"parse a vint" algorithm does not align with "matches the signature for WebM" algorithm

[Gallaecio]

@akshaysharmajs is implementing the standard for Python (https://github.com/scrapy/xtractmime), and we find the webM parsing algorithm rather confusing.

Our confusion did not stem from what was reported in #93, though, but from the return value of the algorithm to parse vint. The algorithm states that 2 values are returned, but only 1 of those values is actually used by the calling algorithm. So the vint-parsing algorithm could be cut in half, and it would not only be simpler, but way more clear.

[annevk]

cc @padenot