Web 2.0 2005 Day 2, Morning Session 2

p. Web 2.0 Conference 2005, Day Two, Second Morning Session. h2. Dave Sifry, Technorati * 18.9 million weblogs tracked * 70k weblogs created per day * as a percentage of posting volume, spam blogs are low * blog finder, lets people tag blogs. tags as classification for content-stream (not a specific page of content, a bookmark) * using tags of "authoritative blogs" now to filter search results h2. Tom Barton / CEO Rackable Systems * built Google's first 10,000 servers * reviews how big, how much it costs to run a largescale datacenter * that they are better at building efficient systems than competitors like IBM, Sun, Dell * deploy servers a rack at a time, ship cabinets with servers in place * price/performance/watt will drive data center compute and storage decisions * post-bubble overcapacity in the data center will be gone in two years h2. Hossein Eslambolchi / AT&T * forces of control vs forces of freedom * passive reception of infomration and entertainment (old) vs active participation in reception of information (new) * moving away from limited content provisions to 100s of 1000s of channels * control your own content * need to change infrastructure to support that * that there isn't a global standard in mobile wireless market [epc: HUH? just because US standardized on xDMA vs GSM doesn't mean there isn't a standard] * moving from proprietary software (at AT&T) to open software. That open software has lower defect rate than private and proprietary software * opening up at&t network os that customers can use web services to do provisioning, network configuration, etc. "turning the network inside out" * from licensed spectrum to unlicensed spectrum * from regulated access to unregulated access * that the true definition of broadband should be 100Mbs, not 1-2Mbs, * traditonal phone number portability - ip address portability, the death of locality * i'm not writing the rest of his points down, i've heard them before. h2. Tim O'Reilly: discussion of open vs. closed systems p. bunch-o-people, didn't catch all of th enames * TO: nature of open vs. closed. what are you going to "own". what is it for each of you * for amazon: we give out huge amounts of data through our web APIs. We "own" the process of letting folks find things and buy them, the customer service and fulfillment. Most of the code we own is not in our web site. * TO: that Amazon created the ASIN and its own namespace, is that evil? * amazon guy: don't know if it's evil, many people use it * yahoo: yahoo owns user's trust * TO: can users take all of their data out of Yahoo!? Clearly once a user puts data in yahoo, yahoo owns the stickiness of the data * yahoo: users can remove the data, but it loses value as it's removed from the collection * yahoo: that there's certain data we can't open up due to contractual obligations * h&q/skype director guy: that people will pay you for the value added services you provide on top of the data that they provide * for VOIP it's a matter of quality so they won't open it up to everyone to protect the quality * TO: would skype interoperate with a Yahoo voice ? * Skype guy: no longer my call but certainly it would be desirable * Cable guy: dependent on third parties for content * TO on developer networks: how do you draw the line between what happens at Yahoo and what happens somewhere else? The sense that the yahoo apis are more about putting data onto yahoo, where google API is outbound, using google data for other purposes outside of Yahoo. * is there an openness to external dna in companies? and is this why many exit strategies are acquisition? * definitely a sense of audition for jobs by doing amazing stuff h2. Jeremy Allaire / debuting BrightCove * future of Video over IP, long tail of video production and content distribution * worldwide video and tv business is USD$340+ billion dollar business * monetization and distribution services * "flash 8 video is the best format for video on the internet today" h2. Jason Fried / 37 Signals * The idea of less, less as a competitive advantage * they have competitor, they need to do more * that this isn't the best way of doing things anymore * to think of "one downing" people, doing less than your competitors h3. Five things you need less of # Less Money. All other people's money is going to do is place you in debt. Debt is not going to make your decisions any better. Money can't buy you time, or passion, and you need a boatload of passion to build products. # Less People. You only need three people to get started building online. If your product needs more than three people it may be too complex. Don't scale up to meet your product set, scale down the product set to meet the people you have. One Designer, One Programmer, One Sweeper with business knowledge and common sense. # Less time. Most time you spend is wasted time, with meetings and dealing with many people. Less time forces you to spend your _less time_ on better things. When you have too much time on your hands you waste it. # Less abstractions. Do less stuff that is not real. Focus on building real things, seeing what happens, and learning as you go. Do not do a functional specifications document. They are purely political, _yes_ documents. Instead build the product and start from the interface first. Build the UI first then wrap that with technology. Have the interface screens be the functional specification. FPFS lead to illusions of agreement. # Less software. Build software that does a lot less. Results in less features, less documentation, less support, less complexity. Less things to explain, less things for people to understand and comprehend. There are a million simple problems to solve before solving the complicated ones. * Want more constraints. Less time, less people, less money forces you to focus and get things done. * Update: Jason's talk online: "Less as a competitive advantage: My 10 minutes at Web 2.0":http://37signals.com/svn/archives2/less_as_a_competitive_advantage_my_10_minutes_at_web_20.php

Posted in Conference Notes