Just wait until open source developers get deluged with emails from the millions of random people that AOL could get signed up for a program like this. You'll see some major usability feats real fast.
...conversely, the developers would also get deluged with E-mail, the entire text of which is like "I cant get my intrnet thing to werk help me pleeze" and they'd be so annoyed and/or behind trying to communicate with people who can't put together a coherent sentence that nothing would get done. Corel's developers might be forced to make the box idiot-proof, but nature would just build a better idiot, and the box would become less useful in the process.
From the article, it seems that these boxes will be used for Web browsing and not much else. Is Mozilla's UI that horrible, that it needs massive amounts of feedback? Isn't the project already getting massive amounts of feedback? Anyway, it will be interesting to see what comes of this.
If this distro is faster, I am going to use it! I think right now the bottleneck of my system is in I/O, and especially disk access. Upgrading to ATA66 certainly would improve things.
Yeah. Right. It's been said before, and in many places, but most IDE drives currently on the market cannot sustain a transfer rate of more than 12M/sec due to limitations in the head/disk assemblies. That won't even saturate a UDMA/33 channel. You might notice a *slight* performance increase at the very beginning of a disk operation, when the drive uses its onboard cache, but for big jobs, forget it. UDMA/66 is really an investment in the future, since in a few years, hard drives might be able to support that kind of I/O speed. Of course, by then, marketing will have us all buying UDMA/256....
It is good that Abit is supporting their hardware in Linux. Why a distro, though? Why not just cut a deal with RedHat/SuSE/Caldera so those distributors get whatever hackarounds Abit came up with and Abit gets a "Works great with $DISTRO" endorsement?
But reading over the posts here, it's clear that in general, people want it both ways.
Everyone should use Linux because it's cool/stable/free/fast/whatever.
Linux is *ours*, and we don't want stupid apps like AOL on it.
The reason (well, one of the reasons) Microsoft has been successful is because they studied what consumers want and delivered it. To some degree, anyways.
Amen to that. I think the problem here is fundamentally related to the way a lot of development for Linux gets done. Many projects get started when a programmer wants to scratch his/her personal itch, and programmers are by nature very different from ordinary users. They want intelligent, general-purpose tools and aren't scared off by arcane interfaces. As a result, the market for AOL-type software in the Linux community is barely being served at all, though it seems like RedHat/Caldera/SuSE/et al. are trying with graphical installations, included support databases and FAQs, printtool configuration programs.... The great strength and weakness of Linux is that there's really no central authority who can tell everyone How To Do Things or What Software To Develop/Sell.
As long as that attitude prevails, Linux is going to have a hard time with nontechnical people.
IMHO, the main reason Linux isn't taking off more than it already is is hardware support. WinXX detects and uses all the hardware in your system right out of the box because--surprise--the manufacturers either wrote their own WinXX drivers or gave MS the specs so drivers could be written. Linux often doesn't have that advantage, so people say, "What good is it if I can't use my FooCard 7000?" And sometimes I can't blame them for that...
As for the flaming thing, if people in the Linux community can't take constructive criticism, they should try to grow up a bit. One well-reasoned, polite defense or critique is worth a thousand flames, if only because it's more likely to be read. It's easy to call someone a f!@#ing piece of s*-t just because you don't agree with his/her statements, but that creates smoke, noise, bad feelings, and nothing else. You'd think people would learn.
Please check out Xmame for the finest in older video game console emulation. If you want to get games, you can probably find them at this place but I assume no liability for any legal repercussions that may arise.
As for why people like them, I guess that it's because you have to put a lot of effort into learning the moves for all the characters, and then you can beat the hell out of your friends and do really cool/difficult things with said characters. Also, they provide more interaction than an RPG--it's just more fun to beat up your friends than beat up the computer.
But really, this looks like more of the same. Ever since Street Fighter 2, it seems like 90% of the games that are released are just the same bloody idea run into the ground over and over again. They change the characters involved, they tweak the controls, they do it in pseudo 3D a la Tekken 3, War Gods, and Virtua Fighter... but the idea's the same and too often the gameplay's repetitive and bloody annoying.
No wonder I end up spending more money on pinball these days.
"MS-DOS... can be apprehended by consciousness"? It fits, I guess, because when you're really zonked, you can only concentrate on one thing at a time and sometimes even that's hard... a case of the human brain reverting to "real mode", sort of.
However, it would not surprise me to learn that the human mind actually ran a hodgepodge of Perl, TECO macros, Lisp, and COBOL (the last being in the R-complex, which pop culture has referred to as the "dinosaur brain".)
I actually found that the article, while it was interesting, was somewhat incoherent... and now, 15 minutes after reading it, I can't remember the author's main point or any insights/cool things that were in the text. Could it be that the author was trying to produce the textual equivalent of a couple of bong hits?:-)
Usenet death? Some are actively killing it...
on
Is Usenet Dying?
·
· Score: 2
Some places that used to help people in getting started on Usenet have stopped that entirely. The U of M used to hand out flyers telling new students how to read/post news using BBEdit, but the flyers (and BBEdit, for that matter) have disappeared from the U's computer labs. To read news at a public station here, you have to telnet to a Solaris box and use tin, or use Deja. For whatever reasons, the U has not chosen to publicize Usenet and/or make it easy for newbies to use.
Of course, it seems like 99% of the users don't even know that Usenet exists. Most of the 1% who do know that it exists aren't interested for various reasons, and prefer IRC because of its immediacy.
This isn't really a good thing, because as others have pointed out previously, Usenet has some things going for it that neither IRC nor Web-based discussion boards have. Unlike IRC, it's asynchronous and persistent. Unlike Web-based boards, it's both centralized and distributed. It's centralized in this way: If you have a question on hardware compatability for Linux, you can easily figure out that you should go to news:comp.os.linux.hardware. On the Web, there's no one place you could go, and you'll spend a lot of time chasing pointers. And of course, it's distributed over N local newsservers, making it that much quicker to d/l your alt.punk fix. And unlike Web boards, the protocol is standardized and the interface is customizable.
If the text-only nature of many Usenet discussion boards scares people off, good. If people can't feel comfortable communicating using plain text, they should probably stay away from computers altogether until someone comes up with a telepathic user interface....
But I suspect that SDMI is here to stay. Ask yourself who you know that is using mp3s... it's mostly geeks like yourself right?
I work as a helldesk minion in a university computer lab. From my unscientific method of walking around looking at what people are doing, 50% of the users are downloading/playing mp3s. Since this particular lab is associated with the Business School, most of these users are not geeks. They are future Suits.
Whether this is good or bad is probably something for Jon Katz to discuss:-) but it's happening. Even the thickest luser can understand that mp3s are (potentially) free beer, not to mention familiar and well-tested and playable almost everywhere.
but I suspect that mp3 will stay fringe
...when every university in the USA has high-speed Net access and loads of music-oriented kids with plenty of spare time, mp3s tend to be everywhere in those universities. And since elementary and high schools in the USA are getting connected as well, anyone passing through any higher educational institution will be exposed to the format and its possibilities. If you think this is "fringe", then I want some of whatever you're smoking.
I suspect [SDMI] will be very convienent to buy on-line.
Yet Napster et. al. have made it very convenient to find free (beer) mp3s on-line, and these search engines/filesharing programs will only get easier to use. Hmm, do I pay $1 for this Beatles song, or spend 5 minutes finding where I can get it for free? 'Tis a no-brainer.
Dell's laptop quality certainly isn't the best that it could be. The U of M Business School had a special deal with Dell where students could buy a laptop with various pieces of software preinstalled. All the Win9x Control Panel stuff was also set up so that students could plug in an Ethernet card and cable, surf, and print to the U of M's networked printers without having to configure anything. About 150 students took advantage of this offer.
First the laptops arrived several weeks behind schedule, then about 20% of them were missing a relatively important piece of software that's used locally, then 4 or 5 of them were apparently set to 800x600 resolution at the factory. (They all had 1024x768 LCD screens, so those looked extraordinarily bad.) I heard a fair number of horror stories from students who dealt with Dell's Friendly Helpful Tech Support, too.
Remember, this was using an OS that Dell is experienced with and knows well. Maybe they've gotten their act together in the intervening 5 months...
Well, it certainly looks pretty, and they talk a good game... but who knows how many of these innovations will actually make the computer more useful? The "genie" effect certainly looks nice... the first time you see it. I think that particular effect, like the infamous dancing paperclip, will be one of those things people think is nifty at first, then later on love to hate/want to turn off.
Likewise for the shrinking/expanding icons in the dock... which brings up another problem: What about desktop space? People are getting more and more used to having $LARGENUM applications open all the time, and switching among them and/or grouping them logically can be a royal pain in both the standard WinXX and MacOS desktops as of now. Most WMs for Un*x do the Right Thing and have virtual desktops, so why doesn't Apple get with the program? Sure, Macs have huge monitors available and can run multiple monitors easily... but Joe User buying his first iMac probably can't afford that.
Apple could even get nasty and say, "This new, totally unavailable-anywhere-else 'Virtual Super SwitchDesk' functionality is almost as good as having 2 or more monitors connected to your iMac. " New checklist feature + increased hardware sales, eh?
Apple shot themselves in the foot when they made floppies an add-on. There are a freaking TON of older computers out there where the best and quickest means of data sharing is via 3.5" floppies. Maybe everything around your home and office has 10/100baseT, but that's definitely not how it is everywhere. Floppies will be around for another 3-4 years if not longer... remember, the IBM PC and its progeny are still around because A) it was easy to port old CP/M stuff to the PC B) all the new iterations of the PC are fully backwards-combatible. (sic)
There are a number of people out there who have no sense of smell, or a greatly reduced sense of smell.[0] Enter "anosmia" into a search engine and see what you find there. If people actually developed this tech. and used it extensively, could the National Anosmic Foundation raise a big stink[1] in much the same way as some blind people did when they found AOL's sites were impossible to view within a text-only browser?
Anyway, I think a better scent for MS's web page would be ethyl mercaptan. CH3-CH2SH, patently artificial, listed in the Guinness Book as being among the foulest odors known to humans. Penetrating and lasts a long time, too--skunks use mercaptans in their personal defense musk.
[0] Yep, I'm one of them. [1] Sorry--couldn't resist.
wireless LANs that allow you to be anywhere on campus with a laptop and hook up to the network - an edge for schools that don't have the coveted Ivy status.
The University of Michigan, who you'd think would be way out in front with a campus-wide wireless LAN, has very spotty coverage and a couple of different wireless networks. There's one network that extends over the space of 1.5 square blocks. They're probably not going to make the coverage area larger because the transceivers are expensive. Max speed I ever observed on this wireless LAN was 400 Kbits/second--nice, but 3x slower than people would get on a bad day with a standard 10bT hooked into a nearly-ubiquitous RJ-45 port.
Security might also be a concern with a wireless net. Wireless sets should have encryption built in and done in hardware, but that adds cost. And when a university sets up a large wireless net, a $10 difference in the price of each transceiver can make a big difference. Better hope those smart CS kids aren't feeling antisocial and h4x0rish...
The future is most definitely wireless; the future is also Not Here Yet. Give it a couple of years...
...so why the announcement that Gateway would start selling systems with AMD processors? I ordered a computer from them in April 1999, specifying a Celeron 400. They called me, asking if a K6-2 400 would be OK instead. I said "Sure."
How was I to know the Micro-Star MS-5185 motherboard that the system was built on was flaky as hell? (Micro-Star doesn't even acknowledge that model's existence on their site, BTW.) Confronting Gateway about the problem resulted in them saying, "It's due to that 'Linux' thing you're running." I ended up replacing nearly half the system components. The original processor is still cranking along, though.
Gateway will probably do well selling systems with AMD processors. They just need to make sure the other components of their systems aren't wretched crap.
However, my mpeg_play and xanim seem to get something related to "interlacing" wrong.
Odd, since I used mpeg_play to verify that the stupid thing got encoded right. I have a very recent version taken from freshmeat, and also called the thing with -dither color and -quality on. *shrug*
The next poster was right about mtv. though it's nagware and closed, I might have to shell out its fee to watch video clips...
Heck, I've been thinking about this for a while. I am currently rendering a small, stupid 3.5 second animation that does the "freeze-n-pan" thing used in the Matrix, all in POVray. After about 3:10 EST, you can go here and see the result. The source code will all be available here so people can see how I did it. I'm sure all the raytracing folks out there know exactly how to do this, but if non-raytracing folks are curious....
The solution that has been taken by most people, including many posters to/., is to hide themselves and hope the newbies go away. With the current situation that has caused the newbies to become what they are, and the amazing pre-background that is not possessed by these newbies that is required with the majority of current documentation and friendly tech support, this solution is in fact a solution, it will make the newbies go away, unfortunately it will also be a great detriment to the operating system itself
Hmm. I don't know where you've been lurking, but on the Linux newsgroups, there are an enormous number of newbies looking for help, and a smaller population of people trying valiantly to help them. I personally visit comp.os.linux.hardware and comp.os.linux.misc every day, looking for questions that I know the answers to and trying to help people even if they don't know what button 2 on their mouse is for. And I know I'm not alone.
I think part of the problem is that Linux/the "geek society" is more of a meritocracy than anything. Anyone can be an expert; all it takes is time and effort. This is not common in the real world, where money/birth/social position/physical appearance are more important. Long-time Linuxers are more used to the "if it works, use it no matter who it comes from" attitude, while others worry about the endless political maneuverings more common in normal human relationships. ("Should I use Person 1's code? It works better than Person 2's code, but Person 2 is influential and maybe I should suck up to Person 2...")
Representative democracies have an inner circle of politicians and Pocket Filler corperations...There is not currently a method of government that has been invented to inform the uninformed, give power to the powerless, and lead the scattered in a manner that is truely their own motive.
"Democracy is the worst form of government, except for all the others."
The problem is that most people don't want to think and make decisions on their own; they want to do the easiest thing, even if that means bowing down to Big Brother. This tendency is probably impossible to conquer. Til everyone on the planet is capable of thinking for themselves and willing to do that most of the time, there will be a minority that will control many aspects of society.
Thing is, in the Linux community, I think those in the "elite minority" encourage independent thought/action/learning. "Study the source, and you too can hack device drivers someday." When was the last time you saw a large media conglomerate encouraging ordinary people to set up their own small private radio stations?
What hype does Windows have? Very little, it is just so well known it doesn't need the hype that is around Linux.
Windows has a bunch of hype going for it... check out any paper magazine aimed at PHBs/Real (L)users and 30-50% of the ads are for one MS product or another. And, of course, when W2K is released, there will be more commercials and banner ads than you can shake a 10baseT at. RedHat/VA/Penguin Computing/whoever will have to fight back with hype of their own, so the hype will not die down anytime soon.
When the hype does die down and Linux is on store shelves all over the world, the Linux community/distro packagers will face a new problem. To borrow from business-weasel jargon, the market will be "mature" or "saturated" and no longer "the Next Big Thing." This means less quick money for those involved, and there will be some sort of backlash when/if Linux becomes really popular. (Maybe the backlash has already started, if you check the more extreme BSD-phile comments posted along with this article...)
Anyway, Linux needs no hype to survive or even thrive, as people can see by looking back to the 0.XX kernel days. This is foreign to marketing peoples' experience, thus they probably can't write about it with any degree of accuracy:-]
Nobody's going to depend solely on ASP for anything important when there's another way. Here's a couple of real-life reasons why:
First, at work, there's a small database that's stored on a few servers 50 feet away from some very competent network guys. One server hiccupped, and for 15 minutes, nobody could log in. This irritated close to a hundred college students even though the problem was fixed ASAP. Several people lost files. And this was in what you could call a best-case scenario--the desktops were all LAN-connected, with the server less than 100 feet away and competent staff very close.
Second, at the U of M (major university, reasonably clued-in IT department) there have been a couple of fires/machine room troubles in the last couple of months, resulting in sporadic, 1-2 day lossages of E-mail (the original ASP 'killer app' IIRC.) Boy, were people pissed off.
Third, clueful folks know you can't depend on the network, the local hard drive, or really much of anything. It's almost essential to have a backup plan that'll let you accomplish X if the preferred method fails. If you can only run FooApp via a remote server, and some klutz with a backhoe severs your business's T1 line....
Others have made valid points regarding privacy, data ownership, and vendor lock-in. I'd say that the privacy and data ownership questions could be solved by looking at the policies of ISPs that handle E-mail (provided there are any ISPs that have rational E-mail policies.) "Does FooNet own that sexually explicit message I sent? Could they read it and distribute it at whim?" (common sense says "Hell no!" to both; let's hope the law follows.)
Lock-in exists in most computing places today, and only widespread adoption of open file formats/+open source could ever eradicate it. (here's hoping.) People have been dealing with lock-in ever since mainframes and will continue to do so. Sad, but true....
Hear hear! The worst headaches in computing come from a lack of interoperability--that's one reason why they have an ANSI standard for C and C++ among other really important things.
The danger people have to beware of here is "Well, it works on my box, and the box I use at work; why does it need to work anywhere else?" That attitude is widespread, although I'd say the Net is taking it apart... slowly. (standard HTML, PDF, JPEG, PNG, XML)
Is it better to have a single species dominate the ecosystem, or is it better to have biodiversity?
Hmm. The situation with computers today makes me wonder whether this should be worded as, "Is it better to have French, German, Spanish, Hindi, Arabic... or should we point guns at people and make everyone speak English?" Native file formats, like native human languages, make sense in some cases. However, there should be some sort of standard format (Esperanto?:-]) which most everything could be translated to/from with minimal information loss. Here's hoping XML fits the bill.
Everybody I know who's installed/tried to install Linux has complained about the horrible mess of partitioning drives, choosing packages, manually configuring X-servers... How about a distro that was designed specifically for those who need to dual-boot Win95/98? (A fairly large market segment, I think.) This should:
Use a boot floppy/boot CD-ROM which automatically ran FIPS to create space for the new OS
Give users control over how things are partitioned (expert mode) or do something "sensible" (64M swap, 10% of disk for/home, rest for/) by default
Give users 3 or 4 preconfigured package install options ("minimal","server","everything") or let the user choose each 'n every package a la SuSE's YAST
up-to-date documentation...
I include the last because the distro I bought back in May said, "LILO and FIPS won't work with the new FAT32 format!" and they most certainly did. Caused me some unnecessary panic there.
Also, how about some standards for where the various config files go? Do I find init.d in/etc or/sbin or what? There is the Filesystem Hierarchy Standard, but it seems like few people follow that--why not start?
Remember - the public are a malleable bunch. Say something loud and frequently enough, they will believe - no matter what physical evidence jumps up and bites them on the nose.
Are they? Doesn't quite work that way; if it did, then there would be no drug problem in the USA--all those "Just Say No" ads would've gotten the crack smokers to give up around 1989.
Admittedly, that's an extreme case, but everyone forgets that it takes time to make a real change in the world. Back in the '60s and '70s, American car companies made enormous bloated monstrosities that kept falling apart. Then Honda came along, and though their ad budget was smaller, people began to notice something about their cars: they worked, and they were cheap. Took people a few years, but they didn't believe the FUD GM/Ford/Chrysler were spreading after their friends bought a Honda and such.
Reality triumphs over FUD, even though it takes a while. Patience doesn't seem to be highly valued in the computing world ("I want my servers set up in 15 minutes! We have a business to run!") but Linux enthusiasts may have to show some...
Maybe hitting a limit to processor power will encourage programmers to reintroduce the concept of "knowing how to write good code." Lord knows processor speed and cheap memory have made it possible for even the best programmers to stop thinking about code quality.
It's not so much a matter of "quality" as it is a matter of time. In a/. article a week or so ago, there was a big story about "Why Software Sucks" that basically said, "programmers don't have enough time to write everything well." When you can code an inefficient, memory-hogging algorithm in 2 hours while a streamlined crashproof sucker takes 2 days, guess which wins out.:-[ And does anyone really want to return to the days of bumming single instructions out of assembly code? The only case where it'd be worth the effort is in kernel design or CPU-hog tasks like distributed.net IMHO.
Maybe Moore's Law should contain an addendum or two:
The time spent by programmers on writing good code halves every 18 months.
The number of idiots using computers doubles every 18 months.
The stupidity of managers and marketroids is proportional to e^(x), where x is the number of 18-month periods under consideration.
This article has opened up one gonzo can of worms. --not surprising, I guess. I find it sickening that a professor can't put his ideas out. If they're stupid ideas, they will be peer-reviewed and consigned to oblivion as such--that's what the physical sciences are supposed to do in most cases.
Unfortunately, it seems like this is a social science question, and we all know how full of idiocy and slippery conditions those are. So I'll go out on a limb and propose a few things below:
Only humans deserve "human rights."
Others deserve a more limited set of rights, which should be wrangled over somewhere.
We need to determine who's human and who isn't.
There's a (semi-crude) test for "humanity" that's so general it even applies to machines.
Give every person over the age of 4 a Turing test. If they pass, they get full human rights. If not.... ????
4 was an arbitrary number, chosen because most human children know a big subset of one human language at that time. People whose minds work will be able to pass the test; people whose minds don't work won't. I'd say that those who don't pass should get at least the same rights as, say, a pet animal.
...not what people are looking for. I bought it at the end of August because I wanted a more general reference than the SuSE manual I had. It's filled that need, sort of, but I don't know in the abstract how newbie-friendly it is. It's nice to see paper tutorials on emacs and vi, though--both programs sorely need those. (hides from flamage)
The non-distro-specific nature of the book is great--after all, a Linux system is configurable to a large extent, and who knows what new oddities Corel/RH/Sun/Debian will put in their latest and greatest things? In the tech world, it seems that if it's documented, it's out of date. Maybe this book will be useful for a bit longer because it doesn't only cover version 0.997 of YaST.
This book also answered one nagging question I had: "How do I switch from 1024x768 to 800x600 to 640x480 under X?" The SuSE manual (and their Web page IIRC) said, "Ctrl-Alt-{+,-}" which didn't do anything. Imagine my surprise when I learned that the "+" and "-" referred to the keys on the numeric keypad.:-( Now if only we could switch color depth on the fly....
I hope these same kids don't grow up thinking that computers are toys.... computers need to be used as TOOLS, damnit!
Computers are at once the most complex and most "changeable" device humanity has created. To borrow from a famous quote, computers are like a Swiss Army Chainsaw that has attachments which allow you to craft your own blades. (Perl, gcc, xxgdb...) You can even use 2 or more blades at once. I think it's good that the cases for these machines are becoming more expressive/decorative, but I don't think it's good to make the case the thing. (*cough* iMac *cough*)
Upthread, someone said that computers marketed to kids should include a programming language. Hear hear! A decent programming language is both tool and toy--ask any 3rd grader who's playing around with Logo. I hope Logo still exists in some form; it'd be just the thing for these PCs to include. Unfortunately, today's kids won't really care much for Logo since it's too hard to use it to draw really cool 3D-rendered stuff...
We should be teaching kids to treat computers as expensive tools
Erm, hang around kids for an hour or so, and you'll see them treat powerful, expensive tools like toys. ("Johnny, no! Don't use that electric drill on the coffee table! AAAAH!") It's just in the nature of humans--we need to play with things to figure out what they're capable of.
In several other threads here, I've seen people talking about "What constitutes a real installation." I think, like it or not, that many if not most of those who install Linux in the near future will be dual-booting between Linux and Win9x. That setup is just a bit more difficult to configure/mess with than only having one OS on the machine, as I'm sure everyone knows.
Partitioning the hard drive can be a royal nightmare, of course. SuSE 6.1 had the time-honored solution (defrag C:, boot into DOS, run FIPS.EXE) but the documentation for doing so was scattered throughout the manual and the first distro CD. Also, there was something in the docs which said, "LILO is incompatible with FAT32 partitions!" which was, fortunately, false.:-)
Setting up LILO to do the dual-boot thing was even harder, as the manual left off a rather important line in the sample/etc/lilo.conf they provided. (mustn't forget "table=/dev/hda".) Well, everyone should be allowed some typos, and after RTFMing, I figured it out...
The real danger here is a lack of good documentation--the SuSE people tried really hard and produced a fairly decent manual, but it wasn't quite there yet. (Have RedHat/Caldera done better in their paper documentation?) It's really no good to tell people, "Read the HOWTOS on the LDP website," because while you're installing, you probably don't have much of a network, and some of the HOWTOS are slightly outdated. Maybe I should write a HOWTO of my own, then....
Slashdot Mirror
From the article, it seems that these boxes will be used for Web browsing and not much else. Is Mozilla's UI that horrible, that it needs massive amounts of feedback? Isn't the project already getting massive amounts of feedback? Anyway, it will be interesting to see what comes of this.
Yeah. Right. It's been said before, and in many places, but most IDE drives currently on the market cannot sustain a transfer rate of more than 12M/sec due to limitations in the head/disk assemblies. That won't even saturate a UDMA/33 channel. You might notice a *slight* performance increase at the very beginning of a disk operation, when the drive uses its onboard cache, but for big jobs, forget it. UDMA/66 is really an investment in the future, since in a few years, hard drives might be able to support that kind of I/O speed. Of course, by then, marketing will have us all buying UDMA/256....
It is good that Abit is supporting their hardware in Linux. Why a distro, though? Why not just cut a deal with RedHat/SuSE/Caldera so those distributors get whatever hackarounds Abit came up with and Abit gets a "Works great with $DISTRO" endorsement?
- Everyone should use Linux because it's cool/stable/free/fast/whatever.
- Linux is *ours*, and we don't want stupid apps like AOL on it.
The reason (well, one of the reasons) Microsoft has been successful is because they studied what consumers want and delivered it. To some degree, anyways.Amen to that. I think the problem here is fundamentally related to the way a lot of development for Linux gets done. Many projects get started when a programmer wants to scratch his/her personal itch, and programmers are by nature very different from ordinary users. They want intelligent, general-purpose tools and aren't scared off by arcane interfaces. As a result, the market for AOL-type software in the Linux community is barely being served at all, though it seems like RedHat/Caldera/SuSE/et al. are trying with graphical installations, included support databases and FAQs, printtool configuration programs.... The great strength and weakness of Linux is that there's really no central authority who can tell everyone How To Do Things or What Software To Develop/Sell.
As long as that attitude prevails, Linux is going to have a hard time with nontechnical people.
IMHO, the main reason Linux isn't taking off more than it already is is hardware support. WinXX detects and uses all the hardware in your system right out of the box because--surprise--the manufacturers either wrote their own WinXX drivers or gave MS the specs so drivers could be written. Linux often doesn't have that advantage, so people say, "What good is it if I can't use my FooCard 7000?" And sometimes I can't blame them for that...
As for the flaming thing, if people in the Linux community can't take constructive criticism, they should try to grow up a bit. One well-reasoned, polite defense or critique is worth a thousand flames, if only because it's more likely to be read. It's easy to call someone a f!@#ing piece of s*-t just because you don't agree with his/her statements, but that creates smoke, noise, bad feelings, and nothing else. You'd think people would learn.
As for why people like them, I guess that it's because you have to put a lot of effort into learning the moves for all the characters, and then you can beat the hell out of your friends and do really cool/difficult things with said characters. Also, they provide more interaction than an RPG--it's just more fun to beat up your friends than beat up the computer.
But really, this looks like more of the same. Ever since Street Fighter 2, it seems like 90% of the games that are released are just the same bloody idea run into the ground over and over again. They change the characters involved, they tweak the controls, they do it in pseudo 3D a la Tekken 3, War Gods, and Virtua Fighter... but the idea's the same and too often the gameplay's repetitive and bloody annoying.
No wonder I end up spending more money on pinball these days.
However, it would not surprise me to learn that the human mind actually ran a hodgepodge of Perl, TECO macros, Lisp, and COBOL (the last being in the R-complex, which pop culture has referred to as the "dinosaur brain".)
I actually found that the article, while it was interesting, was somewhat incoherent... and now, 15 minutes after reading it, I can't remember the author's main point or any insights/cool things that were in the text. Could it be that the author was trying to produce the textual equivalent of a couple of bong hits? :-)
Of course, it seems like 99% of the users don't even know that Usenet exists. Most of the 1% who do know that it exists aren't interested for various reasons, and prefer IRC because of its immediacy.
This isn't really a good thing, because as others have pointed out previously, Usenet has some things going for it that neither IRC nor Web-based discussion boards have. Unlike IRC, it's asynchronous and persistent. Unlike Web-based boards, it's both centralized and distributed. It's centralized in this way: If you have a question on hardware compatability for Linux, you can easily figure out that you should go to news:comp.os.linux.hardware. On the Web, there's no one place you could go, and you'll spend a lot of time chasing pointers. And of course, it's distributed over N local newsservers, making it that much quicker to d/l your alt.punk fix. And unlike Web boards, the protocol is standardized and the interface is customizable.
If the text-only nature of many Usenet discussion boards scares people off, good. If people can't feel comfortable communicating using plain text, they should probably stay away from computers altogether until someone comes up with a telepathic user interface....
I work as a helldesk minion in a university computer lab. From my unscientific method of walking around looking at what people are doing, 50% of the users are downloading/playing mp3s. Since this particular lab is associated with the Business School, most of these users are not geeks. They are future Suits.
Whether this is good or bad is probably something for Jon Katz to discuss :-) but it's happening. Even the thickest luser can understand that mp3s are (potentially) free beer, not to mention familiar and well-tested and playable almost everywhere.
but I suspect that mp3 will stay fringe
I suspect [SDMI] will be very convienent to buy on-line.
Yet Napster et. al. have made it very convenient to find free (beer) mp3s on-line, and these search engines/filesharing programs will only get easier to use. Hmm, do I pay $1 for this Beatles song, or spend 5 minutes finding where I can get it for free? 'Tis a no-brainer.
First the laptops arrived several weeks behind schedule, then about 20% of them were missing a relatively important piece of software that's used locally, then 4 or 5 of them were apparently set to 800x600 resolution at the factory. (They all had 1024x768 LCD screens, so those looked extraordinarily bad.) I heard a fair number of horror stories from students who dealt with Dell's Friendly Helpful Tech Support, too.
Remember, this was using an OS that Dell is experienced with and knows well. Maybe they've gotten their act together in the intervening 5 months...
Likewise for the shrinking/expanding icons in the dock... which brings up another problem: What about desktop space? People are getting more and more used to having $LARGENUM applications open all the time, and switching among them and/or grouping them logically can be a royal pain in both the standard WinXX and MacOS desktops as of now. Most WMs for Un*x do the Right Thing and have virtual desktops, so why doesn't Apple get with the program? Sure, Macs have huge monitors available and can run multiple monitors easily... but Joe User buying his first iMac probably can't afford that.
Apple could even get nasty and say, "This new, totally unavailable-anywhere-else 'Virtual Super SwitchDesk' functionality is almost as good as having 2 or more monitors connected to your iMac. " New checklist feature + increased hardware sales, eh?
Apple shot themselves in the foot when they made floppies an add-on. There are a freaking TON of older computers out there where the best and quickest means of data sharing is via 3.5" floppies. Maybe everything around your home and office has 10/100baseT, but that's definitely not how it is everywhere. Floppies will be around for another 3-4 years if not longer... remember, the IBM PC and its progeny are still around because A) it was easy to port old CP/M stuff to the PC B) all the new iterations of the PC are fully backwards-combatible. (sic)
Anyway, I think a better scent for MS's web page would be ethyl mercaptan. CH3-CH2SH, patently artificial, listed in the Guinness Book as being among the foulest odors known to humans. Penetrating and lasts a long time, too--skunks use mercaptans in their personal defense musk.
[0] Yep, I'm one of them.
[1] Sorry--couldn't resist.
The University of Michigan, who you'd think would be way out in front with a campus-wide wireless LAN, has very spotty coverage and a couple of different wireless networks. There's one network that extends over the space of 1.5 square blocks. They're probably not going to make the coverage area larger because the transceivers are expensive. Max speed I ever observed on this wireless LAN was 400 Kbits/second--nice, but 3x slower than people would get on a bad day with a standard 10bT hooked into a nearly-ubiquitous RJ-45 port.
Security might also be a concern with a wireless net. Wireless sets should have encryption built in and done in hardware, but that adds cost. And when a university sets up a large wireless net, a $10 difference in the price of each transceiver can make a big difference. Better hope those smart CS kids aren't feeling antisocial and h4x0rish...
The future is most definitely wireless; the future is also Not Here Yet. Give it a couple of years...
How was I to know the Micro-Star MS-5185 motherboard that the system was built on was flaky as hell? (Micro-Star doesn't even acknowledge that model's existence on their site, BTW.) Confronting Gateway about the problem resulted in them saying, "It's due to that 'Linux' thing you're running." I ended up replacing nearly half the system components. The original processor is still cranking along, though.
Gateway will probably do well selling systems with AMD processors. They just need to make sure the other components of their systems aren't wretched crap.
Odd, since I used mpeg_play to verify that the stupid thing got encoded right. I have a very recent version taken from freshmeat, and also called the thing with -dither color and -quality on. *shrug*
The next poster was right about mtv. though it's nagware and closed, I might have to shell out its fee to watch video clips...
Heck, I've been thinking about this for a while. I am currently rendering a small, stupid 3.5 second animation that does the "freeze-n-pan" thing used in the Matrix, all in POVray. After about 3:10 EST, you can go here and see the result. The source code will all be available here so people can see how I did it. I'm sure all the raytracing folks out there know exactly how to do this, but if non-raytracing folks are curious....
Hmm. I don't know where you've been lurking, but on the Linux newsgroups, there are an enormous number of newbies looking for help, and a smaller population of people trying valiantly to help them. I personally visit comp.os.linux.hardware and comp.os.linux.misc every day, looking for questions that I know the answers to and trying to help people even if they don't know what button 2 on their mouse is for. And I know I'm not alone.
I think part of the problem is that Linux/the "geek society" is more of a meritocracy than anything. Anyone can be an expert; all it takes is time and effort. This is not common in the real world, where money/birth/social position/physical appearance are more important. Long-time Linuxers are more used to the "if it works, use it no matter who it comes from" attitude, while others worry about the endless political maneuverings more common in normal human relationships. ("Should I use Person 1's code? It works better than Person 2's code, but Person 2 is influential and maybe I should suck up to Person 2...")
Representative democracies have an inner circle of politicians and Pocket Filler corperations...There is not currently a method of government that has been invented to inform the uninformed, give power to the powerless, and lead the scattered in a manner that is truely their own motive.
"Democracy is the worst form of government, except for all the others."
The problem is that most people don't want to think and make decisions on their own; they want to do the easiest thing, even if that means bowing down to Big Brother. This tendency is probably impossible to conquer. Til everyone on the planet is capable of thinking for themselves and willing to do that most of the time, there will be a minority that will control many aspects of society.
Thing is, in the Linux community, I think those in the "elite minority" encourage independent thought/action/learning. "Study the source, and you too can hack device drivers someday." When was the last time you saw a large media conglomerate encouraging ordinary people to set up their own small private radio stations?
Windows has a bunch of hype going for it... check out any paper magazine aimed at PHBs/Real (L)users and 30-50% of the ads are for one MS product or another. And, of course, when W2K is released, there will be more commercials and banner ads than you can shake a 10baseT at. RedHat/VA/Penguin Computing/whoever will have to fight back with hype of their own, so the hype will not die down anytime soon.
When the hype does die down and Linux is on store shelves all over the world, the Linux community/distro packagers will face a new problem. To borrow from business-weasel jargon, the market will be "mature" or "saturated" and no longer "the Next Big Thing." This means less quick money for those involved, and there will be some sort of backlash when/if Linux becomes really popular. (Maybe the backlash has already started, if you check the more extreme BSD-phile comments posted along with this article...)
Anyway, Linux needs no hype to survive or even thrive, as people can see by looking back to the 0.XX kernel days. This is foreign to marketing peoples' experience, thus they probably can't write about it with any degree of accuracy :-]
First, at work, there's a small database that's stored on a few servers 50 feet away from some very competent network guys. One server hiccupped, and for 15 minutes, nobody could log in. This irritated close to a hundred college students even though the problem was fixed ASAP. Several people lost files. And this was in what you could call a best-case scenario--the desktops were all LAN-connected, with the server less than 100 feet away and competent staff very close.
Second, at the U of M (major university, reasonably clued-in IT department) there have been a couple of fires/machine room troubles in the last couple of months, resulting in sporadic, 1-2 day lossages of E-mail (the original ASP 'killer app' IIRC.) Boy, were people pissed off.
Third, clueful folks know you can't depend on the network, the local hard drive, or really much of anything. It's almost essential to have a backup plan that'll let you accomplish X if the preferred method fails. If you can only run FooApp via a remote server, and some klutz with a backhoe severs your business's T1 line....
Others have made valid points regarding privacy, data ownership, and vendor lock-in. I'd say that the privacy and data ownership questions could be solved by looking at the policies of ISPs that handle E-mail (provided there are any ISPs that have rational E-mail policies.) "Does FooNet own that sexually explicit message I sent? Could they read it and distribute it at whim?" (common sense says "Hell no!" to both; let's hope the law follows.)
Lock-in exists in most computing places today, and only widespread adoption of open file formats/+open source could ever eradicate it. (here's hoping.) People have been dealing with lock-in ever since mainframes and will continue to do so. Sad, but true....
Hear hear! The worst headaches in computing come from a lack of interoperability--that's one reason why they have an ANSI standard for C and C++ among other really important things.
The danger people have to beware of here is "Well, it works on my box, and the box I use at work; why does it need to work anywhere else?" That attitude is widespread, although I'd say the Net is taking it apart... slowly. (standard HTML, PDF, JPEG, PNG, XML)
Is it better to have a single species dominate the ecosystem, or is it better to have biodiversity?
Hmm. The situation with computers today makes me wonder whether this should be worded as, "Is it better to have French, German, Spanish, Hindi, Arabic... or should we point guns at people and make everyone speak English?" Native file formats, like native human languages, make sense in some cases. However, there should be some sort of standard format (Esperanto? :-]) which most everything could be translated to/from with minimal information loss. Here's hoping XML fits the bill.
- Use a boot floppy/boot CD-ROM which automatically ran FIPS to create space for the new OS
- Give users control over how things are partitioned (expert mode) or do something "sensible" (64M swap, 10% of disk for
/home, rest for /) by default - Give users 3 or 4 preconfigured package install options ("minimal","server","everything") or let the user choose each 'n every package a la SuSE's YAST
- up-to-date documentation...
I include the last because the distro I bought back in May said, "LILO and FIPS won't work with the new FAT32 format!" and they most certainly did. Caused me some unnecessary panic there.Also, how about some standards for where the various config files go? Do I find init.d in /etc or /sbin or what? There is the Filesystem Hierarchy Standard, but it seems like few people follow that--why not start?
Are they? Doesn't quite work that way; if it did, then there would be no drug problem in the USA--all those "Just Say No" ads would've gotten the crack smokers to give up around 1989.
Admittedly, that's an extreme case, but everyone forgets that it takes time to make a real change in the world. Back in the '60s and '70s, American car companies made enormous bloated monstrosities that kept falling apart. Then Honda came along, and though their ad budget was smaller, people began to notice something about their cars: they worked, and they were cheap. Took people a few years, but they didn't believe the FUD GM/Ford/Chrysler were spreading after their friends bought a Honda and such.
Reality triumphs over FUD, even though it takes a while. Patience doesn't seem to be highly valued in the computing world ("I want my servers set up in 15 minutes! We have a business to run!") but Linux enthusiasts may have to show some...
It's not so much a matter of "quality" as it is a matter of time. In a /. article a week or so ago, there was a big story about "Why Software Sucks" that basically said, "programmers don't have enough time to write everything well." When you can code an inefficient, memory-hogging algorithm in 2 hours while a streamlined crashproof sucker takes 2 days, guess which wins out. :-[ And does anyone really want to return to the days of bumming single instructions out of assembly code? The only case where it'd be worth the effort is in kernel design or CPU-hog tasks like distributed.net IMHO.
Maybe Moore's Law should contain an addendum or two:
Unfortunately, it seems like this is a social science question, and we all know how full of idiocy and slippery conditions those are. So I'll go out on a limb and propose a few things below:
- Only humans deserve "human rights."
- Others deserve a more limited set of rights, which should be wrangled over somewhere.
- We need to determine who's human and who isn't.
- There's a (semi-crude) test for "humanity" that's so general it even applies to machines.
- Give every person over the age of 4 a Turing test. If they pass, they get full human rights. If not.... ????
4 was an arbitrary number, chosen because most human children know a big subset of one human language at that time. People whose minds work will be able to pass the test; people whose minds don't work won't. I'd say that those who don't pass should get at least the same rights as, say, a pet animal.(Let the flaming commence.)
The non-distro-specific nature of the book is great--after all, a Linux system is configurable to a large extent, and who knows what new oddities Corel/RH/Sun/Debian will put in their latest and greatest things? In the tech world, it seems that if it's documented, it's out of date. Maybe this book will be useful for a bit longer because it doesn't only cover version 0.997 of YaST.
This book also answered one nagging question I had: "How do I switch from 1024x768 to 800x600 to 640x480 under X?" The SuSE manual (and their Web page IIRC) said, "Ctrl-Alt-{+,-}" which didn't do anything. Imagine my surprise when I learned that the "+" and "-" referred to the keys on the numeric keypad. :-( Now if only we could switch color depth on the fly....
Computers are at once the most complex and most "changeable" device humanity has created. To borrow from a famous quote, computers are like a Swiss Army Chainsaw that has attachments which allow you to craft your own blades. (Perl, gcc, xxgdb...) You can even use 2 or more blades at once. I think it's good that the cases for these machines are becoming more expressive/decorative, but I don't think it's good to make the case the thing. (*cough* iMac *cough*)
Upthread, someone said that computers marketed to kids should include a programming language. Hear hear! A decent programming language is both tool and toy--ask any 3rd grader who's playing around with Logo. I hope Logo still exists in some form; it'd be just the thing for these PCs to include. Unfortunately, today's kids won't really care much for Logo since it's too hard to use it to draw really cool 3D-rendered stuff...
We should be teaching kids to treat computers as expensive tools
Erm, hang around kids for an hour or so, and you'll see them treat powerful, expensive tools like toys. ("Johnny, no! Don't use that electric drill on the coffee table! AAAAH!") It's just in the nature of humans--we need to play with things to figure out what they're capable of.
Partitioning the hard drive can be a royal nightmare, of course. SuSE 6.1 had the time-honored solution (defrag C:, boot into DOS, run FIPS.EXE) but the documentation for doing so was scattered throughout the manual and the first distro CD. Also, there was something in the docs which said, "LILO is incompatible with FAT32 partitions!" which was, fortunately, false. :-)
Setting up LILO to do the dual-boot thing was even harder, as the manual left off a rather important line in the sample /etc/lilo.conf they provided. (mustn't forget "table=/dev/hda".) Well, everyone should be allowed some typos, and after RTFMing, I figured it out...
The real danger here is a lack of good documentation--the SuSE people tried really hard and produced a fairly decent manual, but it wasn't quite there yet. (Have RedHat/Caldera done better in their paper documentation?) It's really no good to tell people, "Read the HOWTOS on the LDP website," because while you're installing, you probably don't have much of a network, and some of the HOWTOS are slightly outdated. Maybe I should write a HOWTO of my own, then....