> As a point of general design, we should always be extracting the new > smf tools to a temporary location and using those tools, so you can > avoid the need for flag-days of this nature. i.e. the archives themselves > can move themselves past a flag day.
In general, I'm not sure how this could work reliably -- e.g., if those tools depend on new library API functionality which in turn depends on paired kernel functionality. -- meem
