Should we have (srfi srfi-N) modules for all SRFIs?

[Rob Browning]

There are SRFIs like SRFI-6 for which guile already includes an
implementation, but I'm wondering if it would make sense to go ahead
and add a no-op module.

The question is, do we want to make the assertion that you can always
make sure you have a given srfi's functionality by running

  (use-modules (srfi srfi-N))

no matter where the code actually resides?

One argument in favor of that it makes code immune to restructurings.
For example, if we want to move in the direction of having a more
pared down rNrs core, with as much as possible moved to optional
modules (a plan I would wholly support), then having a canonical way
to refer to the srfis would make this easier.

If no one objects, I'll start making this so.

-- 
Rob Browning <address@hidden> PGP=E80E0D04F521A094 532B97F5D64E3930