You've got two cards with two different PCI IDs and bus numbers.
That's not the problem.
Presuming they're doing interleaved frame processing (you simply cannot do scan line interleaving -- things as simple as anti-aliasing make that impossible), then you'll have every other frame rendered by each card.
You realize that the cards don't produce the exact same output, right? You'd probably get a headache from trying to visually process the subtle differences that occurred between each successive frame (but weren't too different from the alternating earlier frame). Ow.
And when it comes to randomized effects, such as used in some pixel shader methods, having two wildly different pseudorandom number generator algorithms would cause interesting effects.
I could see using two different cards from the same processor family, and maybe (but doubtfully) two different cards from the same manufacturer with different chipsets (e.g. -- GeForce3 and GeForce4), but beyond that I can't see a way to do it and not cause problems.
Of course, this presumes they're doing frame interleaving. But I can't think of any other way to do anything like this, and any other possible way would cause more problems, not less.
No, the top-of-the-line $500 ATI card is the new Radeon X800 XT PE (which isn't actually shipping yet, but the $400 X800 Pro is).
You can get an ATI Radeon 9800 XT 256MB (last generation) for around $400, an ATI Radeon 9800 Pro 128MB (1/2 generation back) for around $200, an ATI Radeon 9700 128MB (1 generation back) for around $180. A Radeon 9600 XT 128MB (last generation, mid-level card) is under $150. All are quite capable of running D3 in Windows at adequate frame rates... certainly the best buy is the 9600XT, and it'll fall in price in the coming weeks as well.
Right now is a very, very bad time to buy a video card -- both nVidia and ATI have just paper-released their new cards, so the old cards haven't fallen in price yet and the new cards will be top dollar when you buy them. Unless you absolutely, positively need a new video card right now (like your old one died), then wait until you need to upgrade. If you want to play D3/HL2 with the goodies on, then you'll need to upgrade if you don't have at least a Radeon 9500 Pro (GeForce4's will not cut it).
I agree with your recommendations on which card to buy, although I'd really think twice about buying the last generation of nVidia cards. If you're set on nVidia then spend the money for this generation -- it's a much, much better card and chipset.
If I had to go up against a grizzly bear, I'd rather have nothing but a thong and a Desert Eagle than one of these wacky contraptions.
I can see your point, but I hope your plan is to throw the thong in one direction, the gun in another, and run like hell.
Because any other plan when facing a pissed off Grizzly is going to get you dead. The Desert Eagle doesn't have enough stopping power... at least not unless you're a sharpshooter and happen to know where to shoot. You can unload the entire clip into a grizzly and kill it, sure, but it's going to live long enough to get to you and kill you.
A decent DELL is like $400 and you get the l33t Dell case. Just put in some real RAM, a burner, and a new AGP card.
Er... show me a $400 Dell that's worth upgrading only the RAM, video, and optical drive. At that price you're looking at low-end Celeron units w/ 128MB of memory and built-in video. It's probably cheaper than what I could put together from parts, but my parts will be of considerably higher quality.
As per you, to get that Dell up to speed you'll need to buy a new video card ($130), burner ($80), and memory ($250). Total cost at this point is $910.
Or you can, for $115 more, have an Athlon64 3000+, quiet HSF (Zalman), top end motherboard, an 8-in-1 reader, vastly improved sound quality, faster memory, firewire, more USB ports, no issues with integrated video, a better motherboard, and probably some other stuff I forgot.
Two weeks ago I bought, for $750 shipped, a case w/ 420W PSU, Athlon64 3000+, Zalman HSF, Chaintech ZNF3-150 MB, 1 GB DDR-400 memory, and a DVD+/-RW 8x burner. Toss in another $270 for a HD, video, monitor, keyboard, and mouse and you have a respectible gaming machine. You can quibble with the video card, but it's the same price for either machine to improve it. Except that a higher end video card will rapidly outstrip the Celeron 2400 CPU in the Dell... not so with the Ath64 3000+.
To get a roughly equivalent system from Dell, BTW, is around $2000. If you strip it down and buy the parts you suggest OEM then it still around $1500. Oh, and my case is quite a bit nicer than the one from Dell.
Why dedicate time (and presumably money) to continue the lock-in Microsoft Office and similar apps have in the workplace
Because the lock exists. Wishing it away won't change things. Contrary to popular belief, the world will not beat a path to your door if you build a better mousetrap. Not if everyone uses the old mousetrap, it works "good enough", and your new mousetrap won't fit through standard doors.
In the business world migrating desktops from Windows to anything else is problematic at best. Do you have any idea how many companies run proprietary desktop applications? For which they may not have the source (or it was lost long ago or they no longer have developers for the app)? Or even if they do have the source and the developers, how long do you think it would take to port the app from Windows to Linux? There's also the legion of niche programs that run under Windows, and the volumes of Excel macros and the like. Most businesses will kindly decline to move their desktops to Linux when the finance department informs management that they won't be able to bill customers anymore.
Home use means games -- both educational and entertainment. WINE and TransGaming are getting better, but they're still a long, long way from complete here. And the compatibility list is starting to slide backwards as more games utilize DirectX 9, which hasn't been touched in Linux yet. There's also the minor issue that Joe Schmoe wants to be able to buy a program off the shelf from CompUSA or Amazon or wherever, bring it home, put in the CD, and have it install. Why should they need to know if it's Linux or Windows? Why should they care? You can dismiss Joe Schmoe if you'd like -- it just means that you don't really care if Linux takes over the desktop or not (which is fine, but then this entire thread is irrelevant to you; that's not the subject at hand). Because it won't unless these issues are addressed.
the same stem cell technology is being used to replace teeth can replace hair follicles.
Not necessarily. According to the Guardian piece the stem cells are taken from the patient themselves, but it doesn't say where the stem cells originate. I'm certainly not an expert in the field, but there was a really good episode of the PBS show Innovations on stem cell research recently. It talked about spinal cord repair using nerve stem cells from the nose (yes, you have nerve stem cells in your nose. No, I had no idea either.) and heart muscle repair (post heart attack) using bone marrow stem cells.
Anyway, the deal is that not all stem cells are the same. There are differentiated ones and undifferentiated ones. The differentiated ones cannot be used to grow "any" other kind of cell -- at least, not that we've figured out yet. They have already specialized toward a kind of cell (for instance, nerve cells) and cannot grow other kinds of cells (like blood cells or muscle cells). AFAIK, most of the stem cells we still have after birth are these kind.
The undifferentiated stem cells are pretty much the holy grail. They can (in theory) be coaxed toward creating any kind of cell you want -- blood, muscle, nerve, tooth, hair, etc. Of course, there's the issue of getting them. I think some of the stem cells in the bone marrow are undifferentiated. I'm not aware of any others elsewhere in the body. But, heck, we weren't even aware of stem cells a few decades ago and I'm certainly not a medical researcher, so I could be dead wrong here.
All of that said -- whether or not this could be used for your balding head basically comes down to two things -- 1) are they using undifferentiated cells, 2) can we figure out and replicate the process that causes such stem cells to produce hair cells.
And I very much disagree that the hair replacement market will be a primary funding source -- it's going to be too expensive for some time to come. Surgery, even outpatient surgery, is usually not part of hair replacement, and there's no way to get to stem cells without at least some surgery.
I suspect most of it will come from cardiovascular and cancer research. Stem cell research is already looking extremely positive for heart attack treatment. So far every study done has given back 100% positive results. That's unheard of. And the treatment is relatively cheap to boot.
The cancer research comes in an opposite direction. Do you know what leukemia is? Essentially the stem cells in your bone marrow going haywire. We know that stem cells can regenerate other cells, but we really don't understand how, or why they occasionally malfunction. Which is a danger with using stem cell treatments, at least in theory. But if we can figure out how stem cells actually work then we can make some major steps toward fighting cancer.
Oh, and finally, none of this research is being done with fetal stem cells. It's all being done with the stem cells from the patient themselves. Which is a huge plus as far as rejection goes -- there simply won't be any. The only real advantage of fetal stem cell research is that there's a ton of undifferentiated stem cells in an embryo.
On Win98 you *had* to reformat regularly to keep performance up.
Shrug. I had a Win98SE box go nearly 3 years without a reformat. No noticeable performance degredation either.
With WinXP Pro I reformat when I do major hardware changes, and thats it.
Which isn't necessary either if you do a repair install instead (by "major" I'm presuming you mean a new MB and/or CPU; possibly graphics card seeing as how godawful the graphics card drivers are at properly uninstalling themselves. But there's a solution for that as well).
XP Home (why why WHY is it allowed to exist) are a different story
Uh... would you care to clarify why you feel XP Home is more vulnerable to needing a reinstall than XP Pro? The two have very minimal differences in functionality, none of it in the driver/registry/file system arena.
You can change motherboards in Windows without having to reinstall. You just have to do a repair install part of the way and it should be happy (should being the operative term).
I couldnt get the 2k machine to boot by using the cd or any form of rescue disk
So, uh... how did you reinstall 2k then?
Not that a repair install could've helped you at that point. Once you booted the HD with the new MB you were pretty well screwed.
More info on how to accomplish a MB swap w/o a full reinstall is available here and here.
Yes, Linux does it better. I'm not disputing that. But changing out a MB doesn't necessarily mean you must reinstall (it used to though).
Video game systems manufactured for sale in the United States after 2006 will include some sort of digital TV output. These digital TV sets introduce a significant latency into the chain.
The digital output is likely to be HDMI, which is essentially DVI + digital audio in one (smaller) cable.
The digital sets only introduce significant latency when fed an analog input that they have to up-rez to their native format. If you feed the TV with a native format digital input (most likely 1920x1080i or 1280x720p; 4:3 versions of the resolution will be acceptable as well) then they'll do nothing to the image and will not have any significant delays. That should solve the problem.
What if the audio is coming through your stereo, not the TV?
Then your stereo probably already has this feature. You just have to setup the delay per channel properly.
The inverse problem is a much bigger problem (audio coming out after the video), and actually much more common. Most receivers do a good bit of sound processing nowadays, and some can do so much that they'll end up delaying the audio signal by some fraction of a second. Thing is, they don't delay the video signal noticeably and you wind up with desynched audio/video that way as well. With no way to fix it.
why on earth would they be less-than-forthcoming if they didn't have some sort of adjenda of their own?
Never ascribe to malice that which can be explained by incompetence.
I'm not really trying to defend Diebold here, but a lot of their statements really do seem to be incompetence rather than scheming. They may simply be out of their league here.
Of course, some of the statements made by their CEO and other execs are so inane that we may be faced with a rare thing (at least in corporations) -- malice and incompetence.
If, after a dozen years, I decide to sue every maker and distributer of my corn, do I really have a case?
Yes. If you patented it then it doesn't matter who infringed upon it or for how long -- you can sue for back damages. At least that's true in the US. I believe it's also true in Europe, Canada, and (most of?) South America. I'm not even going to try and guess at Asian or African IP law (note - IANAL, I did take a IP law course in college though... pre DMCA/latest Copyright Act).
That said, if your corn is really everything you claim and it's pretty much eliminated all the competing strains (either by simply being better or by cross-pollination... more on the latter in a bit) then good bloody luck actually getting your money. Most governments would recognize that your corn is simply too vital to their economy to allow IP issues to strangle the farming industry. No, they won't kill you. But they will nationalize your patent. You'll probably get some compensation -- probably enough to make you and your heirs rich -- but you won't be able to bend everyone over and ream them for money.
You see this happen most commonly with drug patents. A good number of African and South American countries have revoked/nationalized the patent on key medicines (generally HIV treatments) because the drug companies were simply charging too much money (at least as far as the populace and government were concerned). Of course, you have to have some way to manufacture the drugs yourself because you can bet the real company isn't going to do it for you. A lot of companies will simply change their pricing structure rather than have the drug nationalized though. Of course, any country that does this is in danger of not only having the pharmaceutical companies pull out, but numerous other industries as well. Corporations really don't like it when you assume IP that they spent millions or billions of dollars in research.
Now, about that lack of a genetic terminator... if the corn is able to cross breed with other strains you may have a very difficult time levying any fines on hybrid strains. I don't know that there's any clear legal rulings on this -- yes, your patented work was used to produce the hybrid, but the hybrid is no longer your patented work. Patents must apply to a very specific implementation, and that hybrid is not the implementation. Patents are not copyrights -- you do not control any derivative works. It'd be an "interesting" case.
The discussion was in regards to standards, so for the purpose of the discussion standards compliance is the "best". Certainly it's not the only benchmark, but that wasn't the discussion.
What do you like so much about VC++?
Nothing. I was merely pointing out that the clueless twit that started the thread was wrong and that VC++ has become very standards compliant (as opposed to attempting to change the ISO C++ standard to fit VC++ 6.0's deeply broken model).
Like I've said several times now, we use g++ for our builds. It's a lot better than the other option we have (VisualAge) for our platform, at least as far as standards go (VA might be usable except that v5 didn't handle templates worth a crap; haven't had the time to test v6. Compile time difference wasn't very significant, and we never tested execution time since our boxes are too damn slow to actually build with optimization enabled. So glad management purchased hardware without getting developer input...).
Frankly, everyone is lucky to have gcc/g++ -- it's an excellent compiler that does a great job of implementing standards. It's not the best, but it's better than a lot (cough - VisualAge - cough).
My statement was simply that -- it's not the best, and that MS VC++ does a better job now. Which it does. If you use boost as a benchmark (and while I'm a fan of boost, and we use it in our codebase, it's simply not a real STL or C++ benchmark) then MSVC++ 7.1 does a better job. From what I understand, it does better on the Plum Hall tests than g++ as well.
I do wish gcc 3.4.0 would be finalized. I'm very interested in running it in our environment (AIX 5.1), particularly since shared libraries don't work properly w/ 3.3.2 (throw an exception in a shared library and watch the program crash), not to mention the precompiled headers and other speed ups. And, no, upgrading to a non-final 3.4.0 is not an option. Frankly, I'm leary about 3.4.0 and may just wait for 3.4.1, since there's inevitably a lot of bugfixes in the first revision -- particularly when talking about less used platforms like AIX (God I hate AIX, but that's not my decision to make...)
Am I the only one that sees this as buying out the competition?
Which is a long held tradition in the business world. With the caveat that you cannot always do it -- there's that irritating little bit about freedom of choice. If someone doesn't want to sell their business to you (or, more particularly in this case) work for you, they don't have to.
And some guy is making the world better by furthering a standard. Let's hire him so that our C++ becomes the only stardard the world must follow.
Of course, the reality here is that VC++ was close to rock bottom when it came to meeting the ISO C++ standards (particularly in regards to the STL). Since Herb Sutter was hired by MS they have drastically improved compatibility with the standard, both in the compiler and in their STL. I don't think they're the most compliant, but they're a damn sight better than a lot of other compilers, GNU g++ included.
As for changing the standard for MS's benefit -- by merely stating that it proves that you have absolutely no clue how the ISO committees work, particularly when it comes to languages.
Oh, and in case you're wondering -- no, I don't use MS VC++. I code in Unix with g++. But I'm not a clueless moron.
Five seconds is probably not long enough to know what really happened.
I seriously doubt that.
For every 10 mph a car is moving it will move 15 ft/s. So a car going 60 mph covers 90 ft in one second. Five seconds gives you 450 ft of data at that nominal speed... at which point I seriously question your observational skills if you needed more time to evaluate the situation.
I'd like to thank Steve Jackson Games and Car Wars for that odd bit of knowledge, and for bothering to even do the physics right in a game that largely ignores physics otherwise.
At what point do you reach a "punishment saturation", and from then on, just start ruining their lives?
I dunno. But why don't we ask the dead guy how his life is going at this point? Or any family members?
This isn't the first time the perpetrator was caught exceeding the speed limit either. The prior incident just destroyed a couple cars -- expensive, but no big deal. Destroying someone else's life is another matter.
Don't get me started about revenge, it's among the most primitive and WRONG emotions a human being can have.
Agreed, but what if it's not about revenge? What if it's about protecting society from someone who is (apparantly) incapable of controlling themselves and/or understanding the consequences of their actions? It's not called "reckless driving" for no reason.
Sorry, seen too many idiots who don't realize that a misused vehicle is just as much a deadly weapon as a gun. Either, when used correctly and appropriately, is fine by me. But this guy didn't use it correctly.
And yeah, I used to be a dipshit driver too... I never did anything close to this (3x the speed limit), but I know I did some stupid stuff. And if I had ever killed someone in the proces I would expect to have been sent to prison for a long, long time.
It's called being an adult and taking responsibility for your own actions. If you're not willing to do so, then I'd suggest giving up the other trappings of adult life -- because you don't deserve them.
Yes, but have you figured out why it's CPU limited? It's not the AI and Physics. There are other games out there with superior AI and Physics (like FarCry, or even UT2k4) that aren't CPU limited.
Here's a quote from a recent thread on BluesNews (see post 48):
2 - Halo is fill rate limited because Bungie did an outstanding job of using the specialized pixel shader hardware on the Xbox.
Since the PC cards lack the specialized pixel shader hardware the CPU has to do it all. That makes the game heavily CPU bound. Of course, if a graphics card -- like, oh say, the GeForce 6 -- can replicate those pixel shader commands programatically then the GPU can take it over and dramatically increase frame rate.
Funny that it's Halo that the GF6 sees the biggest framerate jump...
It's not like those other games are using the hardware shaders yet anyway (or are they?).
They are -- FarCry is probably the most intensive game out there right now, fully utilizing DX9 specs. Halo is no slouch either, although a lot of its speed issues are from wanting to use hardware that simply isn't present (on PCs -- it is on the Xbox; why they didn't port away from this is beyond me).
Aquanox 2, Tomb Raider: Angel of Darkness, Painkiller, UT2k4, BF: Vietnam, and several others utilize DX9 to varying lengths as well. And there's the upcoming games -- Half Life 2, STALKER, Soldner (with an umlaut on the o), World of Warcraft, Everquest 2, and numerous others.
Quake 3 simply isn't a reliable benchmark anymore. It utterly fails to excercise the newer features of the cards -- which are really the only features to bother upgrading for. If all you're going to do is play Q3-era games then a GeForce2 is more than sufficient. If you want to run games already out, and those coming out in the next year, with all the graphical options turned up and at high-res then you'll be best served by either the latest nVidia or (probably) ATI card.
And (most importantly to me, and many others) if you want to get a card that can run new games at reasonable resolutions with most of the graphical bells and whistles on, but at a reasonable price... well, those $400 cards are going to be sub-$200 very quickly now, and the $200 cards are going to drop to around $100.
When I say "Unix Machines", I am referring to Sun, SGI, HP, and IBM Unix based hardware.
Oh yes... IBM hardware is such a deal. I'm so everjoyed that we run on IBM hardware and AIX... so much so that I could just about fucking kill myself.
Our development box was originally a B50 w/ single CPU and 512MB of RAM (it's since been upgraded all the way to a dual CPU B80 w/ 1 GB RAM -- which is still slower than the comparison system). It compiled code at roughly half the speed of a P2-400 running Redhat 7.2. This is using gcc on both platforms. VisualAge was slightly faster, except that it was a total fucking nightmare to actually use because xlC 5 can't generate template functions worth a crap. We would have to maintain the template prototypes ourselves, which became an increasing pain in the ass as our libraries grew.
Of course, AIX's linker is like nothing else on this world, so forget using a lot of open source tools. I'm still waiting on a debugger that will actually work -- both dbx (from IBM) and gdb simply core dump -- usually during running, but occasionally when they just try to load the code. TotalView works, but it's an additional cost. And since we upgraded to AIX 5.1 we've had to revert to static linking -- because dynamic linking is FUBAR'd with gcc (allegedly fixed in 3.4, but we'll see). There's a patch we could try, but compiling gcc on this blazingly fast box takes 20 hours.
Yeah, I'm really glad we spent tens of thousands of dollars on production, QA, and development servers (the QA and dev servers were virtually free, but the production ones were not) that are so abymally slow and infernally different from everything else. Even if a half dozen server class PCs cost the same as these boxes they would've been at least a magnitude faster and we'd have far fewer issues supporting them, not to mention how much developer time would've been saved at this point.
Oh, and we've just moved to (puke) StarTeam for our version control. Of course, it doesn't have a native AIX client, so we're having to work around the fact that the universal java client doesn't recognize executable files as such during checkin. This problem doesn't exist on Linux.
And yes, RedHat does self-destruct at the slightest provocation.
I'm sorry you have such lousy admins. None of the Redhat boxes we have, or that friends have at their jobs, seem to have any of the issues you mention.
I'm not even *going* to mention RPM hell.
Why, because you've found some mystical, magical fix for it? The reality is that RPM hell is a symptom of a larger problem -- that different programs often want different versions of shared libraries. The more programs you install on any one given box the more likely you'll have this problem. And if you don't have a package manager like RPM (or any one of the others) then you're not going to avoid this problem -- instead you'll overwrite a library without knowing that you're causing a problem. Have fun diagnosing that one!
I see a great many people buying hardware from Cisco's competitors in the near-future.
What makes you think that they don't have a backdoor username/pw as well? It may not be hard coded (they could both be strings that are determined by a hash function, based on the date/time or some other changing value), but I'd bet you they're there, at least on any high end equipment. Why? So that the damn thing is supportable remotely... even after some idiot admin screws up everything else. And, no, resetting the firmware on these things to restore the default admin password isn't acceptable -- simply because in doing so you'd lose all the other settings (bad for two reasons -- 1) they usually take hours or days to setup correctly, 2) if you're accessing the box for support, you probably want to see what the hell happened in case it was a bug).
You're presuming (most likely incorrectly) that the new user will keep the computer on 24/7 (and yes, I'm assuming that the MP3 server or network firewall will be on 24/7... which is likely to be true). Most people will only turn the computer on when needed and turn it off afterwards. That dramatically reduces energy usage.
There's the whole "benefit of use" issue as well, but that's murky at best, and telling others how they should use (or not use) their own belongings on a "moral" or "ethical" basis is the most slippery of slopes.
Look, it's pretty simple. If you don't trust the precompiled binaries they have on the machines, then why on earth would you trust the compiler they provide? I'm not talking about not trusting gcc... it would be fairly trivial to produce a hacked gcc that compiles the code in question differently (or simply compiles an embedded version of the code). And you'd have no way of knowing.
Heck, hack the diff tool for that matter. Either reject any schmuck who actually tries to pull this, or replace their source with yours while diffing. How thick is that tinfoil beanie anyway?
Oh, and don't think that a mechanical system is any better either. After all, some machine has to read those ballots and you could just compromise it!
Sure, I guess you could go to hand counting. That'll only take a few weeks to verify the results. No matter... I'm sure the rest of the world will understand that we're just too damned incompetent to use modern balloting techniques. I'm sure it won't have any impact on domestic markets either... nope... just because we don't think our products are good enough to use doesn't mean you shouldn't buy them!
Note that you won't get the full range of the subwoofer with a DVD or CD. Both have a bottom limit of 20 Hz, while the sub in question can allegedly output down to 10 Hz flat (and well below that with fall off).
You'd have to use vinyl to get frequencies that low. And have one hell of a damping system for the player to avoid it being shaken by the sound it's reproducing...
Not a vinyl snob (all of my music is on CD and ripped to MP3), but there are technical limitations. Of course, 20 Hz is far lower than most people's stereo systems can reproduce anyway. Below that you can't really hear it... but you can feel it.
Slashdot Mirror
You've got two cards with two different PCI IDs and bus numbers.
That's not the problem.
Presuming they're doing interleaved frame processing (you simply cannot do scan line interleaving -- things as simple as anti-aliasing make that impossible), then you'll have every other frame rendered by each card.
You realize that the cards don't produce the exact same output, right? You'd probably get a headache from trying to visually process the subtle differences that occurred between each successive frame (but weren't too different from the alternating earlier frame). Ow.
And when it comes to randomized effects, such as used in some pixel shader methods, having two wildly different pseudorandom number generator algorithms would cause interesting effects.
I could see using two different cards from the same processor family, and maybe (but doubtfully) two different cards from the same manufacturer with different chipsets (e.g. -- GeForce3 and GeForce4), but beyond that I can't see a way to do it and not cause problems.
Of course, this presumes they're doing frame interleaving. But I can't think of any other way to do anything like this, and any other possible way would cause more problems, not less.
The 9800 is the top-of-the-line $500 ATI card.
No, the top-of-the-line $500 ATI card is the new Radeon X800 XT PE (which isn't actually shipping yet, but the $400 X800 Pro is).
You can get an ATI Radeon 9800 XT 256MB (last generation) for around $400, an ATI Radeon 9800 Pro 128MB (1/2 generation back) for around $200, an ATI Radeon 9700 128MB (1 generation back) for around $180. A Radeon 9600 XT 128MB (last generation, mid-level card) is under $150. All are quite capable of running D3 in Windows at adequate frame rates... certainly the best buy is the 9600XT, and it'll fall in price in the coming weeks as well.
Right now is a very, very bad time to buy a video card -- both nVidia and ATI have just paper-released their new cards, so the old cards haven't fallen in price yet and the new cards will be top dollar when you buy them. Unless you absolutely, positively need a new video card right now (like your old one died), then wait until you need to upgrade. If you want to play D3/HL2 with the goodies on, then you'll need to upgrade if you don't have at least a Radeon 9500 Pro (GeForce4's will not cut it).
I agree with your recommendations on which card to buy, although I'd really think twice about buying the last generation of nVidia cards. If you're set on nVidia then spend the money for this generation -- it's a much, much better card and chipset.
If I had to go up against a grizzly bear, I'd rather have nothing but a thong and a Desert Eagle than one of these wacky contraptions.
I can see your point, but I hope your plan is to throw the thong in one direction, the gun in another, and run like hell.
Because any other plan when facing a pissed off Grizzly is going to get you dead. The Desert Eagle doesn't have enough stopping power... at least not unless you're a sharpshooter and happen to know where to shoot. You can unload the entire clip into a grizzly and kill it, sure, but it's going to live long enough to get to you and kill you.
A decent DELL is like $400 and you get the l33t Dell case. Just put in some real RAM, a burner, and a new AGP card.
Er... show me a $400 Dell that's worth upgrading only the RAM, video, and optical drive. At that price you're looking at low-end Celeron units w/ 128MB of memory and built-in video. It's probably cheaper than what I could put together from parts, but my parts will be of considerably higher quality.
As per you, to get that Dell up to speed you'll need to buy a new video card ($130), burner ($80), and memory ($250). Total cost at this point is $910.
Or you can, for $115 more, have an Athlon64 3000+, quiet HSF (Zalman), top end motherboard, an 8-in-1 reader, vastly improved sound quality, faster memory, firewire, more USB ports, no issues with integrated video, a better motherboard, and probably some other stuff I forgot.
Two weeks ago I bought, for $750 shipped, a case w/ 420W PSU, Athlon64 3000+, Zalman HSF, Chaintech ZNF3-150 MB, 1 GB DDR-400 memory, and a DVD+/-RW 8x burner. Toss in another $270 for a HD, video, monitor, keyboard, and mouse and you have a respectible gaming machine. You can quibble with the video card, but it's the same price for either machine to improve it. Except that a higher end video card will rapidly outstrip the Celeron 2400 CPU in the Dell... not so with the Ath64 3000+.
To get a roughly equivalent system from Dell, BTW, is around $2000. If you strip it down and buy the parts you suggest OEM then it still around $1500. Oh, and my case is quite a bit nicer than the one from Dell.
Why dedicate time (and presumably money) to continue the lock-in Microsoft Office and similar apps have in the workplace
Because the lock exists. Wishing it away won't change things. Contrary to popular belief, the world will not beat a path to your door if you build a better mousetrap. Not if everyone uses the old mousetrap, it works "good enough", and your new mousetrap won't fit through standard doors.
In the business world migrating desktops from Windows to anything else is problematic at best. Do you have any idea how many companies run proprietary desktop applications? For which they may not have the source (or it was lost long ago or they no longer have developers for the app)? Or even if they do have the source and the developers, how long do you think it would take to port the app from Windows to Linux? There's also the legion of niche programs that run under Windows, and the volumes of Excel macros and the like. Most businesses will kindly decline to move their desktops to Linux when the finance department informs management that they won't be able to bill customers anymore.
Home use means games -- both educational and entertainment. WINE and TransGaming are getting better, but they're still a long, long way from complete here. And the compatibility list is starting to slide backwards as more games utilize DirectX 9, which hasn't been touched in Linux yet. There's also the minor issue that Joe Schmoe wants to be able to buy a program off the shelf from CompUSA or Amazon or wherever, bring it home, put in the CD, and have it install. Why should they need to know if it's Linux or Windows? Why should they care? You can dismiss Joe Schmoe if you'd like -- it just means that you don't really care if Linux takes over the desktop or not (which is fine, but then this entire thread is irrelevant to you; that's not the subject at hand). Because it won't unless these issues are addressed.
the same stem cell technology is being used to replace teeth can replace hair follicles.
Not necessarily. According to the Guardian piece the stem cells are taken from the patient themselves, but it doesn't say where the stem cells originate. I'm certainly not an expert in the field, but there was a really good episode of the PBS show Innovations on stem cell research recently. It talked about spinal cord repair using nerve stem cells from the nose (yes, you have nerve stem cells in your nose. No, I had no idea either.) and heart muscle repair (post heart attack) using bone marrow stem cells.
Anyway, the deal is that not all stem cells are the same. There are differentiated ones and undifferentiated ones. The differentiated ones cannot be used to grow "any" other kind of cell -- at least, not that we've figured out yet. They have already specialized toward a kind of cell (for instance, nerve cells) and cannot grow other kinds of cells (like blood cells or muscle cells). AFAIK, most of the stem cells we still have after birth are these kind.
The undifferentiated stem cells are pretty much the holy grail. They can (in theory) be coaxed toward creating any kind of cell you want -- blood, muscle, nerve, tooth, hair, etc. Of course, there's the issue of getting them. I think some of the stem cells in the bone marrow are undifferentiated. I'm not aware of any others elsewhere in the body. But, heck, we weren't even aware of stem cells a few decades ago and I'm certainly not a medical researcher, so I could be dead wrong here.
All of that said -- whether or not this could be used for your balding head basically comes down to two things -- 1) are they using undifferentiated cells, 2) can we figure out and replicate the process that causes such stem cells to produce hair cells.
And I very much disagree that the hair replacement market will be a primary funding source -- it's going to be too expensive for some time to come. Surgery, even outpatient surgery, is usually not part of hair replacement, and there's no way to get to stem cells without at least some surgery.
I suspect most of it will come from cardiovascular and cancer research. Stem cell research is already looking extremely positive for heart attack treatment. So far every study done has given back 100% positive results. That's unheard of. And the treatment is relatively cheap to boot.
The cancer research comes in an opposite direction. Do you know what leukemia is? Essentially the stem cells in your bone marrow going haywire. We know that stem cells can regenerate other cells, but we really don't understand how, or why they occasionally malfunction. Which is a danger with using stem cell treatments, at least in theory. But if we can figure out how stem cells actually work then we can make some major steps toward fighting cancer.
Oh, and finally, none of this research is being done with fetal stem cells. It's all being done with the stem cells from the patient themselves. Which is a huge plus as far as rejection goes -- there simply won't be any. The only real advantage of fetal stem cell research is that there's a ton of undifferentiated stem cells in an embryo.
On Win98 you *had* to reformat regularly to keep performance up.
Shrug. I had a Win98SE box go nearly 3 years without a reformat. No noticeable performance degredation either.
With WinXP Pro I reformat when I do major hardware changes, and thats it.
Which isn't necessary either if you do a repair install instead (by "major" I'm presuming you mean a new MB and/or CPU; possibly graphics card seeing as how godawful the graphics card drivers are at properly uninstalling themselves. But there's a solution for that as well).
XP Home (why why WHY is it allowed to exist) are a different story
Uh... would you care to clarify why you feel XP Home is more vulnerable to needing a reinstall than XP Pro? The two have very minimal differences in functionality, none of it in the driver/registry/file system arena.
You can change motherboards in Windows without having to reinstall. You just have to do a repair install part of the way and it should be happy (should being the operative term).
I couldnt get the 2k machine to boot by using the cd or any form of rescue disk
So, uh... how did you reinstall 2k then?
Not that a repair install could've helped you at that point. Once you booted the HD with the new MB you were pretty well screwed.
More info on how to accomplish a MB swap w/o a full reinstall is available here and here.
Yes, Linux does it better. I'm not disputing that. But changing out a MB doesn't necessarily mean you must reinstall (it used to though).
Video game systems manufactured for sale in the United States after 2006 will include some sort of digital TV output. These digital TV sets introduce a significant latency into the chain.
The digital output is likely to be HDMI, which is essentially DVI + digital audio in one (smaller) cable.
The digital sets only introduce significant latency when fed an analog input that they have to up-rez to their native format. If you feed the TV with a native format digital input (most likely 1920x1080i or 1280x720p; 4:3 versions of the resolution will be acceptable as well) then they'll do nothing to the image and will not have any significant delays. That should solve the problem.
What if the audio is coming through your stereo, not the TV?
Then your stereo probably already has this feature. You just have to setup the delay per channel properly.
The inverse problem is a much bigger problem (audio coming out after the video), and actually much more common. Most receivers do a good bit of sound processing nowadays, and some can do so much that they'll end up delaying the audio signal by some fraction of a second. Thing is, they don't delay the video signal noticeably and you wind up with desynched audio/video that way as well. With no way to fix it.
why on earth would they be less-than-forthcoming if they didn't have some sort of adjenda of their own?
Never ascribe to malice that which can be explained by incompetence.
I'm not really trying to defend Diebold here, but a lot of their statements really do seem to be incompetence rather than scheming. They may simply be out of their league here.
Of course, some of the statements made by their CEO and other execs are so inane that we may be faced with a rare thing (at least in corporations) -- malice and incompetence.
If, after a dozen years, I decide to sue every maker and distributer of my corn, do I really have a case?
Yes. If you patented it then it doesn't matter who infringed upon it or for how long -- you can sue for back damages. At least that's true in the US. I believe it's also true in Europe, Canada, and (most of?) South America. I'm not even going to try and guess at Asian or African IP law (note - IANAL, I did take a IP law course in college though... pre DMCA/latest Copyright Act).
That said, if your corn is really everything you claim and it's pretty much eliminated all the competing strains (either by simply being better or by cross-pollination... more on the latter in a bit) then good bloody luck actually getting your money. Most governments would recognize that your corn is simply too vital to their economy to allow IP issues to strangle the farming industry. No, they won't kill you. But they will nationalize your patent. You'll probably get some compensation -- probably enough to make you and your heirs rich -- but you won't be able to bend everyone over and ream them for money.
You see this happen most commonly with drug patents. A good number of African and South American countries have revoked/nationalized the patent on key medicines (generally HIV treatments) because the drug companies were simply charging too much money (at least as far as the populace and government were concerned). Of course, you have to have some way to manufacture the drugs yourself because you can bet the real company isn't going to do it for you. A lot of companies will simply change their pricing structure rather than have the drug nationalized though. Of course, any country that does this is in danger of not only having the pharmaceutical companies pull out, but numerous other industries as well. Corporations really don't like it when you assume IP that they spent millions or billions of dollars in research.
Now, about that lack of a genetic terminator... if the corn is able to cross breed with other strains you may have a very difficult time levying any fines on hybrid strains. I don't know that there's any clear legal rulings on this -- yes, your patented work was used to produce the hybrid, but the hybrid is no longer your patented work. Patents must apply to a very specific implementation, and that hybrid is not the implementation. Patents are not copyrights -- you do not control any derivative works. It'd be an "interesting" case.
Best, what? Codegen? Features? Standards compliance?
The discussion was in regards to standards, so for the purpose of the discussion standards compliance is the "best". Certainly it's not the only benchmark, but that wasn't the discussion.
What do you like so much about VC++?
Nothing. I was merely pointing out that the clueless twit that started the thread was wrong and that VC++ has become very standards compliant (as opposed to attempting to change the ISO C++ standard to fit VC++ 6.0's deeply broken model).
Like I've said several times now, we use g++ for our builds. It's a lot better than the other option we have (VisualAge) for our platform, at least as far as standards go (VA might be usable except that v5 didn't handle templates worth a crap; haven't had the time to test v6. Compile time difference wasn't very significant, and we never tested execution time since our boxes are too damn slow to actually build with optimization enabled. So glad management purchased hardware without getting developer input...).
No, I lauded Borland for already bringing the development community a solid (and partially free) tool for use under Windows.
Ah... you mean one of the few C++ compilers that actually has worse standards compliance than VC++ 6.0?
Yeah... great choice of who to laud.
Then you've not been paying attention.
Actually, yes, I have been.
The Free Software community is lucky to have g++.
Frankly, everyone is lucky to have gcc/g++ -- it's an excellent compiler that does a great job of implementing standards. It's not the best, but it's better than a lot (cough - VisualAge - cough).
My statement was simply that -- it's not the best, and that MS VC++ does a better job now. Which it does. If you use boost as a benchmark (and while I'm a fan of boost, and we use it in our codebase, it's simply not a real STL or C++ benchmark) then MSVC++ 7.1 does a better job. From what I understand, it does better on the Plum Hall tests than g++ as well.
I do wish gcc 3.4.0 would be finalized. I'm very interested in running it in our environment (AIX 5.1), particularly since shared libraries don't work properly w/ 3.3.2 (throw an exception in a shared library and watch the program crash), not to mention the precompiled headers and other speed ups. And, no, upgrading to a non-final 3.4.0 is not an option. Frankly, I'm leary about 3.4.0 and may just wait for 3.4.1, since there's inevitably a lot of bugfixes in the first revision -- particularly when talking about less used platforms like AIX (God I hate AIX, but that's not my decision to make...)
Am I the only one that sees this as buying out the competition?
Which is a long held tradition in the business world. With the caveat that you cannot always do it -- there's that irritating little bit about freedom of choice. If someone doesn't want to sell their business to you (or, more particularly in this case) work for you, they don't have to.
And some guy is making the world better by furthering a standard. Let's hire him so that our C++ becomes the only stardard the world must follow.
Of course, the reality here is that VC++ was close to rock bottom when it came to meeting the ISO C++ standards (particularly in regards to the STL). Since Herb Sutter was hired by MS they have drastically improved compatibility with the standard, both in the compiler and in their STL. I don't think they're the most compliant, but they're a damn sight better than a lot of other compilers, GNU g++ included.
As for changing the standard for MS's benefit -- by merely stating that it proves that you have absolutely no clue how the ISO committees work, particularly when it comes to languages.
Oh, and in case you're wondering -- no, I don't use MS VC++. I code in Unix with g++. But I'm not a clueless moron.
Five seconds is probably not long enough to know what really happened.
I seriously doubt that.
For every 10 mph a car is moving it will move 15 ft/s. So a car going 60 mph covers 90 ft in one second. Five seconds gives you 450 ft of data at that nominal speed... at which point I seriously question your observational skills if you needed more time to evaluate the situation.
I'd like to thank Steve Jackson Games and Car Wars for that odd bit of knowledge, and for bothering to even do the physics right in a game that largely ignores physics otherwise.
At what point do you reach a "punishment saturation", and from then on, just start ruining their lives?
I dunno. But why don't we ask the dead guy how his life is going at this point? Or any family members?
This isn't the first time the perpetrator was caught exceeding the speed limit either. The prior incident just destroyed a couple cars -- expensive, but no big deal. Destroying someone else's life is another matter.
Don't get me started about revenge, it's among the most primitive and WRONG emotions a human being can have.
Agreed, but what if it's not about revenge? What if it's about protecting society from someone who is (apparantly) incapable of controlling themselves and/or understanding the consequences of their actions? It's not called "reckless driving" for no reason.
Sorry, seen too many idiots who don't realize that a misused vehicle is just as much a deadly weapon as a gun. Either, when used correctly and appropriately, is fine by me. But this guy didn't use it correctly.
And yeah, I used to be a dipshit driver too... I never did anything close to this (3x the speed limit), but I know I did some stupid stuff. And if I had ever killed someone in the proces I would expect to have been sent to prison for a long, long time.
It's called being an adult and taking responsibility for your own actions. If you're not willing to do so, then I'd suggest giving up the other trappings of adult life -- because you don't deserve them.
Here's a quote from a recent thread on BluesNews (see post 48):
Since the PC cards lack the specialized pixel shader hardware the CPU has to do it all. That makes the game heavily CPU bound. Of course, if a graphics card -- like, oh say, the GeForce 6 -- can replicate those pixel shader commands programatically then the GPU can take it over and dramatically increase frame rate.
Funny that it's Halo that the GF6 sees the biggest framerate jump...
It's not like those other games are using the hardware shaders yet anyway (or are they?).
They are -- FarCry is probably the most intensive game out there right now, fully utilizing DX9 specs. Halo is no slouch either, although a lot of its speed issues are from wanting to use hardware that simply isn't present (on PCs -- it is on the Xbox; why they didn't port away from this is beyond me).
Aquanox 2, Tomb Raider: Angel of Darkness, Painkiller, UT2k4, BF: Vietnam, and several others utilize DX9 to varying lengths as well. And there's the upcoming games -- Half Life 2, STALKER, Soldner (with an umlaut on the o), World of Warcraft, Everquest 2, and numerous others.
Quake 3 simply isn't a reliable benchmark anymore. It utterly fails to excercise the newer features of the cards -- which are really the only features to bother upgrading for. If all you're going to do is play Q3-era games then a GeForce2 is more than sufficient. If you want to run games already out, and those coming out in the next year, with all the graphical options turned up and at high-res then you'll be best served by either the latest nVidia or (probably) ATI card.
And (most importantly to me, and many others) if you want to get a card that can run new games at reasonable resolutions with most of the graphical bells and whistles on, but at a reasonable price... well, those $400 cards are going to be sub-$200 very quickly now, and the $200 cards are going to drop to around $100.
When I say "Unix Machines", I am referring to Sun, SGI, HP, and IBM Unix based hardware.
Oh yes... IBM hardware is such a deal. I'm so everjoyed that we run on IBM hardware and AIX... so much so that I could just about fucking kill myself.
Our development box was originally a B50 w/ single CPU and 512MB of RAM (it's since been upgraded all the way to a dual CPU B80 w/ 1 GB RAM -- which is still slower than the comparison system). It compiled code at roughly half the speed of a P2-400 running Redhat 7.2. This is using gcc on both platforms. VisualAge was slightly faster, except that it was a total fucking nightmare to actually use because xlC 5 can't generate template functions worth a crap. We would have to maintain the template prototypes ourselves, which became an increasing pain in the ass as our libraries grew.
Of course, AIX's linker is like nothing else on this world, so forget using a lot of open source tools. I'm still waiting on a debugger that will actually work -- both dbx (from IBM) and gdb simply core dump -- usually during running, but occasionally when they just try to load the code. TotalView works, but it's an additional cost. And since we upgraded to AIX 5.1 we've had to revert to static linking -- because dynamic linking is FUBAR'd with gcc (allegedly fixed in 3.4, but we'll see). There's a patch we could try, but compiling gcc on this blazingly fast box takes 20 hours.
Yeah, I'm really glad we spent tens of thousands of dollars on production, QA, and development servers (the QA and dev servers were virtually free, but the production ones were not) that are so abymally slow and infernally different from everything else. Even if a half dozen server class PCs cost the same as these boxes they would've been at least a magnitude faster and we'd have far fewer issues supporting them, not to mention how much developer time would've been saved at this point.
Oh, and we've just moved to (puke) StarTeam for our version control. Of course, it doesn't have a native AIX client, so we're having to work around the fact that the universal java client doesn't recognize executable files as such during checkin. This problem doesn't exist on Linux.
And yes, RedHat does self-destruct at the slightest provocation.
I'm sorry you have such lousy admins. None of the Redhat boxes we have, or that friends have at their jobs, seem to have any of the issues you mention.
I'm not even *going* to mention RPM hell.
Why, because you've found some mystical, magical fix for it? The reality is that RPM hell is a symptom of a larger problem -- that different programs often want different versions of shared libraries. The more programs you install on any one given box the more likely you'll have this problem. And if you don't have a package manager like RPM (or any one of the others) then you're not going to avoid this problem -- instead you'll overwrite a library without knowing that you're causing a problem. Have fun diagnosing that one!
I see a great many people buying hardware from Cisco's competitors in the near-future.
What makes you think that they don't have a backdoor username/pw as well? It may not be hard coded (they could both be strings that are determined by a hash function, based on the date/time or some other changing value), but I'd bet you they're there, at least on any high end equipment. Why? So that the damn thing is supportable remotely... even after some idiot admin screws up everything else. And, no, resetting the firmware on these things to restore the default admin password isn't acceptable -- simply because in doing so you'd lose all the other settings (bad for two reasons -- 1) they usually take hours or days to setup correctly, 2) if you're accessing the box for support, you probably want to see what the hell happened in case it was a bug).
You're presuming (most likely incorrectly) that the new user will keep the computer on 24/7 (and yes, I'm assuming that the MP3 server or network firewall will be on 24/7... which is likely to be true). Most people will only turn the computer on when needed and turn it off afterwards. That dramatically reduces energy usage.
There's the whole "benefit of use" issue as well, but that's murky at best, and telling others how they should use (or not use) their own belongings on a "moral" or "ethical" basis is the most slippery of slopes.
That's just insane.
Look, it's pretty simple. If you don't trust the precompiled binaries they have on the machines, then why on earth would you trust the compiler they provide? I'm not talking about not trusting gcc... it would be fairly trivial to produce a hacked gcc that compiles the code in question differently (or simply compiles an embedded version of the code). And you'd have no way of knowing.
Heck, hack the diff tool for that matter. Either reject any schmuck who actually tries to pull this, or replace their source with yours while diffing. How thick is that tinfoil beanie anyway?
Oh, and don't think that a mechanical system is any better either. After all, some machine has to read those ballots and you could just compromise it!
Sure, I guess you could go to hand counting. That'll only take a few weeks to verify the results. No matter... I'm sure the rest of the world will understand that we're just too damned incompetent to use modern balloting techniques. I'm sure it won't have any impact on domestic markets either... nope... just because we don't think our products are good enough to use doesn't mean you shouldn't buy them!
Note that you won't get the full range of the subwoofer with a DVD or CD. Both have a bottom limit of 20 Hz, while the sub in question can allegedly output down to 10 Hz flat (and well below that with fall off).
You'd have to use vinyl to get frequencies that low. And have one hell of a damping system for the player to avoid it being shaken by the sound it's reproducing...
Not a vinyl snob (all of my music is on CD and ripped to MP3), but there are technical limitations. Of course, 20 Hz is far lower than most people's stereo systems can reproduce anyway. Below that you can't really hear it... but you can feel it.