Date: Fri, 30 Nov 2001 23:01:13 +0100 From: "Anthony Atkielski" <anthony@freebie.atkielski.com> To: "Mike Meyer" <mwm@mired.org> Cc: <chat@freebsd.org> Subject: Re: Feeding the Troll (Was: freebsd as a desktop ?) Message-ID: <003301c179ea$8925d270$0a00000a@atkielski.com> References: <15367.37543.15609.362257@guru.mired.org><040701c179af$4bda25f0$0a00000a@atkielski.com> <15367.43943.686638.723011@guru.mired.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Mike writes: > It makes for lousy tech, but it's smart management. If tech alone were the deciding factor, we'd be running FreeMultics, not FreeBSD. > I don't believe the world was ever Apple's > oyster ... There was a time when I and just about everyone else in IT lusted after Macs. They were incredibly cool and light-years ahead of everyone else. But they were very, very expensive, and the price didn't come down, nor did the functionality dramatically increase. Finally Windows came alone, and it wasn't quite as nice, but it was affordable--Apple took no notice of this approaching storm. We went to Windows because we could. And Windows got better and better, whereas the Mac just retreated into its little niche and pouted at the world. And so the Mac faded away, and Windows became the alpha dog. > ... and I helped found a company that sold Apple ]['s > ... That was _before_ Apple's golden age, although I lusted after an Apple II as well, IIRC. Only with a Trinitron monitor, though! > ... actually, we sold the black Apples for licensing > reasons ... Another one of Apple's continuing mistakes has been its desire to keep everything to itself. It even pulled the licenses of some legitimate Mac builders not long ago, as I recall. Once again, Apple shoots itself in the foot. > For Windowing environments, the Mac never caught > up with DOS, so that clearly wasn't it. The Mac was well ahead of DOS and Windows for a few crucial years. As usual, Apple just stupidly ignored the open door, and plodded further towards oblivion. > As soon as MS had a windowing environment that > was good enough - 3.0 - Apple lost the the windowing > environment ... Yup. And it didn't have to be that way. With management like Apple has had through the years, it has never really needed any enemies. > ... though the Mac can still be found in desktop > publishing and graphics markets that haven't converted > yet. That is mostly inertia today. There is almost nothing one can do with a Mac in graphics and publishing that one cannot do in Windows. Even companies like Adobe now develop new releases for Windows first, and the Mac second--because that's what customers are asking for. Furthermore, the Mac doesn't have any other significant domains in its embrace; take away graphics and publishing, and there's practically nothing left. Windows, on the other hand, can do anything. I should point out, while alluding to applications, that while Windows profits on the desktop because there are a hundred thousand desktop applications for Windows, UNIX profits on servers for a similar reason: there are hundreds of thousands of applications that run on UNIX and suit a server or timesharing environment. I daresay that if you need support for the world's most bizarre and obscure network protocol--or even a slightly exotic protocol--you're far more likely to find it for UNIX than for Windows. I've been regularly pleased by the bewildering variety of network and related software available for UNIX. I don't think there is any system in the world that a UNIX system cannot communicate with. > They react to them like they do to most things > that malfunction regularly. The only problem with this is that Windows does _not_ "malfunction regularly." > They all went where microsoft wanted them to go. Where else was there to go? > Proprietary standards lock you down. Open standards > don't. "Open" standards lock you down just as much when the system you are using doesn't support them. > Or Windows. Yes. > That's one of the major disadvantages of not > using Windows: I don't get a FreeBSD driver for > the hardware when I buy it. Yes, although it depends somewhat upon the type of hardware you are acquiring. You're more likely to get a driver for a StorageTek silo for UNIX than Windows, I think. But there are compensating factors that are particularly significant in server environments. For example, PPTP can be a pain to get running on Windows, even though Microsoft was one of the originators of the protocol. PPPoE is even worse--as far as I know, only one or two solutions exist, all of them black boxes from tiny companies. But all of this is standard stuff for UNIX; indeed, _anything_ having to do with a network is standard stuff for UNIX. UNIX already incorporates lots of IPv6 support; I don't expect to ever see that for Windows NT. > On the other hand, the ability to run Linux > binaries meant I'd have a good shot at getting > commercial Unix software running on FreeBSD. Does Linux compatibility in FreeBSD extend to drivers? I wish there were less hype for Linux; it's about at the bottom of the totem pole of UNIX systems, and it pains me to think that it is being held up as a model for all of them. It's like exhibiting a Yugo as a model of the auto industry. > However, there is no single compenent that > can't be replaced by another application > with similar functionality should the need > arise. And how long would that take? > Except, as we discussed in the other thread, > that even Windows NT still suffers from design > decisions made during the Windows 3.1 era > that catered to the single-tasking nature of > DOS. It still emulates certain things for compatibility. But the kernel doesn't require this emulation. > Except that in the other thread, you already > pointed out that NT has an event model that > ultimately derives from the DOS-based Windows 3.1 > environment. It's more complicated than that, actually. The emulation of Windows 3.x is quite strict for 16-bit applications, as they won't run without emulation of that brain-dead system. However, you can, say, run multiple VDMs for your 16-bit applications (as I do) to help keep them from interfering with each other. The 32-bit applications have more autonomy. There is still a lot of message traffic, though. And don't forget, even when applications receive only events that they register for, a lot of applications register for just about everything, thanks to clueless developers who think consistently in terms of a single-user desktop only (and I've seen this kind of mistake even in MS software). > You just agreed with my NT developer friend. He > said that when they added 9x compatability, they > made NT less stable. He must have been referring to NT 4.0. I always resented those changes. It was one step toward the desktop, and away from the mainframe. But, to FreeBSD's advantage, it also made NT less suitable as a server. > What he was really upset about was he that the > level of compatability they got didn't require > compromising stability; they could have gotten > the same level and kept stability where it was. True. A lot of code was transplanted from Windows 9x to NT to get those changes in. And if you know how badly written Windows 9x is, and how well written NT was, it is indeed depressing to contemplate. One reason Explorer sometimes hangs or has other problems is that it was so largely coped from Windows 9x. > Quit confusing the window manager and the server. > The window manager doesn't require any extra priveleges. Without a server, what good is the windows manager? If you are running UNIX as a desktop, you're running an X server on the UNIX machine itself. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-chat" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?003301c179ea$8925d270$0a00000a>