Re: an idea

Ryan <ryan@enteract.com> wrote:=20
> I'm not a http/html guru, so please excuse me if something like this
> already exists, or has been proposed, or is just stupid...
> anyway, it seems netscape & mie always download images (or the whole
> document) even if it's unchanged..  couldn't a CRC type thing be used,
> where the client is told the CRC of the file and can use that to
> determine if it needs to GET it..
> ie, the server could insert a CRC tag
> <img src="http://site.com/junk.gif" crc=ff31cd5f >
> and if the browser has ANY file in cache with that CRC it can skip it

The idea is good but a new tag is not needed. A better implementation would be
for the client to store the computed CRC in its cash directory, and the HTTP
protocol to be extended so that the server MAY return the CRC. Then the client
can decide to show the cashed image if the CRC was returned and is the same.

This extension is compatible, because the client will ignore the CRC if it does
not support it.

Jonathan Rosenne
JR Consulting
P O Box 33641, Tel Aviv, Israel
Phone: +972 50 246 522   Fax: +972 9 56 73 53
http://ourworld.compuserve.com/homepages/Jonathan_Rosenne

Received on Sunday, 22 September 1996 11:40:38 UTC