Egg Freckles: The Newton at 30 | Hacker News Reader

Egg Freckles: The Newton at 30

timemachiner.io

44 pointsmacstainless37 comments

I still have a Newton (a 2100), and although it ceased to be a daily carry device for me somewhere around 2007, I only finally stopped using it altogether just a few years ago, as it was becoming increasingly difficult to move my notes on and off the thing.

The bad handwriting trope has always baffled me. I've never had a problem. I know that the 2000 series Newtons are considerably improved compared to the original generation, but I did also use an original generation Newton, and it was actually fine.

I think the "Eat up Martha" gag from the Simpsons became such a trope that people just assumed it must be awful.

I was always amazed by it. Other PDAs used a special alphabet one had to memorize, but Newton learned and recognized natural handwriting. And while Newton didn't last long, the handwriting recognition did, reincarnated as Inkwell.[1]

[1] https://en.wikipedia.org/wiki/Inkwell_(Macintosh)

And then Microsoft acquired the company that made Calligrapher (which originally ran on the Newtons), and it ended up in the early WinCE PDA devices. I used it with the WinCE devices, and it worked quite well. "Egg freckles", indeed.

> Other PDAs used a special alphabet one had to memorize,

Graffiti was so easy to use that you would see people write with it on whiteboards.

This has always baffled me. Graffiti was generally less than 20 characters per minute for me and I was an early palmpilot adopter. Eventually better options came out resembling the smartphone keyboards we have today.

I never used a Newton but the Microsoft handwriting recognition was pretty good on the iPAQ. Maybe the Palm graffiti approach is ok for people who don't care about feedback and can edit tiny characters on a tiny device.

Graffiti was great to use, but felt like a step down because it required memorizing the input method whereas the Newton's promise was to have full handwriting recognition. Granted I was on as MP 1200 but I don't think the 2000 series improvements were enough.

The motivation for Graffiti was it could be recognized by a small microprocessor meaning the palm devices could be quite inexpensive.

I actually installed Graffiti onto mine and it made a huge difference. I have bad handwriting in general so I could never fault it for misinterpreting my chicken scratch.

Agreed - the hand writing recognition was fine, especially in later versions!

I worked at Apple UK at the time, and for a while did Newton Support!

I loved those devices!

They were so capable. They had IR built in, and you could send files to each other. You could also get TV Remote apps! All amazing stuff to me at the time!

And unlike the Palm devices of the time, didn't erase themselves when the batteries ran out!

One of the most fascinating things about the Newton is that its systems and applications were written in Dylan, a LISP-like language, though that seems to have been replaced by NewtonScript.

https://opendylan.org/_static/dylanwwdc94brochure.pdf

https://web.archive.org/web/20160328123939/http://opendylan....

https://news.ycombinator.com/item?id=22398899

Dylan was never on the Newton. The Newton team had two proposed products, Senior and Junior. Senior was gonna run Dylan, but they kept waiting on the Dylan team to produce something usable, and it never happened. So they gave up on Senior, dumped the notion of using Dyan entirely, and finished Junior using NewtonScript to get the product out the door. That became the Newton.

Yeah, that's what Walter Smith says:

https://news.ycombinator.com/item?id=22400696

Ironically, I see my old comment has two autocorrect typos. :)

I've read somewhat opposite account, that there was a complete Lisp/Dylan system available, but there were issues both political and hardware related that ultimately led to C++ (which was then hot new stuff still) being used with NewtonScript on top.

One of the things that fascinates me about the history of the Newton is that if it weren't for complaints from those advocating C++ that escalated to John Sculley (who ruled in their favor), we may have ended up with mass-market machines that ran a Lisp operating system. There's a wonderful interview between lispm and Mikel Evins, two regular contributors on Hacker News, that can be found at http://lispm.de/lisp-based-newton-os.

I'm quite fascinated by pre-1997 Apple's projects. Before Jobs' return, Apple was an unfocused beacon of innovation; in its attempts to find a successor to the classic Mac OS, it explored various paths, but none of these paths led to a successor until Apple bought NeXT. Apple's more radical explorations during this era seem to have a profound Xerox PARC influence. Xerox PARC is probably most famous for Smalltalk and its GUI. However, there are many people who don't know that Xerox developed its own Lisp environment, Interlisp, which was known for its GUI (not all Lisp innovations came from MIT). These explorations ended once Apple committed to NeXT technology for its foundation in 1997, and today Apple has been pursuing its own direction (e.g., the iOS human interface guidelines, Swift, the gradual inclusion of iOS UI/UX elements and practices in macOS) that is distinct from the Xerox PARC, classic Macintosh, and NeXT traditions.

Apple in the mid-to-late 90s was fascinating. You could see they were trying but had little to no direction. Not that Jobs was the best person ever to run a company, but I believe it was only him who could've saved Apple.

If Apple didn't buy NeXT or if Jobs didn't want to come back, it's an interesting "what if" to determine who would've led them and what they would have done. I could totally see MS _STILL_ investing heavily in them in order to help with their own antitrust issues at the time.

In my opinion, the most interesting alternative timeline of Apple's history would have been had Pink/Taligent been released sometime around 1993 or early 1994, coinciding with the release of the original Power Macintosh lineup and thus beating Windows 95 to the punch by well over a year. In fact, it would have been quite revolutionary: a workstation-grade OS with protected memory, cooperative memory, etc. but installed on consumer-grade hardware; in our real-world timeline, this wouldn't happen until 2001 when Windows XP and Mac OS X were released (yes, I'm aware that the Amiga had protected memory, and yes Windows NT was around, though it was marketed to workstation users). Imagine everyday users having a rock-solid OS in 1993. Pink/Taligent would have also stolen NeXT's thunder since Pink/Taligent also provided an object-oriented development framework, but unlike NeXT, Pink/Taligent would have had a much larger install base. NeXT probably would have either pivoted or folded, and Steve Jobs would've probably been more well known for Pixar.

Then again, pre-1997 Apple courted larger companies like IBM and Sun for mergers. Apple under John Sculley almost merged with IBM; given the AIM (Apple/IBM/Motorola) alliance and also the fact that Taligent and was jointly developed with IBM, I can see a timeline where Apple and IBM merged as a way for IBM to take back control of the personal computer business from Microsoft and Intel. There are all sorts of fascinating possibilities.

Oh for sure. The possibilities are endless. Even with OS2/Warp and BeOS, there was a lot of competition brewing that could've either acquired Apple, Apple acquiring them, or some sort of merger.

I completely forgot about the AIM thing until you mentioned it! Damn those were the wild west days of computing.

Handwriting recognition is a "hard" problem, and the Newton tried to take it on directly, with more or less success. Other companies went for workarounds. Some people retrained humans to write in such a way that the PDA could easily recognize it, e.g. Palm with their Graffiti writing system used with PalmOS. Others just put a keyboard on the device, as in the early WinCE devices, and which reached the pinnacle of evolution on the Blackberry. Then smartphone makers just turned that into a software keyboard on the screen, which is what we see in iOS and Android.

Now, the big thing is voice recognition, which is also a "hard" problem. But we no longer need to seek workarounds the way we did with handwriting recognition, for two reasons: first, because the phone itself now contains much more processing power than the early PDAs did, and second (and more important), the evolution of high-speed wireless networking makes it not only possible, but feasible, to offload the "hard" parts of voice recognition to an even more powerful server somewhere on the Internet. Which enables, not only voice recognition, but software agents like Siri, Alexa, and Google Assistant.

One day there will likely be a breakthrough somewhere that makes the gap between handwriting recognition and voice recognition look like child's play. I'm not sure what that will be. Maybe direct brain interfacing, as in William Gibson's or Jim Strickland's science-fiction universes, or as in John Scalzi's "BrainPal"?

(Taken from some musings I wrote on Facebook back in 2018)

still have a 2000 upgraded to 2100 in a closet somewhere. At some level the iPhone was such a disappoinment after having owned a Newton, becuase the early iPhone, even after the introduction of the app store, felt like a bunch of siloed apps, where the Newton "soup" allowed fairly seamless extensibility and interoperability. (well, at least it felt like that in the 90's.)

I always wonder how a Newton would perform on modern hardware. Considering how well the handwriting recognition worked on 90s hardware it should be pretty great by now. I definitely miss it. The iPhone feels like a step backwards in terms of user and developer friendliness.

Part of the issue is the move from resistive screens to capacitive screens. The iPhone is just not accurate enough for handwriting.

The iPad with pencil is frustrating for handwriting. But the iPad pro with pencil 2 is likely better, but I've not had a chance to try it.

The tablet PC era computers were pretty good with handwriting but the os didn't make enough drastic changes in user interface to make them work well. Choices were using the on screen keyboard a lot or getting a clunky unit that included a keyboard.

Chromebooks with pen are quite frustrating, and I recommend against them. The pens also tend to use disposable AAAA batteries.

Of course the early iPhone apps were siloed, one could not even copy/paste between apps until three versions in. :-) But the iPhone did, indeed, feel like a step back in many ways. Though Microsoft would let it languish and later get stomped by iOS, the early WinCE/Windows Mobile devices had better inter-app communication and integration. I felt like I could get stuff done (and I did) on some of those devices. The early iPhone had a lot of innovation, but there was no last-minute editing of a Word doc on those devices.

Interestingly, the Newton was pretty cheap for its time: released for $900 ($1,844.29 in inflation-adjusted terms). IIRC the original price was supposed to be much higher, but at the last second it was cut lower.

Most first-in-kind technology is usually quite expensive[1] (especially during the 80s/90s era), but Apple has frequently been able to release innovative products (relatively) cheaply, as far as I can tell.

[1] https://simulavr.com/blog/paying-for-productivity/

Something that strikes me, using a Newton today, is just how little has changed in the fundamental things people want to do with a mobile device: messaging, contacts, calendars, and notes. For all that modern devices can do, these four essentials seem more or less the same in concept and function between my MessagePad 120 and my iPad Pro, despite 30 years between them.

Games, stream video content, entertain young children at restaurants and on planes, edit/produce content for social media, ...

It has changed somewhat, but the biggest thing is connectivity. If ubiquitious connections were around then, like they are now, the Newton may have had a very different story.

It is somewhat ironic that I can actually write this on an iPad using Scribble nearly 30 years later and after Apple killed the Newton, reinvented its mobile devices, and declined to add a stylus to them...

Incidentally, it recognizes "egg freckles" just fine.

I suppose it took a bunch of innovations to make it more usable over that past thirty years, though.

- vastly improved software - much better accuracy and integration with "custom" words (such as contact book addresses)

- bigger screen - no hitting the edge every few words

- lower latency/higher resolution - ink appears as you're writing, providing better and instant feedback

I loved my Newton. I loved it so much that even when it was stolen out of my car in ~2003, I got a MessagePad 2100 off of eBay to replace it.

It was only when the iPhone came out that someone was able to pry my Newton out of my hands.

I’ve always wanted to get one of these old Apple Newtons and trying to upgrade the built in handwriting recognition system with something a bit more modern, within the limitations of having 8mb of ram :). Anyone have any experience hacking them?

I think it would be more interesting to run under emulation. The hardware is getting a bit old and creaky these days. Under emulation it could enjoy a considerable CPU and RAM boost.

An iPad with a pencil would be an ideal target.

(Given that it uses an ARM processor, maybe you wouldn't need to emulate it, on A and M-chip devices.)

> (Given that it uses an ARM processor, maybe you wouldn't need to emulate it, on A and M-chip devices.)

The Newton has a 32-bit ARM processor, whereas all of Apple's processors for the last ~5 years have only implemented AArch64.

Here you go: https://github.com/pguyot/Einstein

You need to locate the ROM. The Internet Archive has some: https://archive.org/download/AppleNewtonROMs

Delightful!

My first job after college was maintaining Newton script mobile application used by my employer.

Later we moved the app over to a windows CE device and developed an office management suite in the 4D programming language to compliment it.

Eat up Martha

(one of the 200,000 former Newton owners here)

Reply on news.ycombinator.com