 |
|||||||||||||||||||||||||||||||||||||||||||||||||||
|
Dinner at a pizzeria in Mt View Ten people around the table, two from Boston, one from Toronto, one from Minneapolis, me from Ann Arbor, and a few more folks who I couldn't place geographically. Mostly outsiders to the current crop of everyone goes to everyone else's party Web 2.0 stuff. It was fun. Relatively lots of "how do you keep up with all this stuff" and less "back when I was young this is how we did this with Gopher" or swapping notes on what the latest Valley party was. There's this weird dynamic where lots of people can say "Do you know (random web 2.0 company) ?" and people will say yeah, that's great, but in a lot of cases people don't attach a name to the company. The relatively fewer memorable companies have not just a great idea and great execution but also some memorable people with a real presence on the net. Touring the Computer History Museum I saw a bunch of old gear which brought back a bunch of memories. Some highlights: a circular slide rule like the one my grandfather used; a few abacuses just like the ones I own; a Vax 11/750 just like umix; a very good collection of Ataris, PETs, Osbornes (but no Kaypro?), and other random 8-bit machines; and some very entertaining conversations with a few of the senior docents who talked about building ham equipment from WWII electronics surplus and early APL systems. The 11/750 was the highlight - it looked just like umix which I spent many an hour tending. Aha! Ward Cunningham led a wonderful session on Aha! moments that I have notes (w/Tantek) on - will post a link to that when they are online - subethaedit aha! notes here. There is at least one other person in this world who used LambdaMOO's "gopher slates". Starting to get weary Long day, and a lot of unfamiliar faces. My visual systems are on overload and I have a bit of a headache. The name tags are very hard to read from a distance. Public libraries session at mashup camp Good session, small (about half a dozen of us). I asked everyone in the room what their local public library was an how they used it. We tidied up the AADL hot books display with some better HTML stylings so that will go online next time I'm coding. There's a tremendously fragmented market, so innovations don't move rapidly across all of library-space. I showed off a few Greasemonkey scripts that inserted AADL data into Amazon, Google Books, IMDB. At lunch, more discussions, very brief notes here. Discussion about a friends of the library program in NYC which takes book donations and sells them for $1 ea on Saturdays. Wouldn't it be cool if they scanned in ISBNs for incoming donations so you could see what they had coming in the door? Don't even try to keep inventory, just provide a feed to assist people in hunting. Also at lunch, addressing the problem of alerting you to upcoming books due on hold. Tim Bishop said he wrote a script for the Berkeley library to automate that for his kid; I mentioned Library Elf as a possible solution (don't know if they have support for Berkeley). Open question: would this all be a different world if there was a single library card with borrowing priveleges across the whole metro area? The open space sessions I'm going to - subject to change 10:30: public libraries, Superpatron 1pm: H,.widgets and containers 2:30p:? 4p G: Tools for Ahas, Ward Cunningham Intros The Vacuum Group (that's me), 1 person consulting firm, 200 person advisory group. Public libraries, community and neighborhood and local information. MSN Messenger, presence MSN, Virtual Earth Ning, platform for social software (eh) A9, OpenSearch, mapping API, search API Flyspy, airline search engine (very cool check it out) Amazon eBay Strikeiron, aggregator, microbilling Washington Post EVDB, eventful, events, venues Golfnow, tee times Technorati Six Apart, blogs, open id identity Zend php Civicspace, CivicCRM Yahoo, Eleanor K, Developer API (talk to her) At Mashup Camp, opening session Propose a session, put it into a grid, 30 second intros. Openspace. How many APIs to data? Right now about 170 counted. (I'm sure there are more....) How many should there be? I suggested that about 25 was a good target to shoot for - it's going to be horrific to write code to 170 different specs of varying qualities, and we'd all be better off with common programming interfaces and not inventing a new protocol every time we brought new data online. 300 people. Limited availability of power in building, don't daisy chain power strips. At Mashup Camp, breakfast Saw Susan Mernit. To follow up: mixed public/private library in San Jose, community journalism in Boston, importance of local information in classifieds and personals. Retired physician in Montana who found her in Wikipedia. Mashup camp, the night before Party at the hotel. Geeks, pizza, wine, demos. Mostly demos. I'm approaching each one with the following question: does this make me a better neighbor if I tell my neighbors about this? Does this make me a better father if I spend time working on this? A discussion with the memeorandum guy (sorry, still know more web site names than faces) about OPML and search. We discussed whether it might be possible to build some supergopher (that was the word I used; he had used gopher a long time ago but didn't do any development on it) that was based on OPML, searchable, and reflected local information about things in ways that Google doesn't. My use case was "restaurant reviews in my neighborhood"; I guess I could see that, it would be a short list but I could be thorough about it. Up to the demo room where I came in as someone was demoing a system for wholesaling spare inventory at golf courses. In the best case you are connected live to the golf course's online system; in the usual bad case you're faxing orders in. Neighbor test: no golf courses in Michigan in their inventory. Father test: I don't play golf. Best demo: flyspy, shows you 30 days worth of cheapest fares city to city. Alpha test site is MSP. Way faster than trying to do that same search by hand on Orbitz though I would bet that having seen it you could script the search one user at a time and get something reasonable. This wins the father test (fly to see Aunt Tish) and the neighbor test for some small narrow edge of neighbors who go to MSP. Now if it was a DTW based site that would really be something. Disappointing: edgeio. Um, if I get things right, you take blog postings that have been tagged with "listing" (just like Technorati, and we all know how that goes) and you mash in some metadata to turn it into a giant global classified ads system. With tags. And sliders for geography! But strange sliders that have jumps at "Ann Arbor", "Michigan" and no way to clump A2 and Ypsi. It sort of reminds me of what you'd do if you started with misc.forsale and added some more random things to it but kept the underlying data source as misc.forsale. |
Last modified: Tuesday, February 21, 2006 at 5:14 AM.
Tech resources |
||||||||||||||||||||||||||||||||||||||||||||||||||
