This blog is now hosted at

Wednesday, November 21, 2007

Ribbecke Halfling

I'm having an acoustic bass guitar built by Ribbecke Guitar Corporation, and the first pictures of the build are now available.

If you haven't heard of them, check RGC out-- they make the best acoustic bass guitar I've ever heard. Read more...

Monday, November 19, 2007

Mike Huckabee Uses One Of My Favorite Internet Memes...

So now, in addition to being a bass player (we're all about style over substance here), Mike Huckabee has thrown down the gauntlet. Vote for Huckabee or face the wrath of Chuck Norris.

Chuck Norris doesn't have a chin under his beard-- he has another fist!

Clever indeed, and this video will definitely "go viral", due to the existing viral nature of "Chuck Norris Facts". Read more...

Wednesday, November 14, 2007

Teh iPhone Causes Typos?

Okay, I was right, and I was wrong.

The iPhone isn't significantly slower than other types of keyboards to type on, but it does cause significantly more errors.

An article at ITworld outlines the results of a study into speed and accuracy for different input methods. iPhone users made over twice as many errors (5.6 per message vs. 2.4 for a numeric only keypad).

Apple made a bet on an experimental user interface. Many of those aspects worked well (multi-touch is going to be a de-facto standard over time). Some of them, (the keyboard) didn't work so well.

I know that Steve Jobs hates buttons, but should his jihad continue to the detriment of his users?

Update: Roger Matus covers this in his blog (Death by Email): link

Tuesday, November 13, 2007

Android De-Fragmentation

With Android being released as an open source operating system, one of the risks is that the companies in the Open Handset Alliance could fragment the development of the OS. It turns out that Google has worked to prevent this.

It seems fairly obvious, but balkanization of Android would likely cause significant issues with its adoption. The real issue is: who does balkanization affect?

Answer: the users.

Up until now, the cell phone manufacturers and the service providers have been united in denying user choice. I have high hopes (some days I'm an optimist!) that Android can be the straw that breaks the camel's back in this hegemony.

If android actually lives up to my current expectations, we're due for a serious inflection point-- and another Internet revolution.

Vive la choice!

Monday, November 12, 2007

Android SDK first look

Last week, I took a brief look at the Blackberry SDK. Well, it's November 12th, and it's A-Day. Here are some initial thoughts on the Android SDK.

  • Kudos to Google for producing a cross platform SDK. Win/Mac/Linux is nice, since so many developers today use the latter two as their primary OS.
  • In looking at the documentation, everything seems to follow the web development paradigm, but with a few detours for things like long running processes (like the example they use: a media player that plays music in the background even when the UI isn't showing)
  • The interfaces look very open; you can register for external events (e.g. an incoming call or SMS message)
  • The Content Provider interfaces look nice, and the XPath style access to data is a nice, clean, and open interface.
APIs are nice and all, but what about tools?

I'm happy to note that the tools all seem to be there: Plugin for Eclipse, debugger (both against the emulator, and remotely against a device), plus all the packaging tools. Nice.

Other things to note:
  • Location based services, with integration to Google Maps.
  • 3D graphics (OpenGL). The support looks somewhat immature, but with NVidia as part of the Open Handset Alliance, this might just become a decent homebrew gaming platform.
  • Low level hardware access will be available, but isn't ready yet.
  • XMPP support; this will enable P2P applications and communication (IM, games, etc.)
  • Android has support for recording and playback of audio and video. As a musician, I like the audio recording bit. Nice.
  • They have some really nice documentation available; They even cover some of the things that developer level documentation doesn't normally cover, like how to make an app responsive, how to understand the lifecycle of applications, etc.
With HTC announcing that they are going to release 3 handsets based on Android next year, it looks like Windows Mobile might be in trouble. If they can make a well-integrated, responsive phone that looks nice, they will probably start to cut into the iPhone market as well.

Looks like the smartphone market is about to get a whole lot more interesting. Read more...

Thursday, November 8, 2007

The right to a trial is sacred

Normally I don't talk about things that go on in the office, but the PFY managed to get himself a rather steep speeding ticket.

So, he's doing the smart thing: getting a lawyer!

I've heard all the excuses, I'll ask for leniency, etc. All hogwash, in my opinion. If you're in the Seattle area, and you get a ticket, do yourself a favor and call Jeannie Mucklestone. She's great. Read more...

Tuesday, November 6, 2007

Blackberry SDK

I've been considering a development project for the Blackberry, so I downloaded the Blackberry SDK from RIM. I was pleasantly surprised:

It looked quite complete (having looked at J2ME a few years ago, yuck!)

They included many examples, including some that would be difficult without an example, i.e. Bluetooth Serial

All in all, I'm quite impressed with the level of effort that obviously went into producing this SDK, and it is nice to see device makers actually courting developers.

I'm hoping to do some comparisons or at least an evaluation of Google's Open Handset Alliance/Android toolkit, so watch this space after the 12th of November (when the toolkit is due). Read more...

Monday, November 5, 2007

GPhone Isn't a Phone, It's an Android

The much awaited announcement did actually happen today, it just wasn't what everyone was expecting. Google announced the Open Handset Alliance (OHA). While the gadget freak in me would have liked to have seen a shiny new device, this actually is a great announcement. Read on for more analysis.

The Open Handset Alliance has definite potential to be a market changer. Let's all admit it: cellphones are currently a game where you decide what deficiencies you're willing to live with. Things like:

Windows Mobile: Microsoft has been steadily proving to us that they are incapable of producing a decent quality mobile OS. They always have all the checkboxes (fast processors, great screens, etc.), and always come up short.

iPhone: The device that makes Apple look like IBM in the bad old days.

Blackberry: not quite as powerful or featureful as, say, Windows Mobile

I've long thought that Linux would be great for mobile devices, once they hit a certain level of CPU/Memory capacity. The main downfall with Linux is that nobody seems to go the extra mile to make something that "just works" (and kudos to Apple for doing just that, unless you want decent bluetooth support).

If the OHA can figure out how to put out a bunch of devices that "just work" but are also an open platform that actually invites developers to produce software, we may just have a winner on our hands. It is about time for someone to apply bazaar development practices to the traditionally cathedral mobile market.

As these devices become more powerful, the PC will begin to lose its dominance, and having an open platform upon which to make that transition is a huge win for consumers.

Thursday, November 1, 2007

Bayesian Filtering: Why Not?

The earlier pieces I did on Ron Paul spam (here and here) point to some problems with Bayesian filtering. Read on for some further analysis of the problems with Bayesian filtering.

Bayesian filtering, in a nutshell, breaks down an email into words and or phrases, and then assigns a spam probability to each, based on the word's previous penchant to be spam. For example, let's assume that spammers are sending messages advertising a product called "Bradley". Over time, as those messages are categorized as spam, a Bayesian system would give an increasing spam score to the term "Bradley".

This type of system began to be deployed widely starting 3 or 4 years ago, and was very effective for a couple of years.

Bayesian Poisoning

As seems to always be the case, the spammers switched tactics. They began sending out their spam with a large number of (typically incoherent) words stolen from news sources or literature. This throws off the Bayesian system by

  • Adding to the non-spam score (since there are "good" words in the mail), and
  • Putting the good words in the spam list

Beginning about a year ago, I started receiving spam that only had the good text, no advertisements. This was a deliberate attempt just to poison Bayesian systems.

The ongoing issue with Bayesian systems is that spammers have fairly effectively figured out how to confuse them (either by falsely calling acceptable mail spam, or letting spam go through). Fortunately, the state of the art in spam detection is being pushed forward as well.