Re: Automatic Entry and Forms

hallam@zorch.w3.org
Tue, 27 Feb 96 18:14:30 -0500


From: hallam@zorch.w3.org
Message-Id: <9602272314.AA29520@zorch.w3.org>
To: carl@chage.com (Carl Hage)
Cc: hallam@w3.org, www-html@w3.org, hallam@zorch.w3.org
Subject: Re: Automatic Entry and Forms 
In-Reply-To: Your message of "Tue, 27 Feb 96 13:12:29 PST."
             <9602272112.AA02969@slick.chage.com> 
Date: Tue, 27 Feb 96 18:14:30 -0500


Carl's points about EDI are IMHO very usefull. 

His description of the complexity of EDI is something I can heartily
agree with, like many ISO type standards it is well - like an ISO 
standard. I think that the reason why the IETF has been able to
move fast in many areas has been because it has not embarked upon this
type of spec very often. 

The approach I would like to suggest is that we simply co-opt the EDI
templates by assigning them URIs, URN:org/iso/edi/fooo/

As mobile code becomes more prevalent there many be two routes to achieving
the type of functions that Java attempts. We could adopt the postscript
method of sending an interpreter along with every piece of data. This is what 
Alan Kayes strenuously recommended. I think he is wrong on this point however.
Postscript is a fine printer format but lousy for everything else. You cant
feed postscript into Word and edit it. 

The second and IMNSHO better solution is to try to keep a clear separation
between data and semantics. Rather than intermingle code and data keep them
separate.

A template URI could be used to download an interpreter for data entry or
other processor. Then forms which shared the same fields could share the
same validation code. Libraries of standard validation code could be built
up and shared.

	Phill