2) It takes.9G to prevent humans from having physiological problems.
Excuse me, but how do you know? It seems that all the experience we've had has been with 1 G and 0 G, with a couple short experiments at 1/6 G. Who exactly has spent a year at.9 G and seen serious bone loss?
Oh, well, it's irrelevant:
Steel is not strong enough to provide this in a rotating cylinder,
You want to do the math on this one, and get back to me? Neglecting air pressure (which makes the problem worse, but not much) and endcaps (which makes the problem better, but not much), the maximum radius of a cylindrical O'Neill colony would be
Plug in your 8.8 m/s^2 pseudogravity, and 2700 kg/m^3 density and 185 MPa ultimate tensile strength for a good aluminum alloy, and you get a nice 7.7 kilometer radius. Drop down to the yield strength of structural aluminum, let the colony structure carry it's own weight in radiation shielding and contents, and throw in a safety factor of 2, and you still get a colony a mile wide.
Radiation storms strong enough to kill unprotected people happen frequently. Every few years, we on earth at sea level get an extra chest X-ray from one. They are not predictable, have fast onset. Solar flares are more predictable, but require lots of mass between living tissue and the sun.
Gee, that's funny, I haven't heard about all the Shuttle astronaut and Mir cosmonaut deaths over the past decade. Of course, they're shielded partly by the Van Allen belts, but they're also in a space station whose skin is dozens of times thinner than an O'Neill colony's would be. Would you provide a reference for this assertion?
I want to see us go into space, but all the simple stories like O'Neill's are just silly.
I've heard O'Neill's ideas called uneconomical, and perhaps they're not politically tenable either... but "simple" is one of the last words I would have used...
..whats the point of building and launching these colonies from the moon when it is simpler and more energy efficient to build them on the moon in the first place ?....
First of all, you don't build on the moon. That would require launching assembled items, which in many cases couldn't be done with a mass driver and so couldn't be done nearly as cheaply. You mine on the moon, and do most refining (even the slag is useful for radiation shielding) and assembly in orbit.
Secondly, that refining may be easier to do on a large scale in orbit. What is the biggest mirror you've ever seen? A mirror in orbit doesn't have to support it's own weight, except for the few thousandths of a G needed for attitude control. And at the focal point of that mirror you have a region of space whose contents can be heated up to the temperature of the Sun's surface.
Third, the final resulting colony is really something you *want* in orbit. Solar power, microgravity manufacturing, asteroidal materials, etc. are either only available or are much easier to access from orbit.
Finally, one of the problems with planets is that while you have gravity, you're stuck with it. We do not yet have the medical technology to live more than a few years in free fall without crippling muscle and bone deterioration. Maybe the 1/6 G on the moon will be enough to prevent that; we don't know. But in free fall, we're not limited to 1/6 G; colonies can be constructed that spin to provide artificial gravity; even at a full G of radial acceleration, you can make and spin aluminum structures kilometers wide safely, and unlike on the moon you wouldn't have to worry about bearings or metal fatigue.
Also, as regards shielding, I suspect the 1970s designs didn't take into account the amount of kinetic engery in a meteorite strike - its been shown that standard metal protection is not sufficient to stop even small meteorites over any length of time and that more innovative (carbon-based) materials are required to prevent effects such as spalling
The designs take shielding thoroughly into account: intended to be outside the Van Allen belts, they worry about radiation shielding primarily, but once you've slapped enough feet of glass and slag around the colony to protect against radiation, meteorite shielding is well taken care of as a side effect.
Oh, and did I forget the biggest benefit of having space colonies along side planetary colonies?
We've only got a few planets and moons here, and they all suck in their own way or another... but we have a *lot* of space. The Earth-Moon system may have a little more resources and a few thousand times more available energy than Earth alone... but the amount of accessable material in the asteroid belt and energy radiating out into deep space dwarfs it all. "Mining the Sky" isn't quite as dated as "High Frontier", and is a little farther looking.
One last benefit: A lunar colony will be stuck on the moon, for good; with a space colony if you want to go anywhere you can take home with you. At first that will just mean a few ion drives to push a colony to Mars orbit or the asteroid belt. But God willing, we will develop fusion power sometime in the next millenium, and of course a "fusion drive" is just what happens when you let some of the power plant's plasma leak out the back of the system. Keep doing so until you hit 10-20% of the speed of light, and a century later you're at your pick of hundreds of stars. When that happens, we're going to have a choice to make. Are we going to bother building an interstellar spaceship for the sake of sending a half dozen young explorers to the nearest good-looking star, to arrive old and die there? Why do anything so half-assed, when we could strap an engine and fuel tank to the back of a 100,000 man colony, and go to the stars for real?
Which of the following would be the most effective course of action to get you to release (current!) Slash source code?...
4. Procure many beautiful women dedicated to the open source cause
Are you accepting volunteers yet???
Sorry, not anymore. I accepted one beautiful volunteer, and now she's positively reluctant to even let me interview additional applicants! Go figure.. *g*
Which of the following would be the most effective course of action to get you to release (current!) Slash source code?
1. Post annoying reminders to every/. article 2. Using the word "hypocrites" a lot 3. Slay the Andover Alien Masters and destroy their Mind Control Ray 4. Procure many beautiful women dedicated to the open source cause 5. Crack slashdot.org, download source code, leave kewl hAx0r d00d cracked page to brag from 6. Bribe Hemos 7. Get a lab insider to "innocently" copy code to an unclassified server 8. Dress up as Rob, go to Andover, claim "Hello, my name is CmdrTaco. I misplaced my copy of the Slashdot code, can I make another?" 9. Send the Slashdot gang a free case of beer. 10. Poison the beer, use the antidote as a bargaining chip. 11. Cry 12. Get Nitrozac to cry
I've never heard K&R referred to as "old-skool" C, but that's as good a phrase as I've ever heard. Technically there's not too much different from ANSI C; ANSI just has a few syntactical improvements... but it's been around for at least a decade now.
Saying that something is written in K&R C is like saying it's written in Olde English - not necessarily a bad thing; just if there isn't a more recent "translation" it doesn't bode well for the maintainance of the code.
ISO C is even closer to ANSI than ANSI is to K&R - basically, well-written ANSI C is ISO C.
I agree that knowing C++ helps you write better C... but I wouldn't learn C++ if you don't already know C or Java - it's too big a bite to take at once.
Not just for tulip, but for all his plethora of ethernet drivers.
But tulip is the one I use, not my "almost a tulip chip" linksys card. I wrote to Becker (instead of testing out beta drivers like I should have..) when my card didn't work with the in-kernel driver at the time - he helpfully pointed me to a newer version that works flawlessly.
I sincerely hope that the "Best Newbie Helper" award doesn't go to some website maintainer - as nice as some "Linux Newbie" websites are, they don't hold a candle to the tireless people who work "in the trenches", as it were.
It's the ready availability of newsgroup help that made Mindcraft's "we didn't find anyone to help us tune Linux" lies so blatant and offensive. It's that newsgroup help that makes Linux a bearable transition for people who can't scale the learning curve by themselves. It's that helpful attitude that gives the Linux community the glowing reputation for support it has.
The only non-newsgroup candidate I'd vote for here is Matt Welsh, for obvious reasons. But the LDP is only really helpful once you're comfortable enough with Linux to make use of that detailed information, and only if your question is a FAQ. For all the obscure, unanticipated, or just ultra-newbie questions out there, everybody turns to DejaNews. I'd like to see this award split between, say, the 5 or 10 most prolific posters to comp.os.linux.setup (the most spiritually-draining, most often needed, and most newbie-heavy place to be helpful).
I haven't frequented the group in a while, but years ago (so I'm not eligible; my tirelessness faded into tiredness before 1999 ever rolled around) I and a relative handful of others posted literally thousands of messages (which shocked the hell out of me the first time I searched for "roystgnr" on Deja) over the years that made the difference for people who got Linux running.
Not all newbies stay newbies, either. My last CIVI homework this fall was completed with the indispensible aid of SLFFEA, a program whose author was kind enough to name me and a few other C.O.L.S. folk in the credits. Makes me feel kind of guilty for leaving, now...
I'm still an undergrad, a dangerous state: not educated enough to really know what I'm talking about, but educated enough to think I might.;-)
Anyway, I'm curious: I understand that current JPEG uses Fourier transforms (a full integral transform, or a Fourier series?) to get a spectral representation of the image data, then drops subtle information from that transformed data to get a similar image from the reverse transform.
So I assume the new JPEG (aside from all the quicktime-esque formatting overhead) uses the same technique, just with a different complete set of functions, the wavelets.
So my first question: is there anything about the above that I'm misunderstanding?
And my second question: What are wavelets? Bessel functions? Something I haven't heard of? Is there a simple formula, or a simple ODE generating wavelet solutions, that I could look at or plug into Maple? I gather that whatever they are they approximate discontinuous functions with much better convergence than sines and cosines... but that could describe a lot of things.
I thought I saw OpenVMS on the list of binary releases of M12. Granted, it's still buggier than Communicator, but it's a step up from Netscape 3.
Oh, and I suspect the reason ESR mentions Linux rather than "free operating systems" is the reason he talks about "open source software" instead of "free software" - marketing spin. Don't confuse the lusers with details, just give them a name they can wrap their heads around...
To the people replying to this comment with anti gun control views, you're missing the point.
I agree with ESR's stance against gun control, but I *don't* think that it should be something that gets tacked (unless it's a random sig?) onto the end of what is, essentially, a press release for the Linux community.
"ESR, spokesman for the Open Source community" and "ESR, anti-gun control advocate" are both fine positions to take, but I'd rather he not mix them. If he's going to start a letter summarizing our views and representing the whole of the OSS community, I'd rather he not finish it by making unrelated statements that many OSS users and authors disagree strongly with.
I don't think this specific instance was a problem - the quote was obviously a sig, random or not; and he didn't claim to be representing anyone but himself in the letter. Nevertheless, it makes me uneasy.
Oh gawd, EVERY time an XFree discussion starts, someone says the development is too closed.
Yup. Often it's me. I'm happy to see people are beating me to it now.
You know what? Shut up.
Let me think.. No.
But you have a nice day, too, OK?
I have one word for you and it is NDA.
Damn, that means I need two words to one-up you: Modular programming.
You remember, that non-monolithic X Server design which is one of the biggest improvements in XFree86 4.0? If you've still got a reason why we shouldn't have, say, anonymous CVS access to the X server core and all the non-NDA drivers, then I'd like to hear it.
It's not *HARD* to become a developer, and anyone who complains about it being 'too closed' has obviously not even looked INTO the subject.
You don't seem to get it. I don't want to become a developer just so I can figure out why XFree86 has mouse input bugs on my machine and hardware cursor bugs on my girlfriend's. Another poster who wanted to work on Truetype support years ago didn't want to sign up to "be a developer" before he could even read a mailing list archive and see what was being worked on.
How many people currently doing the heavy work on the Linux kernel started by saying, "You know, I want to become a full-time Linux developer?" I'd like to see numbers, but I'll bet it's not nearly as many as those who started by saying, "I wonder if there's a driver in development for my foo card,", or "This discussion of the unified page cache on linux-kernel is interesting, I think I'll pull down last week's devel source and see how it works.", or even "This looks like a bug. I think I'll check the latest source and see if it's being fixed."
XFree is not a 'closed' development, it's a controlled one.
Linux and FreeBSD are both controlled, but we get frequent releases (and more frequent prereleases) with the former, and anonymous CVS access with the latter.
And it has to be controlled, as NDAs exist, like them or not.
Needless to say, I don't. But this is a red herring; nobody said we wanted up-to-date source code on NDA'd drivers, just on the 99% of XFree86 that is free software.
To be fair to the XFree86 developers, the 4.0 pre-releases are coming more frequently than they have in the past... but the situation still isn't as good as it should be. To be more fair to the XFree86 developers, I'll point out I think it's a shame that they are underappreciated, even as they write a successful free software project that is arguably as complicated as and more important than the Linux kernel.
However, if they want more developers, they're not making enough of an effort to have a project that is open enough to be attractive to tenatively interested programmers.
And they do want more developers. They're asking for them strenuously enough on the web page. What was the initial timeline I heard last year, XFree86 4.0 by June? That's not quite a 100% slip from schedule, but it's close.
And I want them to have more developers! Between the complaints about configuring X, the emergence of DRI 3D and real GLX support, and the major architecture changes in 4.0, they've got one of the most important free software projects in existance on their hands, and I'd like to see them have every hand they can get.
Eric Raymond always uses gcc vs. egcs and FreeBSD vs. Linux to show how even among free software, projects which are "more open" than others tend to be more successful in the long run. I don't think it's worth a code fork to find out, but it's a shame there isn't anything competing with XFree86 to provide a third example.
Honestly, people complain about the "two-party" system, but that's not the problem America has. We can get third-party (and in many cases, fourth-party, fifth-party, etc.) candidates on the ticket, but people, even if they prefer the underdog to one of the major party candidates, still don't vote for him. Why?
Because except for the "protest vote" value, it doesn't pay to vote for a candidate that doesn't have a good chance of winning. Whine all you want about how you're not "throwing your vote away", but this is a real problem.
Hypothetical election: If McCain (who's my favorite of the current candidates, BTW), Gore, and the Perfect Libertarian Geek candidate are on the ballot in the presidential election, I'm probably still going to vote for McCain, lest the Perfect Geek siphon off enough Republican votes to cause Gore to win. If 30% of the voters preferences run "McCain,Geek,Gore", 30% run "Geek,McCain,Gore", and 40% run "Gore,McCain,Geek", then if everybody voted for the candidate they liked best then you'd end up with *more* people unhappy with the election results.
There's an essay in _A Mathematician Reads the Newspaper_ (great book, worth buying or borrowing; sorry I forget the author) about the quirks of voter preferences in multi-candidate elections - in it the author presents a hypothetical election with 5 candidates, and 5 fair-sounding ways (including the plurality vote) of determining a winner.. except that a different candidate would be chosen the winner by each method.
People have talked about "None of the above" votes and other different systems that would force a new election... but we shouldn't *have* to have election after election to choose between the same group of candidates. And besides, while "None of the above" might help in cases where neither party puts up a good candidate, it does nothing for cases where a third party candidate is marginalized because of a close race between the major parties.
What I'd like to see, personally, is an election system where each voter can vote "Yes" or "No" for each candidate, and the candidate with the largest majority of "Yes" votes wins. (Or a new election is held if no candidate gets a majority). That way, if no party puts up a good candidate, then a new election must be held. If a good third party candidate is running in a race where the two leading parties are close in votes, then voters can safely vote for *both* the underdog and their favorite of the leading candidates, giving the third party entry a chance without risking "throwing their vote away". We wouldn't have to worry about the Greens "stealing Democrat votes" or the Libertarians "stealing Republican votes" anymore, and nobody would ever be "throwing their vote away". And if people were really fed up with all the candidates, they could vote "No" to all of them and get a new batch.
Of course, since such a system: a. Is change, and therefore "scaaarry" b. Is threatening to both the entrenched parties it'll never happen.
And of course, what I'd *really* like to see is the ability for people to give 0 votes, 1 vote, or.3245 votes to each candidate, so they could rank them by preference... but stupid people would think that was too complicated, and stupid people have a vast majority vote no matter what system you use to count them...
Maybe I don't understand the scheme you're proposing, so bear with me.
Proving that someone has a "blessed" client sounds theoretically impossible, for many of the same reasons why creating an uncrackable copy protection scheme in software is impossible. You can't have a perfectly hidden private or symmetric key in your "blessed" client, because to use that key the client has to decrypt it sometime, which implies:
A. The algorithm for decrypting it is there, in the client code. Making the client closed source may make it more difficult, but no less possible, to recover/reproduce this algorithm.
B. The decrypted key is in memory sometime. Whether you run the "blessed" client through a debugger, or halt its execution and examine/proc/kmem byte by byte, you can pull out that data somehow.
There are more ingenious techniques, I suppose - someone mentioned running a "blessed" client and using your cheating client as a proxy between it and the server, passing any "key requests" or "checksum requests" to the blessed client while handling most of the gameplay with the cheating program.
Two more points, as long as I'm posting:
There have been near invincible client-side bots on public servers since I started playing Quake over two years ago, with all the aim improving, sight improving, etc. cheats that Carmack outlined. Closed source didn't do anything to stop them. In other words, to the idiot QW player who whined that he was never buying another Id game: shut up; it's not Id's fault.
The best way to prevent cheating from a client-side bot is to have the client-server protocol such that the client is completely untrusted. Unfortunately, this isn't perfect:
Just because it prevents the client from cheating doesn't mean it prevents the client from being a borg or a bot. You can make a Quake game such that borg clients can't see around corners, but you can't make one such that borgs can't have perfect night vision, perfect (aside from dodging projectile weapons) aim, etc.
There are technical complications in realtime games to making the client completely untrusted. Quick example: Not sending the Quake client data on opponents who are around a corner means that the client can't do any local prediction on that player's movement, which means that you're subject to the full 200+ ms modem lag before the opponent becomes visible after he rounds the corner. I've already been too spoiled as a LPB to enjoy modem FPS games; this would just make them near-unplayable.
Having a number of "blessed" clients as you've suggested is the perfect way to prevent cheating, but short of a magic uncrackable piece of hardware to locally verify the "blessed" client status, I don't see any way of preventing people from creating cheating clients. Closed source makes it harder, but no less possible.
Great, so far as it helps with OCR on handwritten pages. Not so great, as the primary interface to a handheld computer. As much as I prefer the Palm Pilot as a PDA, I'd much rather have a tiny, two-finger typeable keyboard than use graffiti (or the onscreen keyboard alternative). Graffiti just doesn't work as quickly for me and feels more awkward. And recognizing "natural" handwriting would be even worse. My handwriting goes from left to right; it doesn't stay within the sensory area on a PDA. The "unnatural" graffiti writing actually helps because it eliminates that instinct to move your hand while writing.
Voice recognition:
Sure, would have been great to dictate homework when I was in the 6th grade and didn't know how to type. But today it would have to get punctuation, spelling of homonyms, etc. exactly the way I wanted it to be marginally better than typing long paragraphs. And as for voice commands? Forget it. It's been like three years since I've had my computer situated out of hearing range of any other computer around, and I don't want to work within some wacky scheme to decide which computer is supposed to be listening to me.
OCR:
Now, this would be useful. I'd love to live a paperless life, but unfortunately I live within a society that doesn't see things my way. It would be nice for my "filing cabinet" to be completely electronic, but today that's not really a good option unless I save.tif files of every document I wanted to scan. I'd move to completely electronic filing (except for those documents where someone/some government agency would want the original copy of something, of course), if only I could throw a search engine and editing capability around it. I'm a pack rat - I've never deleted a nontrivial personal email. But to try and do that with every important paper doc that crosses my path would be impossible without some better way of organizing/indexing them, and that means OCR.
Got this in my mailbox this morning from the Ditherati mailing list:
PICK IT, BOX IT... UM, WE FORGOT WHAT COMES NEXT
"We sat down yesterday and felt that there were some orders that we were not able to ship for Christmas."
Toysrus.com CEO John Barbour, on the lack of fulfillment felt by his company's warehouse workers during the holiday rush, The New York Times, 23 December 1999
It just is fairly new, and isn't used as well as it should be.
I'm referring in particular to capabilities support (in the latest kernels) and ACL support (in beta testing on top of ext2; I hope it gets into the final ext3!)
Look at the past. Linux (and every OS out there, but that's beside the point) has its security undermined fairly regularly by buffer overflows, etc. discovered in various daemons and suid programs. Rough estimate, it seems like there's a remote root exploit every year or two, and more than one local root exploit each year.
Yes, the patches come within hours or days of the exploits, and yes, anyone who can type "rpm -F" can keep their system up to date with those patches. But there's still that window of vulnerability out there, and there's still the (millions of?) Linux systems out there without root users experienced enough to stay up to date with security patches.
Security will hopefully be much improved in the future will be the use of capabilities in priviledged Linux programs and ACLs in Linux distributions to drop all unrequired permissions. Right now, if a program just needs to open a trusted (< 1024) TCP port, or read files with strict permissions, or have raw access to video hardware... then that program gets run as root, with the full array of root permissions.
And then if that program has a security flaw, then anyone in a position to exploit it has root.
With capabilities, a program (or a wrapper program) can be run as root, but can permanently or temporarily drop selected root capabilities. In other words, if a capabilities-aware suid root program just needs to listen at a low port, but it can be tricked by the user into opening arbitrary files... well, then it still won't be able to spit out/etc/shadow to the evil cracker because it'll have wisely dropped it's file access priviledges and switched to a non-priviledged uid immediately upon execution, a feat now possible because it can use the capabilities support to retain the low port opening privileges with the new uid.
Well, that was a mangled description, but you get the idea. There are links to discussions of the subject by people who know what they're talking about at http://www.millenniumproductsllc.com/sjp/
It just is fairly new, and isn't used as well as it should be.
I'm referring in particular to capabilities support (in the latest kernels) and ACL support (in beta testing on top of ext2; I hope it gets into the final ext3!)
Look at the past. Linux (and every OS out there, but that's beside the point) has its security undermined fairly regularly by buffer overflows, etc. discovered in various daemons and suid programs. Rough estimate, it seems like there's a remote root exploit every year or two, and more than one local root exploit each year.
Yes, the patches come within hours or days of the exploits, and yes, anyone who can type "rpm -F" can keep their system up to date with those patches. But there's still that window of vulnerability out there, and there's still the (millions of?) Linux systems out there without root users experienced enough to stay up to date with security patches.
Security will hopefully be much improved in the future will be the use of capabilities in priviledged Linux programs and ACLs in Linux distributions to drop all unrequired permissions. Right now, if a program just needs to open a trusted (http://www.millenniumproductsllc.com/sjp/
It's not that Linux didn't have some of the same fundamental security bugs that Windows did way back when; Linux was just better about fixing them.
When the string of bogus IP attacks against Windows came out two years ago (teardrop & kin) Linux was vulnerable to the first of them too... but the Linux patch was out within a day or two, whereas Windows was vulnerable for months. To boot, the Linux patch fixed all the similar TCP/IP stack problems, whereas Windows ended up being vulnerable to syndrop, newtear, and a sequence of nearly identical attacks.
...hopefully, is that LinuxOne *will* crash and burn, and anyone stupid enough to use it as an anti-linux datum will be quickly pointed to the denunciations on Slashdot and LinuxWorld, from Bruce Perens, and from basically anyone else who knows what they're talking about.
I mean, honestly, LinuxOne's name has been mud from the very day their existance became known! If someone invests in them because "it's got Linux in the name!" but doesn't bother to do so much as a web search to research the company, they deserve what they get.
Quake game mods are written in QC, but turned into data to be processed by the main code. I think the spirit of the GPL would want that code to be released, but it is only a small step from there to saying that every program loaded by a GPL operating system must be GPL, which is clearly not the case.
No, Quake game mods are not required to be under the GPL, at least not if my understanding of the way they work is correct. It's either an interpreter or a JIT compiler, correct? Well, just because an interpreter (like bash) is under the GPL doesn't mean that all the scripts it interprets have to be under the same license. Just because the GCC compiler suite is GPL'ed doesn't mean that the code you compile with it has to be.
The GPL does apply to both statically and dynamically linked binaries, however, so Quake 2 game modifications would have to be GPL'ed if Quake 2 itself was released under only the GPL. If you released Quake 2 in the future under the LGPL then binary-only modules would be fine.
The Quake 2 case is interesting, though, since it's already released under a license that (if not explicitly, than by years of practice) allows binary-only DLLs to be released; and rereleasing it under the GPL wouldn't take away that right for people who purchased the original license. So would people who *hadn't* purchased the original game be allowed to write binary-only modules for a GPL'ed Quake 2? I haven't a clue. Ask RMS.
I must first put up a disclaimer of total ignorance: I'm not an economist, not a historian, and not even a closely interested observer. But I do have a point to make, and a question:
I (like most political pundits I've heard) was under the general impression that the primary cause of Russia's economic problems was simply endemic corruption caused by a too-fast attempt to switch to a free market system.
I'm not so sure anymore. Newsweek claimed recently that over half of the Russian economy was based on barter, as one of many figures they mentioned to demonstrate to readers that Russia's economy is in serious trouble.
I think it has more significance than that, though. People don't think to themselves, "I think I'll trade goods without paper money, because it will symbolize how screwed up our economy is." People barter because they don't have a trusted medium of exchange that they can use instead: either they don't have enough cash, or they don't have any cash they can trust.
We refer to "the Great Depression" instead of "the 1930 Recession" in good part because the U.S. government at the time didn't have a clue how to handle money; when the recession started, in a misguided attempt to preemptively prevent hyperinflation, the government cut back the money supply. Drastically. By over 50% on some indicators (It's been a while since I read this, and I don't so much as remember the difference between M1 and M2 now - the curious can check figures for themselves), and unemployment and GDP (in constant dollars) pretty much dropped in sync with the reduction in the money supply.
OK, I know this is simplistic, I know there were other major contributing factors (e.g. massive trade restrictions/tariffs) to the Great Depression... but the money supply may have been the largest single factor.
And it looks to me like something similar is happening in Russia. *Something* is causing people there to barter, whether it's a black market (which would imply those trade restrictions again), a government reduction in the money supply (trying to fight inflation, perhaps?) or simply the chain reaction of people pulling their money out of the country to avoid what they see as a financially dangerous situation.
Would someone who knows more about the situation in Russia than I do care to comment? Even if this was the problem, I'd have no idea how to solve it. Increasing the supply of fiat money (as eventually happened in the US when the government had to spend big in the late 30s and the 40s) would only work if inflation is in the single digits. Foreign aid and investment hasn't seemed to make a dent so far.
And this is a horrible situation, both for it's own sake and for what it implies for the future. People have already compared 1990's Russia to 1930's Germany, but even if Russia didn't end up in a political collapse, what would their economy's continued failure say about the development of a free market? That it's possible to kick an industrial economy into a depression state from which there is no return? That's almost too frightening to think about.
AMD Will slash Athlon Prices: This is one of the biggest misconceptions about
The Athlon 600 is retailing for as little as $379 now, about half of what they were at four months ago. Granted, it wasn't in one dramatic price slash, but it's a hell of a drop. AMD has kept the Athlons cheaper and faster than same-Mhz PIIIs forever, and I don't see that changing soon. Maybe Intel will get ahead in the Mhz war, but that won't change the market situation for people not buying the absolute fastest chip.
Enough with the oxymoronic names already! "Pentium IV"? Please. You'd think that Intel could make it's chip names correspond a little better to chip differences, too. The only significant difference between a PII and a PIII is the (little used) SSE support, whereas you have to shop carefully between PIII releases to make sure you get one of the uber-cache "Coppermine" (disclaimer: no copper included) models that actually can compete with an Athlon.
It's a shame that Pentium IIIs, while invariably poor consumer buys, are still doing well because of marketing. Want a cheap computer? Get a Celeron. Want a fast computer? Get an Athlon, which will outrun any PIII of the same price.
A neighbor of mine got their first computer last Christmas. Maybe my advice got there too late, because their overpriced, unstable Compaq PII system was coupled with a lousy ATI video card and a winmodem. I had the pleasure last month of trying to figure out driver problems with someone's dual PIII, 128MB (or was it 256?) RAM, Voodoo 3 3500, behemoth system... that was having conflicts with his ISA network card.
Granted, there are a lot of CPU-intensive things (Quake 3 - intense!) out there I'd love to have an 800Mhz Athlon for, but those are the exception (Quake 3 - exceptional!) rather than the rule. (Quake 3 rules!)
And even in 3D games, we're getting to the point where the processor won't be the limiting factor anymore. I mean, once you're getting 60fps at 1024x768, do you shoot for 90fps? No, you crank up your resolution, detail, or color depth. And once you do that you're worrying about your video card's fill rate, texture memory, or texture RAM bandwidth, not about your CPU. Hell, with T&L going on new video cards, we're going to be at the point where the AGP bus and human perception are the limiting factors on visual quality, not the CPU. Maybe games like Halo will have ultra-impressive physics and AI to use all those CPU cycles, but I'd like to see it.
Granted, there are always going to be applications (if only Beowulf clusters) where CPU value increases linearly with CPU speed.. but these are niche things compared to the huge consumer market being sold to today.
What I want to see isn't faster CPUs, it's broadband access! I have a K6II on a 10baseT college dorm connection this year, and if I have to trade it for a quad 1.2Ghz Athlon on a 56K modem when I graduate, it's not going to be a good trade. Whoever thought up Intel's "our CPU makes the internet faster!" campaign should be flogged.
No, you don't lose a point when you're fragged. You can have 200 deaths and 100 kills and you'll still beat the guy who got 99 kills but didn't die. (note: I have a history of getting 90% of the kills and 50% of the deaths of people who regularly beat me, so maybe I'm biased...) And not only are the repercussions next to nothing, but they're being adjusted even lower in Quake 3 for the sake of "game balance" (read: keeping newbies from becoming frustrated). New respawns start with 125 health instead of 100 so they can't be killed in one hit. The default weapon is a serious machine gun instead of a joke blaster. All the non-default weapons are quite good, and easy to get a hold of fast. No more runes in CTF. No more power armor. Even the default Q3test policy of forcing respawns to wait until after your old corpse hit the ground is gone!
They're not just a beige box builder that added Linux to their name, they actually do know the OS and what kind of hardware to put it on well. When we got a Linux workstation last spring, Gateway considered it a "special engineering" cost of ~$500, Dell charged as much as for NT, and both of them offered a selection of adequate but poorly-selected-for-Linux hardware, going with Adaptec over Buslogic, with some poor video card over Matrox... just basically taking their Windows workstation and slapping Red Hat on instead. I'm sure the big vendors are putting Linux prices in line with reality by now, but they're probably still shipping hardware with reverse-engineered drivers.
Slashdot Mirror
2) It takes .9G to prevent humans from having physiological problems.
.9 G and seen serious bone loss?
Excuse me, but how do you know? It seems that all the experience we've had has been with 1 G and 0 G, with a couple short experiments at 1/6 G. Who exactly has spent a year at
Oh, well, it's irrelevant:
Steel is not strong enough to provide this in a rotating cylinder,
You want to do the math on this one, and get back to me? Neglecting air pressure (which makes the problem worse, but not much) and endcaps (which makes the problem better, but not much), the maximum radius of a cylindrical O'Neill colony would be
radius = (tensile strength)/[(density)(pseudogravity)]
Plug in your 8.8 m/s^2 pseudogravity, and 2700 kg/m^3 density and 185 MPa ultimate tensile strength for a good aluminum alloy, and you get a nice 7.7 kilometer radius. Drop down to the yield strength of structural aluminum, let the colony structure carry it's own weight in radiation shielding and contents, and throw in a safety factor of 2, and you still get a colony a mile wide.
Radiation storms strong enough to kill unprotected people happen frequently. Every few years, we on earth at sea level get an extra chest X-ray from one. They are not predictable, have fast onset. Solar flares are more predictable, but require lots of mass between living tissue and the sun.
Gee, that's funny, I haven't heard about all the Shuttle astronaut and Mir cosmonaut deaths over the past decade. Of course, they're shielded partly by the Van Allen belts, but they're also in a space station whose skin is dozens of times thinner than an O'Neill colony's would be. Would you provide a reference for this assertion?
I want to see us go into space, but all the simple stories like O'Neill's are just silly.
I've heard O'Neill's ideas called uneconomical, and perhaps they're not politically tenable either... but "simple" is one of the last words I would have used...
..whats the point of building and launching these colonies from the moon when it is simpler and more energy efficient to build them on the moon in the first place ?....
First of all, you don't build on the moon. That would require launching assembled items, which in many cases couldn't be done with a mass driver and so couldn't be done nearly as cheaply. You mine on the moon, and do most refining (even the slag is useful for radiation shielding) and assembly in orbit.
Secondly, that refining may be easier to do on a large scale in orbit. What is the biggest mirror you've ever seen? A mirror in orbit doesn't have to support it's own weight, except for the few thousandths of a G needed for attitude control. And at the focal point of that mirror you have a region of space whose contents can be heated up to the temperature of the Sun's surface.
Third, the final resulting colony is really something you *want* in orbit. Solar power, microgravity manufacturing, asteroidal materials, etc. are either only available or are much easier to access from orbit.
Finally, one of the problems with planets is that while you have gravity, you're stuck with it. We do not yet have the medical technology to live more than a few years in free fall without crippling muscle and bone deterioration. Maybe the 1/6 G on the moon will be enough to prevent that; we don't know. But in free fall, we're not limited to 1/6 G; colonies can be constructed that spin to provide artificial gravity; even at a full G of radial acceleration, you can make and spin aluminum structures kilometers wide safely, and unlike on the moon you wouldn't have to worry about bearings or metal fatigue.
Also, as regards shielding, I suspect the 1970s designs didn't take into account the amount of kinetic engery in a meteorite strike - its been shown that standard metal protection is not sufficient to stop even small meteorites over any length of time and that more innovative (carbon-based) materials are required to prevent effects such as spalling
The designs take shielding thoroughly into account: intended to be outside the Van Allen belts, they worry about radiation shielding primarily, but once you've slapped enough feet of glass and slag around the colony to protect against radiation, meteorite shielding is well taken care of as a side effect.
Oh, and did I forget the biggest benefit of having space colonies along side planetary colonies?
We've only got a few planets and moons here, and they all suck in their own way or another... but we have a *lot* of space. The Earth-Moon system may have a little more resources and a few thousand times more available energy than Earth alone... but the amount of accessable material in the asteroid belt and energy radiating out into deep space dwarfs it all. "Mining the Sky" isn't quite as dated as "High Frontier", and is a little farther looking.
One last benefit: A lunar colony will be stuck on the moon, for good; with a space colony if you want to go anywhere you can take home with you. At first that will just mean a few ion drives to push a colony to Mars orbit or the asteroid belt. But God willing, we will develop fusion power sometime in the next millenium, and of course a "fusion drive" is just what happens when you let some of the power plant's plasma leak out the back of the system. Keep doing so until you hit 10-20% of the speed of light, and a century later you're at your pick of hundreds of stars. When that happens, we're going to have a choice to make. Are we going to bother building an interstellar spaceship for the sake of sending a half dozen young explorers to the nearest good-looking star, to arrive old and die there? Why do anything so half-assed, when we could strap an engine and fuel tank to the back of a 100,000 man colony, and go to the stars for real?
Are you accepting volunteers yet???
Sorry, not anymore. I accepted one beautiful volunteer, and now she's positively reluctant to even let me interview additional applicants! Go figure.. *g*
Which of the following would be the most effective course of action to get you to release (current!) Slash source code?
/. article
1. Post annoying reminders to every
2. Using the word "hypocrites" a lot
3. Slay the Andover Alien Masters and destroy their Mind Control Ray
4. Procure many beautiful women dedicated to the open source cause
5. Crack slashdot.org, download source code, leave kewl hAx0r d00d cracked page to brag from
6. Bribe Hemos
7. Get a lab insider to "innocently" copy code to an unclassified server
8. Dress up as Rob, go to Andover, claim "Hello, my name is CmdrTaco. I misplaced my copy of the Slashdot code, can I make another?"
9. Send the Slashdot gang a free case of beer.
10. Poison the beer, use the antidote as a bargaining chip.
11. Cry
12. Get Nitrozac to cry
I've never heard K&R referred to as "old-skool" C, but that's as good a phrase as I've ever heard. Technically there's not too much different from ANSI C; ANSI just has a few syntactical improvements... but it's been around for at least a decade now.
Saying that something is written in K&R C is like saying it's written in Olde English - not necessarily a bad thing; just if there isn't a more recent "translation" it doesn't bode well for the maintainance of the code.
ISO C is even closer to ANSI than ANSI is to K&R - basically, well-written ANSI C is ISO C.
I agree that knowing C++ helps you write better C... but I wouldn't learn C++ if you don't already know C or Java - it's too big a bite to take at once.
Not just for tulip, but for all his plethora of ethernet drivers.
But tulip is the one I use, not my "almost a tulip chip" linksys card. I wrote to Becker (instead of testing out beta drivers like I should have..) when my card didn't work with the in-kernel driver at the time - he helpfully pointed me to a newer version that works flawlessly.
I sincerely hope that the "Best Newbie Helper" award doesn't go to some website maintainer - as nice as some "Linux Newbie" websites are, they don't hold a candle to the tireless people who work "in the trenches", as it were.
It's the ready availability of newsgroup help that made Mindcraft's "we didn't find anyone to help us tune Linux" lies so blatant and offensive. It's that newsgroup help that makes Linux a bearable transition for people who can't scale the learning curve by themselves. It's that helpful attitude that gives the Linux community the glowing reputation for support it has.
The only non-newsgroup candidate I'd vote for here is Matt Welsh, for obvious reasons. But the LDP is only really helpful once you're comfortable enough with Linux to make use of that detailed information, and only if your question is a FAQ. For all the obscure, unanticipated, or just ultra-newbie questions out there, everybody turns to DejaNews. I'd like to see this award split between, say, the 5 or 10 most prolific posters to comp.os.linux.setup (the most spiritually-draining, most often needed, and most newbie-heavy place to be helpful).
I haven't frequented the group in a while, but years ago (so I'm not eligible; my tirelessness faded into tiredness before 1999 ever rolled around) I and a relative handful of others posted literally thousands of messages (which shocked the hell out of me the first time I searched for "roystgnr" on Deja) over the years that made the difference for people who got Linux running.
Not all newbies stay newbies, either. My last CIVI homework this fall was completed with the indispensible aid of SLFFEA, a program whose author was kind enough to name me and a few other C.O.L.S. folk in the credits. Makes me feel kind of guilty for leaving, now...
I'm still an undergrad, a dangerous state: not educated enough to really know what I'm talking about, but educated enough to think I might. ;-)
Anyway, I'm curious: I understand that current JPEG uses Fourier transforms (a full integral transform, or a Fourier series?) to get a spectral representation of the image data, then drops subtle information from that transformed data to get a similar image from the reverse transform.
So I assume the new JPEG (aside from all the quicktime-esque formatting overhead) uses the same technique, just with a different complete set of functions, the wavelets.
So my first question: is there anything about the above that I'm misunderstanding?
And my second question: What are wavelets? Bessel functions? Something I haven't heard of? Is there a simple formula, or a simple ODE generating wavelet solutions, that I could look at or plug into Maple? I gather that whatever they are they approximate discontinuous functions with much better convergence than sines and cosines... but that could describe a lot of things.
I thought I saw OpenVMS on the list of binary releases of M12. Granted, it's still buggier than Communicator, but it's a step up from Netscape 3.
Oh, and I suspect the reason ESR mentions Linux rather than "free operating systems" is the reason he talks about "open source software" instead of "free software" - marketing spin. Don't confuse the lusers with details, just give them a name they can wrap their heads around...
To the people replying to this comment with anti gun control views, you're missing the point.
I agree with ESR's stance against gun control, but I *don't* think that it should be something that gets tacked (unless it's a random sig?) onto the end of what is, essentially, a press release for the Linux community.
"ESR, spokesman for the Open Source community" and "ESR, anti-gun control advocate" are both fine positions to take, but I'd rather he not mix them. If he's going to start a letter summarizing our views and representing the whole of the OSS community, I'd rather he not finish it by making unrelated statements that many OSS users and authors disagree strongly with.
I don't think this specific instance was a problem - the quote was obviously a sig, random or not; and he didn't claim to be representing anyone but himself in the letter. Nevertheless, it makes me uneasy.
Oh gawd, EVERY time an XFree discussion starts, someone says the development is too closed.
Yup. Often it's me. I'm happy to see people are beating me to it now.
You know what? Shut up.
Let me think.. No.
But you have a nice day, too, OK?
I have one word for you and it is NDA.
Damn, that means I need two words to one-up you:
Modular programming.
You remember, that non-monolithic X Server design which is one of the biggest improvements in XFree86 4.0? If you've still got a reason why we shouldn't have, say, anonymous CVS access to the X server core and all the non-NDA drivers, then I'd like to hear it.
It's not *HARD* to become a developer, and anyone who complains about it being 'too closed' has obviously not even looked INTO the subject.
You don't seem to get it. I don't want to become a developer just so I can figure out why XFree86 has mouse input bugs on my machine and hardware cursor bugs on my girlfriend's. Another poster who wanted to work on Truetype support years ago didn't want to sign up to "be a developer" before he could even read a mailing list archive and see what was being worked on.
How many people currently doing the heavy work on the Linux kernel started by saying, "You know, I want to become a full-time Linux developer?" I'd like to see numbers, but I'll bet it's not nearly as many as those who started by saying, "I wonder if there's a driver in development for my foo card,", or "This discussion of the unified page cache on linux-kernel is interesting, I think I'll pull down last week's devel source and see how it works.", or even "This looks like a bug. I think I'll check the latest source and see if it's being fixed."
XFree is not a 'closed' development, it's a controlled one.
Linux and FreeBSD are both controlled, but we get frequent releases (and more frequent prereleases) with the former, and anonymous CVS access with the latter.
And it has to be controlled, as NDAs exist, like them or not.
Needless to say, I don't. But this is a red herring; nobody said we wanted up-to-date source code on NDA'd drivers, just on the 99% of XFree86 that is free software.
To be fair to the XFree86 developers, the 4.0 pre-releases are coming more frequently than they have in the past... but the situation still isn't as good as it should be. To be more fair to the XFree86 developers, I'll point out I think it's a shame that they are underappreciated, even as they write a successful free software project that is arguably as complicated as and more important than the Linux kernel.
However, if they want more developers, they're not making enough of an effort to have a project that is open enough to be attractive to tenatively interested programmers.
And they do want more developers. They're asking for them strenuously enough on the web page. What was the initial timeline I heard last year, XFree86 4.0 by June? That's not quite a 100% slip from schedule, but it's close.
And I want them to have more developers! Between the complaints about configuring X, the emergence of DRI 3D and real GLX support, and the major architecture changes in 4.0, they've got one of the most important free software projects in existance on their hands, and I'd like to see them have every hand they can get.
Eric Raymond always uses gcc vs. egcs and FreeBSD vs. Linux to show how even among free software, projects which are "more open" than others tend to be more successful in the long run. I don't think it's worth a code fork to find out, but it's a shame there isn't anything competing with XFree86 to provide a third example.
Honestly, people complain about the "two-party" system, but that's not the problem America has. We can get third-party (and in many cases, fourth-party, fifth-party, etc.) candidates on the ticket, but people, even if they prefer the underdog to one of the major party candidates, still don't vote for him. Why?
.3245 votes to each candidate, so they could rank them by preference... but stupid people would think that was too complicated, and stupid people have a vast majority vote no matter what system you use to count them...
Because except for the "protest vote" value, it doesn't pay to vote for a candidate that doesn't have a good chance of winning. Whine all you want about how you're not "throwing your vote away", but this is a real problem.
Hypothetical election: If McCain (who's my favorite of the current candidates, BTW), Gore, and the Perfect Libertarian Geek candidate are on the ballot in the presidential election, I'm probably still going to vote for McCain, lest the Perfect Geek siphon off enough Republican votes to cause Gore to win. If 30% of the voters preferences run "McCain,Geek,Gore", 30% run "Geek,McCain,Gore", and 40% run "Gore,McCain,Geek", then if everybody voted for the candidate they liked best then you'd end up with *more* people unhappy with the election results.
There's an essay in _A Mathematician Reads the Newspaper_ (great book, worth buying or borrowing; sorry I forget the author) about the quirks of voter preferences in multi-candidate elections - in it the author presents a hypothetical election with 5 candidates, and 5 fair-sounding ways (including the plurality vote) of determining a winner.. except that a different candidate would be chosen the winner by each method.
People have talked about "None of the above" votes and other different systems that would force a new election... but we shouldn't *have* to have election after election to choose between the same group of candidates. And besides, while "None of the above" might help in cases where neither party puts up a good candidate, it does nothing for cases where a third party candidate is marginalized because of a close race between the major parties.
What I'd like to see, personally, is an election system where each voter can vote "Yes" or "No" for each candidate, and the candidate with the largest majority of "Yes" votes wins. (Or a new election is held if no candidate gets a majority). That way, if no party puts up a good candidate, then a new election must be held. If a good third party candidate is running in a race where the two leading parties are close in votes, then voters can safely vote for *both* the underdog and their favorite of the leading candidates, giving the third party entry a chance without risking "throwing their vote away". We wouldn't have to worry about the Greens "stealing Democrat votes" or the Libertarians "stealing Republican votes" anymore, and nobody would ever be "throwing their vote away". And if people were really fed up with all the candidates, they could vote "No" to all of them and get a new batch.
Of course, since such a system:
a. Is change, and therefore "scaaarry"
b. Is threatening to both the entrenched parties
it'll never happen.
And of course, what I'd *really* like to see is the ability for people to give 0 votes, 1 vote, or
...people from faking their client keys?
/proc/kmem byte by byte, you can pull out that data somehow.
Maybe I don't understand the scheme you're proposing, so bear with me.
Proving that someone has a "blessed" client sounds theoretically impossible, for many of the same reasons why creating an uncrackable copy protection scheme in software is impossible. You can't have a perfectly hidden private or symmetric key in your "blessed" client, because to use that key the client has to decrypt it sometime, which implies:
A. The algorithm for decrypting it is there, in the client code. Making the client closed source may make it more difficult, but no less possible, to recover/reproduce this algorithm.
B. The decrypted key is in memory sometime. Whether you run the "blessed" client through a debugger, or halt its execution and examine
There are more ingenious techniques, I suppose - someone mentioned running a "blessed" client and using your cheating client as a proxy between it and the server, passing any "key requests" or "checksum requests" to the blessed client while handling most of the gameplay with the cheating program.
Two more points, as long as I'm posting:
There have been near invincible client-side bots on public servers since I started playing Quake over two years ago, with all the aim improving, sight improving, etc. cheats that Carmack outlined. Closed source didn't do anything to stop them. In other words, to the idiot QW player who whined that he was never buying another Id game: shut up; it's not Id's fault.
The best way to prevent cheating from a client-side bot is to have the client-server protocol such that the client is completely untrusted. Unfortunately, this isn't perfect:
Just because it prevents the client from cheating doesn't mean it prevents the client from being a borg or a bot. You can make a Quake game such that borg clients can't see around corners, but you can't make one such that borgs can't have perfect night vision, perfect (aside from dodging projectile weapons) aim, etc.
There are technical complications in realtime games to making the client completely untrusted. Quick example: Not sending the Quake client data on opponents who are around a corner means that the client can't do any local prediction on that player's movement, which means that you're subject to the full 200+ ms modem lag before the opponent becomes visible after he rounds the corner. I've already been too spoiled as a LPB to enjoy modem FPS games; this would just make them near-unplayable.
Having a number of "blessed" clients as you've suggested is the perfect way to prevent cheating, but short of a magic uncrackable piece of hardware to locally verify the "blessed" client status, I don't see any way of preventing people from creating cheating clients. Closed source makes it harder, but no less possible.
Lets look at the options:
.tif files of every document I wanted to scan. I'd move to completely electronic filing (except for those documents where someone/some government agency would want the original copy of something, of course), if only I could throw a search engine and editing capability around it. I'm a pack rat - I've never deleted a nontrivial personal email. But to try and do that with every important paper doc that crosses my path would be impossible without some better way of organizing/indexing them, and that means OCR.
Handwriting recognition:
Great, so far as it helps with OCR on handwritten pages. Not so great, as the primary interface to a handheld computer. As much as I prefer the Palm Pilot as a PDA, I'd much rather have a tiny, two-finger typeable keyboard than use graffiti (or the onscreen keyboard alternative). Graffiti just doesn't work as quickly for me and feels more awkward. And recognizing "natural" handwriting would be even worse. My handwriting goes from left to right; it doesn't stay within the sensory area on a PDA. The "unnatural" graffiti writing actually helps because it eliminates that instinct to move your hand while writing.
Voice recognition:
Sure, would have been great to dictate homework when I was in the 6th grade and didn't know how to type. But today it would have to get punctuation, spelling of homonyms, etc. exactly the way I wanted it to be marginally better than typing long paragraphs. And as for voice commands? Forget it. It's been like three years since I've had my computer situated out of hearing range of any other computer around, and I don't want to work within some wacky scheme to decide which computer is supposed to be listening to me.
OCR:
Now, this would be useful. I'd love to live a paperless life, but unfortunately I live within a society that doesn't see things my way. It would be nice for my "filing cabinet" to be completely electronic, but today that's not really a good option unless I save
Got this in my mailbox this morning from the Ditherati mailing list:
... UM, WE FORGOT WHAT COMES NEXT
PICK IT, BOX IT
"We sat down yesterday and felt that there were some orders that we were not able to ship for Christmas."
Toysrus.com CEO John Barbour, on the lack of fulfillment felt by his company's warehouse workers during the holiday rush, The New York Times, 23 December 1999
It just is fairly new, and isn't used as well as it should be.
/etc/shadow to the evil cracker because it'll have wisely dropped it's file access priviledges and switched to a non-priviledged uid immediately upon execution, a feat now possible because it can use the capabilities support to retain the low port opening privileges with the new uid.
I'm referring in particular to capabilities support (in the latest kernels) and ACL support (in beta testing on top of ext2; I hope it gets into the final ext3!)
Look at the past. Linux (and every OS out there, but that's beside the point) has its security undermined fairly regularly by buffer overflows, etc. discovered in various daemons and suid programs. Rough estimate, it seems like there's a remote root exploit every year or two, and more than one local root exploit each year.
Yes, the patches come within hours or days of the exploits, and yes, anyone who can type "rpm -F" can keep their system up to date with those patches. But there's still that window of vulnerability out there, and there's still the (millions of?) Linux systems out there without root users experienced enough to stay up to date with security patches.
Security will hopefully be much improved in the future will be the use of capabilities in priviledged Linux programs and ACLs in Linux distributions to drop all unrequired permissions. Right now, if a program just needs to open a trusted (< 1024) TCP port, or read files with strict permissions, or have raw access to video hardware... then that program gets run as root, with the full array of root permissions.
And then if that program has a security flaw, then anyone in a position to exploit it has root.
With capabilities, a program (or a wrapper program) can be run as root, but can permanently or temporarily drop selected root capabilities. In other words, if a capabilities-aware suid root program just needs to listen at a low port, but it can be tricked by the user into opening arbitrary files... well, then it still won't be able to spit out
Well, that was a mangled description, but you get the idea. There are links to discussions of the subject by people who know what they're talking about at http://www.millenniumproductsllc.com/sjp/
It just is fairly new, and isn't used as well as it should be.
I'm referring in particular to capabilities support (in the latest kernels) and ACL support (in beta testing on top of ext2; I hope it gets into the final ext3!)
Look at the past. Linux (and every OS out there, but that's beside the point) has its security undermined fairly regularly by buffer overflows, etc. discovered in various daemons and suid programs. Rough estimate, it seems like there's a remote root exploit every year or two, and more than one local root exploit each year.
Yes, the patches come within hours or days of the exploits, and yes, anyone who can type "rpm -F" can keep their system up to date with those patches. But there's still that window of vulnerability out there, and there's still the (millions of?) Linux systems out there without root users experienced enough to stay up to date with security patches.
Security will hopefully be much improved in the future will be the use of capabilities in priviledged Linux programs and ACLs in Linux distributions to drop all unrequired permissions. Right now, if a program just needs to open a trusted (http://www.millenniumproductsllc.com/sjp/
It's not that Linux didn't have some of the same fundamental security bugs that Windows did way back when; Linux was just better about fixing them.
When the string of bogus IP attacks against Windows came out two years ago (teardrop & kin) Linux was vulnerable to the first of them too... but the Linux patch was out within a day or two, whereas Windows was vulnerable for months. To boot, the Linux patch fixed all the similar TCP/IP stack problems, whereas Windows ended up being vulnerable to syndrop, newtear, and a sequence of nearly identical attacks.
...hopefully, is that LinuxOne *will* crash and burn, and anyone stupid enough to use it as an anti-linux datum will be quickly pointed to the denunciations on Slashdot and LinuxWorld, from Bruce Perens, and from basically anyone else who knows what they're talking about.
I mean, honestly, LinuxOne's name has been mud from the very day their existance became known! If someone invests in them because "it's got Linux in the name!" but doesn't bother to do so much as a web search to research the company, they deserve what they get.
Quake game mods are written in QC, but turned into data to be processed by the main code. I think the spirit of the GPL would want that code to be released, but it is only a small step from there to saying that every program loaded by a GPL operating system must be GPL, which is clearly not the case.
No, Quake game mods are not required to be under the GPL, at least not if my understanding of the way they work is correct. It's either an interpreter or a JIT compiler, correct? Well, just because an interpreter (like bash) is under the GPL doesn't mean that all the scripts it interprets have to be under the same license. Just because the GCC compiler suite is GPL'ed doesn't mean that the code you compile with it has to be.
The GPL does apply to both statically and dynamically linked binaries, however, so Quake 2 game modifications would have to be GPL'ed if Quake 2 itself was released under only the GPL. If you released Quake 2 in the future under the LGPL then binary-only modules would be fine.
The Quake 2 case is interesting, though, since it's already released under a license that (if not explicitly, than by years of practice) allows binary-only DLLs to be released; and rereleasing it under the GPL wouldn't take away that right for people who purchased the original license. So would people who *hadn't* purchased the original game be allowed to write binary-only modules for a GPL'ed Quake 2? I haven't a clue. Ask RMS.
I must first put up a disclaimer of total ignorance: I'm not an economist, not a historian, and not even a closely interested observer. But I do have a point to make, and a question:
I (like most political pundits I've heard) was under the general impression that the primary cause of Russia's economic problems was simply endemic corruption caused by a too-fast attempt to switch to a free market system.
I'm not so sure anymore. Newsweek claimed recently that over half of the Russian economy was based on barter, as one of many figures they mentioned to demonstrate to readers that Russia's economy is in serious trouble.
I think it has more significance than that, though. People don't think to themselves, "I think I'll trade goods without paper money, because it will symbolize how screwed up our economy is." People barter because they don't have a trusted medium of exchange that they can use instead: either they don't have enough cash, or they don't have any cash they can trust.
We refer to "the Great Depression" instead of "the 1930 Recession" in good part because the U.S. government at the time didn't have a clue how to handle money; when the recession started, in a misguided attempt to preemptively prevent hyperinflation, the government cut back the money supply. Drastically. By over 50% on some indicators (It's been a while since I read this, and I don't so much as remember the difference between M1 and M2 now - the curious can check figures for themselves), and unemployment and GDP (in constant dollars) pretty much dropped in sync with the reduction in the money supply.
OK, I know this is simplistic, I know there were other major contributing factors (e.g. massive trade restrictions/tariffs) to the Great Depression... but the money supply may have been the largest single factor.
And it looks to me like something similar is happening in Russia. *Something* is causing people there to barter, whether it's a black market (which would imply those trade restrictions again), a government reduction in the money supply (trying to fight inflation, perhaps?) or simply the chain reaction of people pulling their money out of the country to avoid what they see as a financially dangerous situation.
Would someone who knows more about the situation in Russia than I do care to comment? Even if this was the problem, I'd have no idea how to solve it. Increasing the supply of fiat money (as eventually happened in the US when the government had to spend big in the late 30s and the 40s) would only work if inflation is in the single digits. Foreign aid and investment hasn't seemed to make a dent so far.
And this is a horrible situation, both for it's own sake and for what it implies for the future. People have already compared 1990's Russia to 1930's Germany, but even if Russia didn't end up in a political collapse, what would their economy's continued failure say about the development of a free market? That it's possible to kick an industrial economy into a depression state from which there is no return? That's almost too frightening to think about.
AMD Will slash Athlon Prices: This is one of the biggest misconceptions about
The Athlon 600 is retailing for as little as $379 now, about half of what they were at four months ago. Granted, it wasn't in one dramatic price slash, but it's a hell of a drop. AMD has kept the Athlons cheaper and faster than same-Mhz PIIIs forever, and I don't see that changing soon. Maybe Intel will get ahead in the Mhz war, but that won't change the market situation for people not buying the absolute fastest chip.
Enough with the oxymoronic names already! "Pentium IV"? Please. You'd think that Intel could make it's chip names correspond a little better to chip differences, too. The only significant difference between a PII and a PIII is the (little used) SSE support, whereas you have to shop carefully between PIII releases to make sure you get one of the uber-cache "Coppermine" (disclaimer: no copper included) models that actually can compete with an Athlon.
It's a shame that Pentium IIIs, while invariably poor consumer buys, are still doing well because of marketing. Want a cheap computer? Get a Celeron. Want a fast computer? Get an Athlon, which will outrun any PIII of the same price.
A neighbor of mine got their first computer last Christmas. Maybe my advice got there too late, because their overpriced, unstable Compaq PII system was coupled with a lousy ATI video card and a winmodem. I had the pleasure last month of trying to figure out driver problems with someone's dual PIII, 128MB (or was it 256?) RAM, Voodoo 3 3500, behemoth system... that was having conflicts with his ISA network card.
Granted, there are a lot of CPU-intensive things (Quake 3 - intense!) out there I'd love to have an 800Mhz Athlon for, but those are the exception (Quake 3 - exceptional!) rather than the rule. (Quake 3 rules!)
And even in 3D games, we're getting to the point where the processor won't be the limiting factor anymore. I mean, once you're getting 60fps at 1024x768, do you shoot for 90fps? No, you crank up your resolution, detail, or color depth. And once you do that you're worrying about your video card's fill rate, texture memory, or texture RAM bandwidth, not about your CPU. Hell, with T&L going on new video cards, we're going to be at the point where the AGP bus and human perception are the limiting factors on visual quality, not the CPU. Maybe games like Halo will have ultra-impressive physics and AI to use all those CPU cycles, but I'd like to see it.
Granted, there are always going to be applications (if only Beowulf clusters) where CPU value increases linearly with CPU speed.. but these are niche things compared to the huge consumer market being sold to today.
What I want to see isn't faster CPUs, it's broadband access! I have a K6II on a 10baseT college dorm connection this year, and if I have to trade it for a quad 1.2Ghz Athlon on a 56K modem when I graduate, it's not going to be a good trade. Whoever thought up Intel's "our CPU makes the internet faster!" campaign should be flogged.
No, you don't lose a point when you're fragged. You can have 200 deaths and 100 kills and you'll still beat the guy who got 99 kills but didn't die. (note: I have a history of getting 90% of the kills and 50% of the deaths of people who regularly beat me, so maybe I'm biased...) And not only are the repercussions next to nothing, but they're being adjusted even lower in Quake 3 for the sake of "game balance" (read: keeping newbies from becoming frustrated). New respawns start with 125 health instead of 100 so they can't be killed in one hit. The default weapon is a serious machine gun instead of a joke blaster. All the non-default weapons are quite good, and easy to get a hold of fast. No more runes in CTF. No more power armor. Even the default Q3test policy of forcing respawns to wait until after your old corpse hit the ground is gone!
They're not just a beige box builder that added Linux to their name, they actually do know the OS and what kind of hardware to put it on well. When we got a Linux workstation last spring, Gateway considered it a "special engineering" cost of ~$500, Dell charged as much as for NT, and both of them offered a selection of adequate but poorly-selected-for-Linux hardware, going with Adaptec over Buslogic, with some poor video card over Matrox... just basically taking their Windows workstation and slapping Red Hat on instead. I'm sure the big vendors are putting Linux prices in line with reality by now, but they're probably still shipping hardware with reverse-engineered drivers.