hgm@ncsuvx.ncsu.edu (Hal G. Meeks) (12/19/88)
Article 1061 in comp.sys.next: From: T8M-KAUP@FINTUVM.BITNET (Asko Kauppi) Yesterday, after having read the articles in this newsgroup, I began speculating the situation of the various companies. Will they survive the next round? Here are some of my thoughts. Please comment them, 'cos I *love* a nice discussion (/argue). >1) The Macintosh family > I still cannot help being impressed by the quality of the Mac II colour >screens. It seems that the Apple is doing everything two years in advance >to its competitors: the mouse, windowing, 3"1/2 disks, the HyperCard... Apple currently has a minor dilemna on their hands; the Plus and SE may run the same software as the II, but they must be sorely tempted to extend the capabilities of the native OS to take advantage of the Mac II's color, higher speed, etc. As it stands now, the most signifigant thing Apple has done in the last 2 years is allow Hypercard to be bundled with every Mac sold. A point to ponder is Apple's initial resistance to giving Hypercard away. Is this a trend? >2) The OS/2, IBM and the clones >I have serious doubts about the OS/2. Not that it's that bad an idea >initially, but when will it be ready? I know there's the 1.0 shipping already, >but so what. Exactly. Things do not bode well for OS/2. 3) The Amiga line >The Amiga is "the first home computer with a dedicated, up-to-date operating >system as its standard OS." (I don't recall who said that, but I agree.) Pete da Silva said it. I agree too. >There were the gurus, but that's mostly over and out. And, the Macs and the ST >have their "bombs" as well. So there should be no arguing that the system >wouldn't work. It does. >I just have the feeling that the company doesn't *really* care for this >particular line of products. Perhaps the 1.3 OS will change that feel. >I sure welcome the FastFileSystem. Okay, so I have a bias, since I own an Amiga 2000. Commodore seems to be doing many things right at the moment, in contrast to their earlier stupidity. Support is definitely there; 1.3 can be considered a mature version of 1.2, with extensions that make the OS much more reliable and a pleasure to use. However, the schizophrenic nature of Workbench/CLI needs to be amended. Workbench needs to be beefed up. These things are coming, but probably not until late next year. And I still have to wonder if Commodore can really play hardball with the big boys. At the moment, however, it is the home computer of choice for me. Still an amazing machine. And with the Mac Emulator coming in Febuary, the Amiga will be the only micro capable of running 4 different (and popular) operating systems (6 if you cheat and include OS/2 and CPM ;-) ) 4) The UNIX systems: Sun, Apollo etc. >As the Amiga line is shifting upwards to the higher end of the PC users, >the UNIX seems to be coming "down" from the workstation cathegory to the >PC one. >5) The NeXT (finally...) >Reading the Byte article about the NeXT machine gave me the same butterflies >that the first Amiga notation couple of years ago did. Me want one| I am going to reserve my opinion on the NeXT box for at least a year; so much is up in the air at this point. I suspect it will not be a major player in the mini/micro market, but it will be extremely influential in raising expectations of users as to what a computer ought to be able to do. Just as the Amiga seems to be doing, in a smaller way possibly. 6) The Black Horse (I'm not finished yet..) >Then there's the possibility that something totally unpredictable happens. Like memory chip prices going up drastically. Or market saturation occuring. -- ------------------ "I am living in a condo hgm@ncsuvx.ncsu.edu with Henry Thoreau" netoprhm@ncsuvm.ncsu.bitnet Reagan Years, Part II
shap@polya.Stanford.EDU (Jonathan S. Shapiro) (12/19/88)
I confess to hvaing read this with some interest, and at the risk of getting flamed, I am interested in people's reactions to my own views: In article <2405@ncsuvx.ncsu.edu> hgm@ncsuvx.ncsu.edu (Hal G. Meeks) writes: >Article 1061 in comp.sys.next: >From: T8M-KAUP@FINTUVM.BITNET (Asko Kauppi) >>5) The NeXT (finally...) >>Reading the Byte article about the NeXT machine gave me the same butterflies >>that the first Amiga notation couple of years ago did. Me want one| >I am going to reserve my opinion on the NeXT box for at least a year; so >much is up in the air at this point. I suspect it will not be a major >player in the mini/micro market... I am not sure that the NeXT machine will be a player. The reason is this: The NeXT box is based on technological innovation, but it hasn't got very much of that. Using a greyscale display is not new. An optical floppy is very good for some things, and is a neat idea, but it isn't (today) a viable primary storage medium (too slow), and it certainly isn't a conceptual revolution in storage. The DSP chip shows some real promise, but here again the major areas that it might support have been well understood for some time. The novelty of the NeXT machine, and its success or failure, will rest on NeXT's ability to integrate all of these features in the right way. This was true in much the same way on the Macintosh. Fundamentally, it just isn't clear to me that NeXT knows how to do this. They may, they may not. It is also not clear that they have been able to retain their people in the way that they have to to pull it all off. Now mind you, I hope that the box succeeds, but I think the hype has been overdone in this newsgroup. Steve Jobs *used* to know that it was important to have machines ready when you announce. NeXT hasn't done that. The real outcome on the NeXT machine wont't be clear until it has been around for a while. As for me, I will wait and see (with interest). > >6) The Black Horse (I'm not finished yet..) > >>Then there's the possibility that something totally unpredictable happens. >Like memory chip prices going up drastically. Or market saturation >occuring. Market saturation has already happened. To get something out now you have to reinvent the market and convince people to move. Jon
jbn@glacier.STANFORD.EDU (John B. Nagle) (12/19/88)
The problem NeXT must face is that they've shot their bolt long before
they can profit from it. Everyone now knows what the NeXT machine can do,
and can get busy in providing similar capabilities for the Mac, Amiga, and
386-based architectures. It's entirely possible that many of the capabilities
of the NeXT machine will be working on other machines before NeXT is shipping
iron in serious volume. As of right now, their window of opportunity gets
a day shorter every day until volume shipments to commercial customers start.
It would be interesting to see the NeXT software offered for the Mac
IIx, along with an emulation box which allowed the use of old Mac applications.
But this is unlikely, even though the hardware is reasonably suitable.
John Nagleberes@cadnetix.COM (Tim Beres) (12/20/88)
In article <17911@glacier.STANFORD.EDU> jbn@glacier.UUCP (John B. Nagle) writes: > > The problem NeXT must face is that they've shot their bolt long before >they can profit from it. Everyone now knows what the NeXT machine can do... >[brevity deletion] > John Nagle Scott McNealy addressed this point during his keynote address at SUG. I'm going to paraphrase some of his comments from memory, please politely correct me if I get it wrong [my views in brackets]: 1. Sun can't get away with announcing a product 9 months before FCS. It would kill sales of existing products [admittedly not a prob. for NeXT] and reeks of vaporware promises. 2. Just wait and see what we (Sun) have in 9 months. [Extrapolating from some of the things I saw: NeWS (damn caps!) uses the PS paradigm, with applications forthcoming that subsume pieces of the Interface Builder, etc; the new bus mentioned by Bill Joy and/or SCSI could be used to hook up opticals.] 3. NeXT uses a non-standard optical disk format - the "standard" being worked out by Sony, HP, Phillips, et al is different than that of Canon/next [that's it, no more caps]. 4. McNealy did shrug off the DSP & sound capabilities of the next box. [But the pre-announcement and industry reaction may cause this capab. to be further investigated.] 5. He remarked that mach is a non-standard and buggy choice of OS; furthermore, he somewhat denigrated CMU, their efforts and abilities [Can someone fill in the details, please. I remember thinking - hey, BSD.] 6. He slammed them on their lack of standards: mach, display PS, optical format... 7. He said the enclosure was big and ugly. 8. Sun doesn't do the bundling of SW that next intends, simply because they don't want to lock their customers into the "one database format", as McNealy noted of next. 9. Sun is pushing down and up on the market. Towards lower cost desktop machines and higher cost/performance servers. All running the same software. [Bernie LaCroute (sp.) addressed this subject. Will the lower cost machines in N months be able to compete with next on a price/performance/features basis -me] I do agree with John. Next has given us some advancements, but did they give away the goods too soon? Also - is their selection of features/standards helpful or hindersome? Tim (I want one now - but in 8 months or so, assuming I'm back in school [what the heck], I may want something else entirely) ------>MY SOAPBOX (I speak not for Cadnetix nor any enjoined entity) OK, one more time: This is a frying pan.....this is an egg.... this is an egg in a frying pan....JUST SAY OVER EASY Tim Beres beres@cadnetix.com {uunet,boulder,nbires}!cadnetix!beres
mike@ames.arc.nasa.gov (Mike Smithwick) (12/20/88)
In article <2405@ncsuvx.ncsu.edu> hgm@ncsuvx.ncsu.edu (Hal G. Meeks) writes: > >>5) The NeXT (finally...) >>Reading the Byte article about the NeXT machine gave me the same butterflies >>that the first Amiga notation couple of years ago did. Me want one| ^^^^^^^^^^^^ Me got one! Well, ahem, WE, got one. Our nExt, magically appeared in our lab this morning. I haven't had much of a chance to play with it yet, but here are a few first-hand observations: It is really slick!! "Slick" is a technical term meaning, boffo, neato, keen, etc. A combination of elegance, fun and grace. NeXTSTeP is beautiful. I couldn't get my eyes off of the thing. It'll be hard to go back home to my Amiga now, if it wasn't for the advantage of color. NeXTStep makes the Mac interface look like stone clubs, and flint spears, (read : primitive). The "browser" is as neat as I expected it to be. I'd like to do one for the Amy, but I'm afraid Steve might call me up. . . Haven't played with the Interface Builder, but it looked good. One small detail I don't remember anyone mentioning before about the window motions. When you move a window, you MOVE a window. Not just an outline of the thing, but the entire window. There is a flight simulator that comes with the box. It's called "Stealth", and makes very nice use of NextStep. The performance can't match Interceptor, or the other high-speed Amiga programs. Display PS is much faster than I'd expected. Still, on this release, screen writing speed seemed rather pokey on a line demo they have. The Amiga still shines on this account. Another couple of little things I liked that give the box character is the error feedback when you log on and make a mistake. The screen is dark, and a login box appears in the center. Make a mistake on the password and the box jitters back and forth for a fraction of a second like someone hit it with a hammer. Then there is the Digital Webster error requestor if it can't find a word. I only got a quick glipse over someone's shoulder, but it appeared to have a silhouette of a little man giving you a raspberry. Cute. The "Are you Sure?" requestor has the little man shrugging his shoulders and speading his hands. The Digital Librarian is impressive. We searched for the word "celestial" throughout the works of Shakespeare. It found all 3 entries in what appeared to be less than 5 seconds. Most of these observations are from looking over the shoulders of the crowd in line before me. I'll post more observations when I get some real experiance. -- *** mike (starship janitor) smithwick *** "Scientists say 'Saturn is so light, if you put it in a bucket of water, it would float'. Don't forget, it would also leave a ring". [disclaimer : nope, I don't work for NASA, I take full blame for my ideas]
jgreely@fish.cis.ohio-state.edu (J Greely) (12/21/88)
In article <19728@ames.arc.nasa.gov> mike@ames.arc.nasa.gov.UUCP (Mike Smithwick) writes: > It is really slick!! Well, sort of. I've been playing with our first NeXT for about six hours now, and while I may call *parts* of it slick, other parts are driving me nuts. 1) I *hate* click-to-type. This may be a religious issue, but I just don't like it. 2) The keyboard is not tiltable. Ours may grow rubber feet soon, if it stays in its current location. >"Slick" is a technical term meaning, boffo, neato, keen, etc. A combination >of elegance, fun and grace. "Slick" is frequently replaced by "Beta", so far. "Gee, that's a nice idea; pity it doesn't work. Real Slick/Beta." >NeXTSTeP is beautiful. Beauty is skin-deep, and I think nextstep has warts. I don't like the near-useless dock (for *lots* of reasons), the beta-beta inconsistency of the user interface (yes, this is promised to be fixed in 0.9, but it's really annoying *now*), the "intuitive" nature of many window operations (how many of our Mac/X/NeWS/ Suntools people sat down and scratched their heads over "features" in the windowing system? I think we'll need to take a poll), and some *real* features in icon handling. >Haven't played with the Interface Builder, but it looked good. Haven't had time. I've been busily playing "dumb-user", and managed to go five hours before looking for (fortunately useless) paper documentation. We've got three pages of changes, bugs, suggestions, and gripes already, with lots more to follow (many of them to do with the online documentation, which is in the "sorta" stage). >One small detail I don't remember anyone mentioning before about the >window motions. When you move a window, you MOVE a window. Not just an >outline of the thing, but the entire window. So? This can range from convenient to visually distracting, to screwy when an application doesn't know how to handle refreshes properly. >There is a flight simulator that comes with the box. It's called "Stealth", >and makes very nice use of NextStep. The performance can't match >Interceptor, or the other high-speed Amiga programs. Our licensed pilot played with it for a while, and I'm sure he'll be able to give a better impression of this than I can. Bob? >Display PS is much faster than I'd expected. Still, on this release, screen >writing speed seemed rather pokey on a line demo they have. The Amiga >still shines on this account. Which demo? Most screen handling is quite speedy (that I've seen). The longest redraws I've found were playing with background screens (bitmaps in postscript? yeah, that might be sluggish, depending on how they're stored). >Another couple of little things I liked that give the box character is >the error feedback when you log on and make a mistake. The screen is dark, >and a login box appears in the center. Make a mistake on the password >and the box jitters back and forth for a fraction of a second like someone >hit it with a hammer. Cute, but this falls into the category of "minor touches". There are a lot of these, some good, some bad, some hype. > Then there >is the Digital Webster error requestor if it can't find a word. I only >got a quick glipse over someone's shoulder, but it appeared to have >a silhouette of a little man giving you a raspberry. Cute. The >"Are you Sure?" requestor has the little man shrugging his shoulders and >speading his hands. Various programs use different icons, but a lot are like these (frequently, the same icon is used for a different purpose in different apps; more "fixed in 0.9", from the docs). I have great hope that these will not be 1.0 icons, or they might not stay that way here. Cute is relative, and wears of quickly. >The Digital Librarian is impressive. We searched for the word "celestial" >throughout the works of Shakespeare. It found all 3 entries in what appeared >to be less than 5 seconds. Wow. Try opening the first act of "The Merchant of Venice". Not quite as intuitive, is it? Which documents are stored under "Release notes"? And why do we get a browser rather than a nice scrolling index? Advice for the budding NeXT user: 1) whenever anything goes wrong, just repeat "beta, beta, beta" until the headache goes away. 2) Set up a local newsgroup/mailing list with all of the interested parties communicating. There are enough rough spots that it's nice to have one place to send them to local users (and by all means, send the actual bugs/suggestions to NeXT; things won;t get fixed unless they're pointed out). 3) don't let in the masses until the guru-types have had a chance to figure out how to get it sensible (sendmail, yp, NFS mounts, and other administrative nonsense). Note: if the above comments sound too negative, I'm forced to point out that I haven't logged into my sun since before noon, and while I may not be delirious, I don't hate it, either. Now, if we could only get WriteNow to print onto a non-NeXT laser printer, and have open apps on the dock look different from closed, docked apps, and build a terminal generator app, and slip an "abort" into the browser so you can log out before it finishes sorting the contents of /usr/spool/news/talk/bizarre (or other random large directory that you've opened), and (yeah!) run without NextStep, so I can have a giant vt100 (well, maybe I'm getting carried away now). For fun, type "cd /;cd ..;cd .." into a terminal or shell window, and watch it vanish. Interesting, no? -=- J Greely (jgreely@cis.ohio-state.edu; osu-cis!jgreely) "Give him some orange juice and a sugar cookie, ... maybe a tetanus shot. He'll be fine."
karl@ficc.uu.net (karl lehenbauer #) (12/22/88)
In article <5816@cadnetix.COM>, beres@cadnetix.COM (Tim Beres) writes: > Scott McNealy addressed this point during his keynote address at SUG. > I'm going to paraphrase some of his comments from memory, please politely > correct me if I get it wrong [my views in brackets]: > 4. McNealy did shrug off the DSP & sound capabilities of the next box. > [But the pre-announcement and industry reaction may cause this capab. > to be further investigated.] Based on this and other remarks, they're also shrugging off what the *integration* of all that stuff represents. > 5. He remarked that mach is a non-standard and buggy choice of OS; > furthermore, he somewhat denigrated CMU, their efforts and abilities > [Can someone fill in the details, please. I remember thinking - > hey, BSD.] This will probably come as a surprise to the folks at CMU -- they're running Mach in production on hundreds of machines. From my study of their literature, they seem to have done a good job at producing a transportable VM model that supports tightly coupled (UMA and NUMA) and loosely couple (NORMA) multiprocessing in a Unix environment. When people start plugging their Next machines together and start getting transparent multiprocessing, it will be interesting to see what Sun comes up with to do the same -- any whether or not it is based on Mach. Sequent, Encore, Next, Intel and others are committed to Mach for the solutions it provides in these areas. > 6. He slammed them on their lack of standards: mach, display PS, > optical format... Oh yes, the Sun is *so* open. Just ask the people who were trying to build aftermarket disk controllers for it how open Sun is. > 7. He said the enclosure was big and ugly. Oh yeah, the total reason not to consider buying the machine... > 8. Sun doesn't do the bundling of SW that next intends, simply because > they don't want to lock their customers into the "one database format", > as McNealy noted of next. Great, so by not giving away any free software with their machines, Sun is doing their customers a favor by not locking them into one database format. Look, bundled software is a freebie. Having lots saves users money and is an incentive to buy the machine -- a lesson learned by Osborne Inc, forgotten by all, then relearned by Next. > 9. Sun is pushing down and up on the market. Towards lower cost > desktop machines and higher cost/performance servers. All running > the same software. [Bernie LaCroute (sp.) addressed this subject. > Will the lower cost machines in N months be able to compete with > next on a price/performance/features basis -me] There is more to the Next than an optical disk, bundled software, a DSP, a megapixel screen, 8 MB RAM, 4.2 BSD, Objective C, et al. There is the integration of all that stuff. Even if Sun duplicated the hardware, they would have to duplicate all the functionality, like the look-up-a- word-from-anywhere stuff that, historically, they're not likely to do -- and apparently they'll call not doing it a feature :-) -- -- uunet!ficc!karl "The greatest dangers to liberty lurk in insidious -- karl@ficc.uu.net encroachment by men of zeal, well-meaning but without understanding." -- Justice Louis O. Brandeis
guy@auspex.UUCP (Guy Harris) (12/22/88)
>>One small detail I don't remember anyone mentioning before about the >>window motions. When you move a window, you MOVE a window. Not just an >>outline of the thing, but the entire window. >So? This can range from convenient to visually distracting, to >screwy when an application doesn't know how to handle refreshes >properly. Note that this particular item shouldn't surprise the "NeWS people" mentioned previously (although at some point it ceased to be the default behavior in NeWS, I think).
daveh@cbmvax.UUCP (Dave Haynie) (12/23/88)
in article <29866@tut.cis.ohio-state.edu>, jgreely@fish.cis.ohio-state.edu (J Greely) says: >>One small detail I don't remember anyone mentioning before about the >>window motions. When you move a window, you MOVE a window. Not just an >>outline of the thing, but the entire window. > So? This can range from convenient to visually distracting, to > screwy when an application doesn't know how to handle refreshes > properly. Aargh!! No, no! Please don't say that this system doesn't support "SMART" refresh windows! This is supposed to be a "real" computer with lots of memory. If the application doesn't know how to refresh a window, it shouldn't have to. Let the OS handle all aspects of a window. Just cause the Mac doesn't (or didn't, maybe it does now?) handle such things doesn't mean the rest of the world should stay so barbaric. > J Greely (jgreely@cis.ohio-state.edu; osu-cis!jgreely) -- Dave Haynie "The 32 Bit Guy" Commodore-Amiga "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: D-DAVE H BIX: hazy Amiga -- It's not just a job, it's an obsession
dar@belltec.UUCP (Dimitri Rotow) (12/23/88)
In article <17911@glacier.STANFORD.EDU>, jbn@glacier.STANFORD.EDU (John B. Nagle) writes: > > The problem NeXT must face is that they've shot their bolt long before > they can profit from it. Everyone now knows what the NeXT machine can do, > and can get busy in providing similar capabilities for the Mac, Amiga, and > 386-based architectures. It's entirely possible that many of the capabilities > of the NeXT machine will be working on other machines before NeXT is shipping > iron in serious volume. As of right now, their window of opportunity gets > a day shorter every day until volume shipments to commercial customers start. I enjoy John's pragmatism (almost alone in this group, it seems), but here even he may be too generous. This bit about "providing similar capabilities" to machines that already have capabilities *superior* to NeXT is a real slur on what people are shipping on a daily basis in the existing industry. The Amiga is hot and has been shipping for some time now (likewise Mac). It's just fine at providing text, graphics, sound, and all of the other gee-whiz stuff. Optical disks are old hat on PC's. Why do you think the Japanese always introduce the new models at Comdex with AT compatible interfaces? I don't understand the viewpoint in this group that PC's only run DOS or OS/2. Don't netfolks realize that half of the UNIX market these days (by number of units) ships on '386 AT clones? Why do you think 386 boxes have eaten up the low half of the UNIX market? What do you think '386's will do to the workstation market? Just about anybody's plug in graphics board for UNIX and X will easily outperform the NeXT machine in resolution, "look and feel" display speed, or any other measure. Slam a low-cost Maxtor or CDC at 380 MB into your low cost 25MHZ 80386 machine (not to mention the new 32 MHZ '386 machines or even '486 machines people are announcing) and you will eat the NeXT box alive in speed, quantity of *UNIX* software, low cost, and industry standardization. Try a $2500 25MHZ 386 zero wait state machine with a $1200 380MB Maxtor (Advertised at Fry's at retail here in Fremont), and a $2000 co-processor based 1660 x 1200 19" display, and *then* talk about what a great "deal" the slow-poke NeXT machine is. You know, they may not be elite, but it sounds like Intel processor AT bus machines have already won. If Mac is coming on strong, that's great because it's a real product with a real future and it makes the PC/workstation issue at least something of an interesting race. - Dimitri Rotow
jgreely@diplodocus.cis.ohio-state.edu (J Greely) (12/23/88)
In article <5567@cbmvax.UUCP> daveh@cbmvax.UUCP (Dave Haynie) writes: >in article <29866@tut.cis.ohio-state.edu>, J Greely says: >> ... when an application doesn't know how to handle refreshes >> properly. > >Aargh!! No, no! Please don't say that this system doesn't support "SMART" >refresh windows! This (an application that didn't update itself properly) has happened to me twice, under odd conditions. My suspicion is that it is an alpha-bug, and will be history by 0.9. Most of the real clumsy things are listed as "fixed by 0.9", and the evolving-interface class is on the list (basically, it looks like the user-interface guidelines aren't set in stone yet, so there are mostly-minor differences between applications). I think the whole point of building an application with the object-oriented approach is that a window class (proper Objective C wording? Haven't gotten any docs on it yet to find out) will handle the refresh message itself (ouch, he says, as he opens himself to charges of pure ignorance). Oh well, time to hit the bookstore for a good book on Objective C. Suggestions, anyone? -=- J Greely (jgreely@cis.ohio-state.edu; osu-cis!jgreely) "Give him some orange juice and a sugar cookie, ... maybe a tetanus shot. He'll be fine."
casseres@Apple.COM (David Casseres) (12/24/88)
In article <5567@cbmvax.UUCP> daveh@cbmvax.UUCP (Dave Haynie) writes: >Aargh!! No, no! Please don't say that this system doesn't support "SMART" >refresh windows! This is supposed to be a "real" computer with lots of >memory. If the application doesn't know how to refresh a window, it >shouldn't have to. Let the OS handle all aspects of a window. Just cause >the Mac doesn't (or didn't, maybe it does now?) handle such things doesn't >mean the rest of the world should stay so barbaric. Gee, just what are "all" aspects of a window? Should the OS also do the selection and highlighting of objects in the window when the user clicks the mouse? Maybe it should guess what the application wants to draw in the window, and do the drawing? After all, refreshing a window does in- volve drawing its content. Or maybe you mean that refreshing should only be done by saving the bits of the window's current image, then restoring them when a refresh is needed. Too bad if the content of the window wanted to change before the refresh. Aarch to you too, and happy holidays! David Casseres
anand@vax1.acs.udel.EDU (Anand Iyengar) (12/24/88)
In article <315@belltec.UUCP> dar@belltec.UUCP (Dimitri Rotow) writes: >I don't understand the viewpoint in this group that PC's only run DOS or >OS/2. Don't netfolks realize that half of the UNIX market these days >(by number of units) ships on '386 AT clones? Why do you think 386 boxes >have eaten up the low half of the UNIX market? ... >You know, they may not be elite, but it sounds like Intel processor AT bus >machines have already won. MS-DOS compatibilty presently counts for a lot (and shows how strongly the business environment respects IBM -- and now, Compaq). Other reasonably cheap UNIX boxes have existed for a while, but didn't sell as well. Not everybody is buying '386 machines to run UNIX. Some people just want a a faster DOS machine. It can be argued that MS-DOS helped make many people more ready for UNIX (MS-DOS == UNIX without the good parts). >What do you think '386's will do to the workstation market? Quite a bit I hope. >Just about anybody's plug in graphics board for UNIX and X will easily >outperform the NeXT machine in resolution, "look and feel" display speed, >or any other measure. Slam a low-cost Maxtor or CDC at 380 MB into your >low cost 25MHZ 80386 machine (not to mention the new 32 MHZ '386 machines >or even '486 machines people are announcing) and you will eat the NeXT >box alive in speed, quantity of *UNIX* software, low cost, and industry >standardization. Try a $2500 25MHZ 386 zero wait state machine with a >$1200 380MB Maxtor (Advertised at Fry's at retail here in Fremont), and >a $2000 co-processor based 1660 x 1200 19" display, and *then* talk >about what a great "deal" the slow-poke NeXT machine is. I have a report at home which benchmarks a '386 and a 68020 (similar clock speeds) and finds performance close (each outperformed the other in different parts of the test, but not by much (I think less than 20%, back and forth)). I recall reading a report from Motorola which says that the 68030 can perform 50% better than than the 68020 (2 cycles/mem-access vs. 3 for the 68020). I've worked on 16 and 20 MHz 386 machines for about a year (Compaq's). Both were quite fast until you went to the disk (granted, we only (<-?) had 2.5M of memory, but many times I/O is unavoidable). If the next machine has widened this bottleneck, then it may do well. Also note that the above prices are for the machine with out a real operating system (quench the flames from MS-DOS users -- reality is in the eyes of the beholder). SCO Xenix is $1000 ($700 for a bare-bones system, methinks); you'll want the C-compiler to port X, etc. Also the DSP, EtherNet, D/A, A/D (all of varying usefulness, mind you), and software to drive them add cost. >If Mac is coming on strong, that's great >because it's a real product with a real future and it makes the >PC/workstation issue at least something of an interesting race. It *is* great because it's more competition in the race, but not really that wonderful as new-&-improved-machine value. The Mac OS still doesn't have multitasking -- which the Amiga does (and has for quite a while I'm still waiting to see what Commodore makes of the 2500. Could be interesting). The Mac could also afford to be faster. Anand Iyengar. -- {arpa | bit}net: anand@vax1.acs.udel.edu (<- prefer), iyengar@udel.edu csnet: iyengar%udel.edu@relay.cs.net uucp: ->unidot ->!cfg!udel!udccvax1!anand ->uunet -/ ->!harvard ->udel.edu!iyengar ->!allegra!berkeley -/
peter@ficc.uu.net (Peter da Silva) (12/27/88)
In article <361@internal.Apple.COM>, casseres@Apple.COM (David Casseres) writes: > In article <5567@cbmvax.UUCP> daveh@cbmvax.UUCP (Dave Haynie) writes: > > >Aargh!! No, no! Please don't say that this system doesn't support "SMART" > >refresh windows! Let the OS handle all aspects of a window. I'm afraid, Dave, that simple-refresh has become somewhat of a standard. With all the performance loss this implies. > Gee, just what are "all" aspects of a window? Everything in the window that needs to be managed as a result of other programs' actions. This includes having the window obscured by other windows, depth-arranged, and moved. [ including acknowledgement of the users actions ] To improve effective response time, yes. Certainly the menu system should work asynchronously with the program, and so should statically located gadgets. In the controls industry local acknowledgement has proven very effective in cutting down user's frustration. > Or maybe you mean that refreshing should only > be done by saving the bits of the window's current image, then restoring > them when a refresh is needed. That's right. > Too bad if the content of the window wanted > to change before the refresh. Nope, because since the task is rendering through a a library that handles all the clipping rectangles anyway, it's trivial to have those rectangles that represent the off-screen part of the window point to temporary bitmaps of the right size and shape that are simply copied back in when the window gets screen real-estate again. We're talking about a multitasking operating system that routinely has many windows run by many programs visible at once. Now, suppose you use simple- refresh windows and then do something that requires several programs be notified. Say, you move a middling size window from a place where it's overlapping three other windows to off-screen. With smart-refresh windows, the window mgr runs through the clip-list once, and everything's fine. With simple-refresh windows, the window mgr sends a message to each of the programs, Now, it's got two choices... it can wait for the program to respond, or it can let them update asynchronously. In the first choice, the screen is effectively locked until (in this case) four seperate programs get the message, do their rendering, and reply. this means that any program has to be written so it will never wait more than a second or so before checking the window mgr. This isn't so bad in Mach, with threads, but it's a real pain on the Mac (Mac applications end up looking much like device drivers). In the second choice, the burden of waiting for programs falls on the user. You have to realise that what you see on the screen isn't necessarily real life. In either case the perceived performance of the machine goes way down. My Amiga (with a 68000) seems faster than a Mac-II (with a 68020) simply because it has better real-time response to a user's actions. I find on the Mac that I get in the habit of never arranging windows and sticking to one program (instead of hopping around) because it's so frustrating. -- Peter da Silva, Xenix Support, Ferranti International Controls Corporation. Work: uunet.uu.net!ficc!peter, peter@ficc.uu.net, +1 713 274 5180. `-_-' Home: bigtex!texbell!sugar!peter, peter@sugar.uu.net. 'U` Opinions may not represent the policies of FICC or the Xenix Support group.
pd@sics.se (Per Danielsson) (12/28/88)
In article <361@internal.Apple.COM>, casseres@Apple (David Casseres) writes: >Gee, just what are "all" aspects of a window? Should the OS also do the >selection and highlighting of objects in the window when the user clicks >the mouse? Maybe it should guess what the application wants to draw in >the window, and do the drawing? After all, refreshing a window does in- >volve drawing its content. Or maybe you mean that refreshing should only >be done by saving the bits of the window's current image, then restoring >them when a refresh is needed. Too bad if the content of the window wanted >to change before the refresh. Moving a window should not require redrawing the contents. From the applications point of view it shouldn't matter that the user moves the window. The application shouldn't even know about it. I don't see why this is a problem at all. These kind of problems with window systems were solved years ago. PD Per Danielsson UUCP: pd@sics.se (or ...!enea!sics!pd) Swedish Institute of Computer Science PO Box 1263, S-164 28 KISTA, SWEDEN "No wife, no horse, no moustache."
jesup@cbmvax.UUCP (Randell Jesup) (12/29/88)
In article <361@internal.Apple.COM> casseres@Apple.COM (David Casseres) writes: >In article <5567@cbmvax.UUCP> daveh@cbmvax.UUCP (Dave Haynie) writes: >>Aargh!! No, no! Please don't say that this system doesn't support "SMART" >>refresh windows! This is supposed to be a "real" computer with lots of >>memory. If the application doesn't know how to refresh a window, it >>shouldn't have to. Let the OS handle all aspects of a window. >Gee, just what are "all" aspects of a window? Should the OS also do the >selection and highlighting of objects in the window when the user clicks >the mouse? Maybe it should guess what the application wants to draw in >the window, and do the drawing? After all, refreshing a window does in- >volve drawing its content. Or maybe you mean that refreshing should only >be done by saving the bits of the window's current image, then restoring >them when a refresh is needed. Too bad if the content of the window wanted >to change before the refresh. You misunderstand. By "Smart Refresh" windows, Dave means windows in which off-screen portions are saved. Drawing to off-screen areas goes into those saved areas, so if that window is uncovered, the display is up to date VERY fast compared to most applications redrawing through a damage list. It turns out nExt has something similar, but the off-screen bitmap is the size of the window. Eats memory, but they burn memory like there's no tomorrow. Amigas have something similar, called SuperBitMap windows, where there's an off-screen bitmap, but the bitmap can be larger than the window, and can be scrolled (relative to the window). The window is a window into a larger display. -- Randell Jesup, Commodore Engineering {uunet|rutgers|allegra}!cbmvax!jesup
jesup@cbmvax.UUCP (Randell Jesup) (12/29/88)
In article <2596@udccvax1.acs.udel.EDU> anand@vax1.acs.udel.EDU (Anand Iyengar) writes: > I have a report at home which benchmarks a '386 and a 68020 >(similar clock speeds) and finds performance close (each outperformed the >other in different parts of the test, but not by much (I think less than 20%, >back and forth)). I recall reading a report from Motorola which says >that the 68030 can perform 50% better than than the 68020 (2 cycles/mem-access >vs. 3 for the 68020). An '030 should run (give or take) around twice as fast as an '020, especially if you're using the MMU. The data cache helps as well. > I've worked on 16 and 20 MHz 386 machines for about a year (Compaq's). >Both were quite fast until you went to the disk (granted, we only (<-?) >had 2.5M of memory, but many times I/O is unavoidable). If the next >machine has widened this bottleneck, then it may do well. PClones have pretty dismal I/O rates. On a 68000 Amiga, 7.xx MHz, with a SCSI controller, you can get 650K-800K _bytes_ per second through the file system on large reads, 1.1 MB/s using the driver directly, off good disks like the Quantum 80S or the CDC Wren V. -- Randell Jesup, Commodore Engineering {uunet|rutgers|allegra}!cbmvax!jesup
hyc@math.lsa.umich.edu (Howard Chu) (12/29/88)
In article <5590@cbmvax.UUCP> jesup@cbmvax.UUCP (Randell Jesup) writes: > PClones have pretty dismal I/O rates. On a 68000 Amiga, 7.xx MHz, >with a SCSI controller, you can get 650K-800K _bytes_ per second through the >file system on large reads, 1.1 MB/s using the driver directly, off good >disks like the Quantum 80S or the CDC Wren V. (Not to be outdone...) [sorry for this blatant me-too'ism, but...] I get 1MB/s thru the file system on my ST, using a Quantum 80S. I suspect a good deal of this is due to the drive's track caching, although the speed still holds up for fragmented disk accesses. Also, the Atari ST runs the 68000 at 8.0 MHz, but the little difference can't account for all the difference in throughput. And, on a side note, I have to wonder about what you consider "dismal I/O rates." PCs with ESDI drives (alright, AT clones or somesuch) get 12 Mbit/S throughput, accessing the driver directly. That's easily in excess of 1MB/s. Meanwhile, since we're on this topic... What kind of "winchester" drive is in these NeXT cubes, anyway? They seem godawful slow... (Yeah, I know the optical disk is slow. And clunky. Gotta wonder about this thing...) I thought this mainframe-on-a-chip channel architecture was supposed to make I/O operations fly... Let a "find" loose on the root directory, and watch your performance plummet. (It's pretty odd, when the only way to tell if your machine is still running is to put your hand on the cube to feel if the drive is spinning or not... Even the window manager stops.) -- / /_ , ,_. Howard Chu / /(_/(__ University of Michigan / Computing Center College of LS&A ' Unix Project Information Systems
daveh@cbmvax.UUCP (Dave Haynie) (01/04/89)
in article <2596@udccvax1.acs.udel.EDU>, anand@vax1.acs.udel.EDU (Anand Iyengar) says: > I recall reading a report from Motorola which says that the 68030 can > perform 50% better than than the 68020 (2 cycles/mem-access vs. 3 for the > 68020). Ideally, yeah, and that's ignoring the data cache effects. If you wanted that '020 to run UNIX, you might add a 68851, which adds a wait state, so in more practical terms, the '030 is probably better than 100% faster than an '020 at the same speed. If you let it be. The NeXT has slower system memory, and so takes advantage of another '030 feature, burst fill of cache lines. This lets it eat 4 longwords in 9 cycles (according to the reports of the NeXT I've read, at least), though writes are probably more like 3 or 4 clocks long. In other systems I've found burst mode to increase system speed by about 20% in practice (you don't always use all 4 longwords), but that depends on the memory speed difference between burst and non-burst. With similar memory systems I'd expect an '030 system to very easily outrun a '386 system, especially running UNIX. However, that may not be the real-world case -- there are a number of high-performance '386 system, using external cache and faster DRAM, that may give NeXT a run for the money, at least in terms of CPU performance. Since many of these are based around the otherwise primitive PC-AT architecture, I'd bet that a suitably equipped NeXT would still come out far ahead in I/O performance, though under UNIX, you might not get to see the difference very often. > Mac OS still doesn't have multitasking -- which the Amiga does (and has for > quite a while I'm still waiting to see what Commodore makes of the 2500. > Could be interesting). Glad you think so! > Anand Iyengar. -- Dave Haynie "The 32 Bit Guy" Commodore-Amiga "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: D-DAVE H BIX: hazy Amiga -- It's not just a job, it's an obsession
abe@mace.cc.purdue.edu (Vic Abell) (01/04/89)
Running the dhrystone test (version 1.1), our Sequent Symmetry (80386 at 16 mhz) with write-back cache is about 6% faster than my NeXT. I suspect some of the difference is a result of better compiler technology on the Symmetry.