The CTO of Pandora was nice enough to point out Pandora is now open for everyone (this is why I love blogging). So I gave it a shot.

First off, unlike Scoble, I actually like the flash interface. Works well enough for me. Nothing to download, runs right from the browser.

The interface is good. Very simple, which is exactly the right way to do it. Let’s take a moment to appreciate an interface that could’ve had a million buttons but only has a few. Good.

The basic idea is: starting with an artist you like, Pandora will guide you to other music you might like. It uses a radio metaphor; you create a radio station by specifying an artist or a song, and music streams to you, from that artist and other related artists. You can’t forward or rewind songs or go back to previously played songs. It uses the music genome project to form its recommendations, which apparently is based on fairly deep analysis of the various musical qualities such as temp, style, etc. Pandora will tell you why it chose a particular song to play, quite a neat little feature.

My first attempt was N.E.R.D., a group I’m just getting into. The next immediate recommendation was excellent, but alas I lost it when I switched from the in brower to the pop-up interface. Is there a way to get history?

Interestingly, N.E.R.D. is not available on Pandora. More interestingly, this didn’t bother me. I already know them; that’s what I’m starting from. I’m much more interested in the new artists Pandora can lead me to. In fact, the most useful thing Pandora could do for me is lead me to new music that I don’t already have in my collection, and hence haven’t heard a million times on my mp3 player.

So I tried a few others:

Beatles led me to Billy Joel and Elton John, which I suppose are reasonable suggestions, but I definitely can’t listen to. For some reason I didn’t hit the “I don’t like it” button; I just moved on.

I tried Oasis next. Not sure where that took me, but I gave up on that too,

PM Dawn next. Not great, didn’t really capture what makes them different from the other groups in the same genre. Gave up.

Maxwell didn’t go anywhere good either.

Cheb Khaled and Cheb were not recognized as artists Pandora knows. That’s not good. I guess US music only? That’ll be a problem.

Coldplay was not available but took me to The Perishers, which I liked. Cool.

Ok, so for the most part it didn’t work out for me, but I still like the service. I know I’m a pain to please and have odd tastes, so I’m not really surprised it didn’t work great for me. Generally speaking I dislike a lot more than I like, so it’s hard to find a recommendation I would actually like. That first recommendation on N.E.R.D. was really good, so that could’ve been something.

Now to complaints and suggestions: where’s the button to export my “stations” into a playlist or RSS? Remember, all my data are belong to me.

How about some pre-made radio stations? I know this is based on my preferences, but I don’t want to have to always enter an artist to start things off. Give me a way to turn off my brain and just listen.

Hmm. Second take on the interface: give me a way to rewind. Let me listen to that song agin.

Overall, I quite liked Pandora. Not sure if I’ll end up using it or if it’ll fall by the wayside. It does require you to actively participate in choosing your music, which is a good thing, except that music is usually background to the other 19 tasks I got going on, and I’m not necessarily looking for a 20th.

Would I pay for it? Don’t know. Maybe. I have Yahoo Music Unlimited, but if I didn’t have that, this might be attractive.

I’ll continue to try it out. Hopefully as it learns about me it’ll magically choose great new songs and I won’t have to actively participate in choosing things, returning to my favored vegetative state.

Check out Cory Doctorow’s DRM Rant, via Tim Bray. This thing is excellent. Well worth the read.

I’m trying to buy a couple of webcams, and being the cheap kinda guy that I am, I’m looking for a deal.

So I look on fatwallet and find various deals that carry various degrees of pain. I find one that seems reasonable. All I need is a coupon, $20 off $100.

Now chances are I actually have the coupon somewhere. I get coupons mailed to me at my house, which I promptly “file” into places I will never find them again. I get coupons mailed to me at my various email addresses that I’ve given the various stores to keep my main addresses clean. I have to search those. But those are expired. Damn. Ah, found one. Damn, that’s a different store…

It turns out I could pay a couple of bucks on ebay and buy a coupon. Which would be worth it financially, but I gotta deal with ebay…

So now I’m tired and I give up on the whole thing. I can live without the webcams for a little longer.

Here’s an idea for the various merchants: why don’t you hold my coupon for me at my account on your online store so I don’t have to look for the silly thing? If you’re willing to give me a $20 break on $100, just store that info in my account and apply it automatically to my order. That would actually motivate me to buy from you.

The Scoble, musing about Pandora, writes

I really hate this “everything has to be a Web app” trend that we’re in the middle of. This service is totally crying out for a rich-client application.

Which I would agree with, except it turns out I don’t.

Some time ago when Google Earth first came out, I downloaded it and was entirely impressed. Really quite a nice app. Visually beautiful, well implemented, good stuff all around.

Browsing the blogs, I ran across a posting that told of the Google guy who ran the Google Earth effort actually apologizing that it was not a Web app. Huh, I thought, with such a beautiful app, you don’t have to apologize. It’s so very nice.

Since then I’ve played with Google Earth exactly once, when I was trying to figure out the topology of Yosemite. On the other hand, I use maps almost on a daily basis. Mainly Yahoo maps, because while not as beautiful as Google maps, it’s a bit more usable. Remembers my addresses and such.

So, to my surprise, it turns out my behavior suggests every app actually does need to be a Web app. While I find Earth a lovely application, the overhead of starting up an app, waiting for it to come up, getting what I need, and then shutting it off again is just too much. Put another way, if I can just open a new tab in the browser and get what I need in less time than it takes to start a standalone app, I’m very unlikely to use the standalone even if it’s nicer. The web app needs to be just good enough and I can forego the standalone app.

Which is surprising really. Even more reason to make this AJAX thing more commonplace.

Btw, somebody send me a Pandora invite so I can try this thing out. Please.

Stat that surprised me: I have as many comments as I have posts on my blog. I was expecting a far smaller rate of comments. Is this normal for blogs? Anyway, thanks for the comments, keep them coming.

I just noticed, all the Referrer linkspam I’m getting is going to my Case of the Missing Referrers post. How’s the for irony? I asked for referrers, I got plenty of them!

If whoever’s doing this did that intentionally, that’s pretty clever, good one. If not, quite a coincidence.

I’d been meaning to give Python a whirl for some time, and this weekend the perfect task showed itself. Something that would’ve involved 20 minutes of manually typing information into the computer, or 5 hours of programming to write a bot / screen scraper…

So, 5 hours later, I have my first real Python program. It’s basically a bot that hits a form and scrapes the results into CSV. Nothing complicated, but enough to exercise the basics of the programming language. As a data point, I speak Perl and Tcl (as well as Java and C++), and in 5 hours I feel I’m able to do useful things in Python. That’s a pretty good learning curve.

I quite like Python. My major hang-up before getting into it was the fascist insistence on formatting rules and indention. However, now that I’ve tried it a bit, I really don’t have a problem with it. Makes your code look nice and consistent.

The language itself feels clean. Cleaner than Perl, less strange than Tcl. Once you get used to the Python world view, it does a nice job of playing out the way you expect it to – if you were to guess what a command should look like, you’d typically guess correctly (at least I did).

I have access to a number of Python books, including the O’Reilly Cookbook, Nutshell, and Learning, and I have to say they’ve all been disappointing. Learning is very dry and doesn’t have enough examples. Nutshell is ok, but doesn’t live up to the Perl Nutshell. The Cookbook is the real disappointment; Perl Cookbook is incredibly useful, but I found the Python version didn’t contain anything I wanted to cook.

I also found it harder to find code examples for what I wanted to do, compared to Perl. I can typically find code to copy/paste within the first few results of a search; with Python I’d mostly get the man pages or somebody asking the same question I had.

Which is all a bit surprising, because I thought the Python guys really had their documentation ducks in a row.

Btw, have you noticed how newer languages learn from the old and actually pay attention to tutorials and documentation? The Ruby on Rails guys, for example, have done a great job of marketing their system and providing materials to make it easy to adopt.

One of the advantages of not being an expert in any language is that your barrier to switching becomes much lower. I think I’m actually going to switch to Python as the default for my basic scripting needs. I’m also playing with CherryPy, which looks like a good contender to take over php’s spot in the particular area I’m interested in these days, which is AJAX, and hence requires very little templating. I still intend to give Ruby On Rails a try, but for now the scaffolding is too much, well, scaffolding, to play with.

Great. First the site was down, now I’m getting this genius: host-207-248-240-118.block.alestra.net.mx (207.248.240.118) spamming me with linkspam via referer. He’s been persistently hitting the same URL only to throw his junk referrers in there. He’s got quite a lot of them too, many apparently online pharmacy link farms, such as this one: hpsstudent.com .

I tried blocking him via the apache Deny directive, which worked fine when I denied myself, but had no effect on this guy. Then I tried blocking him at the linux level with:

/sbin/route add -host 207.248.240.118 reject

and that seems to have worked.

I’ve noticed each time my site’s pagerank goes up I get a new set of attacks and junk to deal with. Joy.

Good post on Chris Anderson’s blog on Just enough piracy.

Any protection technology that is really difficult to crack is probably too cumbersome to be accepted by consumers.

zero-percent piracy is not only unattainable, it’s economically suboptimal.

Instead, efficient software and entertainment markets should exhibit just enough piracy to suggest that the industry has got the balance of control about right: not too loose and not too tight. That number is not zero percent

Microsoft had realized that some piracy is not only inevitable, but could actually be economically optimal. The reason is counterintuitive, but intriguing.

The usual price-setting method is to look at the entire potential market, from the many at the economic lower end to the few at the top (it looks like a pyramid), and set a price somewhere in between the top and bottom that will maximize total revenues. But if you cede the bottom to piracy, you can set a price between the top and the middle. The result: higher revenues per copy, and potentially higher revenues overall.

Today the estimated piracy rates are 33% for CDs and 15% for DVDs. The industries say that’s too high, but most anti-copying technologies they’ve brought in to lower that figure have proven unpopular. Would even tighter lock-downs help? Probably not. Maybe 15%-30% is simply the market saying that this is the optimal rate of piracy for those industries, and any effort to lower that significantly would either choke demand or push even more people to the dark side.

This is fascinating. Reminds of the studies that suggest the “natural” tax rate falls somewhere around 20% regardless of where in the world you go. If you try to set the rate higher than that, people will start cheating in mass numbers. Lower, and you’re not getting as much revenue as possible.

I was reading Ross’s blog and noticed his post on dialing into basecamp. He setup a dialup number so you could hear sessions at BarCamp. Excellent idea.

Which got me thinking, now that all conferences have Wifi and we all have Skype, why not do this for every conference? Setup a conf call on Skype, advertise the fact that you’ve got audio of conference X, say via a technorati tag ( “conf:liveaudio” ), and we can all mooch. I mean, join in. It’s easy enough to record, so now we have an audio recording of the conference anybody can listen to whenever they have time. Then we could podcast it, just to cover all the new buzzwords, and I could listen to it on the train.

I suppose this might not be appreciated by the conference organizers. It’s sort of like bootleg recordings of concerts, and there’s a healthy crowd devoted to that. Some bands / labels are ok with it, some are not. I’m sure some of the more enlightened conferences would be ok with this. O’Reilly camp, I’m looking at you. There you go, I’d actually start paying attention to conferences again.

And for the record, at least for me personally, it’d actually increase the chances of physical attendance at conferences if I can be convinced somebody worthwhile is speaking or a worthwhile topic is being discussed, and the way to do that is to let me listen to past presentations.

UPDATE: How cool is this? The good folks of Tech Camp Ireland have taken me up on the idea and are going to be skypecasting their conference. About 24 hours from conception to birth. Nice!

The nice folks over at SocialText have developed a WYSIWYG wiki front end called Wikiwyg, as reported by Joi Ito. You may recall we’ve discussed something like this in the past.

I gave it a try. The editor only has a few features, but it’s quite nice and very usable. Good stuff.

Seems many clients are not setting referrer. In fact most are not setting referrer. Is this a known thing?

Mark Baker kindly linked to my On RSS/Atom replacing SOAP post. Interestingly, he did it by linking to the Webservices tag instead of directly to the article.

That particlar tag was getting practically no traffic before Mark linked to it. Since he linked to it it’s gotten 105 hits. However, I only have 29 referrers from Mark’s blog, and 5 more from his del.icio.us entry. This is all according to awstats.

So what’s going on? Where are the remaining 71 hits coming from? I’m pretty certain they’re from Mark’s blog, but the referrers are not set.

Is this because of the blog reader software people are using? Last I looked (which was quite a while back) referrer was set by the vast majority of the browsers. Looks like it’s not anymore. Are blog readers bad referrer citizens?

This is neat: my article has been translated to Chinese. Here’s the English version, btw. And thanks to all the folks who sent the nice comments, it’s appreciated.

Some years back Alex and I were interviewed by a Japanese technical magazine. There’s a picture of the two of us by Seaport Village with the Coronado bridge in the background, surrounded by Japanese text, which I never have seen a translation of.

It’s always cool to have your words translated to a language you can’t even remotely decipher. I’ve always wanted to pick up a little Japanese and Chinese, but alas…

Unleashed, aka Danny The Dog, is not a very good movie. Avoid it.

The movie has Jet Li and is written by Luc Besson (the Profressioal), so I was expecting something decent. On the other hand, it had been getting bad word of mouth, so I had low expectations. The tipping point was the 7.3 IMDB score (what the heck are you people thinking?).

It’s an imaginative but entirely ridiculous story. Jet Li is his usual self. Kerry Condon is supposed to be 18, looks like she’s in her mid 30’s, and acts like a 6 year old. Bob Hoskins reprises his role from Who Framed Roger Rabit, acting the British gangster. He’s pretty good, but pales in comparison to Alan Ford’s Brick Top Polford (Snatch). The fight scenes are not bad. The rest is an utter waste of time. Avoid it.

Crash. Very good movie about cultures and stereotypes, set in LA.

This was surprisingly good. Not politically correct, actually dealing with interesting topics. A couple of corny bits, but overall very good.

Good performances from unexpected sources, with Sandra Bullock and George of the Jungle doing a nice job. Don Cheadle is good as always.

Go see this one, particularly if you live on either coast or in a culturaly mixed area.

Bad Guy, aka Nabbeun namja. I watched the entire movie. Shouldn’t have. I suffered through it hoping for decent ending, but nothing.

It’s not that I was offended. I just wasn’t that interested.

Avoid this one.

From Paul Graham via Phil Windley

The average office environment is to productivity what flames painted on the side of a car are to speed.

Here’s the basic idea: I want to designate certain documents as Web-backed, meaning they’ll be automatically uploaded to some Web space on a frequent basis. I’d like to be able to access Web-backed docs on my various computers and have the Web backing persist – ie. if I download and modify it on one computer, it should realize it’s Web backed and upload the latest version. When I move to another computer, it should grab the latest version in the background so the latest is always anywhere I am magically.

If you’ve used the Danger Sidekick, you’ve had a taste of the concept. The device automatically puts all your information on the Web, so you can get to your calendar, adressbook, notes, etc, equally well from the Web and from the device. This is probably its best feature.

The OPML editor seems to have a similar feature where your outlines are automatically Web-backed, or “upstreamed”, regardless of what you use to edit them.

The key is that I don’t want to actively “backup” or “upload” the files. That should all be automatic. I want to be able to right-click a file, have a “Web-backed” option in the menu, select it, and from thereon all Web uploading and downloading be automatic. Throw in tagging support so I can find them, and I’m a happy guy.

I want the file to live locally so I can edit it even when I don’t have Web access, and so I can have fast, uninterrupted access to it. I’m often occasionally connected – connected now, disconnected for the next two hours (eg. I’m on a plane), back again, and so forth. While on the plane, the file should simply be there on my laptop, be editable, and on landing and reconnecting, the lastest version should magically get uploaded. When I get home and use my desktop, the latest version should magically be there, should I want to view or edit it.

Akbar Ganji is an Iranian journalist who’s been jailed since 2001 for articles linking Iranian political leaders with a series of political killings in the 90s. In his 51st day of a hunger strike, he is dying.

This is an amazing, courageous man. Years ago, in court for his “trial”, he shouted to the cameras “I’m not going to commit suicide. If you hear I’ve committed suicide in jail, don’t believe it”. At the time the reference was to journalists and political prisoners showing up dead, with the official explanation being they committed suicide.

He’s choosing freedom over life. I can’t imagine the courage.