I first came across this book in a library (a while ago) and basically read the whole thing on the spot. From my entirely biased perspective as someone with a "vintage computer" collection, it's the best book ever. The staggering magnitude of the accomplishments and inventions of PARC are almost as fascinating as fiction. The ineptitude of Xerox management in figuring out what to do with the output of PARC was similarly unreal.
Bell Labs was able to commercialize (or at least license IP) from some of their work with reasonable return, but eventually lost momentum from the series of spinoff-layoff-mergers. How does an organization successfully manage or balance that kind of research productivity with commercialization in the long term? Is there any management literature on that in particular?
I know some places like Microsoft and Google have had research programs for quite some time, although of course Google mostly fails to adequately support products that come out of Google X before killing them off or scaling them back (such as Glass), and I'm not sure how much of what Microsoft Research does is actually commercialized. Notably, Butler Lampson, who was one of the key folks at PARC, has been with Microsoft Research since 1995.
Fumbling the Future: not all that good; reads like a B-school case study
Dealers of Lightning: excellent history
Inventing the Future (by me!): what it was actually like to be there in the trenches. Fictionalized, but all facts are accurate. Foreword by David C. Smith, the guy who invented icons. That's his son on the cover, playing MazeWar on an Alto.
I commented Dealers in parallel, because of the narrative I'd argue it's one of the best tech histories every written. While Soul of the Machine was better reviewed, Dealers was more compelling.
I did enjoy your book by the way - thank you for writing it.
I can understand why - Tracy Kidder made everyone involved with the MV/8000 look bad. The narrative was not compelling either frankly, I've never understood why people found the book so compelling.
Dealers makes everyone look good more or less - like it clearly explained to me why Xerox could never figure out how to commercialize the PARC innovations. The single largest issue appeared to be the incentives the sales force had were not aligned to the kind of products PARC could develop. (though its clear there were other issues with the capacity of PARC to productize something)
A faster Alto with a general purpose operating system and the existing document preparation tools, probably would have significantly outsold the Star - but again, would the Xerox sales force known how to sell that product? I don't think so, and thats the underlying problem.
The Star appeared to have just suffered from the Second System Effect, which compounded the other issues Xerox had.
Very astute analysis. I cover these exact issues as they manifested.
I used to say "Star has done for object-oriented programming what Three Mile Island did for nuclear power."
Obviously not, though, since OOP grew & prospered on its own. I cover our struggles using OOP with a non-object oriented language in exhaustive detail, too.
I'll add the Innovators by Walter Isaacson to the list as a general take on persons involved in the history of computing. It was well written and the stories within were exciting to read.
LOL. Thanks. I made it cheap since it's my first book :)
"The Big Bucks" (about the 80's & the birth of the Internet) is going to be out in a few days, and it'll be more. Why? Because I think for my audience, the price doesn't matter as much, and so my income from writing might turn positive someday (and thus it won't be a "hobby" to the IRS).
Alan Kay, one of PARC's most prominent researchers, recommends Waldrop's book The Dream Machine. Actually I think he has an account here on Hacker News.
My recent study of history leads me to conclude that had Douglas Engelbart[1] been allowed to keep working on NLS[2], we'd have gotten a far better result. Instead his staff was poached, and PARC was the far less than optimal result. History shows time and time again that the best outcome rarely happens. We still don't have a Memex[3] as proposed by Bush in 1945.
"optimal" results are a fallacious indicator: optimal for now can mean a disaster in the near future, think about the actual just-in-time industry crisis for instance.
I do not say that's was happen since I do not really know nor NLS (of course I've seen the Mother of all the demos, but not much more especially in technical terms) nor Smalltalk workstations (I've dig a bit more there, emulation included, but lacking comparison with NLS...) however is a generally valid principles: in nature it does not win the better but the more adaptive. In humans terms some very good stuff might simply fade into oblivion because they are too ahead of their time for most people or are against someone's interests.
To spot a more modern example maybe a bit more easy to know by most HN-ers just see Java history, it's original idea and the subsequent practical fail due to the fact nobody really care the original idea adopting instead just the boilerplate. Plan 9 was another example.
IMVHO the real points was:
- pioneers have seen desktop computers, not services, with the human at the center, not a vendor or a service, more specifically connected people with powerful tool some in politics is named social tissue needed for democracy, something essentially all élites from all countries fear most;
- pioneers have cared about their idea, vision, not much their present time short terms needs, as a result most people have no idea about the future they envisage and have tried to meld what they are told about the future to their present created ridiculous stuff and encouraging some to go that way (see General Magik virtual office for instance);
- pioneers have not cared at all about scaling their gear, witch means how to make them en masse, creating a supply chain, ... as a result profiting from them was essentially an artisanal game no one can really sustain and at government level no one is much interested either.
Another today examples: we have seen in the recent 50+ years past various efforts to make small, cheap VTOL aircrafts, i.e. "flying cars". Now we see even financial institutions embracing them in their sauce, well so far such experiments have proven feasible and to a certain extent VERY interested for all. So far most think they are impossible. Useless. Unsafe. .... As a result the general reactionary behaviors of essentially any large community from élites (who feel well in general and fear loosing their status if something change) to generic human beings (who generally think the change is bad, not needed, they simply fear it if is a real change) annihilate or at least postpone in a very far future any real revolution.
Culture is typically the mean to make things happen but so far we seems to be unable (beside the fact that most do NOT want that happen) to really spread a substantial culture though the masses. We know how to do that in small cohort but not in larger ones. We still miss a way to connect those small cohort to the humanity enough to make the magic happen...
I've studied HCI and Engelbart's work for quite a long time, and I'm not sure I would fully agree that things would have been better had Engelbart continued working on NLS, or if NLS had won out.
One simple example is with copy and paste. NLS had a highly complex and error prone copy-and-paste system. A lot of the commands in NLS were modal (vs modeless), and thus error prone. See this article for more info:
http://worrydream.com/refs/Tesler%20-%20A%20Personal%20Histo...
Another issue was basic usability of NLS. Engelbart's vision was on continuous bootstrapping, continually aiming to create more powerful and expressive systems. This vision was pretty good for technical people, but also comes at the cost of basic ease of use and entry for newcomers. For instance, think about how many problems non-technical people still have with today's PCs and smartphones, many of which strive for "walk up and use" kind of usage.
Having been in research for over two decades now, my position is that we researchers will often identify the right problem and get the right general direction, but not always the right form. For instance, in the late 90s at Berkeley, folks correctly identified cluster computing as the future, and that we should aim to create abstractions to host Internet services to make them more scalable and robust. The specifics went in a very different direction (e.g. cloud computing, Kubernetes, etc), but the general thrust was right. As another example, my colleagues and I correctly identified rapid prototyping and testing tools for user interfaces as a major need, and while we inspired a lot of ideas with our work (pen-based sketching), industry went in a very different direction (see Figma and Axure and InVision).
I think the same is true with Engelbart's work. He was definitely a pioneer and helped open up people's imaginations as to what was possible with computing, but was off on a lot of the details and specifics that really matter for adoption. And that's ok because to a large extent, that's one of the many reasons we as a community and society do research.
I've heard that some people objected to his strong focus on "bootstrapping". In some ways, maybe we avoided a repeat of Babbage and his continual push towards something better at the cost of completion. It's hard to know from this perspective where the truth of the matter lies.
This is due to the different vision of what engineers/researchers and consumers want.
Researchers/Engineers want bottom up progress. They place incremental development at the heart of their system.
Partially since researchers have very little money.
They cannot do anything that requires 100 developers. They want a low hurdle rate, a low barrier to entry. To get started and make visible progress.
Engineers are constantly demanding progress. They want solutions that are state of the art, for various different reasons. The lots of small, incremental improvements to keep the cutting edge.
Consumers want the complete opposite. Consumers want whatever ecosystem has the most money and progress invested into it.
Then the consumer want it to never change. Ever.
They don't want to invest capital expenditure/attention span in learning new things. They have other important problems to deal with.
Then, when there is a deal-breaker scene change, they will jump to the next big thing.
> but was off on a lot of the details and specifics that really matter for adoption.
How can this be sped up? It feels like progress was made as fast as possible but in hindsight it's almost silly that there was not a computer in every household in 1970.
Moore’s Law and Dennard scaling were in full swing during that era. The thing about exponential growth is that most of the gains come at the end. The Alto cost $100K in 1973, the price for equivalent hardware halved every 18 months. That’s exactly what happened. By 1983 we had Mac for $1000 hardware cost / $2500 list price.
I think people forget that it takes 2 types of people to move tech or for that matter a company to greatness. The product is not the star. There's the dreamer that comes up with a product and then there's the one that sells the dream to the rest of us. They are usually BSer's to the nth degree. They will lie to your face and you will love them for that. They are soooooo loved by some that some men and women dream of having their baby. Think IBM,Apple,Ford,TESLA and many more.
Xerox PARC, just didn't have a great BSer in-house.
You forget a more broad stanza: false revolutions happen, like you say, real ones are feared and avoided by most. On one side actual élite (in business, politics etc, a generic tag to say "most well position somewhere in something") fear the substantial change because it might both succeed pushing them aside since they became useless or fail crushing the along if they have tried it; on the other side most people adverse changes.
Anyone might be pushed to buy a phone/something alike, considering it just a small addition to their daily routine, anyone might be pushed to buy a desktop/something else in general like the classic joke from the early web "hi! I want to buy a website" "hum, ok... To do what?" "Oh, I do not want to use it, all start buying one so I need one to remain in the market, is for customers not for me". But a substantial revolution is another story. Just see in daily life how many time you hear people say "I've always do this like that" and so on even if it's clearly an absurd thing or something that can&should be done but in a very different manner. Even small things.
At PARC time was the filing cabinet/suspended folder era, where people deeply think in paper+mechanic terms, the idea of files and folders (mimicking aforementioned filing cabinet and suspended folders) might pass easily, the idea of dynamic documents is simply unthinkable and fearful "I might loose something", "people might cheat", ... similarly the idea to send emails might pass easily, a better fax system, but the idea of computing inside again can't pass. For most a GUI can only pass, even today, as a single sheet of paper, something we can draw, compose, but not something else. Spreadsheets succeed even if they are absurd in IT terms for countless reasons while DB+simple languages for equivalent needs does not for the very same reason.
PARC vision, for the little I know propose a substantial revolution, not an evolution of its present state of thing. That's their real "fail"... To succeed they need to find a way to demonstrate and spread their tech until it became well known enough. Like in the past in France where no peasant have accepted the new vegetable imported from America: potatoes. Rumors spread it's dangerous etc. Things change only when the army drop them around the country in manure piles and soldiers came back after enough time to collect and eat potatoes. Sunflowers in Russia have experienced a similar fate.
I disagree with the lie-to-your-face characterization. Steve Jobs was always telling the truth as he saw it at the time. He was just uncommonly good at making you believe it.
Henry Ford also wasn’t lying at any point, and was not a Steve Jobs figure either. He was widely loved (prior to his anti-semitic statements, to be fair) by the common folk for telling it like he saw it.
IBM, I’m not sure what this refers to. The Watsons were salesmen, but were not widely considered dishonest, certainly not Jr.
Tesla, yes, Musk is full of shit at all points. People don’t seem to love him for it.
"Steve Jobs was always telling the truth as he saw it at the time."
I doubt it.
I'll offer two stories about it.
When Steve recruited me to work for NeXT, he spent an hour or two with me in a conference room, working on persuading me. The experience was fascinating.
One thing Michael Fassbender's portrayal in the film "Steve Jobs" missed was the hard, piercing quality of his gaze when he was pitching. People like to talk about "laser focus"; Steve's gaze is the best example of that I've ever seen in a real, person-to-person encounter. His eyes were hard, vigilant, analytical. He noticed everything.
He tried a series of pitches to get me on board. Now, I did end up going to work for NeXT for a while, but I wasn't buying any of Steve's pitches that day.
He detected my skepticism really quickly each time. He would start a pitch with all the passion and conviction you would expect if you've ever seen one of his keynote speeches or interviews, like his whole being was invested in the pitch. He would immediately detect that it wasn't working, and he'd just drop it, like it had never mattered. He'd move on to another pitch, something totally different, and he'd be just as committed and passionate as in the previous one--until he detected that it wasn't selling me.
It was like he was flipping channels, looking for anything that would work.
The second story is from a couple years later, after Steve returned to Apple and took over again. I was back at Apple, too. Steve held a town hall meeting for the whole company to introduce the new candy-colored iMac (which was basically just the old iMac in new candy-colored clothing). I attended with a good friend who was working on the system software.
Steve came out in full pitch mode, selling Apple's new product to its employees. My impression of his pitch was that it sure was a good thing that he went into selling products and not into politics. Before that, I had always been sort of baffled by the ability of history's great demogogues to move huge crowds to behavior that in retrospect was clearly ill-advised. Steve's internal pitch for the iMac was a sort of "Oooohhhhhh...now I get it" moment for me. He was just extremely good at reading a crowd's reactions, saying what it wanted to hear, moving it to the reactions he wanted it to have.
Walking out after the pitch, my friend was shaking his head, saying, "I believed him. I knew it was all lies. I mean, some of that stuff is my work, and yet I believed him."
My sense is that Steve's objective was always selling his vision. He preferred to tell the truth to accomplish it, but if the truth wasn't working, he was perfectly happy to change channels to anything else that would do the job.
" He would immediately detect that it wasn't working, and he'd just drop it, like it had never mattered. " I've seen something similar. I worked with a sales guy that always seemed like he was 100% invested in what he was saying(selling?). As soon as someone called him out on what he was saying he would drop it all. Similar to dropping a box of bricks. He went from 100 to 0 and move on. His acting and ability to read people were amazing.
Believe me, and I knew/know, none of them. If they sold, they lied. How can you sell a bunch of people a lot of imaginary pie-in-the-sky BS without lying? It's rule number one to get financing and to get people to jump in with out looking and risk their time and effort. For many of these companies the "Dream" came true but founders had no way of knowing that when they made their pitch.
I recently read about Jobs. People that worked with him LOVED him but they also knew he was full of it sometimes.
This article seems to have been digitized with OCR. I've found several glitches: Smalltalk-8O, 'dis play', and other mistakes caused by OCR that should have been caught by a spelling check. Issues with quotes are apparent throughout the article: quotes ending with `“` instead of `”` (see the first paragraph) spaces between single and double quotes (`“ ‘` and `’ ”`) make lines starting with quotes look odd and leave a lines with just a double quote on it at the end of a paragraph (the second paragraph ends with a `”` on it's own line)
I also found this quite interesting:
> One innovative use for the network had nothing to do with people sending messages to one another; it involved communication solely between machines. Because the dynamic memory chips were so unreliable in those days, the Alto also ran a memory check when it wasn’t doing anything else. Its response to finding a bad chip was remarkable: “It would send a message telling which Alto was bad, which slot had the bad board, and which row and column had the bad chips,” Thornburg said. “The reason I found out about this was that one day the repairman showed up and said, ‘Any time you’re ready to power down, I need to fix your Alto,’ and I didn’t even know anything was wrong.”
Dealers of Lightening is a book I recommend to anyone who has worked in technology. It's an important story to understand, how you can have all the right people, and all the right technology, and literally be a generation ahead of everyone else, and still fail.
Having technology that can easily be commercialized at a reasonable price is more important that having the best, newest or fastest technology - you can always iterate later.
The Mac was successful not because it was best on a technical level - it wasn't, but it hit the right price point, and the right level of functionality - and at the right time.
Phenomenal book. Snagged it after a similar thread on HN and honestly woke me up to the previous generations of great minds that paved the path before me.
Demo of Mesa/Cedar system, the lesser known system from Xerox ones, written in memory safe systems programming language of the same name, with reference counting and a cycle collector, and those unsafe code blocks.
Replicates the Lisp and Smalltalk development experience, introduces ideas that were inspiration for Oberon Gadgets, OLE and OpenDoc, had networking, file servers, only downside was being single user workstation.
We as industry are still catching up with that experience and overall system safety.
I'd like to think these ideas have been around and well-known for decades, and we're starting to see the hardware rise up to support them. If you told someone in 2001 that Linux would be rewritten in say, Java, with a garbage collector it would have been dismissed as starkly inefficient. Nowadays we have hardware that makes certain safeties negligible on performance (memory tagging, bounds checking, etc). Nowadays we can make an OS out of Java and it will stand just fine. Or Rust. Faster hardware means we run slower (and safer) software. Not everything gets isolated into an ASIC, of course.
It would still be dismissed as starkly inefficient, because the problem isn't technology, rather mentalities.
Midori powered Asian Bing as proof of its capabilities, and yet Windows team was dismissal of its capabilities and instead went on reinventing .NET with COM, aka .NET Native and C++/CX alongside WinRT. A kind of ironic given that it was .NET original design (codename Ext-VOS), before they decided to reboot COM with a managed runtime, alongside the J++ issues that caused COOL to become C# instead.
Quite interesting feedback from Joe Duffy how everything went down,
> OS and tools for building dependable systems. The Singularity research codebase and design evolved to become the Midori advanced-development OS project. While never reaching commercial release, at one time Midori powered all of Microsoft’s natural language search service for the West Coast and Asia.
Ironically, despite still having the Linux kernel underneath and enough C++, Android and ChromeOS are probably the mainstream OSes that are closer to the overall idea, at least in what concerns userspace applications.
As on the Microsoft side, WinDev seems to always sabotage those ideas, as you can infer from Joe Duffy talks, and from Apple side although Swift is supposed to play a major role going forward, C, C++ and Objective-C still represent the majority of the stack.
>> The Singularity research codebase and design evolved to become the Midori advanced-development OS project. While never reaching commercial release, at one time Midori powered all of Microsoft’s natural language search service for the West Coast and Asia.
> Ironically, despite still having the Linux kernel underneath and enough C++, Android and ChromeOS are probably the mainstream OSes that are closer to the overall idea, at least in what concerns userspace applications.
Singularity uses a proprietary language-based microkernel. Midori was allegedly an attempt at a commercial version of Singularity, and I'm not sure what is different about the two, but Midori also uses a microkernel. Linux famously uses a monolithic kernel. Also, consider, any operating system that uses the Linux kernel... is Linux by definition, though technically Linux is the kernel, GNU/Linux is the OS.
There is a piece of container management software, I think that's what it is, called Singularity, and it uses the Linux kernel running on Linux. Maybe you were thinking of the wrong Singularity.
Yes Singularity uses a C++ based microkernel and everything else is written in Sing#, what is your point?
Midori has nothing to do with Singularity in architecture other than being a second attempt from the same group of researchers at a memory safe OS.
Android and Chrome OS aren't Linux, they use a highly customized Linux kernel, in fact after Project Treble it is so customized that it could almost be considered a pseudo-microkernel, as all Treble drivers exist as user processes, use Android IPC to talk to the kernel and since Android 8 all drivers must be Treble based (Android considers the Linux kernel drivers as "legacy drivers").
Finally I don't see how Singularity and the Linux kernel have anything to do with each other, my examples with Android and ChromeOS are what I mentioned in relation with the Linux kernel.
> Ironically, despite still having the Linux kernel underneath
I misread what you were referring to is all. I thought you were saying Singularity ran on the kernel linux, but you were not saying that at all. You were talking about the GNUless Android and ChromeOS. Comprehension is underrated.
In a sense that is what unikernels/managed runtimes running on top of type 1 hypervisors represent, now if we ever get a consumer OS with that approach, it remains to be seen.
I don't expect to still see it during my lifetime.
Maybe when computing gets rebooted with quantum computing, if you check the languages being used for research (Q#, Qiskit, visual), none of them are classical C or C++ like.
That's on the bare metal side, without disagreeing I was mostly thinking about the ~user/programmer side notion of computing. Tiny blocks of near stateless logical operations streamed (so sync or async or threaded .. no difference). But in a way unikernels are also a way to rethink what a system is.
Yes, Cedar was impressive, like the adult version of the Oberon system. I wonder if the source code of the Cedar environment is available somewhere. There is an interesting report by Atkinson (and Jacobi, who is on the panel in the posted video) where they migrated the environment to Unix using C as an intermediate layer. This is apparently the version used for the demonstration. But I couldn't yet find this source code anywhere. Maybe someone has a link.
Some of the Xerox PARC digital archives have printouts from Cedar, including the bytecode and microcoded CPU instructions, now if everything is there, no idea.
That version of the Cedar source was on a CD-ROM accompanying the blue-and-white version of this SOSP'93 paper:
Carl Hauser, Christian Jacobi, Marvin Theimer, Brent Welch, and Mark Weiser. 1993. Using threads in interactive systems: a case study. SIGOPS Oper. Syst. Rev. 27, 5 (Dec. 1993), 94–105. https://doi.org/10.1145/173668.168627
Thanks for the hint. It seems the additional data is not available from the ACM site. Was really the source code of the whole environment including the Mimosa compiler on the CD? Do you know anyone who has access to it? I might be interested (if feasible) to migrate and run it on contemporary operating systems.
My impression was that people were able to make a living while devoting the vast majority of their effort turning their vision into a reality. If one really has a passion for their work that's about as close as one can get to being paid to play all day. Try doing that just about anywhere in tech today.
What a remarkable list of achievements in the PARC alumni. It's hard to think that amount of computer related talent will ever be concentrated to that level again, especially given the rapid ascent of our field driven in no small part by the innovation that went on at PARC.
My guess at what in the early days of personal computing, i.e., ~1985, really caused the rapid growth: Replacing the typewriters.
The labor of using typewriters was enormous. So, replacing a typewriter with just a PC, maybe with only a floppy disk for file storage, and just an early, crude dot matrix printer saved a lot of money.
Next, electronic spreadsheets saved a lot of money.
Next, with just dial-up connections, we got email which in many cases was a big improvement, speed, cost on what the USPS, FedEx, etc. provided.
...
Now for shoppers, on-line shopping can often yield better shopping results with huge savings in time and effort and also often money.
Apparently now some investors are looking at the Internet and remote work as a way for workers and families to get huge savings on costs of housing and transportation.
Then, nearly everything in the usual parts and pieces of the economy is open to huge cost savings via automation based on computing and the Internet, in simple terms, having some robot do the work instead of a human. E.g., it's spring and time to plow the field and put in a crop of corn. So, click on an icon, and the tractor fires itself up, plows the field, and then puts in the seed. And the tractor was made nearly entirely by robots, maybe including some 3D printing.
Then in the supply chain, the corn goes to feed hogs, chickens, and to fatten up cattle, and the work there also gets heavily automated. Sure, cornmeal can also be used for tasty breading for fried fish and chicken!
Uh, maybe surprisingly soon and quite broadly we don't really need more farm workers but fewer and, then, more developers of the automation ...!
Developing all the needed automation is not so easy now, in particular, while can save big on Opex can have a big initial Capex. So, we need better tools for developing the automation -- robots, artificial intelligence, whatever call it.
Hmm .... With the automation, will (finally) have good numerical data on nearly all parts of the supply chain from a mine, farm, forest, etc. all the way to the dinner table, etc. Then we will notice that with the good, new numerical data, can get some cost savings by the now old subject optimization, mathematical programming, from linear programming, nonlinear programming, ... stochastic optimal control, etc.
And with all the good data on the supply chain, should be able to do some analysis and do well investing the commodities markets.
It used to be, "Machines should work. Humans should think." Maybe it will be "Machines do the work, and humans enjoy life", e.g., pursue family, art, science, understand the universe, etc.
Uh, for that future, we have a lot of software to write!
>My guess at what in the early days of personal computing, i.e., ~1985, really caused the rapid growth: Replacing the typewriters.
In business, word processing really came in with the minicomputer-based word processors from the likes of Wang and the other minicomputer makers. Though they were by no means universal.
When the IBM PC came out in 1981, yes people used it for word processing but spreadsheets, especially Lotus 1-2-3, were really the killer app.
In the mid-term, the replacement of typewriters with computers led to the secretary's job description going away (phone answering machines helped too). Somewhere I read (warning: this may be apocryphal) that secretaries were exactly one of the reasons Xerox never cashed in on the personal computer (or the Xerox Star): Xerox managers had secretaries to do their typing, and couldn't imagine doing it themselves; nor could they imagine giving their secretary a piece of hardware that cost as much as the secretary's yearly salary. (The Xerox Star cost somewhere in the range of $15k, I'm just guessing what a secretary's salary was back then. And never mind that factory laborers used machinery that cost at least that much.)
I had an experience with this a couple decades ago. I was using a piece of Xerox software that came on a CD in a book. The license meant you had to buy a copy of the book for every runtime version, and IIRC it came with some kind of non-commercial use requirement even then. An organization we worked for wanted to use it, so we talked to Xerox, eventually getting to one of their lawyers. He told us it would cost ~$150k _to write the license_. Who knows what it would have cost to actually license the product. Needless to say (so I'll say it), we went with someone else's software product.
The final stage of the creative process is dissemination, and this is what failed Xerox Parc. This is also what failed Kodak, when they failed to capitalise on the digital camera, which they had invented.
The core issue is twofold:
1. Entrenched interests within the company. When Microsoft tried to develop their mobile strategy, the head of the Excel division deliberately made the mobile Excel experience a crappy one as he had no faith in Mobile to begin with.
2. Framing the new product within old paradigms. Kodak had it all: a head start on digital image capture, a partnership with Apply (with the QuickTake 100) and shot tons of money. But despite developing a web-based dissemination platform for digital photos, they could not foresee the rise of social sharing. This was a tragedy, as their Box Brownie (released 1900) had the same effect on the photo industry of the time as social sharing would have on the photo industry. But Kodak was fixated on the mistaken notion that consumers would want to print their 'Kodak moment'.
If Parc had failed at dissemination we'd all still be using the command line.
The windows/icons/mouse/menus paradigm disseminated just fine. It wasn't necessarily completely original - see also, Mother of All Demos - but they showed it could be implemented in real systems with real usability benefits.
Laser printers and networking also disseminated just fine.
What didn't disseminate was the development environment. That's not necessarily a surprise, because an environment that delights creative people with PhDs isn't going to translate well to the general public.
Ultimately Smalltalk, MESA, etc were like concept cars. You could admire them and learn from them, but they were never going to be viable as mainstream products.
Windows and Mac filled the gap by producing dumbed-down and overcomplicated versions of the original vision, which - most importantly - happened to be much cheaper.
Xerox get a lot of criticism for missing the potential, but it's easy to forget that in the 70s word processors typically cost five figures, and minicomputers equivalent to the Dorado cost six figures.
Maybe the future would have been different if someone had said "We need to take these ideas and work out how to build them as cheaply as possible."
But realistically commodified computing in the 70s ran on Z80s and still cost five figures for a business system with a hard drive.
The technology to make a cheaper version didn't exist until the 80s.
The problem since Parc is more that there has been no equivalent concentration of smart, creative, playful people. Technology culture changed from playful exploration to hustle and "engagement", there's less greenfield space to explore, and - IMO - there are very few people in the industry who have the raw IQ and the creativity to create a modern equivalent.
It's pretty much the opposite problem. Instead of exploring tech without obvious business goals, business goals are so tightly enforced it's very hard to imagine what would happen if completely open exploration was allowed again.
> If Parc had failed at dissemination we'd all still be using the command line.
Yes, it was disseminated, but not by Xerox. It was the startups of the time that took the baton.
> The problem since Parc is more that there has been no equivalent concentration of smart, creative, playful people.
Could not agree more. Play is a very underestimated component of productive thought. But what is play? I would say that it is production without responsibility. It is notable that children use play in order to explore their potential and their place in the world. Which is to say, they use it to grow. Need I say more?
The thing is that Xerox mgmt themselves refused to consider downsizing the ideas developed internally towards commodity hardware that people could afford. They were interested only in high margin office systems. So other people did it. Not a bad thing for the industry, but bad for Xerox.
Most famously, the team @ Apple around the Lisa and the Mac. But also Lee Lorenzen was an employee at Xerox in Texas. In 1982 he tried to pitch management on a port of their Star concepts to 8080 CP/M class hardware, to get it into people's hands. Demo footage here, remember this is before the Mac came out [0]. He was rebuffed, and left Xerox to join Gary Kildall @ Digital Research, where they created the GEM GUI (which became the OS for the Atari ST and the basis of Lorenzen's next venture, the successful DTP program, Ventura Publisher) [1]. Xerox was never able to penetrate the consumer level DTP market, but this was a big emerging business in the 80s which Xerox would have been a natural fit for.
This is what I'm scared of with Meta. They have a ton of research going in lots of directions, but will they manage to capitalize on it? At least they've got the Quest that iterates and implements these ideas in a consumer device, but after all these years where are the AR glasses, the watch, the phone, etc.? To me Meta is either going to fail in the same way, or going to boom a la Apple in the next decade.
Scared of? I sincerely hope they suffer from this.
a) Because Meta is frankly a malevolent entity
b) Because VR in the form of the "metaverse" they're proposing is equally malevolent. Especially in the hands of a company who has demonstrated a desire to distract and exploit people's attentions.
They pay people millions of dollars a year to do research that suggests that their software causes mental illness in children and they continue to hock their products as though they'll be the replacement to the Internet in the near future. Thanks for being the frog that boils to death.
I'll now repeat what "dead" sibling commenter said, but without the inflammatory "burn" comment that likely got them downvoted.
Except in this case the problem isn't the technology, but the company behind the technology.
There is nothing wrong with this comment, it is correct. I don't care if you've covered your eyes and ears. Meta is a company that has behaved on the whole profoundly unethically. I think there's a pretty strong consensus for that among techies.
It's pretty intellectually dishonest of you to imply some sort of luddite position when someone presents legitimate criticism of the Meta corporation.
And I personally also have a strong problem with the kind of VR that Meta is advocating for. I think it has a potential for serious societal problems. Especially in their hands, it will likely accelerate many of the societal ills that Facebook has been acused of.
> Instead of exploring tech without obvious business goals, business goals are so tightly enforced it's very hard to imagine what would happen if completely open exploration was allowed again.
Similar could be said of science research and innovation in general. See this [1] HN submission on "When do ideas get easier to find? (freaktakes.substack.com)".
Two concurrent current trends are taking place that militate against this kind of open exploration, James Burke-ian Connections-style "branching". The mania for revenue optimization, and the siren song of walled garden lock-in. Messy, "wasteful", "dangerous" open ecosystems deeply offend both of those sensibilities, but I think those sensibilities are flawed in not recognizing the intrinsic value of human cognition. It is the same kind of oversight I see when modern people assume their time distant ancestors were also cognitively distant.
I don't know how productive this comment will be but thank you for, "getting it." The real, "golden egg" at PARC was the research group itself. It might still be possible to build groups like this again; it's also possible that the cultural moment has passed. I think sans Bret Victor and a few outliers there isn't really anyone to replace people like Bill English, Alan Kay, or the great Dan Ingalls. Hopefully I'm wrong about that.
I think part of what makes it, "very hard to imagine" is that in some sense we would need another, "post-war economic miracle" to get another bright, well educated generation ready to be the next Kay/Sutherland/Engelbart/ect and that seems really far off if that's something that's going to come out of the United States again considering the dismal culture of research here. I think the next Xerox PARC metaphorically will come out of a progressive country in Western/Northern Europe, India, or somewhere connected to the ascending Global South. Imagine Xerox Amsterdam Research Center or the like. Smart people and legal pot shops sounds might be a viable way forward. Time will tell.
...and remember the computer revolution hasn't happened yet!
>But Kodak was fixated on the mistaken notion that consumers would want to print their 'Kodak moment'.
Yes and no. They also invented the PhotoCD.
Kodak doubtless did a lot of things wrong. But transitioning from largely a chemical and film consumables business--they actually owned a chemical company at one point--to digital was always going to be tough. Very little of their expertise and distribution channels was transferrable to digital.
Fujifilm did better by applying its emulsion knowhow to some healthcare products. And also came out with a pretty nice line of mirrorless cameras. But it had tough financial times too and was a much smaller company to pivot.
Yeah, Kodak was a lot less a photography company and a lot more a specialized chemistry company. Utterly changing the core focus of a large company is very difficult to pull off because they spent decades optimizing everything to the original core focus.
With the continuous advances in materials sciences, I wish they had pivoted their expertise towards that direction. Though arguably BASF and similar already occupy that sector, there are various photothermal and similar photochemical advances that need their operational and manufacturing prowess to bring out of the labs.
At the time (~1999) digital photography was starting to take shape on the consumer side, the only option for viewing digital photos was on a CRT computer screen. Also at the time, the idea of anyone other than “nerds” having a computer and wanting to view photos on them was completely foreign to executives in the company (this was the overwhelmingly popular view of almost everyone at the time, not just Kodak executives). IMO, this was by far the primary issue that caused Kodak to miss out.
Thanks. One reason why I come here is to hear original testimony such as this. I present on Kodak in my lecture on creativity and will add your quote to me slides.
Was there anyone you know inside the company that had the potential to change things, to have made a different future?
"Later in the 1970s, Engelbart lost his key engineers to Xerox PARC lab, a lavish and well-funded research center a few miles away. At the head was Alan Kay, 15 years Engelbart’s junior—an upbeat, brilliant guy who knew how to inspire people. The laboratory chief was Engelbart’s former funder from ARPA, Robert Taylor. For Engelbart, networks had always been an inextricable part of his vision. But under Kay’s direction, the engineers created a personal computer, geared toward individual productivity rather than collaboration. Their software included more user-friendly versions of a few of Engelbart’s original ideas, including multiple windows, text with integrated graphics, and the mouse. A cruel joke of the time was that Engelbart’s Augmentation Research Center had been a training program for PARC." [1]
SRI did a boatload of work: USPS zip code automation, BART cards, ATM's, distributed enterprise computing in the 1980's, etc. Unlike its peers, it kept defense contracts below roughly 40% of its overall business. Even PARC was mostly awash in DARPA funds.
Right time, right problem, good management, enormous funds, Bay Area hippie culture, access to Stanford, uninterrupted time, and luck.
At PARC they were able to peer 10 years into the future. That was a big deal for 1969-1979. Looking 10 years forward recently is less transformative. Not much changed between 2012 and 2022.
> I don't know that there wasn't some initial implementation "coded in BASIC" — Smalltalk 71
Dan has said/written before that the "first pass" -- which was done in like a weekend or something -- was done in BASIC just to get a feel for the thing. It wasn't a full blown environment or anything.
'Much to my surprise, only a few days later, Dan Ingalls showed me the scheme working on the NOVA. He had coded it up (in BASIC!), added a lot of details, such as a token scanner, a list maker, and the like, and there it was -- running. As he like to say: "You just do it and it's done."'
"The Early History of Smalltalk" ACM SIGPLAN Notices, Volume 28, Issue 3, March 1993
You've got to appreciate the irony of how an article about historical revolutionary technologies is buried under what has to be one of the most obnoxious web interface I've ever seen. And on top of all that, you realize the website is affiliated with the IEEE.
This thing is not even properly printable, even with Print Friendly. I like to imagine Alan Kay stumbling upon this and what his reaction would be...
I worked for Xerox in 1979-1980. I saw first-hand how the general Xerox organization was too copier-focused to appreciate the treasure they had at PARC. It's ironic that every new hire was given a copy of "The Billions Nobody Wanted" (by John H. Dessauer), a bio of Chester Carlson, the inventor of xerography.
At the top, Xerox saw the "paperless office" as a threat to their copier business. They denied my company access to the technical information we needed to make full use of the $500K ($1.7M today) laser printer we bought from them. A pox on Xerox!
To their credit, every branch had a terminal linked to an SDS Sigma running the Control Program V (CP-V) timesharing system. I used it to write a BASIC program to calculate optimal copier proposals for salesmen.
The politics in academia and research are horrid. It's mainly narcissism that drives people to publish rather than hide their stash. Every such group I've seen have the publishers taking everyone's work and putting their name on it. Smart people go find other smart people, and the smartest make a name for themselves by distilling the collective wisdom.
But the incentives are wrong. They'd rather be known for the idea than do the work to deliver it (and love to blame the bean counters for their own lack of diligence).
Apple today is right to sequester information -- not because it divulges secrets but because it's better culturally to say it's not real until it's a product people can use. They've tamped down the glory-mongering of their pre-Jobs 90's with its bloated Copland and Newton egos. To me that discipline drove Apple to deliver the innovations they have in the 2000's - not Jony's esthetics or Cook's efficiency.
An absolute storm of Creativity where each Genius exponentially builds off of another. The Energy, Enthusiasm and Drive of these Geniuses leaps out at you.
Bur as usual A bunch of horse’s asses who don’t know anything about technology were making the decision (direct quote) and squandered it all.
Would it be possible to recreate a similar environment in any company today? I highly doubt it. Nobody looks at Research as it was done in Xerox PARC, Bell Labs and HP Labs.
Some relevant quotes;
Some researchers say PARC was a product of the 1960s and that decade’s philosophy of power to the people, of improving the quality of life. When the center opened in 1970, it was unlike other major industrial research laboratories; its work wasn’t tied, even loosely, to its corporate parent’s current product lines. And unlike university research laboratories, PARC had one unifying vision: it would develop “the architecture of information.”
Since projects were not assigned from above, the researchers formed their own groups; support for a project depended on how many people its instigator could get to work on it.
“Systems research requires building systems,” he said. “Otherwise you don’t know whether the ideas you have are any good, or how difficult they are to implement.
Since MAXC, the center has built prototypes of dozens of hardware and software systems—prototypes that sometimes numbered in the thousands of units.
“There was a lab where the Altos were getting built, with circuit boards lying around, and anyone could go in and work on them,” recalled Daniel H.H. Ingalls,
If you’re dealing with marketing or planning people, make them kick the tires. All the charts and all the slides aren’t worth a damn
One reason that Xerox had such trouble bringing PARC’s advances to market was that, until 1976, there was no development organization to take research prototypes from PARC and turn them into products.
“The amazing thing about the PARC environment in 1976-77 was the feeling of power; all of a sudden you could create things and make lots of them. Not just one sheet, but whole books,” said Conway.
But some of those who left PARC recalled that a disillusionment had set in. They hadn’t been frustrated with the progression of their careers; rather, they had been frustrated with the rate of progression of their products into the real world.
122 comments
[ 1.2 ms ] story [ 182 ms ] threadBell Labs was able to commercialize (or at least license IP) from some of their work with reasonable return, but eventually lost momentum from the series of spinoff-layoff-mergers. How does an organization successfully manage or balance that kind of research productivity with commercialization in the long term? Is there any management literature on that in particular?
I know some places like Microsoft and Google have had research programs for quite some time, although of course Google mostly fails to adequately support products that come out of Google X before killing them off or scaling them back (such as Glass), and I'm not sure how much of what Microsoft Research does is actually commercialized. Notably, Butler Lampson, who was one of the key folks at PARC, has been with Microsoft Research since 1995.
For more reading:
Fumbling the Future: not all that good; reads like a B-school case study
Dealers of Lightning: excellent history
Inventing the Future (by me!): what it was actually like to be there in the trenches. Fictionalized, but all facts are accurate. Foreword by David C. Smith, the guy who invented icons. That's his son on the cover, playing MazeWar on an Alto.
I did enjoy your book by the way - thank you for writing it.
re Soul: Dick Sonderegger, whom I interviewed at [1], knew the real people at Data General who were featured in it. They kinda despised Tracy Kidder.
[1] https://operationcode.org/podcast#dick-sonderegger-link
Dealers makes everyone look good more or less - like it clearly explained to me why Xerox could never figure out how to commercialize the PARC innovations. The single largest issue appeared to be the incentives the sales force had were not aligned to the kind of products PARC could develop. (though its clear there were other issues with the capacity of PARC to productize something)
A faster Alto with a general purpose operating system and the existing document preparation tools, probably would have significantly outsold the Star - but again, would the Xerox sales force known how to sell that product? I don't think so, and thats the underlying problem.
The Star appeared to have just suffered from the Second System Effect, which compounded the other issues Xerox had.
I used to say "Star has done for object-oriented programming what Three Mile Island did for nuclear power."
Obviously not, though, since OOP grew & prospered on its own. I cover our struggles using OOP with a non-object oriented language in exhaustive detail, too.
A note for folks, the eBook of "Inventing the Future" is $US 3.95 or it can be read on Kindle Unlimited
It's great when people price eBooks cheaply like this.
"The Big Bucks" (about the 80's & the birth of the Internet) is going to be out in a few days, and it'll be more. Why? Because I think for my audience, the price doesn't matter as much, and so my income from writing might turn positive someday (and thus it won't be a "hobby" to the IRS).
Nice Tron reference
1 - https://en.wikipedia.org/wiki/Douglas_Engelbart
2 - https://www.darpa.mil/about-us/timeline/nls
3 - https://www.darpa.mil/program/memex
I do not say that's was happen since I do not really know nor NLS (of course I've seen the Mother of all the demos, but not much more especially in technical terms) nor Smalltalk workstations (I've dig a bit more there, emulation included, but lacking comparison with NLS...) however is a generally valid principles: in nature it does not win the better but the more adaptive. In humans terms some very good stuff might simply fade into oblivion because they are too ahead of their time for most people or are against someone's interests.
To spot a more modern example maybe a bit more easy to know by most HN-ers just see Java history, it's original idea and the subsequent practical fail due to the fact nobody really care the original idea adopting instead just the boilerplate. Plan 9 was another example.
IMVHO the real points was:
- pioneers have seen desktop computers, not services, with the human at the center, not a vendor or a service, more specifically connected people with powerful tool some in politics is named social tissue needed for democracy, something essentially all élites from all countries fear most;
- pioneers have cared about their idea, vision, not much their present time short terms needs, as a result most people have no idea about the future they envisage and have tried to meld what they are told about the future to their present created ridiculous stuff and encouraging some to go that way (see General Magik virtual office for instance);
- pioneers have not cared at all about scaling their gear, witch means how to make them en masse, creating a supply chain, ... as a result profiting from them was essentially an artisanal game no one can really sustain and at government level no one is much interested either.
Another today examples: we have seen in the recent 50+ years past various efforts to make small, cheap VTOL aircrafts, i.e. "flying cars". Now we see even financial institutions embracing them in their sauce, well so far such experiments have proven feasible and to a certain extent VERY interested for all. So far most think they are impossible. Useless. Unsafe. .... As a result the general reactionary behaviors of essentially any large community from élites (who feel well in general and fear loosing their status if something change) to generic human beings (who generally think the change is bad, not needed, they simply fear it if is a real change) annihilate or at least postpone in a very far future any real revolution.
Culture is typically the mean to make things happen but so far we seems to be unable (beside the fact that most do NOT want that happen) to really spread a substantial culture though the masses. We know how to do that in small cohort but not in larger ones. We still miss a way to connect those small cohort to the humanity enough to make the magic happen...
One simple example is with copy and paste. NLS had a highly complex and error prone copy-and-paste system. A lot of the commands in NLS were modal (vs modeless), and thus error prone. See this article for more info: http://worrydream.com/refs/Tesler%20-%20A%20Personal%20Histo...
Another issue was basic usability of NLS. Engelbart's vision was on continuous bootstrapping, continually aiming to create more powerful and expressive systems. This vision was pretty good for technical people, but also comes at the cost of basic ease of use and entry for newcomers. For instance, think about how many problems non-technical people still have with today's PCs and smartphones, many of which strive for "walk up and use" kind of usage.
Having been in research for over two decades now, my position is that we researchers will often identify the right problem and get the right general direction, but not always the right form. For instance, in the late 90s at Berkeley, folks correctly identified cluster computing as the future, and that we should aim to create abstractions to host Internet services to make them more scalable and robust. The specifics went in a very different direction (e.g. cloud computing, Kubernetes, etc), but the general thrust was right. As another example, my colleagues and I correctly identified rapid prototyping and testing tools for user interfaces as a major need, and while we inspired a lot of ideas with our work (pen-based sketching), industry went in a very different direction (see Figma and Axure and InVision).
I think the same is true with Engelbart's work. He was definitely a pioneer and helped open up people's imaginations as to what was possible with computing, but was off on a lot of the details and specifics that really matter for adoption. And that's ok because to a large extent, that's one of the many reasons we as a community and society do research.
Researchers/Engineers want bottom up progress. They place incremental development at the heart of their system.
Partially since researchers have very little money.
They cannot do anything that requires 100 developers. They want a low hurdle rate, a low barrier to entry. To get started and make visible progress.
Engineers are constantly demanding progress. They want solutions that are state of the art, for various different reasons. The lots of small, incremental improvements to keep the cutting edge.
Consumers want the complete opposite. Consumers want whatever ecosystem has the most money and progress invested into it.
Then the consumer want it to never change. Ever.
They don't want to invest capital expenditure/attention span in learning new things. They have other important problems to deal with.
Then, when there is a deal-breaker scene change, they will jump to the next big thing.
How can this be sped up? It feels like progress was made as fast as possible but in hindsight it's almost silly that there was not a computer in every household in 1970.
Xerox PARC, just didn't have a great BSer in-house.
Anyone might be pushed to buy a phone/something alike, considering it just a small addition to their daily routine, anyone might be pushed to buy a desktop/something else in general like the classic joke from the early web "hi! I want to buy a website" "hum, ok... To do what?" "Oh, I do not want to use it, all start buying one so I need one to remain in the market, is for customers not for me". But a substantial revolution is another story. Just see in daily life how many time you hear people say "I've always do this like that" and so on even if it's clearly an absurd thing or something that can&should be done but in a very different manner. Even small things.
At PARC time was the filing cabinet/suspended folder era, where people deeply think in paper+mechanic terms, the idea of files and folders (mimicking aforementioned filing cabinet and suspended folders) might pass easily, the idea of dynamic documents is simply unthinkable and fearful "I might loose something", "people might cheat", ... similarly the idea to send emails might pass easily, a better fax system, but the idea of computing inside again can't pass. For most a GUI can only pass, even today, as a single sheet of paper, something we can draw, compose, but not something else. Spreadsheets succeed even if they are absurd in IT terms for countless reasons while DB+simple languages for equivalent needs does not for the very same reason.
PARC vision, for the little I know propose a substantial revolution, not an evolution of its present state of thing. That's their real "fail"... To succeed they need to find a way to demonstrate and spread their tech until it became well known enough. Like in the past in France where no peasant have accepted the new vegetable imported from America: potatoes. Rumors spread it's dangerous etc. Things change only when the army drop them around the country in manure piles and soldiers came back after enough time to collect and eat potatoes. Sunflowers in Russia have experienced a similar fate.
Henry Ford also wasn’t lying at any point, and was not a Steve Jobs figure either. He was widely loved (prior to his anti-semitic statements, to be fair) by the common folk for telling it like he saw it.
IBM, I’m not sure what this refers to. The Watsons were salesmen, but were not widely considered dishonest, certainly not Jr.
Tesla, yes, Musk is full of shit at all points. People don’t seem to love him for it.
I doubt it.
I'll offer two stories about it.
When Steve recruited me to work for NeXT, he spent an hour or two with me in a conference room, working on persuading me. The experience was fascinating.
One thing Michael Fassbender's portrayal in the film "Steve Jobs" missed was the hard, piercing quality of his gaze when he was pitching. People like to talk about "laser focus"; Steve's gaze is the best example of that I've ever seen in a real, person-to-person encounter. His eyes were hard, vigilant, analytical. He noticed everything.
He tried a series of pitches to get me on board. Now, I did end up going to work for NeXT for a while, but I wasn't buying any of Steve's pitches that day.
He detected my skepticism really quickly each time. He would start a pitch with all the passion and conviction you would expect if you've ever seen one of his keynote speeches or interviews, like his whole being was invested in the pitch. He would immediately detect that it wasn't working, and he'd just drop it, like it had never mattered. He'd move on to another pitch, something totally different, and he'd be just as committed and passionate as in the previous one--until he detected that it wasn't selling me.
It was like he was flipping channels, looking for anything that would work.
The second story is from a couple years later, after Steve returned to Apple and took over again. I was back at Apple, too. Steve held a town hall meeting for the whole company to introduce the new candy-colored iMac (which was basically just the old iMac in new candy-colored clothing). I attended with a good friend who was working on the system software.
Steve came out in full pitch mode, selling Apple's new product to its employees. My impression of his pitch was that it sure was a good thing that he went into selling products and not into politics. Before that, I had always been sort of baffled by the ability of history's great demogogues to move huge crowds to behavior that in retrospect was clearly ill-advised. Steve's internal pitch for the iMac was a sort of "Oooohhhhhh...now I get it" moment for me. He was just extremely good at reading a crowd's reactions, saying what it wanted to hear, moving it to the reactions he wanted it to have.
Walking out after the pitch, my friend was shaking his head, saying, "I believed him. I knew it was all lies. I mean, some of that stuff is my work, and yet I believed him."
My sense is that Steve's objective was always selling his vision. He preferred to tell the truth to accomplish it, but if the truth wasn't working, he was perfectly happy to change channels to anything else that would do the job.
I recently read about Jobs. People that worked with him LOVED him but they also knew he was full of it sometimes.
It's a talent. I wish I had it.
I also found this quite interesting:
> One innovative use for the network had nothing to do with people sending messages to one another; it involved communication solely between machines. Because the dynamic memory chips were so unreliable in those days, the Alto also ran a memory check when it wasn’t doing anything else. Its response to finding a bad chip was remarkable: “It would send a message telling which Alto was bad, which slot had the bad board, and which row and column had the bad chips,” Thornburg said. “The reason I found out about this was that one day the repairman showed up and said, ‘Any time you’re ready to power down, I need to fix your Alto,’ and I didn’t even know anything was wrong.”
Having technology that can easily be commercialized at a reasonable price is more important that having the best, newest or fastest technology - you can always iterate later.
The Mac was successful not because it was best on a technical level - it wasn't, but it hit the right price point, and the right level of functionality - and at the right time.
https://youtu.be/z_dt7NG38V4
Replicates the Lisp and Smalltalk development experience, introduces ideas that were inspiration for Oberon Gadgets, OLE and OpenDoc, had networking, file servers, only downside was being single user workstation.
We as industry are still catching up with that experience and overall system safety.
Midori powered Asian Bing as proof of its capabilities, and yet Windows team was dismissal of its capabilities and instead went on reinventing .NET with COM, aka .NET Native and C++/CX alongside WinRT. A kind of ironic given that it was .NET original design (codename Ext-VOS), before they decided to reboot COM with a managed runtime, alongside the J++ issues that caused COOL to become C# instead.
Quite interesting feedback from Joe Duffy how everything went down,
"Safe Systems Programming in C# and .NET" - https://www.infoq.com/presentations/csharp-systems-programmi...
"RustConf 2017 - Closing Keynote: Safe Systems Software and the Future of Computing" - https://www.youtube.com/watch?v=EVm938gMWl0
> OS and tools for building dependable systems. The Singularity research codebase and design evolved to become the Midori advanced-development OS project. While never reaching commercial release, at one time Midori powered all of Microsoft’s natural language search service for the West Coast and Asia.
https://www.microsoft.com/en-us/research/project/singularity...
Ironically, despite still having the Linux kernel underneath and enough C++, Android and ChromeOS are probably the mainstream OSes that are closer to the overall idea, at least in what concerns userspace applications.
As on the Microsoft side, WinDev seems to always sabotage those ideas, as you can infer from Joe Duffy talks, and from Apple side although Swift is supposed to play a major role going forward, C, C++ and Objective-C still represent the majority of the stack.
> https://www.microsoft.com/en-us/research/project/singularity...
> Ironically, despite still having the Linux kernel underneath and enough C++, Android and ChromeOS are probably the mainstream OSes that are closer to the overall idea, at least in what concerns userspace applications.
Singularity uses a proprietary language-based microkernel. Midori was allegedly an attempt at a commercial version of Singularity, and I'm not sure what is different about the two, but Midori also uses a microkernel. Linux famously uses a monolithic kernel. Also, consider, any operating system that uses the Linux kernel... is Linux by definition, though technically Linux is the kernel, GNU/Linux is the OS.
There is a piece of container management software, I think that's what it is, called Singularity, and it uses the Linux kernel running on Linux. Maybe you were thinking of the wrong Singularity.
Midori has nothing to do with Singularity in architecture other than being a second attempt from the same group of researchers at a memory safe OS.
Android and Chrome OS aren't Linux, they use a highly customized Linux kernel, in fact after Project Treble it is so customized that it could almost be considered a pseudo-microkernel, as all Treble drivers exist as user processes, use Android IPC to talk to the kernel and since Android 8 all drivers must be Treble based (Android considers the Linux kernel drivers as "legacy drivers").
Finally I don't see how Singularity and the Linux kernel have anything to do with each other, my examples with Android and ChromeOS are what I mentioned in relation with the Linux kernel.
I misread what you were referring to is all. I thought you were saying Singularity ran on the kernel linux, but you were not saying that at all. You were talking about the GNUless Android and ChromeOS. Comprehension is underrated.
I don't expect to still see it during my lifetime.
Maybe when computing gets rebooted with quantum computing, if you check the languages being used for research (Q#, Qiskit, visual), none of them are classical C or C++ like.
Carl Hauser, Christian Jacobi, Marvin Theimer, Brent Welch, and Mark Weiser. 1993. Using threads in interactive systems: a case study. SIGOPS Oper. Syst. Rev. 27, 5 (Dec. 1993), 94–105. https://doi.org/10.1145/173668.168627
- Sent from my (not quite) Dynabook
The labor of using typewriters was enormous. So, replacing a typewriter with just a PC, maybe with only a floppy disk for file storage, and just an early, crude dot matrix printer saved a lot of money.
Next, electronic spreadsheets saved a lot of money.
Next, with just dial-up connections, we got email which in many cases was a big improvement, speed, cost on what the USPS, FedEx, etc. provided.
...
Now for shoppers, on-line shopping can often yield better shopping results with huge savings in time and effort and also often money.
Apparently now some investors are looking at the Internet and remote work as a way for workers and families to get huge savings on costs of housing and transportation.
Then, nearly everything in the usual parts and pieces of the economy is open to huge cost savings via automation based on computing and the Internet, in simple terms, having some robot do the work instead of a human. E.g., it's spring and time to plow the field and put in a crop of corn. So, click on an icon, and the tractor fires itself up, plows the field, and then puts in the seed. And the tractor was made nearly entirely by robots, maybe including some 3D printing.
Then in the supply chain, the corn goes to feed hogs, chickens, and to fatten up cattle, and the work there also gets heavily automated. Sure, cornmeal can also be used for tasty breading for fried fish and chicken!
Uh, maybe surprisingly soon and quite broadly we don't really need more farm workers but fewer and, then, more developers of the automation ...!
Developing all the needed automation is not so easy now, in particular, while can save big on Opex can have a big initial Capex. So, we need better tools for developing the automation -- robots, artificial intelligence, whatever call it.
Hmm .... With the automation, will (finally) have good numerical data on nearly all parts of the supply chain from a mine, farm, forest, etc. all the way to the dinner table, etc. Then we will notice that with the good, new numerical data, can get some cost savings by the now old subject optimization, mathematical programming, from linear programming, nonlinear programming, ... stochastic optimal control, etc.
And with all the good data on the supply chain, should be able to do some analysis and do well investing the commodities markets.
It used to be, "Machines should work. Humans should think." Maybe it will be "Machines do the work, and humans enjoy life", e.g., pursue family, art, science, understand the universe, etc.
Uh, for that future, we have a lot of software to write!
In business, word processing really came in with the minicomputer-based word processors from the likes of Wang and the other minicomputer makers. Though they were by no means universal.
When the IBM PC came out in 1981, yes people used it for word processing but spreadsheets, especially Lotus 1-2-3, were really the killer app.
it was a loss for xerox but a win for mankind. Many outcomes of Xerox became patent unencumbered
The core issue is twofold:
1. Entrenched interests within the company. When Microsoft tried to develop their mobile strategy, the head of the Excel division deliberately made the mobile Excel experience a crappy one as he had no faith in Mobile to begin with.
2. Framing the new product within old paradigms. Kodak had it all: a head start on digital image capture, a partnership with Apply (with the QuickTake 100) and shot tons of money. But despite developing a web-based dissemination platform for digital photos, they could not foresee the rise of social sharing. This was a tragedy, as their Box Brownie (released 1900) had the same effect on the photo industry of the time as social sharing would have on the photo industry. But Kodak was fixated on the mistaken notion that consumers would want to print their 'Kodak moment'.
The windows/icons/mouse/menus paradigm disseminated just fine. It wasn't necessarily completely original - see also, Mother of All Demos - but they showed it could be implemented in real systems with real usability benefits.
Laser printers and networking also disseminated just fine.
What didn't disseminate was the development environment. That's not necessarily a surprise, because an environment that delights creative people with PhDs isn't going to translate well to the general public.
Ultimately Smalltalk, MESA, etc were like concept cars. You could admire them and learn from them, but they were never going to be viable as mainstream products.
Windows and Mac filled the gap by producing dumbed-down and overcomplicated versions of the original vision, which - most importantly - happened to be much cheaper.
Xerox get a lot of criticism for missing the potential, but it's easy to forget that in the 70s word processors typically cost five figures, and minicomputers equivalent to the Dorado cost six figures.
Maybe the future would have been different if someone had said "We need to take these ideas and work out how to build them as cheaply as possible."
But realistically commodified computing in the 70s ran on Z80s and still cost five figures for a business system with a hard drive.
The technology to make a cheaper version didn't exist until the 80s.
The problem since Parc is more that there has been no equivalent concentration of smart, creative, playful people. Technology culture changed from playful exploration to hustle and "engagement", there's less greenfield space to explore, and - IMO - there are very few people in the industry who have the raw IQ and the creativity to create a modern equivalent.
It's pretty much the opposite problem. Instead of exploring tech without obvious business goals, business goals are so tightly enforced it's very hard to imagine what would happen if completely open exploration was allowed again.
Yes, it was disseminated, but not by Xerox. It was the startups of the time that took the baton.
> The problem since Parc is more that there has been no equivalent concentration of smart, creative, playful people.
Could not agree more. Play is a very underestimated component of productive thought. But what is play? I would say that it is production without responsibility. It is notable that children use play in order to explore their potential and their place in the world. Which is to say, they use it to grow. Need I say more?
https://ethw.org/Oral-History:Adele_Goldberg#Starting_a_Spin...
Most famously, the team @ Apple around the Lisa and the Mac. But also Lee Lorenzen was an employee at Xerox in Texas. In 1982 he tried to pitch management on a port of their Star concepts to 8080 CP/M class hardware, to get it into people's hands. Demo footage here, remember this is before the Mac came out [0]. He was rebuffed, and left Xerox to join Gary Kildall @ Digital Research, where they created the GEM GUI (which became the OS for the Atari ST and the basis of Lorenzen's next venture, the successful DTP program, Ventura Publisher) [1]. Xerox was never able to penetrate the consumer level DTP market, but this was a big emerging business in the 80s which Xerox would have been a natural fit for.
[0] https://www.youtube.com/watch?v=EMBGRZftS30
[1] https://www.youtube.com/watch?v=6EeOanSInjo
a) Because Meta is frankly a malevolent entity
b) Because VR in the form of the "metaverse" they're proposing is equally malevolent. Especially in the hands of a company who has demonstrated a desire to distract and exploit people's attentions.
https://www.nytimes.com/2021/11/18/technology/meta-instagram...
Except in this case the problem isn't the technology, but the company behind the technology.
There is nothing wrong with this comment, it is correct. I don't care if you've covered your eyes and ears. Meta is a company that has behaved on the whole profoundly unethically. I think there's a pretty strong consensus for that among techies.
It's pretty intellectually dishonest of you to imply some sort of luddite position when someone presents legitimate criticism of the Meta corporation.
And I personally also have a strong problem with the kind of VR that Meta is advocating for. I think it has a potential for serious societal problems. Especially in their hands, it will likely accelerate many of the societal ills that Facebook has been acused of.
"Ubiquitous Applications: Embedded Systems to Mainframe"
https://www.davethomas.net/papers/ubiquitous1995.pdf
Similar could be said of science research and innovation in general. See this [1] HN submission on "When do ideas get easier to find? (freaktakes.substack.com)".
Two concurrent current trends are taking place that militate against this kind of open exploration, James Burke-ian Connections-style "branching". The mania for revenue optimization, and the siren song of walled garden lock-in. Messy, "wasteful", "dangerous" open ecosystems deeply offend both of those sensibilities, but I think those sensibilities are flawed in not recognizing the intrinsic value of human cognition. It is the same kind of oversight I see when modern people assume their time distant ancestors were also cognitively distant.
[1] https://news.ycombinator.com/item?id=31635962
I think part of what makes it, "very hard to imagine" is that in some sense we would need another, "post-war economic miracle" to get another bright, well educated generation ready to be the next Kay/Sutherland/Engelbart/ect and that seems really far off if that's something that's going to come out of the United States again considering the dismal culture of research here. I think the next Xerox PARC metaphorically will come out of a progressive country in Western/Northern Europe, India, or somewhere connected to the ascending Global South. Imagine Xerox Amsterdam Research Center or the like. Smart people and legal pot shops sounds might be a viable way forward. Time will tell.
...and remember the computer revolution hasn't happened yet!
Yes and no. They also invented the PhotoCD.
Kodak doubtless did a lot of things wrong. But transitioning from largely a chemical and film consumables business--they actually owned a chemical company at one point--to digital was always going to be tough. Very little of their expertise and distribution channels was transferrable to digital.
Fujifilm did better by applying its emulsion knowhow to some healthcare products. And also came out with a pretty nice line of mirrorless cameras. But it had tough financial times too and was a much smaller company to pivot.
With the continuous advances in materials sciences, I wish they had pivoted their expertise towards that direction. Though arguably BASF and similar already occupy that sector, there are various photothermal and similar photochemical advances that need their operational and manufacturing prowess to bring out of the labs.
Source: I worked in the Kodak HQ at this time.
Was there anyone you know inside the company that had the potential to change things, to have made a different future?
"Later in the 1970s, Engelbart lost his key engineers to Xerox PARC lab, a lavish and well-funded research center a few miles away. At the head was Alan Kay, 15 years Engelbart’s junior—an upbeat, brilliant guy who knew how to inspire people. The laboratory chief was Engelbart’s former funder from ARPA, Robert Taylor. For Engelbart, networks had always been an inextricable part of his vision. But under Kay’s direction, the engineers created a personal computer, geared toward individual productivity rather than collaboration. Their software included more user-friendly versions of a few of Engelbart’s original ideas, including multiple windows, text with integrated graphics, and the mouse. A cruel joke of the time was that Engelbart’s Augmentation Research Center had been a training program for PARC." [1]
[1] https://www.smithsonianmag.com/innovation/douglas-engelbart-....
http://liveblog.co/users/davewiner/2015/05/06/iWouldHaveHire...
but what actually makes those places so special? why them, not somebody else?
What made the difference between them and the rest?
At PARC they were able to peer 10 years into the future. That was a big deal for 1969-1979. Looking 10 years forward recently is less transformative. Not much changed between 2012 and 2022.
Alan Kay expressed some dismay after his departure that Smalltalk became so much more complex when a compiler was created.
I don't know that there wasn't some initial implementation "coded in BASIC" — Smalltalk 71?
otoh — "Smalltalk 72, which was done in BCPL on a Data General Nova."
https://ethw.org/Oral-History:Adele_Goldberg#Developing_Smal...
Dan has said/written before that the "first pass" -- which was done in like a weekend or something -- was done in BASIC just to get a feel for the thing. It wasn't a full blown environment or anything.
'Much to my surprise, only a few days later, Dan Ingalls showed me the scheme working on the NOVA. He had coded it up (in BASIC!), added a lot of details, such as a token scanner, a list maker, and the like, and there it was -- running. As he like to say: "You just do it and it's done."'
"The Early History of Smalltalk" ACM SIGPLAN Notices, Volume 28, Issue 3, March 1993
https://dl.acm.org/doi/pdf/10.1145/234286.1057828
Privacy Browser doesn't enable JavaScript by default, and reduces frustration with many sites.
I wonder what it is like to live in a society where creativity is actually valued and rewarded.
At the top, Xerox saw the "paperless office" as a threat to their copier business. They denied my company access to the technical information we needed to make full use of the $500K ($1.7M today) laser printer we bought from them. A pox on Xerox!
To their credit, every branch had a terminal linked to an SDS Sigma running the Control Program V (CP-V) timesharing system. I used it to write a BASIC program to calculate optimal copier proposals for salesmen.
The politics in academia and research are horrid. It's mainly narcissism that drives people to publish rather than hide their stash. Every such group I've seen have the publishers taking everyone's work and putting their name on it. Smart people go find other smart people, and the smartest make a name for themselves by distilling the collective wisdom.
But the incentives are wrong. They'd rather be known for the idea than do the work to deliver it (and love to blame the bean counters for their own lack of diligence).
Apple today is right to sequester information -- not because it divulges secrets but because it's better culturally to say it's not real until it's a product people can use. They've tamped down the glory-mongering of their pre-Jobs 90's with its bloated Copland and Newton egos. To me that discipline drove Apple to deliver the innovations they have in the 2000's - not Jony's esthetics or Cook's efficiency.
An absolute storm of Creativity where each Genius exponentially builds off of another. The Energy, Enthusiasm and Drive of these Geniuses leaps out at you.
Bur as usual A bunch of horse’s asses who don’t know anything about technology were making the decision (direct quote) and squandered it all.
Would it be possible to recreate a similar environment in any company today? I highly doubt it. Nobody looks at Research as it was done in Xerox PARC, Bell Labs and HP Labs.
Some relevant quotes;
Some researchers say PARC was a product of the 1960s and that decade’s philosophy of power to the people, of improving the quality of life. When the center opened in 1970, it was unlike other major industrial research laboratories; its work wasn’t tied, even loosely, to its corporate parent’s current product lines. And unlike university research laboratories, PARC had one unifying vision: it would develop “the architecture of information.”
Since projects were not assigned from above, the researchers formed their own groups; support for a project depended on how many people its instigator could get to work on it.
“Systems research requires building systems,” he said. “Otherwise you don’t know whether the ideas you have are any good, or how difficult they are to implement.
Since MAXC, the center has built prototypes of dozens of hardware and software systems—prototypes that sometimes numbered in the thousands of units.
“There was a lab where the Altos were getting built, with circuit boards lying around, and anyone could go in and work on them,” recalled Daniel H.H. Ingalls,
If you’re dealing with marketing or planning people, make them kick the tires. All the charts and all the slides aren’t worth a damn
One reason that Xerox had such trouble bringing PARC’s advances to market was that, until 1976, there was no development organization to take research prototypes from PARC and turn them into products.
“The amazing thing about the PARC environment in 1976-77 was the feeling of power; all of a sudden you could create things and make lots of them. Not just one sheet, but whole books,” said Conway.
But some of those who left PARC recalled that a disillusionment had set in. They hadn’t been frustrated with the progression of their careers; rather, they had been frustrated with the rate of progression of their products into the real world.