ietf-openpgp
[Top] [All Lists]

Re: PGP Message Exchange Formats Internet Draft

1997-08-11 06:48:55
Hello to all,

at 10:12 11.08.1997 +0200 Assar wrote:

[..] let's focus on
the essential point here, and that is compatibility. There is no point in
sending a message out into the world that the spooks can't read, when the
recipient can't read it either because he doesn't happen to have PGP
5.123.2 (or whatever) on his machine.

I'd like to refer to te Tiff-format again: There surely might be
comprehensive readers being able to read&write in many formats and I think
the old standard  (RSA/IDEA/MD5) and a set coping with the actual future
standard (whatever it might be) should be read by all. Therefore slimmed
versions might run on handheld systems with acceptable performance whereas
versions reading and writing any known format will more memory or computing
power when using longer keys.

The choice of algorithm is largely a
matter of opinion and politics, and everyone has his own personal favorite.
If we *really* open this up, then it won't be long before there are an
infinite number of possibilities. (Perhaps we'll even see Rot13 in there -
it's fast and someone out there somewhere loves it.)

:-)

I'm afraid for the moment we are at a point here it is to decide whether to
imply anything into one comprehensive format (like Tiff) or have many
formats parallel formats like a set of (Tiff, Gif, JPEG, TGA, BMP, PCX,
etc.) with a whole lot of compatbility problems in the future.

IMHO the first option should be given the choice.

Perhaps it is possible to come up with some form of 'plugin' architecture,
allowing the enthusiasts to think up any number of options for themselves,
but keeping the core clean and compatible.

Yes, a plug-in-architecture might be the optimum. Especially there
definitely will be future needs similar to those occurred today: DES is no
longer considered secure and there definitely are problems with MD5.
 
Who dares to estimate that factorizing large numbers won't be more easy
within 10 years thus endangering RSA ?  With an universal standard one would
just have to change from RSA to anything else (Diffie-Hellman, El-Gamal or
whatever yet to be developed) and still will provide downward compatibility.

Read you,

Hagen


-- 
_____________________________________________________________________
Hagen Wollert                               
h(_dot_)wollert(_at_)tu-clausthal(_dot_)de
PGPed E-Mail welcome    -    public key available on most key-servers
2048/BBAF99CD   MD5: 11 AD A2 70 67 0A 32 CA  34 D8 F0 CE 67 8A B7 A1