Segfault on startup with missing assets

[Kriper1111]

The expected flow is that it safely exits, of course. The segfault is probably because it tries to unload the shader that haven't been not compiled/linked properly.

It's really a part of a bigger issue with the "framework" and how it (poorly) interacts with RAII - I'm always certain that trying to move any Object around would result in its internal buffer being unloaded.
This is quite critical and I honestly should address it sooner than later.
This and a bunch of other refactors.

[Kriper1111]

Weirdly enough it's because gl context gets unloaded BEFORE the Object destructor goes. Why's that though.

[Kriper1111]

Oh no, that makes sense -- the constructor failed so the Display haven't been constructed, thus there's nothing to destruct, yet the Objects have been initialized, and they'll going to be destructed only when the class unloads, which is after the main code has finished running.

So there are two solutions, I guess:

1. the dirty one, have a ::cleanup() that will just remove the excess shaders and objects.
2. the murky one, have Display have them as non-static members. It's not like we're going to have tons of Displays around.
3. the decent one, if possible - have gl context get destructed last. And probably wrap the whole busyness into like a Application or Window or something. Just wrap it into an object that will get RAII'd destructed last, with the context.

[Kriper1111]

Oh and I suppose the same destiny awaits WorldGridDisplay

[Kriper1111]

The issue is kept open until a framework-level fix is implemented, meaning "I'm 99% sure Objects and Shaders will deallocate only after there's no other same thing left and before context gets destroyed."