Re:request for new forms submission consensus

waterbug@epims1.gsfc.nasa.gov (Steve Waterbury)
X-Delivered: at request of secret on dxcern.cern.ch
Date: Mon, 11 Oct 93 09:19:33 EDT
From: waterbug@epims1.gsfc.nasa.gov (Steve Waterbury)
Message-id: <9310111319.AA06924@epims1.gsfc.nasa.gov>
To: www-talk@nxoc01.cern.ch
Subject: Re:request for new forms submission consensus


Marc writes, 

> All things being equal, I think I'd be happy with the current encoding
> method (name=value&name=value with escaping) just slapped into the
> body of an HTTP/1.0 'SUBMIT' request -- keeps things simple and
> straightforward, avoids creating a new syntax, is known to properly
> handle escaping issues, and servers and code already exist for
> decoding/handling it.
> 
> To be honest, the idea of encoding it in some kind of SGML format does
> not excite me in the least -- what would be the added value?

I think that was my reality check I heard.  

I definitely agree with the "KISS" principle espoused here. 
The stuff I was talking about in my previous message about 
composing requests as globs of HTML with non-HTML tags thrown 
in for logical tagging would unnecessarily complicate what is 
right now an elegant and effective protocol, the 'SUBMIT' 
request.  I take it back!

The application I want to make that uses forms to assemble 
structured documents could still be implemented within this 
context by the use of smart servers to handle the stuff 
submitted with the existing SUBMIT protocol (and it makes 
more sense to direct this task to a special server anyway, 
rather than to gum up what is working now).  

(meanwhile, back on Planet of the Data...)

I'm still interested in more reality checks ... on the idea 
of logical tags, "index" servers (basically servers that would 
maintain the indexes for the implicit database that is the 
whole contents of the Web), "meta-data" servers, etc.  Is that 
too much of a traditional database thing, or do some of you 
have applications that need that stuff?  Does the idea of 
putting in non-HTML logical tags (invisible to HTML browsers 
but visible to indexing filters) sound reasonable or does it 
give people the creeps?  

It would have to be decided how the indexes would be maintained:  
my idea is that it would be at the "Web object" author's 
initiative.  E.g., the author could send the URL to the 
indexing server (or a meta-server that knows which index 
servers to contact) for "registration" and indexing.  

Is this an appropriate forum for this topic, or should I 
take this application-oriented noise to c.i.w3?  
(Go ahead, be brutally honest ... I can take it!)

Steve Waterbury.