Slashdot Mirror
Scientists And Engineers Say "Computers Suck!"
drhpbaldy writes: "At the latest ACM meeting,
scientists and engineers threw mud at computer scientists for not contributing anything useful. They lambasted CS types for developing complex and useless technologies. Some of the fault was placed on VC's for funding only the fanciest and stupidest technologies." Of course, when people say that "design" will save the world, they usually mean their idea of design, which might not jibe with yours or mine.
When engineers sneer at computer science, I just chuckle to myself. Because I know something they don't know: they're just jealous.
Engineers are jealous of programmers. It's that simple. Programmers have an easy life, after all. I only work a few hours a day, get paid big bucks, and for what? For telling a machine what to do. For the act of mere speech. It's Rapunzel's tale incarnate: I spin words into gold.
Engineers have too many constraints; the standards are too high. When the Tacoma Narrows bridge fell down, heads rolled. But when a bug in the latest version of NT disables an aircraft carrier, Microsoft doesn't get blamed at all. Bugs are par for the course in our industry, and we have no intention of changing it. It means higher profits from fixes and lower expectations. How are engineers supposed to compete with those sorts of odds?
I admit I considered going into engineering when I started my college days, but I was quickly disuaded. The courses were too involved, whereas the CS courses were a breeze for anyone who didn't fail calculus. And I don't regret it at all, really.
Programmers might not get the satisfaction of building something useful and might not experience the artistic delight of design, but we at least don't have to work as hard. And when it comes to the bottom line, that's all that counts.
That's because when someone comes up with a useful technology, even something as simple as LZW compression or an MP3 encoder, NO ONE ELSE CAN USE IT in their product. Writing products that use someone else's file format is called a "copyright violation". Standardising on one crypto algorithm is called patent theft. CPUs with compatible MMX instructions gets you sued by Intel. Making DVDs playable on "non-approved" systems gets you jailed, or orders from people halfway around the world.
So yeah, "CS types develop complex and useless technologies." because we have to carefully avoid reinventing someone else's wheel or we get sued into bankruptcy.
One result is millions of different wheels of different diameters, shapes and track widths that are all incompatible with one another. Sounds pretty messy, right? It also happens to resemble what we see today in the computing industry.
The other result is people getting fed up with all the incompatibilities and looking for a standard, any standard. And since the standard is proprietary, naturally this will favor the growth of monopolies, e.g., Microsoft, who thes uses their position as OS "standard" to create other standards, such as Excel and Word formats, whilst actively blocking anyone else from participating in that standard.
IMO both patent lifetime and copyright lifetime ought to be cut to 10 years tops for all things computing related, hardware or software, because stuff in this field ages faster than any other traditionally patented and coyrighted work.
And there needs to be an irrevocable expiration for abandoned patents and copyrights too. It's absolutely insane that Atari 2600 games are still locked away by copyright, while no one is prodcing them. And they'll be locked away for over a century under current IP law. Is this right?
If it wasn't for us software guys, you scientific types would still be writing programs in Fortran.
Oh that's right, you ARE still writing in Fortran. My bad.
Anybody remember the original Apple II?
.. typically these were pieces of firm foam rubber that you would place underneath the RESET key (you had to pry up the keycap) .. resulting in a key that was still "pressable", albeit with a bit more effort.
//c, and IIgs), Apple listened to their users' complaints, learned from their mistake, and required a Ctrl-RESET combination in order to actually trigger the reset. That hard-learned lesson carried over to other hardware and software manufacturers, including the choice of Ctrl-Alt-Delete.
The RESET key, located at the top-left corner of the keyboard, triggered a software reset. This had the effect of (depending on the software you were using) terminating the program and dumping you back to a BASIC prompt or erasing whatever unsaved data you had or doing a hard reboot of the machine. Users quickly found out (the hard way) that this button was way too easy to press by accident. In fact, this problem was so pervasive that magazines such as Creative Computing began advertising for "RESET key protectors"
In later versions of the Apple II/II+ (and in subsequent machines such as the IIe,
We're going down, in a spiral to the ground
Targets of the critics' scorn included convoluted commands such as the common "ALT-CONTROL-DELETE" sequence used to close a program or perform an emergency shutdown.
Put it under F1, see if that makes them happy. You know, there's a reason it's such a 'convoluted' command, It keeps people from accidently executing it!.
I thought my sarcasm was pretty good, thank you very much.
Perhaps my bile was uncalled for, but I'm sick of people implying "good design is easy, why doesn't someone just do it?"
Good design and usability are difficult. Do you think that the industry doesn't know that billions of dollars and instant fame and fortune are at stake here? Do you think that the industry doesn't try really, really hard to get that money?
There's a right way to criticize usablity--one author who does it right is Donald Norman (I'm sure there are others, but on this topic I've only read Mr. Norman's books). He manages to carefully consider what is wrong with a designs, discusses the alternatives, and points out how usablity could be improved.
There's also a wrong way. Say something "Why can't my computer be as easy to use as a toilet?" God, I'm getting pissed off again. What's the feature list of that toilet? And what's the feature list of your computer; can you even hope to cover that list in any amount of detail? In fact, does your computer actually even have a standing feature list, or do you actively install new software (and thus new features) constantly? Dammit, everybody who uses a computer has complex needs--I have a web browser, an email client, a remote telnet session, an mp3 player, and a "find file" all open RIGHT now, and I suspect that I'm pretty tame compared to the average slashdot reader. I'm going to play an online game with friends in another state shortly. I could spend hours describing what I want EACH ONE of these to do. I happen to think that all facts considered, the usability is pretty good. (And I might add: Damn, it's cool to live in the year 2001. This stuff rocks.)
Are things perfect? Of course not. One company has a monopoly on the desktop market and has very little incentive to innovate (in spite of their claims the contrary) and every incentive to continue to keep the status quo. Yes, the "START" button is retarded. Should we strive to improve the state-of-the art? Of course. Would it be awesome if it was easier to do my taxes? Sure, but are you absolutely you want the automated solution you described when it sacrifices transparency (are you sure your taxes were done correctly in that system) and possibly privacy (who's keeping track of all that information flowing between your income-payers and the government?) I actually think that TurboTax made my taxes about as easy as I'd like--it asked me a simple set of questions, I answered, and it was done. Any easier, and I'm not sure I'd completely trust it.
I actually don't know why you're arguing, since in at least one respect, you agreed with me. You said:
Simplicity of interface, sheer useability, takes a lot of talent, skill and creativity.
If you think about it, the article in question basically said these are all trivial, require little skill or talent, and they said it with a condescending attitude. It's actually really really hard. Dismissing the problem is unwarranted and deserves and equally scathing reply.
Dammit, I hate these fuckers.
First of all, they contradict themselves. "Computers are too hard," they whine, but when a computer interface remains consistent and usable for twenty years, "If Rip Van Wrinkle went to sleep in 1982 and woke up today, he'd be able to drive our modern computers with no problem because they're essentially unchanged".
Second, they recommend creating "simpler" and "distributed" devices instead of monolithic boxes that do everything. What the hell does this mean, what devices really need more intelligence? All I can think of is one of those computerized thermostats. Whoopee.
Look. Computers are complex because your needs are complex. Worse yet, my complex needs are inconsistent the needs of others. Try to download mp3s on your toaster. Try to do your taxes while downloading porn while instant messaging your friend in France while checking the weather on one of their great appliances. Try to use that "more intelligent than a computer" airport toilet to write up your Powerpoint slides, you pompus pricks.
Actually, in this case, that might have actually worked.
I just love it when Scientists fling mud and proclaim that the 'real-world' isn't science.
In mathematics, we have the very 'real' Taguchi quality control that revolutionized manufacturing processes, but according to my Math professors, "It's not real mathematics, just some linear algebra application."
On the topic of manufacturing, Metal can now be formed and machined into virtually any shape, Ceramics and metals can be mixed and then burned to form Aluminum tools (molds) for injection molding parts. "That's just an trick to sintering the ceramic" my ceramics professor told me.
My point is that industry types, whether they are applying nueral networks to read handwriting or creating thinner flat panel displays, solve the same complicated types of problems that the more 'scientific' community solves. The scientific community discredits their work because "Theoretically it can be done, so why bother doing it." It's as though the companies that want to enhance their products by funding research shouldn't fund the research that is most likely to enhance their products!
I'm sorry to sound harsh because this strikes close to home for me. I was on track for PhD, but quit and now I'm having a lot more fun developing optimized neural networks to do hand writing recognition.
Keeping
I love my computer enough as it is. If I had a computer that sucked, I'd never leave the house!
-atrowe: Card-carrying Mensa member. I have no toleranse for stupidity.