HI,
While I prefer correct MIB modules, I don't believe the answer is so clear cut. It all really depends on what information you want from a MIB module. What I've seen use of the are the following by programs:
0) OID resolution 1) OID to descriptor mapping 2) Textual convention resolution 3) Data type (value of SYNTAX clause) for each OBJECT-TYPE definition 4) Indexing information 5) variables for traps/notifications
What I haven't seen used by programs (except for display) is the following: 1) status 2) description and reference info 3) object and notification group info (except being used incorrectly by some agent testing systems) 4) module compliance info 5) agent capabilities info
So, if your program doesn't need, say, indexing info, if it is screwed up, then "no problem". But if you do, then IT HAS GOT TO BE CORRECT OR YOUR PROGRAM WILL NOT DO THE RIGHT THING.
The most amazing thing is that some market leading software, such as HP OpenView, doesn't do OID resolution correctly, nor does it do anything correctly with the data types, or indexing in the MIB browser.
If you want producers of MIB modules to write correct ones, then they need to have a motivation to do so. Writing correct (and well put together and USEFUL) MIB modules is not easy.
Regards, /david t. perkins
-- !! This message is brought to you via the `libsmi' mailing list. !! Please do not reply to this message to unsubscribe. To subscribe or !! unsubscribe, send a mail message to libsmi-request@ibr.cs.tu-bs.de. !! See http://www.ibr.cs.tu-bs.de/projects/libsmi/ for more information.