correct to use actual xoshiro128++ implementation

[corwinkuiper]

During transcription there appears to be an error that was introduced such that we ended up using some strange combination of xoshiro128++ and xoshiro128**. This corrects it to use xoshiro128++.

Here is the original C implementation:

static inline uint32_t rotl(const uint32_t x, int k) {
	return (x << k) | (x >> (32 - k));
}


static uint32_t s[4];

uint32_t next(void) {
	const uint32_t result = rotl(s[0] + s[3], 7) + s[0];

	const uint32_t t = s[1] << 9;

	s[2] ^= s[0];
	s[3] ^= s[1];
	s[1] ^= s[2];
	s[0] ^= s[3];

	s[2] ^= t;

	s[3] = rotl(s[3], 11);

	return result;
}

The only error I spotted being that we multiplied by 9 rather than adding s[0] on the first line.

This requires a 0.X bump as we said we wouldn't change this in minor releases (we're still V0, so the third number is considered the minor release).