An external journal article derived from go/service-availability, as a follow up to
the SRE Book and as one way of supporting the larger Google SRE branding effort.
The first half of the article is more or less go/service-availability verbatim with
some small additions, editorial tweaks, and figures (which are pretty rough at this
stage). The second half (Strategies for minimizing and mitigating critical
dependencies) section is new content.