Saturday, March 27, 2004Wallop at the PC Forum. We presented Wallop at the PC Forum this week. Overall response was very positive, but I kept getting the same two questions: 1) when will this be public and 2) how are you going to make money off of it. It really highlighted for me the difference between prototyping as a research project in a large company and product incubation in a small start up company. It's amazing when you think about it the luxury we have, to implement these prototypes, be fairly experimental, and know it is someone else's job to worry about how it will make money. My answers to folks was a) we are in the process of a limited deployment to explore how it will be used by real friendships groups, and b) it's up to the product groups in Microsoft to decide which pieces, or even the whole project, to integrate into their products. Our job is to make sure that when the product groups are ready to pick up on new technologies we have already explored the space.
By shellyf@microsoft.com. [Shelly's Weblog] I tried ot leave a comment at this blog - but it was broken. Here's what I was going to say......
Hey! Sorry I missed your presentation at Etech - but we were scheduled against each other! Lord knows why they would they have the FOAF session at the same time as yours - oh well.
Looking forward to seeing Wallop on Monday.
Coolio - this service has really been moving forward. I wonder how they do the aggregation?
Fotolog.net aggregates an impressive number of moblogs maintained all over the world. With every 15 seconds the Most Recently Updated Fotologs.
Browse by name, country, U.S. state , random fotolog, groupfotologs or create your own free fotolog.
todays statistics
318,484 Fotologgers
6,431,651 Photos
36,488 Photos Today
Monthly Local Meetups are announced with other local Fotologgers.
1,144,000 people signed up for Meetups at local cafes (and other places) in 612 cities across 51 countries about 4,190 topics.
Meetups happen in hundreds of cities around the world. At local cafés, bars, bowling alleys, donut shops, dog-friendly spots, videogame displays, and lots of other places. People vote on where their local Meetup will take place.[Smart Mobs]
Jon Udell on Socialtext
John Udell paints the wide landscape of enterprise social software in his InfoWorld column. You may recall that Jon wrote the book on practical internet groupware some time ago and has an in-depth understanding of the promise of lightweight web-native collaboration.
We are social animals for whom networked software is creating a new kind of habitat. Social software can be defined as whatever supports our actual human interaction as we colonize the virtual realm. The category includes familiar things such as groupware and knowledge management, and extends to the new breed of relationship power tools that have brought the venture capitalists out of hibernation...
Computer-mediated communication is the lifeblood of social software. When we use e-mail, instant messaging, Weblogs, and wikis, we’re potentially free to interact with anyone, anywhere, anytime. But there’s a trade off. Our social protocols map poorly to TCP/IP. Whether the goal is to help individuals create and share knowledge or to enrich the relationship networks that support sales, collaboration, and recruiting, the various kinds of enterprise social software aim to restore some of the context that’s lost when we move our interaction into the virtual realm.
Restoring, or constructing social context is especially important because the vast majority of knowledge work involves remote collaboration, an accelerating trend. Jon also takes on the issues of transparency which enables group memory:
In networked environments, everything we do can be monitored. Absent the natural cues that establish social context — it’s hard to see groups form at the water cooler or hear voices in the hallway through e-mail or IM — social software systems ask us to strike a bargain. If individuals agree to work transparently, they (and their employers) can know more, do more, and sell more...Of course, we humans don’t always need to discover new collaborators. We’re already members of teams. Within those teams, there isn’t a one-size-fits-all social protocol. Outspoken individuals author the blogs popping up on corporate intranets. But other team members may prefer to contribute to a wiki, which is a collaborative space for Web writing. Ross Mayfield is CEO of Socialtext, a company whose hosted workspaces support both modes. “A blog enables people to express their identity,” he says, “while a wiki page de-emphasizes the individual and emphasizes the collective understanding of the group.”
The same person may find both modes useful in different ways. Adam Hertz, VP of technology strategy at Ofoto (a division of Kodak), uses Socialtext to coordinate his development team. During a period when he was traveling a lot, he says he started an internal blog to keep his team updated on his outside activities. It was helpful, but was unnecessary after he rejoined the team.
Whatever the mode of communication, the primary goal, Hertz says, is to create group memory....
The article describes how enterprise social networking services can help in sales intelligence and identifying collaborators. These are tools that take more explicit approaches to building relationships, where connection comes before content. They raise different privacy and transparency issues than tools that encourage people to opt-in to conversations and participation in different ways.
The Socialtext approach is similar to how people network using weblogs in public. There, content comes before connection. Social context supports the decision to connnect and the group memories of conversation enable greater trust.[Ross Mayfield's Weblog]
I myself use four SocialText Wikis now - on a daily basis. Anybody use more?
I call it Digital Lifestyle Aggregation.
Watch this be the new feeding frenzy of 2004. Everybody and their mother will be into this.
But it's not a battle between the TV, cell phone and PC. Real people want them all workign together. No more battles OK - just sweet life and harmony. Om.
Christian Lindholm, director of multimedia applications over at Nokia has some interesting thoughts on the whole concept of convergence and the role mobile phones will play in the evolution of a highly digital and mobile lifestyle. Lets call it convergence 2.0. He found evidence of it at the CeBit.
Integrating voice, image, video and text is becoming possible in an uncompromised form factor known to the broad audience as a mobile phone.
Why because he thinks with tiny flash memory cards, MP3 players, cameras and videocams all being bundled into phones, the only thing missing is the high speed connection and even that is around the corner via 3G technologies.
The integration of Wi-Fi totally changes to nature of a laptop. Laptops are now powerful enough and small enough to really become communications tools
Yeah we knew that and don’t need a phone guy to tell us about these things.
The Home server, rebranded as an entertainment center is the future of the Home PC. Intel was showing off the “Kessler” concept which integrated PC, Wi-Fi, RAID and lots of other technology I still do not understand, but it left a permanent impression on me.
This is convergence 2.0, only I personally think the PC guys are still thinking of it as a PC. Intel and Microsoft despite their billions, and their allies like morons at HP, Dell and Gateway are going to fall flat on their face. I would have loved to hear Christian’s comments on the User Interfaces in phones and user Interfaces in the PC world, and what impact do they have in adoption and usage.
[Om Malik on Broadband]This sounds like a conference for me! Real broadband. I can't wait till they call the crap they sell us today 'mid-band'.
100 megabit DSL at Fast Net Futures.
Next week I will be at the second Fast Net Futures conference in Santa Clara, California. It is organized by Dave Burstein who publishes e-letter, DSLPrime. I will be speaking at>VDSL. I will also be attending the VoN 2004 which is being held simultaneously at the same location.
Deploying in Japan within months, and suitable for 15 million U.S. apartments, as well as many communities with fiber to the neighborhood. Reach today is up to 1,000 feet at those speeds, but Stanford Professor John Cioffi’s keynote (Tuesday 5:45) will predict 100 Mbps on copper will be 3,000 feet in the next few years. That allows coverage of 60% of the U.S. population from existing telco plant. VDSL for the final loop is a third the cost of fiber in existing buildings, reusing existing copper and less than $150 of equipment.
Also looking forward to:
Movies on the Net: Jonathan Taplin of Intertainer, the early leader in movies over the net, promises to break news on Thursday coming from Intertainer’s antitrust suit against the studios (“The Five Families”). Very colorful guy, moved from Bob Dylan’s road manager to Producer of Scorsese’s “Mean Streets” to Merrill Lynch Hollywood to Intertainer.
Indy Fiber: Paul Morris, exec director of UTOPIA, the extraordinary project to bring fiber to hundreds of thousands in Utah, (Wednesday 8:45) is complemented by CTOs of Alcatel (Tuesday 8:45, delivering fiber to Pac Bell) and AFC (Wednesday 10:00, supplier of Verizon’s million home plan) and other experts.
[Om Malik on Broadband]Here's some great insights into how Laszlo is optimizing their platform. You won't find these techniques in Macromedia's copycat.
One feature of the recent LPS 2.0 release is the KRANK feature, for optimizing application startup performance.
Normally, when a Laszlo application launches within a browser, it runs initialization code that creates view and logic objects, binds them to dataset data,and attaches constraints. Some of this initialization depends upon the context of the application launch (query parameters, and the contents of runtime data and media requests). Much of it is the same each time the application is launched.
KRANK launches the application once on the developer's machine, stops the launch process after launch-context-insensitive application state has been created, and snapshots the state of the application at this point. It then uses this state information to create a new executable that reproduces the same application state, but with instructions that optimally create the same memory structures that the original application created by running general-purpose code. This produces an application that is typically larger than the original application, but reaches the point of first user interaction up to six to eight times quicker, because fewer instructions are executed.
This feature is similar to the Windows hibernate feature, where the operating system saves its memory state to disk before turning off, so that it can resume with the same state. It's different in that KRANK snapshots a reusable state that can be run on a different machine (in that respect it's more similar to operating system work on process migration), and that can be restarted multiple times and reapplied to different contexts (in that respect it's similar to continuations).
It's also similar to the image snapshot features of many Smalltalk and Lisp environments (including emacs). Like KRANK, this feature creates a memory image that can be relaunched many times, often within different operating systems. Unlike image snapshot, the KRANK feature is implemented mainly by a computation in process outside the application itself. This eliminates the overhead of the compiler and development environment within the snapshot --- this is important for a client-side web application --- or for tree-shaking techniques to separate the application from the development environment that embeds it.
There are other techniques for optimizing Laszlo applications (and rich internet applications in general). For example, you can toggle whether media and data sets are baked into the application (for a smaller server transaction count and a faster startup experience over broadband) or requested when the application initializes or later (for a smaller initial download size, and a faster dailup startup experience). You can also use deferred instantiation --- a technique we added about two years ago, during the initial implementation of the Behr application --- to declaratively specify that objects should be created in the background, either on a per-instance or a per-class basis.
The nice thing about these techniques is that they are minimally intrusive into the source code. (In fact, KRANK is not intrusive at all, since it's more like a compiler switch such as -o in traditional compilers.) They decorate the hierarchical and functional layout of the source code, rather than requiring it to be rearranged. This is handy for a stairstep development approach, with alternating functionality sprints and performance sprints. It also makes it easier to deploy the same application to both broadband and dialup clients, separately optimized for each.
| New issue of JOHO |
I just published the latest - and possibly last? - issue of my newsletter:
|
The fate of JOHO: Should we carry on? |
Disclaimer: I was one of those who got a ticket to the social software symposium. I'm staying at the hotel and helping out with the festivities (translation: I get to help make sure everyone gets on the bus on time and I get to try to keep Marc Canter out of jail. The Orkut variety, that is. Heh).
[Scobleizer: Microsoft Geek Blogger]
Robert reminds me of some good friends at Apple that helped MacroMind out in the early days - back when Apple had HyperCard and QuickTime and seemed to 'forget to invite us' to a number of key events. These folks would literally sneak me in the side door, we'd set up our color animation tool - working on a funky Mac II - and steal the show.
Scoble makes sure I'm there - just to keep everyone honest. He got me into the Microsoft PDC event and now this. But I've been told to be nice, not to heckle or make the researchers lives hell. Afterall - they have all the budget and resources they want, no need to ship anything and can just sit there and theorize out everything.
I just wonder if anybody will have anything specific to suggest to fix the problems they love to talk about.
"Cause the thing about it is......"
Having Scott Heiferman, Adrian Scott, Amy Jo Kim and Ward Cunningham there will be great. At least somebody will understand it - when I ask "get specific - for us folk who make products".
Everyone comes to their own conclusions on their own time.
Scoible has come to his own - about how social networking, when combined with search and a database of blogs (like Technorati) can then provide an interesting new kind of search experience.
Right on! As I've been saying for a while, social networking does not stand on it's own as some separate kind of product or market. It's like multimedia and the web - a raising of the ante. Across all products and markets.
A new kind of people search is needed.
I've been thinking a bit about social software (Orkut, Linked In, Friendster, and the like) and here's what they are going to be interesting for: finding people.
Now, if you already know the name of someone, Google already works remarkably well, particularly if the person you are looking for has a weblog. Search Google for "Robert" for instance, and you'll find I'm the second Robert. Search for "Dori" and you'll find Smith. Search for "Tim", and you'll find Bray. Search for "Jeffrey" and you'll find Zeldman. Search for Dave and you'll find Winer. Search for Mark, and you'll find Pilgrim. Search for "Doc" and you'll find Searls. Newer webloggers, like my brother Alex, require you to search for both his first and last name, but that's still doable. Yeah, it isn't perfect (if you don't have a blog, or you didn't use your first or last name on it, it's very hard to find you). That's where services like Orkut, Friendster, LinkedIn, etc will help.
One problem: I don't want to search for blogs by name only. I want to look for groups of bloggers. Someone, who asked not to be named on this blog, yesterday asked me "what if I want to find all Egyptian bloggers?" How about "all gay Egyptian bloggers?" Or, how about "webloggers who play bridge?" Or, "webloggers who are also executives?" Or, "webloggers who ski?"
You can see this can go on and on.
Lately I've been doing searches on Technorati. If you search on a term like, say, "quilting" there, you'll find all blogs that mention quilting, and they are in some sort of order based on how many inbound links they have and they bias toward webloggers who mentioned quilting in the past few days.
That's cool, but I want a new kind of search engine that combines the full-text approach that Feedster.com uses with the inbound-link analysis that Technorati does. So, I want to see webloggers who are most likely to talk about quilting in the future.
Does that make sense? Past behavior as a predictor of future behavior. In other words, if I wrote about digital photography several times in the past (I have) maybe I'll write about it in the future (I will).
Anyone working on a search engine like this? Now, imagine if you combined it with Orkut. Now, in the result set, match results up with Orkut profiles. Put pictures next to the weblogs that return. Wow. That might really be disruptive.
Such a system might get millions of people to give up Outlook's contact system.
Hmm, Google owns all the pieces to do this. So does MSN. So does Yahoo. So does Friendster and Technorati, or DayPop and Technorati, or PubSub and Technorati (if they work together). The race is on!
[Scobleizer: Microsoft Geek Blogger]As some of us well know, the Garteful Dead helped craft landscape when it comes to amateur recordings of live performances. For years - the Dead allowed anybody to tape their shows.
Rumor has it that there's a database of all those shows being uploaded and stored at the Internet Archives. There's no official word, but lately these shows have been showing up in the RSS feed of all the recent Archive postings, which is public.
So several times now I've clicked on these links, only to be told that music is not available.
Bummer.
Grateful Dead: 1979-12-26. Live at Oakland Auditorium Arena [Internet Archive]
So here's an image to think about - a social net built up around this database of live recordings, with the ability to leave notes and annotate shows and even songs themselves.... all within the context of a private/public page kind of social network.
"this is when I learned to do the whirling dervish"
"looking for the angel that made me a miracle and gave me two tickets out in the parking lot, I've been looking for you for 24 years."
"oh man, I remember this show - it rocked....."
"this is the show where my wife and I met"
My friends (real friends) and I have been waiting YEARS for these kind of collaborative social network apps. The DeadNet is already happening, so they'd eat this up. But think of the common experiences rug that would be weaved with this sort fo social net.
Many people's lives changed while at Dead concerts or on the scene - and being able to have access to a definitive DeadBase - which would not just be a list of the songs played, but recordings of those shows, and sync that to a message board/social net. Oh man, oh man.
- - -