Feedback

The readers speak out!

ImageFX

Another fine issue! I did want to drop you a line to update you on a few points though.

Nova Design, Inc. was at the Amiga Fest as well. Although we didn't premiere any new products there we did show ImageFX 2.6. Our presenter, Corinna Cohn, gave several presentations and gave away loads of the cool new ImageFX T-Shirts.

Also, in Bill Graham's article on stereoscopic image making, he stated to the effect that all image processors other than 'ADPro' throw away half of an image on deinterlacing. Not so! Our ImageFX puts the other half in its own buffer for easy access. Also, ImageFX has been used to produce 3D images and animations for years in every format from Anaglyph (red/blue 3D) images to 'baseball' card style 3D images. ARexx scripts already exist to do this.

Hope this was informative. Keep up the good work!!

Kermit Woodall
Design Czar
Nova Design, Inc.



The Truth About AAA

I read your comments about AAA in The Amiga Monitor 1_6. Your comments are...less than accurate. Let me take them one by one.

"It's hard to say whether something is 'too slow' or "fast enough" because those are all relative terms."

Not really; there are objective measurements for graphics performance.

"...nobody has ever seen the AAA chipset in action."

True and false. The first pass of AAA was indeed seen in action, at least to the extent that it worked. It didn't work very well.

"...by the numbers...alone, the AAA chipset would blow the doors off the AGA chipset, and easily perform on a par with most 3rd-party graphics boards available for the Amiga today."

Well...somewhat true, but misleading. Yes, AAA would beat AGA; AGA is exceptionally slow in most respects. As for current Amiga boards, AAA might beat some. Not the better, more recent ones.

"It would not take a long time to finish it, because it was very close to completion..."

Again, misleading. The first pass of AAA had been tested. The word from the engineers was that "major subsystems had problems." That is, plenty of things just didn't work. It was too complicated to fix them, so they were "zeroed out" and the changes were written to tape. It would have taken at least one more test pass, but since the last remaining AAA designer had left Commodore, any further changes would have been quite difficult.

"As far as Dave Haynie is concerned...nobody's saying he has to come back to finish the AAA chipset..."

This is probably a good thing, since Dave Haynie never worked on the AAA chipset. Dave did circuit board design, which is quite a different animal.

"It would not take anywhere near six months to introduce a powerful new 680x0-based Amiga."

Well...portions of the OS have to be rewritten for the 060; this isn't trivial. The same thing happened with the 040. And the 060 has now been eclipsed by other processors. You are correct to say that the Amiga "lives and breathes 680x0," though. Exec relies heavily on the interrupt structure of that family.

I have the full AAA spec, by the way, and would be more than happy to answer any questions you might have (a Commodore engineer was kind enough to mail it to me - all 118 pages plus 5 appendixes). You might be interested to learn that the spec was developed from 1988 to 1990, and was formally presented on April 11, 1990. It would have been a killer technology back then, but it's been left in the dust by PCI-based graphics cards with 64-bit (and better) bandwidth.

Jim Meyer

Indeed, there are objective measurements for graphics performance, but the arguments to which I was responding were presented in highly relative terms, and I was addressing them as such. If you want to talk about these objective terms, I'd be more than happy to. While not top-of-the-line in the computer industry today, the AAA chipset's performance capabilities are still quite impressive. This alone, of course, wouldn't be enough reason to use the AAA chipset; I've gone into the others many times before.

You are misquoting me here. I said "for all intents and purposes, nobody has ever seen the AAA chipset in action." This does not include Commodore engineers who, of course, saw what it did as they developed it. I thought it was fairly clear that I was referring to the Amiga community at large.

Yes, AAA would, as I have said, be far ahead of AGA. I would have to say that AGA actually does quite well for itself in most ways, and if anything, has been hampered by the lack of chunky pixel modes, but little, if anything, else. And AAA would keep up with most of what's available for the Amiga now, plus add the versatility of full backwards compabitility, and planar screenmodes which, believe it or not, have advantages in some cases.

Keep in mind that AAA was well on the way to development in late 1993, but when Commodore began to enter its death throes, development funds were drastically reduced, and then cut altogether. Given decent development funding, it would be quite close to completion, especially compared to everything that had to be accomplished in the initial design stages. Again, I believe we are speaking in relative terms here, and given the ease with which the AAA chipset could be integrated into an Amiga, there is plenty of incentive to actually perform that development.

I don't personally know exactly in what capacity Dave Haynie worked for Commodore, so I'll let the letter speak for itself here.

The OS already runs on the 68060, as in the case of certain accelerator boards. If anything had to be changed, it would be relatively simple, especially considering that the OS would probably still be able to run on a 68000, or at least 68020. In reference to the 68040, if you're referring to such things as the 68040.library, we're not looking at major portions of the OS here. And yes, certainly the 68060 has been eclipsed by some processors, like ultra high-end DEC Alphas, but it is still highly competitive when compared to the current Intel and PowerPC lines. I'm not saying it is every bit as fast as the latest Pentium Pro or PPC, but it is good enough, especially considering the speed and efficiency of the Amiga's OS, and the ease with which the 68060 could be integrated into the Amiga, to use it in a new Amiga system to help get the platform back on its feet. I am sure the OS could be adapted to the PowerPC architecture without an act of Congress, but yes, it will take a while to do it right, especially when you consider that the Mac OS is still somewhat 680x0-based, even after years of development, not stagnation as we have experienced.

I certainly don't know 118 pages worth of information about the AAA chipset, but I know its general technical specifications. For those who don't know, a quick list is: chunky pixel display modes, a real 24-bit display, resolutions to 1280x1024, a newer and more powerful blitter (with byte-movement capability) and copper, 64-bit architecture, 16-bit 8-channel sound, and, of course, full backwards compatibility. That is an impressive list of specifications, and although it isn't exactly "cutting edge," it is still quite powerful, and more than enough to be the basis of a new Amiga system, especially considering the other advantages.

It doesn't surprise me in the least that the AAA chipset was conceptually developed from 1988 to 1990. Many people probably don't realize just how long it takes for technology to be planned, designed, and finally developed, particularly in the computer industry. I don't know for sure, but I wouldn't be surprised if the AGA chipset was first conceptualized even before the A2000 was introduced. This does not make the AAA chipset "old." Much technology is designed long before it would even be practical to manufacture and market. In 1988 or even 1990, the AAA chipset would have been like a Ferrari in the days of the Model T. In 1994, it would have been revolutionary. Today, still, it would be a powerful chipset for the Amiga.

I would have to say "left in the dust by PCI-based graphics cards..." is an exaggeration. The AAA chipset could be configured as a 64-bit video system, so the mass market certainly isn't winning the "battle of the bits" on that end. Also keep in mind that PC's have massive, inefficient OS's, and a fundamentally flawed architecture that is a carryover from the XT days. Maybe using these 3rd party chipsets would be a viable option in the future, if the next owners of the Amiga don't have the resources to invest in the continuing development of the custom chipset, one of the Amiga's truly unique and more capable aspects over the years. But for now, they must use what is already established and what works, which means using something powerful that would integrate easily into the Amiga. And that option...is the AAA chipset.

Michael Webb
Editor-in-Chief



I just found your magazine & started reading through the back issues. Very good job, please keep it up. The thing that tweaked me & prompted me to write to your feedback column was all the whining and groaning about the custom chipset, especially the video chips. Sheesh, some of these people need to take a valium. The custom chips have always been what has kept the Ami ahead of the pack (until her owner went toes up); abandoning the concept now would be demonstrating that some people can't learn from history. What is there to learn? Two things leap to mind: 1. The custom chips create more power & what computer user doesn't lust after more power? 2. Let's take a page from the WinTel Weenies (hey, even they can be right sometimes); choice & interchangeability provide versatility & allow things to be updated as needed/desired. You don't like your old video? Just buy a new video card & plug it in. What the new owners need to do is not abandon AAA but update the OS so that manufacturers can build video cards that work seamlessly with it. That way the end user can go AAA or CyberGraphics or whatever. By making the custom chips upgradeable so a person doesn't have to buy a new box to get new features you have the best of everything.

Bob

I agree that trying to declare the custom chipset "unfit for service" is rather a case of looking for a scapegoat for the Amiga's problems. Well, folks, it's not the chipset's fault. The custom chipset has always been one of the best things about the Amiga, in conjunction with the OS and some other nice features. There is no need to jump ship for the multimedia subsystem.

Amiga OS with a truly modular RTG (ReTargetable Graphics) system running on a native Amiga custom chipset would probably be the ultimate configuration. As you say, it would allow full OS support for using whichever video board one desires. Because, of course, one day, the AAA chipset truly will be outdated, and owners of such machines might want to update their graphics power. It might even be wise to consider designing graphics cards based on future custom chipsets, for this upgrade purpose, or even perhaps to expand into other platforms.

But as you indicate, modularity in software is the key, especially since the Amiga truly pioneered that concept.

Michael Webb
Editor-in-Chief



Games Support

I think it unlikely that I'll be the only person welcoming the A\Box this month, after Phase-5 released the specs. It is important that the next-generation of Amiga and compatibles capture the imagination of the general public, and the A\Box offers sufficiently innovative and different equipment to facilitate that. It even looks impressive (should Phase-5 persist with the prototype case pictured on the cover of this month's CU Amiga).

The A\Box will need more than its specifications to survive, though. I don't think anyone doubts the loyalty of Amiga developers - their support for the Power-up program proves a willingness to invest in its future.

I hate say it. I never thought I would say it. But what a new Amiga really needs is games. Actually it has never been entirely untrue - a little more support for the AGA and higher-end Amigas wouldn't have gone amiss a few years back. Quantity is not the primary objective here. 3D is. If the A\Box is released with 20 Amiga circa-1992 type games, the general public will ignore it; moreover, treat it with disdain.

The way I see it, a couple of killer titles, which show the primitiveness of the PC, will make the A\Box a fashionable platform. They may not come in droves initially, but if the Amiga/A\Box's sales curve were to at last point upwards (it wouldn't take many sales) then a re-evaluation could start to take place.

I've e-mailed a couple of games companies asking them to confirm their position on the new compatibles, particularly the A\Box (because it seems to fire the imagination in a way that the others fail to do) and to show a little consumer interest in the new machines.

So far, only Bullfrog have replied. I was reasonable optimistic of their support, being as they were one of the first and one of the last major games companies to support the Amiga, but was disappointed with their response. The reason they dropped the Amiga, they said, was because it could no longer meet the technical requirements of their games. Of course, they did not rule out a return to the Amiga, but said it was unlikely because they do not at present have the resources to, as I requested, convert Syndicate Wars to Amiga.

Just one great title could demonstrate the possibilities of the A\Box to people like Bullfrog. It won't guarrantee success, but it certainly won't do any harm. Meanwhile, we can but support the A\Box and other machines. Save up for one. Students - take out a loan for one (like me). If it fails, run Windows NT and MacOS on it.

In the meantime, whether you're going to be able to afford one or not, start e-mailing companies to express your interest. E-mail every company you can think of, but mainly those who look most likely to be interested - those who developed for the Amiga in the past, those who produce for a number of different platforms now, those who embrace new technology which is not yet standardised (3D-graphics cards for instance, like Gremlin). But basically, e-mail everyone. Including Bullfrog.

I also e-mailed Team17 (proving to be one of the greatest friends the Amiga ever had - show them you're grateful), Microprose (because F1GP2 could be fantastic), the makers of Tomb Raider (their name escapes me, sorry), Gremlin, Virgin (to the webmaster to moan about the lack of Amiga in their survey, and the lack of a direct enquiries e-mail address for unusual requests). I e-mailed a few others, too. I've been busy.

I don't know whether we should be overloading Phase-5 with e-mails (they have enough to do already), but you could e-mail them to press the importance of conversions of select, influential games titles, as their interest seems rather applications-biased.

I agree with a policy of promoting serious Amiga software, because Commodore totally failed in this respect since they saw how many A500's they could sell, but a balance needs to be struck in the software itself (only now a problem) and in the marketing.

This year is make-or-break time and it's rather exciting. The WWW has given me a chance to do something a little more constructive to promote the Amiga than just shouting abuse at narrow-minded ST and PC owners like I used to do. I urge people to do the same. The former and latter (the latter's still the most fun).

Chris Owen

You brought up some interesting ideas, Chris. Many people, at first glance, would declare a game-oriented Amiga philosophy to be insane and suicidal. However, since it is the masses we want to attract to our platform, it certainly couldn't hurt to showcase the Amiga's gaming capability, especially in an era in which Doom, Duke Nukem, and Quake seem to be a primary market motivation for many potential or current computer buyers.

If you think back, actually, the Amiga's aptitude for gaming was what first showed many people its incredible multimedia and multitasking power. When desktop video and computer graphics applications came into being, most people surely made the connection that a machine that could run F/A-18 Interceptor the way the Amiga did could probably render and animate quite well. Some people point out the Amiga's reputation as a "games machine" hurt its success. Momentarily ignoring the fact that Commodore single-handedly killed the Amiga's chances, while some "serious" computer users may have dismissed the Amiga before ever seeing it, quite a few of us started out as game players. Today, this is more the case than ever.

I applaud Phase 5's many efforts to continue to develop cutting-edge technology for the Amiga, and particularly, the A\Box. However, I have to personally place priority on the "real" Amiga. The A\Box could be its successor, but there is not net a need for a successor. In other words, there's still a chance for success. Ideally, the Amiga and A\Box (and any other similar machines) could eventually share success. But there is no need, at least as of yet, for anything to eclipse the Amiga. To be sure, I am not "Anti-A\Box;" rather, I am "Pro-Amiga" above all else. But I wish Phase 5 the best in all of its endeavors, particularly the A\Box.

To those companies that believe the Amiga's technical strengths are not sufficient for modern games, I beg to differ. While there's no question that you couldn't run Quake on a 256K A1000, many Amiga users these days have at least 68020's or 68030's, and an increasing number are upgrading to 68040's and 68060's. Furthermore, any of the Amiga's chipsets provide plenty of power for most games that use low-resolution screens, while the increasingly popular graphics cards improve things further. To unilaterally declare that Amigas are not powerful enough for modern games is hogwash. Aurora Works, covered in the January issue of AM, is a good example of a company that realizes this, as they recently announced their committment to developing game software for top-of-the-line Amigas.

In conclusion to all of this, I agree that we should strongly encourage software developers of all kinds, games included, to develop for emerging Amiga and Amiga-based platforms; but also, to continue to support current Amiga owners. If enough people start contacting major software companies, they are at least guaranteed to notice.

Michael Webb
Editor-in-Chief

Haynie and AAA?

This is in response to the person who wrote in about Haynie coming back and finishing AAA. If I am not mistaken Haynie had nothing to do with the AAA implementation or any of the custom chipset work. Sure, he has worked on putting it on a motherboard and making it all work. Most of his work at Commodore it seems centered around the Zorro bus. So it would take him a long time to get up to speed and actually finish the implementation of somebody else's work. Of course if I am wrong, never mind.

I believe you are right -- I think the reason Haynie is associated with AAA is because he designed the computer that was to play host to it (the "Nyx" prototype). According to my information, he wasn't involved in the actual design of the chipset.

Michael Webb
Editor-in-Chief



Looking Into CPU Alternatives...

Hi All

I'd just like to put forward an alternative CPU to use.

All this talk of PowerPC is just a waste of time. We should go towards a powerful and cheap CPU, namely the StrongARM. This CPU is a wonderfully fast CPU, rated as the 3rd best in the world =)

Anyway, back to reality.

Patrick

Editor's Note: although I believe from the context of this letter that it wasn't meant to be taken entirely seriously, it provoked some responses from the editors that I thought brought up some worthwhile points.

The basis for a new CPU is just based on speed. There IS a LOT of development being done on PowerPC by a LOT of developers. The choice of the PowerPC gives us the opportunity to port over, without significant difficulities, code written for other platforms, thus giving us access to a far greater share of developers.

Greg Noggle
Telecommunications Editor and Hardware Guru



Well, speed is not everything. What of compatibility with older Amiga software? Remember, software sells computers and if you go with a vastly different architecture requiring massive reprogramming of all software, you will have the next generation of Amiga with no software. If you go with the PowerPC, maybe a little slower, you still have the capability of using much of the better-programmed Amiga software. That is very important to any company which invests money buying Amiga Technologies. Assuming, of course, that there is any Amiga software left by the time this is all done. Maybe later, if Amiga software is made truly hardware-independent, then other CPU's could be considered.

Anthony Becker
Executive Editor


Write us!

Do you have something to say about the Amiga, the Amiga community, the Amiga industry, or The Amiga Monitor? Click here to send e-mail to "Feedback", and be sure to include "Feedback" somewhere in the message header. Note that we can't publish all the letters we receive.