To: Ann Janssen who wrote (70371 ) 12/30/1998 9:20:00 PM From: Mary Cluney Read Replies (1) | Respond to of 186894
Thanks Ann for the Article: >>>McHugh calls it "amusing" that Microsoft's Windows NT operating system gets so much press, since it can't handle as many users. "[Microsoft] hopes to be able to scale to 1,400 concurrent users in 1999, while we today handle 250,000 at a time," <<< I never doubted that IBM is great at marketing and most of the article is dominated by IBM marketing. It is not until the end (the last two sentences) that you start to get an inkling that something else is going on: >>> Even Gartner analyst Yefim Natis, a fan of CICS, sees sales eventually slowing, since IBM isn't attracting new customers, just outfitting current ones with expensive upgrades. He expects customers to increasingly combine CICS with other applications servers. If this prognosis is true, CICS may not be a long-term bet. Says Bill Coleman, CEO of BEA Systems, a competitor that sells applications servers for NT and Unix: "It's like somebody in 1905 saying if we just put 20 horses in front of this wagon it will beat your car forever." <<< First of all, I'm not going to bet against Bill Gates (as much as I personally dislike his software) - for MSFT to continue to grow at the current rate he will need to (he must) dominate this (mainframe) market just as he has dominated the desktop market. Secondly, IBM was always good at marketing but not realy all that great in delivering innovative new products and large systems that actually perform to marketing promises. As far as I can remember , going back more than 30 years, I have never known an IBM mainframe system that adequately delivered a system that performed as planned. It normally requires upgrading way before the system is implemented and the upgrades never end. I suppose that when you have committed a large sum of money - it is really difficult to cut off the spending and call the project a failure. There are too many people involved in making the decision and they invariably commit more money to the project to try to make it succeed. I have a Schwab account and I know how miserable the response time is. I just don't know how IBM gets away with it - and they have been doing it for over thirty years. Perhaps the truth is that these dinosaur type centralized systems will always fail because there is no real end to the demands that are placed on a centralized system. They are designed to fail. As far as I can tell, Intel themselves do not have a centralized system that can handle 250,000 concurrent users. Come to think of it - why do they want to put 250,000 users on one system and when that one system goes down - 250,000 workers are rendered helpless. Why not distribute the processing, distribute the DASD, and distribute the telecom load and then provide for plenty of redundancy. If Intel and Microsoft can successfully manage their own business without the need for a mainframe - then I would say that the days of the mainframe are indeed numbered. One more thing, Ann, you are going to have to talk Tom Osborne out of retirement - Nebraska football is jst not the same without him. Regards, Mary