On Tue, Mar 05, 2002 at 09:52:57PM -0800, Guy Harris wrote: > Fortunately, it looks as if "sprint_realloc_value()" doesn't actually > *use* the resulting buffer, so perhaps if "get_symbol()" were modified > to allow a null pointer as the last argument, meaning "don't bother > formatting the OID for this object, please", it might work better. I'll > look at tweaking the SNMP library to do that, and, if it works, I'll > send patches off to the NET-SNMP maintainers (for both 4.2[.x] and 5.0).
Well, that seems to have fixed things. I'll send off some patches.
