To: SunSpot who wrote (39029 ) 3/5/2000 11:29:00 AM From: rudedog Read Replies (1) | Respond to of 74651
SunSpot - I think you are proposing a kind of straw man argument. re: the first company in a technology often is the one to win. Apple did not "invent" the GUI - they just built the one that defined the category. MSFT has never built their business on being the first in a technology. DOS was a warmed - over CP/M clone that MSFT bought. Windows was a "me too" GUI which took 10 years to achieve real usability. SQL Server was originally a port of Sybase SQL Server. The GUI work in the Office apps had been done earlier by others. What MSFT has ALWAYS done successfully is keep plugging until they get the combination of ingredients that works. It is a "fast follower" strategy, and not even all that fast. That has been the whole history of MSFT development, and it has obviously paid off pretty handsomely for them. So the notion that MSFT now needs to be first with hot new technology, when that has never been important to their success in the past, seems a little off-base. Let's look at your points with the "fast follower" model in mind. - Wireless. As the convergence standards emerge, MSFT has only to enable the ones that are working in the market. They have placed some bets through investments and partnerships. Some of those will pay off. - TCP/IP. Here again, "good enough" is good enough. I can detect little difference in TCP/IP overhead or performance between web servers based on Solaris, Linux, and NT - and at the end of the day performance and efficiency are what customers buy. - Handheld devices. MSFT is admittedly weak there, but has IMO correctly decided to concentrate on the business connectivity piece, where they have a much better story than Palm. I believe they should, and will, develop different lines to support lightweight handhelds and larger CE devices, and will win in the commercial space. Commercial buyers outspend consumers by a huge margin. - Network administration. - agree this has not been a focus for MSFT and they need to do a lot more if they intend to move up in the enterprise... - ETHERNET. I have implemented full duplex Ethernet stacks on NT going back several years - no great trick. I can also get 85% of physical pipe capacity on either 10MB or 100MB ethernet using just standard MSFT stuff - better than I can do with any other combination of hardware and software. What benefit would there be to changing that? - Databases - Oracle has a great, well established product which is getting a little long in the tooth and does not do partitioning, parallel queries or clustering very well. DB2 is architecturally leading, does all of the above very well, and has almost no presence outside of "Big Blue" accounts. MS SQL demonstrated the ability to deliver scalability with support for parallel query execution against a partitioned shared-nothing architecture which blew away Oracle and DB2 - a solid technical achievement - and did it at cost levels a fraction of the other two. I know the database business well, and you're smoking dope on that one. MS SQL is already broadly used and will be a lot more prevalent when the next generation hits the street - after years of being a "toy" database it has just about made it to adulthood. - Crypto-technology - an example here would help... Windows 2000 is probably the moment for Microsoft, where operating systems for PCs gives the most revenue ever and forever. In this statement you are echoing what top MSFT executives have been saying. But those revenues are not likely to decline any time soon, despite the trend. Sales of PCs and Intel-based servers will continue to increase for a while, and they will mostly run the MSFT stack...