🚀 go-pugleaf

On Fri, 22 May 1998 00:09:34, mstaben@My.House (Matthew S. Staben) 
wrote:

>#There are real serious implications to this.  For decades now the 
>#process of progress in the computer world has been crash and burn.  Each
>#project started from scratch and was built up and then the wrecking ball
>#would come in before the next project would take over.  
>#
>#Trillions of dollars have been wasted in the process of creating 
>#incompatibilities between old and new systems. 
> 
>I wouldn't go so far to say that.  Microsoft, yes, is guilty of attempting to
>own the world; look at its foray into altering forever the internet protocols
>and standards into a cryptic, Microsoft-only-knows-how, stolen infrastructure.

My opinion is that Microsoft's continued dominance rests on two factors,
both of which are necessary.  The first is they must be the standard 
setter.  Open standards mean head to head competition and that means a 
level playing field that a number of companies will be able to provide 
significantly new and exciting technologies in.

The second is that they need to continue to create incompatibilities.  
If even Microsoft standards become standardized for too long, then there
will be competition there.  Just like the Win 3.1 emulators that started
to make appearances.  Stay too long on a standard and it will get 
cloned.

The net result of this game of tag is trillions in costs for users.
 
>#Eventually, software will need to be standardized.  Now is as good of a 
>#time as any and building upon the current Pentium compatibilities would 
>#make sense.  It is quite possible that unless Merced actually 
>#demonstrates clearly superior abilities to run todays software it will 
>#go the way of other failed branches in the past like the PPC.  It is 
>#easy to criticize the lack of support for a new proposed standard from a
>#company like Intel, but Intel is losing marketshare in significant areas
>#and their ability to unilaterally move the industry in any particular 
>#direction is diminishing.  
> 
>I doubt the PPC is a failed branch.  It is enjoined with MicroChannel, and 
>handling a lot of your own money.

Failed branch as a desktop client or a client/server solution would be 
more appropriate statement.  That is what is being proposed by Steven as
the future for Merced.
 
>#Further, technologies like Java is further eroding the motivation of 
>#customers to chase the hardware standard.  Java is incredible in that it
>#opens possibilities where companies can build software applications that
>#will be timeless in respect to hardware.  This will potentially save our
>#institutions trillions and trillions of dollars over the next few 
>#decades if Microsoft is not allowed to fragment this standard.  
> 
>Absolutely; and Microsoft must not be allowed to fragment Java, or 
>cause a progressive leap into a brand-new computerized world to
>fizzle into an evily inspired caricature of spent impotence.  It is a
>great danger to lose the initiative, especially so in a nuclear world.
> 
>#Of course this point of view is being supported by Intel as well.  About
>#the only resistance to this is Microsoft and some of their minions who 
>#profit handsomely off of minor upgrades to necessary software products. 
>#I read the Intel vision of the future just a couple of months ago and 
>#Intel does not project Merced as anything but a server product.  Of 
>#course that could all change but I think we will have to see the 
>#technology work before anybody can come up with a reasonable prediction.
># Of course Microsoft would embrace it as they need major help at the 
>#high end of Wintel compatibility to compete with the AIX's and Unix's of
>#the world.
> 
>Can't add to this, except I'd like to say that a Merced would be a nice
>thing to talk about at the water cooler.  
> 
>[... deleted some accolades for IBM's consulting business ...]
> 
>[... WinCe & Merced? ... Snip Snip ...]
> 
>[... And some Tiers to Market ... Snip Snip ...]
> 
>[... Argh, and some Spiel on Microsoft's Greatness :) ... Snip Snip ...]
> 
>#The real questions still have to be answered.  Will 1999 Merceds run 
>#software that people are buying faster than 1999 Pentiums from AMD and 
>#Cyrix?  Merced has probably had its Intel architecture designed for some
>#time and the question is will it crank out Wintel compatible software 
>#like a 500mhz Pentium for the same cost?
> 
>I think Merced's going to be another flashy offering that really isn't as hot
>of an upgrade as people are being led to believe.  What's going to be
>hot is what goes on inside.  It ain't going to be Intel.

I think the next decade of improvements is going to come in the form of 
more efficient software and system designs.  CPUs have completely 
outstripped system design for desktop users.  With 500mhz pentiums just 
around the corner and the industry still struggling to get a 100mhz bus 
design working well, there would seem to be room for a lot of 
improvement in system design.  


Bob O - Computing for fun