Re: Netscape v NCSA, Progress?

Daniel W. Connolly (connolly@hal.com)
Tue, 18 Oct 1994 09:45:29 +0100

In message <9410180615.AA00655@ua.MIT.EDU>, solman@MIT.EDU writes:
>You know, mcom has just released a kicking browser to the web community for
>FREE.

Not so fast... this is not FREE in the typical internet sense of the
word "free" -- the vast majority of the intellectual property
developed at Mcom is still inside Mcom. What we got was a few
megabytes of binaries to play with.

Nobody else can patch, enhance, maintain, etc. that code. They didn't
give away much, in the large scheme of things.

Time will tell if they are willing to "give away" as much support as,
say, NCSA or UKansas or CERN has given. It wouldn't make good business
sense to do so, if you ask me.

I don't mean to complain, just to clarify. If the beta release of
netscape solves problems for you, so much the better. But it doesn't
get me anywhere I couldn't get before. (Now if there were a Linux
version that worked over term... :-)

>I think its reasonable for mcom to claim as payment control over the
>protocols used by Netscape.

If they did that, they had better not call it HTML or WWW. They
can take their marbles and play elsewhere, sure. But I don't think
that's in anybody's best interest.

> The fact of the matter is that mcom would
>be perfectly justified in setting the standards de facto, and using
>this control to take over the commercial web server business.

I strongly disagree. Mcom diserves credit for quick turn-around for
what looks like a nice product. But they didn't invent WWW, and they
don't own it.

TimBL conceived of WWW, but the original ideas have been around for
years: look at Ted Nelson, Charles Englebart, and the rest of the
HyperMedia researchers, plus the internet researchers, and the list
goes on and on. Had all these folks done their research behind closed
doors, Mosaic never would have existed.

>Its not showing up at standard setting meetings that entitles you
>to set standards, its writing code that implements them.

I think this is a short-sighted viewpoint. Implementations come and
go. Specifications -- good ones anyway -- stick around. Look at
RFC822. How many implementations do you suppose there have been?
Do you suppose we've seen the last one?

Dan