refactor: remove tox_options_default(...) and make Tox_Options opaque

[sudden6]

tox_options_new(...) already returns a default initialized options
struct and this function makes it impossible to malloc buffers in
Tox_Options, so remove it.

Required for #909

Fixes #273

---

This change is 

[nurupo]

You should not malloc buffers in Tox_Options. See #909 (comment).

[sudden6]

Reopening, resoning in: #909 (comment)

Edit: @nurupo even if you leave out #909 removing this function still makes sense, since it provides no real value (reseting to defaults is done in tox_options_new(...)) and prevents further evolution of the API, because of the uninitialized pointers problem.

[sudden6]

Updated this PR with the logical next step of making Tox_Options opaque.

[JFreegman]

Reviewed 19 of 19 files at r2, all commit messages.
Reviewable status: 0 of 1 approvals obtained (waiting on @sudden6)

---

toxcore/tox.h, line 520 at r2 (raw file):

/**
 * This struct contains all the startup options for Tox. You must tox_options_new to
 * allocate an object of this type and tox_options_free to free it.

"you must use"

and "and use"

[nurupo]

Edit: @nurupo even if you leave out #909 removing this function still makes sense, since it provides no real value (reseting to defaults is done in tox_options_new(...)) and prevents further evolution of the API, because of the uninitialized pointers problem.

Ok, I'm leaving #909 aside for now then. What is the uninitialized pointers problem? Are you referring back to #909 even though you said "if you leave out #909", or is there some other problem I don't know?

Anyway, we do want to make Tox_Options opaque, so that's a nice change to have. tox_options_default() becomes a bit pointless indeed after that change. I guess it could be useful to reset the object to its default values without having to free/new it, but it's not obvious to me why a user would want to do that, doesn't really add much value, so yeah, let's remove it.

[sudden6]

What is the uninitialized pointers problem? Are you referring back to #909 even though you said "if you leave out #909", or is there some other problem I don't know?

My mistake, after making Tox_Options opaque, all pointers are either NULL or valid since we initialize them in tox_options_new(...).

I'd still remove tox_options_default(...) to keep the API as minimal as possible.

@JFreegman thx, fixed.