[Date Prev][Date Next] [Chronological] [Thread] [Top]

Re: Syntax OIDs

At 12:54 AM 8/14/99 -0700, Howard Chu wrote:
>Is it just me, or does anyone else find RFC 2252's specification of
>"noidlen" for attribute syntaxes annoying?  Everything else allows either a
>string *or* a numeric OID, why are syntaxes required to use numeric OIDs?

Raise this issue to LDAPext.  It looks like it was done for a specific
reason to which I do not know.

>It makes attribute type definitions incredibly unreadable, and not much fun to
>write either. Would anyone object if I modified slapd to allow string-form
>OID references to attribute syntaxes in the parser?

For input, as long as the OID for the syntax is available such that
output of subschema rules comforms to the specification, I do not
have any problem with the parser handling such.   However, I think
it wise to use noidlen in the distribution schema rules.  This
allows for information to be "borrowed" for other purposes.