stevej@studsys.mu.edu (jovanovic) (08/04/89)
Dear Apple,
As we all know, the Mac is a unique machine with some incredible software
as of yet unmatched on any other platform. Software, and the Mac interface,
are great, but I'd really like to know what's up in hardware development?
I think that over the past few years you've had enough time to research, so
when are we going to see some new developments.
First, let's talk about multitasking. Why is it that everyone in Apple
Land is swearing that OS/2 & Presentation Manager are slow and horrible,
and that the Mac interface is the ONLY religiously sanctioned way to go?
OS/2 & PM fly!! and PM is almost exactly like the Mac interface (hence the
lawsuit). Nevertheless, behind the interface you need some powerful silicon,
yet you have consistently refused to implement even such a simple thing as
DMA, even though you've had years in which to do it. Multitasking is
impossible without DMA, because virtual memory is a joke without disk DMA.
Aside from the software issues involved in multitasking, the Amiga will
outdo a Mac II performace wise with its myriad of coprocessors. Come on,
guys, everyone who has an Amiga laughs at Mac II owners when they compare
hardware. Even an IBM PC has DMA. It's not a new thing. Also, DMA chips
aren't expensive, but I'm sure Apple Marketing will manage to forget that
small fact when a new Mac comes out.
OK, what about a video coprocessor. You want to put a RISC chip on a
NuBus card to handle video, right? When! We should have had this back
when the Mac II was first introduced! Also, why is it that the Mac II
family's processors are run at 16MHz, but the NuBus only at 10MHz.
NuBus standards, sure, but I think it's time to refine standards when they
impede design progress, don't you. Steve Jobs has got it right. 25MHz
processor, 25MHz NuBus.
Finally, how about putting in a REAL processor in a new machine. No more
16MHz, sorry I'm on my coffee break, 68020s and 68030s. Give us 33MHz
processors, and for God's sake implement a large, fast SRAM cache!
I love my Mac II for the software it can run, but if the Amiga could run
all that software, I'd sell my Mac II and buy an Amiga in a nanosecond!
On the other hand, for its awesome power, I wouldn't mind having a NeXT
machine. How hard can it be to keep up with the competition with your'
massive research and development budget. Either innovate in hardware,
or be prepared to face the market consequences. I'm ready to bet
everything on NeXT!!
Show the readers of comp.sys.mac that you do care about them as Mac users
and will for once do something out of innovation's sake rather than greed's
sake. Give us an answer to the issues raised in this posting, or don't
reply at all--that will definitely say something about the future of Apple
Computer!
Steve J
---------------------------------------------------------------------------
reply to: stevej@studsys.mu.edu
---------------------------------------------------------------------------
dorourke@polyslo.CalPoly.EDU (David M. O'Rourke) (08/04/89)
stevej@studsys.mu.edu (jovanovic) writes: [some stuff about what Apple should do.] What a Jerk!!! You know people are always putting down Apple for not doing the right thing. But you know they are nearly a 4 billion dollar company, and have consistantly offered one the finest personal computers on the Market. Regardless of what Marketing people might think, people just don't throw money at you because you have slick image, at some point that image has to produce. Apple makes money because their products do a pretty good job of living up to the Image that their marketing department puts forth. The Amiga's a nice machine, but it just isn't up to the caliper of the Macintosh when it comes to overall concept. The OS may be multi-tasking, and the hardware maybe nice, but I've yet to see it do anything better than the mac besides games. For better or for worse the Mac community has the developer community behind it, and as long as that continues the software will get done on the Mac first and other platforms afterwards. Talk with any Amiga owner who isn't religous, and they'll tell you the Amiga has it's fair share of short commings as well. EVERY COMPUTER DOES!!! IMHO OS/2 & presentation manager are fast, but given the hardware that they're running on top of they should be a lot faster. One of IBM ps 2's running a full blown OS config. and presentation manger feels like your running a Mac plus, that says a lot for the plus and not very much for the ps 2. Apple doesn't have to answer to this joker or anyone else. The market place will dictate what Apple does and how they do it. I'm not even sure this guy completly understand's DMA and all the ramifications of the different design decisions/choices avalible to Apple. I'd love to see this joker design Apple's next machine and then yell at them because what he designed now doesn't run his favority software package. Prototypes are easy, products are much more difficult. And the Mac community is notoriously picky when it comes to products. As far as Apple sticking with the "standard" 10 mhz Nu-bus, this is good for once. I never thought I'd hear someone complain because the Macintosh was "standard" :-) When you up the clock rate on a machine you up the requirments for all the support hardware and well. More expensive memory, more expensive I/O devices ect... People already complain about the price of mac's, I don't think that upping the clock rate and all of the associated changes that go along with it would lower that price point any. Apple may make mistakes. But basicly people do the best they can, and Apple like any other company is made up of people. But to be quite honest with you I can't think of a personal computer on the market that I'd rather have, and I thank the people at Apple for doing the best they can to keep the best computer on the market. If you really want you're welcome to buy an Amiga, PS/2, NeXT or whatever, but you must have bought the Mac for some reason over the other computers, so quit your whining and try to remember why you bought the Mac in the first place. It was because you felt it was the best computer for your needs, don't try and tell me that behind all of the Amiga's limited techincal gadget's that they don't wish of a developer community as healthy as the Mac's, or wish for as much software as the mac has. I do wish for faster hardware, but then again everyone does, and always will. Apple, IBM, NeXT, SUN, HP or anyone else will never build a computer fast enough for me. And given that realization, I think Apple and it's strong developer community will do just fine for a long time to come. Just my $0.02 worth. I speak for no one but myself. -- \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\|///////////////////////////////////////// David M. O'Rourke____________________|_____________dorourke@polyslo.calpoly.edu | God doesn't know, he would have never designed it like that in the first | |_ place. ____________________________________________________________________|
goldman@apple.com (Phil Goldman) (08/04/89)
In article <577@studsys.mu.edu> stevej@studsys.mu.edu (jovanovic) writes: > [lots of ridiculous complaints about the MacII] > > Steve J ^^^^^^^ Hmmmm...
paul@taniwha.UUCP (Paul Campbell) (08/04/89)
In article <577@studsys.mu.edu> stevej@studsys.mu.edu (jovanovic) writes: >impede design progress, don't you. Steve Jobs has got it right. 25MHz >processor, 25MHz NuBus. Let's settle this once and for all the NextBus is a 12.5MHz NuBus that bursts two words/cycle - this means that (for a 68k family motherboard) you have to have your graphics engine at least partly on the motherboard in order to generate write bursts (68030 chips only generate burst reads), also many graphics ops involve read-modify-write cycles which NuBus doesn't burst at all, you would have to burst in a chunk RMW them then burst them out when you are done which means you can't overlap memory ops with graphics ALU ops. If you can't burst then you are stuck with a minimum of 2 12.5MHz cycles per video access (or 160nS - about the right amount for DRAM cycle time anyway). Apart from this misconception (not helped by Next's marketting :-) the NextBus is really a great piece of work, it is not terminated, except by the source driving impedance (ie it bounces exactly once and then is absorbed back where it came from within an easily determined time which can be taken into account and ignored due to having a synchronous and fixed length bus) this really saves on power and lets you drive the bus with CMOS right out of your gate-array rather than through expensive power-hungry drivers. Clock distribution is different of course (it has to be very clean) and there are (25MHz and 12.5MHz) two of them letting one do a faster state machine on a card without having to fall all over the place recovering 20MHz from 10MHz etc With my software hat on I like the fact that every card has an single known location in which to check for a pending interrupt and to clear it. Paul -- Paul Campbell UUCP: ..!mtxinu!taniwha!paul AppleLink: D3213 "Free Market": n. (colloq.) a primitive fertility goddess worshipped by an obscure cult in the late 20th C. It's chief priest 'Dow Jones' was eventually lynched by an enraged populace during an economic downturn (early 21st C).
ts@cup.portal.com (Tim W Smith) (08/12/89)
Ok, I give up. Where did you get the idea that multitasking requires virtual memory? Tim Smith
davew@hp-ptp.HP.COM (Dave_Waller) (08/15/89)
ts@cup.portal.com (Tim W Smith) writes: > Ok, I give up. Where did you get the idea that multitasking > requires virtual memory? > > Tim Smith > ---------- Well, Tim, it doesn't REALLY, but VM sure makes it a hell of alot easier to implement. With VM, as far as the processor is concerned, EVERY process' Working Set is in memory, allocated to a different piece of a very large virtual addressing space. When it comes time to process switch, the dispatcher need not decide whether or not a process' WS is actually in physical memory, it simply switches to the process and allows the WS to page fault in with the VM handler if it isn't in memory. Ask anyone who's tried to implement a clean multitasking system without VM, and they'll tell you being strapped to a truck tire on a cross country trip would be more fun. Dave Waller Hewlett-Packard Co. Workstation Group Pacific Technology Park 1266 Kifer Rd. Sunnyvale, CA (408) 746-5324 [ucbvax!]hplabs!hpdstma!dave | dave@hpdstma.ptp.hp.com +-------------------------------------+---------------------------------------+ | Standard disclaimer: | "I refuse to put anything in quotes | | The opinions expressed above are | in this space" | | solely my own, and in no way reflect| | | those of my employer. | | +-------------------------------------+---------------------------------------+
mnkonar@manyjars.SRC.Honeywell.COM (Murat N. Konar) (08/16/89)
In article <21223@cup.portal.com> ts@cup.portal.com (Tim W Smith) writes: >Ok, I give up. Where did you get the idea that multitasking >requires virtual memory? > It doesn't of course. It is just a lot nicer if you have a lot of memory to multitask in (virtual or otherwise) just so you can get more applications going. ____________________________________________________________________ Have a day. :^| Murat N. Konar Honeywell Systems & Research Center, Camden, MN mnkonar@SRC.honeywell.com (internet) {umn-cs,ems,bthpyd}!srcsip!mnkonar(UUCP)