I'm digging into the API for the site and I notice when pulling data for a game there are four "cover" sizes within the JSON for the assets. They are "cover-tiny", "cover-small", "cover-medium", and "cover-large".

From what I have seen pulling random games in the API and looking at these, I haven't seen any difference in the URI for each cover asset, they are always identical. I also noticed that there doesn't seem to be a different URI if there is a Cover and a Japanese Cover saved for the game. I couldn't see how a game with only a Japanese Cover would show up in the API, since the only game I am a moderator of has both covers and I can't see that detail on games I'm not a moderator of.

Based on this, when I pull the assets is there a preferred cover size to use? I want to try and future proof my code so that I'm pulling the best option when/if the site uses these different sizes if they aren't already.

The actual dimensions of the image on my end doesn't matter as I'll scale appropriately after pulling it, I just want to try and ensure I will always pull a cover image for any game I that has a cover.

Based on the (horribly outdated) API documentation it seems like the different cover sizes used to actually return different things, but I think something's since changed in the back-end that's made the extra fields obsolete, and the API itself hasn't been changed to reflect that. I imagine that checking any of the fields is fine going forward, but for completeness I'd just check them all. Japanese Covers is also a legacy thing I think, it used to be a thing but the option was removed a while ago. I think some games might still technically have them? But they're few and far between.

In general I don't think there have been any actual changes to the API in at least 4 years, like a lot of things about the site it is in desperate need of an overhaul. Last I heard the current API is most likely going to stay as is for the foreseeable future and there'll be a complete overhaul of it sometime in the distant future. I haven't actually seen anyone mention anything API related in over 2 years, so I think it's pretty low priority unfortunately.