Friday, June 10, 2011

Social Media Best Practices, IV

In this installment, I show how paper.li links in Twitter are just plain annoying and non-conversational.

Wednesday, June 08, 2011

Social Media Best Practices, III

In this video, I really tear into foursquare. I point out how boring and aggravating it is to pollute Twitter with 4sq links.

Monday, June 06, 2011

Cynical Speculation about Motivations behind Linux's Interface Problem

ANALOGY: My wife rented a car. She did not get my advance approval as to the make, model, or year of the auto. I did not even see it until a few seconds before I was supposed to drive it. I went outside, and there was a large minivan in the driveway. I got into the drivers seat and started it up. I didn't even have to think about how to turn it on and access the most critical and basis functions. The human interface from this 2010 model was essentially unchanged from the human interface on the 1970 car I learned to drive in. Across multiple manufacturers from multiple countries, nothing in the interface has radically changed, even though the car's engine & radio & heater & crash protection systems can do far more. I'm guessing the human interface had stabilized long before 1970.

No so with Linux. There are all sorts of problems with every new release. Audio doesn't work. A new desktop environment ships (KDE 4) which makes it impossible to transfer files from one directory to another via the file manager program. Another one (Unity) just befuddles expert users (Brian on Linux Action Show) to the degree that he asks what a Linux enthusiast is supposed to do. Jono Bacon of Ubuntu admitted to "The Linux Action Show" that Ubuntu 11.04 was a "patience release." Gnome 3 threatens worse. Five years ago, distros (OpenSUSE) were alternatively dropping and adding WiFi support. This sucks. What is so maddening, even to those "partisanly" in favor of the ideals of Linux, is that these interface-wrecks show up with a simple "dist upgrade."

Back to the car thing. Right now you can just hand a friend the keys to a car and expect them to be able to hop in your car and drive it. You only have to tell them where it's parked. But imagine cars were like linux. You'd take your car in for a service pak upgrade and BAM!-- you'd be faced with the problem of figuring out where the brake pedal was. The more intelligent among us could probably find the brake pedal with only a small time investment, but the point is why? Why are you fussing with the human interface each release? Are you completely incapable of tuning the engine, or do you see progress mostly as hiding the steering wheel each time? Most of us are very smart and even like to do logic puzzles in our spare time. But we don't like to have to solve logic puzzles when we're trying to do a job with a tool that worked fine yesterday.

I've been stewing about this since at least the first time I tried the first version of KDE4 that showed up with Kubuntu. I have three admittedly cynical speculations as to what's driving this:

  1. Linux devs really are that dumb.
    They are incapable of seeing what a normal human being--no, what a person who needs to do some work right now-- needs to do. They really do see hiding & wrecking the human interface as improvements. Perhaps it may be institutional myopia-- a bunch of blokes sitting in disparate basements across the world might not know the aggregate effect of their tweaking.

  2. Linux devs really are that mean
    They care not about making a working product. I once printed out a forum posting where one said he really didn't care if Linux became an OS for people's grandmothers. IMO, that's the exact same attitude as "don't care if we are ever Enterprise ready." They like hiding things, in the same way that frat boys like to haze freshmen. It keeps it an elite system for the elect.

  3. Business majors running Linux corps really are that cunning.
    Linux is released under the GPL. As such, they have to release the code to you but can sell support services. Anyone can get the code, but they can sell support. Many distros have an "enterprise" version and a "community" version, such as Red Hat Enterprise Linux vs. Fedora. or SLED vs. openSUSE. The community versions are probably what both the uber Linux power-users will seek out and the grannies of the world will stumble on to. These two groups are the ones who will be the most maddened or turned off of Linux by this jerking-around. My cynical speculation is that the business majors running the Linux corps do not care. The more cynical of them might even be able to reason a business advantage out of the rage. If you were in a situation where you were forced to release a free version of your product, could you see an advantage in making your free version buggy & klunky & unpredictable, which would either drive people to get the heck out of there, or to get a business subscription to the corporate version? That's my guess. This possible motivation is more evil, perhaps, but it gives the corps running Linux more credit for having some business smarts.


Disclaimer: my day job involves a company that is involved in Linux, but I myself have nothing to do with this area of the company. I work in computer hardware. Opinions solely my own and based on my own interactions as a hobbyist Linux enthusiast.