Rene,

"Rene G. Eberhard" wrote:

> Ben
>
> > > > Has anyone looked at using the OpenSSL src as a baseline for a WTLS
> > > > implementation ?
> > > >
> > > > I'd be interested in any comments, opinions etc.
> > >
> > > WTLS and TLS are quite different. Beside new algorithms
> > > (ECDH, ECDSA, ..) you also have a different record layer and
> > > tonns of new protocol structures. You also must support
> > > datagram transport layers and the whole state machine looks
> > > more complex. These were only the most significant
> > > differences. There are others too.
> > >
> > > It's not a good idea to integrate the W into the current
> > > OpenSSL project.
> >
> > Why not? OpenSSL is designed to allow exactly these kinds of variations.
>
> I doubt that there's an overall design =).
>
> > Don't forget that despite its name, OpenSSL is really a general-purpose
> > crypto library.
>
> I agree that OpenSSL is a good general-purpose crypto library. But
> in real life SSL, TLS or WTLS are protocols that 'only' use mechanisms
> of a crypto library. There's no need to embedd all crypro related protocols
> just because they use crypto mechanisms.

Who said anything about all protocols? And why not by the way :) Specifically,
why is there no need to add WTLS to the openssl project? Coming from the
embedded market I can only say that it would be very interested in adding
WTLS to the project.

Regards,

Lennart Bang
[EMAIL PROTECTED]


begin:vcard 
n:Bang;Lennart
tel;cell:+46 70 733 14 70
tel;fax:+46 8 732 63 10
tel;work:+46 8 446 34 67
x-mozilla-html:TRUE
url:www.netstream.se
org:Netstream AB;Networking department
version:2.1
email;internet:[EMAIL PROTECTED]
title:http://www.netstream.se
adr;quoted-printable:;;Enhagsbacken 9=0D=0A;S-187 40 Taby;;;Sweden
fn:Lennart Bang
end:vcard

Reply via email to