Revisit at clause

[lhstrh]

An ambiguity in the grammar is resolved with a mandatory semicolon at the end of an at clause that designates where a federate should run. We should eliminate this strange exception (which I take blame for introducing), and review whether we even want this language construct because it's probably too restrictive to specify these kinds of parameters at compile time.

(...) We should follow up on the required semicolon in the atclause and revisit that construct all together. It doesn't appear in the tests and is not very well documented, so I don't think we need to address it before merging this in.

Originally posted by @lhstrh in #1227 (comment)