Date: Tue, 22 Mar 2005 03:32:09 -0800 From: "Ted Mittelstaedt" <tedm@toybox.placo.com> To: <freebsd-questions@freebsd.org> Subject: RE: MS Exchange server on FreeBSD? Message-ID: <LOBBIFDAGNMAMLGJJCKNEEODFAAA.tedm@toybox.placo.com> In-Reply-To: <1344929422.20050322102158@wanadoo.fr>
next in thread | previous in thread | raw e-mail | index | archive | help
owner-freebsd-questions@freebsd.org wrote: > Ted Mittelstaedt writes: > >> There is a third option. Microsoft can simply quite releasing new >> versions of it's established products and go to work creating new >> products that people would want to buy. > > That business model doesn't work, which is why no PC software > company is > using it. > This model works extremely well for computer gaming companies. And if I am not mistaken, gaming companies are the most profitable PC software companies aside from Microsoft, out there. (possibly even more profitable) It also works extremely well for Hollywood, who only extremely rarely will do more than 3 sequels of a popular movie, and quite often will back away from doing upgrades-I-mean-sequels of popular franchises in the fear of wrecking them. For example tons of people are clamoring for more Star Wars movies after the current one comes out this year, but Lucas has refused it. Quite obviously he could milk that cow for another 3 movies and probably get a billion out of it. It would of course completely destroy the franchise as by the time movie #9 came out, people would be so sick of them that it might take a century before you could even explore anything in that universe. > It costs a great deal of money to build completely new products, and > it can take a decade to recover the investment in development, even if > the product does well. Additionally, there is a huge risk of it _not_ > doing well, in which case the losses may be great enough to sink a > company. Overall, it's not nearly as safe or profitable as adding new > bells and whistles to existing products and releasing "upgrades." > Therefore all PC software companies emphasize upgrades, not new > products. > All of that is true. However it is also true that if you don't build new products that eventually your existing product line gets stagnant and a young upstart competitor will come in and wipe you out anyway. Just like Microsoft did to IBM. >> Think of it. Instead of every 2 years yet another tired old bloated >> overengineered version of Windows, they could come out with a brand >> new operating system named something completely different. For >> example, 'doors' > > Even Microsoft can't afford to do that. They were already taking > major risks with Windows NT, but fortunately it worked out. > The only reason NT was a risk (more accurately the NT architecture) was because Microsoft was telling everyone they were going to force the issue on the Windows 9x architecture ie: they were going to stop making Windows 9x versions after ME. They could have easily taken a path where NT-architecture (like NT4, 2K and so on) had NO backwards compatibility and would ONLY run NT-aware apps, and if you had a 32 bit app that wouldn't run on it, you just would keep buying Windows 9x. In fact, this direction was discussed many times pre Win2K. What Microsoft realized at the last minute I think is that since there's no copy protection on Windows 9x, that they could just ignore the issue as people would just keep making copies of it. So, in reality that is what is happening with the Windows 9x-to-NT architecture. People are still in fact installing Windows 98 on brand new hardware, that is why brand new motherboards are still coming with CD's that include Windows 9x drivers. That is why you can still run Windows Updates on a Windows 98 system and get your security updates. That is why new Windows programs mostly still run on it. Of course every year the number of new Windows 98 installs goes down. But they are still happening. > Nobody is going to rewrite 250,000 applications for a new operating > system today. New operating systems must not only be superior to > their competitors in a number of ways, but they must also have vast > amounts of > application software available for them in order to have any hope of > being adopted. The risks for the OS vendor and the application > developers are huge. That's why practically no one writes new > operating systems. That's why Apple couldn't afford to develop a new > OS. That's why Microsoft has consistently kludged its NT-based > systems to support older Windows applications. Even so, NT was > initially hurt by > a lack of > compatible applications. > This describes Linux and Linux according to the surveys is continuing to make inroads to Microsoft's business. Those 250,000 applications are only 250,000 applications today, right now. As the years pass more and more of those 250,000 applications will become obsolete and be retired with new ones. As the new ones are being written, more and more of them will be ported to Linux, assuming Linux continues to take market share. You are talking like a new operating system must be revolutionary. IBM tried that with OS/2 it didn't work. But you can be evolutionary with multiple product lines. And you also keep forgetting the power of a monopoly. If tomorrow Microsoft announced they had abandonded Windows development and the current version of Windows under development would be the last version released, and Microsoft was switching over to Linux, then within 5 years almost the entire industry would be hot and heavy switching over to Linux with the exception of a few BSD holdouts like us and Apple. And I will also almost guarentee that if Bill Gates wasn't running Microsoft, they would be already doing this. It is only because they are so full of themselves and the hydra they have conceived in Windows that they aren't even looking at doing this. But, Gates won't last forever, he's going to die eventually. Then Microsoft will start acting like all other companies. If Microsoft were to come out with a version of Linux today, labeled Microsoft Linux, it would put Red Hat out of business. This is true even if they kept their Windows product line. They would make just as much money as before, probably more in fact. The only reason they don't is they are emotionally tied to Windows. > > Except that nobody would buy Doors, because everyone wants to use > applications that run only on Windows. > Not in the beginning. Nobody wanted to buy OS/2 in the beginning either. But IBM and Microsoft kept selling it and eventually people did start buying it. And ultimately OS/2 was renamed Windows NT and people bought it a lot more. >> But of course, you won't ever see this kind of visionary product from >> Microsoft, because they are a company of mediocre programmers. > > No, you aren't likely to see this from Microsoft (or anyone else) > because the company doesn't want to dig its own grave. > > They've done it in the past: Windows NT being the classic example. > But it gets harder with each generation, and eventually it becomes so > expensive that even a Microsoft doesn't have the financial resources > to bet on such a project. > You don't need a lot of financial resources. If you want to change a large market, you can do it two ways. You can spend a little bit of money every year for many many years, or you can spend a lot of money in a few years. Microsoft doesen't have the money to instantly in a single year switch every PC desktop over to a completely different operating system. But they certainly have the money to spend the next 10 years gradually moving the market over to a completely different operating system. So do a lot of other people, like the Linux people. There are a huge number of markets where this has been done. For example, consider the simple Center Mounted High Brake Light you see on the back of most cars. This was phased in gradually. Today all cars have them. It really didn't cost anything to phase them in because the process took many years, since only new cars had them and it took many years before enough old cars went off the road for most of the cars on the road to have these brake lights. Ted
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?LOBBIFDAGNMAMLGJJCKNEEODFAAA.tedm>