andrew@stl.stc.co.uk (Andrew Macpherson) (03/13/89)
I have been building an almost purely IDA sendmail with 2.1 when I came across a nasty. The c-nrs processor produces an application relay table for one as janet.chn, however that table does not recursively resolve ARs, so a resolution in the chanel table cannot be considered decisive. For instance: uk.co.ist.upper -> uk.co.ist uk.co.ist -> uk.ac.ic.doc Now it *may* be that ni_send resolves this, but that seems an unwaranted assumption, alternatively am I missing something?
jac@doc.ic.ac.uk (Jim Crammond) (03/15/89)
No, not a concept error; but a design decision!! Yes, it is assumed that the right hand side points to a site that ni_send will know about. Ni_send should resolve this since each entry in the nrs tables has a dte address associated with it (The ARs are resolved when the table is produced, so this is static data). If you want to ensure that the RHS of the janet.chn is not itself an AR then it seems to me that c-nrs should do it. I see no benefit of getting sendmail to dynamically resolve ARs when the data is static. -Jim.