Slashdot Mirror
Slashback: Periodicity, Vacuum, Strength
Bratty kids get to sit near the volatile elements. Theodore Gray writes: "About a month ago there was a slashdot lively discussion about my wooden Periodic Table Table. A bunch of slashdot readers sent me elements for it: Thank you slashdot! Two people actually sent me free Ag and Pd, contrary to the jokes in the discussion. I decided the world could stand another periodic table website. Since all the eight dozen other periodic tables on the web have better reference information than mine, I used some Mathematica programs to generate links to many of them for each element. But my site is more beautiful. I'm going for science as art. Mine also has by far the best quality sample photos: High resolution, high quality macro shots of 89 samples so far."
Starts with a crank, too. ripaway writes "With all the recent stories about vaccuum tubes, I find it ironic that I stumbled on this today. Sterephile reports about the Panasonic CQ-TX5500D(link to Japanese site) car stereo that uses a vaccuum tube, with analog vu-meters. It also plays mp3 files 8-) Naturally, this is for the Japan market only."
Sounds like material for a Burning Man tent ... nm1m writes "A superstrong composite developed by Brigham Young University scientists and students has received financing for its first practical application -- mammoth wind turbine towers able to more than triple the electrical output of existing steel models. Read the story here."
We mentioned this interesting lattice-looking material a few weeks ago.
Sucking requires a context to be good or bad. Sun Tzu writes "After the recent discussion on bad software, how about a different reason for why software sucks? Maybe we programmers and users don't have it quite so bad after all."
That dadburn whippersnapper, why when I was a boy ... Junks Jerzey writes "I remember reading about Halcyon Days: Interviews with Classic Computer and Video Game Programmers five years ago in Wired News. Pretty cool stuff, with an introduction by some guy called John Romero. It was available for a long time as a commercial product that used HTML for formatting, but it's now completely online, as reported by the author."
Software development is driven by clueless pointy-hairs, overreaching sales guys who make baseless promises and people who've never had a single software development class or written a single line of code
I realized this at my last company -- I was in a high enough staff position to see the whole tragedy unfold. Features were driven by what the sales team promised, deadlines by what was written into contracts without development's input, and product managers would bypass the release process and give customers internal test versions of the software. The developers were simply issued marching orders and then ignored.
I believe this is the way most crappy software comes about, regardless of how obvious this process is.
Of course, leave it to the geeks and you'll get Mozilla (good, solid, standards-compliant and really, really late). There's a balance between shipping decent software and shipping a product in time to stay alive as a company. id Software has this balance, ION Storm certainly did not.
Rant over. Please go about your business.
Every year during my review, I just pray the words "slashdot.org" aren't mentioned.
I mean, sure there's a market for a high end tube amp in cars. The car, that rather noisy accoustical nightmare that, no matter how you try, you will never ever be able to fit good speakers in. Oh well... but why does the thing have to look so damn.... tacky? Come on. Analog VU meters and the tube exposed, combined with what looks to be a gold finish. Almost as ugly as a Marantz set.
Not that I think modern car stereos look good... give me those they made about 5-10 years ago: decent button layout, single color displays, and no frigging light-shows. *sighs*
If construction was anything like programming, an incorrectly fitted lock would bring down the entire building...
"People are understandably reluctant to add real engineering discipline to software development..."
... well.. ameteurish. But when they make a model that they have lots of reference of, like the starship Enterprise for example, then they look top notch. Even presented with such a startling comparison, they still refuse to do the design work. Why? Because it adds overhead to their project.
I found this 'alternative reason to why software sucks...' to be true with 3D Animation as well.
As a hobby, I assist people entering into the world of 3D art. My goal is to teach them professional methods to achieve their goals. What I've found interesting, though, is that a lot of them are reluctant to actually design what it is they are building or animating.
With new recruits, I can almost never get them to actually sit down with some paper and design the robot they want to build, for example. What they try to do is just sit down and build it. I'll hear stuff like "Oh I can't draw...", or "It's faster if I just sit down and build it. I know what I want it to look like."
The results? Well, the models they invent are
I really think what happens is that they have in inaccurate impression of what being a 3D artist really entails. This is similar to what Ray said in his post about why software sucks. The sad thing is that until they start taking approaches like designing your model, they'll always look like a 3D newb.
Is there a solution? Well, I have an idea as to how to help both the 3D Artists and the Programmers out there: Make it clear that there is more to their job than just poking keys. I had no idea what all a Software Engineer (I used to call them Programmers...) did until I got a job at a software company. I had the impression in my mind that all they did was write code. The thought of them doing things like 'designing the UI' was alien to me.
Heck, before I got a job doing 3D, I thought all I had to do was build a model as fast as I possibly could. I expected they'd give me 3 days to do what would normally take me a week. I had no idea that they'd actually give me time to design and understand my model before building it. I spent over a year trying to be faster in LW, only to find that faster isn't what they wanted.
In short, I think it's very important to alter the perception out there about what a job really entails. If somebody aspiring to be a programmer knows that they need to pay attention to design and UI, then they'll be far more observant about those aspects during their education. If I had known how much learning to draw would help me with my 3D work, I would have done a lot more drawing exercises in high school.
"Derp de derp."
Since it is three times taller than the usual turbines, that means it'll be visible three times farther away, so it will be an eyesore to nine times as many people. Wouldn't you get just as much power by building three of the normal size turbines instead?
Lighting up under black light is neat, but unless I'm much mistaken that is due to fluorescence and not radioactivity.
The second big problem is weak interprocess communication. UNIX is partly responsible for this. Interprocess communication was retrofitted to UNIX in several different ways, most of them bad. The basic problem is that what you usually want is a subroutine call, but what the OS gives you is an I/O operation. If you build a subroutine call on top of an I/O operation, (think Sun RPC, or CORBA) it's slow. This leads to big, monolithic programs that crash all at once, instead of little, intercommunicating ones that contain the damage caused by a bug. It doesn't have to be this way. Take a look at QNX to see this done right.
The third big problem is DMA. The idea that the peripherals see raw address space and can read and write to it dates from the early days of minicomputers, when it required fewer transistors to do it that way. Mainframes had "channels", which connected peripherals to memory in a controlled, secure way. You could take full control of a peripheral on an IBM mainframe, run a driver as a user program, and still not be able to crash the system. With channelized I/O, drivers aren't as privileged. They can only mess up their own peripheral, not the whole system. This improves system stability considerably. IBM tried to put channelized hardware in PCs, but at the same time, they tried to increase their profit margins on peripherals. This killed the IBM PS/2.
Fourth, Microsoft likes a complicated OS. Ballmer has said so publicly. If PCs came with channelized hardware and a microkernel in ROM, the OS would have far less to do, and would be more of a commodity. There'd be alternatives, like KDE and Gnome on Linux, all of which ran the same applications. Standardized interprogram communication, enforced by the kernel and hardware, would make components more pluggable. All this would dent the Microsoft monopoly severely.
Down at the bottom, at the foundations of personal computers, those are the problems. And that's why software sucks.
Posts like this make me wish for a long-overdue moderation option:
-1 Stupid