EDI and WWW

Tim Berners-Lee (timbl@www3.cern.ch)
Wed, 15 Dec 93 10:24:16 +0100


>From: sdw@meaddata.com (Stephen Williams)
>To: Multiple recipients of list <ietf-edi@byu.edu>

I cross-post this to www-talk, as WWW should be aware of EDI.

For wwwwizads: There is discussion
in ietf-edi for doing EDI (electronic commercial transactions) over
the Internet. EDI seems to be like using forms with email submission,
with a forms language which HTML+ _ought_ to be able to subsume.
I suggest someone get hold of the specs and write a DTD which is
equivalent. Whether you like the look of the EDI format or not
(if you are a WWWWizard I can hazard a guess ;-) there is an established
community using this stuff and so the standards must reflact needs in
some way at least. A good test of a new, general, system, is that it
can represent other things without pausing for breath. This is what
WWW rather specialises in. First thing is to get hold of X.435....
We will soon have apparatus for interactive transactions with forms
operating in I suspect a largs scale on the Internet for applications
which are not legally and financially critiacal. Forms with HTTP are
much faster and neater than anything email or even "fast batch"
as they say in EDI. PEM on top of HTTP will introduce the security,
and then there will suddenly be a complete Internet solution for EDI
as a spin off of what we have now. We don't want then that there is
some basica incompatibility with EDI. That's why I suggest someone
check out the specs. I suspect Dave Raggett won't have time soon,
so is there a volunteer?

Now to the message I quote:

>> David [Crocker] said:
>>

>> >Do people (on this list specifically, but in the industry in general, if

>> >they've thought about it) think that X.435 and "internet edi" are in

>> >competition, or that they are just two of many approaches to electronic

>> >commerce?
>
Stephen Williams said:
>They are going to interoperate, if I have to write the software myself...
:-)

Aaaagh, the Internet spirit!

>> In the commercial setting, it is definitely competition. The X.400 VANs
>> already offer non-X.435 EDI and have promised X.435 upgrades. They can
show
>> working, off-the-shelf software.[...] businesses are placing
>> their efforts on X.400 and the INTERNET has lost this race even before
they
>> entered it, except for those organizations that are already on the
INTERNET.
>
>Only until we create it and standardize on it.
>
>We really should have done this long ago. Can anyone provide us with
>stipped down X.400/X.435 code/specs? The X.400 stuff should be
>available from the pp/isode distributions. The X.435 could probably
>be attained just by typing in the ASN.1 specs.

I would be interested in a copy -- anyone prepared to fax
it to me even if its not online? Or is there an ECMA
equivalent number for this ANSI spec?

[...]
>I submit that we should get an X.435 message packer/unpacker package
>together with no transport mechanism bound in and then decide how to
>map that onto MIME. If it's exactly compatible, we should be able to
>catch up quickly.

HTTP is a bit ahead of MIME with the forms DTD here --
as HTTP uses MIME for messages (just does them interactively)
it shouldn't make a lot of difference.

>Adding encryption/signatures, etc. might give us an edge also, ala
>cypherpunks. Remailer techniques could even be used for
>non-repudiation, auditing, or anon support.
>
>What are the requirements for 'fast-batch'? SMTP is pretty fast,
>especially if you run a few parallel sessions (my experience is
>watching the wire with cslip to Internet).

Just watch HTTP.

>> EDI deliver will also have an interesting (IMHO) influence on IIN. The
>> network that can readily carry the commerce of the nation will be the
>> network that the government will want to see grow...

Aside from that, do we really want to draw the line and say that
a particular information item is commercial, and therfore
shough use one paradigm, set of protocols, etc, and another
is not really commercial so will use the same pardigm as
the web, news, and mail? Of course not. Such lines do not
exist. We want to put the features, such as security and
common formats, in in a totally orthogonal way, so that they
can all be blissfully used together. And it shouldn't be
difficult. :-)

>> Bob
>sdw
timbl

__________________________________________________________
Tim Berners-Lee timbl@info.cern.ch
World Wide Web development leader
CERN, CN Division Tel: +41(22)767 3755
1211 Geneva 23, Switzerland Fax: +41(22)767 7155