wayneck@tekig5.TEK.COM (Wayne Knapp) (05/19/88)
Okay people, if you don't program anything other than an Amiga you don't have
much room to talk. When I say a ST is faster hands down, I mean it. I have
both machines sitting side by side, if anyone knows I should. How many of
you have written code for IBM pc's, Atari's (8 and 16 bit), Apples and Amigas.
I'm not a Mac expert so I back down of some of the Mac things. (Except that
the Mac II graphics are slow, that is just untrue. The Mac II is the most
responsive windowing system I've ever used, even better than the Suns around
here. Like it or not Apple did a good job. Not cheap though)
I'm not trying to put down the Amiga. The Amiga is a good computer, I even
own one! What I saying is that a lot of Amiga code seems to depend on the
extra Amiga hardware to get the job done even at the sake of doing a poorer
job than what can be done in plain software alone.
Some great Amiga programs have been written, some code could be better. I
have some experience in programming as I've been doing it for 12 years and
earn my living at it.
If Amiga programmers would put thier effort into writing good sound code, code
that is well thought out instead of code that uses all kinds of gee whiz
hardware tricks, the Amiga would really shine. Other than general hardware
like bitmaps, sampled sound generator and such, I believe the code should be
blind to special hardware. Let the os use the hardware but don't tie the
general code down with stuff like copper lists. Use the hardware to make the
machine look more powerful to a program. Let the program think there are 12
bit planes, don't think HAM. See what I mean.
The Amiga can be a great machine, it just needs a little polishing.
Wayne Knapp
In article <8805170742.AA28361@cory.Berkeley.EDU>, dillon@CORY.BERKELEY.EDU (Matt Dillon) writes:
>
> as comparing the Amiga to a 68000 Mac, I beg to differ. Having run tests and
> used both I see no such discrepancy. In fact, taking the Amiga's worst
> problem (floppy disk speed) the Mac+ writes files to floppies at about the
> same speed. And of course, one need not mention graphically oriented IO.
Test are one thing, finished programs another. I'm amazed by just how
good a lot of Mac software is. Amiga software can also be that good, just
a lot of it isn't yet.
>
> >the Mac software. The Mac os is as far above the Amiga in richness and
> >robustness as the Amiga os is above the Atari ST os. Of coarse I'm not
> >talking about multi-taking here but the Mac also has good solutions for
> >multi-taking. Even the Atari ST can do reasonable mutli-tasking. Multi-
>
> Again, I say B.S.
Look at the programs, Hyper'Card, MacPaint, Labview, on and on. Many great
programs say you are wrong. I think the true power of the Mac os is in the
handling of it program resouces. Text, graphics, controls can all be changed
without recompiling by using a simple program call Res-Edit. Also icons and
every thing are in the same file. The whole system is neat and well thought
out. That is what has help the Mac. Of coarse these things could all be
done on the Amiga. Maybe someone is. If so more power to them!
> exactly take MY breath away. The only think impressive is the CPU and
> CPU speed... what I see on a Mac II screen is very sluggish compared to its
> potential.
I think CPU speed has little to do with it. The software is sound so faster
CPU just make the good software run better. Also when things are updated
faster than you can see how can you call it sluggish. Change floppies on an
Amiga, that's sluggish.
> >Don't depend of hardware to get the job done. Soon the Amiga may be replaced
> >with a machine that has twice the power (without specail hardware) for 1/2 the
> >price. (Maybe even in a couple years) Good code will port easy and soar,
> >hardware dependent code will port hard and often just fade away. Use your
> >heads program great code and make the Amiga a winner!
>
> I disagree. That is, I agree that it is well to hide special
> hardware but I disagree as to your definition of 'special hardware'. What
> is a CRT controller? Do you want the processor to stuff the RGB out itself?
> What is a Disk controller? So now we extend it a little and have Audio
> DMA, a more complex video controller, another processor, etc... Frankly,
> that is the future... first start with more specialized controllers and
> then move on and add more general parallel processing capabilities.
>
Come on, do you think I that stupid. Bitmaps are one thing, halfbrite and
HAM another.
> You might be able to reduce their advantage to a numerical factor,
> but don't be deceived. In many cases a factor of 2 can make the difference
> between perceiving it as slow and perceiving it as fast. What's the difference
> between 20 mph and 40 mph? 30 and 60? Get the picture....
>
Yes, you are finally using your head! Many times I've seen improvements on the
order of 5 to 50 times by improving the code. Try to do that cheaply with
hardware. In fact that is why the Amiga isn't amazing compared to the Mac or
ST. Because not enought good clean thought has been put into the code.
Thinking like: I'll just use the blitter instead of adaption my code to this
better data structure, is going to hurt the Amiga not help it. If the software
is great the great hardware can only help it. The real battle is in software
not hardware. The Amiga has many good features. So let the software access
these, don't make the software depend on them to run.
> > Huh, I own both an Amiga 1000 and Atari ST. I'm porting Amiga code to
> >the Atari right now. I often run then side by side. I know that the are
> >many factors in speed but seldom does the Amiga beat the Atari. I think a
> >lot of the Amiga speed loss is due to memory use. Maybe if I put a couple
> >megs of memory on the Amiga it would inprove a lot but as it is now a 512k
> >Atari kills a 512k Amiga, no doult about it. Also I never seen disk drives
> >as slow as the Amiga's. A the Amiga graphics and sound often beat the ST's
> >thought, but not it's speed. The Amiga should be faster but it isn't.
> >It really fustrates me! So come on lets write better code for the Amiga!
>
> I can laugh at that .. excuse me a moment ... ok, I'm back. Frankly,
> you are dead wrong here. In terms of raw cpu speed the atari is just a
> little faster, and that is assuming you are NOT in monochrome mode. I find
> the well integrated enviroment the Amiga provides is just as fast if not
> faster than all other machines of equivalent stature (read: Atari/Mac. IBM
> is a lower form of life as far as I'm concerned). Perhaps *YOU* should
> invest a little money into a few solutions, of which there are many, instead
> of bitching about your problems. At least when I bitch I create or find my own
> solutions. I personally use REZ and FACC and since then have never
> had to wait for my floppies. Oh, and by the way, I do not currently own
> a hard drive.
>
Laugh all you want, but it is true. The Amiga is often much slower than my
ST. It shouldn't be, but it is. They sit side by side so I know. There is
a lot more to it than just CPU speed. I think that some of the differnce is
that there must be some poor/slow code in the Amiga os or something.
DO YOU use the machines side by side. If not you have no room to talk. I'm
not lying. I want the Amiga to do well, I like it a lot, but I'm not 100%
happy with it.
> For less CPU bound programs, the hardware support and low OS overhead
> easily outweigh such minor processor speed differences.
>
Run the machines side by side for over a year, write programs on both, you
may be shocked at what you'll learn.
> I have very STRONG opinions as to MY FAVORITE machine... so much so that
> I find my particular machine a thousand fold better choice then other
> machines at the time I purchased it (and it's still holding its place after a
> couple of years). We all know which machine *that* is ....
>
It is good that you like the Amiga. So you know the Amiga's strong points.
Open you eyes and see where it is weak. Look at other computers that do
better then improve the Amiga. The Amiga has more potential than any computer
I've ever seen, so don't you think it is time to tap it?
Wayne Knapp dillon@CORY.BERKELEY.EDU (Matt Dillon) (05/19/88)
>Okay people, if you don't program anything other than an Amiga you don't have >much room to talk. When I say a ST is faster hands down, I mean it. I have I have to hand it to you, you are really trying *hard* to 'prove' your point. >both machines sitting side by side, if anyone knows I should. How many of >you have written code for IBM pc's, Atari's (8 and 16 bit), Apples and Amigas. I raise my hand! Everything but the Atari. But of course, that doesn't mean my opinion of it isn't valid.... Hell, when I decided to buy a new computer I examined both the Amiga and Atari quite closely. Even after buying my Amiga, having an interest in ALL micros in existance, I have attemped to give myself an understanding of the Atari's system software (OS and support). I see garbage. I've talked at length to friends and I hear garbage. I read comp.sys.atari.st and I hear garbage and complaints. I've read every friggin holy war argument between Amiga people and Atari people, I've look at Atari and note that they resort to blatent lies and deceptions in an attempt to bolster their line, they introduced their line early to get a jump on the Amiga ... too early, and still haven't fixed the problems. But most notably, I hear arguments that relate to the Amiga of three years ago, not the Amiga today. And what do I say? Get A life! And while I have no intention of refuting everything you say (recognizing that your message is essentially an opinion, as is all our messages). >own one! What I saying is that a lot of Amiga code seems to depend on the >extra Amiga hardware to get the job done even at the sake of doing a poorer >job than what can be done in plain software alone. Perhaps many of the screen hacks do, demos and such, but I have yet to see a professional PD or commercial program that relies on the underlying hardware to the extend you are saying. Rely on the existance of the intuition, graphics, and other libraries yes, but the cases where a program requests use of, say, HAM or EHB mode, can hardly be cause to declare an entire program hardware dependant. After all, we are talking about a very small module of the program. The only *real* definition of being 'hardware dependant' in the world is essentially 'it would be difficult to port the program to another computer because of this'... ALL programs are hardware dependant in one sense or another. I see no such difficulty in the majority (99.9999...%) of non-demo/non-screenhack Amiga programs. The Amiga's library structure effectively hides most, if not all hardware dependancies for most functions a professional or commercial quality program might want to use. You can hardly call a program dependant on the blitter if it makes a call to ScrollRaster(). >Some great Amiga programs have been written, some code could be better. I >have some experience in programming as I've been doing it for 12 years and >earn my living at it. As is true for all computers. I've seen bad code on everything from an HP3000 to UNIX to ANY microcomputer you can name. You have no point here except perhaps to hint that you think the *majority* of Amiga programs are not written well... A statement which is quite incorrect. Good for you. I have some experience in programming also, having been doing it seriously since 7th grade, and I now earn my living at it also. So what? If experience is all that matters to you, then there are more 'experienced' people on this net then either of us, with equivalent opinions even. :If Amiga programmers would put thier effort into writing good sound code, code :that is well thought out instead of code that uses all kinds of gee whiz :hardware tricks, the Amiga would really shine. Other than general hardware :like bitmaps, sampled sound generator and such, I believe the code should be :blind to special hardware. Let the os use the hardware but don't tie the :general code down with stuff like copper lists. Use the hardware to make the :machine look more powerful to a program. Let the program think there are 12 :bit planes, don't think HAM. See what I mean. : :The Amiga can be a great machine, it just needs a little polishing. This is a standard misconception, one I've heard from the very beginning (incidently, brought up in comp.sys.atari.st). 'Other' people have got it in their minds that because the Amiga has a blitter, because the Amiga can handle lots of things in hardware that are done in software on other machines, that this means programs written for the Amiga are DEPENDANT on that hardware. Read my lips: B U L L S H I T. Remember that. You are insulting a lot of people by such a blatent generalization. Not only that, but your premise is based on a rumor brought about by hard feelings. You are taking a couple specialized display oriented programs (e.g. DPaint, other video intensive stuff) and not even looking at all the other stuff that's out there (compilers, Tex, XCad, etc....). And if you aren't doing that, then you ought to explain your reasoning. >> as comparing the Amiga to a 68000 Mac, I beg to differ. Having run tests and >> used both I see no such discrepancy. In fact, taking the Amiga's worst >> problem (floppy disk speed) the Mac+ writes files to floppies at about the >> same speed. And of course, one need not mention graphically oriented IO. > > Test are one thing, finished programs another. I'm amazed by just how >good a lot of Mac software is. Amiga software can also be that good, just >a lot of it isn't yet. I'm amazed that you are amazed. I'm not amazed, and I've got a friend of mine who's a Mac nut with a Mac II even! >Look at the programs, Hyper'Card, MacPaint, Labview, on and on. Many great >programs say you are wrong. I think the true power of the Mac os is in the >handling of it program resouces. Text, graphics, controls can all be changed I agree. The best thing going for the Mac is its resource system. But I don't see how these 'great' programs say I'm wrong. You think I don't recognize the existance of great software? I recognize it as much as I recognize the existance of bad software. >without recompiling by using a simple program call Res-Edit. Also icons and >every thing are in the same file. The whole system is neat and well thought >out. That is what has help the Mac. Of coarse these things could all be >done on the Amiga. Maybe someone is. If so more power to them! In fact, I am... resources that is. I guess that means I get energized. >I think CPU speed has little to do with it. The software is sound so faster >CPU just make the good software run better. Also when things are updated >faster than you can see how can you call it sluggish. Change floppies on an >Amiga, that's sluggish. Fool around with an Amiga running the new fast filesystem on a harddisk and you'll find the 500K+/sec transfer rate quite worthwhile. Bad example? Well, let me put it this way. I don't own a harddisk, I find changing floppies on an Amiga a sluggish but it really doesn't effect how I work on my Amiga. I find the speed of floppies to be plenty fast enough with the proper utilities installed (FACC and REZ). By the way, where did you learn english? As far is your statement 'updated faster than you can see'... I agree. WHEN something is updated faster than I can see I don't call it sluggish. What does that have to do with the argument? Are you relating the concept to the Atari in general? If so, I disagree. >Come on, do you think I that stupid. Bitmaps are one thing, halfbrite and >HAM another. Ah, I see, so you are saying that because EHB and HAM don't exist anywhere else we should not use them? By that thinking you want us to reduce our programs to standards attainable on most machines of the same class..... And I suppose the programs you write for your Atari don't assume the dimension and color capabilities of your machine? And how many programs do you know of that actually use HAM and EHB? Almost none use EHB and a couple use HAM but also allow 'normal' resolutions. In fact, due to its nature, HAM is used almost solely for pictorial output (static pictures or animations), which reduces to simple subroutines in a program ... quite easy to change over to other more standard resolutions, which is all you care about since binaries on one machine aren't expected to run verbatim on the other machine. In fact, of all the programs I've seen only one is specifically dependant on HAM as being an integral part of the program. And you are using HAM and EHB as an example to prove your point? Bad example; think up another one. >order of 5 to 50 times by improving the code. Try to do that cheaply with >hardware. In fact that is why the Amiga isn't amazing compared to the Mac or >ST. Because not enought good clean thought has been put into the code. Maybe your code.... all I can say here is that it's true in some cases, but a damn lie in most cases. This statement is general and applies to a few programs on every computer that exists. >Thinking like: I'll just use the blitter instead of adaption my code to this >better data structure, is going to hurt the Amiga not help it. If the software >is great the great hardware can only help it. The real battle is in software >not hardware. The Amiga has many good features. So let the software access >these, don't make the software depend on them to run. Uh huh. Do you want to know the last time I touched the blitter hardware directly? About two years ago. You are right in that the real battle is in software. Hardware only increases the possibilities in one's ultimate goal. You are wrong in applying a general argument 'These people write better code than those people' to the Amiga community as a whole. HOW CAN YOU DO THAT? It isn't possible to apply such an argument to any sufficiently large group. This one misconception appears to be the core of your entire argument. >Laugh all you want, but it is true. The Amiga is often much slower than my >ST. It shouldn't be, but it is. They sit side by side so I know. There is >a lot more to it than just CPU speed. I think that some of the differnce is >that there must be some poor/slow code in the Amiga os or something. Well then. You say it's slower, I say it's faster. The entire comp.sys.atari.st group will probably back you, and the entire comp.sys.amiga group will probably back me. The two of us have different experiences, thus two opinions. end of Discussion. >DO YOU use the machines side by side. If not you have no room to talk. I'm >not lying. I want the Amiga to do well, I like it a lot, but I'm not 100% >happy with it. No I do not. I *do* have room to talk. Just because I don't have an Atari on my desk doesn't mean I haven't fiddled with one or three. I'm not 100% happy with my Amiga either, but more happy with it than any other micro. Obviously I can't say I have more experience with an Atari than you do, that would be utterly wrong. But because I do not does not mean you are automatically right. >> For less CPU bound programs, the hardware support and low OS overhead >> easily outweigh such minor processor speed differences. >> >Run the machines side by side for over a year, write programs on both, you >may be shocked at what you'll learn. > You keep on telling me I should buy an Atari and will find out for myself that you are right. Yet, I keep up on comp.sys.atari.st, and after comparing discussions between it and comp.sys.amiga for the past 3 years, I find no compulsion whatsoever to waste my money. I look at the programming models for the Atari and Amiga, and the Atari looks like a !@#$ IBM-PC, with the graphics support software slapped on as an afterthought, no multitasking in any true sense of the word, and little support. You expect ME to give up multitasking? That's a laugh. >> I have very STRONG opinions as to MY FAVORITE machine... so much so that >It is good that you like the Amiga. So you know the Amiga's strong points. >Open you eyes and see where it is weak. Look at other computers that do >better then improve the Amiga. The Amiga has more potential than any computer >I've ever seen, so don't you think it is time to tap it? > > Wayne Knapp Open my eyes and see where it is weak. Gosh, I guess you don't know much about me. My first year on comp.sys.amiga was spent essentially flaming C-A on dozens of topics, all related to problems I saw with my Amiga. Guess what? Most of those problems are fixed. Now you open your eyes. I'm sure comp.sys.atari.st would be thrilled if you posted there... why don't you move your comments to that group. -Matt
doug-merritt@cup.portal.com (05/20/88)
Wayne, it is very difficult to understand quite what you mean,
or in some cases why you think so.
Your overall point seems to be that Amiga folks should improve various
things about software. I think that's true on any computer, and there
certainly are lots of specific examples on the Amiga. Likewise, many
of us here are doing exactly that anyway (e.g. Matt's Resource project),
so what's the point?
As to your specific gripes about speed, exactly what are you referring
to that is slower on the Amiga than on the ST? And if you've been programming
both for the last year, how come no comments about how you've written
faster code on the Amiga than what you otherwise see? Since you're not
referring to anything specific, it's hard to know what you have in mind.
And as far as hardware goes, 12 bit planes would be nice, but is just too
expensive. We'd *all* like 12 bitplanes if we could afford them. But
currently, that's out of the question. The Amiga has more plain-jane
bitplanes than the ST, so again the comparison is confusing. The
HAM and extra-halfbrite modes are yet more icing on the cake...yes,
they are more inconvenient than plain bitmaps. But the *only* alternative
is to leave them out altogether; how would that help?
And as far as special purpose hardware like the blitter is concerned,
why complain that it exists? People in the ST world have been waiting
for one for ages; I've seen their comments in the past. If you don't
think blitters are a good idea, even ST fans would disagree!
If you're going to keep arguing these various points, I would like to
see you get more specific about why you have those particular opinions.
It's hard to deal with generalities.
Doug
---
Doug Merritt ucbvax!sun.com!cup.portal.com!doug-merritt
or ucbvax!eris!doug (doug@eris.berkeley.edu)
or ucbvax!unisoft!certes!dougpeter@sugar.UUCP (Peter da Silva) (05/20/88)
In article <2777@tekig5.TEK.COM>, wayneck@tekig5.TEK.COM (Wayne Knapp) writes: > Okay people, if you don't program anything other than an Amiga you don't have > much room to talk. When I say a ST is faster hands down, I mean it. I have > both machines sitting side by side, if anyone knows I should. How many of > you have written code for IBM pc's, Atari's (8 and 16 bit), Apples and > Amigas. (sticks hands up in the air. waves) "Me! Me!" Also: 8-bit CP/M machines, more UNIX and UNIX lookalike boxes than you can shake a stick at, 1802-based boxes, ISIS-based 8080 boxes, Trash-80s, and lots more. Before the Amiga I had an Atari 800, an IBM-PC clone, and an Atari ST on this desk. When I say the Amiga is faster I mean it. > I'm not a Mac expert so I back down of some of the Mac things. (Except that > the Mac II graphics are slow, that is just untrue. The Mac II is the most > responsive windowing system I've ever used, even better than the Suns around > here. Like it or not Apple did a good job. Not cheap though) Suns aren't fast when it comes to windowing. I haven't owned or programmed a Mac (when I saw what you had to do to program it I nearly puked), but I've used it pretty much. Even on a Mac-II I can click on a window to move it and have the mouse in a new location before it's deigned to draw that cute little moving dotted line. And you know what? By the time it's done that it's forgotten where I clicked it and it doesn't *bother* to draw that cute little box. It can take up to half a second before I can move the mouse in a busy screen. This situation is worse on either the Atari or Mac-I. Microsoft windows on an 8088 does a better job. > If Amiga programmers would put thier effort into writing good sound code, code > that is well thought out instead of code that uses all kinds of gee whiz > hardware tricks, the Amiga would really shine. Other than general hardware > like bitmaps, sampled sound generator and such, I believe the code should be > blind to special hardware. Let the os use the hardware but don't tie the > general code down with stuff like copper lists. Use the hardware to make the > machine look more powerful to a program. Let the program think there are 12 > bit planes, don't think HAM. See what I mean. Well, buster, most Amiga programs do a pretty good job of hardware independence at the source level. The ones that play with the copper list are usually clever screen hacks. In fact a major complaint I hear is that programs *aren't* making enough use of the special hardware. > The Amiga can be a great machine, it just needs a little polishing. The Amiga is a great machine. It would be a better one with a little polishing. I'm shining with all my might, personally. But there is no way I'm going to hack all day trying to get a tight assembly loop 1 instruction shorter. If I wanted to code in assembly language I'd have bought an IBM-PC. I will make no bargains with terrorist hardware. -- -- Peter da Silva `-_-' ...!hoptoad!academ!uhnix1!sugar!peter -- "Have you hugged your U wolf today?" ...!bellcore!tness1!sugar!peter -- Disclaimer: These may be the official opinions of Hackercorp.
jason@lakesys.UUCP (Jason) (05/20/88)
It seems sort of strange that you comment negatively about the original poster's generalizations about the Amiga programming community, and then go on to assume that everyone in comp.sys.atari.st would just LOVE to hear flames about the Amiga. Jason (I do own an ST, and I don't own an Amiga, nor do I particularly like the Amiga [based on admittedly small amount of experience with it] - however, I understand that people are generally biased one way or another, in this case leaning towards one machine or the other. Seeing as my postscript is now longer than my post [to make a truly terrible pun], I'll stop this...)
ali@polya.STANFORD.EDU (Ali T. Ozer) (05/21/88)
In article <2777@tekig5.TEK.COM>, wayneck@tekig5.TEK.COM (Wayne Knapp) writes: > I'm not a Mac expert so I back down of some of the Mac things. (Except that > the Mac II graphics are slow, that is just untrue. The Mac II is the most > responsive windowing system I've ever used ... ) The Mac windowing system is actually pretty slow. Consider the lack of "smart refresh" windows, for instance. While programming a Mac Plus with HD I normally have 8-10 windows open in the Finder. (This seems strange to some Mac user friends of mine, who occasionally look over my shoulder and make remarks such as "whoa! you don't know how to close windows?".) Anyway, between the time when you quit an application and you get back to the Finder is usually a good 15 seconds --- time just spent refreshing those windows. You see the empty Finder windows get drawn, one after another, and then, after a short delay, the windows start getting filled, one by one. Every window, even one with a single pixel showing, has respond to a refresh event. Running on a one-meg machine (no MultiFinder), and going back and forth often between applications which do not have "Transfer" commands in their menus can be a nightmare... Running MF of course solves the problem of having to quit/restart the Finder everytime you run an application, but then you start getting a lot more windows on your screen, and you get into the situation where everytime you close a largish window, your system locks up and has to refresh all the windows under it (Finder windows, other application windows, etc). The Mac II solves the problem by throwing oodles of CPU cycles at the windowing system. On a B&W Mac II, this makes the windowing system look fast/decent enough to use. So people then say "oh look the Mac II windowing system is better than the Amigas!". This just goes to show that all you need is $4000 worth of hardware to get a windowing system faster than that of a $600 Amiga. But, of course, who cares about the Mac! I personally bagged my Mac programming position last week. 8-) Ali Ozer, ali@polya.stanford.edu