Original 125 release dates

[Walter Ives]

On 7/6/2023 at 9:34 AM, Lathe26 said:

Was the Home Budgeting you mention related to the Family Budgeting / Budget Program / Home Budget Program (the various names embedded in the code) that was written using KC BASIC?

All I meant to say in the referenced post is that the six programs I listed were on Rochlis' earliest lists of programs to write. The exact names were fluid, for example, Nutrition might be referred to as Weight loss or Art Ulene. Nor was my list intended to be complete, as Federal Income Tax Preparation was also on Rochlis' earliest lists. At that time, the plan was there was to be no BASIC. My suggestion that these were killer aps was sardonic.

 

To that I can add that Conversational French, Physical Exercise, Nutrition, Astrology and Speed Reading were either completed or in progress and were not written in BASIC.

 

As to Home Budgeting, there's a long story here. As I mentioned, Rochlis also wanted a Federal Income Tax Preparation program, and that was a much higher priority. It's easy enough to wave one's hands, or in Rochlis' case one's cigarette, and say you want such-and-such a program, but much harder to come up with an implementable idea that makes sense. The inspiration here was the insight that a tax return and its underlying schedules were in effect a hierarchical collection of multi-column ledgers and sub-ledgers. For example, charitable deductions were listed in a sub-ledger that consisted of four columns: a sequence number, a date, a description and an amount. When you came upon a receipt for a charitable deduction you'd open up that sub-ledger, move the cursor to first field of the appropriate line, add an appropriate item entry and then toss the receipt into a box. Fields were edited using the same editor used for the Keyboard Component command line. As items were added the program would update the running total; when you closed the sub-ledger the amount would be propagated through the rest of the form.

 

Management had a predilection to organize complex programs by creating generalized frameworks on which customizing materials could be hung (witness the EXEC and PICSE) and determined to do so here. It recognized at that time that Federal Income Tax Preparation and Home Budgeting could both be built on the same ledger framework. It further recognized that one could also use that same framework to build accounting programs customized to various different types of small businesses.

 

The underlying design was initially sketched out in the spring of 1979, triggered by the same January 1979 design meeting between GI, Mattel and APh that triggered the Keyboard Component's circuit design. APh was able to complete the design and construction of prototype Keyboard Component hardware by summer, but there was no way the associated cassette operating system was going to be ready in that timeframe. It nevertheless tried to make some progress on cassettes across a broad front (remember, this work was all being done under the auspices of Mattel's Preliminary Design department, whose operating mode was to investigate many ideas and then concentrate resources on the few most promising ones). The initial work, however, didn't even need the Keyboard Component, as much it could be done by pressing a normal VT52 terminal into service as a display device.

 

Parallel development proceeded on the ledger framework, Federal Income Tax and Home Budgeting in a manner similar to the parallel development of the EXEC, Baseball and the first few cartridges. Compared to Federal Income Tax, Home Budgeting was trivial. There were, however, few significant differences. For example, Federal Income Tax had to accommodate at most four column ledgers (line number, date, description, amount), whereas Home Budgeting had to accommodate sixteen, including columns for each month and the whole year.

 

APh often dipped into its library of pre-existing code: here the expression evaluator used to evaluate cells was scavenged from one that was written in IBM 360 assembly language by one of APh's prinicpals many years earlier.

 

Interestingly, the structure of these programs anticipated what we today call spreadsheet programs. The parallel is really obvious in the above description of the charitable deduction sub-ledger. However, if you looked closely you'd recognize that even the ledger for Form 1040 was in essence a spreadsheet whose fields could be positioned to reflect the layout of the actual tax form. The first computerized spreadsheet program, Visicalc, was released in October of that year. It didn't take long for Mattel/APh to become aware of Visicalc: by 1981 lists of Keyboard Component software in or considered for development included a generalized spreadsheet program.

 

As to the BASIC version, Rochlis did not intend the Keyboard Component to be programmable by the user. The decision to add BASIC wasn't made until 1980. The completion of BASIC allowed Mattel to supplement the library with a number of cassettes programmed in BASIC. These programs were rather limited in scope and considered stopgap measures, and were priced very high for the value received. In many cases the bulk of these program didn't need to be developed on using a Keyboard Component; any computer running Microsoft BASIC would do.

 

The BASIC version implemented a subset of what was in the comprehensive version. Internally, it held its ledger field data in BASIC arrays. It only supported specific pre-defined categories of input and expense; these were numerically encoded rather than held as alphanumeric text strings. Data was input in response to program demand prompts rather than by moving the cursor to the appropriate field and typing, and was therefore somewhat tedious.

 

I hope you find the answer to your question in here somewhere.

 

WJI

[Walter Ives]

23 hours ago, DZ-Jay said:

I find all your comments fascinating -- even those castigating my own opinions.

Castigating!? Ouch. Here I thought I was being kind and gentle. I'm glad you're making the distinction between castigating an opinion and savaging its holder, though.

 

To the extent this remark was triggered by my comments on your post on the Audience/Robinson/Daglow exchange, where I pointed out that the exchange didn't comport with facts and timelines you already knew from other sources, you might be interested to know that Rochlis and Krakauer were making statements to the press that were consistent with my further assertion that there was a definite policy of not doing arcade-type games:

One really can't be licensing and peddling arcade games, licensed or not, if that's what you're telling your customers.

 

WJI

[Walter Ives]

23 hours ago, DZ-Jay said:

or as has been suggested to me by at least one individual, "Internet trolling of this small and information-starved community."

 

Well, I guess one of my last posts did happen to include a line that could be seen as trolling Canadians. Caught one, too:

On 7/6/2023 at 7:09 AM, Sinjinhawke said:

On 7/6/2023 at 4:33 AM, Walter Ives said:

Since by the fall the Master Component was also available in Canada, does the use of the term "nationwide" here imply that by then Canada was already generally accepted as being a de facto territory of the United States?

Never, not as long as I own a sharpened hockey stick.  Also you yanks are too religious and fanatical about your political parties.  Just stay down there where we can keep an eye on you.

WJI

[Walter Ives]

On 7/7/2023 at 4:35 AM, DZ-Jay said:

So, in the interest of transparency and of building a coherent, complete, factual, insightful, and authoritative record of the history of the Mattel Intellivision, would you mind offering at least some insight into your sources?

 

No big secret. From my very first post in this series (although I had made a few posts before that):

On 12/3/2022 at 5:00 AM, Walter Ives said:

I induced someone who worked with Chandler back in the day to review this forum with me. I've made some comments you may find helpful. Please allow for the fact that these comments relate to matters that occurred forty years ago and were made without reference to notes. There may be some inconsistencies due to memories returning as they were being refreshed. Also be hyper-aware that I may have garbled some of the information in translation.

 

I'm starting here because this is the first thread we looked at.

I've since spoken with others he's still in touch with.

 

WJI

 

[+DZ-Jay]

3 hours ago, Walter Ives said:

Castigating!? Ouch. Here I thought I was being kind and gentle. I'm glad you're making the distinction between castigating an opinion and savaging its holder, though.

I do make that distinction consciously.  I also appreciated your insistence in trying to correct my understanding in other posts, not in an "I'm right and you are wrong" sort of way, but in the "really, listen to me, this is what really happened, and it is not at all what you are thinking" way.

 

As for "castigating" ... let's just say that I've been known to employ hyperbole for dramatic effect.  Maybe one or twice ... or million times.  😁
 

3 hours ago, Walter Ives said:

To the extent this remark was triggered by my comments on your post on the Audience/Robinson/Daglow exchange, where I pointed out that the exchange didn't comport with facts and timelines you already knew from other sources, you might be interested to know that Rochlis and Krakauer were making statements to the press that were consistent with my further assertion that there was a definite policy of not doing arcade-type games:

One really can't be licensing and peddling arcade games, licensed or not, if that's what you're telling your customers.

 

WJI

For the record, my remark was merely incidentally posted in this thread; it was intended as a general request for a formal authoritative source in order to be able to treat your comments as more than general apocrypha.  Really, it was a general sentiment shared by various individuals in this forum for a while -- I just happen to be the one who stepped up and asked.

 

3 hours ago, Walter Ives said:

 

No big secret. From my very first post in this series (although I had made a few posts before that):

I've since spoken with others he's still in touch with.

 

WJI

 

 

For what it is worth, my opinion is that, in an ideal world, I would like to see these accounts written down and published as an official history of the Mattel Intellivision, complete with named authoritative sources and verifiable citations.

 

However, I recognize that this world is far from ideal, or else I would not have to try so hard to recapture my youth by holding on to my childhood toys.

 

Therefore, I take what I can get, and so I appreciate your posts very much as they are.

 

(Still, it would be nice to know who those people are and hear from them directly, and to know what is your own personal place in this whole story.  I'm not holding my breath, though.)

 

    dZ.

[mr_me]

3 hours ago, Walter Ives said:

I screwed up.

 

I shouldn't have used the exclamation points to highlight the word SuperGraphics. That highlighting is something I invented for that post—Mattel didn't write it that way it back in the day. I know this didn't confuse the regulars, but I don't want to mislead casual visitors to the site. Sorry. The word should be rendered as SuperGraphics.

 

A little more on SuperGraphics.

 

The early 4K cartridges devoted a little over 1.5K to storing graphics leaving 2.5K or less for game play. This often limited the graphical impact of the games. Witness Basketball vs. Basketball II, PGA Golf vs. Chip Shot Super Pro Golf, etc.

 

Mattel started allowing cartridges under development in 1982 to go to 8K and the ECS titles to go to even more. This made more memory available to store graphics, as a result of which the games began to look more graphically impressive. Exceedingly impressed with the graphics that had been designed for World Series Major League Baseball, marketing planned to call the ECS games "Super Games."

The poor sales of Intellivision during the Christmas selling season led to emergency management meetings. Among other things management had the inspiration of leveraging the term Super Games to create the term SuperGraphics. SuperGraphics was envisioned as something that was associated with the cartridge that made the console perform better. Marketing would have liked to call it "SuperGraphics Technology," but "SuperGraphics programming" was as far as legal would let them go.

 

Booth personnel at Winter CES in January and Toy Fair in February, who were not really gaming types, were trained that the 1983 line of game cartridges looked better than those of prior years because of SuperGraphics. They weren't told what that entailed beyond, "it’s a way of getting better graphics out of the system."

The term did not get much traction at that point, partly because the catalogs and other marketing material that had already been printed didn't use it and partly because the sales force was at a loss for words when it came to explaining it to store buyers.

 

When orders for the Intellivision II in the spring were underwhelming, buyers often gave the excuse that they would wait for Intellivision III.

 

In 1977-79 Mattel could simply keep announcing delays to the introduction of the initial Master Component. That wouldn't work for Intellivision III in 1983, and in the run-up to Summer CES Mattel had little choice but to announce its cancellation. Being forced to do that, Mattel really needed something compelling to keep the trade from concluding that it was withdrawing from the market, so it doubled down on SuperGraphics, telling the trade and the press that the capabilities of SuperGraphics made the Intellivision III unnecessary. Management may have served this Kool-Aid, but it didn't drink it, and allowed development of Intellivision III to proceed apace.

 

WJI

 

[This graphic does not belong here]

 

Regardless of what salespeople were telling retailers, there was only one cartridge from Mattel Electronics that I know of labelled "supergraphics" and that's Masters of the Universe.  It's also the only Mattel Electronics published cartridge that I know of that updates graphics at 60fps.  That of course requires more cartridge ROM storage to replace the Exec game engine.  Did APh program any Intellivision games that improved on the Exec 20fps?

 

On 7/6/2023 at 7:28 AM, Walter Ives said:

The following eight cartridges were delivered to retailers in Fresno for sale as part of the initial roll-out: Blackjack & Poker (pack-in), Baseball, Basketball, Football, Armor Battle, Checkers, Backgammon and Math Fun. So you can list those eight as having been released on December 3, 1979.

In 1979 GTE Sylvania reportedly was also selling Intellivisions in their stores in select markets on the east coast.  Do you happen to know when that started?

 

On 6/26/2023 at 2:06 PM, decle said:

So what can we learn from this?  Well, the first thing to note is that according to the September 1982 issue of VGU any release dates originating from Mattel should be taken with a pinch of salt:

 

 

The need for caution is apparent when comparing the Mattel and VGU dates for games released in the latter half of 1982.  Where the BSRs report precise dates, VGU typically only acknowledges games as being with distributors a couple of months later (where no precise dates are given on blueskyrangers.com I push the release date to the end of the period, so a "1982" release date is assumed to be before December 31st 1982).

The article is complaining of Mattel's slow progress developing certain titles resulting in bogus projected release dates.  AD&D, Chess, Pinball, Tron were notoriously "coming soon" titles for many months or years in some cases.   It's also complaining of Mattel's geographically inconsistent distribution, which was a problem not unique to one publisher. 

 

The VGU newsletter is very clear when it talks about projected release dates versus cartidges that are in distribution.  Actual internal Mattel release dates, assuming that's what's on Intellivision Lives, should correspond to a shipping event.  A product's release date is when it first ships to a distributor, which would be well before it shows up on the shelves.  It is possible that there were shipping events prior to the dates on that list.

 

As far as retail advertising availability.   Adverts in magazines are often prepared weeks in advance.  Some of them include long price lists; they're questionable as they often include not yet released titles.  Newspaper ads can be prepared much closer to publishing and often promote specific titles sometimes saying available now.

Edited July 8, 2023 by mr_me

[Rod]

On 7/7/2023 at 4:35 AM, DZ-Jay said:

And yet ... without a clear understanding of your sources and their authority...

     I joined in "liking" this post too at first, until I realized how ironic it was we were all hiding behind pseudonyms.

     The source's name is probably on the distribution lists of memos posted at www.papaintellivision.com. I'm fine not knowing which one.

[Rod]

On 7/6/2023 at 4:33 AM, Walter Ives said:

Mattel's upper management had far too much experience and savvy to put their company into a similar position. [Insert appropriate emoji below.]

     I would have given this comment a  even without the prompt.

     (P.S.: I enjoy your Easter eggs, at least the ones I spot.)

[+DZ-Jay]

6 hours ago, Rod said:

     I joined in "liking" this post too at first, until I realized how ironic it was we were all hiding behind pseudonyms.

     The source's name is probably on the distribution lists of memos posted at www.papaintellivision.com. I'm fine not knowing which one.

 

I am not hiding behind any pseudonym.  This is a nickname I've had for decades, and even my wife calls me "DZ-Jay."  Still, my real name is attached to my work, and I've never been known to shy away from exposing it.

 

Besides, the point of my comment wasn't to "out" anybody.  It was to officially and authoritatively record Mr. Ives' comments as part of the factual history of the Intellivision.  If all this knowledge, insight, and information cannot be corroborated or verfied, it stands as nothing more than apocryphal anecdotes.

 

Now, as I also mentioned in my comment, I am fine with that -- I believe Mr. Ives to be posting truthfully and in earnest.  However, it would be nice to be able to refer to some authority (or authorities) to learn about the history of the console in its proper context, directly from the horse's mouth, as it were.

 

       -dZ.

[First Spear]

On 7/6/2023 at 7:28 AM, Walter Ives said:

 

 

That being said, GI was quite adept at producing ROMs, so those cartridges and others, including Space Battle, Hockey and Roulette, had been manufactured months earlier as scheduled and were being held pending the availability of sufficient working Master Components. Production cartridges were used both at summer CES and in-store promotional events around the country. A few even left Mattel's hands, being given as samples to major retailers.

 

 

Did Tom Mariner have a continued presence at that time?

 

[Rod]

On 10/24/2023 at 7:59 PM, First Spear said:

 

Did Tom Mariner have a continued presence at that time?

 

Mariner apparently worked for GI Microelectronics. He told an interesting story at https://www.paperblog.fr/9322830/lorsque-l-histoire-du-jeu-video-a-appele-player-one-etait-pret/. It's in French, but here's a translation,

 

What would become the standard television interface chip was, at the time, a cabinet full of non-integrated circuits, all wire-wrapped together on one-foot-square boards. Mattel sent a newly hired engineer to judge GI Micro's IC system, and we hit it off right away. He told me about Mattel's plans for a football video game, at the time a major animated graphics innovation. While the rest of the CES hardware geeks showed off their chip-filled boards, I immediately got busy saving the crude animations to the computer's memory. By breakfast the next morning, I had put together some test code – two animated runners, one light blue, one dark blue, on a green background, mimicking what I thought was Mattel's big idea. The little figures chased each other back and forth across the screen, turning at each edge and going the other way (one engineer noted that if I “put blonde hair on the one being chased, we might have a good game there.” Soon, a group of GI Micro technicians gathered and cheered on the lively little runners. The Mattel guy, impressed and excited, called his boss, explaining what he was seeing and the big reaction. Suddenly, he removed the receiver from his ear, looked at the phone with a perplexed expression, then asked sheepishly, “What?" Someone was clearly shouting on the other end, and the engineer went white as a sheet. “He says turn it off now,” he croaked. It turns out that the idea for a football video game and the animation I quickly created was the famous toy company's biggest secret. And they thought it would take months to organize such a demonstration, not a single morning. But they had forgotten to tell their man to keep quiet.

Edited January 28 by Rod

[Walter Ives]

On 7/6/2023 at 4:28 AM, Walter Ives said:

Instead, his program list included Conversational French, Physical Exercise, Nutrition, Astrology, Speed Reading and Home Budgeting, all of which we today recognize as killer aps.

I can't resist commenting on my own post: The term "killer ap" appears to mean something different when applied to the 1982 Keyboard Component. The programs listed here proved to be killer aps indeed.

WJI

[Walter Ives]

On 7/8/2023 at 8:09 AM, mr_me said:

AD&D, Chess, Pinball, Tron were notoriously "coming soon" titles for many months or years in some cases.

The "coming soon" designation was pretty much fiction. Some of the original games were teased before it was clear such games were even possible. I mean who, in 1977, could have told you with any confidence that it would be possible to create a video baseball game in which users controlled the players that was worth playing? The original Roulette, Craps and Slots cartridge was never going to happen because just the graphics storage alone would have required almost the full 4K. That applied doubly to the Keyboard software. I can just imagine Chang saying, in his heavily accented English, "Jeff! You can tease this shtuff, but you sure can't write it!"

WJI

[Walter Ives]

On 7/8/2023 at 8:09 AM, mr_me said:

The article is complaining...

The newsletter was published by a small games store run by a couple of upstarts in North Hollywood. Mattel's management was focused on moving large volumes through established mass market retailers—they just didn't have time to individually communicate with small specialty stores. The only person these guys would have been able to get the regular attention of was their local sales representative, who was too busy processing orders from other stores which weren't pressing him for such details. Furthermore, on the occasion these young whippersnappers were able to ambush someone more senior, that person would have more likely to brush them off than to give them a straight, well-researched answer that would have been of use to the competition.

WJI

[Walter Ives]

On 7/8/2023 at 8:09 AM, mr_me said:

The article is complaining of Mattel's slow progress developing certain titles resulting in bogus projected release dates.

The authors of the cited article had no clue as to how the sausage was made.

The compensation of the top marketing executives, from Rochlis through Morris, and even through Valeski, was determined by their ability to create and execute on a dream. Creating a dream requires making stuff up. Obtaining the position and funding to execute the dream requires demonstrating a demand. Teasing stuff to customers and generating excitement provides evidence that the demand exists. (At the end, when it starts to become obvious that you're not going to be delivering on your dream you have to arrange for some other company to score the coup of hiring you away before your failure becomes public knowledge and are compelled to seek refuge in Pollock, Louisiana or Bryan, Texas. (Cf.: Enron, Theranos). When interviewed later you have to either (1) claim credit if your successors managed to save the project or (2) frame your successors for fumessing it all up.

Humans build on their personal experiences. Rochlis had seen people older than teenagers getting excited playing pinball, so he incorporated pinball into the dream even before GI was selected as a chip set vendor. Rochlis went to Chang and Chang went to APh (remember, Chang and APh were thick as thieves), which did some preliminary work (no surprise here, right? At the time this was all being done under the auspices of Preliminary Design, after all, and that's what the elves in Preliminary Design did, including those elves that happened to be working at APh) which reported that the EXEC framework was inappropriate and that the game would take 8K and at least six months to finish. The "at least six months" didn't register with Rochlis, but the 8K did, so no go. Rochlis erased the game from his list. Same thing with Chess, but Chess needed some extra RAM, too.

APh was able to turn out the first eight cartridges in three months and scheduled subsequent cartridges to be done by students over their summer vacations as well. This left marketing management with the impression that all you needed to do was hire a random inexperienced body, call it a programmer and give it a cartridge name, and it would produce a compelling, saleable game in three months. But the 4K Intellivision games were tiny, under 2.5K of code if you don't include data storage for the graphics. The amount of programming effort a project takes scales as a power law, not linearly. An 8K chess game doesn't use any of that extra space for more graphics, all of the extra 4K goes toward game play. In the absence of other considerations one would reasonably expect such a game to take four or more times as long to program. For Chess one would expect more, for at the time chess algorithms were a topic of current research.

It turns out that Rochlis had written his list of games using reappearing ink, so Chess, Pinball and others like it kept popping up as "coming soon," where "soon" meant that if there was enough market interest programming would start sometime after memory prices dropped in half. Cartridges would become available six months after program completion (after the game was written and tested the object code still had to be shipped to the Far East, be scheduled for ROM production, scheduled for cartridge assembly, returned on a slow boat from China and unloaded from ships by ornery longshoremen ogres). This general practice persisted through subsequent administrations, meaning Kissell, O'Connell, Gillis and Pirner. Morris and Valeski came later and worked on a somewhat different incentive structure.

"Coming soon" also means something a little different for the buyers attending CES or Toy Fair than it does to players and reviewers, for the former are trying to figure out what they're going to be ordering for the next Christmas season. They're looking for product to begin arriving at their distribution facilities in September or October, so to them "soon" means "in seven or eight months," and then only if enough retailers sign up to make production worthwhile.

So when in 1981 Prodromou predicts that 8K cartridges will soon be affordable, O'Connell says, let's really authorize development of Chess and Pinball and put them on the price list. Nobody actually knows how long development is going to take, certainly not Prodromou or O'Connell. They could ask the programmers, but as Fred Brooks says, programmers are young and the young are optimists. By this time what was once minimally restricted preliminary design has magically become a precisely schedulable activity, so beady-eyed little Jawas from the scheduling department begin wandering up and down the cubicles aisles sowing guilt as they enter fictional numbers they thought would be acceptable to management onto goldenrod-colored spreadsheets (in days of yore spreadsheets were made of paper.)

So when O'Connell finally authorizes Pinball, he's thinking the program is going to be ready to send to be masked in three months. Never mind that any Pinball cartridge you managed to write from scratch in three months was almost certainly destined to find its way to the same southwest retirement community as the one to which Atari sent ET. On top of that, when Pinball was resurrected it was given to a recently-hired Mattel employee with a 4K budget and instructions to program it using the EXEC. Talk about being set up to fail! The EXEC was built to handle games in which all eight objects are simultaneously and independently moving and sequencing; it was not the right framework on which to build a pinball game. The fact that this would be fundamentally problematic was not communicated to the programmer and for a long time the problem was not appreciated by her management; this contributed to a need to call in the cavalry (well, more like a Mountie; for management only sent one) and an excessively long development time for that particular cartridge.

WJI

[As to Pinball, the original programmers of the resurrected version still live. Get them to give you a first-hand account.]

[Walter Ives]

On 7/8/2023 at 8:09 AM, mr_me said:

It's also complaining of Mattel's geographically inconsistent distribution, which was a problem not unique to one publisher.

Mattel's sales within the US weren't geographically uniform; it sold to retailers and distributors, and indeed to one or two specific individual buyers at each such retailer and distributor, and it was ultimately up to those individuals to decide what mix of product they wanted to carry and in which stores they wanted to offer it. Some retailers, such as JCPenney were national in scope, others such as Marshall Field served specific regions. The connections between Mattel's sales force and store buyers was dynamic and somewhat haphazard—some were long term, others created at events like CES and Toy Fair. Retailer advertising not only moved product but brought people into stores to buy other things. Faced with a limited supply of a popular product, a national retailer like Circuit City might choose to direct all of that product to a specific region so there was an adequate supply to support the advertising blitz for a particular weekend. That being said, Mattel did have regional salesmen and distributors who targeted individual retailers and their individual personalities and aggressiveness did factor into the equation.

During times of allocation higher volume regular customers had more clout than smaller intermittent ones. Pricing also factored in and at the time the referenced article was published was beginning to get rather volatile. And on top of that everyone, especially smaller individually-owned specialty shops, was trying to manage inventories in a very dynamic market.

To be really specific: the national buyer for JCPenney, Al Nilsen, was better attuned to the video game market than most. He tried to keep his inventory of different titles low enough to both be manageable on a national basis and to not overwhelm customers with too many choices. He might use his judgment to commit at the February Toy Fair in New York to purchasing a certain number of a specific title, say Frog Bog, for his entire national operation. When they proved unexpectedly popular that spring, he might decide to call Mattel and ask for more. He was a good customer and got along well with Richard Hoag*, so he'd get priority in the re-order, maybe in exchange for also featuring Mattel's Synsonics Drums in his ads; this would cause buyers at other companies who were not as quick to lose out. He still might not get all he wanted, so to maximize the value to the corporation of the inventory he could get, Nilsen might choose to concentrate it in the Boston area and advertise its availability to draw people into stores in that area. And he would promise Mattel that the promotion would occur in connection with the Father's Day weekend, which would incidentally benefit all Intellivision retailers in the area. I'm not saying this exact scenario played out—42 year old memories are not good enough for that level of detail in the absence of memory-refreshing documentation—only that it illustrates the flavor of what really went on. I've tried to be careful with my choice of words: "Al Nilsen was a JCPenney buyer better attuned to the video game market than most" vs. "he might use his judgment at Toy Fair" or "say Frog Bog."

Of course, legal kept an eagle eye on everything to ensure that the laws about treating all retailers equally were followed to the letter.

WJI

[*Nilsen and the Mattel sales staff got along so well that the sales staff induced Mattel to hire him. Nilsen can surely give you an actual story rather than a hypothetical one.]

[Walter Ives]

On 7/8/2023 at 8:09 AM, mr_me said:

As far as retail advertising availability.   Adverts in magazines are often prepared weeks in advance.  Some of them include long price lists; they're questionable as they often include not yet released titles.  Newspaper ads can be prepared much closer to publishing and often promote specific titles sometimes saying available now.

You also have to remember that this snippet was written in 1982, when cargo ships didn't leave port until they were full, human longshoremen unloaded them at their leisure and the national distribution network was generally not as efficient as it is today. Mattel Electronics used Mattel Toys' warehousing and distribution infrastructure, which was optimized to ship low-value toys to retailers by most economical means. A railroad-car-sized lot of product could be shipped to a JCPenney distribution center much more quickly than small order to a mom-and-pop specialty shop. The specialty shops could have chosen to pay more for faster delivery, but that would have cut into their profit margins.

WJI

[Walter Ives]

On 7/8/2023 at 8:09 AM, mr_me said:

Did APh program any Intellivision games that improved on the Exec 20fps?

I doubt it. With the typical game cartridge requiring over 1.5K just to hold graphics data, that left less than 2.5K for gameplay, so none of the 4K cartridges programmed by either APh or Baum's group would have had enough room for any substantial gameplay without the EXEC. The purpose of the EXEC was to save cartridge ROM—there was no other reason to use it. If ROM space wasn't an issue you could, and indeed you should, take the EXEC source, put it in your own memory space and streamline it to suit your purposes. At, say, $2.50 per ROM, marketing was loathe to approve 6K or 8K cartridges, so that didn't happen. By the time large capacity cartridge ROMs started to become economical in 1983 the EXEC had served its purpose and APh had moved on to Intellivision III.

That's not to say that there was no internal prototyping that didn't use the EXEC. Rochlis wanted a pinball game early on, but APh determined that the EXEC didn't provide a good framework for such a game and did some development work on one that didn't use it. That project was shelved when it became apparent that the game would need more memory and Krakauer was unwilling to spring for it. When Sohl was seconded to APh to write an Asteroids clone, it was clear that the game was pretty simple and would not tax the 4K ROM budget, so Zwick wrote one in parallel, in the process developing a cartridge-resident framework he called the "Fast EXEC." Rolfe might be able to tell us more about that the next time he visits this site.

APh also programmed the cassette programs. They mostly used PICSE, but PICSE was an animation engine, not a game engine. There were a few cassette games such as Super Football, Bear Run and Wheel of Fortune that had the space needed to incorporate the Fast Exec or an EXEC modified as described above but I have no information as to whether or not they did.

WJI

[Walter Ives]

On 7/8/2023 at 8:09 AM, mr_me said:

In 1979 GTE Sylvania reportedly was also selling Intellivisions in their stores in select markets on the east coast.  Do you happen to know when that started?

Although Sylvania did stick its own label on a few housings in 1979 for show-and-tell, best recollections are that it didn't begin making shipments to its own retailers until after Mattel had enough inventory to support its New York-Chicago-LA splash in the Spring of 1980.

WJI

[Walter Ives]

On 10/24/2023 at 7:59 PM, First Spear said:

Did Tom Mariner have a continued presence at that time?

Mariner was employed as a consultant by GI in Hicksville. Assuming that by "at that time" you mean 1980-1981, most of his time was spent supporting GI's speech processor, both for Mattel and for many other customers.

Also, in late 1981, he oversaw the creation of some demonstration tapes of enhanced graphics for possible next generation systems. The actual creation of the tapes was contracted out to the Computer Graphics Lab at the New York Institute of Technology's Long Island campus, about 6 miles from GI's headquarters. The NYIT computer science department had assembled this lab and rented it out for commercial use. They had a team of animators available to create content for it. These animators edited up frame-by-frame sequences of higher resolution graphics for existing games, rendered them using a frame buffer and recorded them one frame at a time onto 1" video tape using an industrial video tape recorder. This was a time-consuming process which involved displaying a frame in the framebuffer, backing up the tape a short distance, "pre-rolling" it forward and recording exactly one frame of video at the appropriate place, and repeating for each frame of the sequence. You'd then play the tape to see the animation.

WJI

[Walter Ives]

On 1/27/2024 at 4:05 PM, Rod said:

He told me about Mattel's plans for a football video game, ...

This story pretty much matches the one told at Mattel. Mariner was given flip-cards with Dave James' running man animation. There were no hand controllers yet, so the two objects he implemented moved around on the screen by themselves, with no user input.

WJI

[Walter Ives]

On 1/27/2024 at 4:05 PM, Rod said:

What would become the standard television interface chip was, at the time, a cabinet full of non-integrated circuits, ...

This is a picky point that only merits correction because someone may cite the article as authority, say for a class report or a Wikipedia entry: the wire-wrapped boards used "integrated" circuits, not "non-integrated" circuits. That is to say, the board used 7400 family and MOS memory parts, not individual transistors. It's not a mistranslation: the original article does indeed read "circuits non intégrés." Also, don't take Mariner's "one-foot-square" recollection of the boards' dimensions to be particularly precise.

WJI

[Walter Ives]

On 1/27/2024 at 4:05 PM, Rod said:

The Mattel guy, impressed and excited, called his boss, explaining what he was seeing and the big reaction. Suddenly, he removed the receiver from his ear, looked at the phone with a perplexed expression, then asked sheepishly, “What?" Someone was clearly shouting on the other end, and the engineer went white as a sheet. “He says turn it off now,” he croaked.

The "boss" was named Jeff Rochlis.

WJI

[Walter Ives]

On 1/27/2024 at 4:05 PM, Rod said:

It turns out that the idea for a football video game and the animation I quickly created was the famous toy company's biggest secret.

Biggest secret? Hardly. At the time the first handheld games were just starting to hit the market and it was not yet clear whether they would be successful. The "Electronics Division" was nothing more than a pimple on the butt of the Toy company. Of course, Mariner wasn't in a position to know this, so one can't fault his impression.

WJI

PS: I hope y'all don't mind my picky corrections. They are but pimples on the butt of a great story.

[+cmart604]

1 hour ago, Walter Ives said:

 

The authors of the cited article had no clue as to how the sausage was made.

The compensation of the top marketing executives, from Rochlis through Morris, and even through Valeski, was determined by their ability to create and execute on a dream. Creating a dream requires making stuff up. Obtaining the position and funding to execute the dream requires demonstrating a demand. Teasing stuff to customers and generating excitement provides evidence that the demand exists. (At the end, when it starts to become obvious that you're not going to be delivering on your dream you have to arrange for some other company to score the coup of hiring you away before your failure becomes public knowledge and are compelled to seek refuge in Pollock, Louisiana or Bryan, Texas. (Cf.: Enron, Theranos). When interviewed later you have to either (1) claim credit if your successors managed to save the project or (2) frame your successors for fumessing it all up.

Humans build on their personal experiences. Rochlis had seen people older than teenagers getting excited playing pinball, so he incorporated pinball into the dream even before GI was selected as a chip set vendor. Rochlis went to Chang and Chang went to APh (remember, Chang and APh were thick as thieves), which did some preliminary work (no surprise here, right? At the time this was all being done under the auspices of Preliminary Design, after all, and that's what the elves in Preliminary Design did, including those elves that happened to be working at APh) which reported that the EXEC framework was inappropriate and that the game would take 8K and at least six months to finish. The "at least six months" didn't register with Rochlis, but the 8K did, so no go. Rochlis erased the game from his list. Same thing with Chess, but Chess needed some extra RAM, too.

APh was able to turn out the first eight cartridges in three months and scheduled subsequent cartridges to be done by students over their summer vacations as well. This left marketing management with the impression that all you needed to do was hire a random inexperienced body, call it a programmer and give it a cartridge name, and it would produce a compelling, saleable game in three months. But the 4K Intellivision games were tiny, under 2.5K of code if you don't include data storage for the graphics. The amount of programming effort a project takes scales as a power law, not linearly. An 8K chess game doesn't use any of that extra space for more graphics, all of the extra 4K goes toward game play. In the absence of other considerations one would reasonably expect such a game to take four or more times as long to program. For Chess one would expect more, for at the time chess algorithms were a topic of current research.

It turns out that Rochlis had written his list of games using reappearing ink, so Chess, Pinball and others like it kept popping up as "coming soon," where "soon" meant that if there was enough market interest programming would start sometime after memory prices dropped in half. Cartridges would become available six months after program completion (after the game was written and tested the object code still had to be shipped to the Far East, be scheduled for ROM production, scheduled for cartridge assembly, returned on a slow boat from China and unloaded from ships by ornery longshoremen ogres). This general practice persisted through subsequent administrations, meaning Kissell, O'Connell, Gillis and Pirner. Morris and Valeski came later and worked on a somewhat different incentive structure.

"Coming soon" also means something a little different for the buyers attending CES or Toy Fair than it does to players and reviewers, for the former are trying to figure out what they're going to be ordering for the next Christmas season. They're looking for product to begin arriving at their distribution facilities in September or October, so to them "soon" means "in seven or eight months," and then only if enough retailers sign up to make production worthwhile.

So when in 1981 Prodromou predicts that 8K cartridges will soon be affordable, O'Connell says, let's really authorize development of Chess and Pinball and put them on the price list. Nobody actually knows how long development is going to take, certainly not Prodromou or O'Connell. They could ask the programmers, but as Fred Brooks says, programmers are young and the young are optimists. By this time what was once minimally restricted preliminary design has magically become a precisely schedulable activity, so beady-eyed little Jawas from the scheduling department begin wandering up and down the cubicles aisles sowing guilt as they enter fictional numbers they thought would be acceptable to management onto goldenrod-colored spreadsheets (in days of yore spreadsheets were made of paper.)

So when O'Connell finally authorizes Pinball, he's thinking the program is going to be ready to send to be masked in three months. Never mind that any Pinball cartridge you managed to write from scratch in three months was almost certainly destined to find its way to the same southwest retirement community as the one to which Atari sent ET. On top of that, when Pinball was resurrected it was given to a recently-hired Mattel employee with a 4K budget and instructions to program it using the EXEC. Talk about being set up to fail! The EXEC was built to handle games in which all eight objects are simultaneously and independently moving and sequencing; it was not the right framework on which to build a pinball game. The fact that this would be fundamentally problematic was not communicated to the programmer and for a long time the problem was not appreciated by her management; this contributed to a need to call in the cavalry (well, more like a Mountie; for management only sent one) and an excessively long development time for that particular cartridge.

WJI

[As to Pinball, the original programmers of the resurrected version still live. Get them to give you a first-hand account.]

@nurmix