Sitting in the Red Carpet Club in Concourse C in O'Hare, with Jeffrey climbing all over my back. We're midway to Orlando, but not by the original route, which was through Los Angeles from San Jose (one of my least favorite airports). The first leg was delayed indefinitely due to weather in L.A. So we got on a flight to Chicago. If all goes well, we'll only get to Orlando two hours later than originally planned. If we can't get on the flight (we have boarding passes, but were added last and it's overbooked), we won't get in until 2AM.

Jeffrey is climbing all over everything. He tortured the poor woman in the seat in front of us, kicking and pushing with his feet all the way from SJC to ORD. Right now he's standing on his head in a plush chair with his feet against a ficus tree.

Anyway, there may be some clarity coming along on the Google/Deja front. I'm getting more emails from unhappy folks who depended on Deja's service, but I'm not inclined to run anything more until I know the answers to the questions below.

Time for Peerage Journalism

I'm getting emails that I can't square up on this Deja/Google thing. Questions:

1. What exactly did Google buy?
2. Why did they buy whatever they bought?
3. If they only bought the data and not the whole Deja system, how far are they going to go toward restoring any, all, or more than what users lost when the Deja system died?

I don't have time to contact the Google people I know, but somebody needs to come forward from Google and/or what's left of Deja and explain this whole thing. Or just jump into the conversation here where a bunch of journals (and isn't that what blogs are, after all?) together amount to Journalism 2.0.

Off to Orlando.

Hackage

Nice to see Wes digging Don, another planet hacker. It was also good to meet Wes at the P2P thing, but I don't think I made any sense in our conversation. I have excuses (like, extreme indigestion I wasn't talking about and which later kept my butt clamped to the plumbing at the Sutter-Stockton Garage for two hours before I could drive home), but still.

And Wes likes the Ricochet modem: Ricochet 128k is so cool. Unlike most other Net connections that advertise maximum bandwidth, the Ricochet data rate is actually over 400kbps so it's not unusual to get 128k. The external modem is actually not that big, either. So now maybe I want one. Might be a cool way to get online when I'm in a major metro. Might not help in Santa Barbara (where we're moving shortly), though.

Blog slowdown

Dean says he's going to blog a little less for a while. Hope not too much less. Dean's blog is more personal than most — kind of a public diary. I like that.

Actually, I'll be blogging less too. I leave in eight hours for EC4M, the GE Global eXchange event in Orlando. I'll be back home late Thursday. Joyce and Jeffrey are going to, and we'll take in a bit of the Disney thing while we're there.

The continuing end of business as usual

Nice ongoing thread on the Allchin Matter at Tim's blog. Nearby there's a good conversation between Tim and Larry Lessig. The bottom line from Larry:

...our problem is that lawyers have taught us that there is only one kind of economic market for innovation out there and it is this kind of isolated inventor who comes up with an idea and then needs to be protected. That is a good picture of maybe what pharmaceutical industry does. It's a bad picture of what goes on, for example, in the context of software development, in particular. In the context of software development, where you have sequential and complementary developments, patents create an extraordinarily damaging influence on innovation and on the process of developing and bringing new ideas to market. So the particular mistake that lawyers have compounded is the unwillingness to discriminate among different kinds of innovation.

We really need to think quite pragmatically about whether intellectual property is helping or hurting, and if you can't show it's going to help, then there is no reason to issue this government-backed monopoly.

Too few cells in too many scattered places don't add up to a whole animal

Says here (by way of Tomalak) that Metricom is on the ropes. Looks like another one of those services that only work at ubiquity, which means it relies on universal free infrastructure (the Net), massive industrial heft (cellular telephony), or both. Metricom sold best as a kind of cell service for laptops. But it never had the coverage, the bandwidth or the reliability to make it work.

I started using Metricom a number of years ago, when they first rolled out. It was the only way to get an always-up connection to the Net at the time, which was before DSL and cable modems, and when ISDN was prohibitively expensive. The advertised throughput was 33kbps or so, but at its best it rarely beat 14.4. Much of the time it was slower than that. Coverage wasn't an issue where I lived, which was high on a hill with a view to the whole Bay Area. No less than 90 "cells" could see my modem, which sat in my office window. After I moved to another hill with a view, it got worse. In fact it was so much worse that it was borderline useless. Seems my modem established relations with a cell situated on a house about 150 feet away, and throughput with house-based cells was worse than with ones mounted on phone poles. Worse, there was a lot of activity in our local area, which degraded performance. But it there was no way to make it hunt for a better connection. I finally bagged the service after I got DSL to the house.

Now they promise 128kbps of throughput. If this were constantly the case it would about equal my IDSL (nominally 144kbps). Maybe this time it's faster. But I doubt it. Here's what an InteractiveWeek story told just a few days ago:

We tested the Ricochet on the Wireless WebConnect network throughout the New York metropolitan region. Connections were subject to coverage availability, reminiscent of what cellular coverage was 25 years ago: Coverage was pretty solid in Manhattan and quite good within a 5-mile radius of the city. Throughput was variable during the two-week period that we tested; large file downloads via HTTP ranged from 7 Kbps to 105 Kbps, with a median of 71 Kbps. FTP transfers were somewhat faster, ranging from 93.4 Kbps to 205 Kbps, with a median of 112.7 Kbps. Unfortunately, we were occasionally booted off the network and couldn't reconnect for several hours.

The key point is that one about cellular coverage 25 years ago. The price is also reminiscent of that period, when coverage was expensive.

I think the main problem is that it's gotta be everywhere, or darn close. But right now it's not.

But hey: I wish them luck. Their customers too.

discuss