remove rid pointer (but not rid)...

[Garrett Wollman]

In article <20060419.165709.22017808.imp at bsdimp.com> you write:

>in theory with the current API the rid could change, a quick audit of
>the tree I just did shows that no one currently changes the rid.

The perils of generalizing from zero examples.  Mea culpa.

I no longer have even the foggiest idea why I thought that might be
useful.

-GAWollman