Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!utgpu!water!watnot!watmath!clyde!rutgers!husc6!bu-cs!bucsb.bu.edu!madd From: madd@bucsb.bu.edu.UUCP Newsgroups: comp.sys.ibm.pc Subject: Re: IBM new 'standard' Message-ID: <871@bucsb.bu.edu.UUCP> Date: Fri, 27-Mar-87 10:50:19 EST Article-I.D.: bucsb.871 Posted: Fri Mar 27 10:50:19 1987 Date-Received: Sat, 28-Mar-87 15:21:11 EST References: <701@imsvax.UUCP> <879@maynard.BSW.COM> <1029@rpics.RPI.EDU> Reply-To: madd@bucsb.bu.edu.UUCP (Jim "Jack" Frost) Organization: ODO (Organization for the Disorganization of Organization) Lines: 226 In article <1029@rpics.RPI.EDU> chassin@rpics.RPI.EDU (Dave Chassin) writes: >In article <879@maynard.BSW.COM>, campbell@maynard.BSW.COM (Larry Campbell) writes: >First I time warp you back 10 years, and I see you trying to buy a computer >for yourself. Yeegads... the prices are outragous, the machines are pityful, >the OSs non existent. 3 years later, some nice stuff is beginning to appears, >Apple has an innovation... shock waves thru the computing world as a new >field is goudged out: PERSONAL COMPUTING!!! Wow, what a concept, a machine the >I, and only I use. No more logging in, complicated OS, fabulously expensive >machines available only the fabulously wealthy Corps and Research Insts. All >of sudden MILLLLLIONS of people have access to something new. Then IBM >announces a product: *THE* pc. And it's just that, a standard, that left >all other attempts at setting a standard wimpering in the sewers. The PC >is a very different concept from mainframe or even mini-computer technology. >It serves a different group, purpose, and ideology. Then we (PC users) are >told that we are being abandoned. Its not a good feeling... It appears to me that few PC users have seen IBM marketing strategies. IBM does not care about the single PC user. It never has. If it did, the original PC would not have had such a high failure rate. If it did, it would not have abandoned the original PCjr owners. If it did, it would not have abandoned the System/23 owners. IBM saved face with the PCjr by giving away real keyboards -- a good idea for a company whose reputation (at least in part) was built around the Selectric. More recently, IBM failed to inform users that there was a high probability that the hard drive in their new AT would fail in six months or less. Nice work, IBM. Do not honor IBM for introducing the PC. They were looking for a low end market entry. The PC was wildly successful there, mostly, as has been pointed out, because there was nothing else. For the first time, users got to see a fraction of the power of computers. Then, as the PC was seen to be slow and out of date (other personal computers were being introduced that shamed the PC), IBM saw fit to introduce the AT. Fantastic. Another rush. But IBM faced two problems. The AT was potentially powerful enough to eat a chunk out of the System/36 line. So IBM, in its infinite wisdom, decided to be conservative and hold the AT's clock rate down to 6MHz. Don't flame me on this -- I have the original product announcements. The bus was specifically designed with 8MHz in mind. The chips could do it at the time. So why keep it slow? To save money somewhere. This slowed the impact on their other market. The second problem was clones. They were eating bigger and bigger chunks out of PC sales. The introduction of the AT started AT clones. In general, the clone makers did not conform to the conservatism of IBM. They used higher clock rates. They built in more functions. They gave away software for free. They outplayed IBM in everything but documentation, and some even did that. IBM cannot afford to introduce a PC-like model that uses the 80386, unless they substantially improve the state of their System/36 (which they have been trying, you might note). An 80386 machine with a decent operating system will easily outperform the basic System/36. Where would the advantage be then? Clone makers would step in and take over again. Therefore, when IBM announces it's "new" PC, you can expect several things. * closed architecture. It worked in the past. * proprietary operating system/chips. No clones. * expensive. It may replace the /36, so it will be at least half the cost, or about $10-15,000. Compare this to the cost of the Compaq 386. * initial problems. I know of no innovative IBM product that did not have serious problems when introduced. Some, like the /23, were never fixed -- they were just abandoned. * limited PC compatibility. IBM isn't stupid, but they cannot continue to use MS-DOS on powerful machines, for many reasons. The first is that the machine will be mostly redesigned. The second will be that the OS will be multitasking. Look for a VM-style introduction, so that you can run MS-DOS *and* IBM's proprietary OS. You may notice that I said "when". Not "if". I've been hearing an April 2 announcement date, and late 1987 production date. In a week we'll know for sure, won't we? I'm putting my money on it. >I use a Sun 2/120 for any intensive work I do. But the ATs we have serve >an equally valuable purpose. Word processing on any mini or mainframe >is a waste of valuable CPU time, and is an effort. No one needs more >than an occasional side-track anyway. My view of the whole thing is >that if I need a process intensive environment to hell with the PC/XT/AT >type machines. So why is IBM abandoning a proven environment and telling >us that this new one will do the same thing, only better? I say this >is where we (dedicated PC users) are being dealt alot of excrement. This I completely agree with. I also don't see the need for tying up a mainframe to type a simple paper. I also see that you noticed that IBM *is* striking a low blow to PC owners. Like I said before, what else is new? >> The problem here is that there are certain functions that the operating >> system ought to provide, and MS-DOS doesn't. These include: >> >> - reliable file system that scales up to large files well >> - graphics >> - communications >> - multitasking (why force the user to WAIT while the database >> index is reorganized?) >> - interprocess communication >> >> MS-DOS provides none of these. [...] >I disagree with the list in several ways, though indirectly. First the >file system is suited to the type of use. If you want a larger environment >get of the machine. Ah! True. The PC file structure is fine for a PC, with limited exceptions. If you want better, get a different OS. With AT class machines and better, you really don't need a better machine. Just get rid of MS-DOS with its poor i/o handling and file structure. >Second, having ALOT of graphics work (architecture is >the MOST demanding of fields in terms of graphics) in my repetoire on a >wide variety of machines, I would say that graphics support of the type >like MAC, and SunWindows and so on is no good. Its so generic it serves no >function efficiently. I prefer grabing a previously designed system or >database and using it for my own purposes. Graphics is just too broad a >topic to be designed into a system. True. If you really need graphics, the stuff already exists. Graphics is a difficult thing, anyway. When you need it, you need it bad. Otherwise, the better graphics are nice but not necessary. I like the PC method of buying a card to fit your needs. I notice the new Macs are doing that, too. Apple is a smart company. Expect to hear a lot more from them. >Fourth, multitasking. What I've got is >good enough, although I agree it could be better. I don't think it >is necessary to redesign DOS so radically just to plop a program into >a background state. That has always mystified me. Fifth, in a limited >multitasking environment like I think is desired on PC like machines, I see >no need, or room for inter-process communications too be very fancy. Again >it could be done without completely restructuring DOS. No. If you look at MS-DOS, you see that they were attempting to design a multitasking system. It "feels" to me like they decided it just wasn't worth the trouble. They were right, for a PC. Right now they are rewriting MS-DOS completely. Microsoft is not stupid. They know that the days of MS-DOS are numbered. There is VERY limited interprocess comunication on a PC. Pipes classify as "interprocess communication", right? Well, if you have multiple processes, you really need communication between them. The groundwork for this type of communication is very, very simple and could be written into DOS with few changes. The only problem is that you can't write multitasking into DOS without lots of changes. >When you say you're eagerly awaiting a decent applications development >environment, I would answer: if you're not satisfied with the product don't >use it, and I mean PCs not DOS. I'm not saying that the PC and DOS are an >inherently connected gruesome-twosome, I thing there is a great deal of >room for improvement in DOSland, but I think it is major mistake to say >DOS is trash so let's throw it out and start all over. Bravo! You're right on the money here. I really don't want to throw out my copy of PC-Write just yet. I don't say "abondon DOS". I say "improve it." >Too much has been >done in the environement, and for IBM to end the line altogether I think >is a MAJOR MARKETING MALFUNCTION. As long as they continued to make PCs >XTs and ATs there was some control over the quality of machine being made >by the clone makers. A minor point here. Almost all of the clones I have dealt with are of superior quality than IBM. The faults show up when you have a problem -- the PC may be a good product, but often the support is not. IBM can't hardly be given credit for a reliable machine -- theirs is average, no better. Sometime you should get hold of first-use failure rates for different brands of PC's. I won't tell you the results -- you can guess. My personal experience supports the statistics. >Now I feel like I'm being thrown to the wolves, and >I tells me that way deep down IBM doesn't give a shit (please excuse the >language, I try to be civil, but once and while it helps to be explicit) >about the millions of people they conned into believing they cared about >really doing something for the computer industry. Let's not forget that >IBM set the standard in the long run, whether it's hardware or attitude. And >the attitude there new policy suggests is that PC users don't make us enough >money so let's bag 'em. Frankly I'm scared that I'm about to buy a machine >(I want an AT clone) which in two years nobody will be making >anything for. This could be a problem. In the past, IBM continued to support dead machines for some time, at least in a limited sense. However, the PC is a different animal. IBM does not market much software for the PC. It won't support much, then. So when the PC dies, the vendors are likely to move away. Expect software for awhile though, as I'll explain: >Vision of TRS-80, Apple II, et al. Don't you think we deserve better >consideration than that? I've used Apple II machines. They have a damn good support base. New software all the time. It's a matter of installed systems -- how many people have IBM PC's? Lots. Expect support for a long time. The TRS-80 line was different. TRS-80's were bad machines from start to finish. Take apart one sometime. Most of the older ones have cardboard for circuit boards. Not exactly good designing. Also, the operating system they offer(ed) is (was) not so good. >Hopefully there'll be enough of a market to keep >the competition between the clone marker up, and hopefully they won't >try be suckered into trying to work there way into a non-existant market >IBM is trying to make for there new 'baby'. > >These are my opinions, so flame as you see fit... I did :-) These too are my opinions [interspersed with some facts -- anyone caring to look at IBM's history of microcomputers will see that they haven't exactly been nice]. I welcome any discussions/flames. >Dave Chassin > >PS: for those of you who were wondering why we discuss this now, before >the fact: I think it's a damn sight better than after the fact... maybe >enough noise might change IBM mind about dropping the PC line. It probably won't (although I remember the PCjr...), but you're right. If some small business gets a new Mac instead of a new IBM and the new IBM flops, we've saved somebody a hard time. Let's keep this stuff out in the open, where it can do some good. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Jim Frost * The Madd Hacker | UUCP: ..!harvard!bu-cs!bucsb!madd H H | ARPA: madd@bucsb.bu.edu H-C-C-OH <- heehee +---------+---------------------------------- H H | "We are strangers in a world we never made"