JamesD

There was an MSX machine that had a clone of the Z80 but it ran about the speed a 25MHz Z80 would. Search for MSX R800 and you should find info. (or so one web page says) (edit) Ouch... demand is pretty high to get that price! http://cgi.ebay.fr/ws/eBayISAPI.dll?ViewIt...bayphotohosting http://www.faq.msxnet.org/msxtr.html

My memory is fine thanks... I wonder about yours. Check what years that stuff come out vs what year the machine came out. When the computers first hit Atari was pretty protective of technical info. Just because that stuff eventually became available doesn't mean it was always around. The Amiga documentation was available to developers *before* the machine was even released. Then the developer docs were available for purchase by anybody shortly after it's release. My former partner sold the machines from as soon as he could get them. He had signed up to be a dealer when they first showed up at a computer show.

I think the first thing would be one of those educational game systems. For a strictly game system... Gamecube 1st choice... PS2 2nd.

I remember playing the Amiga version... I didn't have the patience to finish the entire thing!

I'm trying to figure out how the Atari was unique in this manner. That describes almost every 8 bit computer I've owned and I'm sure it applies to the C64 as well from the magazines and newsletters I've read. ST easier than Amiga? Oh please... the Amiga wasn't difficult to program at all. It took a little time to learn the library calls which gave it a steeper learning curve than 8 bits but you could bypass that with assembly and Basic was very easy... just like the ST. But then maybe you think it was easier to program the sound chip in the ST since it was something left over from the 8 bit world? The only significant difference was that the ST used the GEM desktop which had been available on the PC for some time which meant many programmers were already familiar with it. As for the memory map in hand... Amiga published EVERYTHING about the Amiga. The Amiga was one of the widest open machines ever created from the time it was released. Just try and say that about the Atari 800. The real reason the ST initially outsold the Amiga was because it was cheaper! Funny how things changed once the Amiga 500 was released.

I have to agree that "advanced follow up system" still could mean portable. I guess even a 7800 could be considered "advanced" when compared to the 2600 and it would be a likely candidate. I personally never cared for the 2600 or 7800 and would rather see the 8 bit but I don't have any say in the matter. Perhaps they are skipping the 8 bit and 7800 in favor of a portable Atari ST. That would have some pretty decent games and if it were fast enough could even emulate the other systems. The question is... could they meet the price point they wanted if it were an ST? The gate count for a 680x0 might be a little much for such a beast.

I've seen circuits that let a chip run double the buss speed while doing internal operations. If you do some searching you might find a design that would work.

Yup... that looks like an early 80 column card for the II+. I think I have one in my II+ as a matter of fact. It was before Apple introduced the graphics card/RAM slot that was introduced with the IIe.

I'm not sure I totally agree with it being a test game... remember, they just needed to convert 800 titles. Other possibilities I can think of: Easiest to convert. Shows off the controllers. Nostalgia for some of the first 2600 games. (wasn't breakout a pack in on the 2600 for a while?) They hadn't cranked up game production until managers gave the final ok but by that time they had to go with what they had ready in order to get the packaging done and ready for the holiday season. I think it's a little of all of the above. If they had done the original packaging, then slapped on a label "Includes bonus game Pac-Man" with a small picture of the game on the sticker and had two games in the pack for a release promotion it would have jumped off the shelves. The loss on the 2nd game would have been well worth it to have a large number of machines out there to attract developers to port their games to it.

Pac-Man might not have been ready to ship when the 5200 was released. The 5200 should have included just about anything other than Super Breakout when it was introduced. The Pong/Breakout type games had pretty much run their course by that time.

(deleted) I hate it when I make a post to the wrong forum.

And how exactly would they do that? Amiga was a separate company and Atari didn't have access to their intellectual property. It was a LOAN... they just assumed they would end up with the rights to market it. <correction> A lot of the cost cutting took place before Tramiel took over. He just continued down the path further. I sympathize with your argument but Atari was still bleeding money due to poor management, a price war in the computer market and lost revenues in the arcades. Cost cutting had to take place. Sadly, it wasn't done in an intelligent manner. They should have checked the status of software projects one by one and anything close to completion should have been finished. The revenue from the game sales would have been worth it... at least as long as there were not E.T.'s in the bunch. Then they should have taken the hardware projects and evaluated them and asked where they needed to be in a couple years. At that point they could decide which engineers were needed and then let people go. Duplication of games in the manner you describe just wasn't practical at that time. It's easy to say something like that now with the advent of cheap hardware, large hard drives, flash memory and the internet but back then... that stuff just didn't exist. For that matter... existing modems were still very slow at that time. If you want to upload, have people submit their stuff through the mail. It would be faster and more practical. I suppose submission through an online service would be possible but I remember what 1200 baud was like. 14.4 was about the first speed I thought was really useful. BTW, one of the early ideas at Atari was to market stuff through terminals. It was a great idea but the technology just wasn't there yet. If flash memory had been available sooner it would have made a huge impact on the industry. Take the Plus/4 for example. If people could have flashed whatever programs they wanted into the memory it might have looked more attractive to consumers. Then you have the updateable bios' or game carts... very cool! Another technology I wish had come out sooner was USB. I was using a protocol almost identical to it on a proprietary embedded system almost a decade before intel engineers worked on it and from what I understand the Atari buss protocol was also similar as well. The problem was that nobody wanted to let anyone else market peripherals for their computer. They wanted people to buy their product instead of whoever was cheapest or closest. Sadly, that's exactly what would have benefited consumers the most and would have allowed all kinds of advances in peripherals. It might have helped computer sales since people would be spending less on the rest of the hardware.

I'm sure opening up many of the chips like the Antic would have allowed some amazing capabilities. The TED chip in the Plus/4 has it's internal registers visible and modifiable by the CPU and people have done some clever things with it. BTW, the Timex Sinclair 2068 added double buffering, higher color resolution, joysticks and an MMU to the Speccy design. Too bad they didn't make an effort for more compatibility with the Speccy out of the box or the later Speccy's didn't borrow from more advances made in the 2068. Porting software from the Speccy to it isn't that bad though. If it hadn't been rushed to market before the engineers were done, it would have included some great stuff. Too bad they didn't give it a better keyboard either. IMHO, it was the most ambitious of the Speccy designs but Timex really fumbled the ball with the American market. I have a 2068 and a Speccy +2A myself. I've never turned on the Speccy though since I need a power adapter. One advantage I have to give to the Z80 is that it supports compilers a little better than the 6502. Yes there are C compilers for the 6502 but I just think it's easier for the compiler to generate efficient code for the Z80. The 64180 offers some enhancements that support compilers a little better. Too bad the Rabbit didn't exist then. It's has a Z80 like instruction set optimized for use with a compiler. Anyway, some of the new games that have come out for the Speccy were written in C.

A 3.5MHz Z80 and a 1.7MHz 6502 are similar in speed. It really depends on whether you write optimal code for one or the other. One of the big problems with many game ports is that the authors that did the porting may have known one of the processors well and not the other. If the custom chips start stealing clock cycles on the Atari that also makes a difference. But then the Atari games can look much more colorful and PM graphics really speed up sprite movement in some games. One real advantage of the Speccy was the graphics display. It has a fast layout for writing graphics to the display on games that can't use hardware sprites. 8 bits = 8 pixels and 256 pixels wide... just the the magic width to use 8 bits in width calculations. This makes the math on 3D games simpler and faster than if a 16 bit value is required. The color handling is a bit ugly compared to systems with the same number of colors where you can set the color down to individual pixels but a lot of games make very effective use of it too. To be honest, I think it's a little easier to code on the Z80. You don't have to do theatrics with page 0 to do the same type of things. However, as a result, I think 6502 coders wrote more optimal code out of necessity from the beginning where you could write downright slow code on the Z80 and get away with it. But then, on the Speccy forums I've seen really good Z80 coders cut half the clock cycles out of what looks like a pretty efficient routine and it's magazines published some pretty optimal routines for many things back in the day so even beginning programmers could start out with some optimal routines supporting their code. Even if their game logic wasn't optimal they could still push graphics to the display pretty fast.

Now imagine how popular it would be if Atari had included it with later machines.

Yeah, but they weren't an upgraded version of Atari BASIC so they wouldn't run the old code as well as new. It should have just shipped with newer machines and people with older versions could buy the upgrade. There's no reason BASIC XL or something similar couldn't have be sold by Atari. Does anyone actually have Microsoft Basic for the Atari? I've never seen the cart myself. The Coco had Color Basic or Extended Color Basic but the overwhelming majority of machines had Extended Basic installed even though it cost more. It was so popular it was standard on later machines. Uh oh... bashing the 800, 64, Apple, Spectrum and all is okay, but you DO NOT want to piss off the PET fanboys!!!! Hey, the PET wins the 'Way Cool Futuristic Looking' award hands down. Funny thing is... nobody makes a computer that looks anything like it now.

The MSX was very popular in some countries and has one of the more active communities in the 8 bit world. http://www.msx.org/msxforum.html He clearly digressed into MSX land making the post perfect for this thread. However... in the interest of dragging the off topic portion of his post more on topic with our off topic thread... But did the MSX have a decent tape interface? Hmmmmm????? And did any of them use a 64180 instead of a Z80? It is 20% or more faster than the Z80 once in native mode. There... that should do it! And Franky say relax! (sorry... couldn't resist the 80s reference) Actually, if you are just number crunching that's true. But then the BBC Micro and some Ohio Scientific machines were faster than the Atari. (I say some because it was an option you had to buy on OSI) So? Does that make them better than the Atari? Once you start coding games and clock cycles get stolen here or there to do this or that... the water gets muddy. If you were to write an arcade game in BASIC, the C64 has a clear advantage do to the more modern sprite hardware. The Atari requires you to move sprite data within the player missile buffer for up and down movement. The C64 BASIC is slightly faster unless you upgrade the Atari. That brings up something I thought Atari really missed the mark on. The upgradeable language on a cart was one of the best ideas they had... and they dropped it rather than creating their own Extended Basic which would have demonstrated the design feature as an advantage. They could have just marketed someone else's upgrade. The floating point speed should have been dealt with early on as well. That killed them in a lot of magazine benchmarks. But then the C64 had plug in BASIC expansions and you could always switch to RAM mode to load a language so I guess that isn't such a big deal. FWIW, the TANDY CoCo also didn't appear to be blindingly fast in BASIC benchmarks either but most machines after the first board revision or two supported a high speed poke where the CPU went roughly twice as fast. The COCO 3 had a ROM version that speeds up BASIC execution by 30% just by changing the way the keyboard was read... which would have made it one of the fastest machines out there once in double speed mode. But I never saw magazines mention any of that outside the Coco realm. The Coco 3 also has an 80 column text mode. The C128 and Apple II series are the only other popular 8 bit gamer systems that had that feature built in that I'm aware of. CPU speed does matter but there are a lot of other factors that determine whether one machine is "better" than another. It's very subjective and depends on what criteria you compare machines by.

Ok... I looked up the patent. It's actually related to what type of tape deck you use. If a cassette deck isn't computer compatible it doesn't invert the signal from the tape to it's proper phase. On the CoCo the load routines used the transitions instead of looking for a specific order of high and low parts of the sign wave and it was able to use about any decent cassette deck instead of just computer compatible ones. I think a side benefit was that if the sign wave on the tape was distorted slightly the machine didn't care as much as other systems. http://www.freepatentsonline.com/4468752.h...amp;stemming=on

I think the CoCo used transitions between low and high rather than peaks or lows to determine what data was on the tape. I believe TANDY had a patent on it and it made the CoCo tape interface less sensitive to noise. The Model I tape interface had been a little flaky even though it was very slow baud rate. Level I BASIC was something like 250 baud and Level II was 500. Tandy obviously learned their lesson.

The 6809 was actually very good for relocatable code and OS-9 took advantage of that. It also had two stack pointers so you could use one for an OS and the other for user programs. Most people just used the user stack pointer like another index register. The 68HC11/12 micro-controllers are very similar to the 6809 but they dropped the user stack pointer and made some instruction changes.

I'm not sure I like the inference here.... Oh please... I talked about kicking a dead horse in the previous post. Talk about this project has gone on for a long time and it sounded like you might still be too busy to work on it. I was just saying it wasn't a dead horse after all.

Yeah... save all your files twice so if one gets destroyed you have another copy.

It's lost because the 6502 doesn't borrow from the 6800. The 6500 borrowed the pinout from the 6800, but MOS was sued for that and then delivered the 6502 with a different pinout. Sorry, check the register models. They dropped an accumulator, added another index register, changed some names and simplified the instruction set. Both CPUs have fast page 0. 6800: ACCA ACCB X PC SP CC 6502: A X Y PC SP CC BTW, the 6809 also reduced the size of the 6800 instruction set. However with a smart assembler it was source code compatible for the most part. It also added another index register, the ability to move the fast page (DP) and allowed combining A & B registers for 16 bit operations. The 6309 added two more accumulators that could be combined the same way as A & B but could also be combined with A & B to deal with 32 bit numbers. A prefetch was also added to reduce instruction cycle times when in native mode. There are other additional registers but they aren't as usefull. The most useful additions (IMHO) were additional instructions with register to register operations and new loop instructions. The 6309 is about 30% faster than the 6809 running the same code when in native mode. The 65816 added a B register and some of the same A & B capabilities as the 6809 but it's less flexible. It did add MMU capabilities internally which is one of the biggest reasons for using this cpu in a design. The only comparable feature in an 8 bit CPU from the time that I know of was the 64180/Z180. The 65816 and Z180 (Zilog's 64180 version) are still commonly used. Hitachi made both the 64180 and 6309... too bad they didn't introduce a 6309 with an MMU.

I'm guessing it was partially a die size issue and it would have required redoing the VIC-II die over at the very least. If chips were still done by hand that's a very big deal. Now it would be trivial. I think they were already adding a tri-state mode to the CPU so rather than redo another chip they added it to the 6502. I'm guessing adding the port to page 0 simplifies the die since you only need to decode addresses on one page and there was probably existing circuitry in the chip already that could be used.

It just shows the genius of Chuck Peddle. He looked at the bus speed of available RAM and IO devices and made a chip that ran instructions quickly within those limits and had an incredibly small die size. It ran completely contrary to the idea that features sold processors. The fact that the 6502 borrows from the 6800 seems to have been lost in this discussion. These were the same people that designed the 6800 after all. What set the 6502 apart was the fact they had designed it to be simpler (cheaper) and to get higher yields from production. It was also about 4 times as fast as a 6800. Before the 6502 started a price war, you could expect CPUs to cost hundreds of dollars. FWIW, the later 6809 seemed to borrow a little from the 6502 and the 65816 seems to have borrowed from the 6809. Just imagine what those engineers could have made if Motorola had funded the engineer's low cost cpu project in the first place. That one mistake may have lead to intel's eventual dominance in the market.