Razor & Blades

The business model for Amazon’s Kindle is very interesting to me.  It’s one of few products I can think of that offers free access to the cell phone network, even if you are using it to line Amazon’s coffers just a little more.

The wireless market is so proprietary and inaccessible in the US that any way of utilizing it that doesn’t involve a two-year contract looks like a breakthrough.  We’re also seeing little pinpoints of light in Verizon’s open data network and unlimited calling plan.  Of course, pulling off iPhone’s Visual Voicemail was also a jaw-dropping accomplishment, even if the concept was blindingly obvious to everyone who ever hears about it.

Are we nearing the end of the wireless phone carriers, and (finally) heading to wireless data pipes, as we need to?  Once we unbundle the device from the network, I think we’ll start to see lower prices and better availability.

Then, the real fun begins.

iTunes Plus Launches (to Unexpected Success?)

There’s no way for the public to know for certain, and I’m sure they’re not going to say a word until WWDC, but the success of iTunes Plus—Apple’s DRM-free music offering with EMI announced last month—seems to be too much for the iTunes server cluster to handle. Even late into the evening, navigating through the iTunes Store is taking much longer than usual, and song previews are frequently stalling while waiting for the servers to catch up.

Honestly, I don’t know if I’d have it any other way. The geek community’s been demanding DRM-free content for years, and it looks like the recording industry and electronics industries are finally starting to listen.

Ars Technica noticed similar iTunes slowdowns while reviewing the new service. I suspect that music upgrades will work more smoothly as Apple increases server capacity and/or the rush slows down somewhat.

In other news, Apple parted $9.99 from me for my first online music purchase ever. See, that wasn’t so hard, was it?

Removing Duplicate Dictionaries from Script Editor

For a few months, I’ve had a number of duplicate dictionaries listed in the Open Dictionary dialog box of Apple’s Script Editor. (This application is used to write and debug AppleScript code — the Mac equivalent to batch files or shell scripts.) While this wasn’t really hurting anything, it was kind of annoying to have the same program show up five or six times, typically after I upgraded the offending apps to newer versions. Since I had trouble with a Google search for "Script Editor" "duplicate dictionaries", I thought I’d share the solution I eventually found.

The solution turns out to be forcing Launch Services to rebuild its database. The database maps file types to applications that are supposed to open them, but sometimes it gets confused and can hang on to old versions of the same program. The magic incantation, according to macosxhints.com, is:

-kill -r -domain local -domain system -domain user

Thanks to this post at Apple Discussions for pointing me in the right direction.

Tracking truth

As I mentioned in my previous post, organization on today’s computers is getting harder and harder, given rapidly increasing disk space and more stuff we want to store.

One of the more important techniques I’ve developed to combat this problem is something I’m calling “designating the truth.” It’s a simple concept with strong real-world parallels, but it often gets neglected in the computer realm.

“Truth” is based on terminology used by Apple’s Sync Services API. The concept of Sync Services is invisible: Mac OS X keeps a variety of data, including calendar events, contacts, and tasks in internal databases that users can’t get to. This data is used to synchronize any software, web sites, and portable devices that want to make use of. Software that the everyday user thinks of as the place for their information – iCal, Address Book, and so forth – are just pretty Sync Services clients.

Once everything has its place, the trick to staying organized is to designate which copy of a given piece of information is authoritative. This copy has a variety of different names. For example, “truth” is Apple’s moniker, while “golden master” is used in software manufacturing.

Here’s where I keep my designated truth for different types of information:

Type Location of “truth” Additional copies
Appointments OS X Sync Services (iCal) Palm PDA
Contacts OS X Sync Services (Address Book) Palm PDA
School work Subversion Working copies on a shell server and my Mac
Music FLAC files on a home server iTunes, Palm PDA, Creative Rio, and more

The beauty of this approach is that I can backup to my heart’s desire, and I no longer lose track of which version is which. This is particularly important for my school work – now, if I want to do an assignment at school, it travels to me in a Subversion working copy on a USB key. When I get home, I check in changes to my server, and no longer have to worry about which copy is more recent.

I also can add new gadgets (the iPod is tempting), and since I know what my truth copy is, I’ll never overwrite it, knock wood.

Where’s My Stuff?

I’ve become convinced that the most pressing challenge of computer usability in the very-near future is organizing the mass of data we’re now capable of accumulating into something we’ll be able to use effectively later. With hard drives becoming larger and cheaper at the same time, the amount of data that a person can accumulate is becoming mind-boggling.

Take me, for example. Between a thumb drive, Memory Stick, SD card, external hard drive, two drives in a Linux server, Gmail account, web hosting provider’s space, and my Mac’s hard drive, I’m about to break the terabyte barrier. (I already have, if you count the dozens of CD-R’s and DVD-R’s lying around). Consider that my first computer had a 20 MB drive (that was an upgrade), and it’s amazing how far the technology has come.

The problem, I’ve noticed, is that with all of this space, things get lost. Developing ways to organize my files has been challenging. Apple helps me somewhat by integrating rich metadata support into the OS, but I’m of the school of thought that says that making an effort to sort files makes for easier retrieval in the long run.

Keeping Synchronized

Another related problem has been keeping track of multiple copies of the same file. This was a nightmare – I work from multiple locations, and having to track changes made to different documents on different media at different times is a monumental undertaking. Thankfully, through the slightly eccentric but surprisingly useful advice of Debian Linux developer Joey Hess, I’ve found a solution that I’ve been adopting: version control.

Version control is useful for programmers, but it has a major strength for non-programmers: it forces users to designate a “golden copy” of a file that all other copies are updated against. Suddenly, the task of figuring out which file has which change disappears, because the version control software (in my case, Subversion) takes care of all that.

I’m not quite up to Joey’s level of reliance on Subversion (he keeps everything in it), but I’m getting there. Currently, almost all school work is in Subversion – the major exceptions being large media files that are almost certain not to change. As I continue working with Subversion, I expect that my setup will eventually look a lot like Joey’s.

Microsoft/Apple: The Letdown

Perhaps the primary reason we’re so far behind in the file organization war is the less-than-stellar tools provided by Microsoft and Apple. Apple currently holds a commanding lead with Spotlight and iLife, but there’s definite room for improvement. For example, opening a shell window to query Spotlight with specific metadata is a usability disaster, as is trying to figure out how to organize images that aren’t photos (importing them into iPhoto seems like a waste here).

And as for Microsoft? If they had spent the last years concentrating on WinFS instead of buggy, draconian copy protection technologies that nobody actually wants, I might not have had to write this post.