It’s interesting that some engineers went on to work with Be, and then a bunch of engineers from Be went (back, for some of them) to Apple early in the MacOS X era. There are things you can trace back to the Newton across companies like that.
Another thing was text selectors, where you could do things with random text, depending on what it was supposed to represent (open the web browser when taping a link, open the phone app when tapping a phone number, track parcels, show a place in Maps, things like that). Nowadays, it is everywhere, but it was a nice feature on the Newton.
IIRC the “data detectors” patent[0] from the Newton era was still valid when the iPhone came out, and was one of the few patents specifically wielded by Apple in court against Android and Samsung.
AFAIK this was the inspiration for Microsoft's database filesystem attempt that almost killed a Windows release and ended up never seeing the light of day. I can't even remember its code name.
I still have my MP2000, and it powers up just fine. Although, it could handle year 2000...it seemed to hit the wall from a date range perspective soon after. Anyhow, I was soooo stoked at the time to get one. Back then, there was not a small, useful, handheld device. People that carried them were weirdos. I doubled-down and carried a giant Moto flip-phone too. As fun as the Newton was, it never became my killer device. Why? No one else had one. I could not easily share all that clever stuff I had in my Notes or my Calendar. I firmly believe that if more people had them...you would have seen more adoption. Casio made a few very colourful WinCe devices, too...same fate. And they were huge... may as well carry a real computer...a laptop.
As soon as cell phones got "apps"... calendar especially, then I knew PDA's were dead. When iPhone was launched...well, I knew that it was the smaller, smarter, more connected Newton that was imagined way back then.
I still have a working eMate 300. Every now and then I start it up and play around a bit. It's neat to see how someone of it still lives on in iPad OS with the Apple Pencil.
Say what you will about Jony Ive but the eMate 300 is a beautiful design that also looks pretty functional.
It also seems to have a high-contrast reflective screen that you could use in broad daylight, something that challenges even my rather bright MacBook Pro.
As you note the iPad is something of a successor to the Newton both in hardware (e.g. ARM CPU, Apple Pencil) and interface (e.g. shape recognition/drawing, long press.)
Apple apparently prototyped larger Newton tablets (VideoPad) as well as pen-based Mac OS devices (PenLite) but never brought them to market. Though macOS does work as a pen based system if you plug in a Cintiq, or an iPad with Sidecar.
I bought an eMate on a whim about 20 years ago and absolutely love it. I do a similar thing and have a play once in a while. Occasionally I’ll use it as a distraction-free writing device, but the batteries need replacing for it to be truly useful. Will get round to it one of these days.
I can’t help but think it would make a really good Raspberry Pi casemod, but it feels like sacrilege to gut a working device.
Incidentally, if anyone has any broken eMates I’d be interested…
Newtons, at least the 2000/2100, and possibly the 130 IIRC, supported PCMCIA Ethernet cards, including both WaveLAN IEEE (first gen 802.11 DSSS) and Proxim RangeLAN (first gen 802.11 FHSS) wireless LAN cards.
But the NewtonOS user experience was very much based around an occasionally-connected model, with syncing between the PDA and a "master" device or beaming between PDAs, and built-in support for inbox and outbox stores that would accumulate items to be dealt with once connectivity was lost or regained.
So although there ended up being a bunch of TCP/IP-based Internet stuff for Newtons, it was very much "bolted on the side", even more so than MacOS and Win95 around the same time, because the interaction model was hard-wired against constant connectivity.
Newtons also supported an infra-red version of AppleTalk, using an adapter connected to the built-in AppleTalk connector. This supported the usual AppleTalk network services -- wireless printing was the most useful.
Unfortunately, I think Apple lost a lot of the institutional knowledge from the Newton project before the iPhone was developed, or perhaps it was never that widespread even inside Apple. But perhaps most crucially, Jobs hated the stylus, and thus we got multi-touch gestural input, which is perhaps the major differentiating feature that the iPhone popularized.
Imagine if Netscape Navigator had arrived first on the Newton, along with a WAN card.
> Unfortunately, I think Apple lost a lot of the institutional knowledge from the Newton project before the iPhone was developed, or perhaps it was never that widespread even inside Apple
In the bad old days, you would have to restart your computer as well. I remember Netscape frequently crashing back in the day and that bomb icon from Mac OS System 7.
> Unfortunately, I think Apple lost a lot of the institutional knowledge from the Newton project before the iPhone was developed, or perhaps it was never that widespread even inside Apple.
What makes you believe that?
> But perhaps most crucially, Jobs hated the stylus, and thus we got multi-touch gestural input, which is perhaps the major differentiating feature that the iPhone popularized.
If the iPhone had supported stylus input from day 1 touch wouldn't have evolved nearly as well. Developers would say "just use the stylus" and been done with it. Modern mobile had to go through an evolutionary phase where touch was the only input method to get where it is today.
Finally, foreshadowing arguments about why it was important to use 64-bit computing in a handheld device, Mike Culbert explains 32-bit advantages in Newton over the incumbent 16-bit of the time: https://beepdf.com/wp-content/uploads/newton/COMPCON-HW.pdf.
Wow. I knew Apple was an investor in ARM, but the specifics are crazy. For those that won't read the linked story, this I found to be fascinating:
> Newton was not successful, but Newton actually made $800 million dollars because Apple eventually sold the 43 percent it owned in ARM, which, by the way, kept the doors open at Apple, just before Steve Jobs came back. It was one of the really important decisions that Gil Amelio [the last CEO before Steve Jobs returned] made, and it gave them the cash to buy NeXT.
This is extraordinary. Picking ARM for Newton basically enabled all of Apple's modern success: keeping the lights on, getting Steve Jobs back, NeXTSTEP evolving into Mac OS X and {i/iPad/TV/mac}OS, and of course ARM-based systems from the original iPod to the Mac Studio/M1 Ultra.
You know, it occurs to me in retrospect that Apple overpaid for NeXT, which is something people said at the time, but I kind of ignored. No one else would have bought them at anywhere near that price. The black hardware had failed. OpenStep was cool but the OS didn’t run super well on PC hardware. NeXT should have been $150M tops. Apple eventually had to rewrite all of the software anyway to make use of it.
Sure, Jobs himself was priceless. But, yeah, Apple did overpay for NeXT at the time.
You know how YC investors pick teams, not products? Apple at the time bought a team, not the technology. They also bought credibility by bringing a founder back on-board.
Apple's biggest problem at the time was bad management, followed by credibility.
They solved both problems and got some technology (ObjC, what would become Darwin, etc) as a free bonus.
The complete NeXT takeover of management made it more like a reverse acquisition in any case and the results speak for themselves.
ISTM that everyone misses a vitally important key part of NeXT.
Yes, there was the OS, but there were other OSes, notably BeOS, and NeXTstep/OpenStep ran on other OSes, notably Solaris.
Yes, there was Steve Jobs.
But NeXT had a core thing that nobody else did: the state-of-the-art development tools.
Apple needed to get Mac developers on-board with the new OS. Sun had a great OS. Be had a great OS. Apple itself had A/UX. OSes abounded; there was lots of choice. Jobs' vision was very important.
But without the NeXT developer tools, notably Interface Builder and Objective-C, the deal would not have saved Apple.
Most of the stuff below the UI layer was the same.
One thing that changed was IOKit, rewritten in C++, but there actually wasn’t a good reason for that and I think everyone agrees sticking to ObjC would’ve been better…
I had a 2100 (I think it was actually a 2000 with the memory upgrade) in the early 2000s. It was a fun toy, but I remember most the challenges of getting to be play like a first-class citizen with my Mac†. I found it to be a cool device in the tinker-y way that I like messing around with Linux and Raspberry Pis, but at the end of the day it required more attention than was worth it.
†Excepting the iPhone, this has been the case for virtually every device I've owned. Palms (after the discontinuation of Palm Desktop), Smartphones prior to 2007, Android phones. 3rd Party vendors make a genuine strong attempt to make everything work, but first-class support really makes using a device normal, instead of a chore.
I owned two Newtons. The jokes about the handwriting recognition are a bit overstated, especially as later apps made it a joy to use. Also, Newton was way ahead of its time, and though it looks quaint today, it was really something back in the day.
I had a Newton back then. Most of those who joked about the handwriting recognition were tech reporters who played with it for all of five minutes. Once you used the device on a regular basis, the recognition got better, similar to today's iPhone keyboard.
The handwriting recognition was running on the CPU so it got much better when they jumped from the 20 mhz ARM on the MP 1x0 to the 162 mhz StrongARM on the later models.
my friend went to a Newton startup.. there was nothing wrong with the product and they did a lot of things right. There were some shenanigans among management but not huge.. the company went bankrupt and the engineers were paid relatively low wages
I spent one memorable Christmas-to-New-Years holiday tracking down a bad bug in the Newton kernel that (for a while) only happened up on those kiosk devices. I spent nearly two weeks tracking down a timing window where an interrupt between two ARM instructions would cause the scheduler to stop scheduling threads. Very Heisenbuggy.
The fix was to swap those two ARM instructions. It's simultaneously the hardest and the most trivial bug I have ever fixed.
Going through this article I realise that apple had tonnes of innovators and a culture for them. No wonder it became what it became after Steve jobs took cover but I feel that the modern giant tech companies (except Tesla probably IMO) do not have this culture now
It’s just the opposite. Apple became successful after Jobs came back because he killed all of the “innovations” and started focusing on products that the market cared about.
Sure QuickDraw GX, OpenDoc, PowerTalk, etc. might have been “innovative”. But they weren’t what consumers wanted.
The iMac on the other hand especially running pre-OS X was technically crap at the time. But it saved Apple.
I think this clip when Jobs killed OpenDoc explains it all.
Innovating too early usually fails. Wait too long and it's not innovation. There's an art to innovating enough that people want to buy what you're selling given the limitations of hardware/networks/ecosystems at the time you bring the product to market.
OpenDoc is kind of an engineers dream. I remember being fascinated by it at the time (having only read about it). But yeah, it was a distraction. Microsoft also made large investment in OLE (object linking and embedding) at the time which was a similar technology. In retrospect these were solutions in search of a problem. Great that you could embed an Execl sheet in a Word document, but it wasn't what made consumers excited at the dawn of the web.
Relatedly, there's Steve Jobs's thoughtful yet off-the-cuff response when questioned at the 1997 Worldwide Developer Conference about Apple killing OpenDoc.
One would hope it isn’t - the problem with component systems is they’re a Conway’s Law violation, which is the reason the app model works so much better. (in terms of actually shipping, not technical correctness)
Never forget that Apple big-footed an individual to take away newton.com from him. IIRC, "Mark Newton" (edit: ?, from memory) had the domain and Apple took it. Can't find a reference to the story now.
Good (and sobering) to also remember the dark side that seems to go along with much (or is it all?) of human achievement.
The sad/interesting part is, that some of those dark moments seem so unnecessary - at least in hindsight. But I suspect that the underlying psychology of ruthlessness may be a requirement for much (all?) larger achievements.
I'm torn on this one, because at some point "first-come, first-served" is not a great thing in the realm of non-fungible domain names. I can't find much information on this subject, but what did Apple do?
This is really strange: I thought maybe I could use the wayback machine to look at the old site, but it only goes back to Nov 1998, where it has Apple content on it. Doing various searches I can't find a trace of the story. I searched the history on slashdot, which is probably where I heard about it, but no reference. Maybe it was Usenet...
But I can't find any references to it anywhere.
Much of the outrage at the time was: Hey, Apple, you already have Apple.com, why aren't you using apple.com/newton ? The owner of newton.com seemed to have as much right as Apple did to use the domain (it seemed to be his legal last name), and IIRC he had it for years before Apple released the Newton.
A bulk of the remaining animosity at the time was HOW Apple went after it. They didn't offer to buy it or negotiate with the guy that had it. Instead they went with: "This is ours now." through ICANN or whatever it was at the time.
Very similar to the nissan.com situation, except, you know, the individual still has nissan.com.
This was probably around the time that an article came out about domain squatting, which ended with: Does this mean that I could just go out and register mcdonalds.com ? E-mail me with your thoughts at: ronald@mcdonalds.com :-)
I built a lot of fun stuff for the Newton. I built a Sherlock clone for the Newton (which 15 years later landed me in the middle of the Apple vs. Samsung patent lawsuit). I also wrote the Java virtual machine available for the Newton. It was a lot of fun for a bored PhD student. My Newtons have recently all died and it's a little sad.
I have a couple here that I could sell you, assuming the shipping isn't awful (I'm in Canada). I tried to put them on eBay some years ago but got low-balled and jerked around. Haven't tried them lately, not sure of their condition.
One is an original MessagePad in box, the other a late model, I forget, I'd have to go look.
Out of curiosity what was your list price? I’ve been starting to collect (got a Mac IIsi with a portrait monitor a few months back) and would be interested in getting a MessagePad in working condition
UPDATE: I'll spend some time in the basement today going through my crap and I'll find the two Newtons, and do some research on what they might be worth based on condition and what others have gone for.
It's time for me to purge anyways. Prefer these go to a nerd's home who will have fun with them, esp if said person has a history with the Newton.
Around 1998 I built an open source application on the Newton called Hemlock which loaded, parsed, and used Sherlock search templates to grab data from internet sites just like Sherlock did. Sherlock can either search from the internet or can search from files. The Newton didn't have a filesystem: it had a searchable database system. One of the things I was working on, and proposed on Usenet, was to search both the internet and the Newton database system in one shot, though I eventually didn't do it as the Newton already had a great search tool for the database system built in.
About 15 years later I walked into my lab and got served with a subpoena by Samsung's legal team for deposition. Everybody's lawyers suddenly wanted to talk to me. They dug through my postings, my software, old files on my laptop, even the old Newtons in my office. Between Apple, Samsung, and the very helpful SFLC legal team, the whole thing was like being dropped into a tank of very nice sharks.
One of the patents Apple was going after Samsung for was 8,086,604, the "604" patent. Hemlock, and my proposals on the Usenet for improving it, predated this patent by quite a bit and easily covered most of its claims. Eventually Samsung decided to argue that they didn't violate the patent, rather than argue that it was invalid. That turned out to be the right decision (they won). But had they gone the invalidation route, I'd likely have found myself on the stand in SF. That patent is still live, and my prior art is still ready to invalidate it.
After the other user mentioned your name I thought it sounded familiar too so I looked you up. I used MASON quite a bit in the past, thanks for all your work on that!
> That patent is still live, and my prior art is still ready to invalidate it.
I wonder how many other invalid patents are out there. It boggles my mind that the patent office will grant anyone a patent without even checking for prior art.
It took a team of very expensive lawyers to find that prior art on some old forum. I don't think the patent office has those kinds of resources for each application.
If they don't have the resources or knowledge necessary to establish prior art, then they do not have the competence to grant patents. Their negligence enabled the illicit enrichment of monopolists and dragged open source developers needlessly into legal procedures, costing them time and possibly money in legal fees.
That is why I think every patent submittal should come with a bounty to be paid to anyone who identifies prior art to invalidate it. The bounty is setup using a bond or something similar. If no prior art is identified within a reasonable time (say 2 years), the bond is reverted back to the submitter.
Maybe the most interesting/memorable part of my brief foray into the Newton was being invited to the Apple Campus and meeting other engineers (I was still so young then). We were each paired with a weird development Newton that was some kind of PCB bento.
It was cool to see how the sausage is made (or at least what dev prototypes can look like).
I remember too the one engineer I chatted with in the evening that told me, "You could work here." That really surprised me — I thought you had to have a PhD or something to work for Apple.
(Not too many years later, I would in fact get a job at Apple.)
Author here! I'm really happy to see people getting into the nostalgia about these weird and wonderful devices. Researching the story was absolutely fascinating, and I loved getting to talk to Steve Capps, who was still excited to talk about the work he did all those years ago.
> And NewtonScript influenced the creation of JavaScript, with its prototype-based object model, dynamic variable typing, garbage-collected memory, and fast interpreted design. Today, JavaScript is the most popular programming language in the world.
The reasons are understandable though (sorry if there are any slight mistakes as I'm writing from memory here).
Brendan Eich was hired to create a Scheme in the browser. Later, the requirements were changed to "make it look like that new Java language". He has said on record that he was familiar with the Self language. JS is often called "Lisp in C's clothing" because under the surface, it is quite similar to Scheme with a prototype object system (Scheme has no object system by default).
Apple created a language in the 1990s called Dylan that was also used on the Newton. It started out as something very close to Scheme + CLOS (Common Lisp Object System though it's more properly MOP -- MetaObject Protocol). Lots of programmers hate the lisp lack of syntax for whatever reason, so they changed Dylan to an infix-style language. OpenNewton is still a thing if you're interested.
NewtonScript kept much of the Dylan-style syntax, but changed out the complex MOP for the more easy to understand prototypal inheritance. I've never encountered even the hint that Newtonscript was on Eich's radar when he started work on Javascript.
There are definitely similarities that become apparent when you look at both language designer's inspirations:
Scheme + Self + C/Java = JavaScript
Scheme + Self + Dylan = NewtonScript
The result is two languages with the same core language ideas from Scheme, the same object system from self, and very different syntax inspirations from C and Dylan.
EDIT: Eich says in an interview he never saw NewtonScript before designing JS
Early Dylan owed more to Common Lisp than to Scheme, especially in its object system. It removed some of the more complicated CLOS mechanisms (e.g. :before, :after, and :around methods) but it still "felt" a lot like Common Lisp before its syntax was de-parenthesized.
The first Dylan compiler was written at Apple Cambridge [Massachusetts] Laboratories in Macintosh Common Lisp (MCL) which was itself an Apple Product. Later, Apple decided they didn't really want Dylan after all so they sold the Cambridge lab, later renamed Digitool. The MCL product lives on today as Clozure Common Lisp (CCL).
The transition from MCL to CCL is a dramatic story unto itself. I should probably write down what I remember about it someday.
I played around with Dylan around a decade ago. Maybe my Scheme comparison comes more from it having just one namespace (lisp-1) rather than the more functional aspects of Scheme. MOP is definitely from CL though.
For the record, NewtonScript wasn’t inspired by Dylan. I did work on Dylan, but only in the early stages when it had normal Lisp syntax. The syntax inspiration for NewtonScript was Pascal.
I suppose if they'd made it the size of an A4 folded along its major axis it would have been easier to hold, though reading text would have been annoying.
I recently attended a sales meeting for group health insurance reps at Blue Cross Blue Shield of TX.
They are going all-in on HMO plans. Of course everyone remembers how unpopular HMOs were in the 90s, deservedly so or not.
The Apple Newton figured prominently in their presentation. Like the Newton, they argued, the HMO was premature to the market, and suffered from lack of supporting infrastructure, despite being ahead of its time in many ways. Today's HMO could be the iPhone.
I wasn't convinced, but it was cool to see the Newton again, out in the wild.
Related, the documentary "General Magic" is quite good, and mentions how Apple's Newton sort of undercut the proto-smartphone being developed by apple veterans at General Magic. So many people saw the future but the tech wasn’t quite there.
Yeah, wondering what Larry Tesler's thought process was here, the people from GM seemed pretty hurt by it in the doc (worth watching but not very critical overall).
Sadly I almost never saw one in the wild, even in the Valley, but TBH, hardware and software hadn't yet reached the point of viability.
I'm glad the author mentioned the knowledge navigator concept video/s (were there two or just one? I can't remember any more). The iphone handily picked up that baton, but wisely didn't try to do so right away. But the NN vision was almost 20 years too soon.
I bought a new Newton after having lunch with Larry Tesler. He was both enthusiastic about the Newton as well as trying to talk me into rewriting my first Springer Verlag Common Lisp book to use the Dylan language. John Koza also had lunch with us, and it was great fun.
Years later, I was cleaning up the closet in my home office and ended up throwing away all Newton materials - my general rule that if I haven’t used something in a few years, I like to get rid of it.
I told him about my experiment of using GA to train small recurrent networks (using what I called variable weight size, that would increase during training). He said that it was a very cool idea but it wouldn’t scale. A few weeks later I agreed with his almost instant analysis. I did use the code as an example in my C++ Power Paradigms book.
Why wouldn’t training small recurrent networks with GA scale? Genetics algorithms are embarrassingly parallel, even more so than training neural networks by gradient descent. It’s not efficient sure but it’s trivially scalable.
I miss soups and NewtonScript. Honestly, it was such a nice little machine that was so close to getting it right. I still think some of the big ideas of that machine would serve us well today. A replicating soup would be fun.
Not feeling old yet?? ... next year is the 40th anniversary of PBS science TV show Newton's Apple, with Ira Flatow. (The one with the theme music 'borrowed' from Kraftwerk.)
184 comments
[ 3.4 ms ] story [ 257 ms ] threadAnother thing was text selectors, where you could do things with random text, depending on what it was supposed to represent (open the web browser when taping a link, open the phone app when tapping a phone number, track parcels, show a place in Maps, things like that). Nowadays, it is everywhere, but it was a nice feature on the Newton.
http://preserve.mactech.com/articles/mactech/Vol.14/14.01/Ja...
[0] https://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=...
https://en.wikipedia.org/wiki/WinFS
https://eeggs.com/items/538.html
It also seems to have a high-contrast reflective screen that you could use in broad daylight, something that challenges even my rather bright MacBook Pro.
As you note the iPad is something of a successor to the Newton both in hardware (e.g. ARM CPU, Apple Pencil) and interface (e.g. shape recognition/drawing, long press.)
Apple apparently prototyped larger Newton tablets (VideoPad) as well as pen-based Mac OS devices (PenLite) but never brought them to market. Though macOS does work as a pen based system if you plug in a Cintiq, or an iPad with Sidecar.
Thomas Meyerhoffer designed the eMate 300. Jony designed the second-generation Newton MessagePad, the Newton MessagePad 110.
(If HN didn't have its stupid edit locks I could edit it to read "say what you will about 1990s Apple.")
In any case, the eMate looks like an amazing design and now I want to see more Meyerhoffer designs!
https://bslabs.net/2018/05/28/wwdc-1997-videos/#500
I'd still like to hear more about Thomas Meyerhoffer though.
I can’t help but think it would make a really good Raspberry Pi casemod, but it feels like sacrilege to gut a working device.
Incidentally, if anyone has any broken eMates I’d be interested…
Merging them with cell phones, cameras, and music players also turned out well.
And that ARM CPU turned out to be a good investment, in more ways than one.
But the NewtonOS user experience was very much based around an occasionally-connected model, with syncing between the PDA and a "master" device or beaming between PDAs, and built-in support for inbox and outbox stores that would accumulate items to be dealt with once connectivity was lost or regained.
So although there ended up being a bunch of TCP/IP-based Internet stuff for Newtons, it was very much "bolted on the side", even more so than MacOS and Win95 around the same time, because the interaction model was hard-wired against constant connectivity.
Newtons also supported an infra-red version of AppleTalk, using an adapter connected to the built-in AppleTalk connector. This supported the usual AppleTalk network services -- wireless printing was the most useful.
Unfortunately, I think Apple lost a lot of the institutional knowledge from the Newton project before the iPhone was developed, or perhaps it was never that widespread even inside Apple. But perhaps most crucially, Jobs hated the stylus, and thus we got multi-touch gestural input, which is perhaps the major differentiating feature that the iPhone popularized.
> Unfortunately, I think Apple lost a lot of the institutional knowledge from the Newton project before the iPhone was developed, or perhaps it was never that widespread even inside Apple
:(
In the bad old days you'd have to restart your web browser because it crashed.
Now you have to restart your web browser because it has consumed all system resources, spun up the fans to maximum, and slowed performance to a crawl.
And those weekly severe security vulnerabilities don't crash your browser anymore - they just enable it to be taken over by criminals.
While MS did do some crazy stuff to kill Netscape, IE was just a better browser.
IE 3.0 for the Mac was native code and the most standards compliant browser in existence for awhile.
What makes you believe that?
> But perhaps most crucially, Jobs hated the stylus, and thus we got multi-touch gestural input, which is perhaps the major differentiating feature that the iPhone popularized.
If the iPhone had supported stylus input from day 1 touch wouldn't have evolved nearly as well. Developers would say "just use the stylus" and been done with it. Modern mobile had to go through an evolutionary phase where touch was the only input method to get where it is today.
In 1996, money to keep things going was important: https://www.nytimes.com/1996/03/28/business/apple-expects-it....
Finally, foreshadowing arguments about why it was important to use 64-bit computing in a handheld device, Mike Culbert explains 32-bit advantages in Newton over the incumbent 16-bit of the time: https://beepdf.com/wp-content/uploads/newton/COMPCON-HW.pdf.
> Newton was not successful, but Newton actually made $800 million dollars because Apple eventually sold the 43 percent it owned in ARM, which, by the way, kept the doors open at Apple, just before Steve Jobs came back. It was one of the really important decisions that Gil Amelio [the last CEO before Steve Jobs returned] made, and it gave them the cash to buy NeXT.
Sure, Jobs himself was priceless. But, yeah, Apple did overpay for NeXT at the time.
Not if you view the transaction as a package deal.
Apple's biggest problem at the time was bad management, followed by credibility.
They solved both problems and got some technology (ObjC, what would become Darwin, etc) as a free bonus.
The complete NeXT takeover of management made it more like a reverse acquisition in any case and the results speak for themselves.
Yes, there was the OS, but there were other OSes, notably BeOS, and NeXTstep/OpenStep ran on other OSes, notably Solaris.
Yes, there was Steve Jobs.
But NeXT had a core thing that nobody else did: the state-of-the-art development tools.
Apple needed to get Mac developers on-board with the new OS. Sun had a great OS. Be had a great OS. Apple itself had A/UX. OSes abounded; there was lots of choice. Jobs' vision was very important.
But without the NeXT developer tools, notably Interface Builder and Objective-C, the deal would not have saved Apple.
One thing that changed was IOKit, rewritten in C++, but there actually wasn’t a good reason for that and I think everyone agrees sticking to ObjC would’ve been better…
for example, NSString, NSApplicationMain, etc.
†Excepting the iPhone, this has been the case for virtually every device I've owned. Palms (after the discontinuation of Palm Desktop), Smartphones prior to 2007, Android phones. 3rd Party vendors make a genuine strong attempt to make everything work, but first-class support really makes using a device normal, instead of a chore.
The fix was to swap those two ARM instructions. It's simultaneously the hardest and the most trivial bug I have ever fixed.
Sure QuickDraw GX, OpenDoc, PowerTalk, etc. might have been “innovative”. But they weren’t what consumers wanted.
The iMac on the other hand especially running pre-OS X was technically crap at the time. But it saved Apple.
I think this clip when Jobs killed OpenDoc explains it all.
https://donhopkins.medium.com/focusing-is-about-saying-no-st...
https://www.youtube.com/watch?v=oeqPrUmVz-o
One would hope it isn’t - the problem with component systems is they’re a Conway’s Law violation, which is the reason the app model works so much better. (in terms of actually shipping, not technical correctness)
The sad/interesting part is, that some of those dark moments seem so unnecessary - at least in hindsight. But I suspect that the underlying psychology of ruthlessness may be a requirement for much (all?) larger achievements.
Was it similar to the Nissan case? https://en.wikipedia.org/wiki/Nissan_Motors_v._Nissan_Comput...
But I can't find any references to it anywhere.
Much of the outrage at the time was: Hey, Apple, you already have Apple.com, why aren't you using apple.com/newton ? The owner of newton.com seemed to have as much right as Apple did to use the domain (it seemed to be his legal last name), and IIRC he had it for years before Apple released the Newton.
A bulk of the remaining animosity at the time was HOW Apple went after it. They didn't offer to buy it or negotiate with the guy that had it. Instead they went with: "This is ours now." through ICANN or whatever it was at the time.
Very similar to the nissan.com situation, except, you know, the individual still has nissan.com.
This was probably around the time that an article came out about domain squatting, which ended with: Does this mean that I could just go out and register mcdonalds.com ? E-mail me with your thoughts at: ronald@mcdonalds.com :-)
One is an original MessagePad in box, the other a late model, I forget, I'd have to go look.
It's time for me to purge anyways. Prefer these go to a nerd's home who will have fun with them, esp if said person has a history with the Newton.
Please, tell us more.
About 15 years later I walked into my lab and got served with a subpoena by Samsung's legal team for deposition. Everybody's lawyers suddenly wanted to talk to me. They dug through my postings, my software, old files on my laptop, even the old Newtons in my office. Between Apple, Samsung, and the very helpful SFLC legal team, the whole thing was like being dropped into a tank of very nice sharks.
One of the patents Apple was going after Samsung for was 8,086,604, the "604" patent. Hemlock, and my proposals on the Usenet for improving it, predated this patent by quite a bit and easily covered most of its claims. Eventually Samsung decided to argue that they didn't violate the patent, rather than argue that it was invalid. That turned out to be the right decision (they won). But had they gone the invalidation route, I'd likely have found myself on the stand in SF. That patent is still live, and my prior art is still ready to invalidate it.
That name rang a bell so I had to look you up. I just have to say, I love your book!
I wonder how many other invalid patents are out there. It boggles my mind that the patent office will grant anyone a patent without even checking for prior art.
Was "Sherlocked" a term coined yet when you did this? Had Apple started getting inspiration from 3rd party developers yet?
Maybe the most interesting/memorable part of my brief foray into the Newton was being invited to the Apple Campus and meeting other engineers (I was still so young then). We were each paired with a weird development Newton that was some kind of PCB bento.
It was cool to see how the sausage is made (or at least what dev prototypes can look like).
I remember too the one engineer I chatted with in the evening that told me, "You could work here." That really surprised me — I thought you had to have a PhD or something to work for Apple.
(Not too many years later, I would in fact get a job at Apple.)
Dual inheritance for the win... _proto/_parent
> And NewtonScript influenced the creation of JavaScript, with its prototype-based object model, dynamic variable typing, garbage-collected memory, and fast interpreted design. Today, JavaScript is the most popular programming language in the world.
The reasons are understandable though (sorry if there are any slight mistakes as I'm writing from memory here).
Brendan Eich was hired to create a Scheme in the browser. Later, the requirements were changed to "make it look like that new Java language". He has said on record that he was familiar with the Self language. JS is often called "Lisp in C's clothing" because under the surface, it is quite similar to Scheme with a prototype object system (Scheme has no object system by default).
Apple created a language in the 1990s called Dylan that was also used on the Newton. It started out as something very close to Scheme + CLOS (Common Lisp Object System though it's more properly MOP -- MetaObject Protocol). Lots of programmers hate the lisp lack of syntax for whatever reason, so they changed Dylan to an infix-style language. OpenNewton is still a thing if you're interested.
NewtonScript kept much of the Dylan-style syntax, but changed out the complex MOP for the more easy to understand prototypal inheritance. I've never encountered even the hint that Newtonscript was on Eich's radar when he started work on Javascript.
There are definitely similarities that become apparent when you look at both language designer's inspirations:
Scheme + Self + C/Java = JavaScript
Scheme + Self + Dylan = NewtonScript
The result is two languages with the same core language ideas from Scheme, the same object system from self, and very different syntax inspirations from C and Dylan.
EDIT: Eich says in an interview he never saw NewtonScript before designing JS
https://wm-help.net/lib/b/book/2520193410/6
Did you mean Open Dylan?
The first Dylan compiler was written at Apple Cambridge [Massachusetts] Laboratories in Macintosh Common Lisp (MCL) which was itself an Apple Product. Later, Apple decided they didn't really want Dylan after all so they sold the Cambridge lab, later renamed Digitool. The MCL product lives on today as Clozure Common Lisp (CCL).
The transition from MCL to CCL is a dramatic story unto itself. I should probably write down what I remember about it someday.
In other words, the size of an A5 sheet of paper. C'mon Ars!
They are going all-in on HMO plans. Of course everyone remembers how unpopular HMOs were in the 90s, deservedly so or not.
The Apple Newton figured prominently in their presentation. Like the Newton, they argued, the HMO was premature to the market, and suffered from lack of supporting infrastructure, despite being ahead of its time in many ways. Today's HMO could be the iPhone.
I wasn't convinced, but it was cool to see the Newton again, out in the wild.
This was years before Steve Jobs' (in)famous big unveiling events.
After that, I wanted a Newton so badly...never did end up getting one. Alas.
https://www.youtube.com/watch?v=2B-XwPjn9YY
I'm glad the author mentioned the knowledge navigator concept video/s (were there two or just one? I can't remember any more). The iphone handily picked up that baton, but wisely didn't try to do so right away. But the NN vision was almost 20 years too soon.
Years later, I was cleaning up the closet in my home office and ended up throwing away all Newton materials - my general rule that if I haven’t used something in a few years, I like to get rid of it.
I told him about my experiment of using GA to train small recurrent networks (using what I called variable weight size, that would increase during training). He said that it was a very cool idea but it wouldn’t scale. A few weeks later I agreed with his almost instant analysis. I did use the code as an example in my C++ Power Paradigms book.
The Newton Application Architecture (1994) [pdf] - https://news.ycombinator.com/item?id=22398899 - Feb 2020 (37 comments)
Newton Storage History (2007) - https://news.ycombinator.com/item?id=13805096 - March 2017 (11 comments)
Why Steve Jobs Killed the Newton - https://news.ycombinator.com/item?id=11486183 - April 2016 (1 comment)
Soup (Apple) – the Newton storage system - https://news.ycombinator.com/item?id=11079874 - Feb 2016 (1 comment)
Why I Carry a Newton - https://news.ycombinator.com/item?id=10877256 - Jan 2016 (138 comments)
Dash Board for Newton OS: a Comic Tragedy in Nine Acts - https://news.ycombinator.com/item?id=10083181 - Aug 2015 (6 comments)
A Guide to the Apple Newton - https://news.ycombinator.com/item?id=10079507 - Aug 2015 (20 comments)
The “personal organizer” we had before the Newton - https://news.ycombinator.com/item?id=9635168 - May 2015 (11 comments)
Patching the Newton - https://news.ycombinator.com/item?id=8125296 - Aug 2014 (17 comments)
Talking with Mikel Evins about the Lisp-based Newton OS from Apple - https://news.ycombinator.com/item?id=7781265 - May 2014 (19 comments)
Steve Wozniak on Newton, Tesla, and the original Macintosh - https://news.ycombinator.com/item?id=5955583 - June 2013 (2 comments)
Ask HN: Is there a FOSS implementation of Apple's Newton OS? - https://news.ycombinator.com/item?id=5643063 - May 2013 (1 comment)
Newton OS running on an iPad - https://news.ycombinator.com/item?id=1703130 - Sept 2010 (6 comments)
Apple Gives Tribute to Newton with New 'What is iPad?' Ad - https://news.ycombinator.com/item?id=1343523 - May 2010 (4 comments)
The iPad, the Newton, & the "Of Course" Model of Innovation Diffusion - https://news.ycombinator.com/item?id=1290914 - April 2010 (31 comments)
A bit of vaporware (or "Microsoft's Secret Newton Killer") - https://news.ycombinator.com/item?id=838496 - Sept 2009 (1 comment)
Mikel Evins about the Lisp-based Newton OS. - https://news.ycombinator.com/item?id=320646 - Oct 2008 (5 comments)
It's time for the Millennials to start scheduling their colonoscopies.
Welcome to old age!