PlayCable research and development

[mr_me]

I was visiting the UK in 1982 and they had some sort of videotex system on the TV. News, weather, sports news on demand; it was great. Not sure why the videotex thing didn't happen in NA. The problem with Playcable was the cost of the Intellivision itself, US$300 was a lot of money in 1981. Most home computer users were pirating their games.

 

The cost of the playcable adapter is more sensitive than the server equipment. For every $12 you add to the device, it adds another month before it's profitable.

Edited November 23, 2018 by mr_me

[carlsson]

Hm. I might have misunderstood but I thought this service main was offered to existing Intellivision owners, or at least that cable operators would be pitched to add it to reach an existing customer group. It would be a little bit of Catch 22 to advertise to people who were looking to buy a home gaming system to go for the Intellivision since their cable operator offered add-on services to download games for it, while at the same time the operators would be told if they offered the service, more people would buy the hardware and eventually become customers.

 

As for videotex distribution in America, I've only read small blurbs along these lines:

 

///

 

Early 1984

IBM, CBS and Sears Roebuch & Co have initiated cooperation to introduce videotex services to the north east parts of USA. Teleshopping, banking, finances, news. They were expecting to be up and running in a couple of years. They were attempting to target a large part of the 7 million home computer users in the US at the time, including IBM PCjr. The long term aim was that a majority of the population would be using videotex in the 1990's, both at home and at work. Early experiments in home banking has shown that people in the US prefer to pay their bills around midnight.

 

The Canadian videotex system Telidon will have invested 280 million dollars by the end of 1984. The Canadian state corresponds to 1/4 of the investment, the rest would be industrial companies. By that time, Telidon were expecting to be profitable in 3-5 years, though they admitted to only have 5000 subscribers, much less than projected.

 

Early 1985

Keyfax in Chicago has invested 20 million dollars over the past 30 months and will now offer videotex services to customers in Chicacgo. The services include functions to order things, news, games, flight ticket bookings etc. The users may buy a Keyfax terminal for $365 or buy a program for $60 and use their own home computer. The monthly fee is between $15 - $30. Keyfax will spend 3 million dollars during 1985 on advertising their services.

 

Knight & Ridder who operate Viewtron, a videotex service in the south of Florida, have only been able to sign up 2800 users compared to the expected 5000 users. Thus Knight & Ridder are forced to fire 25% of their staff consisting of 169 persons.

 

///

 

If anything, it shows that people weren't and perhaps still aren't quite that early adopters of brand new technology if they don't understand the benefits and the price model is borderline greedy.

[+Lathe26]

Late to the party here, but I'll add:

1. The use of FM channels surprised me. The old statements about losing a TV channel made me previously think the PlayCable data was transmitted in the first few scanlines of a TV channel's image. This would have been similar technology to Closed Captioning text or the animatronic Barney the Dinosaur and Teletubby dolls that moved in reaction to their TV shows.
2. I spoke with Rebecca Heineman a few years ago. Atari was working on a competitor to the PlayCable. Here the the notes I wrote at the time (hopefully I have not misquoted or mischaracterized anything):
   She was working on PlayCable for Atari. It was its own box with a Z80 to help with the modem part. ... One piece that she mention was that authentication aspect but communication might still have been one way. The intention being that only subscribers who paid would be able to use it. One of the ideas was that this was going to be for a sup'ed up Atari 2600 where the new hardware had 32KB of RAM and had capabilities beyond the original 2600. The final intention being that the sup'ed up Atari 2600 would have unique games that ran only on it. This was all done with the Warner unit. It got cancelled and the time she didn't know why. ... Later she learned about the Intellivision PlayCable and figured that that had something to do with it.

[carlsson]

Given how many prototype consoles Atari seem to have been working on between the VCS and the eventual reuse of the A8 chipset in the 5200, I would assume that a "sup'ed up" 2600 was related to one of those next gen systems that never got finished. Otherwise it would just add one more layer of confusion if different divisions within Atari were working in different directions to what they would release next. I asked Rob Fulop about that, if he as a programmer had any input on the next generation of console hardware, but he didn't remember anything in particular, rather that the hardware guys would do their thing and the programmers got to work with what they eventually were given, no influence on which features, technologies or solutions would help them make better games.

[decle]

Hey guys,

dave1dmarx has kindly shared the section of his 1983 tape with Buzz Bombers loading on it:

 

buzzBombers.mp3

I think we can say definitively that not only were 8K games possible, but that they were implemented and distributed.

But that is not all. If you listen carefully to the first bit of the tape there are 9 short dings, followed by a double ding and then a further 7 short dings before the first long ding that indicates Buzz Bombers has been found and is starting to load.

In fiddling with the PlayCable I have heard these short double dings before, I generate them every time I screw up my streams and search for game X on a stream containing game Y. For example searching for Baseball on a stream containing Boxing:

 

doubleBing.mp3

Each time the PlayCable finds the start of game Y it dings once to indicate the start of a game header and then again moments later when decides it is not the droid game it is looking for and starts searching again. A double short ding, Hmm.

As the search continues you get one short ding for every 1Kbytes read from the stream, this is why the short dings come more frequently than the long dings, which represent 1Kdecles of game data loaded. A total of 17 short dings suggests more than 16Kbytes, but less than 17Kbytes of data have been read on the stream. That is too much data for Buzz Bombers, which is exactly 16Kbytes long (remember the header is treated separately, hence the double ding). And there are 8 short dings from the double ding (which you include in the count) to Buzz Bombers being found. That is exactly the right amount for a 4K game. Hmm.

I think this is the first evidence that the cable companies might have been transmitting more than one game on a channel. dave1dmarx says that these "glitches" were not unusual when loading games. Very interesting.

<MildSpeculation> If this is correct and the cable companies were double stacking games it has implications for the head end. If there is a 4K game being transmitted with Buzz Bombers it would suggest that there is at least 12Kdecles or 24Kbytes of data on this channel. The broadcast cards used to transmit the stream data supported two channels and must have had enough RAM on board for the complete contents of those channels. Let's assume for the moment that the data was not compressed on the card, it could have been, but let's assume it wasn't. Clearly, it would be insane to fill one channel with two games and leave its sister channel empty. Which would suggest that there is at least another 4Kbytes of RAM to support the second channel (this assumes the 2Kdecle menu program is on that channel), giving at least 28Kbytes of RAM on this broadcast card. This asymmetry in the sizing of the channels might look a little strange, however, it might make sense to pair the menu or a small, popular game, both of which needed to have fast load times, with a fully loaded game channel, in order to maximise the use of the memory on the card. </MildSpeculation>

<SpeculationCity> If I was to hazard a guess I would suggest that the machines originally had 8 broadcast cards (giving 16 channels, initially with one given over to the menu and the remaining 15 carrying 4K games, with no sharing) with each card would then have had at least 16K of RAM. This approach would minimise download times and maximise the user experience. Later, the RAM capacity or number of the cards might have been increased. Given the cost of RAM in the early 80s, it is not unreasonable to suggest that a head end might have ended up with broadcast cards with different RAM configurations, only a few of which had a larger capacity. It is interesting that the Q1 1983 roster only has five 6K games distributed and exactly two are scheduled each month. Therefore, a bit of thought might have been required when allocating games to cards in order to maximise the use of resources and give the consumer the best possible experience by minimising download times on popular titles. So, could it be that the disappearance of the versions of the PlayCable menu with the two longer tunes is broadcast card memory related and linked to the increase in the number of games from 15 to 20 and the inclusion of larger games in 1983? </SpeculationCity>

I'm hoping that dave1dmarx will share the rest of his recordings, it will be interesting to see if anything else can be inferred from the timing of the dings. I think this is another really good illustration of how the most inane things can provide bits and pieces that help to form a picture. Or at least wildly speculate on the possibility of a picture if you squint a bit and hold your head on one side


Cheers

decle

Edited November 29, 2018 by decle

[+DZ-Jay]

Wow! That is some analysis! Bravo!

 

Very interesting indeed!

[decle]

Hi all, Happy New Year!

In keeping with what is looking increasingly like a tradition, I am publishing something on New Year's Day. For 2019 I am making the first version of a technical summary of the PlayCable available (PDF):

 

playCableTechnicalSummary-20190101.pdf

 

Rather like the development description, I hope that this will become a living document that summarises all the research into the PlayCable's implementation. At the moment it mainly contains the work that has been done on RonTheCat's PlayCable adapter over the past 3 months. There is lots of stuff in it that isn't found on this thread, including some hi-resolution images of the PlayCable internals, circuit schematics and details on two PlayCable variations known to exist, so be sure to have a look. I will ensure that the latest version can always be found at the top of this thread.

I am also releasing a ROM image of the PlayCable Phoenix credits stream (this is the unadvertised stream #8 on page 5 of the PlayCable Phoenix replica menu):

 

 

pcCredits.rom

 

And for those that may have missed the ROM image of the menu program itself. Here is it again:

 

 

pcMenu.rom

As always, any feedback is most welcome. In the mean time, Happy New Year once again. I hope that 2019 will be Intellitastic for you.


Cheers

decle

Edited January 1, 2019 by decle

[irvinetom]

Wow, amazing! Incredible work.

[Byte Knight]

Thank you for making the demo available! It would be cool if real roms could actually be "loaded" from this menu.

[decle]

Thank you for making the demo available! It would be cool if real roms could actually be "loaded" from this menu.

 

Sorry, I would like to stay on friendly terms with the BSRs, so unfortunately it's not gonna happen.

 

On the plus side, I do have an update to the PlayCable technical summary:

 

playCableTechnicalSummary-20190107.pdf

 

intvsteve has kindly provided some pictures of his Jerrold PlayCable. Let's just say that inside, it differs significantly from the General Instrument one...

 

 

...and that this design has some fascinating implications for the potential to use the PlayCable for hacking an Intellivision. I can understand why Mattel might have been worried by it. A much expanded Section 8 gives you the low down.

 

We have also dug up some information on how PlayCable fitted into Jerrold's vision of information services in this marketing paper about DataChannel, the infrastructure on which it is built. There is also some circumstantial evidence that Jerrold were considering using related technology in other areas of their network, that was presented at the same NCTA conference in 1980. All very interesting.

 

 

Cheers

 

decle

 

Edited January 7, 2019 by decle

[pimpmaul69]

This is pretty awesome. Thanks for sharing this with us.

Edited January 8, 2019 by pimpmaul69

[Byte Knight]

 

Sorry, I would like to stay on friendly terms with the BSRs, so unfortunately it's not gonna happen.

 

I'm not saying to include the roms, just the ability to load the roms like an LTO Flash! cart. Speaking of that, this would make a pretty sweet "skin" for the LTO Flash!

[+cmart604]

Yikes. I can definitely give you a better scan of the game roster. Not sure where I took that picture. An underground cave perhaps? ?

[JohnPCAE]

The schematics of the Jerrold version will be very interesting!

[decle]

On 6/18/2019 at 3:40 PM, Byte Knight said:

Any chance you could a similar shell scripts with your Playcable demo so we could add games to it ourselves?

It's rather more involved than the scripts, but yes it is possible, here you go: pcFrontEnd.zip

 

Enjoy, Cheers

 

decle

[Byte Knight]

On 6/21/2019 at 3:33 PM, decle said:

It's rather more involved than the scripts, but yes it is possible, here you go: pcFrontEnd.zip

 

Enjoy, Cheers

 

decle

Thank you so much!!  I've been dying to play around with the PlayCable environment since I first learned of its existence...  Even if you could somehow find a real PlayCable, you still can't do anything with it.  So this is really cool!  

[Byte Knight]

After some fiddling around and help from decle, I was able to get 16 of my favorite games working with the PlayCable simulator.  I kept getting "Skipping nameofgame.bin as the size is non-standard" errors when I would try to compile on my MacBook.   decle suggested these changes to package.sh which fixed the error:

replace line 28 of package.sh from this:
 
        binSize=`stat -t "$1" | cut -d' ' -f 2`
 
to this:
 
        binSize=`stat -f%z "$1"`

and you will also need to change line 53 from:
 
binSize=`stat -t "playCable.bin" | cut -d' ' -f 2`
 
to:
 
binSize=`stat -f%z "playCable.bin"`

You are fairly limited as to what games you can add, which is true to the era.  Games that have a larger memory size or long names will not work, such as names like Advanced Dungeons and Dragons, so I had to go with the "Adventure" version.  Likewise, Activision, Imagic, etc games won't work with it.  I was eventually able to get 16 working favorite games.  Here's a video of the final product in action:

 

 

 

[decle]

Hi everyone,

 

I have what I hope is an interesting update of the PlayCable technical description to share this evening, which you can download here:

 

playCableTechnicalSummary-20190918.pdf

 

I suspect you will like the story behind this piece of equipment:

 

 

The big ticket items are:

• p38 - A more detailed description of Joe Jacobs' and Dennis Clark's Development PlayCable adapter
• p6 - A summary of the DCX11A DataChannel card
• p37 - An overview of Jerrold development engineers using hacked PlayCables to play games ripped to audio tapes
• p19 - Resolution of a potential discrepancy in the understanding of transmission frequencies

 

Enjoy

 

decle

 

 

Edited September 18, 2019 by decle

[decle]

Happy 2022 everyone,

 

In the ongoing tradition of publishing something to welcome in the New Year, I have an update to the PlayCable technical summary, which you can download here:

 

playCableTechnicalSummary-20220101.pdf

 

This brings you significant new material in the following sections:

• 6 PlayCable Hardware - All circuit diagrams have been checked and corrected against Jerrold schematics (p13-p19)
• 7.3 PlayCable Game ROMs - more information about which games were broadcast and when (p28-p31)
  • And did you know there was a quarterly PlayCable newsletter, that got to at least issue 3?

• 8.1 PlayCable Adapter Emulator - full circuit diagram of the Jerrold version of the PlayCable (p37-p38)
• 8.2 PlayTape Adapter (p39-p43)
  • Much more information on PlayTape and the changes made to support it
  • A description of DIRTAP a PlayCable demonstration that may have been similar to the menu program used in the broadcast service

   

• 8.3 CYBER Development Kit (p47-p48)
  • Information on the changes made and Joe Jacob's Vector board modification to the Master Component
  • Details on PlayComp, Joe and Dennis' CYBER based personal PlayCable service, quite possibly the best PlayCable variant of all.

 

I've also posted a brief description of DIRTAP and PlayComp, which includes a video and some interesting details of the unreleased games.

 

As always any additions or corrections please let me know.  And if this lot interests you, then you might want to take a peek at an update to the Intellivision development document here.

 

 

Cheers

 

decle
 

Edited January 1, 2022 by decle
Added links to other postings

[Walter Ives]

On 10/28/2018 at 5:50 AM, DZ-Jay said:

An attempt to recoup the Intellivoice investment was made by deciding to include the Orator chip and RESROM in the Intellivision III master component;

Nope. (1) Mattel management in general, and Denham in particular, were not suckers for the "sunk cost fallacy." You can't survive in the toy business unless you learn to take your lumps and move on. (2) No such decision was ever made. Voice synthesis was never included in the Intellivision III, except in some marketing literature.

On 10/29/2018 at 12:39 AM, decle said:

The PDP-11/03 was a the baby of the PDP-11 family. According to this site its LSI-11 processor ran at 0.05 VUP (VAX Unit of Performance - nominally 1MIP). This suggests the 11/03 ran at 50KIPS, and therefore, may have been even slower than the Intellivision!

Com'on, does that seem credible to you? Even the PDP-1 was faster than that. Your source is basing its rating on the "VAX Unit of Performance." The VAX was designed as a scientific computer, so the VUP included in its mix a proportion of floating-point operations representative of that used in scientific calculation. The 11/03 had a (micro-coded) floating-point processor that took about 50 usec/instruction (longer for division), depending on the amount of shifting and normalization required, which suggests that over 20% of the operations in the VUP mix were floating point. Just how do you think the Intellivision would rate if 20% of its operations were floating-point?

 

The 11/03's register-register instructions nominally took 3.5 us when running out of the on-board 4K-word DRAM and so were roughly twice as fast as the Intellivision's. I say nominally because the board's oscillator wasn't crystal controlled—a certain class of users used to have fun seeing how much they could increase the clock speed before the board started failing. Off-board memory was nominally asynchronous so executing out of that was somewhat slower--you could even use core memory for non-volatile applications.

Deliveries of the PDP-11/03 started in February 1975. By the time Intellivision started shipping FIVE YEARS later DEC had upgraded its low end offering to the considerably faster PDP-11/23.

On 10/29/2018 at 12:39 AM, decle said:

It also seems it could only address 64KB (32K 16bit words) of RAM.

What do you mean, only?

[+DZ-Jay]

4 hours ago, Walter Ives said:

Nope. (1) Mattel management in general, and Denham in particular, were not suckers for the "sunk cost fallacy." You can't survive in the toy business unless you learn to take your lumps and move on. (2) No such decision was ever made. Voice synthesis was never included in the Intellivision III, except in some marketing literature.

 

 

Fair enough, as you also pointed out in another post.  I guess, in the dark of more authoritative information, we all fell for the marketing myths as fact, along with some insinuations to this effect made by the late Kieth Robinson.

 

Thanks for providing the context and correcting the record.

   dZ.

[mr_me]

Voice capability may not have been in the prototype at APh but it wasn't just in marketing literature.  It was in the Intellivision III target specification at David Chandler's group.  Keith R wrote that in 1983 the price of voice synthesis chips dropped dramatically, that they considered including them in game cartridges.

[Walter Ives]

On 12/7/2022 at 4:55 AM, mr_me said:

Voice capability ... was in the Intellivision III target specification at David Chandler's group.

My previous answer conveyed the essence: here's the wordy version: The specification to which you refer is a draft by a well-meaning, newly-hired minion that summarized his effort to collect, process and circulate various ideas that were floating around. That is a generally worthwhile exercise, but the first cut of any such document can be expected to be, shall we say, somewhat less than definitive. As you can see, Chandler marked up the list of objectives on his copy to show, among other things, that the voice capability would be provided by the Intellivoice module, a not unreasonable position for him to take in view of the facts that the Intellivoice was just reaching store shelves and that he was planning an expanded version for use with Intellivsion II which he expected would also work with Intellivision III. The thrust of the body of this "specification," however, is not the feature list on this first page, but the idea that the Intellivision III should be implemented with an entirely new set of custom chips, including a CPU, that used DRAM-style RAS/CAS addressing. That concept was not original with Mattel, but was a total fantasy that if attempted here would sabotage the whole project. As it happened, this document had no influence whatsoever on the design of Intellivision III, which was not done by Chandler's group.

 

WJI

[Walter Ives]

On 12/7/2022 at 4:55 AM, mr_me said:

Keith R wrote that in 1983 the price of voice synthesis chips dropped dramatically, that they considered including them in game cartridges.

Well, the cost of fabricating the voice synthesis chips didn't drop between the time of the Intellivoice's introduction in 1982 and 1983—any new parts produced in 1983 would have been made on the same proven line as the old, so the statement makes no sense, eh? But it was nevertheless correct, for Mattel had far more devices on hand than it knew what to do with. That inventory represented a sunk cost: the parts were now worth zero, a dramatic "price" drop indeed. Could anything be done with them other than bury them in the sands of New Mexico or, geographically more practically, offer them as fill for Hong Kong's planned new airport? Absolutely: they could be incorporated into new voice game cartridges for the cost of soldering them in. Even better, at least as far as those who envisioned Mattel as a software-only powerhouse were concerned, they could be used to remake the B-17 Bomber and Bomb Squad cartridges for other platforms. While not exactly strokes of genius, both ideas were worth considering. As with many, many other good ideas, these were not immediately acted on. The arrival of the savior Morris made them irrelevant. When that savior didn't work out, the next savior, Valeski, decided against their resurrection because they didn't comport with his intention to tease out the last remaining value in the Intellivision line with the least possible risk or new investment.

 

There are another two paragraphs of weeds here beyond the scope of the comment to which this is a reply.

 

WJI

[decle]

The latest update of the PlayCable research document can be downloaded here:

 

playCableTechnicalSummary-20230907.pdf

The main addition is a significant expansion of the description of the PlayCable headend system, based on a Jerrold datasheet:

 

This provides us with quite a bit of detail regarding the hardware used, confirming what your $11900 of cold hard cable company cash bought you, and that the PlayCable games were distributed on tape.

Enjoy!