Slashdot Mirror
Linux's Difficulty with Names
JohnTyler writes "This article at XYZ Computing takes a look at Linux's strange naming practices. When compared to their Window's equivalents, the names of many Linux programs are difficult to recognize and even tougher to remember. This may seem like splitting hairs, but it is actually an important usability issue. Just think, if you had to do a bit of graphic design which would be easier to pick out of the menu, GIMP or Photoshop? Or if you wanted to play a song, Media Player or xine?" The article is a bit thin, but it raises an excellent point.
ls, rm, df, du, etc . . . did any of the engineers at Bell Labs type 10-fingered?
But then again, you click the "Start Button" to shut down in Windows :)
This message printed on 100% post-consumer recycled electrons.
This is really more of a software designer's issue than a strictly Linux one. As we speak, I am looking for my copy of Daemon Tools on my computer, but I can't find it because it's named in the start menu by the software's manufacturer, not the name of the program. This is the case for many windows apps and I view it as a similar problem.
I suppose Outlook Express is the ideal name for an email client...as is Outlook. Acrobat is the perfect .pdf viewer or creator. Excel instantly draws to mind spreadsheets [now, but 20 years ago?]. I could go on, but why bother.
The article is just more crap slinging between two apes vying for dominance.
"This article at XYZ Computing takes a look at Slashdot's strange naming practices. When compared to their Web 2.0's equivalents, the names of many Slashdot URLs are difficult to recognize and even tougher to say. This may seem like splitting hairs, but it is actually an important usability issue. Just think, if you had to do a bit of news which would be easier to tell your friend on the phone, digg.com slash technology or linux dot slash dot dot org slash, no not linux dot slash dot dot org, i said linux fullstop ess ell aye ess haych dee oh tee fullstop oh arr gee?
The article is a bit thin, but it raises an excellent point.
liqbase
Names don't matter, it is all about training and then familiarity.
What's more intuitive, "Matt", or "Coffee Boy"?
Oh, and what does Exel and Outlook do? Does Outlook Express do it any faster?
As a technical discussion, names as handles to objects or ideas don't matter (excluding downright misleading names, like a boy named Sue): it gets down to user training. To write that "Whatever the reason, desktop Linux's usability is hindered by its naming practices" is just silly: in a work enviornment, users will use what they are trained on. At home, Grandma is going to use whatever will let her get her polaroids out of her new camera.
And Windows isn't particularly easy to use; rather, everybody has had some exposure to it.
As for your examples... once you know what they stand for ("list","remove","disk free", etc.), those commands are a hell of a lot quicker to type (and less prone to error) than spelling the words out.
You're right that many people will probably regard this as splitting hairs, and this in itself is a pretty big issue. Names (from "top-level" names like application titles down to the names of lowly index variables) are critically important in usability, as is documentation.
Yet try as I might, with the notable exception of Python, I've never been able to pick up an open-source product of any complexity that I'm not familiar with, without buying an O'Reilly book or something of the like. Flame me if you will for "not trying hard enough," but it seems to me like having to try hard goes against the definition of usability in some ways. This makes for a pretty big hidden cost.
Open-source projects are the products of engineers working on something they feel is personally important, and it's perhaps unsurprising that communication with the end user (at least on the level of completeness and polish that larger companies need to demonstrate) is not given much priority. But the end users are what will drive the victory or loss of Linux on the desktop, and I think they are already voting with their mice.
And say what you want about Microsoft - but the level of effort they put into this front (from the easy-to-understand language in setup to the MSDN) is way ahead of what I've seen from the Linux world. I think they are the ones to be applauded in this case.
But it makes sense if you think about it.
Maybe if you think about it drunk.
You can't take the sky from me...
But your attitude is EXACTLY what is holding Linux apps back from popular adoption.
.pdf viewer or creator. Excel instantly draws to mind spreadsheets [now, but 20 years ago?]. I could go on, but why bother.
I suppose Outlook Express is the ideal name for an email client...as is Outlook. Acrobat is the perfect
The name of an app is not meant to be Literal!! It's meant to make you want to own it! If you had a choice between two toilets, the Open GNUFeces gtkSepticPort, or a CrapThrasher 3000, is there any question which you would select? Calling a graphics program The GIMP (yeah, I know it's meant to be a snarky acronym; newsflash: after the age of 16, nobody cares.) is like naming your son Susan. In fact, I've introduced the GIMP to new users (all of whom look like they'd rather be anyplace in the world than in that room at the time) with a, "Hey, look, with a name like The GIMP, it's got to be good, right? Right??"
For serious 'flagship' Linux applications, allowing the "coding community" to name them is right in line with allowing the "marketing community" to write them. It screams "Hobbyist," which is fine, if that's all you want it to be. In the early '90's, when nobody knew any better, it was not unusual for an organization's HTML jockey to also be responsible for creating the site's look and writing its content. Then, the medium matured, rapidly. When I see the names for a lot of these (very, very fine and well-coded) linux apps, I get the urge to crank Nine Inch Nails, order a double-mocha-latte, and re-read SnowCrash...
Right out of HS in 1959, I attended the Barns School of Business to learn "computer programming", which was really just using jumper cords to connect holes in a patch board to a neutral bus board on the IBM 402 Tabulator. We used the 540 Gang Punch to enter data onto punch cards for sorting in the 402 Tabulator. The purpose of "programming" was to sort punch cards so the tabulator could tabulate them and sent the results to a printer. But, I looked too young and couldn't convince employers to hire me.
So, after bumbing around on a few jobs for a few years, I went to college. Nine years after my first "programming" training, in 1968, I took Fortran-64 programming in grad school. We used the KSR-133 tele typewriter which supposedly poked along at 10CPS, but if you tried touch typing at 10CPS you couldn't push the keys hard enough to make it punch a hole in the yellow spool of tape into which your program was punched. After spending a couple hours typing in my solution to the quadradic equation I'd put my reel of yellow tape in a brown envelope and the prof would mail it to the CDC6600 computer center in a town 120 miles away. A week later I'd get a printout showing the errors in my typing and the process would start all over. If the printout contained a printing of my program followed by the answer then I had completed the task. IIRC, we completed only 3 or 4 problems that semester. The next semester the physics dept had a contract with the local bank and their B200 computer. We still used the same KSr-133 but the turnaround was the next day because we could go to the bank after hours and see our programs being run. If the errors weren't too bad we could use their KSR-133 and do a "quick" correction on the spot.
I never used Fortran after that class, and it was only ten years later, in 1978, that I got back into programming using Apple II BASIC. In 1980 I resigned my teaching job and I've been programming every since. In 46 years I've seen "programming" go from patching a breadboard to using tools like Eclipse, KDevelop, MSVC++.NET on PERSONAL computers that are millions of times faster than the 402 Tabulator, or even the IBM 1400 series transistorized computer, which was just on the horizon as I graduated from Barnes. Now, researchers are beginning to get a handle on optical and quantum computing with 10's to 100 GB of RAM and all solid state storage replacing mechanical HDs.
What a ride it has been!!!
Running with Linux for over 20 years!
It has a lot to do with the fact that open source geeks can't afford trademark lawyers. A name like "gwksprt" may be horrible, but at least you're unlikely to be sued over it.
Uh, because that's basically what the article was doing, comparing Windows menu entries to aan opensource application's executable.
Looking through my menus in Gentoo, it is fairly clear what the program does either by the name, or icon. There are some exceptions, like GIMP is just called "The GIMP", but at least it is under the Graphics menu group so I've a pretty good idea what its function is even without recognizing the application. That actually applies to most applications under my Gnome desktop actually. Everything is grouped by it's function, unlike Windows where typically applications are grouped by manufacturer.