[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[cme@cybercash.com: Re: Comments on SPKI draft of 25 March 1997]
Return-Path: <cme@cybercash.com>
X-Sender: cme@cybercash.com
Date: Sat, 29 Mar 1997 13:59:28 -0500
To: rivest@theory.lcs.mit.edu (Ron Rivest)
From: Carl Ellison <cme@cybercash.com>
Subject: Re: Comments on SPKI draft of 25 March 1997
Cc: spki@c2.net, blampson@microsoft.com
In-Reply-To: <199703291833.AA22655@swan.lcs.mit.edu>
Mime-Version: 1.0
At 01:33 PM 3/29/97 EST, Ron Rivest wrote:
>I suggest that the BNF have the five types as above, but that the actual
>syntax use the same object type for the last three, as SDSI does:
> ( ref <key-or-key-hash> name ) for <simple-name>
> ( ref <key-or-key-hash> name1 name2 ... ) for <fq-name>
> ( ref name1 name2 ... ) for <relative-name>
>and that the implementation check for the appropriate number and type of
>arguments (e.g. that there is a real key-or-key-hash when necessary, etc.)
I'd like to hear from other developers on this issue. I agree with the need
for these three, but I prefer having different object types for them, so
that the determination of kind of object doesn't require a look ahead.
>Also: shouldn't the names allowed be general byte-strings and not just tokens?
>What if a Chinese user wants to give appropriate names to the public keys
>of his colleagues? Shouldn't he be able to used unicode, e.g.
> [unicode] chinese-name
>(or, following my previous note)
> ( display-type unicode chinese-name )
>instead of just tokens?
Yes, they should.
- Carl
+------------------------------------------------------------------+
|Carl M. Ellison cme@cybercash.com http://www.clark.net/pub/cme |
|CyberCash, Inc. http://www.cybercash.com/ |
|207 Grindall Street PGP 2.6.2: 61E2DE7FCB9D7984E9C8048BA63221A2 |
|Baltimore MD 21230-4103 T:(410) 727-4288 F:(410)727-4293 |
+------------------------------------------------------------------+