To: rudedog who wrote (45649 ) 5/30/2000 4:20:00 PM From: Andy Thomas Read Replies (1) | Respond to of 74651
I don't think Silverberg was responsible for "Cairo." Are you sure you've got your facts straight on this one? I will tell you that the NT guys for the most part had nothing but disdain for the win3.x and win95 products from the get-go, even though these "PSG/PSD" products were the company's cash cow at the time. Office eclipsed personal systems (by then probably IPTD) in revenues but only a couple of years later. When win95 was completed, some of the team were taken and put under winnt. Though some integration work between win and nt had taken place earlier, I think these guys were supposed to "finish Cairo." The winnt guys treated the win95 newcomers badly, at least from what I've heard. Part of their effort was the WDM (Windows Death March) driver model which came out for Win98. The rest of the win95 team was made into the IE team and Silverberg managed to get MSFT programming languages under his umberella. So while I don't know the exact facts on Cairo (except to say I don't recall it as having been Silverberg's main thrust), I will mention that MSFT's first Java JIT was done by Silverberg's (and Slivka's if memory serves) people. The funny thing was that MSFT's first Java JIT blew the doors off of Sun's and was more compliant with the spec than anything Netscape had ever done with Java. NSCP continued to write away from the spec but Sun never complained. As time went on Sun did complain, even though it was more supportive of the spec than NSCP's. That Java JIT was probably written by another small team, much the way IE3 was written. I don't imagine Silverberg was too happy with the way his win95 team was treated by Gates et al after the release of the product. You had to have been there to have appreciated Silverberg's attention to detail. He was testing DOS comm apps while win95 was being invented in order to make sure the serial communications were acceptable. He found a couple of bugs himself. Most VPs are not detail-oriented so when one such as Silverberg comes upon a given scene, people take notice. I don't think the MSFT detractors are often willing to admit what an incredible piece of work win95 was - it tried to be everything to everyone and did a darned good job of it. It was however slowed by Exchange which was written by another, less-capable group. We used to argue in management meetings about getting the exchange people to take their code out of win95 and sell it seperately, so win95 could stand out on its own as being very quick... at least after bootup. I had an NFL simulator and it took forever to run through the league statistics program through plain dos. Up until then the assumption had been that anything run from a dos box inside of windows would be slower than anything run in native dos. With win95 that changed... dos apps doing extensive hard disk work were actually much faster running from a dos box under win95 than they were running natively under dos. I guess it's all water under the bridge so I'll stop rambling... Andy