[Cryptography Team] First Message

[Sean Glazier]

Actually with the stuff coming out for our current release it will be one,
*Uno* asn1 implementation.

What happened is I hacked what I needed together for x509 first round
another engineer took another approach and then Len decided to take a
different and better approach to the ASN1 implementation. It is faster and
more flexible and will deal nicely with what we have to deal with. The
crypto code has changed quite a bit. Api's are better and we attacked the
ASN1 problem so we can move x509 forward. Our latest code and builds I will
send a link Monday or Tuesday to our current so people can have the latest
and greatest and take a look through it.  

Yes asn1 is hard to wrap around and the reason we had so many was time to
market. Implement what we needed to get it out and then do the rest later
and tell people hey this will change. So the other implementations are gone
and things for 7.4 will be on one asn1 implementation that works.


Sean

-----Original Message-----
From: cryptography-bounces at lists.squeakfoundation.org
[mailto:cryptography-bounces at lists.squeakfoundation.org] On Behalf Of Cees
De Groot
Sent: Friday, October 21, 2005 4:24 PM
To: Ron at usmedrec.com; Cryptography Team Development List
Subject: Re: [Cryptography Team] First Message

On 10/21/05, Ron Teitelbaum <Ron at usmedrec.com> wrote:
> The ASN.1 is a precursor to x.509v3.

That's not how I remember it. Note: this is all from some serious
crypto code hacking in the VW crypto library a few years ago, when I
had to make Netscape 4.5 browsers believe that our appserver was a
valid SSL server with valid certificates and stuff, with some
smartcard authentication threwn in for good measure :)

As far as I remember, x.509v3 is a data structure that uses the ASN.1
for its default external representation.

And ASN.1 is a horrible piece of design-by-committee excrement,
apparently extremely hard to write a good clean OO parser for - at
least, that's why I think that a loaded VW image typically ends up
with 2 or 3 ASN.1 implementations (one for certificates, I think
there's another one for SSL handshaking, and I'm getting old but I
recall vaguely that I stumbled over a third one). I don't know why but
I tried my hands at a parser once, and horribly failed :)
_______________________________________________
Cryptography mailing list
Cryptography at lists.squeakfoundation.org
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/cryptography