Miramar 3D: The Ad Hoc "Systemery" of Innovation
posted by Eleanor Wynn on September 24, 2007
Systemery is etymologically related to “truthiness”. I am not going to say systemic because it has been coopted.
This is really about Miramar. Miramar was recently featured at Intel Developer Forum as part of Justin Rattner’s keynote speech, featuring key developer John David Miller (aka “JDM”). I posted on Miramar in February as a chronicler was kind enough to note. Cynthia Pickering brokered this whole thing over the last 5 years and prepped the demo. See below, and also the comments, for more credit to Cindy, of which not enough can be heaped on. Her marathon training really shows.
Well, Miramar made it to IDF keynote status. But there is a long tale to tell here and it is the true story of how innovations happen—sort of like how some soldiers survive combat I guess. I don’t know how many of you read Gunther Grass’s amazing New Yorker story about his short and extremely lucky time in the SS, mostly slogging through woods, hooking up with resourceful companions, changing trucks before they got blown up, and getting the heck out of dodge before the bullets arrived. I personally forgive him as I would anyone caught up in a period of history where you just do what you have to do, especially if you are seventeen. But, I digress.
fThe Intel Miramar 3D user interface has lots of compelling usage scenario material. We discovered a strong hook-up with the Croquet game architecture, which was our first point of contact with a scalable architecture. Qwaq is the company formed to market the concepts of Croquet. Miramar plus Qwaq was a keynote topic at Intel Developer Forum. You might think this invention came out of nowhere and the brilliant foresight of the people who presented it, but think again my friend.
We are going to start at the beginning. Why? Because there are people who would have you believe that there is a formula for innovation, that it is rationalizable and that you can just crank it out without ruffling a single corporate feather. That stuff ain’t innovation! That is “invention”—frippery, doodads, cuteness. Some of which could become an innovation.
Gene Meieran, revered Intel Fellow and person of more wisdom than many of his peers, has said that innovators are a challenge to manage and I have heard someone else say they are a “pain in the *ss”. Those of us on the distaff (feminine gender) side don’t get around to pain-in-the-posterior status until later in life, so it took me a long time to dare to stand up for what I believed in, imagining that, as in graduate school, surely my superiors knew better. Until I got to be their age or more.
Here is the tale of Miramar, fabulous 3D user interface that after several years of development was “zbb’d” (zero-based budgeting acronym) by Intel Architecture Labs before it became Intel Research. (That manager had the decency to cry at the meeting where he told many people their jobs were lost.) Miramar just idled away getting more out of date by the year.
Enter the Virtual Collaboration Team from IT, founded by Nathan Zeldes and Chuck House. I was on that team, as was Cindy Pickering. That team talked and talked and talked about core issues in collaboration software that would serve the needs of a globally distributed corporation. There was plenty of “control vs. freedom” ideology discussion, and a feature set that we worked out in terms of what was needed for a globally distributed corporate with 2/3 virtual collaboration as the engine of productivity.
But no matter how well we described our vision, everyone who heard it compared it to something already existing. Which it was nothing like! So, I decided we would need to prototype it, starting with a Flash demo that would at least show the look and feel of something that addressed the needs. I did the first annual Virtuality Index study to figure out what were the key variables for distributed workforce and how prevalent they were at Intel: time, space, organization, culture, software tools, and multiple responsibilities or multiple teaming (2/3 of workforce). Multiple teaming looked like a characteristic that needed software innovation as it was not even in the organizational model of any software producer in the collaboration software industry.
In Miramar, and in our first Flash demo, we focused on the idea of “team spaces” and we also hinted at the concept of object-linking so that an individual could share instances of an object across multiple teams without having to replicate them in multiple team sites.
Collocation does play a role here. I had a lovely workspace designed by my former colleague John Light (who cleverly figured out the total cube space we collectively had available, including two labs he needed, and shaved feet off of each cube to create a little sitting room next to a full-length window—major perk for Intel cube land). John shows his spatial relations genius in many ways.
And John was one of the developers on the original Miramar. After Keith Feher, the guy we funded to just design the Flash demo for the virtual collaboration workspace decided to use some 3D in his design, we all got onto the Miramar bandwagon, and John Light was right there to be the way the truth, and the Light. John Light and John David Miller, going back further, were managed by Marc Millier as part of “sense-making” initiative. Marc deserves to be in some hall of fame for his combination of brilliance, insight and modesty. He now works as design architecture manager.
His view of Miramar came from insights about how his father, an accountant, managed multiple projects—physically surrounding himself with them, keeping them all within reach, but focusing on the present work front and center. This is something the tree-based file structure does poorly—maintain visual memory. Which is why so many people have twenty open windows on their desktops—fear of forgetting. Except that in 2D, these windows still get buried, just like paper.
Another feature of Miramar was object-linking so that anything connected to a project, whether it be a person, a document, an application, could be explicitly connected to it via context retention. The TaskTracer research we conducted with OSU showed how critical this was. The two original developers on the project, John David Miller and John Light, had done super-slick renditions of the navigation characteristics needed to manage a multi-dimensional space. This originally supported standard office tools.
Meanwhile I was lurking about in Croquet, which started looking more and more interesting. It had the Alan Kay imprimatur on it and Alan Kay acolytes that I once knew at Xerox were still working on it. And John Seely Brown, who I also knew at Xerox said it was state of the art. (Do you see how socially embedded we have become in this very short account?) I proposed some of those folks to do a panel with me, which never happened, but I wound up meeting Rick McGeer from HP Labs, reconnected with Ted Kaehler, met David Smith and indirectly met David Reed from MIT Media Labs. Croquet made a great scalable architecture for Miramar.
All along I was working with Cynthia Pickering, the person who could actually make things happen due to being a Principal Engineer, well organized, very insightful and willing to take risks. Nonetheless there was an argument among the VCRT team members as to whether we would not be able to move faster developing on our enterprise collaboration environment. The answer is “No, we would run up against a dead end there”—here is where the pain the *ss part comes in, executed by yours truly—moreover, we already learned that anybody we described this to would just claim they already did that with their “groovy” software. Right!
So then came some organizational hurdles. We were this close [] to a Joint Invention Development with HP where the Croquet team was working when, oops! The whole Alan Kay contingent was eliminated from H-P, on the brink of success. Cindy and Chuck House pursued them to Southern California and kept the dialog open.
Chuck, now MediaX Industry Relations Director at Stanford is a major wheeler-dealer. So that conversation stayed open across a divide of not much happening. Not much except that our IT group hired John David Miller who disappeared down a rabbit hole intensively developing for two years……
….then along came Qwaq, Inc., the formalization of Croquet. Suddenly John David Miller (JDM) was living at Qwaq, commuting weekly to Palo Alto. That deal was consummated by Luciano Oviedo, also in IT and who I happened to meet via his wife being so pregnant she looked like she had a basketball under her dress…at Santa Fe Institute!
So, recently, Justin Rattner, VP of Corporate Technology Group (and he will have to tell how sudden his was) wants to go into “Virtual Worlds”. Hello! We got virtual worlds! We got user interface, architecture, usage scenarios and even more, compelling needs (remember multi-teaming and context retention?) In a matter of maybe six weeks JDM, Cindy and Justin’s very adept and insightful Technical Assistant Ketan Paranjape, had created a keynote presentation just-in-time for presenting our Virtual Worlds initiative. Just-in-time after about 10 years of prep, that is. Handed off from one foster parent to another. With a few dogged believers picking up after the original team had been disbanded.
What is really unusual about Miramar as we know it now is that it aligns across all levels: it has a very robust usage model based on four years of workforce trending data, machine learning data, and ethnographic data; it has a phenomenal user interface that, previous to this usage model was an application in search of a need; and it puts a pretty navigational face onto a state of the art deeper scalable architecture, that without that UI and that usage model was just a cool open source developer playground. They all line up, which is an example of how technology innovation really ought to work. It can take time to get that convergence—just as there was a long interval between the development of TCPIP and the full-blown Internet as we know it.
“The wheel in the sky keeps on turning….”
Comments (19)
tagged: 3D, context retention, Intel Developer Forum, Miramar, multi-teaming, navigation, virtual teams, virtual worlds
Comments
Sep 25 | marc said:
Nice post Eleanor. You perhaps give me too much credit and JDM and John Light too little, but your history matches my memory pretty well.
Folks may be interested to know that the Miramar interface actually was designed to solve a specific user interface problem. We were working on an “information ecosystem” concept where the “right” information would automatically be brought forward based on the user’s active context. The typical WIMP interface enforced arbitrary boundaries between visualizations of the information that destroyed the value of creating the users working context.
JDM and John’s design for a seamless interface in which to capture and maintain all of the information for a particular user context was the true innovation. We did not set out to create a 3D ui (unlike some other efforts at the time), we set out create a powerful way to help information workers understand what they already knew. It came out in 3D because the “deep” space was necessary to solve the “context centric user interface” problem.
Thanks again for documenting this 10 year “overnight” innovation…
Sep 25 | jdm said:
Hi Eleanor: I appreciate the spirit of this - yes, innovation is messy, non-linear, and often achieved only after years of perseverance despite being told, “No.”
That said, I do remember some details of this a bit differently and can fill in some of the others….
Miramar was “born” on 2 Oct 1997, as evidenced by the first pencil sketch of “Mirimar” [sic] in one of my lab notebooks, done in preparation for a visit to a potential customer. (The joke: Miramar was the Navy “Top Gun” aviator training school, and with our Miramar we wanted to prove that 3D was useful for more than just flying around and shooting things.) At that time, I had just returned from a researcher-in-residence stint at MIT Media Lab (in John Maeda’s lab) and had come back with a prototype 3D timeline browser, code-named, Grand Canyon. Grand Canyon is an interesting way to immerse in personal and public photos, headlines, and so on, but it isn’t something you would use all day long for your every-day work. Because of Grand Canyon and the other 3D visualizations we had created, we recognized the need for a 3D workspace in which to contain and relate them all in The Big Picture.
It’s worth pointing out that these visualizations were the front-end of a personal, multi-agent info system code-named Escher that we had constructed earlier, as part of our work in information relevance and what is now sometimes called text mining. In Escher we had a pluggable architecture, where we could write a little code to sense or reason about a specific user behavior, document feature, and so on and have all those bits share that info with other agents. In December 1996 we demo’d Escher publicly at an invitational research forum we held in Oregon, at the Jones Farm campus where we were. Some of that code was later packaged up into a Relevance SDK, but by and large, that work sits idle and forlorn. Innovation wasted.
Going back even further to the late ‘80’s up to May 1991, John Light and I were at a little Beaverton company, Graphic Software Systems (GSS) that was looking for the next big thing, as their graphics business was becoming commoditized. I designed a self-organizing information system I called “MySelf” (with a nod towards the Self programming language) and proposed that as a potential new business for GSS. That turned out to be a bit farther off the current path than GSS wanted to go, so I left GSS to form my own company. The intent was to do just enough contract work to pay the bills while we developed our own products, but after 4.5 years I joined Intel. The stuff Intel was doing in Intel Architecture Labs (IAL) was that compelling, fostered by Steve McGeady and other leaders who understood how to create and nurture a creative, innovative organization.
The point of this digression into ancient history is really to underscore how important a supportive organization is to innovation, and that if you have what you think is a great idea and the passion to make it happen, you have to press on through the unsupportive organizations to get to the good ones. GSS wasn’t the right place to make MySelf, Escher, or Miramar happen; IAL under Steve McGeady was. Having spent time at both MIT Media Lab and IAL in their heyday, I have to give the nod to IAL - for one thing, it could take an idea end-to-end.
From Oct 1997 to early 2001, we developed a half-dozen or so Miramar prototypes, each one probing the user-interaction model, navigation model, plugin application model, and so on. Early versions sat atop Escher, which provided a rich, visual discourse between user and computer. The last versions were built on a simpler document model, largely because we didn’t want users to have to install Escher, and also because we found users wanted to collaborate, and sharing a file via email was a lot easier than importing/exporting pieces of the Escher database. When the project was resurrected in IT Research as “Miramar 2” (2005-present), specifically to support collaboration, Mark Hancock (U Calgary) and I developed a protocol to synchronize instances of Miramar, both for real-time collaboration and offline/online syncing. (Mark also got a paper published RE a clever interaction technique he invented for resolving sync conflicts.)
While we were developing Miramar 2, Eleanor arranged the first meeting with the guys destined to become Qwaq, then at HP Labs. We hit it off, saw we had complementary pieces of the puzzle, but before we could formalize a working relationship, HP “right-sized.” After they reformed as Qwaq, we were finally able to ink the deal that was announced last Thu at IDF. (Btw, credit for the IDF demo goes to Cindy Pickering for the Miramar demo content, and yeah, I provided the technical glue to demonstrate a combined Miramar / Qwaq Forums, but my hat is off to the Qwaq guys for moving heaven and earth to get all the pieces from other content vendors and integrate them all in time for the show. Bravo!)
What next? I’d love to tell you, but we’ve got to save something for next IDF, right? Seriously, tho, with our collaboration with the smart guys at Qwaq firmly in place and Justin Rattner et al now beating the drum for “3D Internet” (or, if you prefer, as I do, the term we sometimes use internally, “Connected Visual Computing”) the best is truly yet to come….
Sep 25 | Nathan Zeldes said:
And then there’s this other innovation model where bright people like the above have great ideas, and they immediately get funded to implement them without hindrance.
Nah… just kidding. Eleanor captures the inevitable complexity and the required interplay of minds and resources very well. All’s well that ends well!
Sep 26 | Michael Molin said:
Hello Eleanor,
With virtual worlds I have a feeling that we are going to picture a sci-fi movie like in 30s in the last century about Moon landing. We are just don’t know what we are talking about - even now since the launch of Sputnik 50 years ago.
Also, this reminds me about the recent popular opinion (not about those UFOs and War of the Worlds - or Coast to Coast radio show now) that this world is somewhat a simulation a la the Matrix movie. This is PR like an iPhone or gPhone.
Michael
Sep 26 | Michael Molin said:
Hello Eleanor,
With virtual worlds I have a feeling that we are going to picture a sci-fi movie like in 30s in the last century about Moon landing. We just don’t know what we are talking about - even now since the launch of Sputnik 50 years ago.
Also, this reminds me about the recent popular opinion (not about those UFOs and War of the Worlds - or Coast to Coast radio show now) that this world is somewhat a simulation a la the Matrix movie. This is PR like an iPhone or gPhone.
Michael
Sep 26 | Eleanor Wynn said:
Thanks to all for filling in more details of this history. So it goes back even farther….JDM, how do you think IT Research picked up Miramar? Chuck and I stood up for that plan. Some other folks wanted to prototype in another platform that was more “mature”. It was Keith Feher who had the idea to use the Miramar look and feel for our VCRT Flash demo long before you came back into the picture, in about 2003 or earlier. That is because it had the capability to support multiple project spaces. We had project spaces, team spaces and personal spaces in that demo, which was shown to IT Staff at IT SLRP. Miramar was in mothballs and would have stayed there. John Light was sitting next to me, a constant reminder; you were over on another campus while much of this prep was going on, including the fight over whether or not to use Miramar. Brad Anders or Don Meyers will remember that discussion and maybe have some better dates. I did not come along just at the Croquet juncture. The whole VCRT effort led by Nathan and Chuck led up to this. That is the point. It takes connectors, it takes a compelling usage model, ie, one that cannot work without that technology, and it takes a tag team. I guess another piece of the story is that all the actors are not aware of all the parts, as I was not aware of your and John Light’s experience with the graphics company. I just wish I could get John’s comments on this.
Sep 26 | Eleanor Wynn said:
Michael, in this case we are talking about an information management collaboration user interface that uses 3D. Virtual Worlds was as you say the PR part of it—sounds good. But people are starting to do things in World of Warcraft & Second Life. However, Miramar is far more beautiful and also has work-related information management characteristics that mean it should be taken seriously. I wish you could see it. If you ever go “west” as far as Nizhny Novgorod, let me know and we will see if we can get a demo to there where we have a site.
Sep 27 | chuck House said:
I am “well along” on a history of Hewlett-Packard (Stanford Univ Press, 2008), with 4.5 years of research and writing to date. It has provided a huge opportunity to appreciate innovation in all of its messy intersections, deadends, resurrections, and cross-currents. A small example — yesterday I was invited (only ex-HP person) to the Smithsonian 40th anniversary of the “first handheld calculator invention” and listened to an esteemed curator explain that TI “invented” the scientific and graphing calculators as well as the four-function machine. Laughable, except you want to cry. But then I got the actual inventors of the TI stuff to tell the curator “their version” which matched mine! Satisfying on one small level. History gets written by those who write it, not those who did it necessarily.
Sorry, I too digressed. On topic: it is a major thrill for me to see Justin touting Miramar/Qwaq at IDF. I was lucky enough to be invited to this IDF by Gene Meieran, including a lunch with some of the Intel Fellows, and the YouTube clip of Greg Nuyens and Miramar w Justin is great good news.
I have heard both the sagas as Eleanor and JDM shared them, several times, so have no reason to doubt that all of that happened. Re the “rekindling” effort that Nathan and I started — it was an underground activity that JJ encouraged when I joined from Dialogic, and I met Nathan soon after (first email between us was Sept 3, 2001, eight days before 9-11). WIth Brian Gorman, we hatched the idea of the VCRT group, in a PP slideset and email trail of Oct 29, 2001.
At Brian Gorman’s autumn 2001 IT Productivity conference, Nathan and I presented the idea to Mary Murphy-Hoye of doing a “skunkworks” project, and we got a “birds of a feather” time at the meeting at noon Nov 14. Cindy Pickering, Tammie Hertel, Brad Anders, Don Meyer, Eleanor and Mark Chuang were the volunteers who stayed more than two meetings. By midsummer 2002, we had a Concept Car idea, sponsored by Cindy, and this turned into a Flash demo around screens designed primarily by Mark C. that were ready for “prime time” in April 2003 with many more people involved, including Phil Tierney and FLEX, etc. Doug Busch, JJ, and even Andy Bryant liked the stuff, and we did show it to Otellini, Tennenhouse, and Rattner at the time
Much earlier, I had met JDM thru Alex Marquez in Corporate Capital, for some K-12 tools. Jeannette Harrison told me to go see Richard Beckwith (not sure of the name) who intro’d me to Mark Millier, who sent me to JDM and the Grand Canyon stuff, and as a sidebar, Miramar. I had managed the Motif team at HP years earlier, which did the 3D widgets for OSF, winning three straight IDSA medals from 1990-1992 for screen designs, so it was easy to perceive the Miramar contribution. That was the magic moment for me, to try to get it de-mothballed. This meeting was Jan 23, 2003. Mark Chuang and I had email dated Jan 28, 2003 where I “wax” about it, and he cranked the ideas into the VCRT sketches without the “3D” element or Z-axis.
When the Flash Demo (and subsequent Video (in two forms — 2 min and 10 min) “played well”, I got permission to start a project, and Cindy (as always) knew where to go to get $$$. And somewhere in there, we got connected to JDM more tightly. It wasn’t until the Collaboratory got set up late in 2004, though, that we could actually hire JDM, and by then, Eleanor and others were changing groups. Brad and Cindy were the primary folk at this time left from VCRT, and Brad actually was in an adjacent group as well
Re the HP/Intel Joint work, Eleanor did begin that connection, and I was able to help promote it, The first email trail I have is Feb 22, 2005 from Rick McGeer at HP who wrote to Cindy, Eleanor, and JDM, cc’ing me, and suggesting a joint project. We eventually got far enough to have agreement for a JIP project between the two companies, and also Craig Samuel (HP) and I submitted a joint proposal to CRA’s Grand Challenge on Aug 12, 2005 that was supported by HP more than Intel (Tennenhouse didn’t like it even tho Justin did)). This was a merge notion of Croquet and Miramar, as Eleanor mentioned. And then Alan Kay’s group got shot at HP.
I kinda lost interest in Croquet at that point, but when Justin and Abel Weinrib hired Alan Kay as a consultant in November 2005, Cindy and I “jumped” on the chance to rekindle things again. While interviewing Alan Kay for my HP book January 20, 2006, I got to tour the Applied Minds facility, and renew acquaintance with both Bran Ferran and Danny Hillis. I was impressed enough to write on Jan 31 to Andrew Chien and Justin that we should open discussions w Applied Minds, and that led to bringing Gene Meieran, Cindy, and others through the facility in Feb 2006 (later for Justin and Andrew); at the 2nd meeting, Applied Minds and Alan’s Viewpoints group were “spinning out” the group that became Qwaq. Alan and I are Advisors for them still.
As I was retiring from Intel, one question was how could we nurture this Qwaq/Intel relationship, and Cindy and Martin Curley and I, along with JJ, put considerable effort into trying to place JDM into that environment, which has worked out well to date
I think this whole thread points out that many people help on good ideas, and that yes indeed it is messy, with lots of false starts and considerable backtracking. The real joy is to see something come to fruition, and to celebrate the many contributions that helped.
And by the way, none of this could be traced as easily if we all followed the Intel policy of erasing all history…
Sep 27 | Eleanor said:
Thanks Chuck, the plot spreads. One point of correction. From the VCRT, I was the person who had the idea to do the Flash demo because I could see that people were translating our words, and even Mark Chuang’s foils (thanks for bringing his name back into this) into what they “already knew”. Mark and had several intense working sessions in JF3 Cafe to come up with those foils. I had the verbiage and concepts, he had the PP skills. And, Mark was funded to do this by whatever group Gregory Bryant was running at the time—they had a collaboration charter for a while. Mark’s direct manager was Frank Soqui. They ran a tight ship and I recall going in with Mark to support the project he was doing that led to this. Mark was key in developing the first visual representation. I proposed the Concept Car that resulted in the Flash demo, and Cindy funded it. I managed the key messages and worked with the designer. We did get Phil Tierney on board to PM it, thanks for that entry as well. The key messages were key as we started with a mile-long list of messages but ended up with 3: “all my teams in one place”, expressivity, asynchronous and persistent. There may have been one more, I have to go back and look. As Brian Gorman said at the time “this demo put collaboration tools ahead 3 years in ten minutes.” Cindy and I worked to pare that 10 minutes down to 3 for a preso that Doug Busch was supposed to give, but never did. But we ran that 3 minute demo over and over again at many venues. Cindy is the elastic that held this all together. She not only had the vision to pick up on the various suggestions and keep them going, she funded them, she managed all the organizational interfaces, she co-authored various papers, she did tons of hard work on the JIP, the prototype itself with the applications we devised, the Research at Intel Day showcases, the Country Fair (Intel Research as opp to IT R&D)showcases. She hired JDM after you (Chuck) and I lobbied for him against another candidate. One more thing—in the Virtuality Studies we tracked the success of the mainstream collaboration tool that was rolled out and decided not to go up against it. The Virtuality Survey, thanks to some deeper diving by Steve Graves, told us that the people with the most pain were Design Engineers, so we went towards modeling that environment. Cindy did all of that—populated Miramar with DE scenarios and also the Digital Health scenario. These things made Miramar believable and necessary for certain segments and not just a futuristic risk for the whole workforce. The point I would like to get across is that technical excellence may “seem” to be the cause of innovation but without a compelling scenario, which is NOT JUST USABILITY, and a lot of organizational muscle, patience and supporting research, it may not see the light of day.
Sep 27 | Keith Feher said:
It was love at first sight; I remember seeing Miramar at the IAL country fair in 97 or 98. From that first glimpse at the Country Fair I thought this is the way computing should be done. To step back for a moment, I have an extensive background in 3D graphics and animation and worked as a graphic artist/UI designer for the User Centered Design Group in IAL. 3D graphics or even AutoCAD require extremely complex user interfaces to house all the complex tasks and options that a user requires and more often than not a user would have so many windows or tool pallets open on a project that the actual view of the model he was working on was so obscured that they were constantly shuffling the UI elements around to do the simplest of tasks, this was the same issue when it came to designing the collaboration tool UI. In late 2002 I was fortunate enough to work with IT research to help develop a “concept car” demo of their collaboration tool research. After many many long discussions with our Human Factors Engineers and Information Architects in Flex Services I decided the only way I could graphically represent everything that was to be in this demo was in a 3D interactive environment, there was just too much info for typical Windows interface with windows and palettes… at least for a demo. So that’s when Miramar came to my mind, excitedly I explained my idea to the team and everybody hated it. It made perfect sense to me, but most people couldn’t get past the concept that 3D was for video games and you don’t do video games at work. The 3D environment was a perfect way to visualize and organize massive amounts of data in a logical and fast way, but a 3D interface was considered excessive and opulent by most. It wasn’t till Eleanor introduced me to John Light that we started getting people excited about the possibilities of 3D and Miramar. It finally seemed like it was going to happen… till I scoped the project and got estimates on what it would take to update Miramar to a dot Net framework and with a price tag of around $250K that’s when the idea was mothballed at least for my involvement since I was in the IT Flex pay per use model. But as things would have it Eleanor and team forged ahead and almost 4 years later I hear Justin talking about Miramar at IDF… Woohoo!
Sep 27 | Eleanor Wynn said:
Keith, yet another personal saga connected to this. You were a link in the chain, because until you brought up 3D I didn’t think of it, but when you did I had a reference point for it. The Flash demo stayed the way you left it for a couple of years until Cindy got the Collaboratory going and we decided we could apply a headcount to building a prototype. Then came the discussion over what platform to build it in/on? The upshot of that discussion was that Cindy took on JDM to do that massive effort of updating Miramar. As I said, “just in time” to be ready for Justin’s demo. So, it was non-trivial to get Miramar up to date, per the figure you stated.
Sep 28 | Brian Gorman said:
For me, the significance of Miramar circa 2003 was that it showed the way to visual and workflow integration of collaboration partners, work objects, and tools. The best part was that the application was “out of my way”, and the promise was that I could touch these things directly, instead of feeling like my interactions were being mediated, muted, and even dicated in their form, by a thick layer of application goo and narrow channels of allowed behavior. The Miramar graphics and especially the flash demo were stunning in their ability to change people’s thinking, to help people get past the blinders imposed by our mainstream UI’s. 4 years on, it’s a thing of beauty to see Miramar stay true to that vision. I think what I appreciate most of all in the evolution of Miramar is the minimal compromise in its execution, which is a tribute to the artistic sensibilities of its creators.
The fundamental premise of our work in the 2002-2004 period was that collaborative work processes in an enterprise were in many ways the very definition of our culture. Given that premise, it was unimaginable that we would be willing to allow those processes to be designed by a software vendor - we needed to turn the standard story for “office automation” on its head…decide what we want our culture to look like, and then assemble or build applications to instantiate that. That this proved to be hard work is hardly surprising. But Miramar was pivotal in helping free us to envision that future…and it may yet help us to actually create it.
This blog has triggered a nice set of memories Eleanor…thanks for posting it…I sure do miss working with Chuck!
Sep 28 | Eleanor Wynn said:
Brian, thanks for reminding me of that strategic view that we had. We felt that no vendor could understand our globally distributed workforce and the fact that we depend for our daily, even hourly, productivity on collaboration tools. That is another reason why we did the Virtuality Study: we didn’t even have the measure of our own “virtuality”. The results proved more dramatic than we ever anticipated and further drove the need for new solutions. A key finding from the study was that tools were the biggest factor in performance. Earlier research to the contrary, people trusted each other on virtual teams. The rollout that you did of the standard collaboration tools improved satisfaction with IT collaboration tools in the short run (not short enough—there is a lag) and performance metrics finally went up in the fourth year of the study. But there is still room for improvement. The fact of multi-teaming and round the clock called for the features of context retention and team spaces which Miramar offered. And the Croquet/Qwaq piece offers the “deep structure” of scalability across a large and distributed enterprise.
Sep 28 | John Light said:
I don’t have much to add to the story after all the other comments. I want to corroborate JDM’s telling of the origins of Miramar and Marc’s telling of the larger context we were working in.
In that line I will add that JDM went to the Media Lab with a huge amount of work by our group in his mind. I’d say he had the whole problem fully exposed after man-years of research in our group and others, so something was likely to happen in such a creative environment. We all knew there was a “a pony in there somewhere”, so when he came back from MIT with a compelling idea, the whole team was ready for it.
The point I’m making is that the “unproductive ” couple of years before MIT were critical to the “insight”.
Barely mentioned is the wealth of user feedback stored under JDM’s desk from the four rounds of user testing. The feedback was astounding in how ordinary user’s anticipated, accepted, and demanded access to what we were showing them. It even now brings tears to my eyes.
Sep 28 | Eleanor Wynn said:
“tears to my eyes”—employees put so much of themselves into creating innovations, but look what it takes to make them actually see the light of day. That is what is missing in the standard accounts of innovation…people say they take “passion” but passion requires personal involvement, which tends to get underrated when it comes time to cut costs…
Sep 29 | Michael Molin said:
That’s wonderful, Eleanor. Thanks very much. But my path is to give the world a personal device for interaction with our real world - Cell PC Plarform (really I have been working with Intel Software Network on this for two years already). And right now, it’s really interesting with Moorestown platform launch.
Oct 01 | Cindy Pickering said:
First of all I want to thank Intel Research and Justin for giving us our moment in the spotlight during Justin’s recent IDF keynote on 3D virtual worlds. We represented the “enterprise” angle among a broad spectrum possible usages that included virtual surgery, multi-player online gaming, education and social/consumer virtual worlds such as Second Life. It’s great to get this kind of visibility (even though I was only backstage).
I decided that I would focus on the IT Research and Innovation side of the story and chime in with some of my perspectives that havent already been covered in the earlier comments from various contributors to the collective work. Much has already been said about Miramar and its IAL origins. Mention was also made of the Virtual Collaboration Research Team (VCRT) in IT. These are the 2 primary origins of what culminated in our signing the Miramar Qwaq deal July 2007. While Miramar was on the shelf circa 2001-2004, there was a great deal of attention being paid to defining the vision and strategy for global team collaboration in IT. For example in 2001 Collaboration was a key topic in IT SLRP for the first time and I was a primary content expert for that topic. I also masterminded the IT Productivity Game Plan in Nov 2001, where we set the goal for 100M savings goal from Productivity and decided to form the VCRT. After its launch, the VCRT met regularly to discuss their vision of Collaboration based on internal research, business requirements, industry trends and technology advancements. 2002 was a very significant year for me — one in which I established the IT Concept Cars Program, under the sponsorship of the IT Research Committee. I also led the definition of the collaboration architecture for IT products and services, and worked with the VCRT to define the Vision. 2003 was even more productive. In 2003, the VCRT got its arms around the vision when we decided to leverage the Concept Car Program to build a concept animation that enacted what we meant instead of trying to show it with powerpoint slides. At the 2003 IT SLRP we unveiled the Vision Concept which was 7 minutes long, and somewhat esoteric in that it was based on a reasearch team like ourselves whose participants were on multiple teams, and whose days alternated between going to meetings and working individually on team related assignments. Doug Busch, our CIO at the time, asked for a 3 minute snapshot, that he could use externally a week or so later. Since we had done our homework to discover collaboration needs in the enterprise (Eleanors Virtuality Index and IT Segmentation Studies), I thought our scenario should focus on a typical business problem in a real world setting, so I rewrote the script with Eleanor, and did daily code review drops with Keith Feher until we achieved our goal, which was wildly successful and by the way, also narrated by Larry Shoop .
We submitted an Invention disclosure form and patent application (H2 2003) for our Multi-team Immersive Integrated Collaboration Workspace innovations. The embodiment of the invention was a seamless collaborative workspace in which individual users could keep track of all of multi-team activities as a whole, work without time and location boundaries, interact expressively with remote team members and move effortlessly among applications and team workspaces.
We also extended the Collaboration Concept animation to include business intelligence and mobility that year, which influenced the 2003 IT Product Line Business Plan. Meanwhile the Digital Office Program at Intel began getting intererested in what we were doing and asked us to participate in their needs assessments and market segment analysis, as well as ongoing Product Opportunity Planning sessions. In 2004 we did a joint Concept Car with the Digital Office team to illustrate integrated asynchronous meetings and converged communications; the demo was used in their quant and qual studies that year. Doug Busch presented the Collaboration Vision Concept Car at the 2004 Spring IDF in Beijing. We also took the concept demo to the 2004 Research at Intel Days where Gartner analyst Tom Austin asked for a follow up interview and included our work in a presentation at the Gartner IT Expo/ Symposium. This paved the way for future inclusion in the 2005-2007 Research at Intel Day showings. Eleanor’s and my November 2004 Intel Technology Journal paper, “An Architecture and Business Process Framework for Global Team Collaboration” provides a good summary of work up to this point.
Meanwhile the VCRT had decided to disband and Brad Anders, Chuck House and I were cooking up the Collaboratory. We wanted to make the team official and establish a research lab with a real mission and charter. About this time, IT was undergoing some organizational changes and we hooked up with Martin Curley, director of IT Innovation and Research who help us champion the Collaboratory with IT Staff and for funding in plan 2005. Of course we didn’t get everything we asked for but we got enough to begin moving us to the next level. We established the lab and a few key programs — one of which included vehicle for evolving the collaboration concept into an interactive research prototype that people could begin using. Ever at Eleanor’s encouragement, I hired JDM in February 2005, to do the research prototyping using Miramar as our vehicle. Of course this involved significant upgrades to the Miramar codebase to comprehend .Net, and other “advancements” that had occurred since 2001/2 when it was put on the shelf. More importantly, Miramar was initially designed for single user, as a desktop application and we needed for it to be a multi-user app. JDM worked miracles and at the end of 2005, we had the rev 0 multi-user version of Miramar and were able to recreate a generic team example that represented the VCRT collaboration vision. During that time we took Miramar to Research at Intel Days and also appeared at 2005 Fall IDF on the demo showcase floor where users voted us third place in the top Intel pavilion Demos. In 2006 JDM worked with a few contractors to incorporate Skype, document versioning, a robust installer and a developer SDK. In 2007 we began working on the Qwaq deal, because we decided that we need a path to productization if we were going to get to the next stage since we didn’t have the internal resources to productize our prototype. (You already know about the Qwaq deal based on what others wrote. We had tried out the idea previously with other prospects but it didn’t work until the 3rd try, and we couldn’t have done it without Chuck House and Luciano Oviedo).
While JDM is off working the combined product with Qwaq, I continue to champion its internal usage through our IT Innovation Zone, which is a community of innovators inside Intel who share HW/SW innovations via an internal IT website that provides secure legal downloads and community support. Martin Curley continues to sponsor us and provide the infrastructure of the worldwide innovation centers and personnel to help showcase our ideas and also contribute to potential usages and requirements from their customer base. For example in the Ireland Innovation Center, Tom Byrne masterminded several usages scenarios in the areas of electronic medical record, IT operations monitoring, and Factory automation systems monitoring. He and Alec Leckey, also from Ireland, worked with Eugeny Samsonov, a developer in the Russia Innovation Center to further develop the Operations Monitoring capabilites. Eugeny used the Miramar SDK and his web service development expertise to integrate Miramar visuals with events from traditional IT monitoring apps. For example, an error or issue with a server gets detected by MOM. The affected server and other impacted components can be represented visually in Miramar by blinking and changing color, bringing them to the attention of the IT operators, who could then collaborate to solve the problem. Eugeny also figured out how to populate Miramar programatically from a Web Service, making it easier to integrate with legacy IT systems. Their work was used in the IDF keynote demo. Thanks Guys!!!!!
As you can see, this entire effort has grown and evolved significantly. Over time, people have come and gone; some have persisted and dedicated themselves to making things happen throughout. There is still lots of work to do, but prospects look more promising, especially now that Intel is turning some attention and interest to this area, as we saw at the recent IDF.
Oct 01 | Eleanor Wynn said:
Thanks Cindy, you have the full story. The moral here is that it “takes a village” tho it also takes a consistent leader, like you, to keep the follow-through and the organizational coordination. Renee James, Intel VP of Software Solutions Group wrote an internal post on “communities create innovation”. She is right on target, and here is the case example.
Oct 05 | jdm said:
Yes, it “takes a village” - of people willing to stick their necks out, and keep them out, even when something else would be safer.
As the person most directly benefiting from other people’s involvement over the years, I’d like to publically thank y’all for helping keep the dream alive - John Light, Marc Millier, Cindy Pickering, Chuck House, Brian Gorman, Eleanor Wynn, Keith Feher, Gene Meieran, and Rick McGeer, without your constant and ferverent evangelism and support, we wouldn’t be where we are today.
Let’s also remember the technical contributions by Marc Millier, Matt Grenby, Sunil Kasturi, Mark Hancock, Jim Dowling, Mike Jeronimo, Dick Bowler, Greg Bohrn, Jason Shigley, Jim Sunderland, Lee “bogus” Boekelheide, Tim Roberts, Tom Byrne, Eugeny Samsonov, Keith Fehrer, and especially John Light (with apologies to anyone I forgot to mention).
And let’s not forget the HFEs, Gene Lynch and Doug Sorrenson.
Let’s also thank Qwaq as our extraordinary new collaborators on collaboration, esp. Greg Nuyens, Remy Malan, David A. Smith, Andreas Raab, Josh Gargus, and Brad F Fowlow, along our Intel bizdev driver, Luciano Oviedo.
Finally, out of that entire list (and I’m sure I left some out…), I must especially recognize John LIght and Cindy Pickering as not only supporters of Miramar, but equal partners in the endeavor.
Onward!
— jdm