Jeremy Allaire: Internet Platforms

Maven Networks Launches

Mon, 22 Sep 2003 15:44:05 GMT

A company I have been working closely with over the past six-months, has finally taken the wraps off its products. Maven Networks today launched an exciting new end-to-end system for delivering high-quality video applications to broadband-connected PCs. I've written up some thoughts on Maven Networks.

PingID supports WS-Federation

Sat, 20 Sep 2003 11:05:58 GMT

I heard about this, but was happy to see Marc Canter comment on the fact that PingID is releasing a prototype of WS-Federation. Ping have shown themselves to be a visionary company that is agile in addressing emerging standards. I've looked closely at the Liberty Alliance Phase II specifications and WS-Federation, and they are remarkably similar, which creates a tough choice for customers looking to federated identities between websites. Knowing that there is a software company that will protect your investment by supporting all the standards in a federation gateway is a big plus. I'm headed to DigitalID World in a few weeks in Denver, the center of the universe for digital identity issues.

Internet Convergence 2.0

Fri, 22 Aug 2003 13:54:23 GMT

There's an opinion / trends article I wrote posted on Macromedia's DevNet website. It gives a cursory overview of ten trends that are driving a new Internet convergence. Enjoy and would love comments!

Broadband Bounty

Thu, 08 May 2003 20:56:59 GMT

British broadband brouhaha British Telecom is being criticized for lowering its wholesale broadband pricing to a point that leaves a £1 ($1.57) a month difference between its ISP fees and what it sells bandwith to rivals carriers. Talk about a thin margin. It only goes to show that the bits are the commodity and content is the value creator. Here's why AOL and MSN, as they face the transition from dial-up customers to broadband, are in such a fix -- they've been selling bits, that's what the marketing has been all about: "Get connected." Well, now folks are connected and the reason they'll pay for faster connections is more/different content and services. But, because of actions like British Telecom's, connectivity is too cheap to earn a significant margin and they are being forced back on content. The message in the BT move is that the content challenge is sweeping the globe. Yahoo can (and has) taken advantage of this by working with BT to sell broadband services directly to consumers in the United Kingdom and collecting a bounty for new customers, as it has with SBC in the United States. It's already built its business around aggregating content and as long as it doesn't charge for content will be able to take providers' money for bringing an audience, as well as collecting broadband bounties.

[via RatcliffeBlog: Business, Technology & Investing]

Good stuff from Mitch.

Intel's wireless vision

Thu, 08 May 2003 20:28:05 GMT

Intel's wireless vision The Register: "The endgame, says Intel, is 'reconfigurable, intelligent CMOS radios'."

[via Werblog]

This is what MillennialNet is doing.

Wireless Standards in the Home

Thu, 08 May 2003 20:27:00 GMT

Ubiquitous Wireless Broadband for home media The ultra-wideband community is looking to position UWB as the dominant technology for moving media streams between devices in the home by making it into "wireless Firewire" (Firewireless?). UWB has some characteristics (including high bandwidth and low power) that may make it better than WiFi variants for these uses. It's quite possible that four classes of unlicensed wireless technology will all thrive -- WiFi, UWB (aka WiMedia), Bluetooth, and WiMax (802.16) -- though Bluetooth may be the odd standard out.

[via Werblog]

Interesting take from Kevin. In looking at opportunities for home media redistribution, several folks have argued with me as to the appropriateness or robustness of Wi-Fi for reliably moving high-bandwidths. Today, they're right, but tommorow as 802.11g and improved antenna and access point technology come to the fore-front this may no longer be an issue.

Interestingly, at the same, another unlicensed spectrum standard is emerging around RFID tags, dubbed ZigBee. ZigBee uses IEEE 802.15.4 RF as the physical network and provides a services layer on top of it. One of our investments is in MillennialNet, an MIT-derived company building a software networking platform on top of RFID for sensor-connected devices. They're also helping to drive ZigBee, along with Motorola and others. Millennial's software+hardware reference designs can enable intelligent, networked devices that have 5-10 year battery lives. Examples of ZigBee applications include:

Wireless home security
Remote thermostats for air conditioner
Remote lighting, drape controller
Call button for elderly and disabled
Universal remote controller to TV and radio
Wireless keyboard, mouse and game pads
Wireless smoke, CO detectors
Industrial and building automation and control (lighting, etc.)

PingID ships open source Liberty SDK

Thu, 08 May 2003 20:14:36 GMT

PingID has just shipped the 1.0 release of SourceID.Java, their open source SDK for Liberty-enabling any J2EE application. They're also a few weeks away from a full .NET implementation of SourceID, which will mean J2EE and .NET applications will be able to federate authentication using the Liberty Alliance 1.0 specifications.

Unlike Passport, Liberty specifications, and SourceID in particular, make it relatively straightforward to enable single sign-on across websites both inside an enterprise and across the public Internet without a central identity repository or service.

This is a great step forward for identity management and should help to give Liberty more momentum.

Talk by NTT DoCoMo Managing Director of Strategy

Sat, 26 Apr 2003 00:03:00 GMT

Takeshi Natsuno, Managing Director for DoCoMo Strategy

Keynote this morning from NTT executive, talking about i-Mode.

He's showing us subscriber data for wireless Internet in the world. NTT dominates in the world, and then next five are Japan and Korea. Everyone else, including Europe and US are tiny in comparison. For many reasons, this hasn't been repeated in other geographies.

He outlines the functional footprint of the 504 series. Interesting, the Nokia 3650 which is $150 in the US now has more functionality. He's showing powerful Java-based games, it's pretty sweet. He says the game software industry for phones is becomming as big as older game industries.

They just announced the series 505 phones. He's really pushing Macromedia Flash as the major breakthrough for their phones. But these devices have a richer display, 1.3 megapixel camera, memory stick, and bio-metric authentication input device! Phew. The Flash demo's are really nice --- smooth, vector graphics compared to the nastly bitmappy Java apps that he just showed.

38 million active iMode subscribers. More and more iMode has VPN support and can integrate into corporate environments. Therer are over 2,300 official commercial content/service providers in their network, with over 62,000 voluntary iMode sites. His 38 million number means that they are the biggest ISP in the world, now bigger than AOL.

Consumers pay for devices and access because they've made it safe and easy for content providers to provision services into their network. It becomes a virtuous circle as subscriber growth happens more content companies and higher quality content comes online.

They went from zero subscribers and content to this world given three key strategies:

1) Technology. Selected best available content technology -- real HTML (cHTML), MIDI, Java, Flash.

2) Business model. Designed business model around content service providers, not for the needs of telecom operators or handset manufacturers. Micro-payments model, user friendly portal for content selection, easy install and use process.

3) Marketing. Drove marketing around ordinary people, appealing to a mainstream audience not techies.

What he's not noting is the degree to which a) their incredible centralized control enabled the rapid and easy establishment of standards, and b) the advanced nature of their installed network technology enabled them to move much faster than US and European markets.

More data. Aggregate ARPU has stayed strong only because of iMode. Voice ARPU has continued to decline while data ARPU has made up for that decline.

Two years ago, strategic decision was made to add Java to the phones. They now have 17.4 million active Java subscribers, with over 600 Java app suppliers. There is also a market for Java-based corporate applications. There are 6,000 independent Java content/app sites available over iMode. Despite Sun, they made Java in devices successful.

As a result, they say significant packet usage boosted by Java handsets, in addition to higher value/priced applications where they get a commision. Latest Java phone subscribers are using 4x the packet usage than normal phone users.

He expects that with the introduction of Flash in their phones, they'll see a very significant boost to packet usage beyond Java.

Overall, digital content marketplace on iMode is about $1B in 2002, and the percent of customers paying for digital content is about 53% of subscribers.

He's comparing Japan versus other markets. He says it's fundamentally about a content-centric value chain all the way through the model. In other markets, the fragmentation in operator, handset vendor and content value chain makes coordination difficult. Operating a controlled but dynamic eco-system is the basis for NTT's success.

Going forward, always trying to improve the evolution in the handset and network capabilities; in content capabilities; evolution of the core user experience. But these evolving dimensions need to be synchronized.

What about 3G and Broadband?

The network speed component is only one dimension of the value-chain. The entire eco-system needs to evolve with it, and that can only happen when enough user critical mass is available. It's back to the chicked-and-egg problem that they faced when they started iMode.

On their 3G platform, they offer MPEG4 video, and it's evolving into being a multi-point video phone and better platform for content. 384Kbps currently. But they only have 380,000 users of 3G today because not enough applications have been developed for it. But they're driving customers over with a cost-savings value proposition, and then layering in specific 3G content services. 505i phone is the first to come close to enabling 3G applications. But it will take 3 years to get to 2 million users, largely do to lower area coverage.

Predictions:

1) US/Europe follow iMode model with content-centric value chain model.
2) Big migration from 2G to 3G starts in Japan in 2004.
3) Deeper convergence between Internet and telco
4) De factor standards most important.

My question: will devices evolve to be true consumer electronics devices that combine stereo music, DV/camera and video communications, eventually displacing portable music and camera markets?

He thinks it will in the long run, but it's no where near making this transition today. The music and camera capabilities are very far behind, and are really mostly applied in the special context of mobile communications. That will change over time.

Building the Infrastructure for Broadband, Panel

Thu, 24 Apr 2003 18:35:39 GMT

Jeff Huber, VP of Tech for eBay

eBay generates $681 per second. How does ebay compare to other commerce players? They're now 31st largest commerce provider, bigger than GAP, ToysRUs, and others.

Where do they focus? Front end (e.g. new and scarce products), end of life, and used/vintage products. Asser that this is nearly a $2 trillion market opportunity. Continue to see strong growth in most categories, but big in cars, home and garden, clothing and accessories. Operate 27 sites globally. Strong international growth.

<jeremy>Not sure what the relationship of this is to broadband infrastructure?</jeremy>

They're seing 6.5 million web service API calls per day! Sending 21 million emails a day, conducting 80 million searches a day. 30% of inbound products are coming in through virtual supply chain --- e.g. web services.

With regards to broadband, it's the always-on nature that drives the nature and velocity of usage of eBay, and the ability to have an immediate, notification driven relationship with buyers and sellers. From a bandwidth perspective, they can really improve the richness of the product views and details with broadband focused content. And broadband empowers more individuals to create eBay-based businesses.

They'd like to become an alternative distribution channel for any manufacturer or retailer, and using multiple pricing models, including fixed price increasing dramatically.

Question: web services adoption; how's it growing/going? what standards are you using and are you ahead of the standards?

We haven't been as high-profile about PR, but activities are ahead of expectations in terms of adoption. Went for a least common denominator approach --- e.g. it's not SOAP, but XML/HTTP and a RESTful API.

Question: what impact on broadband content will happen?

I think one concept that came early today is the idea of a tipping poing, but they're still only at 50% of their end-users having broadband, so sellers are reluctant to use broadband rich media content. Typically, the more expensive and advanced product areas are using broadband ready content.

Question: how has the reputation system on eBay supported commerce?

One of the key issues they see is customer safety around the experience. People just don't trust the system in general. Trust and safety systems, including reputation, are a big part of that. Fraud management systems are a part of that, too.

David Labuda -- CEO Portal Software.

Broadband will surprise us. The apps that take off were never predicted. Examples --- eBay and commerce; SMS and Mobile; ring-tones and mobile. Nobody knows what the killer apps will be. Must be able to launch services quickly and inventively, and fast experimentation.

The ability to intuitively price these new premium services is a real challenge, especially for communications companies used to just charging for data rates and time. There really isn't a mapping b/w bandwidth used and value of the service. This is a real sea change for a lot of these companies. This is value-based pricing. This could enable QoS models and cost structure savings to increase margin on high-value items.

Security and fraud prevention is a major issue still. This is paramount because otherwise the communications companies will have terrible experiences and will back away.

Question: In favor of value-based pricing, but some of the pricing will be above and below the cost of bits.

Lots of people will try and beat the system, and you can try and reduce that problem, but it won't be the norm. Agree there will be a variety of low and high value transactions, but that for consumers you can't have bandwidth-based pricing.

<jeremy>We're mostly talking about the mobile market, and the drive to paid content for services rather than raw data as the unit for pricing/billing. The discussion could have benefitted from this clarification, as many people think we're talking about broadband suppliers charging for access to third-party apps and content</jeremy>

Rick Rashid -- SVP Research at Microsoft

It's been 10 years since Mosaic. It was really a tool for people to interact with computers at remote locations. Driving change will be computers interacting with computers, and then servicing the consumer. So with broadband in the home, the computers need to be intelligent and able to get data and services on the network all the time.

Today TV is more like browsing and you get data on the fly. Increasingly, machines in the home wil pull content over the network and assemble programming for you on the fly, when you want it. We've seen this ourselves in our own research environments. For example, TerraServer went from being a browser app to a data service, and more and more its apps communicating with the the server for data.

Effecitvely, what the user sees are specific apps/services that are local on their machines, with communication and data happening in the background using XML web services. So this is the real change and trend, this new distributed computing infrastructure. Next is defining how we do trust, communication, discovery, and so on.

Question: value of the data is the real deal, but the communications linkages are not the value?

Yes, that's right. You're seeing very low costs for data communications --- dark fiber is getting dirt cheap. What's the real value? It's in what you do with the data and services around it.

Question: doesn't the web services strategy, with horizontal services, wont they also be commoditized?

Yes, that's right. You'l get to a point...what is the specific value of the info/service you're providing? If it's unique you can drive some revenue from it. In many cases, if anyone can do it, then it will be a commodity. Web services does help this commoditization. The other side of the coin is that it enables new kinds of apps to be built. People are building data federations --- linking past databases using web services, creating a facade of data services with higher-order value (e.g. aggregation and re-composition of services).

Question: how does one maintain their brand identity in the web services space?

eBay (Jeff Huber) -- right now its additive, not detracting. This is mostly on their supply-side versus consumer side.

Question: are consumers comfortable with computers acting on their behalf?

Paul Florack (VeriSign) It is a key issue, and trust and authentication are crucial for consumer acceptance. We'll leverage certificates for verification across a number of fronts involving web services.

Rick -- we've been concerned about this notion of there being no one place that can be trusted. consumers will trust some things and not others, and thus defining web services security, is the need for multiple forms of certification and proof, and the idea of federation of trust, so you don't have to rely on a single trust mechanism.

Panel with Glaser, Seigelman, Eric Schmidt -- Broadband Applications

Thu, 24 Apr 2003 16:30:04 GMT

We have a distinguished group of panelists.

Russ Seigelman -- Kleiner-Perkins, ran MSN for Microsoft
Bob Meyers -- President of CNBC
Salvador Arias -- IBM consulting, communications practice
Rob Glaser (via video) -- RealNetworks CEO
Eric Schmidt (via video) -- CEO of Google

Thoughts from each panelist.

Russ -- big believers in broadband. what happens next? hard to know. what's going to happen to usage? broadband drives more use of the Internet. believes that 1:1 video calls will be a big growth driver, will happen in mobile and at home. business model of broadband is different --- lots of infrastructure investment will mean that there will be more tolls in front of the consumer. it's the only way this will happen.

Eric -- the industry has been in a disaster for 2-3 years. the issue has to do with the fact that computers are getting more powerful, and the value equation hasn't kept up with PC capability. great thing about broadband is that it's the only technology that can meet the capacity of PC/CPUs. mostly broadband has to do with persistence of connection (always on aspect) means a very different usage. last year world switched from analog to digital devices in the consumer world. new generation of digital devices, combined with broadband persistent connection, is the killer app. vast majority of broadband adoption will be non-traditional, self-published audio and video. result will be new works of art.

Bob -- work at CNBC, part of NBC, part of GE. seek out environments where there are new challenges. challenge of working on point-to-point, interactive versus broadcast world. who's gonna pay for Internet delivered content? who's gonna make the transition from traditional media into networked, IP broadband environment? in the old day you have a dumb receiver (TV), receiving from a dumb but efficient network (VHF, Cable). now the receiver and the network have more intelligence. and the programming itself has a lot of intelligence.

Rob -- want to add to the comments already discussed. Real's focus is on delivering a/v apps on the Internet. started with narrowband, but focus was to design for the broadband world. our sense is that in the mid-term, the medium will have the same transformative impact that cable had over broadcast, in terms of impact on business and programming models. Real sees growth from dozens of channels that are profitable, to tens of thousands of profitable new channels. Result will be increase in consumption. There will be a substitution effect. Global medium, no channel or spectrum constraints, spans consumer and business equally. Takes issue with Eric's idea that it won't be traditional programming/video that has distribution. Real is now broadcasting 50 MLB games live over the network today, way more capacity than any cable provider. ABC news is now doing 24/7 live news, all delivered over the Internet. This is meaningful from a traditional media distribution perspective.

Salvador -- still focused on what's happening with the pipes and the economics of consumer adoption. four key factors: 1) competitive dynamics b/w suppliers are driving prices/offers down, 2) need to have enough bandwidth to enable compelling apps (e.g. above 1.5MBs, esp video centric apps), 3) pricing elasticity having an impact, lower cost structures helping, higher demand equates to economies of scale 4) government intervention --- what will happen with government support to drive adoption; still uncertain, mostly industry driven now. momentum is building for government supported investment protection (e.g. tax breaks, subsidies, etc.).

My question --- original programming, what will it be, will new companies emerge.

Rob. Lots of tries during dot com era, and most tries were too early. Will play like cable tv, like early days when MTV and others started with low-cost programming. Even ESPN and CNN were very parsimonious with cost structure. Kinds of programming --- number one thing that you can do that you couldn't do in the past, is to add interactivity as an intrinsic component to the video. Customization of programming, personalized information tools. Other applications include community integration and real-time communications around the programming. Social environment in addition to a programmed environment.

Bob. Do you want to make it a business? It's hard to make a business out of a much smaller niche service than we have today with large-scale media already. As far as adding elements of interactivity, to realy create a new channel that will be video based, person-to-person; not sure how to do it, how to find it, but he can't see a way to a breakthrough that this will be a new business for them.

Russ. There are now cable channels that are connected to live events on the Internet --- polling, video, community interaction.

Question: how much control will consumers have over timing of viewing of programming? hard to imagine Tivo'ing CNBC, he says. need to watch it live. says the key for broadcast is getting people to want to see it live.

Rob. RealityTV interesting use case. Doesn't have script format like other programming; it's the environment where media companies are combating traditional programming constraints; aggresive integration of marketing and advertising into programming. Media companies will integrate marketing/ads into programming to combat Tivo style recording where you can skip advertisements.

Russ. Time-shifting will work and people will pay for it.

Question: what will happen with paid content?

Rob. We jumped into paid content big time. As advertising online collaposed, broadband grew, so the only business model became paid content, with a set of free content around it. Pragmatic approach today. 900,000 paid subscribers with a broad range of paid content. Free consumption is still there and promotion and advertising supported, and its at an all time high in use. These two models reinforce one another, because the free model creates an audience, which in turns drives more paid premium content, which in turns draws more people to the medium. A la carte in music is just about to happen, its the play to save the music industry.

Eric. He disagrees with a few things. Says he's a huge fan of CNBC. Majority of broadband entertainment will be ad supported media. The advertising model is broken because of lack of ability to do sophisticated targetting. Tivo model enables the ability to interlace targetted advertisements. Online delivery in general supports this. High-levels of personalization will mean a fragmenting of audiences, with a deeper mixture of blended advertising, marketing and programming (as Rob suggests). Sees a huge change brought on by precision in advertising, and this is the revolutionary change on broadcast.

Question: cost of technology and running a media system is still extremely high, much higher than even labor costs. Couldn't break even on their projects.

Rob. Two different cost drivers. One thing you saw in the bubble was that there was lots of capital, and so people were used to spending tons of new capital. Real change since then and people aren't willing to do this so are looking for outsourced usage because its more cost efficient for the broadcaster/supplier. Second thing is the rapid drop in bandwidth cost. This variable cost is getting cheaper and cheaper, dropping 50% per year. And it's accelerating. So things like video on demand are now becomming feasible.

Russ. It's true at the backbone, but not at the end-points. This is still too costly and for not enough speed.

Rob. The guys going bankrupt are the CLECs, but the ILEC/encumbant, they're doing fine, and actually are growing and profitable. Those price points are a chokepoint on growth. This is a glass is 2/3 full situation. Broadband growth is happening. Cable guys are nailing their numbers every quarter. They're not seing a saturation at $50/month. We'd love to see $20/month and the dramatic growth that would happen with that. But this is a temporary inhibitor, and we're still seeing strong growth. Government should have a policy to drive prices down with subsidies, and yes, two player oligopolies don't price compete.

Question: broadband is turning the communications and media industry inside out. And that it's not having a lot of relevance for corporations?

IBM guy is talking about how all is good...not sure what he's really talking about....trying to incorporate IBM OnDemand marketing pitch...not connected....

Question: we're really dealing with duopoly. isn't the real issue price? won't there be real price competition.

Russ. The real competition will be from wireless for the last-mile, perhaps powerline. Real question is whether there is a price umbrella that will drive demand? We're doing well now in terms of adoption, but it's still a long way before we get to 70% penetration. Some scenarios could come in and really change the model, such as WiFi mesh networks. Mostly theory now, but could happen quickly with real business model innovation.

Eric. If you look at the adoption rate for tech as a proxy for the future, one of the fastest adoption curves is 802.11b/g. Everyone is embedding this uniformly --- MANs, corporations, homes, etc. Still some sharing, billing issues with mesh networks. It's pretty clear that the change to add the third wire will be distruptive, and will affect the pricing model in unanticipated ways. The way to get it down under $10 will be by users bearing some of the capital costs, and then participating in the mesh network.

Rob. These new environments become real new platforms. For example, I picked up a Cisco ATA router and have a VoIP phone device, and it really works. This is one of a number of apps that switch over to the broadband pipe. Big fields like education and health care could switch over to this medium.

Salvador. Telco thinks, if I got to an IP based platform, can I really simplify my cost structure positively. Lots of investment in their back haul and there's a concern about a displacment impact.

Question: the conversation has been around innovation around broadband, and talking about continuous video streams. the Internet showed that a very servicable pipe happened when the browser happened and people had power at the end-point, and the innovation was around what you could do at the end-point, not the pipe itself. real innovation is in the richer combination of media, communications, interactivity at the end-point.

Eric. There are a couple of historic transformations. The cost of published information --- any media type --- have dropped dramatically because of the digital revolution. How do we get more useful media? useful is in the eye of the beholder, and there will be thousands of capabilities. Blogging is the other trend, which is a constant stream of thousands of perspective. The Internet model gives the end-user more power and this is the real transformative impact on the broadband media world. On another note, people are using broadband to transfer to physical media -- e.g. CD and DVD burning.

Rob. DVD and CD burning, etc. The role that broadband plays in catalyzing media delivery in and around the home. Once you have it in the home, it gives the user much more control to redistribute to fixed media, to devices, to get it to other fixed end-points (e.g. stereo and TV).

Question: how many users will we need for the dramatic impact to society?

Critical mass will be 50% of households.

Question: how do you se this playing out in a nice near-term number --- say 2008? what will we be saying?

Russ. It depends on what your expectations are for the outcome. Maybe I'm just a dreamer. I want to walk around everywhere and have a video phone and have it be high-quality and everywhere. We'll be there after 2010. It must mean that the majority of people have that. It's not a 2008 kind of thing.

Bob. It doens't matter. We've reached a tipping point. The comparison between TV to Cable change isn't appropriate. We're talking about a communications medium, a revolution, becuase it's two-way. The things people do on broadband they already do on narrowband. Broadband just lets me do it better. Nothing radical happens in the next five years, we'll just keep evolving and the network effect will take hold.

Eric. In 2008, we'll all live in a parallel world. At the same time, a subset of us will have new digital device with all of the worlds information available immediately, on the device. I think there will be many thousands of chanells, derivative of what we see today, but highly personalized, and the vast majority will be small entities, with small audiences. Much more information and end-user empowerment.

Salvador. 2008 the cable world will have significant improvements, such as one network that does IP, video, voice, etc. It will clearly wipe-out dial-up and a broadband IP majority in the US. It's not going to be video as we think of it, lots of community oriented stuff, real-time communications, gaming oriented content. The world will be more on demand for us. Broadband will get there.

Rob. There will be multi-channel access to video and programming. He shares Eric's vision of the portable revolution, with highly reliable, caching enabled devices, driven off WiFi hotspots. Financial markets will see a great come back because of the inevitable progress of the Internet.

Go to CF_Europe, Europe MX Conference

Thu, 24 Apr 2003 14:27:06 GMT

I just wanted to plug the forthcoming Macromedia-focused developer conference in Europe, CF_Europe. I'm not sure why they didn't change the conference name, as the speakers and agenda are really very broad on all MX products, with a solid focus on building RIAs. Sister conference MXNorth in Canada was great, so I expect this to be as good.

The role of visionaries in the future

Thu, 24 Apr 2003 13:51:13 GMT

Rob Austin did a bunch of research and interviews with the creators of modern computing, like Bob Taylor, Alan Kay, Doug Englebart and Bob Metcalfe, among others. Here's some of the things he learned.

Bob Taylor is one of the early innovators in networking technologies and the Internet. Rob Austin, conference chair, talks through several examples from the late 1960's when Bob brought the idea of packet switched wide-area-networks (Internet), and IBM, ATT, and Xerox all passed. What would have happened had they adopted the idea back then?

Bob Taylor: "people tell me that the Internet happened fast; that's crasy, it took forever."

Some people can tell the future, gives examples of Vannevar Bush (hypertext), JCR Licklider (Man-Machine Symbiosis), Bob Taylor (The computer as a communication device), and Doug Engelbart (mouse, gui, real-time communications and collaboration). These were ideas developed in the 40s, 50s, and 60s.

He reviews some interesting quotes from his interviews, and interestingly nearly all of them share some things: they all are driven by the desire to help humanity; they all don't think we've nearly reached their vision, despite the progress that has been made since the 1960s.

Discusses the tension between invention/vision and short-term business objectives. There is an internal tension in business firms that attempt to reduce long-term vision and R&D because they way in which it can cannabilize current products, prices and profits. Cites Veblen (19th century economist) and John Kenneth Galbraith (contemporary nobel winning economist, advocate of "great society" economic policies under JFK).

Broadband Explosion: Opening Keynote, Internet 2

Thu, 24 Apr 2003 13:11:41 GMT

I'm attending the Broadband Explosion conference hosted by Harvard Business School. It's a great line-up of speakers and keynotes, and a small intimate crowd. Rob Glaser, Sky Dayton, Russ Seigelman, Kevin Werbach, Geoffrey More, Clayton Christianson, Rich Rashid, Eric Schmidt, and lots of other good folks.

Today's opening keynote is Doug Van Houweling, President and CEO of Internet2, the government sponsored research initiative to create an NG-Internet.

He's giving a little history on the emergence of the commercial Internet, via NSFNet and a partnership with IBM, MCI and the state of michigan. I remember this very well, and knew the folks running policy at NSFNet. It was clear back in 1993 that this was going to be turned into a commercial backbone, and the rest is history. But now the same model (government/commercial/higher-ed collaboration) is being used to create a next-generation Internet -- explosive broadband.

He's recalling the early strategic decision making that went on with IBM around NSFNet, and the controversy it had with IBM, who were marketing SNA networks as an alternate WAN topology. At the same time they had built core router technology for this early Internet, before Cisco even had backbone routing technology. IBM decided not to productize their backbone routers, and this opened the door to Cisco dominating the backbone routing infrastructure.

Internet 2 Charter: enable new applications, re-create R&E network, transfer technology to production Internet. It's recreating the NSFNet model, etc.

202 universities have joined the Internet 2 project, 50 corporate members, 30 government and research labs, and over 30 international partners.

Compares time to download The Matrix DVD using various connectivity tech. 56k model is 171 hours, DSL is 25 hours, T1 is 6.4 hours, and on tests on Internet2 it's 30 seconds. Sounds good to me. :)

Blasts today's broadband. "Real Broadband", as Marc Canter would say. Multi-megabit bidirectional, and services rich -- e.g. multicast, IPv6, deep facilities for measurement, monitoring, metering.

What current experimental applications are being deployed on this today?

Media distribution --- away from sattelite to IP digital distribution; HDTV over broadband. Super quality video. Experiments with 4x quality of HDTV.
Telepresence. Extremely high-quality remote presence and monitoring. For example, underwater observatory with mobile monitoring, cameras, etc.
Robotics and bio-feedback. Devices that record behavior -- e.g. a surgeon working with instruments, and end-users who can re-experience that movement with simulated devices, wearing VR goggles.
Real-time collaboration. High-quality multi-location video conferencing with screens the size of whiteboards.

The infrastructure needs must change dramatically. Today's Internet is "best efforts" Internet -- e.g. make a best attempt to get data there quickly. Internet 2 is guaranteed level of quality, which translates to a much larger set of pipes, routers, etc., but more importantly is a services infrastructure in the network geared towards quality of service such as IPv6 and multicast.

Also believes Internet 2 needs to embrace federated policy models for security and trust. They aim to build proof-of-concepts for security models that work across enterprises. They're working with the Liberty Alliance and Microsoft, and trying to get something working today in the Internet 2 architecture.

Biggest challenge is the last mile, given the role of the end user in distruptive change and value on the Internet. Internet 2 can enable Gigabit to the home, but this requires a partership between government, community and commercial service providers.

Working with Dave Winer

Thu, 24 Apr 2003 03:18:15 GMT

Dave and I caught up last week over lunch at Henrietta's Table, a nice lunch and meeting spot at the Charles Hotel. He was one of the first victims of my new Nokia 3650 device (left picture was beamed to my PC when I was within range).

As he noted about our conversation, we worked together in the past on the early genesis of "web services" as found in WDDX and XML-RPC. That was a fun time, and it's rewarding to see these concepts flourish on the Web today. One thing that both WDDX and XML-RPC had in common was that both used relatively simple data models, which made them more readable, interoperable and easy to implement.

Now that Dave's back east and at Harvard, we've got lots of good connections to make, as he noted.

Intelliphones

Wed, 23 Apr 2003 21:15:36 GMT

Russ raps it out Delineating Devices: Current Generation, Multimedia Mobiles and Intelliphones.

I was struggling to explain to my wife recently what I was working on and the devices I was targeting. In that discussion, I ended up coming up with a sort of categorization that I'll talk about here. It's really helped me understand what I'm working on and for who.

The idea was that I was trying to explain the difference between the super-powerful Nokia 3650 and Jim's new Siemens S55. They have so many things in common that it's hard to differentiate to the non-Mobile-obsessed. I was explaining to Ana that Nokia is doing something very weird in my mind by pushing the 3650, which is a phone with incredible potential and capabilities, as a competitor to phones that have much less power.

Briefly here's how I how deliniated the different phones that are out there. I just divided them into levels, though maybe this is a bad idea because of the confusion with "generations", i.e. 2.5G vs. 3G, etc.

Anyways here's my thoughts:

Level 1: "Current Generation". These are the mobile phones that most everyone has now. They all have grey LCD screens and only support vanilla GSM or CDMA. Some of the later models may have polyphonic ringtones and WAP, but for the most part these phones are used for calls and SMS messages. The advantage of these phones is their simplicity and recently their size. My last phone was an Alcatel m5510 which was sold for less than $100 and weighed only 75g - everything in comparision to that phone seems like a brick. But most of these phones are a bit older and are still pretty hefty. Example phone: Nokia 3310/3330.
Level 1.5: "Current Plus". These are phones that may have some of the features of the next generation phones, but not enough to push them up the ladder. A good example is the Siemens C55 which has GPRS, but a monochrome screen and no MMS.
Level 2: "Multimedia Mobiles". These are the next generation mobiles that are going to be pushed by just about everyone. Vodafone is basing their Vodafone Live! around these phones, which have a minimum functionality of the following:

Color Screen
Camera: Integrated or Attachment
MMS - Multimedia Messaging
J2ME - Java games (maybe BREW or Mophun)
WAP 1.2.1 (needed for MMS)
Tri-band/GPRS/Higher speed CDMA
Polyphonic Sound
These phones are generally heavier than the Current Generation, but not as heavy as Symbian phones.
Level 2.5: "Multimedia Plus". I think these are the phones that add a bit more to the package:

Bluetooth
MP3/FM Radio/Stereo Playback
WAP2/XHTML-MP
Email Client
PC Sync / SyncML / iSync
3G Connection over UMTS/CDMA2000 (i.e. phones from Hutchenson's 3 service).

Level 3: "Intelliphones". These mobile phones are more akin to small computers or powerful PDAs than anything that preceeds them. Though they MAY have many or all of the capabilities of Multimedia Phones, these devices normally have an ARM-compatible processor running in the 100s of MHz, a real OS (Symbian, Palm, Linux, M$), a real file system and usually much more memory and/or expansion slots. You can install custom software on these phones like on your PDA, and do everything from play multi-player Bluetooth 3D video games to using a real web browser or viewing streaming video. Example phones: Nokia 3650, SonyEricsson P800.
Level 3.5: "Intelliphone Plus". These don't exist yet - or at least aren't commonly available - but are Intelliphones plus the higher bandwidth of 3G or WiFi or with GPS or with megapixel cameras, etc. Samsung's coming Symbian-based 3G phone will fit this category.
Level 4: "Future Phones". These are phones you usually see mocked up in 3G promotional brochures. You can safely ignore this category for at least 5 years.

So the problem with these classifications is that they're far from absolute. The specs for each phone range all over the place (the SonyEricsson T68i for example has everything it needs to be a Multimedia Plus phone, yet doesn't have Java or Brew). However this is meant just to be a guide and I think it's pretty obvious where the sweet spot is - the same place that Vodafone is aiming right now, at the Multimedia Mobile phone market. MMS, Java, Ringtones, WAP. It'll be interesting to see what great apps arrive for the intelliphones and to see how far they go - and to see if Nokia is able to Trojan Horse enough 3650s into the world so that the intelliphone market gets a major boost...

In general I've decided to target level 3.5 phones with my latest efforts, under the idea that you don't develop for what's available now, but for what's going to be available soon. Otherwise you end up finding work arounds and solutions for hardware that will soon be sitting in someone's drawer somewhere.

-Russ [Russell Beattie Notebook]

I now think of Russ as our 'mobile guy. He's working on something we ALL can use! Thanks Russ!

[via Marc's Voice]

Thanks for finding this Marc. This is good view of the richness and growth trajectory for mobile phones. I'd like to add two things to this. First, I just became an "Intelliphone" convert with the purchase of a Nokia 3650 with ATT Wireless service. This device is just awesome. Russ outlines the functionality above. ATT is subsidizing this with a $150 rebate, so this device is now just $150 in the US, so there really no reason not to buy. :)

The second thing is noting that, presumably, at some point we'll have a version of Flash in all of these devices. Some people noticed the announcement earlier this month about NTT DoCoMo integrating a light version of Flash into their new phones this year. We can only assume that that will find its way to all good Symbian devices, just as Flash 5 is in Nokia 9x00 series devices. I've looked at J2ME apps, and they're gritty and hard to build. WAP 2.0 and XHTML is just not a great form-factor and viewing experience. I want vector graphics on these little engines, and a richer model for integrating sound and graphics into content and applications.

Centrino: Trojan Horse for Future Cell Data

Tue, 22 Apr 2003 21:31:22 GMT

Centrino: Trojan Horse for Future Cell Data Intel's trying to become the life of the Wi-Fi party: My analysis of Intel's Centrino plans appears in today's Seattle Times. My thesis, in a nutshell, is that although putting 802.11b in a modern laptop isn't a stroke of genius, Intel's actual, stated goal (if you listen to them closely) is much clearer: to provide an end-to-end experience for purchasers of Centrino laptops (configuration, troubleshooting, connection) and to have a path for them to offer ever more advanced wireless technology in the same form factor to manufacturers who are already happy with the process. You could see within a year or 18 months a Centrino offering, for instance, Wi-Fi (g or a/g) plus GPRS or a 3G flavor with Intel's branding on the service. You might be buying Intel Centrino Service by Cingular or T-Mobile Centrino for Intel or whatever the combination of brands is. But it would be a single brand promise and an end-to-end promise, too. Right now, there's a lot of finger pointing when you buy and try to configure a Wi-Fi adapter. Most of the time, it works. When it doesn't, who do you complain to or even get tech support from? When Linksys and Orinoco cards I purchased didn't work in a Sony laptop, I sent email to five different companies and received 15 to 20 suggestions. Fortunately, the last of these, which trickled in, had the solution (Wireless Zero Configuration was turned off). In the same circumstance, if I had a Centrino laptop, I could call the laptop maker, and if I wasn't happy with their help, Intel has a staffed Centrino support line I could call. Both tech support operations are supporting the entire chain. I'm not going to get (I hope) a cock and bull story about it being Microsoft's fault, the driver's fault, the hardware's fault. Let's take one alternative, too: Dell is offering Broadcom's g and a/g solutions under its own TrueMobile name. The g card is a zero-cost sidegrade from a Centrino system to an identical Pentium-M/855 system. If something goes wrong with this combination, I'm entirely reliant on Dell. (Dell is selling a/g and g into the enterprise mostly. If you're planning to upgrade your WLAN to g later in the year, buying a Centrino now might be foolhardy -- and I'm guessing a lot of companies who have thoughts of g or a/g are holding off on...

[via Wi-Fi Networking News]

Good analysis on Intel's long-term ambitions. The likely collision of WiFi+2/3G cellular with multimedia communications will become a great place for hardware, software and service platform brands to establish themselves. Nokia's doing it in their universe. Intel would love to broaden out beyond their current role as hardware supplier up to a higher-level brand and offer for consumers, and mobile appears to be a big play in that direction.

iPronto Home Automation Client

Tue, 22 Apr 2003 21:10:34 GMT

Look what Don Norman hath wrought! Information appliances!

Philips iPronto Does It with Linux [Slashdot]

In addition to providing a traditional infrared interface, the iPronto also implements Wi-Fi (802.11b) wireless connectivity and even an Ethernet port for LAN or broadband Internet access. There's also a USB port as well as an MMC/SD card slot for connecting to other devices (digital cameras?) and numerous other expansion and upgrade purposes.

The iPronto's "customizable" graphical user interface consists of a "high-resolution" 6.4-inch touchscreen LCD, along with a group of stylish control buttons. The device also has a built-in microphone and stereo speakers which Philips says will allow users to listen to MP3s streamed from the Internet (and, presumably, from other home systems via some sort of home network), and which may some day enable futuristic uses such as voice recognition and IP telephony. [LinuxDevices]

$1,700 seems like a bit much for this device, but if this is what the Linux desktop looks like - then cool! Where's the hard drive - why shouldn't this be a Wifi server as well?

[via Marc's Voice]

I'm a user of the prior release of the Pronto, and I have to say it's a nice device. Having the Internet-connected version doesn't seem to add much. The folks who'd use this device already have a WiFi enabled laptop, so they're not going to use it for accessing Internet services. And while it can theoretically talk to other home platforms, virtually none of the mainstream consumer electronics speak any standard protocols or let alone IP. Even mega-automation companies like Crestron (high-end consumer and corporate a/v automation) advertise an "open platform" based on IP and Ethernet, but when pressed the sales folks couldn't point to any other platforms (lighting, security, a/v gear) that did too.

Cisco to Ship WiFi Phone

Fri, 18 Apr 2003 15:48:38 GMT

I picked up this tidbit from a forwarded email, not sure of the original source:

A Cisco executive on Wednesday showed a Wi-Fi mobile phone to channel partners at the Cisco Partner Summit in Las Vegas and said the device will begin shipping in June. The phone, which will carry the model number 7920, communicates only with IEEE 802.11b technology and is designed for use within enterprises rather than totally replacing a cell phone, said Charlie Giancarlo, senior vice president and general manager of switching, voice and carrier systems. However, Cisco is in talks with cell phone makers about the possibility of adding cell phone capability to such a device, which might carry the Cisco brand.

WiFi versus 3G

Wed, 16 Apr 2003 19:26:17 GMT

3G.co.uk picked up on a report from German market research firm Metrinomics:

"WLAN will establish itself as the dominant mobile wireless infrastructure within the next five years at 3G’s expense according to a report published by Berlin based market researcher Metrinomics. Almost 90% of participants in the study also believe that WLAN will be the infrastructure of choice should their employers decide to implement an in-house infrastructure solution within the next five years."

Lots of debate on this, but it's nice to see more and more analysts stretching to this conclusion.

AOL gets videophone religion

Mon, 14 Apr 2003 14:18:51 GMT

AOL gets videophone religion

AOL Tests Video Instant Messaging [Slashdot]

OK - so since AOL refused to open up the AIM protocol and stops folks like Jabber from inter-connecting their IM system to "the rest of us" - they've come up with a sneaky way of moving forward.

Currently in beta testing, the new AOL feature will enable the service's subscribers to activate an additional pane on their IM client. Within that pane, users can record video clips of themselves via a Web cam, and then send that clip to the buddy with whom they're chatting. Users will have to click to open each new clip they receive.

AOL agreed to the limitations of their "advanced multimedia messaging" - as long as they were the predominant IM client. So they've instituted "push-to-talk" - which is really all that can be done today, as opposed to keeping a real-time video connection on all the time. This sort of videophone compromise will get them into the game, as apparently Microsoft has struck a deal with Logitech to co-brand their Video Companion software.

As many times as I told people that videophoning would be the killer app of broadband, I was told "it can't be done". Well now it is. Though it won't ever be more than 20% of revenues garnered, I predict that this sort of "store-and-forward" videophoning will be the catalyst that will kickstart broadband into overdrive. Kids, grandparents and broadband enthusiasts will eat it up.

Then mainstream business guys and gals will figure out they that'll never have to get onto a plane again - and they'll like that. If mainstream AOL picks up on this - buy stock in camera, teleconferencing and ISP companies. And sell your airline, hotel and rental car stocks. Broadband - here we come.

[via Marc's Voice]

I spent many months looking into practical consumer applications for async video communications. It will be interesting to see what the AOL implementation feels like, but the technology is definately there and the approach (async versus real-time) is the right one. My own experiments enabled people to easily record a video from a webcam or DV camera, and then send or publish through IM --- we used JBuddy to provide presence awareness and messaging into all the IM networks -- email or post to a blog using MetaWeblog API. It all worked great and makes tons of sense for broadband suppliers and consumers.

ENT 1.0 and RSS Metadata

Mon, 14 Apr 2003 14:14:00 GMT

ENT can provide great new functionality - NOW RDF still looking for a killer (set of) app(s)?.

<snip>

Lots of pratical kinds of apps and services could benefit from ENT and since it's RSS2.0 - all the current aggregators can access it. It'll be one of those quick adds, and give tool vendors all sorts of new ways to provide great value to their customers.

Maybe these tool vendors can start to charge more :-)

[via Marc's Voice]

Marc has been tracking some great action around a new proposal for adding topic metadata to RSS feeds. I've read through the spec and like the fact that it's a lightweight, simple to understand approach to topics, and it can scale as external/URI-based topic maps emerge. Would love to see popular RSS tools support this and see what can get done. This helps to open up the range of application types that RSS can support.

Hydra: Collaborative Text Editor

Mon, 14 Apr 2003 13:46:07 GMT

I picked this up from my comments on the Social Software post. Hydra is an OSX-based text editor that uses Rendezvous protocols to facilitate real-time collaborative coding. It's from a dev team (Coding Monkeys) out of Germany. Real-time apps continue to excite and impress, and it's great to see different approaches --- using Rendezvous, Flash Communication Server, Groove, or the new Microsoft P2P SDK.

Powell on Powerline

Thu, 10 Apr 2003 20:12:04 GMT

"This is within striking distance of being the third major broadband pipe into the home," Powell said. "I'm a little bummed it's not (available) in my area."

I think he means the forth major broadband pipe, as fixed and unlicensed wireless will be right in there too. But, it's great to see Powell pushing Powerline.

New Social Software Products

Wed, 09 Apr 2003 13:59:42 GMT

Social Software Making Progress The smaller the group, the more immediate value in the relationship. That's one notion behind an emerging phenomenon called ``social software"

[via Dan Gillmor's eJournal]

Dan's got some thoughts on the emerging category of "social software", a phrase Clay Shirky has been promoting. Another interesting company solving similar problems --- and one who was just on a panel moderated by Clay, also including SocialText --- is Providence, RI-based Traction Software, who's "enterrprise weblog" software is really a powerful distributed communication and publishing tool for information professionals.

Analysts: Wi-Fi a 'Positive Disruption'

Wed, 09 Apr 2003 13:38:59 GMT

Analysts: Wi-Fi a 'Positive Disruption'

[via 802.11 Planet]

"Not since the earliest days of the Internet have we seen a technology capable of creating such positive disruption and change and we expect that 802.11 and its derivatives will only increase in importance over the next five years," said Jupiter Research Director Michael Gartenberg