I'm using LibreTube for a very long time and I often switch between different instances when the one I use stops working. I noticed that instances without CDN work more reliably for me and they don't even have this recent "sign in to confirm that you're not a bot" error. But on the official wiki it says that instances with CDN are preferred. What is this thing and are there any concerns when using an instance without it?
CDN stands for Content Delivery Network, and is essentially a network of servers that act as distributed caches to make accessing various assets faster for the end user. Say you have an application that needs to load a bunch of images and scripts, instead of going to the home server to fetch all that content (which could be on the other side of the world from you) the idea is that you save time by accessing those assets via a nearby CDN, reducing the chance of routing errors etc.
There are downsides to CDNs as well. End users might have privacy settings that block CDNs, they may cost money to make use of, might make deployment a bit more complicated. I my biggest issue with CDNs is that they may not respect the end user's privacy, so if I develop an application and I really care about my users' privacy, I'd likely endeavour to not make use of a CDN as I can't guarantee that the CDN won't be a weak link in my chain. It doesn't matter how much I respect people's privacy if the CDN stores logs and tracks my users.
Huh that is much different to what I thought it was. Thank you very much for the explanation and add it to the official documentation when?
Also I have another question: is CDN mandatory for the LibreTube's proxy system to work?
My apologies, I should've prefaced that my experience with CDNs come as a developer, but not for LibreTube. Think of it more as an overarching explanation of what a CDN is, and not for how LibreTube makes use of it as I don't possess that information.