I just read somebody else's comment about click-through counters. Post any arguments about that as replys to that comment. (It's a reply to one of the top 5 posts, subject is descriptive. You find it.)
I propose that we launch a massive Slashdot statistics program. It would involve a significant amount of invasion of "privacy", i.e., almost every link off of/. would be logged, but we could get a lot of raw data to look at in interesting ways. Here are a few of my ideas:
Karma tracking: This itself would not need any data-gathering things. Grab stats based on user's karmas. Examples:
Number of requests to main page by karma
Number of comments posted per (unit of time) vs. karma
Average moderation vs. karma
Number of users at a certain karma
Number of comments containing the words "FIRST POST!!!" vs. karma
Collary to the last one: Number of times it actaully is first post vs. karma
Stats:
How many times users check their stats (could also study karma on this).
Number of times other users check a user's stats
Areas: Record popularity of different areas of Slashdot (Funny, Red Hat, GNU, etc.) by how many users read articles and of those, how many post comments, and the average moderation of those comments (and average karmas of people who post there; what "the elite" visit).
Referring=Slashdot: Quick redirects embedded in each link off of Slashdot (at least on the main page, maybe even on users' comments, and don't forget those *@$# ads) could could hits by link. This would make studying the Slashdot Effect even easier. If the Slashdot servers were fast enough, and you could analyze the ethical issues of doing so, links could link to a cgi script that would request the web page, and this way the banner ads and other stuff would be preserved, but you could do more counts like whether the server was overloaded or (ethical question mark) track links from the linked-to site.
There are more! Let someone else suggest them, though. I think that analyzing the raw data gained by these approaches would be a good statistics project for any upper high school or college student. And hey, I like Slashdot, and if I get credit for it, even better! (Well, it would be mostly raw data, not the real site, but still...)
Ken
PS - I would do such a project. Anybody who wouldn't?
Linux is so similar to other Unices in the development environment that you could argue that everything targetted specifically for Linux (except for some of the low-level stuff, and very few people are working on that) is really targetted for the entire Unix community. You can't see the exact motives for writing the software. But what happens is that most of the people who actually use the software happen to be running Linux. How Linux got to that point is beyond me because I did not personally experience it (Hey, what can you expect from a 14-year-old?) but I think it has something to do with the driver support, being free, and being fun, but that's just what I glean from other people. I just can't accept a 99.98% figure without a little contesting.
Business computing is the kind of computing that makes money. There are a lot of people who spend their spare time messing around with their computers and trying OSes like Linux, but what do they do to get their money? They work. And what do they use at work to (hopefully) boost their productivity? Computers. And there are a lot of people who use computers at work who couldn't care less about them once they walk out to their car. If you want to address the largest number of users, you have to get into the field of business computing. There are exceptions to everything.
Ken is done arguing.
PS - What goes up and never comes down? Tom C.'s karma! Are you some super-nerd? You have something to say about practically every news story in the past few weeks at least, if not the last year or so. How do you get a karma of 74 (last time I checked)? Well, for someone who wrote a book on Perl, that's sort of a given, but...
What's wrong with a Microsoft accent? I think that Linux will eventually become everything that's good about Microsoft's OSes and more without the bad parts of MS.
True, nothing I list is really inherent to the kernel, which is what Linux technically is. I was targetting the specifics of Linux on the desktop. As I said, I'm sure that I missed a few things. Linux on the desktop is what the average Joe Computer User is going to see, not the kernel and stuff like that. If Linux is to succeed, Joe has gotta be happy with it. It seemed like the remote apps thing didn't hit the right target. Remote applications could be the future of business computing. Each user could have even a spare 386 running the SVGAlib version of the VNC viewer connected by a decently fast network to a few powerful computers managed by people who really know what to do with really powerful computers. The question is return on investment. It is not very difficult to add another user with VNC server on the main server and put a spare 386 in their office, but in today's business world where most everyone has their own computer, installing a new computer means going through complicated setups and other time-wasting things and then having to upgrade it when users start using an app that requires more speed. And of course there are calls for help and stuff that require the IT folks to go to the employee's office. Wtih a centralized server, upgrading would be done to only a few server computers. Money would be saved because the fast processors would almost always be in use, instead of wasting clock cycles in employees' computers when they are not in full use. Everyone would get the same, powerful computer (although it would probably be possible to set process priorities to favor the executives) and any calls for help could be answered on the phone while the IT guy looks at the employee's desktop on their own computer. It's all about return on investments. Think about it, guys. And remember that the VNC approach works almost flawlessly on Linux, okay on Win95, and (from my experience) barely acceptable using MacOS as the server.
About the free OSes: sure there are a lot of Unices that are free besides Linux (FreeBSD comes immediately to mind), but Linux has the widest user base among them, and the most developers concentrating specifically on it, and a few other good things that make the three contenders really MacOS, Windows, and Linux. I can see a future where Linux is on the top. Microsoft could sell Windows 2000 for free and still have hoardes of money. But they don't, for now.
In "anti-Microsoft-monopoly," I meant for monopoly to apply Microsoft. I'm sorry that I didn't make this clear. Linux is the opposite of a monopoly; there are so many different organizations/companies/individual hackers that drive Linux, and so many different distributions (a new one comes out almost biweekly) that Linux could be thought of (by the DoJ & Associates) as a model of how a non-monopolistic 'organization' can succeed. I can't see into the minds of the big guys in this case, but I can guess that a few of them at the DoJ are thinking that if Linux can succeed as an assortment of different individual entities, why not make Microsoft into the same thing? Imagine if there were twenty different distributions of Windows, each by a different Baby Gates company.
ken
PS - Again, just my opinion. You are permitted to disagree.
Microsoft probably won't want to try to conquer Linux. Sure they have the money and the influence, but Linux, to the DoJ and friends, embodies anti-Microsoft-monoply, and if Microsoft starts messing with Linux, DoJ will jump on them. Microsoft knows this.
Microsoft's high end is the desktop, and that's the last markey Linux will take over. That's a telling point... only after all the other markets have fallen.
GNOME and KDE have made significant strides on making Linux a contender in the desktop environment. In some ways Linux has already overtaken Windows:
Customization: How many window managers are there for Windows? How many are there for Linux?
Remote Apps: What's X designed for? And why does the VNC server work best on Linux? Remote connectivity was a core design element for X.
Stability: People get mad when their apps crash. People notice that for some reason their apps don't crash as much when they run them under Linux. Netscape seems to be an exception, at least for me.
Cost: If two things were equal in every way except that one was $90 and one was free, which one would you buy? And what if it turned out that the free one was better?
I'm sure I missed a few. Sure, Win has its advantages. But they are probably not going to stay for long.
Remember back in the good old days when there was no secret information to exchange and everybody got along on the Internet? You know, the world from which rsh, etc. came from? Well I don't. Today everybody is so concerned about protecting their information that the opposite is happening: they are revealing their information, under penalty of law. The problem with most encryption methods is that it is theoretically possible to get almost anything out of the encrypted text, given a suitable input (there are exceptions). It is not necessary to find out what that input is, but only to prove that it exists. The same thing applies to stenography: given almost any stegotext and a suitable algorithm, you can get any content you want out of a harmless-looking file (of course length matters here; under most circumstances you couldn't extract a 5-page document from a bullet graphic). But in any case, it's your information, whether you legally own it or not, and it shouldn't be somebody else's business to go tampering around with it, especially not to incriminate you!
I have to ask if the people who run the British government are mentally stable, benevolent, and fit for positions of power. Maybe they are, maybe not. Anyway, I'm glad I live in America (for now).
I want to see what's inside of those packets. tcpdump doesn't show the meat of the stuff; just the headers. I don't think there's much stuff going past my node, though.
You can't expect someone to live only off the Internet. That's not how life was meant to be. People were designed for casual personal contact with one another; that's why we have vocal chords. Spending the day locked in and restricting your talking to typing is just plain stupid. You can't reasonably expect to survive that way. Sure, it's possible that you can fufill your physical needs, but your social needs are far from there. There is nothing like standing near somebody and talking through the air with plain old sound waves. Sure, I surf a lot, but I have real-life friends, too. I think one of the focuses of this article was exactly what I just said -- that it's a stupid idea.
As for restricting yourself from the telephone, there are Internet telephone products that can call regular phones, provided you have a sound card and a decent speed connection. But still, why restrict yourself? If you wanted to simulate conducting your entire life at home, the telephone is an important element. This "study" could have been more useful as an analysis of telecommuting, where the telephone is an important element (I said that before, didn't I?).
Final point: Isn't such a thing horrible for your physical health as well? He didn't say anything about exercising. His muscles must have been practically gone by the end of the week. Get out and run/jog/bike/walk/hike/rock-climb/etc. a little. There is no shortage of options. If you're really bored, go travel somewhere and do something there. But don't just stay in your house!
I wish I could watch video in real-time over my net connection. Forgive those with 38.4's (maximum, that is).
This article isn't lame (not LAME either). It's an important message (although not directly spoken within it) that we have too many people spending too much time on the web.
If I met this guy during his week, I'd tell him to get a life, and a couple extra to spare in case he lost that one.
Tell me if you disagree. (Or is asking you for that stupid? Be mature, folks!)
Okay, I read farther down and found something that works like a circular treadmill, but I of course forgot what it was, so you'll just have to go look yourself. But I didn't understand how it worked anyway so the question still stands.
Just after I posted this I saw the EST next to the dates and realized that out-of-time-zone was a possibility. And real time depends on your point of reference. It just so happens that on Earth the generally accepted universal time coincides with GMT. If you go by the sun being directly overhead, my time is the real time and so is yours. But since we can be talking to each other in real time, it's necessary that we establish a common time. Don't brag because it's your time zone. And why am I making such a big deal about it?
Walking doesn't have to be non-vehicular -- but the question is not what is virtually walking, but how you control it. By vehicular I meant the kind of thing that remotely acts like a car (aka truck, bicycle, SUV, airplane, etc.), that you control with some sort of steering mechanism instead of your own body. Although I have to say that using your own body's movements to control something, eliminating the steering wheel layer of abstraction, is kind of neat.
Is it just me, or does this seem a bit far-fetched. Generally if I can't visualize it it's far-fetched. To grow a crystal accurately from nothing other than an inert gas (that the user can breathe?) quickly enough to be a representation of reality and then be able to change the structure of the crystal in real time is a couple thousand km over my head. Now if you figure out how to do such a thing, e-mail me a sample of it because by then you will have figured out how to do that, too. Or am I just not understanding your idea (which is totally possible)?
I don't trust technology enough to allow it to mess around with my neural signals. Many other people in this discussion agree with me (or I agree with them, whatever). A mechanical solution like one you suggested is what is necessary.
And how would these arms you talk about be attached to the body in a way that would not limit movement like bending your joints while still supporting your weight? And it would be a pain to hook up. And I would not want to debug it.
How the heck are you planning to make one of those? If you have a board thing on rollers, the user will eventually walk off of the board. And you can't use the conventional treadmill idea because circles, unlike lines, have a center. Think about it.
Those are interesting suggestions, but the harness idea (mentioned elsewhere as well) has the problem of the feeling of opposing force (also mentioned elsewhere). If your legs are dangling, it's going to feel like your legs are dangling, not like they are touching the ground at every step. It is going to feel almost the same as fake walking. To increase the complexity (but since some outrageously complex ideas have been proposed, it seems that complexity doesn't matter so long as it works) and realism, why not have some line actually be the ground? And to maintain the ground level in a crouch or in a landing after a jump, the harness should be held up by movable wires that could change the vertical position. In general the wires would be relaxed so that the used could "walk" on the "ground", but as the user left the middle of the device, the wires would gently pull him/her back to the middle. Going along with that is that the "ground" be made up of a bunch of movable rod-like things that could approximate uneven terrain by moving up and down. The ground would be covered by stretched plastic to even it out. Then when the user was wandering away from the middle of the device and the wires were pulling him/her back, the rods could tilt slightly back toward the middle so that the natural thing to do would be to go back a little bit. If the plastic was just slippery enough, the pulling back could be accomplished without the user knowing. Of course the whole system would have to be intellegent to allow for the initial acceleration forces felt when beginning to move or stopping. This could be accomplished by delaying the pull-back as much as possible in the necessary circumstances. This seems really complicated, but most everything proposed is, and with today's fast processors, the complexity can be dealt with. This solution also better accounts for response time because most of the actual movement such as jumping, crouching, or walking is left to physics (which is the fastest processor I've ever seen), and the things the computer has to worry about is the compensation, most of which is repetitive (pulling back mostly the same amount while walking) and thus can be optimized with local, less powerful processors. The only really hard things are dealing with sudden motions, such as when you stop running. However, the delayed response time would approximate inertia in this case, adding to the realism.
The only problems I can see with this approach are common to most every approach. First, it would be a pain in the butt to get into because of all the wires and sensors and stuff, so it probably wouldn't work for a public video game arcade kind of situation, and it wouldn't be able to deal with bending over very well. Having a harness could also allow for (but not accurately simulate, nothing proposed can) jumping off of a high ledge in a game where such a thing can't really hurt you (like a lot of them). It would also allow for the simulation of some other thing moving them around, like hitting a wall could cause an abrupt force backwards, and you could approximate the up-and-down forces of, say, a ski lift. It would be enormously flexable.
But it would be (also like most other things) prohibitively expensive.
Come to think about it, this sort of resembles how Star Trek NG did it on the holodecks, execpt it was a force field that pulled them back towards the middle of the simulator instead of a harness and cables. They called it a treadmill force field or something like that, except that it couldn't have really shared that much with a treadmill because treadmills only go in one direction.
Anyway, why bother in the first place? VR is always going to be virtual, and that means that there will have to be some restrictions. Besides, vehicles are more fun anyway; you get to go faster.
Kenneth
PS - What's wrong with the real thing? And what were you doing on Slashdot at 6:42 in the morning? I was sleeping:-|____.
I cannot possibly be expected to read all of these comments.
The makings of some of the worst programming problems in the world in one little web page. I propose that you try these suggestions with Scheme (never heard of it? Try appending a.org):
In Scheme, variables have no type. That can really get fun. You can pass functions as arguments to other functions just as easily as any other data. You don't even have to define it somewhere else and pass a pointer; just put it there.
You can put stuff wherever you want. For example, you can put all of your methods in one big list, and instead of using the function's name, use a reference into the list. As an added bonus, redefine the list somewhere in the middle of the code, deep inside some function. Your functions can be designed to automatically delete themselves when called! (You can do this even if your functions aren't in lists.)
Sort of like #2, you can redefine identifiers such as functions once they are declared. A function could do one thing at one point in your program at one time, and another thing at another time.
Scheme allows your functions to recurse as often as they please. No stack space concerns! That means two things: One, you can have hideously complex algorithms, and two, you can split every operation into tiny little pieces, and scatter these functions all over your code. Have them change their purpose often.
Stuff that gets executed can look a lot like stuff that isn't supposed to get executed. The way for calling a function is (name arg1 arg2...). The way for making a data list is '(thing1 thing2 thing3...) Oops! That isn't a function call, that's just data! Confuses the heck out of people who are not you.
Your code maintainer likely doesn't know Scheme. If your boss asks why you are using this language he's never heard of before (it's a dialect of LISP), you can tell him that Scheme is a functional, properly tail-recusive dynamically-typed data processing language and it is great for parsing complex things and that it is very flexible, and it allows you to write elegant and simple code, all of which are true. Of course, you don't have to write elegant and simple code, do you?
Quasiquoting is a very good way to make things a lot more complicated. Basically it allows you to execute stuff while constructing a list, and put the results into the list. There is a very useful variation, however, and it's called unquote-splicing (whatever that means). If the function that you want to put in returns a list, it strips the parenthesis off of the ends and inserts each element as an element in the list. So your lists can vary widely in length from one time to another, depending on what your functions return.
There are many different implementations of Scheme, and they are substantially different in I/O. So the code that compiled just fine on your computer using one implementation will fail miserably when your code maintainer comes in and tries to execute it on his implementation.
Fun Thing To Do With Scheme Number 10: Make your functions do something different depending on how many arguments they are passed and how many times they have been called. The possibilities are almost endless: a little, innocent-looking loop in your program calls the function about 10 times, and each time it is called, it does a different thing. To discourage the one who plays around with your functions, have it delete your master data list and then undefine itself if someone calls it more than the number of times you know that it should be called. Of course, you could have a secret parameter embedded deep within your arguments that allows you to call it however many times you want.
I think I had better stop it at 10; the fun could go on, but it doesn't. Just another little morsel: Scheme comments are semicolons. Play tricks on people who are used to shell scripts. And there are some important cases where #'s are legal, specifically to quote an atomic (bunch of characters put together that isn't a string) element that includes spaces. Tada! Your comments look like code, and your code looks like comments!
I use Scheme how it was intended to be used: elegantly and flexibly. But there is a lot of room for error--Exploit it!
Kenneth Arnold
PS - Scheme really is a neat language, if you have enough time to learn it. It's a lot like Perl in that it's great for parsing stuff, but it's more flexible, for example, the function rules are greatly relaxed (see #1.)
This is really a reply to several comments, so my choice to put it here is a bit arbitrary.
As a person who is looking at a bunch of web sites every day (i.e., typical end user), I have a couple things to say about what I like in web sites:
Content: I have a 38.4 connection, at best. I HATE waiting for huge images to download! Trust me, there is an alternative to big images. A well-placed, small image that repeats itself often throughout the site if possible (so I don't have to download it again and again) is more like it. No images is often bad. Make a simple title graphic, a "bullet" graphic, and then think very carefully before adding anything else. If you need to add a big image, play around with the resolutions and file formats to minimize the size. Skillful use of common fonts (for Windows users, Verdana comes to mind, put an "alternate" font for those without it, maybe Arial) can provide good visual effects without a very long download. I've seen some very nice stuff done with just a CSS. But, a web page with a whole bunch of black and white going on forever is not very informative because my brain tends to automatically switch into skim mode. Use indentation, horizontal rules, and a table of contents on any page that usually renders longer than one and a half screens. And don't put everything on one page, but don't put too little information on a page to be useful. I guess the word I am searching for is balance.
KISS: The comment on Lynx is right on the money, for a different reason than I think was intended. If you lose significant content or organization (such as an image map without text equivilants), than you are targetting your design incorrectly. Remember the purpose of images. Lynx also cannot take advantage of some newer features. Try to make your page work the best it can in Lynx, and then add snazzy new stuff only if it significantly adds to the informational quality of your site. And Lynx also can have trouble with tables. I personally hate tables. Most browsers, especially the Netscape that comes with RH6, have to wait until the entire table is downloaded before displaying it. When you have a slow connection like me, I sometimes go grab a snack waiting for Slashdot, which is basically one big table, to load. I have heard offhand that Mozilla renders columar table content as it gets it, and that is one reason why I am downloading it now. But don't count on everyone having enough patience to wait for your site.
HTML: (This was not on your list.) The first word of that acronym is Hypertext. Hypertext has links. I like it much better when Hypertext has appropriate links. Place the links over the one or two key words that describe the linked-to page, and put enough in your page that people can freely explore related sites, but not so much that they get overloaded with possibilities. Again, the word is balance.
All code has bugs: Instant question: why use code? If you can avoid it, don't use it. Keep your troubles to the minimum and use static pages where static pages are due. This comment is static, isn't it? Okay, you can use server code, but make sure that it runs fast enough that I am not kept waiting.
I, as an end user, feel that I am devoting too much time to writing this comment. I gotta go.
Kenneth Arnold, web surfer
PS - I guess this took more of the form of a web page design lecture, but I don't have time to go back and change that.
Good luck, future webmaster(s)! Tell me/us the site name when it is done, and we can test how well your servers stand up to the Slashdot Effect.
Thinking back on it, I was once a Linux newbie. But I had a 486 to fool around with. So guess how long it took to recompile that beast. And configuration was a mess! I blew it away, so does anyone know if they had menuconfig back in the 1.x kernels? That really would have helped me a lot if I had known.
Enough reflections. My two computers that run Linux run 2.2.12 and 2.2.13. I use menuconfig. I compile my kernel in 5 minutes. I have fun.
I was a bit distressed to see that the article left out modules. When I got RH6, I eventually got over my party over how great it was in comparison to a 3-4 year-old Slackware distro and got to recompiling my kernel. If they even had modules back in 1.x, I had never needed them because I had so little hardware. But now I have an ethernet card that likes to know its hardware address, a ppa Zip drive that I use so infrequently that it shouldn't remain in my kernel, and a few other things, so I started to think about how I could reduce my kernel size a bit. I saw these things called modules. Okay, got it, something that plugs into something else. Well huh. Why do I care, I asked myself. After parsing the kernel docs about modules a couple of times and then recompiling about ten times, not to mention crashing my poor computer to the point where the three-finger kick (that fourtunately tells init to bring the system down nicely) to fix it up. I finally figured modprobe out, and kerneld/kmod, and conf.modules, and/lib/modules/`uname -r`/ and depmod and all that stuff, but not without a lot of headaches. If this had not been during my summer break I would have forgot about the whole thing. In short, I could have used an introduction to modules like the article introduces kernel compiling.
Kernel compilation is one of the things that, at least for me, really separates Linux from Windows. In Windows, you get a kernel and a whole bunch of drivers that sort of plug into it but generally stay in their own everything-space and have more often than not been one factor causing it to crash. In Linux, I get complete control over exactly what goes into the most fundamental part of my computer, and I know what gets "plugged in", why, and where the stuff goes. I can put stuff in, take it out, tweak parameters with it without a restart, and, as usual, have fun. Linux gives me just the right mix of power to make it do whatever I want, to the point of changing how the core of my OS works, but at the same time the power to forget about it if I'm not in the mood. Most other stuff with Linux is like that. If you want to mess with it, you can, and you can do most anything you want with it, but if you don't want to mess with it, and you have a relatively normal configuration, you can just forget about it and get to work. Speaking of work, when are they going to have a Microsoft Word for Linux? StarOffice just isn't cutting it for me, although after working with emacs for a while, I appreciate its advanced features. Never mind; I see an ad for Applixware up at the top of my Netscape window. That's another thing that newbies could use: a good office suite bundled with the distribution. "They" should "standardize" the office suites for Linux and get one good, familiar, and cool office suite that is integrated as well as MS Office, retain full file-format compatibility with MS Office applications, and bundle it with every distribution pre-"installed". (I was puzzled to see that StarOffice actaully had an installer--RPMs work just fine for me. I know, not everybody has rpm.) You don't want to hear me talk about what else in Linux should be standardized. Believe me.
This is getting long, and people don't like long posts very much. Replace this sentence with a closing remark.
Kenneth Arnold
PS - I have homework to do. Don't make me feel like taking time away from it to write this was a waste of time.
PPS - Don't yell at me, but I'm getting really annoyed at Netscape and Slashdot. Netscape keeps crashing (twice today), and Slashdot has these huge tables that load up even if I don't want to see them. But I've found that the reduced mode reduces it too much for my tastes. Could we strike a balance?
There could certainly be an interest for developing DSP products for PCs, in such areas as sound cards and graphics boards (although I'm not completely sure about the usefulness of the second one. A DSP that is already very connected with personal computers is the so-called Velocity Engine on the G4. (I'm getting this information from someone else, so I may be wrong.) It employs advanced vector processing to make a very fast and efficient SIMD (single instruction multiple data) CPU design. I also mentioned graphics cards. A digital signal need not be a sound. It may be possible (I haven't researched this) to implement very fast image transformations using DSPs. Given a suitable implementation, a DSP could accomplish many things a CPU of a computer could; the Velocity Engine proves this.
The text editor / web browser stuff was supposed to be funny but obviously wasn't. What was funny was a web browser for a TI-8(something) calculator. You had to hook it up to a laptop computer to get the Internet connection, so why not just use the laptop computer?
I have heard a lot about Fourier transforms, but I don't really understand them (that is where I really am confused). I know that they are used for COFDM (see my Digital Television Transmission Standards article from last Sunday) to create thousands of individual carriers without having thousands of different oscillators and the like. Could you provide a link to somewhere that explains this technology (since you seem to know about this)? If it's a good site, it's a good chance that Nerds would be interested in it.
I have gotten tired of telling Anonymous Cowards to get a Slashdot user account. You can do that easily from users.pl on the Slashdot main site. Then you can do all sorts of cool stuff like customize your main page and make your comments start at Score 1. I've seen plenty of good AC comments down at the bottom of the comments because they had low scores.
I'm not a tech person, but I did understand a good portion of what I read about the so-called Velocity Engine for the G4. The vector processing stuff is what gives that processor its power. So a vector engine is a DSP? Sounds reasonable. If someone could get a definition (reply to some other comment; this one is too deep) in here, maybe that could help relative newbies like me get a clue what is going on, particularly (1) What a DSP is, exactly, and (2) Why should I care?. Then I will be happy, and some other people will be happy to learn a little interesting tidbit of knowledge
Ken
PS - Will an Intel DSP make my modem faster? Seriously, will they be any better then their competition? Am I asking too many questions? Am I right to say yes to my own question?
Slashdot Mirror
I just read somebody else's comment about click-through counters. Post any arguments about that as replys to that comment. (It's a reply to one of the top 5 posts, subject is descriptive. You find it.)
Ken
I propose that we launch a massive Slashdot statistics program. It would involve a significant amount of invasion of "privacy", i.e., almost every link off of /. would be logged, but we could get a lot of raw data to look at in interesting ways. Here are a few of my ideas:
There are more! Let someone else suggest them, though. I think that analyzing the raw data gained by these approaches would be a good statistics project for any upper high school or college student. And hey, I like Slashdot, and if I get credit for it, even better! (Well, it would be mostly raw data, not the real site, but still...)
Ken
PS - I would do such a project. Anybody who wouldn't?
Linux is so similar to other Unices in the development environment that you could argue that everything targetted specifically for Linux (except for some of the low-level stuff, and very few people are working on that) is really targetted for the entire Unix community. You can't see the exact motives for writing the software. But what happens is that most of the people who actually use the software happen to be running Linux. How Linux got to that point is beyond me because I did not personally experience it (Hey, what can you expect from a 14-year-old?) but I think it has something to do with the driver support, being free, and being fun, but that's just what I glean from other people. I just can't accept a 99.98% figure without a little contesting.
Business computing is the kind of computing that makes money. There are a lot of people who spend their spare time messing around with their computers and trying OSes like Linux, but what do they do to get their money? They work. And what do they use at work to (hopefully) boost their productivity? Computers. And there are a lot of people who use computers at work who couldn't care less about them once they walk out to their car. If you want to address the largest number of users, you have to get into the field of business computing. There are exceptions to everything.
Ken is done arguing.
PS - What goes up and never comes down? Tom C.'s karma! Are you some super-nerd? You have something to say about practically every news story in the past few weeks at least, if not the last year or so. How do you get a karma of 74 (last time I checked)? Well, for someone who wrote a book on Perl, that's sort of a given, but...
What's wrong with a Microsoft accent? I think that Linux will eventually become everything that's good about Microsoft's OSes and more without the bad parts of MS.
True, nothing I list is really inherent to the kernel, which is what Linux technically is. I was targetting the specifics of Linux on the desktop. As I said, I'm sure that I missed a few things. Linux on the desktop is what the average Joe Computer User is going to see, not the kernel and stuff like that. If Linux is to succeed, Joe has gotta be happy with it. It seemed like the remote apps thing didn't hit the right target. Remote applications could be the future of business computing. Each user could have even a spare 386 running the SVGAlib version of the VNC viewer connected by a decently fast network to a few powerful computers managed by people who really know what to do with really powerful computers. The question is return on investment. It is not very difficult to add another user with VNC server on the main server and put a spare 386 in their office, but in today's business world where most everyone has their own computer, installing a new computer means going through complicated setups and other time-wasting things and then having to upgrade it when users start using an app that requires more speed. And of course there are calls for help and stuff that require the IT folks to go to the employee's office. Wtih a centralized server, upgrading would be done to only a few server computers. Money would be saved because the fast processors would almost always be in use, instead of wasting clock cycles in employees' computers when they are not in full use. Everyone would get the same, powerful computer (although it would probably be possible to set process priorities to favor the executives) and any calls for help could be answered on the phone while the IT guy looks at the employee's desktop on their own computer. It's all about return on investments. Think about it, guys. And remember that the VNC approach works almost flawlessly on Linux, okay on Win95, and (from my experience) barely acceptable using MacOS as the server.
About the free OSes: sure there are a lot of Unices that are free besides Linux (FreeBSD comes immediately to mind), but Linux has the widest user base among them, and the most developers concentrating specifically on it, and a few other good things that make the three contenders really MacOS, Windows, and Linux. I can see a future where Linux is on the top. Microsoft could sell Windows 2000 for free and still have hoardes of money. But they don't, for now.
Gotta go, Ken
In "anti-Microsoft-monopoly," I meant for monopoly to apply Microsoft. I'm sorry that I didn't make this clear. Linux is the opposite of a monopoly; there are so many different organizations/companies/individual hackers that drive Linux, and so many different distributions (a new one comes out almost biweekly) that Linux could be thought of (by the DoJ & Associates) as a model of how a non-monopolistic 'organization' can succeed. I can't see into the minds of the big guys in this case, but I can guess that a few of them at the DoJ are thinking that if Linux can succeed as an assortment of different individual entities, why not make Microsoft into the same thing? Imagine if there were twenty different distributions of Windows, each by a different Baby Gates company.
ken
PS - Again, just my opinion. You are permitted to disagree.
Microsoft probably won't want to try to conquer Linux. Sure they have the money and the influence, but Linux, to the DoJ and friends, embodies anti-Microsoft-monoply, and if Microsoft starts messing with Linux, DoJ will jump on them. Microsoft knows this.
GNOME and KDE have made significant strides on making Linux a contender in the desktop environment. In some ways Linux has already overtaken Windows:
I'm sure I missed a few. Sure, Win has its advantages. But they are probably not going to stay for long.
Ken
Remember back in the good old days when there was no secret information to exchange and everybody got along on the Internet? You know, the world from which rsh, etc. came from? Well I don't. Today everybody is so concerned about protecting their information that the opposite is happening: they are revealing their information, under penalty of law. The problem with most encryption methods is that it is theoretically possible to get almost anything out of the encrypted text, given a suitable input (there are exceptions). It is not necessary to find out what that input is, but only to prove that it exists. The same thing applies to stenography: given almost any stegotext and a suitable algorithm, you can get any content you want out of a harmless-looking file (of course length matters here; under most circumstances you couldn't extract a 5-page document from a bullet graphic). But in any case, it's your information, whether you legally own it or not, and it shouldn't be somebody else's business to go tampering around with it, especially not to incriminate you!
I have to ask if the people who run the British government are mentally stable, benevolent, and fit for positions of power. Maybe they are, maybe not. Anyway, I'm glad I live in America (for now).
Ken
PS - That's just my opinion.
Get a kernel that fits in 1k. But first, figure out how to compile for it.
What useless platforms has Linux been ported to anyway?
We all know what the author means. Why waste your time arguing about it? I know what it is but I'm not telling.
ken
Good luck fitting it into 1k! The networking code, whatever that would mean for such a computer, would alone take up at least that much.
I want to see what's inside of those packets. tcpdump doesn't show the meat of the stuff; just the headers. I don't think there's much stuff going past my node, though.
You can't expect someone to live only off the Internet. That's not how life was meant to be. People were designed for casual personal contact with one another; that's why we have vocal chords. Spending the day locked in and restricting your talking to typing is just plain stupid. You can't reasonably expect to survive that way. Sure, it's possible that you can fufill your physical needs, but your social needs are far from there. There is nothing like standing near somebody and talking through the air with plain old sound waves. Sure, I surf a lot, but I have real-life friends, too. I think one of the focuses of this article was exactly what I just said -- that it's a stupid idea.
As for restricting yourself from the telephone, there are Internet telephone products that can call regular phones, provided you have a sound card and a decent speed connection. But still, why restrict yourself? If you wanted to simulate conducting your entire life at home, the telephone is an important element. This "study" could have been more useful as an analysis of telecommuting, where the telephone is an important element (I said that before, didn't I?).
Final point: Isn't such a thing horrible for your physical health as well? He didn't say anything about exercising. His muscles must have been practically gone by the end of the week. Get out and run/jog/bike/walk/hike/rock-climb/etc. a little. There is no shortage of options. If you're really bored, go travel somewhere and do something there. But don't just stay in your house!
I wish I could watch video in real-time over my net connection. Forgive those with 38.4's (maximum, that is).
This article isn't lame (not LAME either). It's an important message (although not directly spoken within it) that we have too many people spending too much time on the web.
If I met this guy during his week, I'd tell him to get a life, and a couple extra to spare in case he lost that one.
Tell me if you disagree. (Or is asking you for that stupid? Be mature, folks!)
Ken
Okay, I read farther down and found something that works like a circular treadmill, but I of course forgot what it was, so you'll just have to go look yourself. But I didn't understand how it worked anyway so the question still stands.
(K(en(neth( A(rnold)))))
Just after I posted this I saw the EST next to the dates and realized that out-of-time-zone was a possibility. And real time depends on your point of reference. It just so happens that on Earth the generally accepted universal time coincides with GMT. If you go by the sun being directly overhead, my time is the real time and so is yours. But since we can be talking to each other in real time, it's necessary that we establish a common time. Don't brag because it's your time zone. And why am I making such a big deal about it?
Walking doesn't have to be non-vehicular -- but the question is not what is virtually walking, but how you control it. By vehicular I meant the kind of thing that remotely acts like a car (aka truck, bicycle, SUV, airplane, etc.), that you control with some sort of steering mechanism instead of your own body. Although I have to say that using your own body's movements to control something, eliminating the steering wheel layer of abstraction, is kind of neat.
Ken
Is it just me, or does this seem a bit far-fetched. Generally if I can't visualize it it's far-fetched. To grow a crystal accurately from nothing other than an inert gas (that the user can breathe?) quickly enough to be a representation of reality and then be able to change the structure of the crystal in real time is a couple thousand km over my head. Now if you figure out how to do such a thing, e-mail me a sample of it because by then you will have figured out how to do that, too. Or am I just not understanding your idea (which is totally possible)?
[adverb form of confused], Ken
I don't trust technology enough to allow it to mess around with my neural signals. Many other people in this discussion agree with me (or I agree with them, whatever). A mechanical solution like one you suggested is what is necessary.
And how would these arms you talk about be attached to the body in a way that would not limit movement like bending your joints while still supporting your weight? And it would be a pain to hook up. And I would not want to debug it.
Ken
You forgot one: blowing up the world! I've done that.
Of course, there is the problem with realisticly simulating going to the bathroom... oops!
cya, Ken
How the heck are you planning to make one of those? If you have a board thing on rollers, the user will eventually walk off of the board. And you can't use the conventional treadmill idea because circles, unlike lines, have a center. Think about it.
Ken
Peak Performance
VNSIII
Whole Body Kinesthetic Displays
Ken
Those are interesting suggestions, but the harness idea (mentioned elsewhere as well) has the problem of the feeling of opposing force (also mentioned elsewhere). If your legs are dangling, it's going to feel like your legs are dangling, not like they are touching the ground at every step. It is going to feel almost the same as fake walking. To increase the complexity (but since some outrageously complex ideas have been proposed, it seems that complexity doesn't matter so long as it works) and realism, why not have some line actually be the ground? And to maintain the ground level in a crouch or in a landing after a jump, the harness should be held up by movable wires that could change the vertical position. In general the wires would be relaxed so that the used could "walk" on the "ground", but as the user left the middle of the device, the wires would gently pull him/her back to the middle. Going along with that is that the "ground" be made up of a bunch of movable rod-like things that could approximate uneven terrain by moving up and down. The ground would be covered by stretched plastic to even it out. Then when the user was wandering away from the middle of the device and the wires were pulling him/her back, the rods could tilt slightly back toward the middle so that the natural thing to do would be to go back a little bit. If the plastic was just slippery enough, the pulling back could be accomplished without the user knowing. Of course the whole system would have to be intellegent to allow for the initial acceleration forces felt when beginning to move or stopping. This could be accomplished by delaying the pull-back as much as possible in the necessary circumstances. This seems really complicated, but most everything proposed is, and with today's fast processors, the complexity can be dealt with. This solution also better accounts for response time because most of the actual movement such as jumping, crouching, or walking is left to physics (which is the fastest processor I've ever seen), and the things the computer has to worry about is the compensation, most of which is repetitive (pulling back mostly the same amount while walking) and thus can be optimized with local, less powerful processors. The only really hard things are dealing with sudden motions, such as when you stop running. However, the delayed response time would approximate inertia in this case, adding to the realism.
The only problems I can see with this approach are common to most every approach. First, it would be a pain in the butt to get into because of all the wires and sensors and stuff, so it probably wouldn't work for a public video game arcade kind of situation, and it wouldn't be able to deal with bending over very well. Having a harness could also allow for (but not accurately simulate, nothing proposed can) jumping off of a high ledge in a game where such a thing can't really hurt you (like a lot of them). It would also allow for the simulation of some other thing moving them around, like hitting a wall could cause an abrupt force backwards, and you could approximate the up-and-down forces of, say, a ski lift. It would be enormously flexable.
But it would be (also like most other things) prohibitively expensive.
Come to think about it, this sort of resembles how Star Trek NG did it on the holodecks, execpt it was a force field that pulled them back towards the middle of the simulator instead of a harness and cables. They called it a treadmill force field or something like that, except that it couldn't have really shared that much with a treadmill because treadmills only go in one direction.
Anyway, why bother in the first place? VR is always going to be virtual, and that means that there will have to be some restrictions. Besides, vehicles are more fun anyway; you get to go faster.
Kenneth
PS - What's wrong with the real thing? And what were you doing on Slashdot at 6:42 in the morning? I was sleeping :-|____.
I cannot possibly be expected to read all of these comments.
The makings of some of the worst programming problems in the world in one little web page. I propose that you try these suggestions with Scheme (never heard of it? Try appending a .org):
I think I had better stop it at 10; the fun could go on, but it doesn't. Just another little morsel: Scheme comments are semicolons. Play tricks on people who are used to shell scripts. And there are some important cases where #'s are legal, specifically to quote an atomic (bunch of characters put together that isn't a string) element that includes spaces. Tada! Your comments look like code, and your code looks like comments!
I use Scheme how it was intended to be used: elegantly and flexibly. But there is a lot of room for error--Exploit it!
Kenneth Arnold
PS - Scheme really is a neat language, if you have enough time to learn it. It's a lot like Perl in that it's great for parsing stuff, but it's more flexible, for example, the function rules are greatly relaxed (see #1.)
This is really a reply to several comments, so my choice to put it here is a bit arbitrary.
As a person who is looking at a bunch of web sites every day (i.e., typical end user), I have a couple things to say about what I like in web sites:
I, as an end user, feel that I am devoting too much time to writing this comment. I gotta go.
Kenneth Arnold, web surfer
PS - I guess this took more of the form of a web page design lecture, but I don't have time to go back and change that.
Good luck, future webmaster(s)! Tell me/us the site name when it is done, and we can test how well your servers stand up to the Slashdot Effect.
Thinking back on it, I was once a Linux newbie. But I had a 486 to fool around with. So guess how long it took to recompile that beast. And configuration was a mess! I blew it away, so does anyone know if they had menuconfig back in the 1.x kernels? That really would have helped me a lot if I had known.
Enough reflections. My two computers that run Linux run 2.2.12 and 2.2.13. I use menuconfig. I compile my kernel in 5 minutes. I have fun.
I was a bit distressed to see that the article left out modules. When I got RH6, I eventually got over my party over how great it was in comparison to a 3-4 year-old Slackware distro and got to recompiling my kernel. If they even had modules back in 1.x, I had never needed them because I had so little hardware. But now I have an ethernet card that likes to know its hardware address, a ppa Zip drive that I use so infrequently that it shouldn't remain in my kernel, and a few other things, so I started to think about how I could reduce my kernel size a bit. I saw these things called modules. Okay, got it, something that plugs into something else. Well huh. Why do I care, I asked myself. After parsing the kernel docs about modules a couple of times and then recompiling about ten times, not to mention crashing my poor computer to the point where the three-finger kick (that fourtunately tells init to bring the system down nicely) to fix it up. I finally figured modprobe out, and kerneld/kmod, and conf.modules, and /lib/modules/`uname -r`/ and depmod and all that stuff, but not without a lot of headaches. If this had not been during my summer break I would have forgot about the whole thing. In short, I could have used an introduction to modules like the article introduces kernel compiling.
Kernel compilation is one of the things that, at least for me, really separates Linux from Windows. In Windows, you get a kernel and a whole bunch of drivers that sort of plug into it but generally stay in their own everything-space and have more often than not been one factor causing it to crash. In Linux, I get complete control over exactly what goes into the most fundamental part of my computer, and I know what gets "plugged in", why, and where the stuff goes. I can put stuff in, take it out, tweak parameters with it without a restart, and, as usual, have fun. Linux gives me just the right mix of power to make it do whatever I want, to the point of changing how the core of my OS works, but at the same time the power to forget about it if I'm not in the mood. Most other stuff with Linux is like that. If you want to mess with it, you can, and you can do most anything you want with it, but if you don't want to mess with it, and you have a relatively normal configuration, you can just forget about it and get to work. Speaking of work, when are they going to have a Microsoft Word for Linux? StarOffice just isn't cutting it for me, although after working with emacs for a while, I appreciate its advanced features. Never mind; I see an ad for Applixware up at the top of my Netscape window. That's another thing that newbies could use: a good office suite bundled with the distribution. "They" should "standardize" the office suites for Linux and get one good, familiar, and cool office suite that is integrated as well as MS Office, retain full file-format compatibility with MS Office applications, and bundle it with every distribution pre-"installed". (I was puzzled to see that StarOffice actaully had an installer--RPMs work just fine for me. I know, not everybody has rpm.) You don't want to hear me talk about what else in Linux should be standardized. Believe me.
This is getting long, and people don't like long posts very much. Replace this sentence with a closing remark.
Kenneth Arnold
PS - I have homework to do. Don't make me feel like taking time away from it to write this was a waste of time.
PPS - Don't yell at me, but I'm getting really annoyed at Netscape and Slashdot. Netscape keeps crashing (twice today), and Slashdot has these huge tables that load up even if I don't want to see them. But I've found that the reduced mode reduces it too much for my tastes. Could we strike a balance?
To try to save myself from sounding stupid:
There could certainly be an interest for developing DSP products for PCs, in such areas as sound cards and graphics boards (although I'm not completely sure about the usefulness of the second one. A DSP that is already very connected with personal computers is the so-called Velocity Engine on the G4. (I'm getting this information from someone else, so I may be wrong.) It employs advanced vector processing to make a very fast and efficient SIMD (single instruction multiple data) CPU design. I also mentioned graphics cards. A digital signal need not be a sound. It may be possible (I haven't researched this) to implement very fast image transformations using DSPs. Given a suitable implementation, a DSP could accomplish many things a CPU of a computer could; the Velocity Engine proves this.
The text editor / web browser stuff was supposed to be funny but obviously wasn't. What was funny was a web browser for a TI-8(something) calculator. You had to hook it up to a laptop computer to get the Internet connection, so why not just use the laptop computer?
I have heard a lot about Fourier transforms, but I don't really understand them (that is where I really am confused). I know that they are used for COFDM (see my Digital Television Transmission Standards article from last Sunday) to create thousands of individual carriers without having thousands of different oscillators and the like. Could you provide a link to somewhere that explains this technology (since you seem to know about this)? If it's a good site, it's a good chance that Nerds would be interested in it.
I have gotten tired of telling Anonymous Cowards to get a Slashdot user account. You can do that easily from users.pl on the Slashdot main site. Then you can do all sorts of cool stuff like customize your main page and make your comments start at Score 1. I've seen plenty of good AC comments down at the bottom of the comments because they had low scores.
See 'ya later, digital friends: Kenneth
I'm not a tech person, but I did understand a good portion of what I read about the so-called Velocity Engine for the G4. The vector processing stuff is what gives that processor its power. So a vector engine is a DSP? Sounds reasonable. If someone could get a definition (reply to some other comment; this one is too deep) in here, maybe that could help relative newbies like me get a clue what is going on, particularly (1) What a DSP is, exactly, and (2) Why should I care?. Then I will be happy, and some other people will be happy to learn a little interesting tidbit of knowledge
Ken
PS - Will an Intel DSP make my modem faster? Seriously, will they be any better then their competition? Am I asking too many questions? Am I right to say yes to my own question?