SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : All About Sun Microsystems -- Ignore unavailable to you. Want to Upgrade?


To: Michael Watkins who wrote (9734)5/7/1998 11:45:00 PM
From: Bob Drzyzgula  Read Replies (2) | Respond to of 64865
 
The OS and the development tools are the same as are generally available in most Linux platforms; the development tools in particular are available for free for every Unix platform, including Solaris/SPARC. Corel Computer's contribution here is (a) to port it to a new platform, and (b) to use it in an a design experiment to make a turn-key end-user platform. Here you have a box without a shred of x86 technology running nothing from Microsoft, and yet you can plug it in, power it up and have a fully-functional machine that has the potential to do anything from be a simple web browser to run stochastic simulations. If perchance this thing were to sell well and they get the volumes up, you might imagine it on the shelf in Montgomery Ward's Electric Avenue for $395, including operating system, a dozen programming languages and a good stock of application software, including maybe Word Perfect and Quattro Pro. As a computer for kids in school, this would offer unprecedented functionality for such a price.

Since kids love to tinker with things like this, it could result in a signficant expansion in the number of people who learn how to program on a non-Microsoft-based platform (this is already happening with Linux in general, but making a cheap, Mom-and-Dad friendly box like this could take Linux in new directions). Since the source code for the compilers and the operating system and most application software would be included, there would be virtually no limit to the depth at which they can investigate. This is something that neither Sun nor Microsoft offer.

Since they have some control over the platform, Corel could have signifcantly better luck making Word Perfect stable and bug-free; they will even be able to modify the OS source code to make Word Perfect work better, without the trouble that Microsoft is getting themselves into... if the OS source is there for all to see it's kind of hard to argue that they are witholding information from competitors.

*If* they can make it a solid, stable platform for Word Perfect, I could also see law offices using it. Word Perfect on Windows is OK, but one does get rather tired of the crashes. It isn't possible right now to tell how much of this is because MS put Word Perfect killer code in their OS and how much is because Corel's code is just plain buggy. This platform could help to answer that question.

We all know that Corel is to the point of watching the last few drops of blood drip from their wrist. They are in deep, serious trouble financially and a good share of the problems are their own damned fault. Thus there is, of course a huge question mark hovering over all of this.

But what I'm seeing is also that the open source/free software world is giving a new avenue of attack for vanquished foes of Microsoft. Ray Noorda kind of started the trend with Caldera, buying up all the old Digital Research assets and tying it up with Linux; You can now get the DR-DOS kernal source for free from Caldera's web site, and they are using it as a base to develop a "kiosk" computer platform with a DOS-based web browser and TCP/IP stack. It is also integrated into Caldera's Linux distribution, so that if you type "DOS" at the Unix command prompt it starts up an X86 virtual machine running DR-DOS. (You get the virtual machine with other versions of Linux but you have to supply your own copy of DOS). They've also released the source code and compiled binaries for CP/M (so why not a CP/M-based microwave?). Of course Netscape joined up with the Mozilla source, and Anderseen yammering on about how Mozilla and Linux are going to slay the Bill-dragon. Now Cowpland is piling on, selling a unique Linux hardware/software platform and porting all the Corel applications.

It's easy to dismiss this as just a bunch of stubborn mules who refuse to give up the fight. But Caldera's OSs are selling reasonably well, and they are expanding their embedded applications group. You can get the DR-DOS source together with an SDK for building embedded systems with DOS in ROM or Flash memory for under $1000, including a bunch of run-time royalties. Netscape's Mozilla project has captured hundreds if not thousands of developers world-wide to start pounding out a better browser. Another company, Cygnus Solutions, has ported the development tools to a bunch of embedded platforms, including microcontrollers like the Hitachi H8.

Once upon a time, companies like Novell and Corel thought that what it would take to defeat Microsoft was another Microsoft, with a soup-to-nuts portfolio of operating systems and/or applications for the PC platform. What Sun knows is that Microsoft owns the PC platform and will garrote anyone who tries to take it from them, which is in and of itself a good enough reason for them to stay away from building an X86-based machine. Novell and Corel have learned that lesson the hard way.

But now, a couple of decades of cooperative development over the Internet by some people even more obsessed than the likes of Noorda and Cowpland have created what many consider to be the ultimate weapon. An operating system so portable that it can be hosted on a new architecture with a few months of work, so lightweight that it can run on ten-year-old hardware (it is even being ported to the Palm Pilot) but so sophisticated that it routinely bests NT by most measures. An API that is not perfect but is intimately understood and deeply loved by millions of developers. And most of all, a totally novel idea: It cannot be owned. Linux cannot be owned by anyone, not even Linus. The GNU General Public License is viral... anything it touches is un-ownable. Thus, anything that Caldera or Corel do to the Linux kernal must be shared with the world, for free and in the form of source code. Beyond patented inventions inside Corel's Linux-based hardware or trademarks on the name, anyone can take Corel's work on the OS, build a machine based on the DEC StrongARM, and sell it on the same shelf as Corel's unit. Caldera did much of the work to add Novell's IPX protocol to the Linux kernal and to improve the SMP capabilities. All that stuff is now part of everyone's Linux. Red Hat developed a great package manager called RPM that made their distribution stand out, but it was released under the GPL and now several Linux vendors use it (Caldera and S.u.S.E, for example) and it has been ported to most other versions of Unix.

In the end, by sharing all of this stuff, the Linux developers have avoided the bickering and divisiveness that has plagued the remainder of the Unix industry. Yes there are disputes and differences of opinion, but they are healthy disputes and in the end usually one or two alternatives rise to the top. Where would Unix be today if, for example, Sun, DEC, HP and SCO cooperated at that level?

No, Linux isn't perfect. It has a pile of kinks and rough edges. But it is getting better on a daily basis (take a look at road-kill.freshmeat.net -- seriously, its not what it sounds like). But there is one huge difference between Linux and NT: Linux can be understood, NT cannot. What you can get out of Linux is in direct proportion to the amount of effort you put into it. Virtually anything that is broken can be fixed if someone cares enough. NT is an infinite labor and time sink, and many things will never be fixed no matter how annoying they are, for the simple reason that Microsoft likes them that way.

For grins, read Henry Petroski's "The Evolution of Useful Things" ($10.40 at Amazon.com). Petroski's central thesis is, in effect, that imperfection is the mother of invention. People invent things because the things that preceded them didn't work right. Man's perpetual state of dissatisfaction with the world is what drives us to create and to improve. Linux is, in my opinion, a fabulous example of exactly what he is talking about. If you don't like something about it, you fix it. If it doesn't bother you much, you leave it alone and find something else to do. From this perspective, the reason that NT is so bad an operating system is that so few people have an opportunity to fix the things that are broken. Most of us can only complain, and Microsoft just doesn't care as long as we keep buying it.