Path: utzoo!news-server.csri.toronto.edu!cs.utexas.edu!uunet!mcsun!unido!gmdzi!strobl From: strobl@gmdzi.gmd.de (Wolfgang Strobl) Newsgroups: comp.sys.mac.misc Subject: Re: give me solid facts: why is the mac better than MeSsy DOS/WINDOWS Message-ID: <4245@gmdzi.gmd.de> Date: 7 Mar 91 19:56:49 GMT References: <4176@gmdzi.gmd.de> <29159@cs.yale.edu> <4196@gmdzi.gmd.de> <29227@cs.yale.edu> Organization: GMD, Sankt Augustin, F. R. Germany Lines: 361 favorini-francis@cs.yale.edu (Francis Favorini) writes: >In article <4196@gmdzi.gmd.de> strobl@gmdzi.gmd.de (Wolfgang Strobl) writes: >> Windows applications >>don't have to resort to all these little tricks which where necessary >>to squeeze a program into the 640k box, and which made these programs >>incompatible with each other. >> >>It does quite a good job to emulate the old DOS environment, if somebody >>wants to run an old application. Sometimes it fails, but so what? Do >>you judge the Macintosh environment by looking at its PC emulators? >>Surely not! >My point is that the fundamental problem with MS-DOS/Windows systems is that >most people using Windows also want to use DOS programs. As you say later, it is a *transient* problem, not a fundamental one. > Switching back >and forth is a hassle, despite the DOS box in Windows, which very often >doesn't allow you to use a DOS program to the fullest, if at all. Why do >people want to use DOS programs? Because plenty of useful programs are not >available in Windows versions. Not yet, mostly because there are so many useful old DOS programs. > And I'm not talking about old programs, >but current stuff put out by MAJOR companies. (E.g. Paradox by Borland) I like Paradox, too - I still keep a copy of the original article from Zloof in a '77 IBM System Journal, where he described Query-By-Example, which is the idea on which Paradox is built upon. But both the monolithic implementation and the user interface are a bit outdated. I would not call it "current". >Now you might say there are several databases/word processors/speadsheets >available. But if the one I've been using for the past three years isn't >among them it doesn't do me a whole lot of good. I have too much time >and effort invested in XYZ program to convert over easily. I am not >blaming either the companies who write the apps or the Windows folks at >Microsoft. Programmers are not clairvoyants or magicians. The point >is that PCs are in a transitional phase, and it can be hard on the users. You are absolutely right here. But I think it is better to make the transition now than later. If we make it later, it will be even more difficult and costly. Character based, "integrated" DOS applications with their own user interface are a dead end. >Ultimately, users will benefit, of course. But the intervening hassle is >of an indefinite length, and in my opinion that is one area that puts PCs >behind Macs. It's the users choice. The intervening hassle is as long as the user chooses. There is nobody forcing the users to stay with old DOS applications, and there is nobody forcing them to switch to Windows, immediately. This is indeed a problem Mac users don't have. >>>Switching from DOS to Windows and back is a pain. Sometimes Windows >>>trashes the machines memory/ports/environment enough that you must >>>reboot to use another DOS program. HASSLE! Of course all this is >>>to use an inferior GUI anyhow. See below for details. >> >>Inferior to what? To a mixture of Paradox and your favorite menu >>system? >I think it's obvious that I mean it is inferior to the Mac. That's what this >comparison is all about. It is not so obvious. The standard argument against Windows in this thread is that it handles old DOS applications not very well. While I don't buy the precondition, the whole argument is flawed in my opinion, because it doesn't say anything about the quality of Windows as a GUI at all! >>>[Discussion of the fact that Windows is slow on machines that are otherwise >>>fast. E.g. 20MHz 386s] >>The bottleneck for the GUI operations you describe as to slow (scrolling, >>menu pull-down, ...) is the video card, which is slow and dumb, usually. >>You did not mention what brand of video card you use, but I would >>assume the problem there. >I use an ATI VGA Wonder Card with 512k on-board video RAM. Now this may >not be the fastest card available, but it is not anywhere near the slowest. I don't have access to such a card, so I can't compare it with others. But from reading comp.windows.ms, I remember that there where many complaints about the Windows 3 drivers of the ATI VGA Wonder to be very slow. >But the point is that if you need to get the fastest (i.e. expensive) video >card to make Windows perform acceptably in the display of its basic GUI >elements, then plenty of people will have to get a hardware as well as >software upgrade to use Windows. Also the prices people quote when saying >how cheap PCs are should reflect that. This is a point against PCs. No, it isn't. You don't need to get the fastest, but sometimes a card which performs well under some conditions - using character based applications or static graphics in this case - does not perform not so good under changed conditions - if you switch to a GUI, for example. I have an old ATI EGA Wonder somewhere, which I bought because it, while beeing a bit slower, is able to change personality on both sides, both to the software and to the monitor. This made it possible to buy a cheap B&W monitor and use it as a cheap, temporary replacement for a otherwise required color monitor. Under character based applications, the difference in performance is hardly noticeable. Under Windows, the card crawls. Fast SVGA cards supporting 800*600 screens are in the price range of 100$ to 200$, so we aren't talking about that much money, here. >[I talk about File Manager and Program Manager.] >>You're right, both programs leave something to be desired. Other companies >>have started to sell replacement shells. So far, I like none of them, so >>I stay with the above two programs. I don't share your opinion that >>having *two* such programs is a bad thing. >Conceptually, two programs might work, if they cooperate well. I don't think >Program and File Manager do. Can you give a specific example? I have some complaints myself, but would like to hear where the cooperation between Program and File Manager is bad, in your opinion. >[I complain 640k DOS memory and other RAM problems.] >>See above. All this has nothing to do with Windows. Windows has no >>640K limit. It does not use different kinds of memory. It emulates all >>kinds of quirks / bugs / features / interfaces an old DOS application >>may expect, mostly successfull, sometimes unsuccessfull. But this >>emulation feature is not the central feature of Windows, it's an >>ad on, a migration path. >It may have nothing to do with Windows, but it has everything to do with >PCs. The point here is that you have to be compatible to your previous >mistakes, unless you can afford to alienate your current user base. This is misleading, because there is nobody alienating their current user base, here. PCs are manufactured by many different hardware producers as part of their computer equipment line, while Microsoft is a software house selling application software and operating system software for these and other systems. So a PC is the result of the combined effort of different companies, which is quite different from the Macintosh, where the essential parts, the hardware and the system software come from exactly one producer, Apple. Anyway, if beeing compatible with previous mistakes means repeating them, I'd rather be not compatible and make an advance. >Obviously, both Apple and Microsoft feel they can't. However, I think >Apple has been more succesful at maintaining a smooth transition to new >versions of their system software than Microsoft has. (Granted Microsoft >has more to handle, but they can't blame anyone but themselves.) Sure they can. They didn't invent the machine for which they wrote the system software, for example. And they can blame the necessity to be compatible with the tremendous amount of usefull CP/M software they had to be compatible with. >PC users have to deal with this, and it is a problem. You can't both have the cake and eat it. A PC user has the choice to select from a rich supply of hard and software of different price and quality. With this freedom comes the ability to make the wrong choice, of course. >>> 3) Different video modes. I don't want to have to configure >>> my software everytime I move from my VGA monitor to the machine >>> in the other room running with EGA or MCGA or Hercules, etc. >>Windows programs don't need that. In fact, a Windows program has no >>way of obtaining the video mode it is running in. >That's fine. But PC users still have to worry about video cards and modes. as long as they continue to use old DOS applications. >[I talk about (Non)Standard File in Windows, and general klunkiness in its >GUI.] >>I have changed my opinion about this point a few times now. >>A standard file dialogue is simpler both for the programmer >>and for the the user. But this "one size fits all" approach >>does not encourage the development of spezialized file dialogues, >>which are optimized to the actual context. Windows offers a set >>of building blocks for file dialogues, which in fact allow all the >>variations you describe above. >I'm not talking about exotic features. I'm talking about the simple >mechanics of picking the file you want to operate on. This one of the >most basic interface elements, and Windows programs can't agree on how >to do it. I might also add that not a single one I've seen matches the >cleanness and simplicity of the Mac's version. I'm not talking about exotic features, either. But I don't think that only because the Mac does it a certain way, and because it is written down in the Human Interface Guidlines, Windows has to do it in an identical way. But let me ask a specific question. What is wrong with the file dialogs used by the standard Windows programs, say notepad or calendar? >>> 6) Windows crashes much more than my Mac at home. And I have >>> 27 INIT running at home! Windows will crash about once or >>> twice a week during seemingly normal and mundane operations >>> like typing scrolling. It crashed even more often after you >>> have installed new software (I mean just plain old apps) until >>> you have tweaked WIN.INI, CONFIG.SYS, and AUTOEXEC.BAT enough. >>> This is a major hassle. I have never had my Mac crash more >>> often just because I installed a new applications (INIT maybe, >>> app no way). >> >>Such statements are very hard to verify. A fair comparison would be >>running Windows with Windows applications only and then comparing. >The crashes I'm talking about occur when I run only Windows programs. I see. My experience is a bit different. There are of course buggy applications which can crash the system, because Windows doesn't use the memory protection of the protected mode fully. But if I avoid them, and if I don't run old DOS applications - which may interact with the system in a strange way - Windows runs the whole day long, 7 days a week. From what I see from people using Macs here, a faulty application can crash the Mac OS easily, too. >>Many old DOS applications are buggy in the sense that they contain >>spurious reads or writes which hit ROM other places where it does no >>harm, so that these bugs get unnoticed. Windows in Enhanced mode >>is able to detect such illegal operations sometimes, and tries >>to terminate the offending application, sometimes successfully, >>sometimes not. >Here again, I have to say that PC users suffer from the fact that DOS and >Windows beat each other up sometimes. You describe the situation as symmetrical, but it isn't. Windows is built on top of DOS by using its file system (which is indeed a drawback) and throws away everything else, by managing the system resources itself. PC users suffer only if they choose to. >[I bemoan laborious and ill-documented installation procedures for apps.] >>This is really a problem. But I don't know how it should be solved >>without going the Apple or the IBM way, i.e. having one hardware >>supplier controlling the system. >I don't know what you mean here. Apple and IBM control the hardware and >operating systems (if you count OS/2), but certainly not the vast majority >of software. And Microsoft certainly controls DOS and Windows. The problem >is not control, but designing an OS that allows third party software to >be installed without mucking around with lower-level aspects of system >configuration. Apple has Microsoft/IBM beaten by a mile on this one. Apple surely has beaten IBM in this area. Apple so far never lost control over the essential parts of their system - the hardware and the system software, while IBM never had the control over the system software, lost the control over the hardware to clone makers and failed miserably to regain control via the switch to a closed - i.e. not clonable - architecture, the Micro Channel. Microsoft controls Windows, but not much more. It certainly doesn't control DOS. There is a complete, enhanced DOS clone from Digital research, which is cheaper and runs Windows, for example. Designing an OS that allows third party software to be installed without mucking around with lower-level aspects of the system configuration is trivial (ok, let's say: simpler) if one party builds both the OS and the essential hardware interfaces. It is not so simple if an OS designer has to take foreign hardware including changing interfaces into account. The problem gets worse if those hardware was not designed for automatic installation and new hardware continues to do so. On the long run, this problem will go away because hardware supplies start to build their hardware to be compatible with the software instead the other way around. But this will take a while. The shorter solution would be the way IBM intended it to go: PC's are built only by IBM, add-on cards have to be approved by IBM, IBM builts a special version of OS/2 with additions nobody else has. I prefer it the way it is now. >>Tools which move the 640K barier a little bit are in common use for >>some time now - what Microsoft now integrates in DOS 5.0 is not at all >>new. And it is in no way related to Windows itself, only to its ability >>to run old DOS applications. I don't understand what that ability >>of Windows has to do with the level of the Mac 5 years ago. You compare >>apples and oranges. >I know all about memory managers and have been using them for years. They >can cause all kinds of problems and the sooner they aren't needed, the better. So we don't disagree, here. >As I have said repeatedly, users of PC do have to worry about these things >if they want to use both DOS and Windows programs. And most will have to, >at least for now. The whole point of this discussion is to compare PCs >and Macs. And if a current PC feature is comparable to an old Mac feature, >it just shows that the Mac is more advanced in that area. I think it is fruitless ;-) to compare Macs with PCs that way. The Mac started with nearly no hardware: a CPU, a little bit of memory, a few peripheral circuits and a small monitor. What the Mac made the Mac was its innovative software, which broke with most concepts which where in common use at that time in the microcomputer area. The innovation was to squeeze inventions made earlier and elsewhere (at Xerox Parc) into a small box. The IBM PC instead was an earlier attempt to built some kind of "super CP/M machine" by selecting an 16 bit processor instead of the common 8 bit processor, and by adding hardware - video, DMA controller, floppy controller, optional numeric coprocessor and the usual serial and parallel interfaces. The hardware was a step forward, the software wasn't. It makes sense to compare the system software of the various Macs with Microsoft Windows, because they are on the same level, if looked at from an architectural point of view. It may make sense to discuss whether the Intel processors or the Motorola processors are better for a GUI of the Mac or Windows sort. But it is not very productive to compare "the" PC and "the" Mac, because the results are arbitrary and only reflect the current prejudices of the people. >> >>>Things I like about Windows: >>> 1) Better multi-tasking than Multifinder on the Mac. I.e. >>> regular programs can run in the background, not just >>> print spoolers, etc. (Note that I mean actually run, not >>> just sit there as in Multifinder.) >> >>Huh? My understanding was that background applications continue to >>run under the Multifinder, too. >What I mean is that old Mac apps did not include and background ability, and >many still do not. Thus, when they are swapped out by Multifinder, they >don't do anything useful. I wouldn't blame the Macintosh of not beeing a multitasking machine because of that. >>Many of Windows problems are related to DOS. But you don't have to >>live with most of them, because they only occur if you continue >>to run old DOS applications. If you don't, the only thing Windows >>inherits from DOS is the file system. >> >>Microsoft already tried the switch to a new operating system. This >>seems not so simple as it sounds. >This is exactly the point. The mish-mash of operating systems/interfaces/ >memory models/video standards etc. is a serious headache for people trying >to do real work in the real world on PCs. I.e. tens of thousands of people >who recently bought Windows still have to run their old DOS programs too. >What do you think they did with all that software? Throw it out? Not at all! They should do what they actually do: continue to run what is useable, and throw out what is unuseable. This works very well for many people. > Even >if you could replace all of your old DOS software with spiffy new Windows >software, it would cost lots of money. I repeat my assertion that until >DOS is long gone, PC users will still be weighted down with a nasty old >boat anchor, even if they have nice Windows programs to look at. Of course, this applies only to people who already are computer users having a software collection, not for starters. Many of those computer users started using computers long before the Macintosh hit the market. So your statement reduces to the simple truth that there is an advance in the area of computers, and that those who start later have the advantage to be able to learn from the experiences of those starting earlier. This applies to both system designers and users. Wolfgang Strobl #include