Hacker News new | past | comments | ask | show | jobs | submit login
The Jonathan Computer (storiesofapple.net)
80 points by rbanffy on Oct 26, 2021 | hide | past | favorite | 31 comments



The idea of the 'reverse Trojan horse' reminds me of the later fate of OS/2, which shipped with such good Win 3.1 compatibility that nobody felt the need to build native OS/2 apps.

The Jonathan is probably my favorite Apple concept but I agree that the business case for it would have been pretty terrible for several reasons.


The Windows 3 compatibility was licensed from Microsoft. That way, in order to offer Windows 3.1 compatibility, IBM needed to pay Microsoft per OS/2 seat.

IBM also refused to bundle OS/2 with their own computers at a lower price point than the ones with Windows bundled.

All things being the same, OS/2 was more expensive to run, while Windows 3.1 was crappy but got the job done.

In Apple's case, they could just not make the MS-DOS compatible module. If someone else made it, Apple would still make money out of licensing.

Since all Mac software relied on the Toolbox, it wouldn't be that hard to offer some form of multi-processing more or less the way we used X in the early days - by having a puny workstation (!) on our desks and running the heavy lifting on a big machine (that didn't even need to run the same flavor of Unix, or Unix for that matter). The difference would be that any "network" between the boxes would be on the backplane bus and VERY fast.


>IBM also refused to bundle OS/2 with their own computers at a lower price point than the ones with Windows bundled.

Not only that, but, IIRC, they charged more for a PC bundled with OS/2. This was a result of all the groups within IBM having been moved toward being run like individual businesses. So, the OS/2 business charged the PC business (whom they treated like any other customer) more than Microsoft charged for a copy of Windows.

At least that's what I heard at the time.


Besides being more expensive, Windows 3 apps were (in my memory, at least) every bit as unreliable under OS/2 as they were on Windows (because the compatibility was an almost full Windows runtime), further negating any advantage OS/2 could have.

As you point out, even the version of OS/2 that didn't have the Windows compatibility malware was more expensive than Windows.

It's really hard to believe IBM would be so short sighted, but Microsoft is incredibly lucky in that regard - every company that tried to compete head-on with them, with the possible exception of Apple, made similar boneheaded moves.


One reason I've never gone to work for Apple (before they became Evil) was because I saw so many people who had poured their hearts into great products only to have them buried.


I mean the core premise behind it not being launched is that it might have posed a significant risk to the business. It’s all about cost benefit analysis and trade offs. Every business will have products that seem cool or features that seem cool that get shelved for one reason or another and Apple is no exception.

Sure, this concept might’ve been awesome and ignited a revolution. Or it might have sent Apple to the grave.


It's a cool, revolutionary concept [1] but it would have overshadowed the Mac, which would have gone from being the star attraction to an optional add-on.

There would also have been potential problems with manufs copying the bus spec and making their own hardware.

Working through whether it would have been a good thing or a bad one is a very interesting case study. I suspect it could have been pushed towards being a good thing, but not without a lot of friction and changes in a company that was basically set up for commodity single-box products.

[1] Actually it nearly isn't. It looks like a modular bus broken out into boxes - so basically S-100 or ISA, but with separate cases instead of cards. Which makes me wonder about potential speed issues. But it looks like a revolutionary concept, because it's much cleaner and more designer-y than anything else that was around.


And the beautiful design hid the fact that this would have been a nightmare to make it work in software. Plug-and-play just was not a thing in that era, and it's barely there now.

I can't see how you would make the idea of "unplug the Apple ][ box and plug in an 80286 box" work without with a mountain of software to assist.


> Sure, this concept might’ve been awesome and ignited a revolution. Or it might have sent Apple to the grave.

The IBM PC almost did that - as soon as clones begun to appear, IBM was relegated as just another PC maker. They tried to make a software-compatible PC that would not be easy to clone in the PS/2, and they failed badly. Now IBM doesn't even make PCs or PC-based servers.


This is common at any Big Tech company. There are dozens to hundreds of “prototype” projects going at any point in time, and while working on them you get paid a ton and have excellent job security, but no promise that your work will see the light of day.

It’s not a terrible deal, and it’s a great learning experience. For some, it’s a great entire career. For others the not really shipping your work part is hard and thus it’s only good for a season of life, to learn and to pile up savings. But that can set you up well for leaving and trying your own startup later.


That could literally happen at any company. Just because you have an idea, it doesn’t mean it’s commercially nor strategically viable.


Well I've mostly worked for startups where we had to build the thing :-)

But in Apple's case it was much bigger - lots of projects, people spent years built chips, whole new platforms, new technologies, 2d/3d rendering libraries - FireWire came so close to being still-born that its team was shopping it around 3rd parties trying to get them to drum up internal Apple support


Companies make these decisions all the time. What’s remarkable about Apple is how far they were allowed to take some of the prototypes.


A similar concept was the Burroughs B-20 / Convergent AWS.

https://en.wikipedia.org/wiki/Convergent_Technologies

Converget Technologies did a LOT of pretty brilliant designs.


I would buy a computer that worked that way today! It's like building your own PC except way, way easier.

If implemented correctly you could upgrade the necessary parts over time. My only concern would be that eventually the "bus" part would become out of date.

Still, it looks amazing.


Are you sure? I bet the parts would be more expensive than a design with one enclosure hosting multiple cards (somebody has to design and build those extra enclosures).

Also, communication between parts likely would be slower than for an all in-one design.


Am I sure of what? I never said I thought it would be cheaper or faster.


maybe framework will follow up with a desktop machine!


“The design’s militaristic look with smooth surfaces, sharp corners, vertical ribs around the base and the use of a dramatic black color with white product graphics was unlike anything done before at Apple.”

This really wouldn’t look out of place on the Death Star, or even in Vader’s meditation chamber.


I'd like to have seen the world where this actually came out. I think more than a few of the MS-DOS users of that era might've migrated to that.

It probably could have convinced my father and I to switch, if implemented sufficiently well to allow a good upgrade path.


In reality there were things like OrangePC cards that let you run DOS on Macs. https://en.wikipedia.org/wiki/Orange_Micro


In a way, this is it: https://frame.work/

But really, these mockups are absolutely beautiful. IMHO easily among the best looking designs I've seen in computing.

I think the framework's approach is a good one, it would be very interesting to have a cheap data transfer backbone that you plug modules into -- to include adding more processors. It would be a better form factor than wires and dongles all over the place for things like external hard drives and external GPUs.


This isn’t correct:

> some key design elements of Jonathan, such as the ribbing and sharp corners, were incorporated by frogdesign in the first unified look of Macintosh products, known as SnowWhite.

The IIgs, mentioned at the very top of the article as Fitch’s recently-completed project, was part of the Snow White design language. There’s a time travel violation here.


This has the look and feel of modular PLC architecture.. one base computer, easily extended by various plug & play modules.


I love this concept, and I love the execution in these concept machines. I know a lot of things don’t make sense from a business perspective, but I wish more of these ideas had seen me light of day.

In some ways we see the execution of at least the ideas in the modular Pismo PowerBook, and lately the FrameWork laptop.


Killing that was probably a good move for Apple: "reverse Trojan horse" is right.


That keyboard looks really sleek for 1984!


Note it’s just a black version of the actual Apple IIgs keyboard.

https://deskthority.net/wiki/Apple_Desktop_Bus_Keyboard


Someone should do one with modern low-profile mechanical switches. I LOVE those keycaps.


Best using Alps mini switches, like on the TRS-80 Model 100 and siblings! (Well, those are not manufactured anymore.)


Mechanical compatibility has a way of outliving its users. There is something available today that probably has the same physical dimensions as those.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: