Re: RFC3161(TSP): doubts about tcp protocol

["todd glassey" <todd.glassey@xxxxxxxxxxxxxxxx>]

The problem Adrian is that now there are many more patents than the ones
that were included in the original statement afterall its been how many
years since then?

Todd

----- Original Message -----
From: "J Adrian Pickering" <jap@xxxxxxxxxxxxxxx>
To: <ietf-pkix@xxxxxxx>
Sent: Wednesday, April 16, 2003 12:37 PM
Subject: Re: RFC3161(TSP): doubts about tcp protocol


>
> At 15:43 16/04/2003, Stephen Kent wrote:
>
> >As noted in my response to Todd, his assertions re problems with
> >implementing RFC 3161 are subject to dispute. The RFC does alert folks to
> >possible IP issues,
>
> and I did so while it was in discussion

As did any number of people - but once the initial statement of a few
patents were added the furor over this disappeared and 3161 was back on
track to standardization. The problem is that this effort didnt take a year,
it took seven so far that I know of and in that time any number of new or
other derivitive patents have been issued.

>
> >  but at least one court case in the U.S. where a patent holder tried to
> > sue based on a simple time stamp protocol implementation that is very
> > analogous to what 3161 calls for, provides a basis for believing that
> > very broad patent claims in this area may not be upheld in court
>
> I am glad to hear it.

I dont think you really are - this is the US Court saying that Bell Labs
cannot patent all forms of Digitally Signed Time and really nothing else.
All the other claims held and are still good as far as I know.

> The one that really concerns me is the Haber et al
> one in which time stamp data is extended by restamping the (hash) records.

That is part of RE34954 and that also is the reissue patent from Bell Labs.
But that claim withstood the trial but was not one of the primary claims
that were sought to be invalidated.

> This is neat but nearly obvious.

yes again I think  its a repackging of existing math and process as
crypto-speak to accomplish the making of a claim of a new and novel process
and that's pretty much it. Kinda sad that it gets passed the USPTO but it
does now and then.

> I look forward to this being challenged in
> court as it needs to be - it nearly cripples the principle of archiving.

Agreed. And this was originally done in the mid 60's by IBM in their code
management tools.

> This is what worries me about whatever is behind TAP.

yes - but TAP can use a modular regenerating system. TAP's biggest worry is
that it needs 3161's TST as its time setting process and that is inherently
flawed - also the other issue is that TAP does not really take into account
that UTC is a proclamation that happens once a month from the BIPM and not
some magically uniform piece of data. Also that as new timebases are
developed they allow us to retroactively plot Delta-T between what we
stamped something as and what it really was. And that als has to be figured
into the bigger picture here. remember that what the TAP is really for is
digital evidence.

>
> Adrian Pickering/
> University of Southampton, UK
>