> Given that disks, disk controllers, and disk array controllers all
> have big RAM buffers and report writes as done long before they
> actually are complete, I'd say: let's flush both sync-file and
> sync-file-system from SRFI 170.

And there was much rejoicing :)

> This form of cutting the Gordian Knot works for me.

Seconded.

> It could someday be the remit of a SRFI or library.

Not sure if this is what you mean, but disk syncing is definitely enough
of a black art that it could merit its own SRFI that contains nothing
but a sync procedure (and a lengthy Rationale section full of caveats
about it :)