Re: Schema OID definition limits
Hallvard B Furuseth <[email protected]> Fri, 8 Apr 2011 00:59:09 +0200
| Newsgroups | gmane.ietf.ldapext |
|---|---|
| Message-ID | <[email protected]> |
I wrote: >Mark Hinckley writes: >> Is there a limit defined in the LDAP RFCs anywhere for the length of >> the OID value for a schema definition, for example inetOrgPerson is >> 2.16.840.1.113730.3.2.2, whereas organization is 2.5.6.4. > > I don't think so. Object identifiers are defined by ASN.1, the LDAP > RFCs just refer to it. Whoops - poorly placed sentence. To clarify: Nor do I know of limitations in ASN.1:-) > Finally, your OIDs might meet a quick-and-dirty implementation which > represents OID components as fixed-size machine integers. So it may be > advisable to stick to components < 2**31, and 2nd component < 2**31/40. > Or nowadays, perhaps I should be saying 63 instead of 31 bits. (The > final bit would be for the sign, or for a "this is the last OID > component" flag.) For the positive sign, I should have said. For languages or code which use signed integers. -- Hallvard