Re: [Isms] fingerprint TCs and hash types

[Wes Hardaker <wjhns1 at hardakers.net>]

>>>>> On Fri, 11 Sep 2009 06:51:36 +0200, Juergen Schoenwaelder <j.schoenwaelder at jacobs-university.de> said:

JS> In general, I like this format. The downside of course is its length
JS> compared to an OCTET STRING representation:

JS> sha-1	 20    65
JS> sha-256	 32   103
JS> sha-512	 64   199

Yep.  I thought about it.  But the upside of human readability and
string management might outweigh it.  Most management tools do deal with
hex <-> binary conversion fairly well, but this is still probably easier
in the end.

Another win, though, is that the prefixes are already standardized and
we don't need to do anything else.

JS> The second column shows the length of the hash and the third column
JS> the length of the fingerprint (including the label). With sha-512, we
JS> are still well below the common 255 octet boundary but the overhead is
JS> significant and even the sha-1 fingerprint easily allows to ship an
JS> additional varbind while still saving space.

If you're counting bytes in terms of bytes on the wire, you probably
need to include the type encoding as well for the second column that
holds the type which will be 2 OIDs in size length (adding in the length
for the column OID and the value OID).  However, the results will still
be small compared to the ascii encoding.

-- 
Wes Hardaker
Cobham Analytic Solutions