response 1 of 14:

Aug 10 21:49 UTC 1995

At an Ann Arbor Computer Society meeting a couple of years ago,
the speaker claimed that, outside of IBM, the computer industry
had yet to return a significant profit to investors.  Oh, there were 
always "bubbles", where somebody made a killing selling stock in 
a hot company at just the right time, but over the long haul
computer firms had been a bad investment, and someday
the investors were going to catch on to this.

If I remember correctly, the theory was that products were not 
lasting long enough in the marketplace.  (Remember when Gopher
was the hot new Internet access tool?  How about Mosaic? 
These weren't commercial products, but they illustrate the 
point.)

If someone comes out with a tool which eclipses Netscape, and 
starts giving it away, what's all that Netscape stock going to 
be worth?

response 3 of 14:

Aug 11 05:57 UTC 1995

Well, MicroSquish has certainly done well for itself.  I mean, even
at its peak they could have bought NetScrape without even putting 
a terribly big dent into their ready cash.

Right now, NetScape has the potential to make money on two fronts:

- The MS-Windows browser market
- The server market, particularly for online commerce

Those are the only places.  Very few Unix users pay for their browsers,
and the Mac market is also somewhat small and heavily slanted towards
educational institutions that currently get the product for free.  Their
secure NNTP server seems unlikely to displace Lotus Notes as a secure
groupware product.

With its announcement of support for Web stuff, it seems likely that
MicroSoft will further erode the already not terribly strong position
NetScape has in the Windows market.  I mean, sure a lot of people use
their browser, but few of them pay for it.

On the server market, the market is going to get commoditized.  The
software in question is simply not that complicated; an average CS
undergrad given some time could do it and, in fact, did.  Anybody
who thinks the technical expertise that left the Mosaic project at NCSA
to found NutScrape was particualrly good should look at the code for
Mosaic for UNIX.  For example, here is how Marc Andreessen deletes a
tempfile:

  command = (char *) malloc(strlen(filename) + 16);
  sprintf(command, "/bin/rm %s", filename);
  system(command);

He claims it's better to do this than use the obvious unlink() because
it's faster and more portable.  If anybody here who knows anything about
UNIX agrees with this assessment, please let me know so I may drop my
jaw to the floor.

So, competitors (I have one in particular in mind :-) will start coming
up with products that do what people want from NetScape for a lot less
money and under much more reasonable terms vis-a-vis support and the
like.  Once products are commoditized and big margins go away, the free
ride ends.  The question is whether NetScape will have been able to leap
forward with something new by then; maybe it can.  It's certainly
managed to ride the hype well; as with MicroSoft, most people don't
realize that there aren't really any features NetScape added that other
people hadn't already implemented elsewhere (and usually better.)  They
just put them together into a slick package and sold it effectively to
people who may or may not have known what else exists.  I mean, how many
people think MicroSoft invested compressed filesystems?  How many people
think NetScape invented inlined JPEGs?

The proof of the pudding will be NetScape's value around, say,
Christmas, after the hype has worn off, Windows95 and MSN are out of the
blocks and doing whatever they're going to do, and the many shares of
stock held by people in on the ground floor of the big N start trickling
out into the market.

(This is also based on the supposition that the Internet won't turn out
to be like CB Radio, where after a while the fad dies out and most
people leave the medium to the few who started using it in the first
place.  There are ways in which I think that would be a Good Thing.)

(Disclaimer:  I speak for myself only, blah blah yada yada.)

response 5 of 14:

Aug 11 15:08 UTC 1995

agora 99 linked to internet 95.  Thanks to remmers for pointing this item
out to me.

response 7 of 14:

Aug 12 00:06 UTC 1995

re 4:
        What kind of trouble are you having.  I have Netscape running just fine
under both Linux and Windows For Workgroups.

response 9 of 14:

Aug 12 04:14 UTC 1995

(Actually, he put an & after the /bin/rm, so it would be faster by running in
the background, except that the fork and exec for doing a system call and 
running a shell and all that is way slower than just doing the unlink would
have been.  He also does check for the system call to fail, except since
it's in the background he can't really check the success of the removal
anyway.  I just simplified it to post it here.)

response 11 of 14:

Aug 12 05:40 UTC 1995

(Oh yeah, I think he also forgot to free() the memory after he was done with
it.  Fun fun.  Fortunately, as I understand it he spends most of his
time being a generic media darling and crud and doesn't write actual
code any more.  Some people said they thought his newfound fame was
going to his head, but he brushed aside their concerns and rebutted:
"Who are you to call me arrogant?")

response 13 of 14:

Aug 14 13:55 UTC 1995

Netscape is overpriced. I do not see a barrier to entry in that market.