There’s a fascinating post at the StormDriver blog comparing the first mass-produced portable computer, the Osborne 1, with today’s hot-selling portable computing device, the iPad 2. The Osborne was launched 30 years ago this week, and of course the comparison with the iPad is about as stark as you would expect. The iPad is literally thousands of times better in some measurable aspects, costing about one eighth as much in inflation-adjusted dollars.But as the blog post points out, even though the Osborne company didn’t succeed, it proved that there was a market for portable computers, and the next couple of generations of portables stuck to the suitcase form factor that Osborne pioneered.
These days, the Osborne name is mostly known in marketing circles for the “Osborne Effect,” wherein anticipation for an upgraded product kills demand for the existing product. There’s no better modern student of this phenomenon than Steve Jobs himself, who has proven to be absolutely fanatical about keeping a lid on Apple product updates in order to avoid undermining the market for existing Apple products.
What this comparison got me thinking about, however, was whether the hardware or the software has been the more important driver of the computing revolution of the past 30 years. Obviously, it was impossible back then for a handheld device to be powerful enough to do more than simple arithmetic, and that put a severe limitation on what people could do with computers, especially portable ones. I don’t think anyone would disagree that big high resolution color screens, GUIs, 3D shading, and LiPo batteries don’t represent a huge leap forward for technology.
But as I was contemplating the Osborne software specs: CP/M, WordStar, SuperCalc, dBASE II, and thinking about what could be done with such limited hardware, I started to think about the radical software advances over the past decades: TCP/IP and the popularization of the internet, email, messaging, the web, search, social networking. You could actually take a 1980s-level hardware infrastructure and layer some software with modern capabilities on top of it and get something really quite amazing. You could search for information with Google, keep up with your friends on Facebook, share your ideas over Twitter, collaborate with your colleagues using email and messaging, play an Angry Birds-like game, make a complex spreadsheet, prepare a presentation for a speech. You’d have to give up VoIP, digital photography, 3D gaming, and all but the most rudimentary GUI, but you’d have a decent computing experience.
Now let’s consider the other scenario: We have modern multi-gigahertz, multi processor hardware, but 1980s level software: It’s too horrible to even contemplate.
Obviously, neither of these scenarios is plausible, because advances in software and hardware have gone hand in hand, but I’d venture to say that, especially when you consider how pervasive the internet has become since 1981, and the explosion of innovation that it spawned, the bigger revolution has been on the software side.
I recall going to an Osborne sales pitch with my Father. It was my first exposure to a spreadsheet. The machine was cool, but the screen and column limitations were telling. The price, however, was amazing.
I personally had much more success with an early Tandy Model 100, still one of most amazing machines created. I would pack it in my tank bag on my motorcycle, and travel “blog” with it on the road. Then upload the stories to the school BBS over its built in modem. With no moving parts, it easily handled the vibrations that typically accompany motorcycle travel. My cassette tapes, they’re not so lucky. They’d almost always have to be rewound to retighten them up after a ride so they could be played.
I always loved the keyboard on that thing, I fashioned cool little feet to give it a nice tilt by cutting the eraser ends off of a couple of pencils, and it really did run forever on those batteries.
The best feature about the word processor was that when you upload the story to someplace else, you could tell it a different column to wrap the lines at (say, 72) vs its 40 column display. You just put the BBS in to “post” mode, and hit the send key, at 300 baud.
Worked like a champ. Simple, fast, light. It would be less bulky today, but it was tough and a real workhorse.
I don’t know why, but your post reminded me of this comic: http://xkcd.com/533/
🙂
The original article reminded me of what this family had in 1967:
http://www.youtube.com/watch?v=aJ6SbvrjxZA
They already had cloudcomputing. 😉
Still a bit expensive, but I’m sure some people pay that for their broadband.
Ace link, some of it is true in the video … truely a first for Tomorrow’s World …
I haven’t worked out the numbers for the Osborne and the iPad but I did give a lecture a few years ago at a local university wherefor I estimated how much larger a problem could be handled by a recent computer than a 1980s one. All interesting problems are NP complete so I assumed an exponential order of algorithm (processing time and memory) and found that the modern computer can handle problems about 40% larger than a 1980s computer. Not really a substantial increase for all that extra memory and processor clock speed.
The basic problem is the modern obsession with digital computers. Most interesting problems are either analogue or close to analogue and the idea of converting the problem into the digital domain, running an inefficient algorithm in that domain and then converting the result back will always place a practical upper limit on the speed of computing.
So, yes, the iPad is faster at linear things (adding up a grocery bill, compiling my latest program, etc.) than the Osborne but it’s not much faster at interesting things. And, frankly, even most of the linear things (e.g., my word processor) are limited by my typing speed, for which the Osborne was fine.
But I agree that it is more portable.
Uhm, I hope you are joking? Most interesting problems are not NPC. Most interesting problems only become NPC when you phrase them as decisionproblems where you seek optimal solutions. In reality pragmatic solutions can get very close to the optimal solution and belong to P.
I vaguely remember seeing Osborn in a store, I think it was Jordan Marsh, Boston Downtown. But the first portable, or better luggable I actually used was Compaq, it has CRT display and it was a dedicated Sniffer box. Pretty heavy, but actually fairly high quality.
One of my friends had an Osborne. I had a lot of fun with it. It had a chess program that I could actually *beat* …..
Good times!
“You could actually take a 1980s-level hardware infrastructure and layer some software with modern capabilities on top of it and get something really quite amazing. You could search for information with Google, keep up with your friends on Facebook, share your ideas over Twitter, collaborate with your colleagues using email and messaging, play an Angry Birds-like game, make a complex spreadsheet, prepare a presentation for a speech.”
Being done as we speak, this is being made primarily for computers that need low power mainly for monitoring environments (e.g. water level in a Water tower) and has a wireless capabilities to talk to home. I believe they have a full TCP/IP stack for CP/M implemented.
http://hackaday.com/2009/01/10/n8vem-single-board-z80-computer/
Edited 2011-04-06 00:14 UTC
Yeah, there are plenty of old boxes from the 80s that you can now get online with and do just about everything the article talks about… granted, most of them are later 80s machines, but still.
Personally, I would consider something like Epson HX-20 (1981) or Tandy TRS-100 (1983) as a better contender for the first true portable! Those machines had excellent autonomy and were very light.
Triple negative gone wrong
. I think you want to ditch that last “don’t”. Or better, ditch the negatives alltogether.
Well, quite amazing for the limited capabitilites, yes. But the almost implied “we could’ve done that in the 80s” is of course not true – the infrastructure that’s needed behind all those on-line capabilities is of course humongous, and not achievable with 80s hardware.
Well, e.g. SymbOS has a fine GUI on 80s hardware.
It really depends on your definition of a fine UI
plenty os UIs out of the 80s were amazing for their time. And, dependent on exactly how much you had to blow on hardware, not all that different from todays.