[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
MIB lint web page available
- To: mibs@ops.ietf.org
- Subject: MIB lint web page available
- From: Bill Fenner <fenner@research.att.com>
- Date: Thu, 7 Dec 2000 23:04:35 -0800
- Delivery-date: Thu, 07 Dec 2000 23:06:17 -0800
- Envelope-to: mibs-data@psg.com
- Versions: dmail (solaris) 2.2g/makemail 2.9a
For some time now, I've been extracting everything that looks like a
MIB fron RFC's and Internet-Drafts and they've been available at
http://www.aciri.org/fenner/mibs/
This is probably the first public announcement of something I've been
doing for a month or two, which is running the "smilint" program
(from libsmi) on each of these extracted MIBs. The intent is that
MIB authors can check their drafts for syntax simply by checking a web
page, instead of by having to install some software. (Yes, it's a small
barrier, but it's obviously still a barrier, given the number of errors
in published MIBs, so might as well lower it...)
http://www.aciri.org/fenner/mibs/htmllint/
and
http://www.aciri.org/fenner/mibs/htmllint/bydate.html
It seems that some aspects of RFC2578 are generally ignored (like
the suggested 32-character limits on object names), and other aspects
have evolved (like RFC 1442 and RFC 1902 suggest replacing INTEGER
uses with Integer32, but RFC2578 says that they're indistinguishable,
but the web page still warns that you should use Integer32 instead of
INTEGER). If there's consensus that parts of RFC2578 are not relevant,
I can update the page to skip certain warnings.
Feedback is welcome. Yes, a search-all-the-mibs feature is on the list =)
Bill