Slashdot Mirror

Some years back, one of the former department heads at Xerox's Palo Alto Research Center (whose reputation for innovation is nearly unmatched in history) wrote a book on this subject. I recently read it and enjoyed it greatly. It's called Breakthrough: stories and strategies of radical innovation. I highly recommend it.

Apple Research

Hi all, This is a great comment thread -- really! A few clarifications:

        PARC was not sold off but is an independent subsidiary (incorporated in 2002). This means we work with multiple other clients -- from Fortune 500 companies (including Xerox) to startups to government agencies -- practicing an open innovation model where we provide custom R&D services, technology, expertise, IP, and more.

        Our clients commercialize our technology for their markets -- we have never directly commercialized our own tech. But because we work closely with them, and have our own business development embedded inside the company, we get to bring that commercial knowledge in not to mention carry it across different domains (but protected for different "fields of use" for different clients/applications).

        Indeed, one huge advantage corporate R&D has over university and government labs is its industry focus. So you know how to turn visions into realities because you know how to work within industry constraints to find solutions. But we also get to use government funding to help jump start the really early stage stuff.

        And yes, we are very different from a patent troll! We're also not just contract research. We invest in our own research and development, and we create IP strategically (measured not just by patent number but in portfolio coverage and reduction to practice), and our clients who co-develop with us leverage that IP to reduce their own risk (say, from doing it alone internally) and accelerate their time to market. These aren't just buzzwords: this is make-or-break for a lot of companies!

        Not sure there has ever been "pure" research for research's sake... Even since its inception, Xerox PARC had the founding charter to invent the "office of the future" (which it did). And today it's multiple futures: networking of the future (see for example www.parc.com/ccn), cleantech of the future, novel electronics of the future, and so on. But the secret sauce (diverse disciplines, industry focus, etc.) is the same.

If you're interested in learning more on how we resolve the seeming paradox of "business"... "breakthrough" (our tagline today is The Business of Breakthroughs), do check out some of our posts at http://blogs.parc.com/blog/topics/business-of-innovation/ -- we'd love your comments!

~PARC Online
@parcinc

PARC is now an independent entity, which is not the same as being sold off. It now researches on behalf of other entities besides Xerox, but Xerox remains its largest customer.

Closer, but still not correct. PARC is an independent subsidiary, but fully owned by Xerox.

For a tech savvy bunch, we can't seem to use the internet well.
http://www.parc.com/content/newsroom/factsheet_parc.pdf

There are two ways anyone can visit PARC:
1. PARC Forum every Thursday http://www.parc.com/events/forum.html
      Not a guided tour, but you get to ask questions. And the talks are available for viewing afterward.
      I've asked questions of Guido van Rossum (a famous Dutchman no doubt you know) and Jill Tarter (SETI), and dozens of others.
2. Art exhibits
      There are art exhibits occasionally and they have guided tours of the art on specified days.
      You don't get to ask any questions; it's just an art exhibit space.

Intel has a small museum you can visit, and the Computer History Museum in Mountain View is a must-see.
The Tech computer museum in San Jose is iffy even if you have kids (exhibits aren't well maintained) though the imax theatre there is nice.

Now, what can I see in Amsterdam ;-)

This has already been done in a research context to study behaviors in WoW. See, for example, http://www2.parc.com/csl/members/nicolas/documents/CHI2006-Alone.pdf.

Here ya go.

Cue cries of "This article is very dense. Does anyone have a link to a summary?"

(said with "King & I" diction) Et Cetera, et cetera, et cetera.

Now compare that with some other companies:

http://research.google.com/pubs/papers.html

http://research.microsoft.com/apps/dp/pu/publications.aspx#p=1&ps=36&so=0&sb=d&fr=&to=&fd=&td=&rt=&f=ms&a=&pn=&pa=&pd=

http://domino.research.ibm.com/library/cyberdig.nsf/recent (last 30 days!)

http://www.parc.com/publications/

http://research.nokia.com/

Two conference publications by Apple employee is a joke for a company the size of Apple. Apple doesn't even have a site where they show their research.

(Apple used to have a research lab with real researchers and publications in the 1990's, but they closed it.)

And the poster session is not the output of R&D by Apple, it's people talking about using Apple products in their work.

> It's just yet another technology invented in a lab for academics' sake.

Yeah, and what's that ever given us?

http://www.physics.ucla.edu/~ianb/history/
http://www.research.ibm.com/about/past_history.shtml
http://www.research.ibm.com/resources/awards.shtml
http://www.parc.com/about/milestones.html ...

It is a pity that people talk about virtual reality and related fields without even understanding the basics - but that is the consequence of media hype surrounding this field, together with people calling non-immersive, often even non-interactive applications "virtual reality". Computer games, SecondLife, QuicktimeVR are not VR, period - you cannot really achieve meaningful feeling of presence there. Of course, it sounds and sells better if you stick a gee-whizz sticker on the box ...

I love when people decide their definition of a phrase is the right one and that the rest of the world is wrong.

Word!

http://www-vrl.umich.edu/intro/index.html#NonImmersive
http://portal.acm.org/citation.cfm?id=1029964
http://www.fiu.edu/~mizrachs/VR.html
http://www.agocg.ac.uk/reports/virtual/37/chapter2.htm
http://www2.parc.com/istl/groups/uir/publications/items/UIR-1993-07-Robertson-Computer-NonImmersive.pdf

Mod parent up! Wikipedia is a battleground between small groups of highly focused editors (and I'm being kink by not using some medical term here), and the occasional blokes with too much time on their hands. The fate and content of any article is usually decided by no more than a handful of people. Don't take my word for it, check out http://wikidashboard.parc.com/.

The contested articles have all the scars of a battleground, which seriously impacts readability. Try reading articles affected by Israel-Palestine, Korea-Japan or Creationist-Scientist wiki-wars.

The uncontested articles often read like pure puff, fancruft or obituaries. Try reading some pages about anime characters, or the biographies of obscure individuals that don't get deleted because of their close affiliation with Wikipedia administrators. Did I mention the vanity pages kept by most admins? They used to get indexed by Google until this month.

The well written and informative articles on notable topics are quite rare.

We're working on it. :)

the closing of PARC

Eh?

You're referring to Polybot, which has Generation I listed at 1997. Polybot is a bit more flexible (har har) due to the fact that each module is entirely self-sufficient (aside from power, which they can channel from module to module as needed), although I haven't seen a demo of it handling water....

"Already done" notwithstanding, it's nice to see a robot succeed so well in such varied scenarios.

You're referring to Polybot, which has Generation I listed at 1997. Polybot is a bit more flexible (har har) due to the fact that each module is entirely self-sufficient (aside from power, which they can channel from module to module as needed), although I haven't seen a demo of it handling water....

"Already done" notwithstanding, it's nice to see a robot succeed so well in such varied scenarios.

WoW will go down in history as a classic game.

I would be curious to see a comparison of total man-hours spent enjoying WoW or EVE vs total man-hours spent watching a production of a Shakespeare play. Wow has about 6.5 million players, if we assume a safe average of 100 hours played per player WoW has been played for 605 million man-hours. Meanwhile, In 1600 the population of London was 200,00 by 1700 the population of London was about 600,000 So assuming every single person in London saw two productions of Shakespeare every year, that's only about 200 million man-hours of Shakespeare enjoyed in 100 years. I would say that by some measures WoW is already a greater cultural influence that Shakespeare.

I really roughed in these numbers (but do have sources), if someone who is better at figuring these things would be so kind as to try to supply some better total estimates I appreciate it.

For me the exception in this list is the Xerox PARC work on Content-Centric Networking. I have been following this for some time and feel that it is good research. Xerox has always been at the edge and many people never realise the fruits of their research as the projects are often spun off as separate entities. This seems to incorporate many of the ideas behind Cougaar, Jini and Jxta but using the discovery process in a different way.

Good overview http://www.parc.com/research/projects/networking/c ontentcentric/default.html

Also Van Jacobson, the man behind the project http://en.wikipedia.org/wiki/Van_Jacobson was responsible, in part, for traceroute and other goodies, so there is probably quite a bit of traction there.

Eeek.... the link somehow got omitted on my post...

Modular Robotics

Since having posted i've learnt this thing isn't even a smart eliza program. It's just an actor on a microphone talking through speakers on a toy.

This isn't even really worth discussing

That was actually the original idea behind worms, which, like so many other things, came from Xerox PARC

Put the data in several secure areas and on the USB encrypted.

Print out your keys using the Xerox http://www.parc.com/solutions/dataglyphs/

Its proprietary but damn simple to print.

Ever looked at an internet map? While it is indeed an international infrastructure, most of those nations use the US as a jumping off point. Maybe we should just cut off the rest of the world, most of the US wouldn't notice a difference (except maybe a lack of some BBC news, and a lot less spam)... let you guys figure out your own version, then you don't have to deal with our rules :)

Regarding soul shards: this is a strange request. All activities you're usually involved in WoW fall into 2 categories:
a) those that don't require shards ( crafting, running around, etc );
b) those that generate shards ( killing mobs, PvP ).
I can't think of any reason why you'd need shards when you're out of combat. ( OK, with a possible exception of trying to solo instances, because you don't get shards from gray elites ).

Regarding VW, I also disagree. My VW has tons of armor, 20% more health than I ( despite the fact that I have ~+60 worth of +sta modifiers on me at lvl 40 ), and it can usually keep a mob busy until I put 4 dots and a shadow bolt on it. Add more DPS and taunting ability to it, and it will be like always fighting in a group with a warrior, except you get double the XP. Warlock is already the most soloable class in WoW ( http://blogs.parc.com/playon/archives/2005/06/grou ping_ratio.html ); and warlock/succubus pair typically deals more damage in PvE than any other character of equal level, including rogues and mages. Adding anything else will make the game unbalanced.

having a laptop in class never really helped me. sure, i could type up notes, but i don't take many to begin with anyhow. i found that in the majority of my classes, i ended up using pen and paper anyway, as a math/cs major. diagrams, flow charts and little visiual queues greatly out numbered raw code that was generated in my classes.

now, the one technological advance that would benefit EVERY STUDENT, 100%, no matter their computer literacy, typing speed, course load or distraction threshold, is a simple, affordable ebook reader. make the viewable screen 8x10 or 11, use e-paper so it can run for months on a set of watch batteries and pad the living shit out of it so it'll be more durable. distribute recquired school texts as PDF on CF cards and you've just solved one of the biggest problems in american schools: students who have to lug 50lbs of text with them throughout the day because they don't have time between periods to stop off at dormrooms/lockers for the next round.

this is where the inovation should be. add a simple input interface and you could have information cross referenced between documents and suppliments. figure out how to make ultra-low energy draw wifi cards and you could link them to a national database for easy inquiries on specific topics. hell, you could have two models, the Standard that displayed and cross-referenced the info on the CF card, and the deluxe that allowed the user to "take notes" and link it to a specific page of a text (to accommodate those of us who like to write in the margins). if you had one of these designed like a portfolio with one screen on either side of the fold you could make one side the "book" side and the other the "note" side..

we're really missing the boat on this one, folks. students in general don't need full fledged laptops. all they need is an easy and convenient way to read and reference text.

in this scenario, they students who would truly benefit from laptops could still have one. but i'm willing to bet that the percentage of stuents nowadays who could really use a laptop to broaden their education to be around 5-7%, no greater than 15%. the rest would either use it in the fashion i describe for the e-book or use it to dick around in class while the teacher isn't looking..

In a paper I wrote last year, I had a look at e-mail as a tool for knowledge workers and argued that it is being used so intensively, because it is both personal (it belongs to the user), private (no one else can access it if it is not shared) and personalisable (it can be configured to ones personal needs and work style). Most importantly however I find is that it combines storing information with sharing it: There is no other system in which people can organise their stuff in a personal space and share it with others so easily and flexibly.

I wonder what the consequence is: Shall we improve email clients further and further to make them even more flexible tools that then "can do everything"? (This paper from Xerox PARC would call that "overloading".) Or will we develop other tools? Or is the question irrelevant, as the new ways that we invent to share digital information will be integrated into what we call "email" today and the boundaries will be blurred?

theoretically the robot could add more sections or shed damaged ones without compromising functionality
This is how PARC's modular robots worked, which is the really cool part. The modules could be snapped apart and then reassembled into a different-shaped robot. Some models could even reconfigure themselves without intervention.

The connection between each section would include an umbilical capable of transmitting data, and perhaps even extra voltage ...
Some of the PARC modules did exactly that. And they had limited demos with specialized attachments instead of generic modules, also as you mentioned.

My housemate that summer was working on the control system of these robots. The computer fed them simple script-like commands, so there was no real AI backing it up. Given that, it was impressive what these things managed to do. But of course, they were working toward taking input from each module and applying complex processing to react to it.

As far as your suggestion about distributed processing across modules, other groups at PARC were doing work in that field too. Awesome place to intern!

theoretically the robot could add more sections or shed damaged ones without compromising functionality
This is how PARC's modular robots worked, which is the really cool part. The modules could be snapped apart and then reassembled into a different-shaped robot. Some models could even reconfigure themselves without intervention.

The connection between each section would include an umbilical capable of transmitting data, and perhaps even extra voltage ...
Some of the PARC modules did exactly that. And they had limited demos with specialized attachments instead of generic modules, also as you mentioned.

My housemate that summer was working on the control system of these robots. The computer fed them simple script-like commands, so there was no real AI backing it up. Given that, it was impressive what these things managed to do. But of course, they were working toward taking input from each module and applying complex processing to react to it.

As far as your suggestion about distributed processing across modules, other groups at PARC were doing work in that field too. Awesome place to intern!

Check out Xerox PARC's PolyBot; each segment contains its own motor and PowerPC processor. This was on Slashdot some time ago.

Ms Calendar: Honestly, what is it about them that bothers you so much?

Giles: The smell.

Ms Calendar: Computers don't smell, Rupert.

Giles: I know. Smell is the most powerful trigger to the memory there is. A certain flower or a whiff of smoke can bring up experiences long forgotten. Books smell. Musty and, and, and, and rich. The knowledge gained from a computer, is, it ... it has no texture, no context. It's there and then it's gone. If it's to last, then the getting of knowledge should be tangible, it should be, um... smelly.

Ms Calendar: Well! You really are an old-fashioned boy, aren't you?

This explain anything? That said, there really is something about having an acutal piece of paper in your hands. Maybe if electronic paper ever gets developed enought that might help.

Error correction codes are an important facet of DataGlyphs when one is trying to store a message. However, in this case there isn't any message stored - every glyphmark is devoted to purely positional information. This is called an address carpet, which is described in more detail in this
IEEE article.

Neat, but not amazing. You have to read the article to realize that the system only works if all the puzzle pieces have been printed with special marks, DataGlyphs. It's like printing registration marks on all the pieces. Sort of. The dataglyphs actually have more interesting properties, but the point is that this isn't the vision system you expect. It isn't even a general puspose puzzle solving system. As soon as the system recognizes the glyph marks it knows exactly where the piece belongs. It doesn't "solve" anything. It doesn't have to figure out where the pieces go. You couldn't show it pieces from a puzzle off the shelf and have it solve it.

A Glyphsaw Puzzle starts out as a computer graphics file generated by the PARC DataGlyph Toolkit. The image is sent to a professional jigsaw puzzle manufacturing company, which creates cardboard puzzle pieces. From a distance, the pieces look similar to those from any other jigsaw puzzle. Up close, one can see individual glyphmarks.

...but let's remember that he opposed the use of GUIs

Whatever gave you that idea? According to his book and Jef's own website he was one of the early spokesmen of WYSIWYG. He even invented drag'n'drop - building on other's design ideas.

What he IS opposing is the way current GUIs use the mouse and modes - even alot of his own designs from the time when he was working for Apple.

An accelometer wouldn't sensibly be used to replace the input style / use context of keypads. (Except perhaps in case of accessibility issues and people with disabilities.)

Instead, novel input techniques have been researched for quite a while. Check out these few example publications:

http://sandbox.parc.com/want/papers/mui-cacm-2000. pdf

http://www.csl.sony.co.jp/person/rekimoto/gwrist/

http://www.csl.sony.co.jp/person/rekimoto/tilt/

http://tangible.media.mit.edu/papers/Graspable_Dis plays_CHI97/Graspable_Displays_CHI97.html

http://research.compaq.com/wrl/projects/RocknScrol l/RocknS.html

Yes, they do (or did) and called it ubiquitous computing.

I suppose this is another example of Microsoft getting credit for company's innovations [apple.com]?

I'm getting tired of just about every discussion about Microsoft being used as an opportunity by Apple fans to promote their favorite company. Keep that sort of stuff to the Apple groups, please. Whether or not Microsoft copied a feature from Lotus Improv has nothing to do with Apple.

Furthermore, it is stupid for Apple fans to point fingers when it comes to copying: without copying other companies' innovations, Apple wouldn't exist; they copied the very core of their platform from others (SRI, PARC, Alan Kay). Apple does have better taste than Microsoft in what they copy, but I hardly think they are more original.

Now, with that said... I should point out that the real error in this system is that spammers will just build a database of known hashes.

If the postage is 20 bits, then that's only a search space of 1 million. Just precompute them all (would take less than 12 CPU-days) and you can answer any question in O(1).

Raise it to 25-bit postage, and the spammers spend 32x longer computing 32x more keys, and it'll take them 34 CPU-years to populate the database. With a cluster of 1-million zombie pcs, they'll have that cracked in less than 20 minutes! Whoops.

Raise it to 30-bit postage, and suddenly it takes 17 CPU-minutes just to send one mail. Ouch. Meanwhile, the spammers need 34800-cpu-years. But fortunately for them, their army of zombie computers will have that cracked in ~12 days. Now the spammers are the only ones who can send mail without a 17 minute delay.

Just for grins, let's consider 35-bit postage. That might actually deter spammers, since it would take them 34 years to build the database with their army of zombie PCs. But *nobody* is going to be willing to waste 9 cpu-hours to send a single mail.

Except that that wasn't the first time. This kind of visualization comes from Xerox PARC and is called "perspective wall". The idea is that the tilted views off to the side give you some context without taking up too much space.

Looks like Apple is still getting their ideas from Xerox PARC, even 20 years after the "original" Macintosh.

The X developers should rewrite the server from scratch using the Aspect Oriented methodology and for example the AspectJ programming language. Many of the X extensions really touch all parts of the server which is exactly the kind of problem aspect oriented programming was designed to solve.

Using AspectJ, an extension such as the Damage extension could be written in a weekend.

Also rewriting the server in AspectJ would allow the developers to leverage the full power of the Java language. With Java reflection the core dispatch code in the server could be replaced by just a few lines of code. The RENDER extension could be completely removed from the server and replaced by using the delegate design pattern to forward X requests to Java2D

The Fresco project had huge potential, but never managed to escape the legacy language C++. It seems everybody working on window system is stuck in the software engineering practices of the seventies.

Simulating real-world devices to make computers more usable is a common idea, but not a very good one. Physical devices have lots of limitations and painful user interfaces (sometimes literally). Have a look at IBM's attempt at this. Some of the best attempts at using 3D as part of regular user interfaces probably come from these people; you can judge for yourself whether their user interfaces are useful.

These kinds of attempts at general-purpose 3D user interfaces have the smell of failure--companies desparately trying to look "hip" and "modern", but without anything real to show for it. To me, it's an indication how far behind Sun really is. Good user interfaces should be unsurprising, simple, fast, and use the medium they are presented on well. In the case of computers, that's a 2D, low-resolution, high color depth screen. Design for the medium.

See also the PARC history page.

Aspect Oriented Programming. In depth information is here.

Today it's implemented in a lot of different langugages, maybe sometime in the future a whole development system is created using this technique.

Don't mistake Pointcuts as the only feature of AOP...

Another long list of links and a comprehensive introduction of the pioneers.

The technology referred to is called dataglyphs

This really makes me wish i knew more about coding theory.

"PARC develops a unique approach to the visualization of information that uses people's perceptual and cognitive capacities to help them deal with large amounts of information. The approach is used in 3-D Rooms and is an integral technique used in the Xerox product Visual Recall. It results in the invention of the hyperbolic browser and other focus-plus-context visualization techniques that give the user three-dimensional views of text databases. These visualization techniques offer a revolutionary way for people to access information on the Internet and will later result in the formation of a PARC spin out, Inxight Software, Inc."

And, another description:

"Xerox spins out Microlytics to commercialize PARC's early compression technology research by bringing artificial intelligence spell-checking software, linguistic and data compression technologies to market. Based on an understanding of the deep structure and mathematical properties of language, linguistic compression technology is used for visual recall, intelligent retrieval and data compression. This work has a major impact on the automatic processing of language structures and is one of the key research areas underpinning Xerox's multilingual suite of products."

You can find the history here.

Here's a brochure

= 9J =

The GUI itself could have been patented if software patents had been around at that time. In 1981, just one of the innovative ideas coming out of the PARC project, the Xerox Star 8010 was unveiled at a Chicago trade show as the first computer with a GUI. The Apple Lisa came soon after, having much better commercial success. If these guys can't patent such things, why should anyone else be allowed to?

you can always Print out the info. If you're really worried about lightning you can use a laser printer and marble

Ah... there's my account ;-)

anyway, this is already done with PolyBot:
http://www2.parc.com/spl/projects/modrobots/polybo t/polybot.html

each module is autonomous, and it can make legged as well as snake-like configurations. I assume that this worm i similar.

yes, you can:
http://www2.parc.com/spl/projects/modrobots/polybo t/polybot.html.

PolyBot can do snake forms, as well as form legged versions of itself. Each module is autonomous.

This is close to true, but there are many years of research on lightweight audio communication suggesting that open connections can actually lead to behaviors very like push-to-talk (short bursts of talk) - at least, if you are connecting people who want to be connected. After a while, the open connection is no longer considered as much of a "social contract" to pay attention (compared to the phone), and the formalities still go away.

For some recent social research on young adult use of Nextels, see http://www.parc.com/audiospaces/ptt.htm.

It'll be interesting to see how well some of the non-iDEN providers do with this. Having to wait several seconds for the receiving phone to wake up and poll the network for incoming connections seems sort of crippling for PTT.

...you can still get some info and a download here (well, for the moment anyway).