Re: HTML Streaming

Peter Flynn (pflynn@imbolc.ucc.ie)
29 Aug 1997 20:30:03 +0100


Date: 29 Aug 1997 20:30:03 +0100
From: Peter Flynn <pflynn@imbolc.ucc.ie>
To: www-html@w3.org
Message-id: <199708291930.UAA19388@imbolc.ucc.ie>
Subject: Re: HTML Streaming

Albert writes:

>I have never heard of a bot that truncates a HTML file at the download of 
>the keyword tag. 

I don't think any of the big public indexes do that but several small
in-house ones do, where the use of META with Keywords is enforced.
 
>DTD. The entire language would need to be reviewed to allow each tag to 
>stream. 

I think what I don't understand is what needs to be done to it. What do
you have to do to <P> to "make it stream". By streaming, do you mean
constant refreshing of the data? The term is usually used wrt realtime 
data like audio or video, not serial data like text.

>Frankly, I don't think it is that big of a problem. I think the HTML 
>file can be streamed with the addition of tags that give a description of 
>the HTML file. The colgroup and other individual descriptions meant to 
>stream an individual tag would not be required. I don't think protocols for 
>the organization of tags, in the head or body, would do as much harm as you 
>imply.

I didn't mean to imply they'd do any harm, just I still don't understand
what you need to do to the file structure. Can you give examples of what
you'd need to add to make it stream?

///Peter