Lars Pind

Economist.com in The future of journalism:

Matthew Hindman, a political scientist at Arizona State University, found that the top bloggers are more likely than top newspaper columnists to have gone to a top university, and far more likely to have an advanced degree, such as a doctorate.

This captures the power of grassroots journalism pretty well.

I’m a strong believer in the agile manifesto’s “Responding to change over following a plan”. So much so that I do everything I can to make it a lifestyle — and an ever more feasible one, thanks to mobile phones, the internet, laptops, and wifi.

So when we woke up in New York last Sunday, after staying our first night at a bum hotel on Broome, we went straight to a wifi-augmented cafe and hopped onto Craigslist, where we found an apartment that had just been posted moments earlier. We called up the guy, went to see it an hour later, and then we had a place for the week!

Subletting an apartment is so much nicer than staying at a hotel, and a much better deal for us, too, although this particular apartment just went up from $125/day to $160-175/day (the page quotes both prices).

We also looked for sublets on craigslist from home, but leasers faced with a choice between someone they’ve met in person and someone six timezones away with a funny name, tend to come out … not in our favor. Not to mention the cash-up-front problem.

So next time I’m in town, I’m definitely going to do the hotel-for-the-first-night and then go hunting for an apartment once we’re in town.

It’s striking how frequently strangers have stopped to comment on my laptop since I got a mac – it’s happened on airplanes, in cafes in Amsterdam, and just now at the Pickmeup in New York. They’ll ask about the switch, about how large the monitor is, or they’ll just come over to tell me that they, too, have a mac. It’s quite entertaining.

I’m having some mail problems with lars at pinds dot com currently, so email me at larspind at fastmail dot fm if you want to reach me, while I get this thing resolved.

UPDATE: Resolved.

My phone number through May 15 will be +1 (646) 257 9980.

Thomas just announced reboot7, and it is looking to be an exciting event indeed. I’m particularly looking forward to meeting and hearing Douglas Bowman and Doc Searls, whom I haven’t met before, and Cory Doctorow, who is always a treat. I’m looking forward to this.

I’m going to the US for a bit, and would like a local San Francisco number for my cell, so people can call me without paying an arm and a leg. (When I lived in the states and didn’t have a credit rating, AT&T wanted $800 in deposit to open up international calls on my cell!)

And I already have a quad band phone that I’m reasonably happy with, thank you very much, so I’m not interested in a new phone.

All I need is a SIM card with a local number, and a pay-as-you-go plan. Even better, if I can order it online and have it shipped to where I’m staying in New York, so it’s ready for when I arrive.

I thought T-Mobile had this (saw it in a store in NYC last I was there), but I can’t find phone-less pay-as-you-go services on their site.

Does anybody know of such a service?

Holy Macaroni, this Google Maps and Craigslist (works in Firefox but not Safari) combo is kicking ass. We’re about to go to San Francisco for a few weeks, and have been looking for a place to stay. preferably in the Mission where the good burritos are, and where we lived before.

Going through Craigslist figuring all this out is tedious. But with this brilliant web app, it’s a breeze. The dragging just works, and the little pins even have photos and contact info right there.

I’m blown away by the possibilities. (Thanks to Peter Merholz for the pointer.)

My last post reminded me of a conversation with Christian Dalager last week, which took place when he was listening to this talk:

What is it that IT Conversations does for us? I read Clay Shirky’s blog, and I can read faster than he can talk, so you’d think I’d get more information transfer from reading. So why do I get so much out of listening to the talk?

There’s a couple of obvious reasons, one being that voice communicates many things other than the words themselves, most notably emphasis, and some more intangible cues about the person talking. Another being that the person’s getting paid what’s probably a significant sum of money to talk, so is likely to prepare well and do a good job.

But there’s a more profound thing going on as well.

When you’re writing for the web, you can take your message and its packaging as a given, and let your audience self-select around that: Those who care about what your topic, who know enough of the background to understand what you’re talking about, and who understand your references will get it, and those who don’t will just not show up.

When you’re giving a talk to an audience that is physically present, you have to take the audience as a given, and design the talk from that, making an effort to get everybody on board. You have to start from the beginning. You have to tell us why it’s important. You have to explain more of the background. You have to focus on the big picture and avoid the nerdy details.

That’s why, as a listener, there’s a greater chance that you’ll find inspiration on IT Conversations than in a blog post.

Just finished listening to Clay Shirky’s Ontology is Overrated talk from Etech, and it’s great stuff. I’ve been a fan of Clay’s writing for 6 years now, and it’s good to hear him talk for the first time. Not surprisingly, he does that very well, and I love his point about library categorization schemes optimizing for shelf space. What characterizes a great talk is that it’s so obvious once it’s laid out for you.

Now here’s someone who gets what open source is all about:

ad_proc -private ad_try {code args} {

  @author rhs@mit.edu
  @creation-date 2000-09-09

  Executes $code, catches any exceptions thrown by ad_raise and runs
  any matching exception handlers.

<strong><em>  If you use this I will kill you.</em></strong>
} {
   ...
}

My emphasis. From exception-procs.tcl in OpenACS :)

Greenpeace.org is in the air, the culmination of 6 months of hard work, challenges, and plenty of fun along the way. It’s a pleasure to work with such smart and motivated people.

It’s time for me now to (tie up some loose ends and) take a nice long break.

Harboe has out-innovated its competitors: They offer the first legally available Cola in 2 liter bottles. Their competitors must be frightened!

Seriously, though, it does seem like the capacity of innovation in the Danish soft drink and beer market (which, unlike in the US, consists of the same companies) is limited to the packaging. The contents are never-changing and dull as ever.

The main product innovation coming out of beer giant Carlsberg in the past 10 years or so has been a constant stream of new packaging – plastic, cans, new glass bottles, new cases – and each time, they’ve given the price a notch up. Their latest innovation: Reverting to the old bottles!

What are they thinking, over there in the Carlsberg silo? Back in the late 19th century they were at the forefront of brewing technology. Today you can get the same beer in 33 cl old design glass bottles, 33 cl new design bottles, 33 cl cans, 38 cl plastic bottles, 50 cl cans, and 50 cl bottles, in six-packs, 24-bottle cases and 30-bottle cases. Seriously, do they think that’s going to make the product more interesting?

It probably does sell more for a short while, after which they have to come out with new packaging, while still keeping the old ones on the market. Doesn’t sound like a winning strategy to me. It just gradually increases the costs of managing it all, but I doubt that it has any permanent effect on profits. (Particularly frightening for Carlsberg must be Heineken with only one container on the Danish market, a bottle that’s way cooler than any of Carlsberg’s.)

As an aside, I was in the supermarket the other day, and wanted to buy beers. But I just couldn’t get myself to get Carlsberg – the thought of it made me feel like an alcoholic. The only reason to buy Carlsberg is its low price/alcohol ratio. Not that that isn’t a good enough reason.

While listening to Larry Lessig’s 1 hour 36 minute long podcast, my iPod Shuffle starting cracking up and being really slow to respond to pausing, forward, rewind, etc. It wasn’t the track, because it was fine when played on my computer. I thought I’d done something to break it, but it turned out that when I changed to another track, it was fine.

So my conclusion is that the software in the Shuffle has a bug for very long tracks. Not too terrible, but I do hope they’ll fix it in an update.

Time-shifting is very cool, a great way to enjoy the medium, the wave of the future, etc.

But isn’t it ironic, how the broadcasters have most of these programs as individual pieces on tapes or hard drives and spend considerable effort putting them together so they fit together snugly one after another in the broadcast, only to have us spend considerable effort tearing them apart into separate programs on a hard drive afterwards.

That’s going to have to change.

As an aside, Larry Lessig mentions Japan, where people have 100Mbit/s ADSL connections for USD40/month (1:13:37). That kind of speed increase (I’m on a 512Kbit/s connection) opens up a whole new class of possibilities.

I just realized that I haven’t had two consecutive weeks in the country since New Year’s, and that after these two weeks, it’ll be end of May/beginning of June before it happens again. Phew!

But then we’re going to have a baby and I’ll probably stay around for a bit :)

Inspired by the Ruby construct where you can embed something like

#{"<em>some string</em>" if <em>condition</em>}

inside a string, I have started using this syntax in Tcl from time to time:

DELETE FROM gcms_rels
WHERE revision_id = :revision_id
[if { [exists_and_not_null rel_options] } { 
    set x "AND related_item_id IN ([join $rel_options ", "])" 
}]

It is pretty much exactly the same thing as in Ruby, and also very similar to the use of ad_decode that we’ve been using in OpenACS for a while. But it is superior to ad_decode because it is more readable, and because you can have elseif, switch, or any other construct in there.

The only ugly thing is the dummy set x part, which is required because, unlike Ruby, Tcl will interpret "foo" on a line by itself as the name of a command to be executed. I also tried using return, but that blows out of the entire proc, so isn’t usable.

If you have a cleaner solution, I would like to hear about it.