[4228] in cryptography@c2.net mail archive
Re: PGP SDK
daemon@ATHENA.MIT.EDU (Thomas Junker)
Sat Feb 20 13:05:40 1999
From: "Thomas Junker" <tjunker@phoenix.net>
To: Jon Callas <jon@pgp.com>
Date: Sat, 20 Feb 1999 05:07:24 -0600
Reply-To: Thomas Junker <tjunker@phoenix.net>
Cc: cryptography@c2.net
In-reply-to: <v04003a12b2f39c0664b7@[161.69.48.157]>
On 19 Feb 99, at 14:55, Jon Callas wrote:
> You're not using the PGPsdk. It sounds like what you are using is the old
> ViaCrypt libraries.
Yes, that has become clear, notwithstanding the fact that NAI sent this to us
in late 1998. I think that was because we had insisted on something we could
use from Visual Basic, and Viacrypt/PGP/NAI seems never to have considered
that tons of people can and do throw together large amounts of useful
functionality in VB. If a VB programmer can make use of Windows and other
APIs, there is no reason why the same can't be done with crypto APIs, unless,
of course, the crypto stuff has been put together in such a way as to preclude
use from VB. I understand that callbacks are a requirement to use the PGP
toolkit, and that VB either can't do them or couldn't do them in earlier
versions.
I'm still scratching my head over why a dependency on callbacks would be built
into what logically should be a set of crypto subroutines with a clear and
simple programmatic interface. The difference between using the old DOS
command-line PGP from within an application and using what I see described in
the PGP V6 SDK manual is like the difference between bicycling to the store
for a loaf of bread and building an all-terrain vehicle with GPS navigation to
get you there and back. The objective is _still_ just to get a loaf of bread,
and quite possibly it's a very tiny piece of the overall application, a piece
that doesn't warrant such complex work to implement.
It's becoming clearer to me why crypto-enhanced apps are as scarce as hen's
teeth.
Regards,
Thomas Junker
tjunker@phoenix.net
