Jump to content
Sign in to follow this  
gozar

YAWI (Yet another what if...)

Recommended Posts

I recently dove into a 600XL to upgrade it to 64K. While I had the cover off, I started to wonder why the 800XL ever existed...

 

So, for the first part of this what if, what if the 800XL never existed? Two scenarios (both assume the monitor jack is added):

  • Atari releases a 16K 600XL and the 1064 64K upgrade
  • Atari releases a 16K 600XL and a 64K 600XL (possible naming it the 800XL)
  • Atari releases a 64k 600XL

Concentrating on one model would bring down the manufacturing costs, and allow them to increase their profit margin. Also, to my untrained eye, it appears that the 600XL had fewer components and would be cheaper to manufacturer (for example, only 2 ram sockets instead of 8).

 

Second part: Adding integrated 80 columns to the XL line

 

One of the biggest advantages that the Apple //e line had was the 80 column option. Would the Atari computer been more competitive with the Apple //e if it had 80 columns built in?

  • Like 1

Share this post


Link to post
Share on other sites

I assume that eight 64Kbit DRAMs were somewhat cheaper than two 256Kbit DRAMs, at least until the time the 2-RAM chip XEs came out (when they went to the trouble of redesigning the 130XE motherboard around two or four 256Kbit DRAMs). The 600XL was obviously cheaper to produce, but one can only assume that a 600XL containing 2 x 256KBit DRAMs was not cost effective compared to an 800XL containing 8 x 64Kbit DRAMs... at least in the early Eighties. That's pure speculation, of course. :)

Share this post


Link to post
Share on other sites

but one can only assume that a 600XL containing 2 x 256KBit DRAMs was not cost effective compared to an 800XL containing 8 x 64Kbit DRAMs... at least in the early Eighties.

No need to assume, the internet knows all!

http://phe.rockefeller.edu/LogletLab/DRAM/dram.htm

 

That big chart confirms that 64kbit DRAMs were cheaper than 256kbit DRAMs until 1986. At the time of the 800XL's introduction in 1983, 2 256kbit DRAMs cost as much as 37 64kbit DRAMs - almost 5 times the cost!

 

- KS

  • Like 1

Share this post


Link to post
Share on other sites

Thanks! That's what I was looking for, but I was too inept to find it. Glad I wasn't talking rubbish. :)

Share this post


Link to post
Share on other sites

Maybe its just me but I never wanted a 600XL, it just looked like a runt of a computer, computer to me meant big, substantial, like the difference between the 400 and the 800, one was a computer but looked a little like a toy while the other had a classic style, proper keyboard (still my fave KB of the range).

 

Perhaps Atari felt the same and thought they needed a bigger flagship version?

Share this post


Link to post
Share on other sites

I don't particularly like external Ram expansion solutions. They're either/both unreliable or take away from further expansion possibilities.

Not to mention the space factor. I had a 600XL and it turned out cheaper to buy an 800XL and sell off the 600XL than to buy a 1064.

Add one and the 600XL goes from compact to cumbersome.

 

Interesting chart - I didn't even think 256 kbit Rams even existed before 1984 or so. Logically, cost would have played a huge factor in which ICs were selected. 8 chips vs 2 does mean extra space but the plastics and PCB extra cost would have been miniscule vs the cost of denser RAM.

Edited by Rybags

Share this post


Link to post
Share on other sites

I recently dove into a 600XL to upgrade it to 64K. While I had the cover off, I started to wonder why the 800XL ever existed...

 

So, for the first part of this what if, what if the 800XL never existed? Two scenarios (both assume the monitor jack is added):

  • Atari releases a 16K 600XL and the 1064 64K upgrade
  • Atari releases a 16K 600XL and a 64K 600XL (possible naming it the 800XL)
  • Atari releases a 64k 600XL
Concentrating on one model would bring down the manufacturing costs, and allow them to increase their profit margin. Also, to my untrained eye, it appears that the 600XL had fewer components and would be cheaper to manufacturer (for example, only 2 ram sockets instead of 8).

 

I think one of Atari's biggest problems was they fragmented their own market by selling multiple machines at the same time with slightly different capabilities.

 

Replacing the 400/800 with a single model that can be sold with 16K or 64K would make a lot of sense IMHO.

The CoCo supported such a RAM upgrade. You just swapped chips and changed some jumpers for the RAM upgrade. (original CoCo's weren't quite so easy)

 

I think the most important thing is that it should have been released when the 1200 came out.

It needed time to establish itself before the C64 came out.

 

Second part: Adding integrated 80 columns to the XL line

 

One of the biggest advantages that the Apple //e line had was the 80 column option. Would the Atari computer been more competitive with the Apple //e if it had 80 columns built in?

Atari was seen as a games company and I don't know that an 80 column display mode would have changed that.

Companies could have released software that displayed 80 columns on existing hardware, they just didn't.

 

This would either attract a lot of business software or turn out to be something that cost money to implement but doesn't really draw from the Apple II market.

Even with 80 columns you have to remember that the the Apple II+ still had slots and the IIe was expandable to 128K. You could add even more RAM to either one with third party RAM cards. The Apple market was moving beyond 64K while Atari was playing catch up.

 

I think it would be a bit like the CoCo3 which added 80 columns. You'd get a bunch of business software but not much of the major name stuff.

There is some potential there though just because it would have been much earlier than the CoCo3.

Still, popular business stuff would probably stick to Apple and IBM.

Plus you have to remember that some of the most popular Apple apps were written by Apple themselves.

 

The 1200XL came out over a year behind the IBM PC so the Atari wouldn't be seen as an upgrade path over an Apple at the late date of the 600xl's release.

For this to work, any replacement Atari with an 80 column mode almost has to be released before the IBM which means 1980. That means the XL series almost has to be the original machine which is unrealistic.

 

Perhaps an CTIA/GTIA upgrade.in 1980 would have made more sense. That way every Atari can support the new mode.

I think an 80 column text mode was certainly feasible in time for the release of the GTIA, maybe push back the GTIA a couple months.

At that early date the Atari would be very attractive vs the Apple II and has time to attract business software before the PC comes out.

I don't know if it would attract a big business market, but it eliminates one major advantage of the Apple II while adding one over the C64.

 

Whether or not it makes the Atari more successful, you have to admit there would be some sort of impact on the market as a whole.

Suddenly 80 columns looks like a standard feature for any new machine.

It might have pushed Commodore to add an 80 column mode to the VIC II.

 

At the very least, the Atari might be the machine of choice in countries where Apple has little presence.

That has to attract some business software.

Edited by JamesD

Share this post


Link to post
Share on other sites

I've never owned a 600xl, it seems like an odd beast in time. I'm trying to remember sequence in time, I think it came after another bunch of end of life and earlier systems that had been replaced by more successful derivatives. For example, the VIC 20 was replaced by the C64. The original Timex, which you could get a 16k expansion pack for, was replaced by the Spectrum. Apple did have the Apple III which was targeted more towards business, but didn't get a lot of penetration.

 

There was an attempt by Commodore to penetrate the business market with the C128. 80 columns and CP/M was not enough to derail the IBM PC by then.

 

I don't want to say all of these were unsuccessful, just that they weren't the most successful which is an entirely different thing. To me it seemed like the 600xl was a step backwards since you could get more memory in all its contemporaries and many of its predecessors.

 

There wasn't anything that would have beat the IBM. I worked for a Fortune 500 company at the time and the anti anything except DEC and IBM was like they had cooties. I built a home brew computer on my own time for some lab automation projects and remember being told by my supervisor "This is why you will never go anywhere in this company." It was considered a sign of weakness that you had any interest in things other then test tubes. Once my super boss, VP in charge of R&D, yelled at me when he found out I changed my own oil. He said I if I wanted to work on weekends I should pay to have my oil changed and he would give me a list of things to do at work that were important vs. wasting my time on things an idiot could do. When we finally did get some IBM PCs a senior vice president went though all the divisions he was in control of telling the managers and directors "If you see anyone playing games on the computers, I want them fired immediately!" He didn't tell the employees, wanted it to be a surprise. Probably an axe to grind like he was one of the administrators that was against bringing any form of automation into the work place. We were chemists and engineers after all, not IT workers. Kind of funny in retrospect.

 

You see price and performance didn't matter so much. It wasn't your money, it was the companies. I think Lotus 123 was about 5 times the cost of Visicalc but it was the companies dime so who cared?

 

Once that happened, IBM in the work place, everyone wanted an IBM at home. The standard software piracy went on for the home computer. I think the difference was larger sales since once again, you would actually buy a copy for work since it wasn't your money.

 

So what was left? Nintendo was doing well. Sega, Saturn, Turbo 16, Lynx did OK. Market became everything IBM, games and applications, and toys/game machines. Like I said, I've heard estimates that Atari sold ~2 million 8 bits into this climate and the C64 did even better. It's hard to call these failures. Even the 600xl is a relative success compared to some of the other computers at the time. Anyone ever hear of the Bluebird? :)

 

There's some interesting stuff if you look hard enough. I've been piecing together a lot of computers lately. You look at the motherboards and see memory/SD cards, just like the 800. You see expansion slots and think 800, 800xl, 130xe. You boot from a USB floppy and think POKEY/SIO. Atari actually had so many ways of doing things right, it is almost funny. They missed on a few important upgrade paths like processor and video. For the most part the things they had right were abandoned in their later computers. They did get on the IBM bandwagon w/o the common sense and future thinking they had originally. There is no reason the same path taken by other manufacturers like Dell could not have been done at Atari.

Edited by ricortes
  • Like 1

Share this post


Link to post
Share on other sites

Would the Atari computer been more competitive with the Apple //e if it had 80 columns built in?

 

 

Yes.

 

Word processing was a big deal in the early 80s. Most couldn't justify buying a computer to play games, but word processing was a big excuse/justification to make computers "practical" in the home.

 

By 1980 my father wouldn't pay for any computer that didn't have 80 columns. That's why we had various Apple II computers, then PCs, then Macs.

Share this post


Link to post
Share on other sites

I think the 80 column issue was once discussed elsewhere and one conclusion reached was that there isn't much point in having an 80 column display on a computer primarily used (at the time) with a TV set. To this day, software 80 column modes displayed via Y/C on a monitor aren't the easiest on the eyes, even when using an LCD. If 80 columns had been complemented by some kind of analog or digital RGB output, then yes - it would have made sense.

Share this post


Link to post
Share on other sites

Anyone interested in 80 column had a monochrome monitor like a green or amber screen. Presto, what was fuzzy 80 column text on an analog CRT became crisp, perfect text on a monochrome monitor. No other output or standards necessary.

 

You can still play games on monochrome monitors, they're just in varying levels of brightness (grayscale, greenscale, amberscale, etc.) instead of color.

 

Monochrome monitors were also significantly less expensive than quality color monitors.

 

The only thing missing from people enjoying 80 columns on their Atari computers was decent 80 column output. If that had been available people would have jumped all over it. We would have had at least one Atari in our house instead of various Apple II models. The Atari was a better computer and cost much less, but it couldn't do decent 80 column for my father to do work on.

  • Like 1

Share this post


Link to post
Share on other sites

The C128 wasn't exactly unsuccessful at about four million copies sold....it arrived 2 years after the 800XL and while it had 80cols built-in there's not much software that supports it and it isn't really "integrated" into the system. 40col and 80col output are generated by two different video chips and require separate connections to the monitor using different signaling (composite or S-Video for 40 col and RGBI for 80col, making the latter hellishly difficult to use on modern hardware.) It's probably safe to guess that many if not most 128s were used as high-end C64s and hooked up to TVs or regular "40col" color monitors.

 

80 columns were available with the Bit3 for the 800 well before the 1200XL and 800XL arrived but apparently did not attract as many users as 80col-cards for the Apple (too lazy to compare historic prices). Maybe Atari considered the market penetration of the Bit3 (and similar) when designing the 1200XL (for which it would have been a logical upgrade). For some reason the 1200XL was designed without any upgrade path and as the 800XL was built to be cheaper than the 1200XL it was rather unlikely that it would gain any functionality.

 

I don't really remember if I ever played an Atari hooked up to a mono monitor but I certainly tried this on some contemporary computer and it just wasn't the same. 80col compatible color monitors were so hellishly expensive at the time that they would probably have cost twice or three times as much as a complete 800XL/1050 setup. Mono was cheaper than regular color monitors but using two monitors isn't that practical. I tried it for some time on the ST but that was only affordable because the color monitor was shared with another machine (and prices had dropped somewhat by that time).

 

I vividly remember reading about the first IBM PCs and wondering why anyone would want a machine with such meagre offerings at outrageous prices. But then you have to remember that those PCs replaced dedicated word processing systems ("storing typewriters") costing two or three times as much. At least over here Atari never had a dealer network that was able to sell working packages to businesses. (I don't think Apple was either, so you simply saw less PC use in small businesses before the arrival of the PC clones).

Share this post


Link to post
Share on other sites

Just remember that even with 80 columns, that doesn't solve all of Atari's problems in the business market.

Atari BASIC wasn't exactly known for being a good version for business because of the string handling, lack of the PRINT USING command, etc...
This may not seem like a big deal but a lot of custom business software was written in BASIC.
Maybe another BASIC would have stepped up for the business market but Microsoft BASIC for the Atari didn't come out until '82.

Then you also have the Atari 810 disk drive. It only held 88K where the competition held close to 150K.

The cost of Atari drives was also greater than Apple II drives.
By the time you buy a dual drive system, any cost savings on the main computer vs the Apple II is largely eaten up and you still only have the storage of a single drive Apple II.
Replacement drives didn't arrive until the XL line in '82 which is after the IBM PC is released and again... the Atari doesn't look so much like an upgrade over the Apple II. (was a double density upgrade available sooner?)

And then you need an interface to hook up a parallel printer or serial port for a modem other than Atari.

So I think there are a series of issues Atari needed to deal with if they really wanted a part of the business market.
80 column support, an upgraded BASIC and a drive upgrade are all doable without a new machine.

Edited by JamesD
  • Like 2

Share this post


Link to post
Share on other sites

Since I was so young at the time, I went to archive.org and read the Dec. '83 issue of Byte. I never realized how big CP/M was, no wonder Atari was looking at creating the 1060. Atari and the rest of the 6502 machines had already lost the business world by Dec. '83.

 

Anyway, a 800XL with a 1050 disk drive was ~$650. An IIe, 64K, and disk drive was ~$1,400. A C64 and 1541 was ~$500. PC clones were over $2K. Not only would 80 columns put the XL more on par with the IIe, but it would also help it compete against the lower priced C64. Even though the business world was lost, the home market was not.

 

 

p.s. That issue also has a review of the ATR-8000 by David and Sandy Small along with an article about Microsoft Windows.

  • Like 1

Share this post


Link to post
Share on other sites

Since I was so young at the time, I went to archive.org and read the Dec. '83 issue of Byte. I never realized how big CP/M was, no wonder Atari was looking at creating the 1060. Atari and the rest of the 6502 machines had already lost the business world by Dec. '83.

...

I believe CP/M came out in '75 but I don't think CP/M was a big seller in the personal computer market until the introduction of machines like the Osborn 1 and Kaypro in '81.

The IBM PC also came out in '81 so opportunities for Atari to expand into business drop off rapidly after that.

 

FWIW, a lot (most) of those business apps weren't available on CP/M much before the Atari was introduced.

Wordstar didn't get released until '78 and by '80 had sold a whopping 5000 copies. By June of '83, Wordstar had sold 650,000 copies for the PC. That should tell you where the business market was after '81 and it wasn't CP/M. I'm sure the Osborn, Kaypro and TRS-80 Model IV helped CP/M numbers significantly, but nothing like PC numbers.

 

Based on that, even *if* Atari had part of the business market, you can't be sure there would have been millions more machines sold.

I think an 80 column mode would be more important just to set the Atari apart from other sub $1000 machines than to become a player in the business market.

 

Share this post


Link to post
Share on other sites

I said 80 columns could be added through a new GTIA but that may not be true.

It may require different timing, a different ANTIC and possible different analog circuitry.

That would have to be something planned from the start or part of a new design.

Share this post


Link to post
Share on other sites

To a large extent, we are a evolved SIG. By that I mean the we are not typical users and the things important to us are not necessarily important to > 90% of the people who bought Atari. For a time snapshot back in the day, we were programmers, professionals, BBS users. 80 columns was then and is now important to us.

 

Now thirty years later we are still the same SIG. What has changed is: Instead of one person in two hundred intercommunicating, it's more like 1 in 2. We are still 1 in 200, it is just that now several billion people have joined us in the telecommunication part. Meh, if the 600xl was the odd man out back in the 80s, it probably seems like it was made on a different planet by most consumers now. Doesn't have a built in touch screen, wifi, flash storage, can't make phone calls, and won't fit in your shirt pocket.

 

Once again, Atari did have some talent and could see the way things were going. The 600xl had a substantially smaller footprint and volume that foretold the direction consumers would go. Atari was also into video telecommunication which I see as a direct predecessor to Skype and Youtube.

 

Just a few historical notes that seem important to me and probably won't resonate with anyone else. :) The original PCs actually had a cassette interface! Could you imagine using Wordstar running from cassette! Of course i don't think it ever came out on cassette, just illustrates how badly IBM saw the market. The first PCs only came with ~CGA and mono adapters. The original Lotus 123 could not even show a graph preview on our text style monitor. We ended up with 'special' computers that had both monitors for graphics and text cards, i.e. two monitors sitting on the desk. The mono graphics monitor would light up and display the graphs when called on. This lead to the IT crew buying a ton of Hercules graphic cards which could do text and graphics with the same card. Soon as the AT came out with EGA that could do it all, the existing PCs were turned over to area secretaries for word processing use or data terminals to access the DEC mainframe.

 

Real tangent, been assembling retro hardware to do some mischief. One of the computers I got was a ~10 year old Dell. I did some research on it and it turns out Dell stopped including floppy drives ~10 years ago. They still had the connector for one, but proprietary cabling and no existing mounting hardware bracket unless it was ordered with the floppy. The bios does not even include anything other then 1.44 meg 3.5" drive. We are fast getting to the stage where kids under the age of 20 will not know what a floppy is. I think when I do a retro game night with my grand kids I will try telling them my Atari computer came from a Pharaohs tomb in ancient Egypt and see if it flies. :)

 

Just a by the by comment on 80 columns. The early designs needed to do things like PLL to up the system clock to be fast enough for use as a dot clock. Bob Wooley in particular did designs for 80 column displays back in the 80s. He took a different approach by IIRC just using a 14.318 MHz crystal/resonator which directly feed the parallel to serial dot clock chip and used a single TTL counter chip to divide it down for the system clock. The 130 XE which was already out at the time, and I am sure Bob owned one, had a 14.318 crystal in it and would have been an easier hack. As always, Bob was predisposed to 1200XLs. Probably would have been relatively trivial for Atari to add a 6545 video chip for 80 columns at that time of the 130XE but could'a, should'a, would'a. An off the shelf 130XE with 80 column video may not have become a world beater to the general public but I am sure it would have been to a number of people here including me.

Share this post


Link to post
Share on other sites

I suspect that a 130XE-only 80 column solution released in 1985 or later would not have had much impact. The Apple II had many easy and compatible 80 column expansion options by 1980. By 1983 80 columns was built-in to the Apple IIe, then the Apple //c in 1984. A decent 80 column solution had to be released no later than 82-83 with the XL line to compete effectively with other computers, build a library of compatible software, and cease (or at least slow) the mass exodus of software publishers from the A8 that happened in 1985. 80 column support had to happen early enough to encourage more consumers to buy Atari instead of Apple, which in turn would have lead to the A8 being supported for longer like the C64 and Apple II.

Share this post


Link to post
Share on other sites

80 Columns needed to be added to the XL line as early as possible for it to have any affect. Maybe even create a compatible card for the 800? But since it wasn't even in the plans for the 1400XL, Atari apparently didn't view 80 cols as a necessity. That's too bad. It would have made it a lot more appealing to students, especially those heading to college.

  • Like 1

Share this post


Link to post
Share on other sites

No need to assume, the internet knows all!

http://phe.rockefeller.edu/LogletLab/DRAM/dram.htm

 

That big chart confirms that 64kbit DRAMs were cheaper than 256kbit DRAMs until 1986. At the time of the 800XL's introduction in 1983, 2 256kbit DRAMs cost as much as 37 64kbit DRAMs - almost 5 times the cost!

 

- KS

Does anyone know why the intermediate sizes (32K, 128K, etc.) were not available? It would make sense to have the DRAM chip sizes in multiples of two.

Share this post


Link to post
Share on other sites

Does anyone know why the intermediate sizes (32K, 128K, etc.) were not available? It would make sense to have the DRAM chip sizes in multiples of two.

 

Due to DRAMs using a multiplexed address bus, adding one pin gives two bits of address.. So, capacities grew by 4x every time...

Share this post


Link to post
Share on other sites

Does anyone know why the intermediate sizes (32K, 128K, etc.) were not available? It would make sense to have the DRAM chip sizes in multiples of two.

 

Intermediate sizes were available, but they were never popular and thus never very cheap. You have to know a bit about DRAM to understand why:

  • The actual bits, or "cells", are the cheapest part of the DRAM chip. (Even back then, they used redundancy to make cells yield near 100%.)
  • Besides the bits, there is a bunch of costly overhead per chip - the pins themselves, "row drivers", "column sensors", the plastic package.
  • The largest chips are the cheapest per bit - they have the most bits per overhead cost.
  • BUT, there is a limit to how big a chip can be (based on the "reticle size"), and the largest possible chip is approximately square.
  • So, the cheapest DRAM chips put their bits in a square grid of rows and columns. For example, 256 rows * 256 columns, or 65,536 bits.

To fit in the recticle, a 128Kb rectangular chip would require the same type of transistors ("process") as a square 256Kb chip. But, it would have more overhead. So, it might be something like 65% of the cost of a 256Kb chip - not a good price per bit.

Every couple of years, they invent a new "process" that enables more transistors to be packed in the same square. The transistors shrink in height and width, so the next generation is usually four times the bits - going from 256*256 bits to 512*512 bits. When a process is new, it is a lot more expensive until the equipment is paid off and competition ramps up. For that reason, the cheapest chips are usually one "process" older, or 1/4 the size, of the largest chips at that time. The chips that are two processes old are rarely worth buying, because their overhead is too high for the bits you get. So, around the time 1Mb chips become common, 256Kb chips become the best deal.

 

(This doesn't apply to modern DRAMs. Modern DRAMs are no longer laid out as simple grids - modern transistors are too weak to support a single grid of, say, 65536x65536 bits. So, now chips are made of several smaller grids - this lets you have an approximately square 2Gb chip using, say, 8 smaller 256Mb grids.)

 

- KS

  • Like 1

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
Sign in to follow this  

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...