BARRETT%FOREST.ECIL.IASTATE.EDU@cunyvm.cuny.edu (Marc Barrett) (03/27/90)
FLAME ON!!!!! Once again, Commodore has let Apple slip ahead of them in the area of technological innovation. Had Commodore released (or at least publicly shown) their video card, developed by the University of Lowell, they might have captured some headlines for new innovations. But I doubt that the introduction of the video card will get any attention now that Apple's new 24-bit graphics accelerator has been shown publicly. The Lowell card pails next to this thing, which will do animation as fast as the Lowell card, but with full 24-bit color. Also once again, Commodore has let Apple slip ahead of them in the area of new technological innovatiion. Had Commodore released (or at least publicly shown) the Amiga 3000, they might have captured some headlines for developing innovative new computer systems. But I doubt that the introduction of the Amiga 3000 will get any attention now that Apple's 40Mhz 68030-based MAC IIFX has been shown. The Amiga 3000 pails next to this thing, which included a plethura of newly developed custom chips. A total of ELEVEN custom chips are used to give the MAC IIFX impressive speed while relieving the CPU of a lot of I/O and sound duty. Unlike the Amiga 3000, however, all of these custom chips run at the full speed of the microprocessor. flame off
daveh@cbmvax.commodore.com (Dave Haynie) (03/27/90)
In article <15003@snow-white.udel.EDU> BARRETT%FOREST.ECIL.IASTATE.EDU@cunyvm.cuny.edu (Marc Barrett) writes: >FLAME ON!!!!! > Once again, Commodore has let Apple slip ahead of them in the >area of technological innovation. Had Commodore released (or at >least publicly shown) their video card, developed by the University >of Lowell, they might have captured some headlines for new >innovations. Commodore has publically shown the ULowell card. The last show I attended that had it on display was the World of Commodore show in Toronto, where it was running the X windows, in color of course, under AMIX. >But I doubt that the introduction of the video card will get any attention >now that Apple's new 24-bit graphics accelerator has been shown publicly. The Apple card sounds pretty good, even at the $2000 price tag. But it's mainly good for Apples, of course, since this is the first card of it's kind (eg, general purpose QuickDraw engine) shown for an Apple machine. But a big part of the announcement is that Apple would have you believe this is the first time hardware graphics acceleration has taken place. Sort of like the way they invented multitasking with Multifinder was released. Of course, this is an old technique, pioneered by IBM, who, by the way, invented the 3.5" disk several years after it was used by both Apple and Commodore, and also managed to invent multitasking, several years after Commodore used it and zillion after it was invented, or even available on a personal computer. >A total of ELEVEN custom chips are used to give the MAC IIFX impressive speed ... Actually, a normal, every-day 32k external cache has more to do with the Mac IIfx going fast than anything else. What Apple terms "custom chips" are simple gate arrays. They are, for instance, doing some kind of DMA transfer for hard disk I/O, rather than the 8 bit programmed I/O they've used in the past. Pretty much what we've been doing all along. While most of the Mac IIfx does go faster than the 7.16MHz of the A2000, everything I've seen so far indicates only the cache is running a real 40MHz 68030 cycle. Apparently Apple's hype "works good". I have yet to see a hard-core architectural description of the IIfx, just the stuff being discussed on the Apple net groups. NeXT's "mainframe on a chip" made things sound real good, until you opened the hood and took a look around. >Unlike the Amiga 3000, however, all of these custom chips run at the full speed >of the microprocessor. Well, you certainly seem to know quite a bit about the Mac IIfx and the A3000. Care to reveal some real sources (eg, print articles) so's I can read up on 'em too? BTW, for those who haven't been paying attention, all the IIfx display cards run over NuBus. The CPU slams on the brakes every time it talks to NuBus. There is one card that provides an AMD 29k at 30MHz to execute QuickDraw commands, and costs $2000. Really not bad for a video display with it's own CPU -- you can pay over twice that for a similar unit on a PClone. Of course, you can also buy A2000s for most rooms in the house for the price of one decked out IIfx. In any case, this can really make QuickDraw scream -- Apple claims 5x to 20x speedups. Of course, if you're using one of those programs that's working internally thinking in Postscript, and just uses QuickDraw to put the pixels up on the screen, you won't go any faster on this Mac setup. -- Dave Haynie Commodore-Amiga (Systems Engineering) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy Too much of everything is just enough
karl@sugar.hackercorp.com (Karl Lehenbauer) (03/27/90)
In article <15003@snow-white.udel.EDU> BARRETT%FOREST.ECIL.IASTATE.EDU@cunyvm.cuny.edu (Marc Barrett) writes: >FLAME ON!!!!! > > Once again, Commodore has let Apple slip ahead of them in the >area of technological innovation. Had Commodore released (or at >least publicly shown) their video card, developed by the University... 1. The 2FX is vaporware in that it will not ship for months. 2. Apple will not be able to sell a whole lot of them at the quoted price. 3. Apple will be struggling to compete against Windows 3.0, etc, as the PC is closing the gap in graphical user interfaces The Amiga may have some of the same struggle on 3, but as even the $550 A500 has a realtime preemptive exec, shared libraries and a rabid following of talented hackers who are intent on making use of that stuff. I am still hopeful. Further, Macs are expensive. Amigas are not. Further, Macs crash and burn. Mac people don't like to talk about it, but the ones I know have a lot of mystery crashes. Plus, small sample size here, but they tend to catch fire (because of no fan, perhaps?). A buddy of mine has shelled $180 four times in the last three years to replace his digital board (can't really call it a motherboard 'cause there're no slots on the affordable ones, right?) after his analog board smoked and fried it. Another guy I just met paid $180, too, for the same reason. Ask around. And after Apple screwed everyone who bought Apple Threes, then screwed everyone who bought Lisas, then screwed everyone who bought 128K Macs, then sued Microsoft, etc, etc, you're not going to get me or a lot of other hackers (yes, I fancy myself one) near one, other than to projectile vomit at how slow the 68000-based ones are at window manipulation and how hard they are to program. -- -- uunet!sugar!karl "As long as there is a legion of superheros, all else -- can surely be made right." -- Sensor Girl -- Usenet access: (713) 438-5018
es1@cunixb.cc.columbia.edu (Ethan Solomita) (03/27/90)
In article <5463@sugar.hackercorp.com> karl@sugar.hackercorp.com (Karl Lehenbauer) writes: >3. Apple will be struggling to compete against Windows 3.0, etc, as the > PC is closing the gap in graphical user interfaces > I wouldn't count on it! I am repeatedly amazed at how much money people can be made to spend on a Macintosh. I never thought they could sell a Mac Plus, but I am surrounded by them. It's like with IBM, people just buy them without thinking. >The Amiga may have some of the same struggle on 3, but as even the $550 A500 That depends what the A3000 costs. If the A3000 has a built-in flicker fixer, and is cheap, ... -- Ethan Ethan Solomita: es1@cunixb.cc.columbia.edu "If Commodore had to market sushi they'd call it `raw cold fish'" -- The Bandito, inevitably stolen from someone else
robin@sabre.austin.ibm.com (Robin D. Wilson/1000000) (03/28/90)
>In article <15003@snow-white.udel.EDU> BARRETT%FOREST.ECIL.IASTATE.EDU@cunyvm.cuny.edu (Marc Barrett) writes: >>Unlike the Amiga 3000, however, all of these custom chips run at the full speed >>of the microprocessor. I thought that on the A1000/500/2000 line the custom chips ran 2x as fast as the main processor. So Apples manage to slow their down, and set some new records for speed; hmm. The real speed increase is in the rate at which their marketing mumbo-jumbo has proliferated. BTW, how do you know that the A3000's custom chips don't "run at the full speed of the microprocessor"? Has this stuff been announced already and they left me out? +-----------------------------------------------------------------------------+ |The views expressed herein, are the sole responsibility of the typist at hand| +-----------------------------------------------------------------------------+ |USNail: UUCP: | |2323 Wells Branch Pkwy., #G107 cs.utexas.edu!romp!ibmchs!auschs\ | |Austin, TX 78728 !sabre.austin.ibm.com!robin | |Home: (512)251-6889 ^^^^^^^^^^^^^^<-MUST BE INCLUDED| +-----------------------------------------------------------------------------+
fiddler@concertina.Sun.COM (Steve Hix) (03/28/90)
In article <10363@cbmvax.commodore.com>, daveh@cbmvax.commodore.com (Dave Haynie) writes: > In article <15003@snow-white.udel.EDU> BARRETT%FOREST.ECIL.IASTATE.EDU@cunyvm.cuny.edu (Marc Barrett) writes: > > Once again, Commodore has let Apple slip ahead of them in the > >area of technological innovation. Had Commodore released (or at > > Commodore has publically shown the ULowell card. The last show I attended > > >But I doubt that the introduction of the video card will get any attention > >now that Apple's new 24-bit graphics accelerator has been shown publicly. > > The Apple card sounds pretty good, even at the $2000 price tag. But it's > mainly good for Apples, of course, since this is the first card of it's kind > (eg, general purpose QuickDraw engine) shown for an Apple machine. SuperMac and Radius both are selling QuickDraw accelerators. RaterOps may also be shipping one. > But a big part of the announcement is that Apple would have you believe this is > the first time hardware graphics acceleration has taken place. They and AMD both ignore to above mentioned products...even though one of them uses the AMD 29000. Marketing trolls are curious thingies. > >A total of ELEVEN custom chips are used to give the MAC IIFX impressive speed ... > > Actually, a normal, every-day 32k external cache has more to do with the Mac IIfx > going fast than anything else. What Apple terms "custom chips" are simple gate > arrays. They are, for instance, doing some kind of DMA transfer for hard disk > I/O, rather than the 8 bit programmed I/O they've used in the past. Pretty much > what we've been doing all along. While most of the Mac IIfx does go faster than > the 7.16MHz of the A2000, everything I've seen so far indicates only the cache > is running a real 40MHz 68030 cycle. At least two of the chips are, essentially, versions of the 6502 running some of the I/O for the 68K. Weird memory that does latched read/writes. Expensive. Only from Apple so far. But a bit faster. > BTW, for those who haven't been paying attention, all the IIfx display cards run > over NuBus. The CPU slams on the brakes every time it talks to NuBus. There is > one card that provides an AMD 29k at 30MHz to execute QuickDraw commands, and > costs $2000. Really not bad for a video display with it's own CPU -- you can > pay over twice that for a similar unit on a PClone. Of course, you can also buy > A2000s for most rooms in the house for the price of one decked out IIfx. In any > case, this can really make QuickDraw scream -- Apple claims 5x to 20x speedups. Just about the exact same speedup shown by the SuperMac accellerator. ------------ "...Then anyone who leaves behind him a written manual, and likewise anyone who receives it, in the belief that such writing will be clear and certain, must be exceedingly simple-minded..." Plato, _Phaedrus_
fiddler@concertina.Sun.COM (Steve Hix) (03/28/90)
In article <5463@sugar.hackercorp.com>, karl@sugar.hackercorp.com (Karl Lehenbauer) writes: > > 1. The 2FX is vaporware in that it will not ship for months. Like the NeXT? (Surely not *that* long!) I seem to remember some wait between announcing the 2500 and actually seeing a real one. > 2. Apple will not be able to sell a whole lot of them at the quoted price. Sadly, they'll probably sell them as fast as they can produce them. > 3. Apple will be struggling to compete against Windows 3.0, etc, as the > PC is closing the gap in graphical user interfaces Windows n.0 is tracking a moving target. I'd place it a distant fifth in usable GUI's. > The Amiga may have some of the same struggle on 3, but as even the $550 A500 > has a realtime preemptive exec, shared libraries and a rabid following of > talented hackers who are intent on making use of that stuff. I am still > hopeful. Further, Macs are expensive. Amigas are not. Then why aren't they selling that much better than Macs? Maybe price isn't everything, and maybe CBM ought to consider what's not working. > Further, Macs crash and burn. Mac people don't like to talk about it, but > the ones I know have a lot of mystery crashes. Plus, small sample size > here, but they tend to catch fire (because of no fan, perhaps?). The smoking Plus. They don't catch fire, much less burn. One component, marginally specc'ed, fails and takes out the analog board (which cover both power supply and video driver duties). The fix is to get a better part, which fix has been well known for quite a while. Not by the dealers, though. (Sounds familiar, somehow.) Later Pluses, btw, use a heftier part and don't seem to smoke. Most of the remainder of the ab failures involve failing solder joints. > A buddy > of mine has shelled $180 four times in the last three years to replace his > digital board (can't really call it a motherboard 'cause there're no slots > on the affordable ones, right?) after his analog board smoked and fried it. > Another guy I just met paid $180, too, for the same reason. Ask around. They need to find a real tech to fix the problem right. Probably for less than $100. Ask around. (The Plus is the one that dies. Adding a fan usually does an endrun around the problem. Blame Steve Jobs for that one. The rest have fans.) > And after Apple screwed everyone who bought Apple Threes, How? By replacing all the original "fine line" ///'s with the fixed "coarse line" version for free? (Marketing/executive staff did some reall dumb things with the ///, but engineering was trying to do it right.) ------------ "...Then anyone who leaves behind him a written manual, and likewise anyone who receives it, in the belief that such writing will be clear and certain, must be exceedingly simple-minded..." Plato, _Phaedrus_
FelineGrace@cup.portal.com (Dana B Bourgeois) (03/28/90)
Who is this guy? The Human Torch?
There have been lots of messages here on the net pointing out:
1. You can't get the hardware yet...it is sales-talk only
2. While improved, the specs don't exactly blow anybody away.
3. The price is in work-station territory, not desk-top
4. The A3000 has not been announced so we don't exactly know what 'our'
side has up its sleeve.
So if you aren't the Human Torch, do you work for...
...SATAN??!?
Dana Bourgeois @ Cup.Portal.Com
"...or Apple Marketing??!??"dksnsr@nmtsun.nmt.edu (Dr. Mosh) (03/28/90)
Plus the fact that The Mac IIfx is aimed at the Workstation market, it has competitions galor. With the introduction of the IBM System 6000's... who wants a Mac??? -Dino Khoe -- ============================================================================= Dr. of Moshology | | | / \ /\ |/ Any system dksnsr@nmtsun.nmt.edu | |---||____|| |\ can be New Mexico Tech Computer Science | | || | \/ | \ cracked...
seanc@pro-party.cts.com (Sean Cunningham) (03/29/90)
In-Reply-To: message from BARRETT%FOREST.ECIL.IASTATE.EDU@cunyvm.cuny.edu Excuse me, but do you know something about the 3000 that isn't public knowledge??? The thing isn't even out yet, and you're saying it "pales" in comparison to the Mac...PLEASE! >From what I have on good authority, the A3000 has nothing to worry about. The new Mac is, yes, impressive. And that 24-bit board is nice. But let's keep things in perspective. 11 coustom chips? You mean those 6502s they're using to control I/O??? I see nothing impressive about that. The chip count on the motherboard is the SAME as previous Mac ][s. I'm not sure what they took away, but that's a fact. Big deal, they've got a couple of Apple ][s or C64s controlling their I/O...I'll stick with the Denise and the Paula, even at 7.16MHz. Sean //////////////////////////////////////////////////////////////////////////// UUCP: ...!crash!pnet01!pro-party!seanc | ARPA: !crash!pnet01!pro-party!seanc@nosc.mil | " I drank what? " INET: seanc@pro-party.cts.com | -Socrates | RealWorld: Sean Cunningham | Voice: (512) 994-1602 | | Call C.B.A.U.G. BBS (512) 883-8351 w/SkyPix | B^) VISION GRAPHICS B^) \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
seanc@pro-party.cts.com (Sean Cunningham) (03/29/90)
In-Reply-To: message from karl@sugar.hackercorp.com Nice flame there...hehe I'm a true believer that one good flame deserves another...the Amiga doom-sayers can bite silicon! Sean //////////////////////////////////////////////////////////////////////////// UUCP: ...!crash!pnet01!pro-party!seanc | ARPA: !crash!pnet01!pro-party!seanc@nosc.mil | " I drank what? " INET: seanc@pro-party.cts.com | -Socrates | RealWorld: Sean Cunningham | Voice: (512) 994-1602 | | Call C.B.A.U.G. BBS (512) 883-8351 w/SkyPix | B^) VISION GRAPHICS B^) \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
mithomas@bsu-cs.bsu.edu (Michael Thomas Niehaus) (03/29/90)
> Actually, a normal, every-day 32k external cache has more to do with the Mac > MacIIx going fast than anything else. What Apple terms "custom chips" are > smiple gate arrays. They are, for instance, doing some kind of DMA transfer > for hard disk I/O, rather than the 8 bit programmed I/O they've used in the > past. Pretty much what we've been doing all along. While most of the > Mac IIfx does go faster than the 7.16MHz of the A2000, everything > I've seen so far indicates only the cache > is running a real 40MHz 68030 cycle. Here's some real info from the Apple spec sheets. My commentary is added in brackets: % 40Mhz 68030 microprocessor and 68882 floating-point coprocessor. The Macintosh IIfx runs up to 100 percent faster than the Macintosh IIci, and up to 300 percent faster than the Macintosh IIcx and Macintosh IIx computers. [ This processor runs most of the time at zero wait states by using the cache in conjunction with the 80ns memory. This memory is also specially designed to allow burst access, and simultaneous reads and writes, creating an additional performance boost. ] % Built-in 32K static RAM (SRAM) cache stores the processor's most frequently used instructions. [ This is required. Otherwise you would need somewhere in the neighborhood of 40ns memory to get 0 wait states out of the machine. ] % dedicated SCSI DMA (Small Computer Systems Interface/Direct Memory Access) channel which reduces the workload of the main processor and speeds performance of the SCSI bus. [ Actually, there is nothing special about this chip. It is a standard controller. It does effectively double the speed of SCSI on the Macs. Also, with the IIfx, this is the first time that DMA was supported by the operating system. ] % dedicated I/O processors increase system efficiency. [ These I/O processors are really 6502 processors, the same as used in the old Apple II line of computers. They handle the floppy drive, Apple Desktop Bus (mouse and keyboard), and serial ports (which includes Apple's LocalTalk networking). This takes quite a load off the CPU as well. ] % six NuBus expansion slots can Accommodate multiple video, communications, networking, and other expansion cards. [ Same as all of the other Mac II models. No changes in speed or capabilities. ] % Processor Direct Slot (PDS) allows direct access to the system bus. [ This is new to the 6-slot machine as well. It gives the possibility of adding cards that run at a much higher speed than NuBus. I'm sure that someone will think of a good application for this. ] % six built-in ports: two serial ports, two Apple Desktop Bus ports, stereo sound jack, and a DB-25 SCSI interface. [ No change here either. ] % 512K ROM on SIMM includes support for 32-bit QuickDraw and A/UX-supported functions such as: 32-bit addressing and virtual memory, which extends the computer's internal memory by transparently treating the hard disk as though it were RAM. [ Same as the IIci. ] Here is some more information from the Apple releases. All marketing mumbo-jumbo has been removed wherever possible. Q: How many new Application Specific Integrated Chips (ASICs) are on the Macintosh IIfx board? A: There are a total of seven new ASICs. Two are Peripheral Interface Control (PIC) Chips, which are dedicated to controlling low-level routines such as serial communication and floppy disk access. There is one Small Computer System Interface/Direct Memory Access (SCSI/DMA) chip, which controls the SCSI bus. There is one Operating System Support chip, which is a multipurpose ASIC, dedicated to managing interrupts from external sources so they are synchronized with the internal bus. There is one Bus Interface Unit 30 chip and a Bus Interface Unit 2 chip, which drive NuBus processes and bus synchronization, respectively. There is one Faster Memory Controller chip, which controls access from the cache, ROM, and RAM to the system's main processor. Additionally, there are four ASICs which also appeared in the Macintosh IIci design. Q: What sort of advantages does the Macintosh IIfx get from the Peripheral Interface Control (PIC) chips? A: The PIC chips are important to the overall performance of the Macintosh IIfx system, as they balance the increase in clockspeed to 40 MHz by ensuring that I/O processes keep pace with the processor's speed. Additionally, they free the main processor from processing low-level interrupts. As a result, the PIC chips provide high levels of dedicated and predictable performance for standard I/O routines. For example, network routing techniques on the Macintosh IIfx are completely transparent to users performing any number of Macintosh applications. In future versions of the Macintosh operating system, the Macintosh IIfx will take greater advantage of the PIC chips' capabilities. That should give you enough information to dwell on for a while. -Michael -- Michael Niehaus UUCP: <backbones>!{iuvax,pur-ee}!bsu-cs!mithomas Apple Student Rep ARPA: mithomas@bsu-cs.bsu.edu Ball State University AppleLink: ST0374 (from UUCP: st0374@applelink.apple.com)
mithomas@bsu-cs.bsu.edu (Michael Thomas Niehaus) (03/29/90)
> 1. The 2FX is vaporware in that it will not ship for months.
I don't know about that one. The latest product availability listing from
Apple:
Item 1602065 20-March-90 15:21PST
Sub: Central Product Availability
March 20, 1990
PRODUCT AVAILABILITY
CHICAGO
----------
CURR.
FILL PROJ.
NUMBER DESCRIPTION EDOTEO DATE F/DATE
------- ----------- -------------------
M0031LL/A Mac SE 2/40 2 WK 3/14 3/28
M5392 Mac SE 030 1/FD 3 WK 3/16 3/16
M5510LL/A Mac IIFX 4/FD 4 WK 3/14 3/14
M5515LL/A Mac IIFX 4/80 2 WK 3/15 CURR
M5520LL/A Mac IIFX 4/160 7 WK 3/19 3/19
M5610 Mac IICX 1/40 1 WK 3/12 3/26
M5660 Mac IICX 1/FD 1 WK 3/14 3/26
M5710LL/A Mac IICi 1/FD 2 WK 3/01 3/26
M5745LL/A Parity Mac IICI 4/80 4 WK 10/09 10/09
M5790 Mac IIX 1/FD 3 WK 3/12 3/12
M5860 Mac IIX 4/160 3 WK 3/12 3/12
EDOTEO stands for Estimated Delay on today's entered orders. CURR FILL DATE
is the date of the oldest unfilled order. PROJ F/DATE is Apple's projected
fill date.
According to this statement, the Mac IIfx is currently shipping. There
is a 2-week backlog on the 4/80, which should be eliminated by the end
of the month, a 4 week backlog on the 4/FD (which won't improve by the
end of the month), and a 7-week backlog on the 4/160 (which also won't
improve).
-Michael
--
Michael Niehaus UUCP: <backbones>!{iuvax,pur-ee}!bsu-cs!mithomas
Apple Student Rep ARPA: mithomas@bsu-cs.bsu.edu
Ball State University AppleLink: ST0374 (from UUCP: st0374@applelink.apple.com)karl@sugar.hackercorp.com (Karl Lehenbauer) (03/29/90)
So Apple has discovered coprocessors, eh? Ironically, I recall a flame war a couple years back with a bunch of Mac people arguing that the Mac not having coprocessors was a *benefit*! Now that Apple has invented them, I gather they are the slickest things since laser-etched toast. Further, as the vendors of various PC boards have been learning, wimpy coprocessors can be a substantial impediment to performance. Many vendors are abandoning their 8088, 80186, etc, processors in favor of much higher performance ones in, for example, networking and multiport serial boards. Any Mac apologists care to compare the performance of a 6502 to the very-obsolete-and-never-particularly-attractive 8088? -- -- uunet!sugar!karl "I hate quotations. Tell me what you know." -- Emerson -- Usenet access: (713) 438-5018
coxc0010@ucselx.sdsu.edu (David Tse) (03/29/90)
Vaporware? The MacIIfx was in our school computer store 2 weeks ago. They got the portable Mac very early too. (before the retail store much sooner). the educational price at out school is: MacIIfx, one high density FD $6399 MacIIfx, 4 MB ram, 80 harddisk $6999 above w/ 160MB harddisk $7399 ?? as in most Mac, video card, keyboard, and monitor extra. 13" color monitor $705 keyboard $100 to 200 video card (256 color) $350 ?? new video card true 24 bit color $650 I have seen the demo file in it. and play around a bit. Is Amiga coming out with a true 24 bit graphics machine (not Targa or transputer) itself? When? I have been waiting for ECS and 1.4 for 2 years and when is it coming out? I am quite disappointed about Amiga being catched up by others, and Commodore is not improving it fast enough. Plus the clone market is vastly dropping in price quickly, amiga is losing even the price/performance ratio slowly. It took me a long time to decide buying my A500, my friend told me for a whole year about it, he doesn't even afford to buy one. Is there any circuit simulation software similar to MICROCAP II? I know the PD spice requires too much memory, am I right? I am always a proud Amiga owner. But by the way Commodore market the Amiga, when are they going to sell enough? so there will be good software covering all areas (CAD, Mathematica, business, and all the smaller area, Ada complier) ? flame? I am on the Amiga side! but I want the most useful all rounded for a good price ,(poor student, ya know). so I don't have to buy an Intel machine just to run Microcap II. If somebody can tell me when the ECS and 1.4 will be out, I appreciate. sincerely, David Tse
coxc0010@ucselx.sdsu.edu (David Tse) (03/29/90)
(somebody looked dowm on the 6502 based I/O controllers on the MacIIfx) Ok, fine, but please check out some good design besides this. look at the 24 bit graphics, look at the memory fetching, do you think by ignoring the only we can do is to hope Commodore can catch up before others catch up on amiga. david Tse
coxc0010@ucselx.sdsu.edu (David Tse) (03/29/90)
In article <1990Mar29.073036.18428@ucselx.sdsu.edu> coxc0010@ucselx.sdsu.edu (David Tse) writes: >(somebody looked dowm on the 6502 based I/O controllers on the MacIIfx) > >Ok, fine, but please check out some good design besides this. look at the >24 bit graphics, look at the memory fetching, do you think by ignoring the technical facts of our competition, you become a more loyal Amigan? >only we can do is to hope Commodore can catch up before others catch up >on amiga. > >david Tse > note: i miss the line in between.
bscott@pikes.Colorado.EDU (Ben M Scott) (03/29/90)
Remember also that the 6502s Apple is crowing about were the very heart and soul of... you guessed it, your friend and mine, the Commodore 1541! Marc has repeatedly claimed he has 'pure motives for stirring up trouble' here. I doubt there is such a thing as a pure motive for causing problems. I use the Amiga to do freelance animations for a video post-production house. I'm not an artist; 99% of my work is mechanical flip-n-tumble of the corporate logo and such. But one real advantage of the Amiga that almost NO other system can aspire to is REAL-TIME animation. Even the very highest-end systems do not do this very well; most of the fancy animation you see is single-frame recorded on to videotape. This makes for fine animation, but is not so great when you have a client sitting in your studio asking what you can do for him. Nothing compares to booting up DPaint III, roughing out something in 5 minutes and showing him what he's getting for his $60/hr or more... On a related topic, I have a request: I need a videotape or print versions of the infamous Mac 'Helocar' commercial. I am planning the "Helocar Killer" demo, and have many of the necessary primitives done or at least thought-out. It involves the Mac Helocar slowly taking off at a whole 3 frames-per-second (about all the Mac in the commercial could manage) and just as it's about to fly, in comes a futuristic antigrav-powered Amiga car which shoots the Mac into dust with twin Boing-ball cannons... trouble is, I've forgotten what the Helocar looks like. If anyone has a source for a picture, please mail me. . <<<<Infinite K>>>> -- _______________________________________________________________________________ | | | Someday, I'm going to make up a clever .sig file like everyone else has... | |_____________________________________________________________________________|
mfi@serc.cis.ufl.edu (Mark Interrante) (03/29/90)
In article <5479@sugar.hackercorp.com> karl@sugar.hackercorp.com (Karl Lehenbauer) writes: >So Apple has discovered coprocessors, eh? Ironically, I recall a flame >war a couple years back with a bunch of Mac people arguing that the Mac >not having coprocessors was a *benefit*! Note the difference in the mac coprocessor scheme, the processor is a 30mhz risc chip that uses downloaded code. It is not a custom chip that has hardwired graphics operations. >Further, as the vendors of various PC boards have been learning, wimpy >coprocessors can be a substantial impediment to performance. Many vendors >are abandoning their 8088, 80186, etc, processors in favor of much higher >performance ones in, for example, networking and multiport serial boards. > >Any Mac apologists care to compare the performance of a 6502 to the >very-obsolete-and-never-particularly-attractive 8088? The 6502 is used as an embedded controller for operations that are inherently low speed; ie. serial ports and keyboards. BTW the 6502 controller runs at 10mhz. ----------------------------------------------------------------------------- Mark Interrante Software Engineering Research Center mfi@beach.cis.ufl.edu CIS Department, University of Florida 32611 ----------------------------------------------------------------------------- "Imagine what it would be like if TV actually were good. It would be the end of everything we know." Marvin Minsky
hrlaser@crash.cts.com (Harv Laser) (03/30/90)
In article <3471@pikes.Colorado.EDU> bscott@pikes.Colorado.EDU (Ben M Scott) writes: > >On a related topic, I have a request: I need a videotape or print versions >of the infamous Mac 'Helocar' commercial. I am planning the "Helocar Killer" >demo, and have many of the necessary primitives done or at least thought-out. >It involves the Mac Helocar slowly taking off at a whole 3 frames-per-second >(about all the Mac in the commercial could manage) and just as it's about >to fly, in comes a futuristic antigrav-powered Amiga car which shoots the Mac >into dust with twin Boing-ball cannons... trouble is, I've forgotten what the >Helocar looks like. If anyone has a source for a picture, please mail me. > >. <<<<Infinite K>>>> Dig thru your oldish issues of INFO, Ben... they had a little blurb on a free 15 min. VHS tape that Apple's been giving out for a few months which extolls the virtues of Mac "multimedia" stuff. Most of the Helocar commercial animation is on that tape. If you have an Amiga Framegrabber you could just grab a frame (or even the whole animation) of the Helocar, slap it into a paint program, cut out the pieces you need, clean 'em up, and then add your Amiga antigrav Boing-cannon car, sit back, and wait for Apple to come and sue you :-) No, but all seriousness aside, there's an 800 number to get this tape and they don't ask any embarassing questions when you call for one.. check back issues of INFO or else maybe someone else here knows it.
daveh@cbmvax.commodore.com (Dave Haynie) (03/30/90)
In article <1917@awdprime.UUCP> robin@reed.UUCP (Robin D. Wilson/1000000) writes: >>In article <15003@snow-white.udel.EDU> BARRETT%FOREST.ECIL.IASTATE.EDU@cunyvm.cuny.edu (Marc Barrett) writes: >>>Unlike the Amiga 3000, however, all of these custom chips run at the full speed >>>of the microprocessor. >I thought that on the A1000/500/2000 line the custom chips ran 2x as fast as >the main processor. Essentially true. The Amiga chips run a memory cycle in two 7.16 MHz clock cycles, the 68000 in those systems runs a memory cycle in four 7.16 MHz clock cycles (for NTSC; the PAL machines run 7.09 MHz clocks). Since the A2500/30, the CPU subsystem speed was fully decoupled from the Video subsystem speed. When the CPU needs to access anything in the video subsystem, it synchronizes to video speeds. You can expect all future Amigas to work basically this way, regardless of the video and CPU clock speeds in any given machine. All Mac II's except the Mac IIci have a similar split. The video subsystem is on the NuBus, and the CPU must sync up and take wait states to communicate over the NuBus, which runs a minimum memory cycle of two 10MHz clocks (though most NuBus devices require at least three 10MHz clocks). Older Mac IIs (the Mac II and Mac IIx) use a 15.6 MHz CPU clock, twice that of the Mac 512/Mac Plus 7.8 MHz clock, probably because they use a number of the same gate arrays, all of which were designed to run at 7.8MHz. The Mac IIci runs a CPU clock of 25MHz, and slows down to talk to some of it's I/O chips. The Mac IIfx runs a 40MHz CPU and cache clock, and slows down for reads of main memory any other access of I/O, whether on the motherboard or NuBus. I haven't read enough on it to know just how much of a slowdown there is for various motherboard resources. That's no big surprise, though; DRAM isn't as fast as SRAM, but it's a heck of a lot faster than most things CPUs communicate with, like I/O chips and ROM. -- Dave Haynie Commodore-Amiga (Systems Engineering) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy Too much of everything is just enough
daveh@cbmvax.commodore.com (Dave Haynie) (03/30/90)
In article <133557@sun.Eng.Sun.COM> fiddler@concertina.Sun.COM (Steve Hix) writes: >In article <10363@cbmvax.commodore.com>, daveh@cbmvax.commodore.com (Dave Haynie) writes: >> In article <15003@snow-white.udel.EDU> BARRETT%FOREST.ECIL.IASTATE.EDU@cunyvm.cuny.edu (Marc Barrett) writes: >> > Once again, Commodore has let Apple slip ahead of them in the >> >area of technological innovation. Had Commodore released (or at >> The Apple card sounds pretty good, even at the $2000 price tag. But it's >> mainly good for Apples, of course, since this is the first card of it's kind >> (eg, general purpose QuickDraw engine) shown for an Apple machine. >SuperMac and Radius both are selling QuickDraw accelerators. RaterOps may >also be shipping one. Not all of them are general purpose, though. At least, when I read the write-up of the ARM based accelerator (Radius?), the implication was that it would only speed up the operations of software that specifically knew about the board. In other words, it didn't replace the QuickDraw routines on the Mac, it simply provided a speedup for programs (like the CAD package sold by the same company) that knew about that card. According to the BYTE article, RasterOps does make a similar unit. >They and AMD both ignore to above mentioned products...even though one of >them uses the AMD 29000. Marketing trolls are curious thingies. Well, that's to be expected. Even if one of these does the same basic thing as the Apple card, you better believe Apple will imply they did it first. IBM pretty much made this Standard Operating Procedure, and Apple has picked up on it really nicely. >At least two of the chips are, essentially, versions of the 6502 running some of >the I/O for the 68K. 6502s are in many standard cell libraries. I imagine the I/O tasks weren't all that big, and they were byte oriented. You can gain quite a bit by going to an I/O processor like that. Commodore's A2232 seven port serial card uses a 4502 for I/O, and manages to keep up with seven RS-232 lines at 19,200 baud each. The 4502/6502 pumps bytes as well as a 68030, and responds to interrupts much quicker. >Weird memory that does latched read/writes. Expensive. Only from Apple so far. >But a bit faster. Well, yeah, if you latch your writes, you can run one no-wait-state write. If the very next cycle is a write, you probably have to wait, but if it's a cache hit or a read, it's probably a win. >Just about the exact same speedup shown by the SuperMac accellerator. Maybe that one does a full QuickDraw interpreter. I haven't heard anything about it. -- Dave Haynie Commodore-Amiga (Systems Engineering) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy Too much of everything is just enough
stevem@sauron.Columbia.NCR.COM (Steve McClure) (03/30/90)
In article <22773@uflorida.cis.ufl.EDU> mfi@serc.cis.ufl.edu (Mark Interrante) writes: |In article <5479@sugar.hackercorp.com> karl@sugar.hackercorp.com (Karl Lehenbauer) writes: ||So Apple has discovered coprocessors, eh? Ironically, I recall a flame ||war a couple years back with a bunch of Mac people arguing that the Mac ||not having coprocessors was a *benefit*! | |Note the difference in the mac coprocessor scheme, the processor is a |30mhz risc chip that uses downloaded code. It is not a custom chip |that has hardwired graphics operations. | ||Further, as the vendors of various PC boards have been learning, wimpy ||coprocessors can be a substantial impediment to performance. Many vendors ||are abandoning their 8088, 80186, etc, processors in favor of much higher ||performance ones in, for example, networking and multiport serial boards. || ||Any Mac apologists care to compare the performance of a 6502 to the ||very-obsolete-and-never-particularly-attractive 8088? | |The 6502 is used as an embedded controller for operations that are |inherently low speed; ie. serial ports and keyboards. BTW the 6502 |controller runs at 10mhz. The 6502 is now a 30Mhz RISC chip??? I assume you mean Mhz not mhz. I think the previous argument still applies here. The 30Mhz RISC is on the 24bit graphics board correct? What difference does it make if the operations are in hardware or software, they are basically the same functions? Steve ----- Steve.McClure@Columbia.NCR.COM GO JACKETS!!! ICS '86 The above are my opinions, which NCR doesn't really care about anyway! CAUSER's Amiga BBS! | 803-796-3127 | 8pm-8am 8n1 | 300/1200/2400
daveh@cbmvax.commodore.com (Dave Haynie) (03/30/90)
>>Any Mac apologists care to compare the performance of a 6502 to the >>very-obsolete-and-never-particularly-attractive 8088? >The 6502 is used as an embedded controller for operations that are >inherently low speed; ie. serial ports and keyboards. BTW the 6502 >controller runs at 10mhz. For byte pumping, the 6502 is probably better than any 8088. Based on the memory cycle time, a 10MHz 6502 is roughly a 40MHz 8088. Certainly an 8088 would do some things better, but if all the IOP is doing involves "interrupt -> grab a byte -> rti", the 8088 may even be less efficient. And you don't find the 8088 as a standard cell in everyone's standard cell library.... -- Dave Haynie Commodore-Amiga (Systems Engineering) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy Too much of everything is just enough
lphillips@lpami.wimsey.bc.ca (Larry Phillips) (04/07/90)
In <MWM.90Apr7000929@raven.pa.dec.com>, mwm@raven.pa.dec.com (Mike (Real Amigas have keyboard garages) Meyer) writes: > >The flip side of this is that for the programmer, a preemptive system >has one less thing you have to worry about doing right (allowing other >tasks to run often enough that users on an 8MHz 68000 won't complain), >so you can spend more time making the rest of the program better. But >the number of Mac programmers vs. the number of Amiga programmers >probably makes up for that. There is one more thing about non-preemtive multitasking though, that is worth noting. You will suffer a speed penalty by having to call a system routine in a loop that might otherwise not need to call a routine. An example might be during ray-tracing, matrix calculations, and so on. -larry -- Entomology bugs me. +-----------------------------------------------------------------------+ | // Larry Phillips | | \X/ lphillips@lpami.wimsey.bc.ca -or- uunet!van-bc!lpami!lphillips | | COMPUSERVE: 76703,4322 -or- 76703.4322@compuserve.com | +-----------------------------------------------------------------------+
mwm@raven.pa.dec.com (Mike (Real Amigas have keyboard garages) Meyer) (04/07/90)
In article <620@lovelady.cs.utexas.edu> s320@cs.utexas.edu (Spring 90) writes: >A premptive OS and protected memory have nothing to do with each other. The >Amiga has a preemptive OS and no protected memory. A "preemptive" OS is >TOTALLY TRANSPARENT to the application programs: the OS gives each program >a "quantum" of time and the switches, with the application not even able >to find out about it. Get any good book on Operating Systems and "read >more about it" :-) Guess again!!!! That's right, you get a quantum of time for each program. But what happens if certain things are not protected, like the TIMER???? Probably the same thing that happens on _any_ computer without memory protection when some program does something stupid to the OS - you wind up rebooting the machine. [Example of program crashing a system deleted.] So, we need protected memory if we have preemtive multitasking. It is a good idea, anyway, so my program doesn't overwrite your program's space, either on purpose or accidentally... Actually, your reasoning applies to _any_ system, no matter what kind of tasking it does. After all, doing that means that only the broken program needs to die. Even on a single-tasking system, that means you don't have to restart the OS. Much as some Amiga users may hate to admit it, the difference between preemptive and non-preemptive multitasking isn't that great from the users point of view. Non-preemptive just means that there are more ways (just one?) for a program to be "multi-tasking unfriendly". Once you have such a program enter the twilight zone, you take the same action on either kind of system, said action depending on the rest of the system. The flip side of this is that for the programmer, a preemptive system has one less thing you have to worry about doing right (allowing other tasks to run often enough that users on an 8MHz 68000 won't complain), so you can spend more time making the rest of the program better. But the number of Mac programmers vs. the number of Amiga programmers probably makes up for that. <mike -- Here's a song about absolutely nothing. Mike Meyer It's not about me, not about anyone else, mwm@relay.pa.dec.com Not about love, not about being young. decwrl!mwm Not about anything else, either.
navas@cory.Berkeley.EDU (David C. Navas) (04/08/90)
In article <MWM.90Apr7000929@raven.pa.dec.com> mwm@raven.pa.dec.com (Mike (Real Amigas have keyboard garages) Meyer) writes: >Much as some Amiga users may hate to admit it, the difference between >preemptive and non-preemptive multitasking isn't that great from the >users point of view. Well, I take exception to that on a theoretical level at any rate. Now, it's true that WP programs may not need preemptive multitasking, but then WP programs don't need multitasking. Now, when do *I* use multitasking? WHen I'm doing downloading and/or playing music [on my computer, silly] while programming, etc. Under these circumstances, MultiFinder doesn't cut it for the obvious reasons. I think that it's not so much that multitasking programs don't need preemptive multitasking, but that the current manner of programming doesn't teach you to think in concurrent ways, and so preemptiveness isn't so obvious. > <mike David Navas navas@cory.berkeley.edu "Think you can, think you can't -- either way it's true." Henry Ford
karl@sugar.hackercorp.com (Karl Lehenbauer) (04/08/90)
In article <MWM.90Apr7000929@raven.pa.dec.com> mwm@raven.pa.dec.com (Mike (Real Amigas have keyboard garages) Meyer) writes: >Much as some Amiga users may hate to admit it, the difference between >preemptive and non-preemptive multitasking isn't that great from the >users point of view. Perhaps. Nonetheless, as an example, the Hackercorp Cybernetic Sequencer (approaching beta test within a few weeks) consists of separate cooperating tasks for MIDI playback, MIDI recording, standard MIDI file loading, standard MIDI file saving, the MIDI clock, the SMPTE clock, Amiga sample playing as a MIDI synth, and the user interface. Although the loader and saver could be nonpreemptive, and the clock programs really just set up interrupt handlers, the MIDI recorder, MIDI player and sample player could not be tasks under a non-preemptive or non-realtime exec -- their functionality would have to be encoded into an interrupt handler. Granted, these mainly affect the developer, as you said, but they do buy the user some things. What? Well, the ability to play a song while simultaneously loading other ones, not requiring the saver or recorder to be present in a playback-only (i.e. performance) environment (could be handled with overlays, but ick), the ability to use the recorder and player simultaneously in kinky ways, the eventual ability to (hopefully) do simultaneous animations, IFF picture displaying, fading, wiping, etc, i.e. multimedia and, as long as the priorities are set up correctly, the ability to run other programs simultaneously with these programs without their affecting the hard-realtime time-critical aspects of those activities. And the preemptive multitasking may not be apparent to the user, but I would argue that its presence on the Amiga (in conjunction of course with the blitter) substantially contributes to the Amiga's higher windowing performance than, say, the Mac, for a given CPU type and clock speed. In other words, having to constantly explicitly pass control to other tasks can substantially decrease performance, which does ultimately affect the user. -- -- uunet!sugar!karl "I hate quotations. Tell me what you know." -- Emerson -- Usenet access: (713) 438-5018
s320@cs.utexas.edu (Spring 90) (04/09/90)
In article <23800@pasteur.Berkeley.EDU> navas@cory.Berkeley.EDU.UUCP (David C. Navas) writes: >In article <MWM.90Apr7000929@raven.pa.dec.com> mwm@raven.pa.dec.com (Mike (Real Amigas have keyboard garages) Meyer) writes: >>Much as some Amiga users may hate to admit it, the difference between >>preemptive and non-preemptive multitasking isn't that great from the >>users point of view. >Well, I take exception to that on a theoretical level at any rate. Now, it's >true that WP programs may not need preemptive multitasking, but then WP >programs don't need multitasking. >Now, when do *I* use multitasking? >WHen I'm doing downloading and/or playing music [on my computer, silly] while >programming, etc. >Under these circumstances, MultiFinder doesn't cut it for the obvious reasons. Ah, guy, I just (yesterday) downloaded a BUNCH of stuff with my comm program in the background in multifinder....and Bard's Tale and some other games (Pipe Dream, also, more that I don't remember) play sound FLAWLESSLY when they are in the background. So MultiFinder obviously cuts it... Please, folks, don't make pronouncements like this unless you know what you are talking about...notice I didn't say anything about the Amiga, as I am NOT very familiar with it, having used one only a few times... -- Ted Woodward (s320@cs.utexas.edu) Someone shot the food...
dave@cs.arizona.edu (David P. Schaumann) (04/09/90)
In article <23800@pasteur.Berkeley.EDU>, navas@cory.Berkeley.EDU (David C. Navas) writes: ]In article <MWM.90Apr7000929@raven.pa.dec.com> mwm@raven.pa.dec.com (Mike (Real Amigas have keyboard garages) Meyer) writes: ]>Much as some Amiga users may hate to admit it, the difference between ]>preemptive and non-preemptive multitasking isn't that great from the ]>users point of view. ] ]Well, I take exception to that on a theoretical level at any rate. Now, it's ]true that WP programs may not need preemptive multitasking, but then WP ]programs don't need multitasking. ] ]Now, when do *I* use multitasking? ]WHen I'm doing downloading and/or playing music [on my computer, silly] while ]programming, etc. ] ]Under these circumstances, MultiFinder doesn't cut it for the obvious reasons. ] ]I think that it's not so much that multitasking programs don't need preemptive ]multitasking, but that the current manner of programming doesn't teach you to ]think in concurrent ways, and so preemptiveness isn't so obvious. ] ]> <mike ] ]David Navas navas@cory.berkeley.edu ]"Think you can, think you can't -- either way it's true." Henry Ford I'm supprised that's all you do while downloading. When I'm downloading a group of files (usally from a Fred Fish Disk), I start the download, then click to my shell window, format a new disk, and begin unarchiving the files as they arrive (all while the download is taking place). Also, I often look at the contents of text files while waiting for the next file to be completed. I'll never be happy on a system without true multitasking and a ramdisk or very fast hard drive again. Fortunately, this turns out not to be a problem ;) Dave Schaumann dave@cs.arizona.edu
navas@cory.Berkeley.EDU (David C. Navas) (04/09/90)
another Mac and Amiga discussion. No flames, some facts. Most of you may want to hit 'n' right now, though.... In article <626@lovelady.cs.utexas.edu> s320@cs.utexas.edu (Spring 90) writes: >Ah, guy,I just (yesterday) downloaded a BUNCH of stuff with my comm program in >the background in multifinder....and Bard's Tale and some other games (Pipe >Dream, also,more that I don't remember) play sound FLAWLESSLY when they are in >the background. So MultiFinder obviously cuts it... My examples were from frustrated Mac users. My experience is on a Mac plus, mileage may vary on more advanced machines. Actually mileage *will* vary on different machines. However, even in the regular finder on a Mac plus, I end up hitting the drag bar several times to move a window -- exactly because I'm used to the Amiga's *instant* response, rather than the diddly response I get from the Mac plus [an obviously outdated machine]. In fact, my *personal* experience is that diddly won't run on a 1 meg system set up, and I'm lucky to get Word up under multifinder with a 1+ meg system. Whereas I could play some simple music scores *and* program (from a harddrive, of course) without worrying on my Amiga -- but we won't get into the memory allocation scheme differences of the two systems.... >Please,folks, don't make pronouncements like this unless you know what you are >talking about...notice I didn't say anything about the Amiga, as I am NOT very >familiar with it, having used one only a few times... I actually *do* know something about Macs, having written much on one, and having played around extensively with it via A-Max. I *don't* know the ins and outs of getting multifinder to work (and confess that the downloading problem was reported to me rather a long time ago), and frankly I don't think it's worth my time figuring it out. I do know that there are substantial differences between preemptive and non- preemptive scheduling, otherwise no one would go to the bother of writing preemptive kernels. Never mind that Apple has been quoted in Byte as saying that their operating system is not sufficient to take advantage of the MacIIfx architecture, and conjecture that 7.0 may solve some of the deficiencies [sp?]. Frankly, this ought to be continued in e-mail, I'd love to hear stories on how to get multifinder working (and stable), at the very least to help my friends out. At any rate, both systems have their weaknesses and kludges. We have HAM, you have multifinder. Let's all grow up and admit our problems and grow and learn from each other -- isn't that what CS is supposed to be all about? I'd hate to think that misguidedness is dealt with by flames. >-- >Ted Woodward (s320@cs.utexas.edu) > >Someone shot the food... No flame intended, no flames responded to. :) *diddly is (C) advertising agency of you know who :) :) :) David Navas navas@cory.berkeley.edu "Think you can, think you can't -- either way it's true." Henry Ford
mgenius@wam.umd.edu (Sandro M. Fouche) (04/09/90)
In article <212@caslon.cs.arizona.edu> dave@cs.arizona.edu (David P. Schaumann) writes: >In article <23800@pasteur.Berkeley.EDU>, navas@cory.Berkeley.EDU (David C. Navas) writes: >]In article <MWM.90Apr7000929@raven.pa.dec.com> mwm@raven.pa.dec.com (Mike (Real Amigas have keyboard garages) Meyer) writes: >]>Much as some Amiga users may hate to admit it, the difference between >]>preemptive and non-preemptive multitasking isn't that great from the >]>users point of view. >] >]Well, I take exception to that on a theoretical level at any rate. Now, it's >]true that WP programs may not need preemptive multitasking, but then WP >]programs don't need multitasking. >] >]Now, when do *I* use multitasking? >]WHen I'm doing downloading and/or playing music [on my computer, silly] while >]programming, etc. >] >]Under these circumstances, MultiFinder doesn't cut it for the obvious reasons. >] >]I think that it's not so much that multitasking programs don't need preemptive >]multitasking, but that the current manner of programming doesn't teach you to >]think in concurrent ways, and so preemptiveness isn't so obvious. >] >]> <mike >] >]David Navas navas@cory.berkeley.edu >]"Think you can, think you can't -- either way it's true." Henry Ford > >I'm supprised that's all you do while downloading. When I'm downloading >a group of files (usally from a Fred Fish Disk), I start the download, then >click to my shell window, format a new disk, and begin unarchiving the >files as they arrive (all while the download is taking place). Also, I often >look at the contents of text files while waiting for the next file to be >completed. > >I'll never be happy on a system without true multitasking and a ramdisk or >very fast hard drive again. Fortunately, this turns out not to be a >problem ;) > >Dave Schaumann >dave@cs.arizona.edu The fact is that I can do all of these things while downloading on the mac. I'd rather be making out with my girlfriend, then be so tied to my computer that I'm formatting disks and playing video games while downloading. -Sandro Fouche'
gilmore@vms.macc.wisc.edu (Neil Gilmore) (04/09/90)
In article <212@caslon.cs.arizona.edu>, dave@cs.arizona.edu (David P. Schaumann) writes... (whole lots of multi-theory deleted) >I'm supprised that's all you do while downloading. When I'm downloading >a group of files (usally from a Fred Fish Disk), I start the download, then >click to my shell window, format a new disk, and begin unarchiving the >files as they arrive (all while the download is taking place). Also, I often >look at the contents of text files while waiting for the next file to be >completed. I do all this as well as run a shell just for programming or a game while I wait. In fact, I never actually run my terminal program (vt100 if any one cares, I have to use kermit, you see), I just execute a script which sets up everything for me (including loading another script into ram: so that I can zoo to disk from ram:) and shuts it all down when I'm done. Now if I just got arexx, and made a macro which would do my downloads in the middle of the night.... +-----------------------------------------------------------------------+ | Kitakaze Tatsu Raito Neil Gilmore internet:gilmore@macc.wisc.edu | | Jararvellir, MACC, UW-Madison bitnet: gilmore@wiscmac3 | | Middle Kingdom Madison, Wi | +-----------------------------------------------------------------------+
mithomas@bsu-cs.bsu.edu (Michael Thomas Niehaus) (04/09/90)
David Navas writes: > Now, when do *I* use multitasking? > WHen I'm doing downloading and/or playing music [on my computer, silly] while > programming, etc. > > Under these circumstances, MultiFinder doesn't cut it for the obvious reasons. What are these obvious reasons? I download in the background all the time, while checking my hard drive for viruses and talking to our VAX host. (I have two serial lines.) Heck, even MS-DOS can do things like this (with Windows or various other TSRs). > I think that it's not so much that multitasking programs don't need preemptive > multitasking, but that the current manner of programming doesn't teach you to > think in concurrent ways, and so preemptiveness isn't so obvious. Huh? The only time that programming needs to make you think in concurrent ways is when you are working on a parallel processing machine. What we are talking about here is the possibility of two distinctly separate programs running simultaneously. Since there is no interaction between the two, multitasking is not taken into account when writing the programs (assuming of course that the operating system has some multitasking abilities). You are referring to a concept that really isn't related (although it is dependent on the availability of multitasking). Separate "threads" are the concept that most programmers do not consider when programming. They are not taught to think concurrently in this case. You can have preemptive multitasking without such threads being feasible (take VMS for example -- takes to long to create the thread, unless you "pre-create" it and use some type of IPC to send it commands). Unix has the ability to do this partially (with forks), but the only operating system that I have seen that does this well is OS/2. (PageMaker is the first application that I have seen that really takes advantage of this by reformatting in the background while the user continues to edit.) -Michael -- Michael Niehaus UUCP: <backbones>!{iuvax,pur-ee}!bsu-cs!mithomas Apple Student Rep ARPA: mithomas@bsu-cs.bsu.edu Ball State University AppleLink: ST0374 (from UUCP: st0374@applelink.apple.com)
papa@pollux.usc.edu (Marco Papa) (04/09/90)
A message from the NET POLICE: The message I'm responding is a good example of the WRONG way to respond to a message: you quote EVERYTHING (40-50 lines) and your message is 4 lines long. PLEASEEEEE! Keep only what is needed from the quoted message. Everybody pays for the extra, unnecessary verbiage and it takes FOREVER to get to YOUR response. I myself, if I don't see part of the response on the first screen, will press 'n' automatically. It takes some extra effort to clean up the reply, but it will undoubtedly widen your audience. Thank you for listening. -- Marco -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= "Xerox sues somebody for copying?" -- David Letterman -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
gilham@csl.sri.com (Fred Gilham) (04/10/90)
Michael Niehaus writes:
==============================
Unix has the ability to do this partially (with forks), but the only operating
system that I have seen that does this well is OS/2. (PageMaker is the first
application that I have seen that really takes advantage of this by reformatting
in the background while the user continues to edit.)
==============================
I don't know anything about OS/2 (half an operating system???) but I
remember reading a year or so back that a well known word processing
group bought amigas to help program the OS/2 version of their word
processor, because they found the amiga implementation of tasks etc.
easier to use.
My own experience with the amiga is that it taught me to think in
terms of concurrent processes. It is very easy to use them on the
amiga.
--
Fred Gilham gilham@csl.sri.com
If it can be shown that the machinery has come into the world as a
curse, there is no reason whatever for for our respecting it because
it is a marvellous and practical and productive curse.
-G. K. ChestertonFelineGrace@cup.portal.com (Dana B Bourgeois) (04/10/90)
Ted Woodward says his Mac can download in the background and do other stuff also. I believe him. But can a Guru on Mac/Amiga explain why? How does the Mac do this and how is the technique the same/different than the Amiga? I for one would like to know what the Mac is doing and what effect does it have on the user and how does the multifinder technique appear to the user as opposed to the Amiga. My guess is that it takes well behaved programs to work together under multifinder. More work for the programmer but when running, maybe they appear as multi-tasking as well-behaved Amiga programs. Email replies to me if you like. I will summarize responses to save bandwidth. Dana Bourgeois @ Cup.Portal.Com See? Only 1 line in my .sig! Well, maybe two. Or is it three...
Classic_-_Concepts@cup.portal.com (04/12/90)
All stock Amigas can multitask. Right out of the box. Even the 256K varieties (early 1000s). To use MultiFinder on the Macintosh, you require at least a Mac Plus, at least version 5.0 of the operating system and at least 1 meg of RAM. BUT the instructions explicity warn that you will need at least 2 meg if you plan to run more than one large application program at a time. Amigas multitask from the moment you power up Workbench. Multifinder must be specifically configured for startup. This is not a significant difference but more a detail of convenience. Multifinder doesn't work with all Mac programs. Running applications can fragment memory on the Amiga or the Mac, but **Macintosh applications require contiguous blocks of memory**. To clear up a fragmentation problem on the Mac, close down everything. Rerun everything. To be fair, I have also found this to sometimes be necessary when running the Amiga version of Deluxe Paint. I always give it a pretty hard workout and at some point find I can't even load small brushes. Rerunning the software cleans up the problem. On the Mac, to set aside memory for running an application under Multifinder, you can trust the default or tweak it from a utility to increase or decrease the size. Personally, I like the operating system to worry about this for me rather than juggling memory menus. That's enough for one posting ... LadyHawke
ddev@wam.umd.edu (Don DeVoe) (04/12/90)
In article <28812@cup.portal.com> Classic_-_Concepts@cup.portal.com writes: >To use MultiFinder on the Macintosh, you require at least a Mac Plus, at >least version 5.0 of the operating system and at least 1 meg of RAM. BUT >the instructions explicity warn that you will need at least 2 meg if you >plan to run more than one large application program at a time. First, Apple hands out OS revisions for free, so who cares if multifinder requires 5.0? Second, of course you need at least 2 megs to run more than one large program! Heck, I've got plenty of applications which require 2-3 megs all to themselves. >Amigas multitask from the moment you power up Workbench. Multifinder must >be specifically configured for startup. This is not a significant difference >but more a detail of convenience. Specifically configured? What this means is you have a choice between finder or multifinder. Changing between the two takes only 2 mouse clicks... >Multifinder doesn't work with all Mac programs. And neither do all programs multitask on the Amiga. If properly written, there is no problem. As I said, if a program doesn't work, just 2 clicks and you're back in finder. >On the Mac, to set aside memory for running an application under Multifinder, >you can trust the default or tweak it from a utility to increase or decrease >the size. Personally, I like the operating system to worry about this for me >rather than juggling memory menus. Gee, and I thought that one of the biggest complaints most Amiga owners have about the Mac is that the OS worries about doing too much for the user, thus taking away the user's control over the system... -- Don DeVoe ddev@epsl.umd.edu