LinLog

If you're a masochist like me, you may have see OSNews's link to Sal Cangeloso's recent item on Linux application names. His premise is essentially that the usability of Linux systems is hindered by bad choice of names, where application names in Windows are easier to associate with the application function. To support this, he gives a list of application categories with some of the more popular Windows and Linux applications that fit them. The problem is, he cherry-picks his examples and completely misses the real issue.

If you read Cangeloso's list, there are nine application categories. Five of them are related to multimedia, while the important "office" category is completely missing. Hardly a diverse sample to begin with. Worse yet, in four of the multimedia categories, he chose Windows Media Player as one of his Windows programs. While I'll readily concede that the name "Windows Media Player" clearly indicates an application related to media, hammering on a single example over and over again does not establish a general trend. In fact, it just makes the argument look weaker when you consider some of the popular office applications that didn't make the list (presumably because they didn't support the argument). Examples such as Access, Outlook, Excel, and Acrobat come to mind. I find it hard to believe that anyone who hadn't already heard of those applications could work out what they're supposed to do from the name alone.

Of course, any such list will be much to small to establish any real pattern, as there are a plethora of good and bad examples on both sides. But you know what? It doesn't matter, because application names don't really have much to do with usability. In fact, application names have nothing to do with usability at all. Application names are about marketing. Would it really make Outlook easier to use if it were called "Microsoft E-mail and Personal Information Manager"? Of course not. In fact, the only thing it would do is make Outlook harder to talk about because of the unwieldy name.

The real usability issue here is in how the user opens the application. It doesn't take much thinking to realize that this need not necessarily involve the name of the application at all. For instance, one of the more common methods for opening a program is to double-click on an associated file, e.g. double-clicking on a document to open it in Microsoft Word. Likewise, many applications install custom context menu items that do not make reference to the application name, e.g. the "Add to playlist" item used by Windows Media Player. This is good usability - it clearly communicates what the item does without adding any unnecessary information.

What Cangeloso is really concerned about, though, is the system applications menu. The usability issue here is different. It is discoverability, i.e. how easy it is for a user to find a new application of which he isn't already aware. However, as the image on the first page of Cangeloso's article so nicely shows, this is already a solved problem. The image depicts the KDE applications menu with program descriptions turned on and items nicely arranged into broad function-based categories. In other words, rather than the user simply seeing "Firefox" and wondering what on Earth it could be, he will see an "Internet" category that contains an item that reads: "Mozilla Firefox (Web Browser)". Given this extra information, it is hard to see how changing the application's name itself would make things any easier for users.

As I said earlier, names are for marketing, not usability. The idea is to make the product recognizable and get users to remember it. A descriptive name can help in this regard, but it is far from a necessity. If we're going to criticize the names used by free software, let's at least address real problems, like names with bizarre pronunciations or names that consist largely of abbreviations. Now those a valid criticism. But to complain about an application's name because it harms system usability is just bizarre. If changing the name really makes the system more usable, I'd venture that you have much bigger problems to worry about.

I got a great new toy yesterday. It's a $13 cable. But not just any cable - it's a data cable for my cell phone. Combine it with BitPIM and you have the solution to a problem that's been bothering me for a while now.

You see, my wife and I both have Verizon cellular phones. If you don't already know, the deal is that you sign a two-year service contract with Verizon and they give you an obscenely huge discount on any phone that you buy at the same time. I'm talking like 50% or more off the normal retail price. And sometimes they even have brands that are buy one, get one free on top of the discount. So unless you're filthy rich, you can't afford not to get a new phone every time you renew your contract. This is kind of a win-win situation, since you get new hardware at a good price and Verizon gets to keep down the amount of old hardware on their network.

The only thing that bugs me about this is that every time you get a new phone, you have to sit down for an hour or so and enter all your contacts into the new phone. Sure, if you buy the expensive phone they'll give you a data kit that can sync your calendar and contacts with Outlook, but as a Linux user that doesn't do me much good.

Another annoyance surfaced a year ago when I got our current phones. This time I got camera phones - a pair of Samsung SCH-A670 camera phones, to be precise. I actually only got them because they weren't much more expensive than the non-camera model (like $20), but it turns out that having the camera is kind of handy sometimes. The only problem is getting the pictures off the camera and onto the computer. Sure, you can sign up for Verizon's picture messaging service which has some kind of web interface, but that adds something like $10 to your monthly payment. You can also send the pictures by e-mail, but then Verizon charges you $0.25 per message. That seems pretty silly when all I want to do is move the pictures the three feet from my camera to my hard drive. I ought to be able to do that for free. And it turns out I can.

A couple of weeks ago, I saw a comment on digg.com that made reference to BitPIM. It's basically a PIM application that's designed to sync with cell phones. It can sync calendar and contacts, download pictures, ring tones, text messages, and various other things. There's even a feature that lets you browse the phone's file system. And the best part? It's written in Python with wxWindows, so it works on Windows, Linux, and MacOS.

(Note: There is no Ubuntu package for BitPIM, so I had to download the RPM and convert it with alien. The only problem I had was an error about being unable to find libtiff.so.3. The correct way to fix this is probably to install that version of libtiff. However, I used the "cheap symlink" fix, which was to create a symlink named libtiff.so.3 that pointed to the version that's actually on my system, which was libtiff.so.4. It worked for me, but your mileage may vary.)

So now I can get all the data on my phone onto my PC. BitPIM even has an option to import and export vCards and other formats, so it's possible to sync my phone with Kontact, even if it is in a round-about way. Now all I have to do is find out if there's any way to make KDE talk directly to my phone. The ability to sync with KAddressBook and KOrganizer would be great. An ioslave to access the phone's filesystem would be great too. I think I feel a new project coming on....

I upgraded to KDE 3.5 a couple of days after it came out. My reaction? A pleased yawn.

Let's face it: there's nothing terribly interesting about this release. For me, there's even less than there was in KDE 3.4. I see a few more icons, some better menu grouping, and the addition of a volume manager which doesn't do quite as much as I would have hoped. Of course, those aren't the only changes, but the major ones listed in the announcement don't effect me, as I don't browse with Konqueror, don't use Kopete or SuperKaramba, and don't care about the edutainment packages.

That said, this release is still worth the upgrade. For one, they finally seem to have fixed that annoying problem with desktop icons moving around between sessions. The volume manager is also nice, although I would have liked a little more granularity, e.g. the ability to set actions for particular volumes, not just MIME types. In other words, another incremental improvement:good, but not exciting.