There's an excellent -- really -- article available for free download from Skeptic.com called "Why This Universe?" by Robert Kuhn. The title, disarmingly broad as it is (one thinks of "Love and Death," Woody Allen's send-up of overly grandiose Russian novels) is nevertheless too narrow. The article is a synopsis, with a good amount of detail, of each of the major cosmological and philosophical issues surrounding the existence, and perceived characteristics, of the universe.

image: alon
Included in this multiverse of issues are "Meaningless Question‚" (nature and its parameters are a "brute fact," as philosopher Robert Nozick called it); "Necessary/Only Way‚" (the universe is the way it is as the result of "deep essence" of physical law); "Almost Necessary"; "Temporal Selection"; "Self-Explaining" ("the universe is self-created and self-explaining"); "Multiverse by Disconnected Regions"; "Multiverse by Cycles"; "Multiverse by Sequential Selection‚" "Multiverse by Quantum Branching," and so on, and on.
It's really a nicely comprehensive primer on the mind-bending fecundity of mind-bending possibilities, and excellent summer reading for folks interested in Foundational issues.
It's also good news this week -- or bad, if you were hoping for Relativity to take it on the chin -- for the question of whether fundamental constants have changed across time. The ratio of electron-to-proton masses has been carefully tested by an Australian lab comparing light from a quasar to the same type of light produced in a lab, in the thought that, one of these signals being a few billion years old, fluctuations across cosmic timescales would be observed. Physicsweb has the article, wryly titled "Fundamental constant is pretty much constant." The short version: plus ca change . . .

image: orangeacid
Question: Einstein was smart enough to realize that if the clock is slowing down by the same amount as the yardstick is shrinking, the inertial observer won't be able to tell that anything has changed. Is it possible that the fundamental constants could be fluctuating in sync with the parameters required for measuring them, so that they always appear to be inviolate?