Re: solidifying a really neat forms hack

Tony Sanders <sanders@bsdi.com>

Mail folder: WWW Talk Oct 93-present
Next message: Peter Lister, Cranfield Computer Centre: "File size indication"
Previous message: Mr. Tom Cozzolino: "Request: Oracle/Sybase dirt-simple example?"

Errors-To: sanders@bsdi.com
Errors-To: sanders@bsdi.com
Message-id: <199311241630.KAA02555@austin.BSDI.COM>
To: www-talk@nxoc01.cern.ch
Subject: Re: solidifying a really neat forms hack 
In-Reply-To: Bert Bos's message of Wed, 24 Nov 1993 12:35:28 +0100.
Errors-To: sanders@bsdi.com
Reply-To: sanders@bsdi.com
Organization: Berkeley Software Design, Inc.
Date: Wed, 24 Nov 1993 10:30:00 -0600
From: Tony Sanders <sanders@bsdi.com>

> Marc Andreessen writes:
>  |NAME="isindex" result in query strings that look like:
>  |        action?query
>  |...and NOT:
>  |        action?isindex=query
...
Bert Bos replies:
> I agree, but it can be independent of the implementation of ISINDEX.
> E.g., specify that the "=" is omitted when the INPUT NAME attribute is
> empty or omitted:

Exactly, all Marc is really saying is that a form with a single
NAME="isindex" should generate url?foo instead of url?isindex=foo.
Browsers are still free to do whatever they like with <ISINDEX>.

However, let me also say that I personally would rather see the query
encoding specified in a generic way rather than special case the NAME
(I think we did this already).

I did my little demo to show what you could do if you could query multiple
existing databases from a single document.  I would prefer that we don't
special case NAME="isindex".

Just for the record:  ISMAP was a hack also, the proper HTTP/1.0 way of
doing this is to serve the image with attribute "Public: SPACEJUMP".
However, when all servers were HTTP/0.9 there was no other choice.

--sanders