Re: Question: volume generated by WWW vs. Gopher

Rich Wiggins <>
Message-id: <>
Date:         Thu, 21 Oct 93 00:36:57 EDT
From: Rich Wiggins <>
Subject:      Re: Question: volume generated by WWW vs. Gopher
To: Gabriela Segarra <>,
In-reply-to: Your message of Tue, 19 Oct 93 15:30:52 CDT


It depends entirely on what you do with the Web versus Gopher.

Our root Gopher menu is maybe 3K in size; our home page for WWW is about
the same, including our university logo in color. But due to the very
richness of WWW, and the liberal use of images within documents in
particular, many Web documents are large; a single click can cause
several hundred K bytes (or more) to be sent down the pipe. The full
motion weather satellite loops delivered as MPEGs by my colleague Chuck
Henrich can be a meg or larger.

So it's really a question of what you put under the Web. In a sense, if
you don't do anything interesting you won't have a problem. :-)
Seriously, a Webified Gopher tree shouldn't cause any more load than
delivery via Gopher. Conversely, you could offer large files via a Gopher
(images, software libraries) and generate lots of load that way.

You might look at John Franks' Gn server as an interesting bridge
between the two worlds.

/Rich Wiggins, Gopher Coordinator, Michigan State U

>To the WWW people, I have a question about the volume of messages
>generated by the WWW. Our University has been running a gopher server
>for some time and we have now a brand new WWW server. The people
>running the gopher server are reluctant to this new initiative and are
>asking me wether the WWW transactions will increase the traffic on our
>almost saturated network. So, I was wondering if somebody has any
>numbers to help me defend the WWW, possibly comparing the amount of
>data transferred in a transaction by gopher and the WWW. I think it
>should not make any difference, but still ... I would appreciate any
>information or pointer to the proper place. Thanks, Gaby