|
|
Netscape made the front page of the New York Times this morning. Netscape Communications Corporation has just gone public and listed shares with stock exchanges. Trading on Netscape stock began yesterday and traders went wild -- originally priced at $28 per share, it opened at $71 and climbed to a peak of $75 before noon. By the end of the day the price had fallen to about $58, but this was still up $30 -- over 100% -- from its original pricing. Netscape's chairman, James Clark, owns about 25% of the stock; his shares are now valued at a half BILLION dollars. Quite a showing for a company that is only 15 months old and has never yet turned a profit. The Times story indicates that this was the best opening day for a stock issue of this size in the history of Wall Street. I guess cyberspace isn't just for rocket scientists and nerds anymore.
14 responses total.
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?
i know someone involved in just that, or close.
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.)
Wall st. has never had to install and debug the beast... Mine is still getting its act together. The article in Dr. Dobbs on html hotspots is the sort of thing for the small minds of the investor. It would be nice if we had one style of browser, and netscape is as close as we can imagine a good engine today. I doubt microsoft has much more than a full terminal package, they always underdo every product they ship. Finally, I am happy that a sofware firm is getting credit and it would be nice if this lead to better products that last a long time and mature as we enter the future.
agora 99 linked to internet 95. Thanks to remmers for pointing this item out to me.
(Full terminal package? What exactly does that mean in this context?)
re 4:
What kind of trouble are you having. I have Netscape running just fine
under both Linux and Windows For Workgroups.
<gregc's jaw drops to the floor>
My gawd, that's awful code:
1.) It's NOT more portable, it makes the (bad)assumption that "rm" lives in
/bin on every system it's built on. Or even, that the system uses "rm"
to remove a file. Wanna bet on what happens if this were ported to VMS?
From a ANSI C library standpoint, unlink() is the defined portable method
because the C library is the transparent layer to hide whatever is
necasary to remove a file on a given system.
2.) It's NOT faster. unlink() is a system call. system() is a library function
that has to load a shell, which then has to load the rm command which
*then* calls unlink(). At a rough guess it's at least 100 times slower.
3.) He assumes the malloc always succeeds. Very Bad assumption.
4.) He assumes the system() always succeeds. Another bad assumption.
Part of my specialty is designing portable C code. The above makes me want
to slap the author silly.
(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.)
A lump of shit by any other name...
(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?")
No one said you had to be a good programmer to make a lot of money. Assuming they can find and hire good programmers, Netscape will live or die by their vision of what Web browsers and Web servers should do and how well they can sell that vision. All these people who bid up the Netscape stock are betting that they have the right vision and they can sell that vision. Time will tell.
Netscape is overpriced. I do not see a barrier to entry in that market.
fwiw - i have it on reasonable authority that the Oct AACS meeting will feature some one doing a dog-n-pony show with MSN - film at 11, i think.
Response not possible - You must register and login before posting.
|
|
- Backtalk version 1.3.30 - Copyright 1996-2006, Jan Wolter and Steve Weiss