In-Reply-To: <199503282045.OAA18605@riverside.mr.net>
Message-ID: 
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII



On Tue, 28 Mar 1995, Jeremy D. Allaire wrote:

> In response to this general sentiment (expressed by many folks in many forums,
> though typically here) that there is not "One True Way", etc, a few comments.

It is fun to see the "Religeous Debates" that get started when one discusses
the Operating Systems or Networking Protocols.  Engineering is only secondary
to the agreement and trust that Vendors, Corporate Customers, and Individual
users have for the technology.  CLNS is superior to IP in many ways, but
IP has the agreement.

Why didn't SNA or IPX take the market?  IBM made several changes in protocol
that were dependent on IBM applications.  LU-2 was dependent on 3270 display
terminals (or emulators).  Even when they came out with APPC, IBM discouraged
basic conversations in favor of mapped conversations (Implying CICS 
dependencies).

Novell eventually released a reference implementation of IPX to BYU, but
there are many restrictions and limitations on "reverse engineering" IPX
applications and Netware Network management protocols.

Still, fueled by several billion dollars in advertizing revenue, the 
technical press touted the superiority of the proprietary products,
while the "rodents" of MIT, Berkely, CMU, and later SUN, HP, and Soulborne
quietly ground out the TCP/IP, UNIX, X11, and GPL standards and 
implementations.  Over 1 million of the top engineering people in the
world have been contributing regularly as part of their normal work.

> While no one is looking for One True Way, they are looking to capitalize on
> major trends in the online industry, and are willing to grow with that.

Every year or so, someone tries to take ownership and "corner the market"
on "Open Systems Standards", by adding proprietary extensions.  
Unfortunately, the nature of the Internet is that when someone tries to
"Own" a particular enhancement, others come up with GPL enhancements.

Sun and HP have made a very successful business out of taking the standard
applications and bundling them up into an easy to use comprehensive package,
with minimal proprietary extensions.  Even their enhancments (NFS, RPC,...)
are put out in GPL for those who want to include it.

IBM and DEC are beginning to join the fold.  The 10-15 Linux distributions
are quickly turning PCs into high powered workstations.  The "Open Systems
Standards" are quietly growing at about 15-20%/month.  About the same
way that the Internet grew in the 1991-1994 period, and for the same reasons.

> Now, for whatever reasons (I don't want to articulate it again, as it has been
> done frequently and elequently by others) the Web has emerged as the de facto
> client/server system for online information services.  That's a fact.  There are
> better platforms from a technological point of view, and, perhaps even from
> a administrative/supporting model point of view (see Interchange and MSN).

> However, the openness and widespread activity by thousands of software companies
> is driving things far faster than these proprietary systems can keep up with.

The openness is driving the industry, but the proprietary systems vendors 
are feeding the press.  If Bill Gates spends $2 billion on advertizing, with
the understanding that articles favorable to the Internet/Web are damaging
to his business, the press won't bite the hand that feeds it.

> Incidentally, there is an interesting comparision in the OS business.  For the
> past two years there is little doubt in anyone who seriously considers the
> matter
> that OS/2 is a superior operating system to Windows, and even, perhaps to
> Windows95.

IBM killed OS/2 when they tried to tie it to MicroChannel back in 1988.  
Since then they have lost the trust of vendors and developers.  In addition,
the entire interface was incompatible with other standards that were 
available or being developed at the time.  IBM didn't really promote OS/2
1.x, and OS/2-2.0 had some technical problems.  Until WARP, the thinking was
that they could sell a crippled system and rope 

> It's been around for quite some time, yet any rational software developer
> knows that
> it is a relatively dead system with the exception of vertical market
> applications.

One of the interesting things about OS/2 was that it is possible to get
the standard GNU compilers and tools to work with it.  Of course, on
IBM's Power/PCs, they are putting AIX on it as the operating system of 
choice.

> This has to do with a lot of factors, including the openness of DOS and
> Windows and the
> huge installed base of users.  Similar principles apply in the
> Internet/Online world,
> and to miss them is to miss the boat completely.

Dos is OPEN?  Windows is OPEN?  Ever tried to trace a memory leak into
a foundation class?  Ever tried to figure out why your screen locks up when
you cover the abort dialogue?

There are two bases here.  There are the 60+ million 386 or better 32 bit
CPU's which are currently running in 16 bit mode, and there are the 60 
million Windows Users who don't want to spend $1400 for RAM to run NT
and are tired of waiting for a real multitasking operating system that
can actually handle communications via modem.  They want to upgrade to
something better, don't want to spend too much, and have certain minimal
requirements.

Microsoft has kept UNIX out of the market for 15 years through 
advertizing and media hype.  When Bill Gates announced Windows (vaporware
at the time) Sun and X11 were multitasking windows on UNIX and passing
pixmaps around the Internet.  In fact, UNIX has been 2 to 5 years ahead 
of Microsoft for at least 10 years.

> I'll give Interchange and MSN high marks in terms of technology (certainly,
> for now,
> superior to the Web in most respects), but I have yet to hear a cognizant
> and compelling
> argument for them as lasting technologies, though with MSN I would accept

If Bill Gates is willing to spend $5 Billion in advertizing to get the
media to sensationalize every hacker or intruder, MSN can probably keep
the internet at bay long enough to kill any potential commercial competitors.
They will start by giving it away for $50/copy, then they will raise the
price to $100/copy until the user base reaches critical mass, then they
will switch to "Perceived Value Pricing" of $400/copy.  Just like they
did for Word, Excel, PowerPoint, Project, Access, and Visual C++...

Meanwhile, the "Rodents" will be lunching on dinosaur eggs (Andrew, Linux...)
User supported software starts out free, but you are willing to pay for
subscriptions to quarterly updates.  I get better support on the "Unsupported
Software" than I got from Microsoft, Sun, or even IBM.

> the idea that
> it will eventually integrate with the Web or be a real compliment of sorts.

Yes, and Microsoft Word will do WordPerfect, RTF, and Wordstar formats.
For an extra $600, you can even get to do HTML.  Of course, you can't get
your document to use all the features of word and then save the document in
a foreign format.  Oh Well.  If history repeats itself, Microsoft will
make a document "Converter" that will take every Web Page you grab, convert
it to some proprietary Microsoft format, and then store it on your machine
so that you can ONLY read it with the Microsoft Product.

We already know Microsoft is opening it's own News Service, which should
eliminate the need for most of the Wire services, local newspapers, and
on-line lists.  Microsoft will take contributions, but will probably charge 
a fee for the service.

Wanna buy a bridge?

> Jeremy Allaire

	Rex Ballard


From rballard@cnj.digex.net Wed Apr  5 10:56:24 1995
Status: O
X-Status: