Harry,
< Hence the demise of the monopoly of the mainframe which was not flexible enough to handle the needs of individual users and the needs of the enterprise. >
I would agree that the MF *environment* was not flexible enough (at that time), I would disagree that the MF itself was not flexible enough.
I watched many MIS types in management positions resist the PC paradigm. They refused to let the apps people revamp the apps to take advantage of the *new paradigm*. The apps could have been written, and have since been written that could have kept the MF much more in the "main stream" than it currenlty is. This would most probably have slowed down the growth & advancement of the PC.
Today, the MF cannot provide what the PC is providing to the individual user. In fact many PCs today have more computing & storage capacity than many of the MFs in existence when this revolution kicked off.
As to the NC, it may *solve* some of the problems that currently face large companies but it will at the same time introduce other problems that, IMVHO, will be even worse. The bandwidth will have to be greatly expaded so that *all* software can be downloaded from the servers. Also, the question arrises of whether or not the server can handle the increased load. With just those two considerations, you can eat up in upgrades what you *might* save in maintenance. There are even more problems of which I can think that would increase the debit to the bottom line if companies wish to go the NC route.
There are places for the NC. The most prevalent are those which dedicated terminals are used today. Placing them there will actually reduce much cost as you do not have to run a dedicated line from a MF to them. They can attach to the network, get the client software then talk to the MF. Shortly before I left the Military, we implemented PCs all over the place and were able to remove several *tons* of dedicated cabling. This simplified the network as well as lowered the cost of maintenance.
Just my 2› worth,
dmg |
