances because it can be a security issue depending on use case, but encryption before compression should never be done because it will almost always increase the size of the file

Depends on if you’re using lossless or lossy compression. Lossless compression will usually make it bigger, because it relies entirely on data being formatted so their are common patterns or elements that can be described with fewer parts. Like, an ok compression algorithm for a book written in English and stored as Unicode would be to convert it to ASCII and have a thing that will denote Unicode if there happens to be anything that can’t convert. An encrypted version of that book would look indestinguishable from random characters, so compressing it at that point would just put that Unicode denoter before every single character, making the book end up taking more space.

Compress the encrypted data. You’re talking about encrypting compressed data, this was talking about compressing encrypted data.

And the fact that it can grow data means you should really put a test to make sure that the compressed data is actually smaller… I once had something refuse to allow me to upload a file that was well below their 8Mb file limit while it was claiming it was above the limit, and I’m assuming it was because they were testing the size after compression and that file grew from 6Mb to above the limit.

If that’s true, what’s to stop someone else from just compressing it themself and opening the same attack vector?