General Magic is a film about the ‘90s startup that imagined the smartphone | Hacker News Reader

General Magic is a film about the ‘90s startup that imagined the smartphone

theverge.com

100 pointsrbanffy41 comments

I did a look back at using a Palm V as a young techie back in 1998/1999. It's interesting since I still had some records of screenshots from real-world apps I used on a daily basis back then.

http://www.pixelmonkey.org/2012/12/30/mobile-in-1998

The Palm V wasn't that far off from modern smartphones, when set up correctly. Of course, lacked color and cellular connectivity. But, form factor was very similar to modern smartphones. And use cases have evolved in roughly the same direction. Plus, due to the grayscale screen and slow clock speed, the battery life was pretty darn good!

The two major problems that hindered widespread adoption of the Palm V: (a) The requirement to use Palm Graffiti with a stylus for text and input[1], which had a steep-ish learning curve. (b) The slow serial port (!!!) sync to desktop for pretty much all "connected" functionality.

AvantGo[2], in particular, was very ahead of its time, and using it did feel like "living in the future". And, in fact, I was -- it allowed me to read news articles and digital web content 'on the go', even in the period 1998-2001.

The iPod would be released in 2001 and the iPhone in 2007. So, it turns out it was about ~5 years too early for the hardware/infrastructure tech to catch up to the (conceived) mobile software use cases.

[1]: https://en.wikipedia.org/wiki/Graffiti_(Palm_OS)

[2]: https://en.wikipedia.org/wiki/AvantGo

It's easy to overlook how critical infrastructure is in the ecosystem. In 1998 pervasive Wifi wasn't a thing, and virtually all noteworthy apps were desktop oriented. Generic cloud storage wasn't a thing yet, and household internet was connected directly to a PC anyways, so a serial connection to a PC for sync makes sense -- USB was still very brand new, while serial was pretty universal.

To me the interesting question is, given they had a functional device, how did they fuck up the next decade? Palm could have easily introduced many of the same features iPod had, and rescued 3Com/USRobotics from the dialup modem graveyard. Instead they spun off Palm for a forkload of cash, after driving away the founders to Handspring. It wasn't until 2004 that they supported any sort of sound beyond piezo buzzers with the Zire 31.

I haven't thought of WiFi as a available infrastructure until this year, the development was so slow to arrive I missed it actually coming.

And this is the City Of London, I live close enough to consider the city home.

Ironically, I only noticed the ubiquity of WiFi in the city as a result of testing the lowest end phone and prepaid options possible. This is in reaction to the most ridiculous dispute with my contract network provider, over the fate of the premium number I've used for a long long time being held hostage. I only wanted it to move to another account..

I remember the Palms very well. At the time I carried a treasured HP 209LX, secretly wishing for a Psion 5 keyboard but not their systems environment. (I had the HP a while if you note the overlap)

I wish I could find the story again, but I think the sadness was just the classical combination of blunt management and silicon valley egos.

This is the canonical answer for every technological near miss, isn't it?

I mean cost driven aggressive B School management plus idealistic impatient but always sufficiently talented to change the world software and hardware Engineers, Always Always Always the same, WHY DO YOU GUYS KEEP DOING IT??!!!

I honestly care about the answer, because I thought this was the past and not the normal, when I learned to program, oh, thirty five years ago...

As for the infrastructure WiFi...

The Result has been I learned that I am almost entirely independent of any cellular carrier network for my voice connectivity.

I haven't quite yet, but I actually could drop the Networks out of my life!

The UK is blessed with a ISP called Andrews and Arnold, www.aaisp.net who will port my (once freed) precious memorable number to a SIM card that resides on their MVNO. This cuts my recurring payment to merely £2.40pcm! And, as desired, I can have more lines on that number. I must avoid starting in how good AAISP really is. If you want to be recognised by voice when you call, and get enterprise grade services, please mention John owes Phil Boddy one, and hadn't forgotten...

Oh, and my Networks- free telephonic life?

I am seriously contemplating whether it's worth the hassle of applying to install a VHF/UHF D-STAR Repeater, which our building management seems happy to let us do, to cover the Bank / St. Paul's areas.

AAISP let you port your number (any number, cellular numbers are new) to their VOIP services. I can route them as I please, so apart from the slight lack of duplex, why not over HAM RADIO? I have figured out how to use a freephone number to call in and take over the call, from a payphone, and the idea amuses me greatly. But the fact is, I'd rather carry a dedicated WiFi access device that provides me with use of a headset and the radio /audio processing power possible in a shoulder strapped battery and module, than I would carry another$1000 phone obsolete the moment the manufacturer cares to not update it.

I guess I'd just love to be serious about putting my call sign on my business card!

I had a Kyocera 7135[1] back before the Treos took off, pretty incredible for the time. No concept of a "data" plan, all my bits came from dialing up a cellular line like old school modems.

I'd actually argue that the precursor to the modern smartphone was the Sidekick. It was one of the first ones with push IM, email(even for pop3!) that worked well with consumer's interests in mind. Sadly it never went anywhere other than t-mobile in the US. Andy Rubin did go on to build Android with large parts of that team so not all was wasted.

[1] https://www.phonescoop.com/phones/phone.php?p=173

The Sidekick was huge among the deaf community because that keyboard was great for texting.

Yeah, that keyboard was awesome and the sound/feel it made was solid when you let it snap out.

A lesser-known awesome feature of AvantGo was that it also functioned online. In early 2000 I was using AvantGo on a Palm Vx, IrDA connected to my cell phone, getting Internet access via my carrier's WAP gateway (PPP over CSD? I forget the specifics), to post to my 'blog and browse the web while riding the train.

It always made me think of those AT&T You Will commercials. I never sent a fax from the beach... but I sent a few from trains.

I love my palm III xe. My single application is to record go games (I still use it). A friend had bought one and was using it to take notes during meetings. I think it was a good product with a small market and with no obsolescence (my palm works as well as when I bought it). The market could have increased a bit by having better app stores, better apps, better features for the price (the price tag was high).

Great technology is not a product.

Let me repeat that: Great technology is not a product!

Founders would do well to remember. If you fail to find product-market fit, if you're thinking too far ahead, or if your product has too many compromises: you can invent ground-breaking technology that completely revolutionizes the industry then promptly file for bankruptcy and walk away with nothing.

Again let me repeat that: You can invent ground-breaking technology, be 100% correct about the future, and be 5-10 years ahead of everyone else... then watch your dreams go up in smoke as you walk away with zero dollars and without even a wikipedia blurb mentioning your accomplishments.

Great ideas without execution are worthless.

Great ideas plus poor execution are a deadweight loss on humanity.

> Great ideas plus poor execution are a deadweight loss on humanity.

That's overstating it. In general, the failed attempt helps propagate the good idea, and helps that idea find a better execution in the future, as well as being a cautionary tale.

I call this knowing if you're working on the 3rd or 4th wave of innovation in a space, or not.

Executing the right amount of your innovation for the corresponding wave of technology that the market is ready to uptake is critical. Or you'll be too far ahead, or not ahead enough.

Otherwise, the market usually waits for the 3rd or 4th generation of iPhone, Samsung Galaxy Note, Nexus, Apple Watch, iPad, etc before they become widely usable.

I love the stories of General Magic and NeXT. Even though the businesses basically failed, the technology and experience from them would go on to completely change the world :)

For a look at another company in this space around that time, Jerry Kaplan wrote a memoir about his experiences at GO: https://www.amazon.ca/Startup-Silicon-Adventure-Jerry-Kaplan...

Is there a case where ground breaking innovation resulted in a successful business? Usually innovation is too costly and the first incarnations lack mass appeal.

Most ideas are neither new nor novel. Even if you're the first the odds are high that you are far too early and the technology won't be ready for a decade or more.

Even the first toilets were far from successful. Indoor plumbing required a whole industry and multiple compounding innovations to become successful.

Many people were working on flying machines. The Wright brothers simply got lucky in timing: internal combustion engines finally got reliable and useful enough for the idea to work. If they hadn't done it someone else would have within the following 5-10 years (and many others were in fact trying at the time).

Many people had MP3 players but none of them managed to design a product ordinary people would want to use, nor the muscle to make the record industry play ball with an online store. You can scream until you're blue in the face that Apple didn't invent MP3 players but that just makes you irrelevant and a bad potential founder/employee.

I would argue that most truly ground-breaking innovations do not result in a wildly successful business. Finding a way to pair ideas with execution and product-market fit are more important.

The bicycle is yet another example.

A very nice 10-minute video covering the story of General Magic, the company:

https://www.youtube.com/watch?v=Opcuy-8VO64

LGR Tech Tales has a lot of these types of videos. Also Company Man. But Company Man is more general and not all Tech Companies.

One fad in the 1990's was agent-oriented computing. You'd have scripts that carried code and data from place to place. For instance, you might send a bidding agent with your strategies to eBay's marketplace where it would constantly look for stuff you'd be interested in doing only deals you'd want. In day and age of 28Kbps links, there were advantages to not having to constantly move data back and forth. One of most interesting options came from General Magic:

https://en.m.wikipedia.org/wiki/Agent-oriented_programming

https://en.m.wikipedia.org/wiki/Telescript_(programming_lang...

Telescript was fun, but those machines were awful buggy and the PC client was a bit odd. I still think agent oriented programming has a place with large program organization. We regularly run other people's code on remote machines every time we visit a web page so one of the objections to mobile agents isn't exactly different than today.

The opportunities for emergent behavior (or misbehavior) are abundant. That's a great approach to make systems that act in completely unpredictable ways.

Maybe, but after all the talk of training and not knowing why decisions are being made, I really want to keep anything resembling emergent at arms length. I very much more interested on multi-agent systems effects on resource allocation and maintainability.

I see your line of thinking as great and interesting, but I'm now in the camp that sees a lot of people blaming "algorithms" for computer decisions that have negative effects (banning or denying resources) on people and thinking that's ok they cannot explain their program's decisions.

At the end of the 90s my brother had a Spectronics phone that was amazing. It was controlled with hidden buttons on its sides, took some practice. But it had everything a smartphone would have many years later.

Looking at their wikipedia page I believe it was in the TS2000 series.

When did this General Magic stuff take place? The article keeps using the phrase 1990s over and over.

There's a Wired article (April 1994, their fifth issue) with more background and some timelines: https://www.wired.com/1994/04/general-magic/

Ok. SO it's interesting to note that 5 years later a Swedish company had a finished product. Even though the Spectronics design was very unique and not at all adopted by anyone else in the later smartphone era.

This reminds me of my Palm Pilot Pro from ~97. It came with a modem add-on for dialup internet. It was a smartphone, otherwise.

Handspring, an offshoot by Palm's creators created the first cellular addon (Springboard) for PalmOS shortly after. Worked OK.

PalmOS, on a Treo 90 (color touch screen in 2002), with Datebk6 remains the most productive device I've ever used by a wide margin. Touchscreen, and keyboard in one form factor cut down on grafitti use.

The best part? One shortcut key (fn+calendar) took the device from power off right into a cursor in to add a task. When horsepower didn't exist, productivity made up for in oodles. It had a usable keyboard, well before blackberry showed up. Nothing today touches this.

The Palm Treo 600 came along in 2003 to me were the first real, usable smartphones. No app store, but apps could be synced from the computer. It combined a physical keyboard with a touchscreen on a smartphone to boot.

I bought a Datarover from eBay on a whim a few years ago.

The device itself is a brick but Magic Cap is a delight: well designed, intuitive and fun in a way modern systems are not.

If the backlight hadn’t fizzled out I’d be using it regularly as an over-engineered landline.

I also did that some years ago. I used to be a heavy newton user and then, later in early 2000, I decided to check out some other old PDAs, so I purchased a data rover and some others. Magic Cap is a very good surprise. I wish we had more stuff like that these days instead of the boring stuff we do have now.

I agree. Using the Newton was fun in a way the modern devices simply aren't. The same happened with desktop operating systems. The original Mac was more fun than what we have now.

I look forward to seeing this. There was a lot of exciting and creative stuff going on in the 90s around handhelds.

The palm pilot is still probably the best series of PDAs ever.

Although maybe nostalgia is fogging my memory.

Early palms had a lot of issues - I remember swapping AA batteries very frequently for example - but the products and company showed an incredible balance between technological vision and practical reality. The glyphs are a great example; most of the letters mirrored the English alphabet but a few were quite different. With a little practice you could write accurate notes quite quickly. Compare this to the oft-mocked Newton. Sync also worked quite well. I wrote a number of apps that hooked into the sync process and uploaded data to web-based systems. I still find it kind of incredible that in ~2000 I could take notes on my palm and then dock, sync and view it in a web portal. Pretty underwhelming now but not when most websites were still geocities homepages!

For me my Pilot was a better organizer than any current phone. It was lean and very fast. The Newton 2000 was also very good.

Back when I was a kid in the late 90's I followed the emerging .COM boom and stock market avidly. I kept a virtual stock portfolio of companies I'd invest in if I could. General Magic was in there. Like many promising companies at the time though, it didn't survive the crash, so my stock pick turned out not so good (good thing I was only playing pretend). Nice to see they weren't completely forgotten though.

> Some inventions seem genuinely prescient, like a collection of animated proto-emoji stickers. Some are impractical but fascinating, like a “town” computing interface with buildings for apps.

I wonder how many other have never heard of Apple’s eWorld.

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

I wonder how many never heard of Quantum Link’s Habitat from 1986.

Also, the Packard Bell Navigator. I had so much fun in that interface when I was a kid.

These sorts of interfaces were very much of the atmosphere of the mid 1990s. Another somewhat better known one (although known as a flop) was Microsoft Bob. https://en.wikipedia.org/wiki/Microsoft_Bob

A friend of mine recently developed robert.js, a library for manipulating Bob's datafiles.

https://github.com/Treeki/robert.js

For example, here are the Bob rooms, rendered in your browser: https://wuffs.org/bob/room.html

The COMPAQ iPaq was in many ways ahead of its time. If they would have invested in a better OS, or convinced Microsoft to build something more "mobile", they could have beaten Apple possibly by 3-4 years.

https://en.m.wikipedia.org/wiki/IPAQ

FWIW, everybody in the PDA space was influenced by General Magic. Two others that come to mind were Rex https://en.wikipedia.org/wiki/REX_5000 and Geos https://en.wikipedia.org/wiki/GEOS_(16-bit_operating_system)

Oh, cool. I haven't heard about the REX. Thanks for sharing.

Reply on news.ycombinator.com