[foaf-protocols] Why exponent/modulus

Nathan nathan at webr3.org
Sat Sep 18 03:22:01 CEST 2010


Henry Story wrote:
> Creating an ontology for PEM and DER is not the big issue. It's a few lines of rdf
> in a file. The issue is whether it simplifies things to tie people to publish
> ASN.1 strings. My guess is that it won't get rid of the need to parse those files
> anyway. So we might as well work on tweaking the tools to parse them correctly.

I could debate about whether it simplifies things, and the obvious 
answer is that technically ASN.1 simplifies nothing. However, in reality 
I'd suggest that ASN.1 tooling for the common usage is abundant, those 
DER/PEM base64 encoded strings are the defacto usage around the net and 
the asn.1 is locked away nicely in black boxes. So perhaps by pushing 
people to actually get in to the ASN.1 and rip out chunks which aren't 
commonly used we may be making this more complex than needed. Which 
seems like false logic to some extent, but in reality give somebody a 
PEM/DER and they stand a chance of knowing what to do / recognising it, 
give them a modulus/exponent and say it's a public key and they'll 
probably wtf a little.

Before getting any more in to it though, might be worth asking the 
question..

Would anybody object to using PEM/DER representations of the certificate 
(as sent over http+tls) in the Web Id protocol?

and conversely, does anybody prefer this, think it has benefits, think 
it could simplify rolling out / understanding / implementations?

(best to view this as optional if it were added for now I'd think..)

Best,

Nathan


More information about the foaf-protocols mailing list