using internal rand.Rand?

Idea from Brett Lawson...

Noticed this in CBGT:
"NOTE: This library initializes math's random seed (rand.Seed(time.Now().UTC().UnixNano())) for unique id generation."
Why not use an internal rand.Rand?

The issue identifies CBGT's use of math/rand.Seed(time.Now().UTC().UnixNano()) for unique ID generation, but names no file or test. Start by locating that seed call in the moved cbgt repository and determine how random values are used; done would mean the proposed internal rand.Rand approach is implemented without the global seed side effect and is covered by the relevant tests.

Written by the indexing model from the issue text.