The Insanely Great Story of the Apple Newton

[Electric Monk]

I was planning on writing a computer industry timeline, but now if I write one it will look like crap in comparison to this one...

A very interesting timeline, and the OTL updates are helpful because I don't really have much knowledge of Apple in this period.

Sorry about that .

I figured OTL updates might help. I'll try and keep throwing them even as it diverges so you can see what happened the first time and ITTL.


I'll note here that Scully being better at his job will help somewhat—especially in building a more solid foundation—but Apple is also going to both do and look better than IOTL, so the fall might be harder.

Think of the overreach inherent in stuff like the Pippin (insanely dumb Apple game console) and then imagine a richer more successful Apple's arrogance in the late 90s.


Some reason why I think Dylan (programming language) deserves to live:

The language itself is cool because it is very competently designed, has a lot of features for developing applications rapidly, and still allows for fairly efficient machine-code compilation. It's cleaner and more elegant than Java/C#, higher-level and more powerful than Python or Ruby, and fairly well-understood implementation techniques allow performance closer to C/C++.

The Apple Dylan environment was neat because it married the basic features of Lisp and Smalltalk environments, which to this day are the standard for IDEs. Apple Dylan was built on top of Macintosh Common Lisp, which gave it incremental native-code (68k and later PowerPC) compilation. This gave the environment the capability of "live code update", allowing interactive evaluation of code fragments and the ability to update the code of the application while it was running/being debugged. Even modern C++/Java environments do not properly reproduce this feature!

It was also built on a database model of source code, rather than a file-model. Instead of storing code in files, it stored individual definitions (class definitions, method definitions) in a database. Windows called "browsers" could give you different views into the code, grouping together definitions in different ways, without being limited by the fixed groupings that arise from a file-based model.

It also had a ton of creature-comforts, from very complete tools for viewing the data structures of the program while it was running to an integrated disassembler, that would let you click on a method and see the 68k or PPC code generated for it.

--- Links ----

Dylan Wiki: http://wiki.opendylan.org/
Apple Dylan Environment: http://wiki.opendylan.org/wiki/view.dsp?title=Apple Dylan (screenshots and feature lists)

 

[Electric Monk]

[OTL]
Defying Gravity: The Making of Newton
Doug Menuez, Photography & Markos Kounalakis, Text.
Beyond Words Press; Hillsboro, Oregon: 1995.

By no means, he [Gassée] assured Sakoman, would marketing people be allowed to enter and defle the holy temple of the scientific priesthood. Sakoman, aware that the option of leaving his job and developing a start-up company was fraught with its own perils and compromises, shook Gassée’s hand and closed the deal. He dusted off a code-name he had been aving for just the right project: Newton

Sakoman chose Newton not only because the original Apple Computer logo depicted the 17th century English scientist sitting beneath a tree, but mainly because he believed that “Newton shook up people’s ideas about the way things are.”

[…]

“We have to wipee the slate clean. Typing on a keyboard isn’t a natural thing. Most people never learn it and the people who do learn it don’t learn it until quite a few years into their existence. But grabbing a pencil and pointing with it and scribbling and making marks is a skill that you learn really early, and therefore it feels a lot more natural.

“Why can’t we write things the way we were taught to in school? Why can’t the computer deal with things that we’re not good at? When we draw a square, we ought to be able to look at it and say, ‘That’s a square.’ A computer ought to be more natural, it ought to be more directly manipulated, it ought to take much more of an effort to do what you meant instead of what you did. All those are very hard things.”

[…]

When the core Newton team approached Capps [Steve Capps, a brilliant programmer who had left Apple a few years earlier], he was busy building his own music software company and inventing a new guitar-synthesizer instrument. They told him about the new, great machine they wanted to develop, one with not only a pen interface but advanced communications abilities too, all in a lightweight package. Sakoman emphasized that Newton would be the next new productivity tool and that portability would be important because “tools, to be useful, have to be with you. Unless a tool’s accessible, it’s worthless.”

Capps listened to the pitch and expressed his major concern: “I don’t want to make a yuppie toy.” […] Sakoman convinced Capps that his goal was to promote pure research […] Capps was hooked.

[…]

The team brainstormed as Culbert [electrical engineer late of AT&T] asked defining questions: “Okay, we’ve got this fun computer called Macintosh. What do we do next? What is the next computing device that is going to be interesting in the industry?” In their rarefied research environment, they were asking engineering questions rather than trying to respond to consumer needs. So perhaps it was little surprise when their initial goal turned out to be an engineer’s dream and a consumer’s nightmare. In 1988, the Newton group was researching the ultimate $8,000 machine.

[…]

In 1989, the second year of the project, Sakoman spent much of his time “selling our continued existence” to the company. Gassée dictated most of Newton’s existence, but bureaucratic machinations at Apple could grind down executive decisions made by fiat. Gassée told the rest of the company simply to give Newton resources and then leave it alone. But Sakoman had to take care of the details involved in actually procuring those resources and justifying Newton’s isolation. His work became increasingly thankless as the rest of the company’s prejudice against Newton grew obvious. Employees outside Newton had become suspicious of the unknown. A scorched-earth policy of product development was the company tradition, a Jobs legacy. It was even a stated Scully goal that a modern company should make its own products obsolete. As a result, there was an unspoken fear throughout Apple that Newton would devour the Macintosh, just as the Mac had gobbled up the groundbreaking Apple II.

Sakoman tried bravely to maintain the Gassée-promised separation of Newton’s church from the Apple state. As time passed, though, he got tired of the balancing act and the struggle to protect Newton. Gassée was a loyal patron, but Sakoman wearied of trying to satisfy other reluctant benefactors at the management level while keeping his own little team unaware of conflict. As the liaison responsible for representing Newton to the rest of Apple, he had to allay the company’s fears that Newton, like the Macintosh group before, might “raise the pirate flag. But there’s a fine line between being open and spending you whole life doing demos.” He resented “the religion” at Apple, and “the religion,” as Sakoman likes to say irreverently, “is Macintosh.” To any in-house paranoid or fanatic (and there were plenty of both in those days), Sakoman’s Newton team could be seen as a cult of blasphemers aiming not merely to reform, but to revolt.

Sakoman finally revolted against the bureaucracy that had thwarted his original goal to run a genuinely independent start-ip. He quit. Gassée, for other reasons, had also had enough—and the feeling was mutual; Apple had had more than enough of him. By spring 1990, both the Newton group’s leader and its patron were gone. Uncertainty reigned. It was now open season on the freshly orphaned group.

 

[Electric Monk]

Retcon: Replace the last paragraph of the above post (and change the tag to [Adapted OTL]) with this:

Sakoman was sick of this, but didn't want to leave Newton behind and join Gassée who had recently quit. He decided on one last appeal, going to boss. He met Sculley, Apple's CEO, in March of 1990 and asked him to consider Newton. Sculley gave them a month for a demonstration to him and the board, but also promised no interference if they approved… as long as Sakoman allowed marketing types and a technical overview by Larry Tesler. Sakoman, unwilling to see Newton go down the drain, agreed.

Sakoman reluctantly took one more for the team and, along with Capps, worked hard towards the April demonstration. Prior to the April board presentation, Capps and Sakoman showed Sculley the Newton prototype. They had deliberately dimmed the lights in the conference room for dramatic effect. Capps thought the darkness of the chamber enveloped "this amazing, cool, backlit screen, and made the thing look golden. It was this magical thing." Sakoman stood there proudly holding the glowing, golden device. He handed it to Sculley. "Well, John you've had a problem understanding what the Newton was all about. Here's what we want to show you."

Sculley looked at the machine as it displayed some preprogrammed HyperCard stacks running simulations of functions he had always heard about, but was seeing now for the first time. The Newton simulated an architect's sketch pad and cleaned up the imprecise drawings, straightening lines and snapping them into symmetrical alignment. Newton was longer just a concept visualized by the mind's eye. Sculley looked up, inspired and relived that he would not have to cancel the project. It clicked.

"Okay, I get it!"

Sakoman breathed a sigh of relief, the hard part was over. Sculley was enrolled.

------

[OTL Version]

Sakoman finally revolted against the bureaucracy that had thwarted his original goal to run a genuinely independent start-ip. He quit. Gassée, for other reasons, had also had enough—and the feeling was mutual; Apple had had more than enough of him. By spring 1990, both the Newton group’s leader and its patron were gone. Uncertainty reigned. It was now open season on the freshly orphaned group.

[The above paragraph has been replaced by the fifth last of the Adapted OTL version. Stuff following that is slightly modified from the OTL book, but is basically the same]

-----

I've decided to keep Sakoman. IOTL he leaves with Gassée, ITTL's somewhat better Apple (in internal culture) he is more willing to make a last ditch attempt.

If he actually has no interference he decides that's better than the "no interference" of the previous two years, and accepts marketing types in order to get this.

Among other things Bob Herold, designer of the NewtonOS's kernel, will stay as the project (changed direction or not) will remain 'more fun'. I don't know if Herold himself matters, but it is apparent the work environment took a moderate downward turn under deadlines & Tesler. We still have both, but we also have Sakoman to counter some of that.

As both Sakoman & Larry Tesler believed in the larger (tablet sized) Newton I get no benefit there from keeping Sakoman in charge. It will still take some time before Newton group switches to the smaller Newton.

However I imagine ITTL that Tesler will love Newton as IOTL. With Sakoman still in charge I'd wager he takes over as a guiding angel (running interference with management, for one), but Sakoman stays in direct control. With any luck Newton still gets ARM, instead of the Hobbit, and Tesler pushes Dylan even harder than OTL.

Without being utterly Newton focused Tesler can work with Pink on Dylan (and thus gather momentum to save Dylan) and perhaps is seen as less tied to the Newton so he can better argue for more time (backed up by Pink/Dylan people).

------

Comments anyone?

 

[Electric Monk]

I'm trying to give some idea of what using a Newton is like, so I present the following (from this dusty corner of the Interweb—more stuff at the link):

An even better and more natural way to create a new note is a real stroke of genius: just draw a horizontal line across the current note and a new note will be created. Because the note software has a limit to the size of an individual note, this little feature makes it extremely easy to just keep on writing, note after note.

In addition to the line-across-the-note trick, there are several other gestures for editing text that make the notepad significantly easier to use than keyboard-based text entry software. For example, to join two words (or two letters) one simply draws an upside down caret below the two characters to be joined. Very natural. Want to add a space? Draw a little caret below the line with the point where you want the space. This would be awkward with a keyboard, but remember you are using a pen at this point.

There are many more gestures, really too numerous to post, that all add to the beauty and utility of the notepad. To split a line, just draw a carriage return, so to speak. Select a work? Double tap it. Want to move it, just drag it around with the pen. Want to select a whole paragraph? Draw a box around it. Convert text to all uppercase? Select it then draw a vertical line up through it's middle. Lowercase? Same action but drawn down from above the text. Perfect for a pen-based interface.


Right now I don't have any applications running except for Backdrop+. It has a feature that lets me enter a quick note. Not the notepad application per se, just a slip that let's me write a short note. Let's try "Lunch with Bob" and see what happens. I actually wrote "Lunch w/ Bob" and the Newton got it right because I had defined "w/" to be shorthand for "with".

The Newton is now displaying a slip that shows a meeting from 12:00 to 1:00. The list of invitees properly includes Bob's full name (I only have one Bob in my list of names). I can now just press the "Schedule" button or adjust the time or other facts about the meeting. I happened to do this at 9:30 this morning. Had I done this between 12:00 and 1:00 the meeting would have been scheduled in the next half hour. After 1:00 and it would have been scheduled for tomorrow (assuming I already ate lunch today). Isn't that clever?

Here's a fun Assist trick: Daydreaming about a vacation and want to know the time in some other part of the world? Just write "Time in Melbourne and tap Assist. You get the little time application showing the current time in Melbourne, Australia (1:46 am Fri 5/23 as I write this).


What makes the Newton so insanely easy to use? I'm not talking about whether there are keyboard shortcuts for functions, or whether the Newton's arrow keys are better than scrollbars, or whether the Newton interface is "intuitive". I'm talking about the whole Newton experience. Why is it that the Newton is easier to use for day to day activities than any other computer on the planet?

The single best feature which contributes to this is the instant on of the hardware.

As an example, suppose I need to jot down a quick note; perhaps a ATM withdrawal or a phone number. Open the lid, press the power switch, and faster than I can grab the pen the device is waiting for my input. It's amazing what this does for makeing the device so accessible. I think even waiting 5 seconds would diminish the utility of a handheld, so unless it's on instantly it's just not good enough. I've never once thought "Nah, I won't enter it in the Newton because it's just too much trouble". It's never too much trouble to enter data into the Newton.

What else makes the Newton so easy to integrate into daily activities? The filing system. It's so transparent it's almost scary. I'd wager it's darn near impossible to lose data, *any* data, from a Newton. Everything is automatically saved. Writing a note at a traffic light? When the light turns green just flip the power switch off and drive. Next light, turn the Newton on and you are right where you left off. This is absolutely wonderful, and part of what makes the Newton a Newton.

 

[Electric Monk]

Oof. Let me point out I've now managed to misspell John Sculley's name as John Scully about a million times. Whoops.

I can point out that in the 1990's there was an Apple employee with the unfortunate name of John Scull .


Ok. A sneak peak, at a decision that's going to alter the Mac:

In Sculley’s most fateful decision he directly bucks Jobs, and instructs that the Mac will ship with 512k of memory[1] (OTL 128k), that its price will remain $1,995 (OTL $2,495), and that Apple will reduce their margin target to 30% in order to engage IBM over market share. IOTL, like TTL, he is unwilling to override the ‘experts’ on moving Apple to the Intel x86 IBM compatible platform, and go head to head with IBM/Microsoft on the OS/software front. However he does know market share, and decides to go after IBM as hard as possible. This demands both a perfect Mac launch, and the sub-2K price. Hence a Mac with more memory but at the same price point, achieved by sacrificing margins.

IOTL Sculley was focused on selling the company to someone (even IBM), ITTL he decides to engage IBM & focus on strategic acquisitions (his back-up plan, which he ignored IOTL). Notably Compaq could have been bought for $100 million then as it very cash poor and needed support, and even came to Apple.


[1] Amusingly one of the engineers designed the Macintosh to hold 512k of memory by end running around Jobs, so the Fat Mac that shipped with 512k a little while later didn't require a redesign or anything, just a few extra memory chips.



Comments? Questions? Responses?

 

[Nicole]

Very interesting- I suppose the whole Apple Lisa thing is unaffected?

 

[Electric Monk]

Very interesting- I suppose the whole Apple Lisa thing is unaffected?

The Lisa was heading out the door as Sculley came in. The money was sunk, the advertising was sunk (badly: they used lifestyle ads to sell to business) and the (over) production capacity was set. Sculley can't really change that.

The Lisa was also expensive (20k in today's USD), slow, and IBM PCs were coming out at the same time that were much cheaper despite the fact they were still running DOS while Lisa had a GUI.

I don't think I can salvage Lisa, I'm afraid, but I've been wondering if it would be possible to port Lisa's programs to the Mac—or at least provide for the use of Lisa file formats, and then upgrade some Mac programs to use Lisa's features. At the least that would provide a transition strategy for the people who did buy Lisa (according to Wikipedia NASA used them for project management) and keep them more loyal to Apple.

In Sculley’s most fateful decision he directly bucks Jobs, and instructs that the Mac will ship with 512k of memory[1] (OTL 128k), that its price will remain $1,995 (OTL $2,495), and that Apple will reduce their margin target to 30% in order to engage IBM over market share.

I'm responding to myself, yes, so I can add a little more information. IOTL Sculley understood that high margins = high profits = high share price = hot commodity for someone to buy.

ITTL he doesn't want to sell, he wants to build and compete. Therefore, as he also understands marketshare, he cuts margins (overriding Gassée in the process) and accepts the $100 million Mac marketing campaign as a one-time brand awareness type deal rather than directly paying for it (with the $500 increase) out of units sold.

 

[danielb1]

One possibility: what if, in the early 1980s, Apple had made an effort at ensuring Mac and Lisa compatibility before they were launched?

 

[Electric Monk]

One possibility: what if, in the early 1980s, Apple had made an effort at ensuring Mac and Lisa compatibility before they were launched?

They were completely different operating systems, I'm afraid, and the teams hated each other.

So, not much chance I think and my POD of a slightly different John Sculley doesn't change that.

It would have to be after the fact to get the Lisa base to upgrade onto Mac. Same problem with Apple ][/III as the Mac really was a completely different system that just happened to offered by the same company.

 

[Nicole]

Well, arguably the Macintosh XL and MacWorks could be seen as an OTL Lisa-Macintosh "bridge"...

 

[Electric Monk]

Well, arguably the Macintosh XL and MacWorks could be seen as an OTL Lisa-Macintosh "bridge"...

It was more of a way for LisaOS to run Mac programs, when it needed to be the other way around (plus they sold regular Lisa's at the same time).

Like most things Apple when they stop paying attention (operating systems from 1990-6, Newton post-launch, Mac industrial design, etc.) it was kind of half-hearted.

Heck they even disliked the Apple II despite the fact it had financed the Apple III/Lisa/Mac programs, the slow sales experienced by them all, and all the other going on things at Apple.

 

[Electric Monk]

Can anyone suggest a good name for Pink/Taligent when it's released?

In OTL MacOS went from System 1 to System 7, then MacOS 8 to 9, and finally OS X (ten, not x). Codenames were Pink/Taligent, then American composers or derived names (i.e. Copland & Gershwin, Rhapsody) and then big cat names (Puma, Leopard, etc…)

ITTL I imagine a System 8 of some sort will be released in 1993 or so, but in 1995-6 so will Taligent. I could call it MacOS 9, but that doesn't really break hard enough like OS X does. OS IX doesn't have quite the same appeal.

Most notably is that Taligent approaches the entire operating system metaphor entirely different from anything else, so I don't feel sticking with the System or MacOS numbering scheme feels right. I'm obviously not going to adopt MacOS 1996 like Windows, either .

Perhaps MacOS 'name' but I'm not sure.

So ideas? Either with a decent way to phrase it, or a name for the MacOS 'name' slot. Thanks in advance (also, if you comment about the timeline I'll be your friend ).

 

[Electric Monk]

I've had an interesting thought.

IOTL Steve Jobs leaves Apple shortly before the desktop publishing revolution (sparked by Aldus Pagemaker, Adobe PostScript, and Apple's LaserWriter & AppleTalk networking) but it appears that he would have been handed that if had just toughed it out for a couple months instead of (basically) being a big baby and mounting a poorly though out coup.

Steve Jobs is almost certainly the perfect man to push desktop publishing and establish utter Mac dominance in the field.

Let's say he has a slightly better relationship with our slightly different Sculley. Or at least butterflies mean he's going to be there slightly longer.

He gets handed desktop publishing and probably does a fantastic job, especially given our moderately higher Mac marketshare (from the lower price & 512k of memory).

Now he's certainly still going to leave Apple (I can't imagine him staying much longer than OTL, even diverted by desktop publishing) however does he still establish NeXt? Perhaps not.

If we're trying to set up an alternative in computing to Windows & MacOS than there's plenty of options—but everything remained too fragmented.


So what does Steve Jobs want? He wants to show up Apple for kicking him out of the Mac and ITTL he probably gets away without the non-compete clause in the deal he signs with Apple.

What are his options?

He can software, hardware, or both. He loves hardware design, certainly, but he also understands that software is important. This would tend towards a both option.

However OTL's both option, NeXT, failed utterly despite some pretty hardware. However NeXT did make an excellent programming environment (Openstep) and a great web developing platform (WebObjects).

At this time we have:

• Commodore Amiga, about to make their last foray in being a factor in the computing world. They have a pretty good operating system, and pretty computers.
• In a year or two Gassée will start up Be Inc. and make BeOS, a better operating system than NeXTSTEP ever was.
• Sun will end up using Openstep on Solaris/SPARC in the mid-90s OTL.
• Both IBM & Intel would like an operating system. IBM to move PowerPC chips, Intel so they aren't stuck with Windows and no competition.
• This is RISC chips (in the form of SPARC, POWER, MIPS, and so forth) last chance to shine below the workstation level but they're facing Intel's 90% dominance of the market.

If we wanted to bootstrap BeOS the AmigaOS isn't a bad way to go about it, with Openstep (the programming environment by NeXT, as BeOS was hard to program for) for programming and have them run on Commodore boxes using PowerPC? chips

Both Commodore Amiga's, because of speciality hardware, and BeOS, because of software, were good in the high performance video/graphics field and yet Commodore Amiga's were some of the cheaper boxes around.

If anyone wants to offer up combination suggestions to get a third computing platform in the 90s, I'm listening. (If AmigaOS/BeOS can use Solaris or BSD or something as their base for future development we could even co-opt much of OTL's Linux movement onto them like OS X has in OTL, and Pink won't ITTL.)

 

[Electric Monk]

We’re taking a slight sidetrack from Apple & Newton to cover Steve Jobs and what he does.

[Mixed: ATL & Adapted OTL]
Infinite Loop: Behind the Legend, The (Incomplete) True Story of of Apple Computer, Inc.
Michael S. Malone.
Subtext/Doubleday, New York: 2002

Steve Jobs had been on the verge of leaving Apple for months. Sculley had forced him out of the Mac group, and it was clear that Jobs—the founder of Apple—had no place left at his company. It was never clear why Jobs stayed on. He mounted an aborted attempt at a coup, dying before it came out into the open. He tried to muster support to remove Sculley and replace with somebody else, but that didn’t go anywhere.

However his heart didn’t seem to be it, perhaps his last vestiges of a friendship with Sculley and others at Apple keeping him a little calmer then he had been. Luckily for Apple, the Mac, and Steve Jobs himself he did stay.

You see, the desktop publishing revolution was about to start.

Although the Macintosh had done fairly well in sales Sculley and his team had reached the terrifying conclusion that the Mac dominated no markets, it had no place to call home. So when Paul Brainerd brought his idea for desktop publishing to Apple (championed by a young Harverd MBA with the bad luck to be named John Scull) he was warmly welcomed, and his company, Aldus, soon had their Pagemaker program heading towards stores.

Combined with Apple’s LaserWriter printer & the AppleTalk network Macs could now run a page layout program and Apple clearly saw what it had, throwing advertising dollars behind Aldus as they tried to draw distinctions between the IBM PC and the Mac.

There was a third player in this nexus. Adobe. Founded by ex-Xerox PARC researchers led by John Warnock, who just happened to be one of the very few people Steve Jobs looked up to. Warnock had invented Imterpress a Xerox laser printer compatible programming language. Xerox had kept it’s perfect streak in making stupid decisions and had spent years hemming and hawing. Warnock finally had enough and quit.

Adobe used Warnock’s understanding of graphics to turn fonts into mathematical equations, allowing computers to pick from scores of typefaces, adjust their size, and switch to bold or italitc at the touch of a key. It was revolutionary in the time of a single typeface in a single size that could only be capitalized or underlined.

Combined with Pagemaker and Apple’s LaserWriter printers Apple finally had something to call their own—especially when buyers realized that hooking a dozen computers up to one LaserWriter using AppleTalk made everything affordable.

Apple, recognizing a good thing, bought 30% of Adobe for $5 million dollars. It was Steve Jobs second last great act, and a reminder just how valuable Jobs could be.

Steve Jobs’s last great act at Apple was taking over the desktop publishing division and bringing it to everyone: from news divisions to small business, from the world’s largest corporations to individual graphic artists; Apple would completely dominate the area of desktop publishing thanks to Jobs.


One person who did understand the import of desktop publishing was Bill Gates, and it scared him every way but witless. As with his arm’s-length partner at Intel, Andy Groove, one of Gate’s greatest skills was that he responded to every real or imagined attack with an instant, all-out counteroffensive. He never allowed his more inventive competitors to consolidate their gains.

He had been forewarned and forearmed with VisiOn—a graphical user interface (GUI) for IBM PCs in November of 1983. VisiOn was a disaster in everything from design, to alienating developers, but it kicked Gates into gear.

It would take two years to release Windows, Microsoft’s own GUI, so in the meantime he announced that Microsoft would meet VisiOn with their own GUI which shut down VisiOn’s sales and gave him the time to release Windows.



[ATL]
Excerpted:
Bits & Bytes: News From The Valley
Computing Worldwide, Volume III, Issue 7: July 1989

In a surprising move last Wednesday rumor swirled around the Valley that Steve Jobs’s company, Next Inc., was under consideration for sale. Speculation centred around Sun Microsystems known to be considering the use of Next’s OpenStep software development platform.

Next has not yet finished their “revolutionary” (their press materials) operating system, NextStep, or its companion programming development environment, OpenStep. However anticipation was high, and besides Sun Microsystems a number of companies were considering NextStep and OpenStep—such as IBM.


Excerpted:
Bits & Bytes: News From The Valley
Computing Worldwide, Volume III, Issue 8: August 1989

Commodore has confirmed that they’ve bought Next Inc. for an undisclosed price. Their stated primary purpose is to establish sole control over their next generation operation system, NextStep. They had previously considered licensing it, but had decided instead to buy Next outright.

Commodore will reorganize operations, bringing all in-house software development into the Next Group run by Steve Jobs. Currently unknown is the fate of WorkBench, the Amiga operating system.

Apple, Sun Microsystems, and Microsoft had no comment when we went to print.

No further information is currently available. This reporter might speculate that Steve Jobs has decided he needed a larger company for whatever his future plans are, but that could be going too far.


Excerpted:
Bits & Bytes: News From The Valley
Computing Worldwide, Volume III, Issue 12: December 1989

Commodore has announced that NextStep will be able to emulate programs from both Commodore Amiga’s operating system, WorkBench, and the Commodore 64. It appears that Commodore has finally had some sense shot into their head. Computing Worldwide congratulates them.

This reporter never speculates on future hardware but he imagines that this sets the stage for Commodore to reorganize and clarify their product lines.


Excerpted:
Bits & Bytes: News From The Valley
Computing Worldwide, Volume IV, Issue 1: January 1990

In today’s unlikely news from the Valley Atari & Commodore have announced their intended merger. I wager this makes a great deal of sense given that they’re both competing in the same home computer / console / game space at the low end, but I’d never have thought that these two companies could actually get together.

Well.

Never underestimate Steve Jobs, is the lesson everybody must always have drilled into their skull. Perhaps a tattoo, on the arm, for the sake of convenience.

Jack Tramiel is once again CEO of Commodore (as the companies will stay under the Commodore name). Steve Jobs has taken over computer hardware and software design, moving up from just software.

Atari will refocus on consoles, like their current Atari Lynx, and will absorb Commodore efforts in that direction. The Amiga line will incorporate the Atari STs.


Excerpted:
Bits & Bytes: News From The Valley
Computing Worldwide, Volume IV, Issue 6: June 1990

In what this reporter regards as truly stunning speed Commodore has announced their new line-up. It seems like Commodore has pulled a rabbit out of their hat. Let’s go through it all, shall we? (Yes, we usually don’t do this in this column, but I have a certain feeling about this whole thing)

In the holding pattern is the Commodore 256. It works with all C64 & C128 programs, and sells for $495. It’s basically the same old Commodore. I imagine that the company is basically printing money off of these.

From their Atari group we have the Atari Jaguar, obviously positioned to complement their handheld Lynx system and to compete with the Sega Genesis and the forthcoming Super Nintendo. It goes for $195 and ships with a couple games and a controller. Assuming Nintendo doesn’t run them over it looks pretty good.

Upscale a little bit is the Commodore Amiga at $695. This guy handles both Commodore 64/128/256 programs, plus Atari Jaguar games (interesting…) and runs a new version of AmigaOS. I would hazard a guess that this guy is a placeholder.

I suppose you can see the hand of Steve Jobs a little, forcing simplicity. Give him a minute though.

More importantly are Commodore’s two NextStep based computers, running on Motorola’s 68k chips. We have the NextBox for $1,995 (not coincidentally the original price of the Macintosh) and the NextCube for $3,995. They compare well against their equivalent Macintosh models (and have better operating system internals—though people tell me MacOS remains easier to use than NextStep) and against IBM clones. The NextCube, especially, is stunning in black.

Sure, you may ask, why the heck is Commodore (of all corporations) making these machines? Steve Jobs, basically, and let’s see if he can pull it off. He’s got a large marketing budget, people may have told me.


In other (entirely unrelated news) Motorola & Sun Microsystems have announced they are working on a new chip design. Loosely based on Sun’s SPARC and the Motorola 88000 series the project is currently named SparcLite (cute).

This is, as far as this reporter can see, implausible at best. What do they want from each other? Nothing. No terms were disclosed, nobody has said anything, and this reporter is getting somewhat disturbed by the truly bizarre things happening in the Valley.

Perhaps we should all watch Commodore. Scrap that, just watch Steve Jobs. Oh. And you might want to note that little birdies have told me NextStep requires you to develop CPU independent programs.

I wonder what Bill Gates is up to?


Excerpted:
Bits & Bytes: News From The Valley
Computing Worldwide, Volume V, Issue 1: January 1991

Well Steve Jobs has done the impossible. He merged Commodore and Atari, released the #2 console (the Atari Jaguar trails the Sega Genesis marginally, and both beat the NES), boosted sales of the Atari Lynx (close on Nintendo’s Gameboy) and has had a very good season in computers.

The Commodore 128 has been pushed into developing markets and sources say is getting great sales in places like Russia. The Commodore Amiga has doubled its sales (mostly from incorporating the Atari ST) and has become the gaming box of computers.

The NextBox & NextCube struggle in North America against the more entrenched PC & Mac markets, but have seen very good sales in Europe. Notably, however, is that they’ve taken over the graphics/video niche of the old Amiga 2000 and people seem to love developing for NextStep.


In other (entirely unrelated news) Sun Microsystems has announced that they’ve licensed Commodore’s OpenStep development environment. I couldn’t even guess as to what they need it for.


In other (even more entirely unrelated news) Sun Microsystems & Motorola have announced that IBM has joined them on their SparcLite project and features from IBM’s POWER will make it in.

In this reporter’s opinion this appears to be a crazy attempt to take on Intel’s x86 chips in the consumer space. But where could they possibly get an operating system?


Excerpted:
Bits & Bytes: News From The Valley
Computing Worldwide, Volume V, Issue 10: October 1991

In a move that shocks no one Sun Microsystems and Commodore are entering into merger talks. At the heart of it is Sun Microsystems desire for NextStep and a need to make enough machines for the Sun/Motorola SparcLite project to work.

Commodore is a large company but has persistent cash flow problems. Sun is a very large company, but isolated in the workstation/server market. I’m not sure what, if anything, they’re going to do with Atari.

In other news the next round of Commodore computers have hit the shelves. Although they remain the same price both the NextBox and the NextCube are a little bit faster, and both now have the ability to play Atari Jaguar games (previously a $100 add-on). Early data says sales are picking up.

From all accounts both the Commodore 128 and Commodore Amiga continue to sell fairly well, essentially owning much of the low-end market.


Excerpted:
Bits & Bytes: News From The Valley
Computing Worldwide, Volume VI, Issue 9: September 1992

For those of you following along this has been a crazy trip Steve Jobs has been on. He started Apple Computer, made the Macintosh, created the desktop publishing revolution and then left Apple. He founded Next Inc., made NextStep and OpenStep (an operating system better than Windows and, if a bit lacking in user interface, much more solid internals than MacOS paired with a great programming environment) and then sold Next to Commodore, going along for the ride. He promptly got Commodore to merge with arch-rival Atari to consolidate the low-end market, and then released a successful new game console and new computers running NextStep.

Meanwhile Sun and Motorola worked on a new chip design based off the SPARC & 88000 series, and then were joined by IBM who brought some of their POWER features into SparcLite, as the new chip design is known. Sure IBM & Sun working together is weird, but they like each other more than they like Microsoft and Intel.

Finally Commodore and Sun Microsystems merged and (today) released the new $9,995 workstation NextStation. Along with a new iteration of the NextBox and NextCube. Faster, stronger, able to leap tall buildings… wait. Just faster.

Even the Atari unit is doing well, with the Jaguar locked in a three-way race with the Super Nintendo and the Sega Genesis. Atari Lynx, the handheld, is far back in second place to the Nintendo Gameboy but is still making a good deal of money.

Software talk from the company is that NextStep is incorporating the features it likes from SunOS and will soon be supporting both SPARC and (whenever it gets done) SparcLite.

If I were Bill Gates what would be the one thing I’d fear most? A development environment that lets one create programs for multiple CPUs and multiple operating systems (IBM compatibility would cease to matter for those programs). Who has such a thing? Steve Jobs with OpenStep. Sources tell me OpenStep has recently been ported to x86, and currently works on NextStep, SunOS, and… Windows.




Comments & Questions?

 

[Nicole]

Fascinating... Though I wonder about the ATL Amiga... it seems like a more standard computer in the ATL, as opposed to the OTL's highly specialized chipsets... Is that an accurate guess at it?

In any case, quite interesting, I find Commodore and Atari very interesting companies and was wondering what would happen to them in TTL.

 

[Redem]

So do we have to endure the "trendy" image mac keep trying to give itself in this TL too?

 

[Electric Monk]

Fascinating... Though I wonder about the ATL Amiga... it seems like a more standard computer in the ATL, as opposed to the OTL's highly specialized chipsets... Is that an accurate guess at it?

IOTL the Amiga was either a games box (on the low end) because of price for capability (colour, bit-depth, sound, etc…) or a video/graphics box (on the high end) because of its specialized hardware chipset.

ITTL Steve Jobs reorients the Amiga solely as the low-end games box (with the added benefit of playing Jaguar games, and doing basic computer tasks). Therefore if your kids want a console you can buy them an Atari Jaguar. If you also want a cheap computer you step up a few hundred to the Amiga.

The higher-end Amiga spot (the 2000/3000 of OTL) is taken over by the NextBox running NextStep, though it retains some of the features the OTL Amiga in hardware chips (as the Next hardware also did IOTL, in a different way) and continues to dominate the video/graphics niche. However with Jobs marketing it, and with the shiny new NextStep OS the NextBox has taken some PC marketshare and sold quite a bit more than the OTL Amiga 3000.

If I recall correctly Apple sold around 1.3 million Macs in 1990. The Amiga 3000, according to wikipedia, sold a million that year. ITTL Apple will probably sell around 2 million Macs, and the NextBox (first edition) will sell around 1.5 million.

NextStep's user interface is OTL's user interface—which is better than Windows, but worse than Macs. However in the internals like multitasking and so forth NextStep is ahead of both Windows & MacOS so they're attracting a mix of higher end PC people for the interface/ease-of-use (not a Mac toy) and Unix people for the internals outside of their old video/graphics niche.

Think of OTL Unix geeks moving to OS X because it's much more usable and still has BSD/Unix underpinnings.

In any case, quite interesting, I find Commodore and Atari very interesting companies and was wondering what would happen to them in TTL.

I wanted a way to create a third computing platform. Commodore Amiga and the Atari ST are the only halfway decent selling options. Combine them and you have a pretty good platform. Add in a proper console release to compete with Sega Genesis/Super Nintendo in 1990 (instead of 1993 and being edged between 16 and 32 bit consoles) and that offers a way to separate Commodore computers from other peoples computers.

Steve Jobs is crazy and unpredictable enough that I can use him for anything. ITTL he decides that the only way he can create his dream computer is to have a platform big enough to do so. That means taking on Microsoft. The only way to engage Microsoft is to do so broadly.

• NextStep: A very good operating system, better than Windows.
• OpenStep: A software development environment that can work on multiple platforms, so that people can write for both NextStep and Windows with the same program. It also is good enough to get people to use it.
• SparcLite: Jobs would be fine with x86, but using SparcLite is the only way to get Sun/Motorola/IBM to support him.
• More SparcLite: It lets Sun/IBM have a processor that they'll sell lots of, with features from SPARC/POWER so it helps their high-end development.
• Commodore Sun: Commodore needs a sales force to break into business and out of games/video/graphics. Sun has a very good sales force and can match up Commodore computers with Sun servers. Commodore also has a history of having chip/computer in the same corporation, so this is comfortable for their company culture.
• Sun Commodore: Sun needs an operating system for SparcLite, and would prefer to have it in-house. SunOS isn't terribly suitable for computers (but, interestingly, NextStep could move upwards to become a server OS). Sun also would love to have an awesome programming environment (OpenStep) to beat Microsoft with.
• Motorola: Will get to make NextStep clones, and gets free design work for SparcLite.
• IBM: Will get to make NextStep clones, and will get to use OpenStep to make Windows/NextStep programs with one programming environment. Also it hurts Microsoft.
• Microsoft: Is confronted by Commodore/Sun, Motorola, and IBM NextStep clones with IBM/Sun sales forces to break NextStep clones into business, and OpenStep to break the IBM/Windows compatible stranglehold in business.

At the end of the day using OpenStep a company can make programs that work on both Windows & NextStep, and NextStep is the better operating system on (for the moment) the much faster SparcLite chips.

 

[Electric Monk]

So do we have to endure the "trendy" image mac keep trying to give itself in this TL too?

What do you mean by that? The way Apple advertises OS X & the Mac (the Think Different/Switcher/Mac Vs. PC campaigns)?

ITTL the Mac will retain somewhat higher marketshare throughout and because of dominance in desktop publishing it will be more widely used in business (at least when they don't "require" IBM compatible computers).

Furthermore with NextStep clones as a viable platform MacOS won't be seen as the "other" platform, and will adopt somewhat different positioning as a result.

ITTL marketshare of personal computers (I exclude, say, ATM boxes that might be running Windows) might be something like this:

• MacOS 15%
  Oriented towards the consumer, owns desktop publishing, strong in education. Is the easiest to use.
• NextStep 15%
  Consumer/business mix, with a strong hold on the video/graphics market & probably the best gaming platform because of Atari console compatibility on the consumer boxes.
• Windows 60%
  Business because of legacy IBM compatible, and cheapness of IBM clones, some consumer because it's typically cheaper than MacOS.

Note that ATL Windows is much better than OTL Windows because of increased competition, though both MacOS and NextStep remain better operating systems in different ways.


Will Apple still do trendy advertising? Probably, it's in the company DNA, but it won't be the smug (IMO) type of OTL. (I loved Think Different, myself, but don't really like either the Switch or the Mac Vs. PC campaigns.)

 

[Electric Monk]

I hope people are still enjoying the timeline, even with the somewhat implausible rise of Commodore & NextStep (but not impossible . Rest assured I had to put together just about every anti-Microsoft company (except Intel…) in the valley, just to give NextStep a chance at winning marketshare.

Next up is stuff about Pink (I still need a name for it, hint hint) and operating system talk. A few things are given away about the Newton, but I've been pretty careful to not actually give you an idea of how Apple & the industry is doing outside of broad stuff.

After this I have a number of Newton posts to cover the development of the Newton in more detail from 1991 to 1996, and I still have to cover Apple history from roughly 1985 to 1996. Apple history stuff will also cover NextStep, Sun Commodore, Windows, and so forth in passing. Finally I plan on using the Bits & Bytes columns to cover the rest of the industry and an 'at-the-time' look at events because the books are all years after everything occurred, and the Bits & Bytes column is within a month of events so things will look different.

Once everything hits 1996 I have to sit down again and figure out where things are going. Up to 1996 I have a pretty good idea, overall, past that I don't know.

One small retcon: Defying Gravity is now published in 1996, not 1995. (So much cleaning up work will be required for Timelines & Scenarios.)

There's not that many possibilities, but the name of the person giving the Macworld speech at the end has been left out . If you like guessing at things, Sculley's leaving Apple in 1992 for one of the several jobs he was up for….

Please speak up if you happen to be enjoying the timeline, if you'd like anything to be made more clear, if you want to suggest things, if you have questions, if you have comments, heck I'll even take queries!

 

[Electric Monk]

This one is particularly kind of a rough draft at the moment so feedback would be nice, but here we go anyway:

[ATL]
Apple Confidential 2.0: The Definitive History of the World’s Most Colorful Company
Owen W. Linzmayer
No Starch Publishing, San Francisco: 2007.

Operating Systems Timeline, Part II: 1987-1996

1987
A meeting is held. Blue, Pink, and Red (pinker than pink) index cards are used. Blue is for System 7, what short-term improvements can be made to MacOS. Pink is for the longer term improvements they’d like to do to the core architecture such as preemptive multitasking, protected memory, and the like. Red is for the crazy stuff, like voice command.

The sprawling Apple research labs also have other projects underway in the Advanced Technology Group. Therefore the decision is made to set up Pink as a mix of research lab and clearing house. They will begin work on Pink itself as well as evaluating ongoing research projects to see what should be made part of Pink. Most notably Blue will retain first call on resources, and Apple engineers are strictly prohibited from jumping ship to Pink.

Around this time the Apple File System (AFS) team starts up at the Advanced Technology Group. Their goal was to study the current state of files systems, identify what needs fixing, what users want out of a file system, what an advanced operating system needs out of a file system, and then build a new file system. A tall goal, but the combination of Jean-Louis Gassée & John Sculley means vast vaguely defined research projects with lots of money.

NextStep also start development at this time now that Steve Jobs has left Apple for Next. Object oriented but based on BSD & the Mach microkernelt combined with an innovative new programming language (objective-C) and development environment (OpenStep).


1988
Pink, like Next’s NextStep, is on board with the object oriented model. However they will be object oriented from the kernel up, and will start with a clean fresh sheet on the entire operating system. This results in them essentially handing off to Blue group any of their work for general improvements, and Pink refocuses on its goal as an entirely new operating system

In a new twist for Apple Jean-Louis Gassée establishes the Developer Outreach group. This group is to work as much as possible with developers to improve MacOS, as well provide cover for when Apple chooses to include programs with MacOS (as Claris, a software subsidiary of Apple, is heading for independence Apple will no longer be able to hand-off applications to them). Apple’s relations with third parties, historically frosty, will gradually improve following this move.

The Apple File System group is progressing, but inventing a new file system takes a lot of time and/or resources and so they’re smart enough to move outside Apple. Once they get permission they settle on Plan 9 From Bell Labs, a distributed operating system designed to replace Unix, as Plan 9 is working on a filing system. Apple and AT&T (owner of Bell Labs) have had the occasional talks and this provides enough of a bridge. The Plan 9 file system group establishes collaborative links to the AFS group. AFS’s goals include speed and data integrity, Plan 9’s are centred around communication and ‘everything is a file’.

Notably Plan 9 attempts to remove the difference between local & remote files (except for latency) and treats everything as a file (including networking, graphics and the like). Plan 9 is a communications focused file system. Among other things is an interesting security model: the user owns his local machine. Privilege to use other things, such as printers or other people’s computers, is done through the concept of groups. In a lucky stroke for Apple this will end up meshing well with Pink’s user interface metaphor, and the Open Collaboration Environment.

Windows NT starts development at this time. Unlike Pink & NextStep it is intended to have Unix comparable features (multitasking, processor independence, and the like) and serve as the future of Windows after Windows 3 or 4 as DOS is not suitable for a modern operating system.


1989
By 1989 the Blue group is progressing well on what will become System 7. Blue has carefully been kept as the priority and Pink remains fairly small. With the release of System 6, Blue group officially becomes the MacOS group and picks up most of System 6’s engineers. They also lose a number of their best people to Pink, as Pink group is now exiting start-up mode.

Pink has finalized their goals as a new microkernel based OS, with a 90% compatibility layer for System 7. With the benefit of hindsight they will indeed incorporate their plan to investigate user interface concepts and explore active assistance (courtesy of Newton group). Their 1993 planned release is reasonable at the time, and perhaps if they hadn’t widened their scope they might have reached it. They start work on the microkernel, named NuKernel, having abandoned their old preliminary choice of Mach. Notably they intend to follow a strict minimalism approach to get by Mach’s performance limitations[2]. This will eventually make for an excellent microkernel.

The innovative Dylan language & developing environment has been underway in development for only a short while at this point (under the codename Ralph) but both Pink and Newton groups identify it as something they want. Therefore the Pink and Newton agree to share Dylan, which will lead to increased co-operation between the two groups.

Sometime in the summer Apple’s Open Collaboration Environment (OCE)[1] begins development. As this is the height of Sculley’s ‘don’t bother me with technology, as long as I see results’ era almost any mid-level manager can start a project under the aegis of Apple’s Advanced Technology Group. OCE starts with a vast overview of the current email system, and decide that users want to move everything around between computers, not just email. They conclude a more general ‘post office’ metaphor is required with users having a single inbox and global address book and able to send files and the like without needing email.

Outside Apple at a little company called Next, Steve Jobs’s design for a new operating system, NextStep, and it’s accompanying programming environment, OpenStep, is moving along nicely and Jobs begins to put together the right partners for it. The MacOS is about to get a wake-up call.


1990
MacOS group is heading for the release stretch for System 7, and begin thinking about what they need to do for System 8. They plan for a two year release cycle for System 8, that is 1993, after the four years it will have taken to get System 7 out the door.

Pink spends 1990 writing NuKernel, and continuing to look around at Apple projects for items to include in Pink. Notably they target a fairly wide swath of talent for recruitment to Pink, both outside and inside Apple.

Jaguar and Cognac projects started. Jaguar is porting MacOS to RISC chips (the Motorola 88000 series, in this case) without maintaining backward compatibly to build a workstation type computer while Cognac seeks to maintain backward compatibly. Cognac starts up sometime after Jaguar, because they view abandoning the legacy of Mac apps as crazy. Apple’s ability to attract great engineers, and their reservoir of talent, means they can do both at the same time

QuickDraw GX[3] begins development. It is intended to replace the QuickDraw 2D rendering system with advanced features such as resolution independence, and an object oriented retained mode framework making it much easier to program for. Although this project will swerve a lot in its history, it will eventually wind up an important part of Pink. Comparisons to NextStep’s Display PostScript usually note the eventual outcome is a very similar system.

In what will be a related development TrueType[3] is started. It is intended to handle the computer-to-screen and computer-to-printer translation of fonts in a much superior fashion to PostScript by creating new TrueType fonts, as well as escaping the expensive license required to use PostScript fonts. In the end, of course, this will mostly be used to break Adobe’s monopoly pricing of fonts & PostScript.

With the release of Commodore’s NextStep Microsoft announce ‘Cairo’[4] their own object oriented next generation operating system. The Pink group is mildly chagrined by NextStep but reason that it’s the same MacOS/Windows metaphor as always just built better. Pink is given somewhat more resources, though care is taken not to slow down Blue’s System 7. Pink decides to devote more effort to a new metaphor for the computer. They now believe the only way to truly compete is to totally shift the user interface model.

Windows 3.0 is released, sales are quite good but the availability of NextStep on Commodore machines (widely viewed as superior) slows adoption on new computer purchases. The legacy base will, however, upgrade faithfully.

Microsoft is disappointed, and will devote further resources to Windows NT. Windows NT has been in development since 1988 and is not Cairo. Unlike object oriented NextStep or object oriented/new user interface metaphor Pink, Windows NT has been mostly designed simply to be a modern much more solid non-DOS based version of Windows. This will begin to change in early 1991.


1991
Pink & Newton groups formalize co-operation as both are aiming for similar release dates (technically the Newton group, at this point, is still aiming for an April 1992 release, but they know 1993 is more likely. Pink remains projected for 1993 as well). Most importantly this brings the Newton concept of ‘soups’ over to Pink. That is the idea that all data should be accessible by all applications, if need be.

MacOS group releases System 7, and begins work on System 8[5], intended as the final version of the original Mac operating system, for a 1993 launch. Their intent is to push the Mac operating system as far as it will go. System 7 slows sales of Windows as well as providing a slight uptick above projected Mac sales. However as System 7 is the first operating system upgrade that Mac users have to pay for some protest is registered. Apple responds by giving away free copies to anyone that bought a Mac in the last 3 months.

In the spring of 1991 Pink is branching out, with work starting on an emulation layer for System 7 and a hardware abstraction layer. Pink group is uncertain about the future CPU of the Mac platform, and so they decide not to worry about it by making a hardware abstraction layer. The Newton group agrees, and NewtonOS will also be processor independent. At this point Pink begins to worry about the file system, as they’re unsure if the current Mac file system will give them the support they need—especially considering their use of Newton-like soup data storage.

Shortly thereafter Pink group finds the Apple File System group. Through AFS Pink learns of Plan 9 From Bell Labs. Although they are approaching things from different directions both operating systems share a similar focus on communications. Apple signs a development deal with AT&T and Pink begins to systematically raid Plan 9 for everything they can use. By incorporating large parts of what is essentially a research operating system Pink will wind up in a very different place than the other major operating systems of our time: NextStep & Windows.

AFS is made the file system for Pink, with a 1992 preliminary (but stable) release and the remaining advanced features by 1994. As it is now obvious that Pink will slip from their projected 1993 release date, the new date is set as 1994.

TrueType meets with the QuickDraw GX group, and they decide to join forces as a form of empire building. TrueType GX will enable advanced use of fonts in morphing and ligatures. This will also extend to the use of TrueType & PostScript in printers & enable desktop printers. A side effect of this project is improved printing controls, which is adopted by the MacOS group for System 8. They decide, however, that actually incorporating QuickDraw GX would be too much work.

Microsoft approaches Apple about breaking the Adobe font monopoly (as their own PostScript replacement effort is faltering). Apple mentions that they might have something to do that with. Apple, however, decides that Commodore should be a part of this because of NextStep, and they extend feelers to Commodore.


1992
By March of 1992 Pink finished NuKernel and had a stable file system, and so they began targeting previously identified higher level pieces of technology for incorporation. Among them include the Open Collaboration Environment and QuickDraw GX. Around this time Blue was considering both of those pieces of technology for incorporation, if they had grabbed them it’s unlikely that their (probable) release in 1994 would have seen them either completely done, or being able to run fast enough on ‘94 hardware.

Diagram of Pink in 1992:

User Interface (Undisclosed)(1)
Dylan Programming Environment & Language
Open Collaboration Environment(2) / QuickDraw GX
Plan 9 / Apple File System / Hardware Abstraction Layer / Emulation Layer (for) Original MacOS
NuKernel

(1) At this point, in fact, the Pink team had not settled on the now-familar metaphor and interface of today.

(2) Excluding, of course, the PowerShare server element as that’s a separate piece of software.

By early 1992 the NewtonOS was getting closer to completion, but the mid level stuff, notably handwriting, remained poor—although it was entirely usable overall. So ParaGraph, the Russian company that invented Newton’s handwriting recognition (Calligrapher) is bought out by Apple and handwriting recognition moves in house under new cursive & printing divisions. Recognition divides into scanning words from the dictionary, from letter recognition, and learning (the user can define their letters in the program, or it can simply build an outline from practice translating the users writing).

The Star Trek project—porting System 7 to x86 chips, at the behest of Novell—finishes its job in an amazing six months.

Apple was therefore faced with a tough choice.

On the one hand the Sun/Motorola/IBM alliance had a new chip under development, SparcLite, with the promise of being much faster than Intel’s x86. They’d certainly love to have Apple on board as MacOS has a far greater marketshare than Sun Commodre’s NextStep and potential clones. At the time NextStep’s marketshare was fairly low, and nobody knew if forthcoming NextStep clones could expand the market.

On the other hand Intel owned 90% of the personal computer processor market, and x86 would free Apple from ever having to worry about CPUs again. Intel wanted a counter to Microsoft Windows as they clearly saw the problem of having their main operating system customer rules the market, and they’d likely have been willing to work with Apple on x86 processors.

SparcLite was faster then current x86 chips, better emulated their old Motorola 68k chips and had some features from Motorola’s 88000 chips that Jaguar/Cognac were targeted on so it wouldn’t be too much more hassle to port. Of course Star Trek was up and running on Intel’s x86 chips.

In the meantime talks with Microsoft about Adobe led Apple & Microsoft to confront Adobe with TrueType GX. Adobe clearly sees the problem, and offers them a deal. In return for Adobe licensing TrueType GX technology & Apple/Microsoft abandoning their TrueType font project, Adobe will reduce licensing fees for PostScript fonts, and redo all their fonts into a new PostScript format that will use the advanced features of TrueType GX and publish it openly as the new standard for fonts. Apple & Microsoft agree to this, because they also see that TrueType GX technology wouldn’t be highly used if Adobe wasn’t on board. Apple retains TrueType GX and licences is to Adobe and Microsoft—although as it isn’t done yet they’ll have to wait to use it. As Sun Commodore had earlier been informed (by Apple) about this they too join in and license TrueType GX for NextStep.

Microsoft, now on better relations with Apple, approaches Apple about a proposal to collaborate on software component based design[6]. Apple brings Sun Commodore on-board and counters with ‘we need standardized document formats first’ and quickly signs up most of the industry sick of working with Microsoft Office document formats and under heavy pressure Microsoft agrees. This is a major victory for MacOS & NextStep, as well as application developers.

Lost in the shuffle is any plan for software component/document centred application design. However Pink group is intrigued by the idea and incorporates it into Pink. Although the 1994 release is looking less likely, that remains the official internal company line.


1993
With Apple’s new CEO in place he decides in favour of SparcLite.

The Star Trek program is cancelled, and most of them move to Pink (to work on the hardware abstraction layer, so Pink will never have to worry about CPUs), a couple move over to Newton, and the rest quit.

The software people in Jaguar/Cognac are merged in with Blue, and System 8 is pushed back to 1994 so it will run on SparcLite.

The hardware people of Jaguar/Cognac move to the hardware side of things, and begin working on Apple’s first couple of SparcLite machines.

Developers are informed that the forthcoming System 8 has been delayed until 1994, and Apple is now targeting it for SparcLite chips. However, courtesy of Cognac, System 8 will have very good emulation and so old applications should run reasonably well.

As Plan 9 From Bell Labs continues in research paper mode the Pink/AFS groups divorce from them. However much of Plan 9’s applicable work has been used, and Pink has solid UNIX/Plan 9 derived underpinnings including very robust communication subsystems.

Pink begins integration work on the back-end for their various technologies, and the release date slips to 1995.

NewtonOS has now been finished but remains too slow on current hardware[7]. The Newton group convinces Apple to double their investment in ARM (since their first investment in 1991 ARM has done quite well), so they can just dictate the chips they want from ARM. Some discussion begins with the Reference Consortium[8] about scaling SparcLite down into the embedded processor space. This new delay will give time for more work on handwriting, optimization, and included applications. However without Sculley the Newton group faces a cutback in resources, and the project downsizes somewhat. Nevertheless the new Apple CEO does believe in Steve Sakoman (Newton project lead), and progress continues.

Dylan the programming language has been finished, and both Newton and Pink begin using it for application development. Although the LISP like syntax is unfamiliar, programmers widely consider it the best environment they’ve ever worked in. OpenStep, the old gold standard, has been beaten—but nobody knows it yet.

NextStep 2 is released and is widely seen as a solid improvement. NextStep 2 is SparcLite native. Furthermore OpenStep has been x86/SparcLite/68k & Windows/NextStep cross platform(s) since late 1992 so application transition will be very simple.

Windows NT blows past the planned release date of 1993. With NextStep and System 7 (MacOS) picking up marketshare against Windows 3, and Windows 4 coming out in 1994, they’ve decided to spend more time on the internal technology to better match NextStep and the rumours from Apple of a new operating system. At this time the Cairo project is raided for technology by both Windows NT & Windows 4.


1994
SparcLite is released as product, though it has been standardized for some time now. The chips, at comparable or lower prices, are noticeably faster than Intel’s x86 chips. Intel steps up their campaign against Microsoft to break compatibility so they too can have a ‘clean’ design type chip that can compete with SparcLite. Intel is stuck with x86 mostly because Microsoft refuses to do the work needed for Windows 3 to run on multiple CPUs.

Apple releases System 8, which is SparcLite native. They also announce their first SparcLite machine, the PowerMac[9], with intentions to transition both laptops and desktops to it.

Although they don’t beat Sun Commodore to market as the first with a SparcLite machine, they are second ahead of both IBM & Motorola NextStep clones.

Pink now has their various technologies working together with the file system, NuKernel, and general underpinnings finished. However most of the middle and upper layers of the operating system need work, and application development is progressing slowly as people come up to speed on Dylan. Pink’s user interface concepts & metaphors have been hammered out but work remains to implement them.

NewtonOS has seen a lot of work over the last year and is much faster on the same hardware, with 85% handwriting recognition, and a lot of polish. However the group feels that a more work could be done and they hold off Apple management as they continue to develop Newton’s features.

The QuickDraw 3D group starts up, but soon finds itself split on the question of low level programming. Roughly half the group would like to develop a high-level application programming interface (as there were none at the time), and the other half would like to develop a low-level application programming interface, essentially duplicating OpenGl or Microsoft’s Direct3D. In the end the decision is made to simply use OpenGL as the low level and stick with developing a high level interface.

In a stroke of luck for Pink, Windows 4 is pushed back another year[10] to polish the user interface. With the release of System 8 the original MacOS is widely considered the easiest to use; NextStep 2 is also considered well ahead of Windows on user interface grounds, and ahead of both Windows 3 and System 8 on internals & overall performance because of that.

Windows NT is shipped as Windows Server 1994. With a similar interface to Windows 3 it does not bring anything new to the table for consumers, but as server software it’s quite good and sales are steady. Windows NT Consumer continues development, aiming for a 1996 release date.


early 1995
At the January Macworld in San Francisco Apple announced their forthcoming Pink operating system with release in one year at the same Macworld. It was quite a show. They ran through the new ‘People, Places, Things’ operating system metaphor, demoed the new document based application development, showed off QuickDraw GX’s advanced typographic features (as they knew that Windows 4 & NextStep 3 will also include support for them), demonstrated the sophisticated Open Collaboration Environment, and also let people play with Pink’s hardware abstraction layer (by running Pink on a SparcLite and a x86 box) and emulation (by running System 8 programs).

The press at the time is heavily mixed, many loving the new way to approach the user interface, and the rest hating the way it deviates from the old metaphor of files and folders. The features of it are widely liked, however, and the technical discussion of internal systems is shocked by the Plan 9 like underpinnings, and the NuKernel.

Dylan is released as a programming language & development environment running on System 8 & Pink and picks up widespread praise. Furthermore development guides for Pink, QuickDraw GX, and Open Collaboration Environment are published and Apple announces that they’re porting Open Collaboration Environment to both Windows and NextStep and that it will also launch at next January’s Macworld. They gloss over that Windows & NextStep will support QuickDraw GX features but trumpet the fact that Adobe has made all PostScript fonts GX compatible.

Although Newton was intended for this show, Apple management concluded that they want the focus to be Pink this year to counter Windows 4 and the forthcoming NextStep 3. Newton group wound up losing some people, but gained a full extra year of development—although they have to be ready to be shown off in August. They promptly begin rewriting as much of NewtonOS as possible to improve performance, focus on near-perfect handwriting, and continue pushing ARM for better CPUs. They throw away their current case design and restart fresh. Finally they also begin prepping a new smaller Newton model[11].

Overall the Newton released in 1996 bares little resemblance to the Newton that might have been released in 1993 or 1994. Likewise a Pink release in 1993 or 1994 would have simply been a Windows NT type release, a MacOS with new internals and a few new features but otherwise basically the same.


late 1995
Windows 4 is released (including Apple’s TrueType GX technology) and conducts a marketing blitz in an attempt to win back marketshare from MacOS/NextStep.

NextStep 3 also sees release (including Apple’s TrueType GX technology) and conducts a marketing blitz aimed solely at Windows.

The press generally see the two launches as the coming of age for personal computers. Windows 4 finally has a user interface ‘close enough’ to System 8, and NextStep finally has enough marketshare to compete on a level playing field in mindshare with Windows & MacOS.

What overshadows both, however, is the announcement of the Newton. Shown at Boston Macworld in August with hundreds of units on the floor to play with the press loved it. Further with Dylan already released they only have to add the Newton module to it & release the Newton development guide, and people now knew what to develop for. Newton’s launch in January is to give developers a few months to write programs (and to give Apple time to make some Newtons, of course).

The latest version of Pink is also shown off at Boston Macworld and with more time to play with it the user interface does get some better reviews—although opinion remains love it/hate it. As before, and despite TrueType GX now part of Windows 95 & NextStep 3 the general review of Pink’s features sets it head and shoulders above the other operating systems. Apple shows a few features of Pink they didn’t reveal earlier in the year such as Active Assistance (similar to the Newton’s use of it), everything is always saved, syncing with Newton and more detail on lower level features like the file system & microkernel. Finally, for developers they unveil the new QuickDraw 3D API and promise to ship it with the first update to Pink and to go cross platform.

(By 1997 the OpenGL/QuickDraw 3D combination will force Microsoft to end development of Direct3D as developers declined to use it, given its shortcomings against OpenGL/Q3D.)

Diagram of Pink in 1995:

People, Places, Things
Dylan Programming Environment & Language
Open Collaboration Environment(1) / QuickDraw GX / (QuickDraw 3D)(2)
Plan 9 / Apple File System / Hardware Abstraction Layer / Emulation Layer Original MacOS
NuKernel

(1) Excluding, of course, the PowerShare server element as that’s a separate piece of software. Will also launch in mid-1996 for Windows/NextStep

(2) Will launch in the second half of 1996 for Pink, first half of 1997 for Windows/NextStep.


early 1996
The release of Pink, and the release of the first NewtonPad at Macworld San Francisco, in January. Also at that expo was a first look at NewtonNote: a new smaller Newton model, planned for the end of 1996.



Macworld San Francisco, January 1996.

In 1987 Apple started our next generation operating system. It seems to be that kind of year as NextStep began then, and Windows NT began the year after. We called it Pink, because of the index cards. There was a meeting where we sat down and talked about the future of the Mac’s operating system. Blue was for the easy features, and that became System 7 and eventually System 8. Pink was for the more complex features, mostly relating to stuff like protected memory and multitasking and so forth—internal features that improve the experience. Red was pinker than pink—that is more in the future—for stuff like voice command.

Pink started like Windows NT. Build a new base and put the original Mac user interface atop that, perhaps with a few new features. Don’t get me wrong, Windows Server 1994 is the only Microsoft operating system I’d use, if forced to, because it’s stable. However it’s the same old user interface, a poor copy of MacOS.

Then there’s NextStep, which also produced modern internals and an updated user interface and it’s quite good, I freely admit, but still basically a different interpretation of the original MacOS user interface with quality underpinnings.

By the time Windows 4 & NextStep 3 came out they’ve produced a pretty copy of our user interface, we’re not looking quite as brilliant. However it’s been a decade since then we started the modern graphical user interface with the Mac in 1984, so it took them a while. We decided that rather then mild improvements in the user interface, like System 7 & 8, we’d use our opportunity with Pink to create an entirely new interface metaphor for computers. We think it’s a better one.

Pink is the future. A next generation operating system for the Macintosh, built from the ground up but still compatible with all System 8 programs. Pink is object oriented from the kernel upwards, and is buzzword compliant from protected memory to preemptive multitasking.

I’m going to go from the bottom up. Starting there we see our microkernel, NuKernel. Building on all the lessons learned both from Apple projects, the Vanguard microkernel, and many outside microkernels to make a very fast and lightweight kernel, that also has high performance.

Like I said Pink supports all the buzzwords of a modern operating system, from protected memory to preemptive (both background and user interface) & symmetric multitasking. It also has a hardware abstraction layer, meaning programs written for Pink—as well as Pink itself—can run on SparcLite or x86 or future advanced chip designs. It’s very simple to port Pink, and properly done applications will move right along with it.

We have a new filesystem to go with Pink. We didn’t originally plan on this, which is one reason we’re a little late, but the benefits were hard to ignore. Our new Apple File System, AFS, features data protection and speed. It’s very fast, and we’ve had it benchmarked against other file systems but what I feel is more important is the way it handles data. We’ll go into more detail in the sessions, but suffice it to say your data is very safe. It detects data corruption and, if you have a back-up, seamlessly replaces it with an older version. If you have several hard drives everything can be duplicated, allowing recovery with the current version. Furthermore everything is treated as a ‘file’ from the networking stack to the graphic system, which allows for simple and effective manipulation.

QuickDraw GX handles what you see on your screen, from sophisticated font handling to system-wide anti-aliasing to make sure everything looks crisp and beautiful. One thing to also note is that QuickDraw GX brings resolution independence to Pink, no matter the resolution of your monitor you can easily scale Pink to match what you feel like working with: with no loss in quality onscreen. Infinite zoom, with everything resizing to match, although System 8 programs cannot take advantage of this.

The font/print element of QuickDraw GX, TrueType, is something of an industry standard, already, as Windows & NextStep both use it their most recent releases. Furthermore Adobe has made their PostScript fonts work with GX technology. Just fantastic looking typography.

Now OpenDoc is interesting. What OpenDoc mandates is that applications must talk to each other. This allows for narrowly focused applications that seamlessly transfer data to another program. It’s a combination of Newton’s soups, allowing a general storage of accessible data, and a document focused application and communication model.

Perhaps you like to use a certain word editor for your writings, but need to also use big complex spreadsheets, and then have to use Photoshop to adjust graphics and afterwards words, data, and graphics are put together into a page layout program. Currently you’d have to copy and paste and lose most formatting, or perhaps just use the page layout application to do everything poorly. Now all four applications would talk to each other and allow you to drag and drop (or tell it to transfer) data in without any lose of formatting and without any problems.

Developers benefit from now being able to write specialized programs—perhaps a word processor for the legal field—and can now break larger programs into separate components. Obviously we’ve spent a great deal of time on our frameworks and development environment to make this as easy as possible.

Dylan, our innovative object oriented programming language, is made for easy and fast application development and I promise that you’ll love it once you get used to it.

Now I’d like to talk about our Open Collaboration Environment. This is an end-to-end operating system independent delivery system. Some features are part of Pink, but others are not. Essentially this gives you a global mailbox & global address book synced across multiple devices with some other features we’ll get into.

For example I may have a Mac running Pink. However I also have a Newton, and I even have a Windows & NextStep box. Normally keeping track of just email would be difficult, let alone documents and perhaps I like to draw on my Newton and send them to people. Under the Open Collaboration Environment model it becomes simple.

PowerShare is the server software, PowerTalk is the general transportation layer, and Open Directory is your contacts and user interface modelling. Together with encryption & digital signatures (managed by the Keychain in Pink and the Newton only) this makes up the Open Collaboration Environment.

Let’s say there is a server somewhere running PowerShare, which is specifically designed to work on Pink, Windows, or NextStep servers. On each device I have a PowerTalk/Open Directory compatible client. PowerShare will automatically update all three devices mail and contact list to keep them in sync with each other, and of course you not only have your files on the server but also in three separate independent backed-up places.

Something interesting to keep in mind is that PowerTalk prefers peer-to-peer transfer. If you both happen to be online the document will skip the PowerShare server. Essentially it will be from you to Bob to Bob’s PowerShare server so that things can be synced. If the person is offline it follows a more normal You to Bob’s Server to Bob pattern. This also means you don’t need the server, although if someone sends you something when you’re offline it will fail, and of course you won’t be able to sync multiple devices without connecting them directly online at the same time.

So that’s email and contacts, so you’re always working off the same data. What else? Well everything is a place or a person or a thing; if you want to send a document from the Newton to the Mac you can, or if you want to send a document from the Newton to a printer you can. PowerTalk is also a transportation layer. It functions with any networking standard and sends/receives with any email standard as well. Obviously Apple is making PowerTalk clients for the NewtonOS and Pink, but a Windows & NextStep versions could be made by anyone as well.

You could, if you wished, send a document to a friend (without any use of email) and also print that document on your networked printer. One person, one thing; both work. Of course if you’re sending something to a someone not using PowerTalk the system degrades gracefully. Documents are turned into email attachments, for instance, and the benefits of being able to send someone a contact and have them press a button to add to their contact list are lost, but the basic data still comes through.

Furthermore the entire system features optional transparent encryption end-to-end via a public key system. You create the key with a public part, where anyone can encrypt anything, but only you retain the private part to decrypt things. This includes a digital signature, so you can be sure of who sent you things. Now this doesn’t verify a person’s identity, but you’ll always know that it’s CyberDog52 sending you stuff as his signature will remain the same & the signature will let you know if the file has been tampered with after it was signed.

If you’re using Pink or have a Newton you possess a Keychain, that handles all your passwords and encryption tasks which is itself covered under a master password—it even tells you how to create a good password and rates them for you. The Keychain can be synced (though it requires you to use encryption) between devices as well.

To sum up: Nukernel, a sophisticated microkernel. A hardware abstraction layer, meaning Pink can run on any powerful enough CPU such as x86 or SparcLite. An emulation layer for your old System 8 programs until they have new versions. The Advanced File System which is faster then competing file systems, and also works to keep your data safe and backed-up. QuickDraw GX for onscreen graphics and print/font handling. The Open Collaboration Environment for superior handling of email & documents across networks, along with a global address book, a global mail address, and encryption for everything if you so choose; we even include a global password manager. As mentioned the TrueType part of QuickDraw GX is a standard is used by Windows, NextStep & Adobe, and is available for license. The Open Collaboration Environment is an open standard, as the name applies, and anybody can build compatible client interfaces, the PowerShare server software is a reasonable price standalone and can be licensed for incorporation into your email server software.

I think it’s clear that the underpinnings are fantastic, but what about the graphical user interface?

You might have noticed I threw in a reference to People, Places, Things in the Open Collaboration Environment section. That’s how Pink is built. Document centred, multi-user networked, direct manipulation interface with infinite session undo under a principal interface paradigm of People, Places, Things. People are your contacts, Places are environments that can be like folders today or be structured around tasks and projects and users and you can even make them collaborative, and Things are fax machines or printers. For example you could all work on notes for class together in a Place, and then each person could send it to a Thing to be dealt with—whether the thing was a printer, or a fax machine—or simply leave it be and then you can go look at it again in your place with any changes spread back across the other users of the Place using PowerTalk. Or everyone could subscribe to a shared calendar editable by all you choose to allow, and every time you go alone your calendar is updated to match the current revision.

Finally, and I think this might be the best part of Places, PowerTalk allows real time collaboration. If two users of PowerTalk clients wish they can establish a direct connection to each other using their global mail address. This can be as simple as chatting back and forth, or if you both had the relevant PowerTalk program you could say draw things together or take notes together. If you’re not using Pink you don’t get the full power of Places, but you can still collaborate using PowerTalk clients.

Pink also moves to a document centred interface using OpenDoc, where the document is what matters and you put together your personal combination of applications to do the job rather then be forced into monolithic programs that try and do everything. Oh, and of course—like the NewtonOS—you never have to actively save anything, unless you specifically want multiple copies. With an infinite undo that’s saved when you close your document you can open it back up and either start hitting undo or open the graphical undo interface and scroll until you find the version you want. That’s one of the useful things our new Apple File System does for us.

I hope you enjoy trying out the release version of Pink, and don’t forget to check out the Newton. Both Pink and Newton are available now here, or anywhere Apple products are sold. Have a good Macworld everybody.



NOTES

In general, ITTL, Apple is much better about not talking about all their cool projects. With NextStep and a better than OTL Windows they’re much more worried about other people copying them. Furthermore under Sculley’s ‘don’t bother me with technology, as long as I see results’ (IOTL, just the first part of that was true) management pressures groups to get much farther along before talking about their projects.

Therefore burning developers with every cool thing that never quite works (OTL’s QuickDraw GX, QuickDraw 3D, Open Collaboration Environment, Pink, OpenDoc, Taligent, Copland) doesn’t happen ITTL.

[1]
This is similar to OTL’s Open Collaboration Environment. However the various problems with OCE in OTL are solved ITTL because of their merger with Pink.

It’s similar to an OTL’s IMAP (where messages are stored on the server, and computers are sync their local copy) or perhaps an OTL Exchange server. It features the added benefit of optional peer-to-peer mail, and general delivery (documents, print jobs, etc… can all be ‘sent’) combined with encryption, digital signatures a single unified email address (back when there multiple email formats/services), and a global synced address book.

The global password Keychain (OTL’s Keychain) of OTL’s OCE is just a regular part of Pink ITTL.

[2]
This NuKernel resembles both OTL’s NuKernel and OTL’s L4 microkernel.

[3]
QuickDraw GX & TrueType are similar to OTL’s projects except that Adobe/Microsoft adopts GX technology in return ditching TrueType fonts and QuickDraw GX is much easier to program for.

ITTL Gassée does not piss off Adobe with the old version of the project (Royal). Instead it’s more of a theoretical exercise in improving fonts & printing, that can eventually be used to break Adobe’s hold on fonts in a much quieter way than OTL.

[4]
Microsoft Cairo was never more than a collection of technology, and was never released though bits and pieces made it into later operating systems. It was announced IOTL to counter System 7 and NeXT’s NeXTSTEP (both actual out the door OSs) with vaporware to scare off people from moving to them (since Windows 3 was only partially doing that job). ITTL it’s a reaction to Commodore’s Nextstep operating system.

However it will have something of a different fate ITTL.

[5]
TTL’s System 7 is delivered at about the same time, but is both much more solid and much more stable. IOTL System 7 suffered because Pink drained all their staff (they even had to hire the entire graduating class of Dartmouth to work on System 7) and ITTL that doesn’t take place as Pink staffs up in a more sensible fashion.

System 8 will be the consolidation and solid improvement that 7.5 (sort of) was IOTL, except starting from a much better base.

[6]
OTL’s OpenDoc, which will not come to pass ITTL. As with OTL Pink incorporates an OpenDoc light style of applications and (hah!) they’ll call it OpenDoc.

[7]
The Reference Consortium is TTL’s version of the AIM (Apple-IBM-Motorola) alliance of OTL. It consists of Sun Commodore, IBM, Motorola, & Apple. Several clone manufactures are interested in joining as members to get the SparcLite CPU and a license for the NextStep operating system and a cloning license. Sun Commodore gives away OpenStep separately to encourage Windows/NextStep cross programs.

Sun Commodore’s contribution is: SPARC + design/fab experience/facilities, NextStep + Commodore, NextStep + cloning license, & OpenStep. They also have Sun servers running NextSolaris (NextStep Sever), & Atari consoles (and a requirement that any computer running NextStep must either incorporate a slot for Atari games, or have it as an add-on).

IBM’s contribution is: POWER + design/fab experience/facilities, & NextStep clones.

Motorola’s contribution is: 88000 + design/fab experience/facilities, & NextStep clones.

Apple’s contribution is: CPUs sold. They buy SparcLite processors and use them in their machines and as Apple’s sales in 1994 for SparcLite machines are more than everybody else combined that’s a big contribution.

[8]
Newton, ITTL, has around the same level of resources poured into it. However they also realized the handwriting problems which pushed them back from 1993 to 1994, and work with Pink on syncing has slowed development some more—as has creation of a syncing program of System 8, Windows, and NextStep. Finally OTL Newton didn’t really hit it until NewtonOS 2.0 and MessagePad 2x00. ITTL delaying until early 1996 will mean NewtonOS will be better than OTL NewtonOS 2.0, and the hardware will be much closer to the MessagePad 2x00 of 1997.

ITTL Sculley didn’t talk about his Knowledge Navigator so expectations aren’t really high for the Newton (in fact not many people know it exists). Further the Newton wasn’t shown in 1992, and therefore PDAs as a market haven’t really taken off, and nobody else has handwriting recognition.

The Reference Consortium is the group centred around SparcLite and NextStep. Reference is used because they are very much mounting a challenge to Microsoft, and want the slight rhetorical edge that calling themselves the Reference Consortium can bring.

The Reference Consortium, at the time, consisted of: Sun Commodore, Motorola, IBM, and Apple.

[9]
Derived, as IOTL, from the PowerBook line of 1991 not—as is sometimes believed—from the PowerPC chip.

[10]
As IOTL Windows 4 (née Windows 95) ends up being two years late. ITTL it will stick with Windows 4, as Windows NT will be getting the Windows Year format names.

[11]
The first Newton will be similar to OTL’s MessagePad 2000 except somewhat slower/less memory. However the extra operating system time means ATL Newton will actually perform better than OTL’s MessagePad 2000, or the 2100 for that matter.

It will resemble a MessagePad 2000 and have the same screen (in size, resolution, 16 grayscale) but less plastic around the screen and hence somewhat smaller and lighter—in size and weight roughly like the original MessagePad.

However the extra time means that a smaller Newton model (Palm Pilot ish) is also in the works as was planned IOTL.




Comments, questions, some kind of discussion perhaps?