RE: Embedded certificate image

["Jim Schaad" <[e-mail]>]

RFC 2397  The data URL scheme.  Limitations on the length of the URL are
application dependent.

> -----Original Message-----
> From: [e-mail] [mailto:owner-ietf-
> [e-mail]] On Behalf Of Stefan Santesson
> Sent: Thursday, July 30, 2009 10:15 AM
> To: Peter Gutmann; [e-mail]; [e-mail]
> Subject: Re: Embedded certificate image
> 
> 
> Peter,
> 
> In-line;
> 
> On 09-07-30 10:01 AM, "Peter Gutmann" <[e-mail]>
> wrote:
> 
> >
> > Stefan Santesson <[e-mail]> writes:
> >
> >> This is the case for RFC 3709, which is the standard we would use to
> bind a
> >> cert image to the certificate. RFC 3709 does only offer a URL as
> means of
> >> referring to the actual image, it does not offer any other means of
> local
> >> storage.
> >
> > Why not use one of the type-and-value options, OtherLogoTypeInfo or
> something?
> > This just seems like a horrible kludge, like pounding a nail with a
> scredriver
> > because that's what was lying around.
> 
> We have concluded that the RFC 3709 syntax does not allow any such
> option or
> extensibility. It's either in the URL or nothing at all.
> 
> >
> > How do you distinguish real URLs from not-a-real-URLs?
> >
> > Peter.
> >
> 
> That is my question to Jim. Apparently there is an RFC that would tell.
> 
> In any case, this is just an interesting idea that we should
> investigate. I
> have scary feelings too about this but I would like to turn the stone
> and
> look at it before throwing this option away.
> 
> /Stefan
>