Honest query: I've thought of this myself, and wonder what problems there may be to implementing a UI toolkit (such as GTK or QT) so that it uses GLX to render the widgets, rather than the current method. (Since making a window using GLX is fairly close to a no-brainer; but how to take an app like GIMP or Konq and putting the entire app onto the GLX extension is a mystry to me...)
So is Windows' GDI32.exe. So is OS X's Quartz. If you want to believe either is a memory featherweight, I got a few investment ideas for you...
impossible to configure without a lot of help
The exact same can be said of both Windows and Macintosh -- and PicoGUI. Configuring hardware is not, nor has it ever been, easy. The only reason OSX or Windows seem so 'easy' to configure is that Apple or Microsoft wrote programs to provide "a lot of help". X doesn't have anywhere near the amount of automation built-into it. PicoGUI doen't auto-configure itself either; with a few exceptions, the only way you'll get PicoGUI to work is through X.
The thing here is that X is just a graphics subsystem. It does not contain configuration utilities, or its own widget sets. Strip off the (many, many) shiny "computer administration for dummies" configuration programs, and you've got the ease-of-use of Linux. (Actually, linux is probably much easier to configure manually than GDI32 or Quartz would be.) It isn't a fault of X -- Nobody's getting paid $90k a year to write these config utilities. Font configuration problems? What?!? You mean you have to do it yourself!!! Oh no!!!
Look - on some level, there will always be a need for people who understand how to configure GUI systems by hand, manually -- somebody has to know how to do that if they are to write a 'config for dummies' applet. Don't blame your refusal to learn on X. If you find the current system inadequate, then fix it yourself. Around the time you figure out how to make a spiffy auto-configure, you'll most likely have gotten past the need for such a utility, and won't be as motivated to 'fix' a system that was never really broken -- just misunderstood. (Of course, it helps that when people try to explain how confusing X is, and then try to straighten things out, usually end up spouting off incorrect information -- they're confused because they never knew the facts to begin with, and others take it as the truth, and the cycle repeats, only worse each time, until X has a terribly bad wrap.)
Something smaller faster and more elegant And you compare X to GDI32 or Quartz Extreme? They aren't smaller, or faster... It's all in the graphics driver itself there -- not the GUI system. All of this eye candy people crave requires a great deal of resources -- Anti-Aliasing, Themes, gradients, pixmaps -- all of them take significant resources, and ever-increasing numbers of processor cycles to execute. Hell -- Microsoft's own 'theme management' for WinXP is several times slower & larger than many of the alternate themeing systems (such as StarDock's WindowBlinds) Quartz offloads everything that it can to the video card-- earlier releases (before 10.2) used the same 'primitive' drawing techniques that X uses, with similar performance.
As far as elegance goes -- Elegant to use, or elegant to program? Win32 widgets are a nightmare to program. I've not had a chance to do OS X, but I understand they are elegant. And X widgets (at least QT and GTK) are quite elegant to program.
As far as 'elegant looking' -- blame the writers of the applications.
Most commercial apps have graphical design teams who specialize in GUI design placing the widgets. Linux apps have a programmer placing them in a way that seems logical to him/her, and attempts to match it as best as possible to various GUI standards (such as Apple's or KDE's).
Standardization on the GUI is something that has met rather vehemant opposition -- look at the KDE-GNOME wars!!!
But X itself is not the problem. X only provides the ability to have a keyboard, mouse, and graphics. Everything else is higher-level, and not a part of X.
Ogg would require a very expensive general-purpose DSP, a discrete DAC chip, and glue circuitry in addition to paying the normal patent license fees (no, MP3 is not going anywhere). That would increase the cost by more than $50 just for the hardware, and several dollars more to amortize the cost of developing a software decoder for the DSP.
Do you have any clue at all how ludicrous this statement is? I design chips, and integrate them into systems. This statement shows a clear lack of understanding of both the way hardware is built, and of the requirements for decoding OGG, WMA, and MP3.
Here's the run-down: Ogg would require a very expensive general-purpose DSP
First, decoding MP3 or WMA requires a very expensive, DSP as well. Supporting anything other than MP3 requires a general-purpose DSP; the only reason MP3 does not require a general-purpose DSP is that there are expensive, single-purpose MP3 decoding chips on the market. But the instant you throw in anything but MP3, you have to use a generic processor. There are no single-chip WMA decoders, nor single-chip realaudio, or AAC decoders.
A discrete DAC chip: This is required in every case: MP3, WMA, or anything else. The raw, decoded PCM that comes out of the decoder is then piped into a dedicated DAC to convert it to an analog signal. There are several reasons to do this, noise and distortion top the list. Besides - a 16-bit DAC chip is -- quite literally, at volume, $0.02.) And it's already required for the MP3 or WMA decoding anyway -- so that isn't adding anything to the cost.
glue circuitry: This is again, standard practice for all players, and is a trivial (and in most cases, zero-cost in time, effort, and money) problem.
But all this is a moot point: Most of the tasks required to decode MP3, OGG, WMA, AAC, or any similar codec are identical. The tasks are performed by a single low-cost chip dedicated to the task. (FWIW, I did a cost-analysis of an Ogg Vorbis decoding chip; it would take one engineer about 5 weeks to design. From there it's the initial fab cost that is the problem... the chip itself would be a couple of pennies to make per unit.)
Two of the central operations of both MP3 and OGG Vorbis are the Discrete Cosine Transform, and Huffman encoding. There are many, many makes and models of DCT and iDCT chips on the market, as well as huffman decoding chips. Nearly 99% of the hardware cost in decoding MP3 and OGG chips can be done by these two chips; and it works equally well for both MP3 and Vorbis. The microcontroller used will no doubt be more expensive in an OGG decoding solution when compared to an MP3-only decoder-- but the same can be said of WMA.
The more modern MP3 players (unless you count the $60/16MB MP3-only POS models) are all small computers. The microcontroller sends the commands off to the iDCT and huffman decoder chips, which do most of the work. OGG requires 2-3k more memory than MP3-- a trivial hardware cost of less than $0.01/unit.
And, finally, there are currently-existing MP3 players that were designed for MP3 (but also play WMA or RealAudio), that have had vorbis ported to it. (One prime example is the ill-fated Iomega HipZip, which was a PocketZIP/click drive + MP3 player. The vorbis porting was done under NDA, and never released after Iomega killed the HipZip altogether) Since Archos released specs for its jukebox, there is a vorbis development effort there as well.
In nearly every new MP3 player, the player itself requires no additional hardware. It's all in the firmware program. And with a pre-existing, 0-cost integer decoder, the development cost would be minimal. In all reality, it's no more difficult to make the hardware in an 'MP3' player play Vorbis than it is to support WMA or Real.
Lastly, MS is the only one who's at least claiming that they support multiple channel audio support, mathematically lossless compression scheme (at 2-3:1)
Um? Ogg Vorbis has supported multi-channel audio for years (yes, years. It has always been able to support 256 distinct audio channels) You just don't see too many 5.1 OGG files because, well... you don't see many 5.1 albums in the store, do ya? There is working support for 5.1 (and maybe 6.1 by now) Ogg Vorbis, and has been for quite a while now.
And for the compression claim: FLAC and Monkey's Audio both achieve similar compression scores -- and more importantly, it actually exists. This high-compression lossless WMA is still more vaporware from the company that invented vaporware.
There are plenty of sites who have independantly reviewed Vorbis vs WMA vs MP3Pro in double-blind tests. MP3Pro generally comes out on top, with Vorbis slightly behind. WMA is way behind, with AAC between WMA and Vorbis.k And WMA comes out better than traditional MP3.
And, finally: The techniquies that MP3, WMA, and Vorbis use to compress audio are all quantative: It can be mathematically proven which is 'best' for a given signal. The devil in the details is that Human ears just plain suck, and can easily find a more accurate reproduction to sound 'worse'.
I espescially liked the snobby, elitist "pimple-faced linux geeks listening through their awesome *cough* Logitech speakers" I really wonder how many stereotypes are in this one statement. Pimple faced linux geeks? Odd considering the majority of Vorbis software is for Windows. And that its primary supporters are those with college degrees in math, science, and engineering-- Professions well above the median income bracket. Implications of crappy Logitech speakers? While not the best, double-blind tests have shown many of their models to be far better than any shelf system, and at least as good as most of the so-called 'home-theatre' systems. After you get past that price range, the room you're listening in has as much of an effect as the electronics do. (And power line noise, etc. etc...) And, no, I don't have Logitech speakers, but I DO read reviews. I've heard them myself. They aren't at all bad-- I'd put them against any teenie-bopper's $5k "Thud" car audio in a heartbeat.
But, what do you expect from a world where the best and brightest are outcasts, and the druggie popularity-whore is lauded?
Re:Features & Why OGG is not better supported
on
Another iPod Competitor
·
· Score: 3, Informative
It basically comes down to CPU power. Most of the embedded chips do not have enough processing power to support OGG decoding.
This is quite false. In fact, many of the required decoding operations (IDCT being the primary one) are so close to identical it hardly matters. Moreover, decoding of Vorbis is nearly as efficient as MP3 decoding (give it the same kind of optimization time MP3 has had, and we can revisit the statement.) Since WMA is closed, I haven't seen any good papers on WMA decoder efficiency, but I think there's little reason to think it behaves much different to MP3 or Vorbis.
And there's one key point that makes the CPU power argument fall flat on its face: Encoding into either MP3 or WMA, requires massive (at least to an embedded system) resources; espescially when compared to the amount of processing power needed to decode the same piece. The difference is several orders of magnitude.
If the thing can record and encode to MP3, (espescially in real-time), Vorbis decoding is a no-brainer. Espescially with a fully-integerized decoder freely available.
Frankly, I think the thing should have the following decoders (in order of 'significance')
MP3 -- The de facto standard. AAC -- The 'new' MPEG standard (MPEG4), which not only encompasses MPEG4, but also RealAudio8+
WMA -- Pandering to Microsoft hasn't been bad for Creative in the past, and since there are millions who just use WMA because it's built-in to Windows, it makes sense to support it.
Vorbis -- The hacker favorite, but also lacks the licencing fees required for MP3, AAC, or WMA. MP3Pro -- Less popular than Vorbis, but it has the blessing of Thompson and Frauhaufer.
Ideally, it should be able to have decoder 'plugins' that you download, and are stored on the disk with the music (ie. not in firmware).
AS/400s dont use apples regular PowerPC chips moron. They use PowerPC/400 chips with proprietary cores and instructions.
This is like saying that the Pentium4 is a cpu with proprietary cores and instructions. The P4 has SSE2 instructions, where earlier Pentiums (and any AMD CPU) do not have them. Until the Palomino core (AthlonXP & MP), AMD had no support for Intel's SSE instructions (which Intel introduced with the P3). On the other hand, the Athlon is also a proprietary core, with its own proprietary instructions (3DNow!), that are not available on any Pentium.
PowerPC is like saying x86 or SPARC - a family of compatible processors, with a set of identical instructions.
The difference between the PowerPC/400 and the PowerPC "G4" that Apples uses is that the G4 (not the PPC/400) has the proprietary instructions & core (Altivec is a Motorola-proprietary tech).
More likely they'd just come out with new Macs that are functionally identical to the current machines, but with an Intel or AMD chip inside instead
By functionally identical, I assume you mean that the user doesn't see the difference?
And, of course, it's pretty obvious that Apple may use an AMD or Intel chip-- but it won't be x86. More like Itanium2/3 or a modified Opteron (that is not x86 compatible).
But there's one thing that many fail to see: Unlike x86, with an Apple, you can't just pop out the CPU and replace it with a faster one. The cases aren't exactly 'standard' either. If they switched to an Intel or AMD chip, the Apple motherboards likely wouldn't allow for CPU upgrades at all.
And, of course, for the overclocking crowd (whose only desire seems to be to run OS X on their cheapo hardware): Forget about overclocking anything from Apple if they go Intel/AMD. Apple values stability far more than speed-- they won't push any of their hardware past its design, nor will they allow it to happen. Locking the clock speed to the motherboard/cpu is how Apple can ensure that shady dealers won't try to sell an overclocked (and unstable) Apple, while claiming that it's not overclocked. The whole 'unlocking Athlons' is a moot point: x86 motherboards are capable of multiple clock speeds; primarily so it can handle CPU's of multiple speeds (and many are designed to allow overclocking of the CPU). But in an Apple design- expect to see the motherboard and CPU speeds permanently locked (Easy to do, espescially if the clock generator is part of the chipset, and requires 100's of surface-mount de&re soldering to replace; a task that is nearly certain to ruin the motherboard)
But since it is Unis and the military who were conected, it meant students and military people got free beer net.
You make it sound like there were other ways to access the internet. There WAS NO 'internet' in 1969. The military ARPANET WAS IT. Service(wo)men didn't have access, nor did students. It was years later that either were granted access to use ARPANET at all. (For that matter, it was decades later that either had any real access to computers; even then they were restricted to VT-100 and other 'dummy' terminals.)
It consisted of newsgroups, e-mail and FTP. There were different, conflicting methods of addressing email and other name resolution. The modern versions of these standards came not from cooperation, but from one method becoming the 'defacto' standard because of its popularity.
It's easily arguable that the only reason the tech was 'free' is because software didn't qualify for copyrights and patents at the time.
The first newsgroups were moderated and censored before anything could be passed on. It was a military network, and subject to military rules and security. Only individuals (nearly always scientists and engineers) with security clearance were permitted to use it at first. It allowed "Manhattan Project" like collaboration in a decentralized form; no single bomb could kill all of the country's best scientists at once. It allowed a higher level of dissemenation for 'sensitive' information that normally would have been classified and distributed on a "need to know" basis.
It wasn't until much later that the military released control, and the internet became 'free'. There was no 'Web', no HTTP-- It wasn't until '83 that telnet began. All servers had to be approved by the US military.
still it had a tradition of free beer and free speech
There's a significant difference between "a tradition of freedom", and being guaranteed it. If you give a slave "a tradition of freedom" s/he's still a slave-- their freedom is not secure, and can be revoked at any time. Such was the case with DARPAnet. The free speech didn't exist until years after the network existed. The "Free Beer" (in the form of software) existed because non-free software (and espescially software patents) were either completely or nearly non-existent.
That's a patently false joke. The internet was meant to be a massively distributed fail-safe way to transfer research information, so that in the advent of a nuclear war, the remaining research institutions could collaborate with each other (and develop tech to 'win the war'). Distributing porn probably ranked second on the list.
The internet was originally DARPAnet - A U.S. military research network. It linked universities whose research was funded by the military.
The internet was meant to research warfare and weapons. Being 'free' had nothing to do with it.
No one faulted the MPAA etc for encrypting DVDs. That was fair and all. We cracked it, but it was fair. We faulted them for making it illegal to try to crack it.
The difference is that the MPAA has a leagal (and, in the U.S. a constitutional) right to protect their content; to prevent its copying. The MPAA paid for the research and development of CSS. (although they didn't get their money's worth). The MPAA invested millions to make DVD's a widely-accepted standard. The MPAA recoups the investments a few pennies at a time for every DVD sold. Consumers pay for it in the end-- but they get a real product in return. There's nothing forcing us to buy the movies. We aren't required to pay for DVD's encryption if we don't use it.
Spammers & spam is a totally different story. Consumers (and businesses) are involuntarily forced to shoulder nearly all costs (free netzero here, a hacked server there; no bills, no money to pay), and recieve absolutely nothing -- no product, no service -- in return.
The only other form of involuntary, forced payment is taxes and fines-- but the payment provides products and services (education, roads, police...). While taxes & fines take, they give back.
Secondly, micro-pay. We see this tossed around a lot. Now, I don't like the idea of paying for email, but perhaps that's what it will take to remove spammers. Could someone create a "premium" email service which would require that senders pay.001 cents to send a message TO the account? Would this discourage spam? I think so.
A big reason why people don't like micropay is that it requires personal information (how else to pay?). The other problem, of course, is the idea of paying for the service (the same one as outlawing weapons): Only legitimate, leagal users would respect the system.
However, since spammers already use forged headers and hacked servers -- not to mention the con jobs involved -- there is no reason why they would send via a cost-per-message service.
In addition, there's the other, real problem: What's to stop spammers from creating their own servers to (Which they already do) to bypass the fees and restrictions involved?
Think of it like waste water (sewage):
There is the leagal, lawful way that costs money, and for which there is a bill. In some cases (chemical plants, hospitals, food processing plants) the sewage is monitored, and 'dirtier' water costs more.
There's also the storm drain, which is free and untraceable.
Spammers often create their own software; perfectly secure, patched software only means they won't be using my server. It doesn't mean they can't roll their own software that flat-out breaks e-mail addressing standards: Leave incorrect routing, source IP address, invalid everything to make it untraceable. Our e-mail system is designed to deliver mail -- not to be traceable. Even if everything else is bogus/forged, as long as the message contains a valid destination, the message will be delivered. Creating a system that ensures the message arrives at the recipient is easy; having a valid, traceable trail is much harder-- espescially when everything can be forged.
The only real method I can think of is a forced abandonment and change of the entire e-mail system to a completely closed, centralized, audited, encrypted, and controlled email system, with an unforgeable key for each user-- and that would only reduce spamming until the system is compromised.
And how do you force everyone to abandon an old, free email system for a new, metered system of dubious security/privacy & de-spamifiedness?
So, making it leagal to target spammers (both the spamming bodies, and the organizations/people they send spam for.) is the only fair way to combat it-- if advertisers actually have a financial penalty per victim, it won't be the rampant problem we now have. And if it's more expensive than alternatives, then advertisers will choose a cheaper method.
Technology is expensive and even worse than our imperfect leagal system. Spam-blocking often also blocks valid emails (like when I opt-in to a list that tells me a patch is available for my software). Using a techical solution still forces the consumer to shoulder the full cost of blocking spam. It doesn't cost spammers anything if the spam is blocked. They don't pay for the spam-blocking software. They don't have to invest time and talent in implementing it.
When I see ads on TV, in print, or hear them on the radio-- there is a real, desireable product (media, news, etc.) delivered in exchange for the chance to advertise to me. I accept the intrusion of advertisement in exchange for the products and/or services I recieve between commercials.
At least NetZero gives me free internet access in exchange for having ads pushed to my desktop.
However, spam provides me with absolutely nothing-- no product, no service. What's more-- I am the one shouldering nearly all the cost! I have to take the time to download (even parts of) them. I pay for the bandwidth. I pay the price for the email box; either directly (ISP) or banner ads (Hotmail). And I have to go to the effort of identifying and deleting them. The spammer's only cost is time.
Advertising is a heavily regulated & legislated industry for a good reason. Legislating spam is the continuing efforts of lawmakers to regulate the advertising industry. Thanks to legislation, advertisers cannot:
Drive down the street with a thunderous PA system blaring their piece at maximum volume (ice cream trucks can only play music-- but they can't have any lyrics or other non-instrumental content)
Solicit wares door-to-door (without a licence to identify them and provide leagal accountability)
Telemarketers can only call during certain hours-- and they bear a great financial cost to make the calls.
And in all 3 of the above cases, there is a significant operating cost to 'reach the audience'. In the case of spam, there is no financial difference between one and one million targets, and they provide no service or good for the privelige.
Try to remember: Every law can be broken-- but every technology can be defeated. The difference is breaking a law has a penalty.
The problem is that in many cases, the counter-measure, while being released far sooner (perhaps even decades), can very easily come too late.
What good is a countermeasure / vacciene / cure to a deadly virus, if because of that little tidbit being published openly, terrorists were able to infect and kill a few hundred thousand?
It doesn't matter that the countermeasure came out sooner because of the publication. The countermeausre would not have been necessary to begin with. The disease would have remained a 'niche' disease contained to a small geographical area, and a small population. But with premature publication, the disease was easily modified into a weapon; and used before a countermeasure existed. It is still too late for the dead or dying by the time the countermeasure is developed.
If by witholding the information, lives can be saved... it's worth the extra time. Even if it saves only one life, it's worth the extra time.
It is like the close watch the nuclear countries have over non-nuclear countries (but who have fission reactors for electricity). We still have no countermeasure against a nuclear bomb in a city. But if we can keep any more bombs from being made, there is a greatly reduced chance of a bomb ever detonating in a city.
I'll say this: If nothing else, at least you can see that the Itanium != Opteron competitor. The intended markets are so *completely* different... I finally gave up in apathy from the rabid responses from Opteron fans.
I suspect they aren't so much Opteron fans as they are cheap, and prefer to put up with poorly designed, inexpensive, rushed-design hardware; they are unwilling to fork over the extra cash for a computer with a better and more elegant design and architecture. (I'm not saying that the Opteron isn't an excellent piece of technology. I'm saying that if the same technology & effort that was spent on the Opteron went to something non-x86, things would be even better)
Another hallmark is they want Apple to release OS X for their cheap x86 hardware; somehow believing that it would be at least as stable on the horiffically diverse x86 (and its much more erratic (and often terrible) quality control, to say nothing of quality) than the tightly controlled & heavily tested hardware it currently runs on.
Even when shown the facts they don't want to believe that Apple is a hardware company first, and software second-- and when Apple allowed Mac clones, it almost bankrupted them.
I think he meant go 32-way with a system that uses x86 processors, and is fully x86 compatible.
I know there are systems that are more than 32-way using x86 processors -- but they aren't x86 compatible at all.
Re:about time..The Day Z-D Journalism Died
on
Ziff Davis Teeters
·
· Score: 2
The funny thing about PC Rags reviews is that the data and graphics often are accurate but the text and titles are not.
Yeah, but at least you could get real facts if you looked for them. I knew they were way off base around 1990 -- and I was 12 at the time. I mean - I was leagally stupid/gullible back then, and I could see the lies as plain as day. But at least I could find something useful in the tables & graphs shown (and at least more factual).
I tried PC magazine about 3 years ago. I was expecting software comparisons, hardware reviews... things like I was used to seeing.
What I got was 'We compare this shareware html editor to Microsoft FrontPage'. There was nothing to do with PC's -- it was more like "Web Page magazine" Not at all what I was expecting. And even the comparisons were far from fair (a cheap $30 shareware utility to a $500 Microsoft utility?) The funny thing was how well the $30 shareware did against Microsoft's behemoths.
But the problem was still there: It wasn't so much that they weren't objective; I knew that a decade before. But they also gave up on being a technical magazine. They didn't even provide comparisons for anything useful. All 4 'major' office suites had a release (Corel WordPerfect, Lotus SmartSuite, Microsoft Office, and StarOffice) -- but Microsoft office was the only 'review'. Absolutely nothing -- not even a mention of any kind -- of the other office packages.
Nothing comparing the speeds of AMD vs. Intel. Nothing with speed or reliability of Drives, motherboards, video cards. Nothing but web hosting and web authoring tools & services.
So, I gave up on ZD, and pretty much tossed the whole tech magazine journalism altogether; at least in the broad sense. Linux Journal and DDJ are still good (although, at least in the case of LJ, it's targeted to Linux peoples, rather than a more general magazine) here's hoping they keep their technical content.
The only problem with the argument is that the world doesn't have a well-organized research structure; concurrent development of a technology without any sharing of information is a reality. Leibnitz and Newton both developed calculus around the same time, independantly of each other. (Both methods work fine, but are *very* different.)
So, to be honest, a fair amount of the time, the exact same technology is researched at the same time, completely independant of each other; the discovery is literally made twice before anything is published. It's also paid for twice. That doesn't mean that there was waste involved.
Frankly, I can deal with slower development of some technologies if its censorship increases my personal safety. This is espescially true in the case of leathal chemical agents, or deadly (but uncommon) bioagents such as ebola or anthrax.
Yes you can; it's been done before. In many cases, keeping a secret is much easier than the problems the information's misuse can cause. Some things should be kept secret, and world is a better place for it. By keeping the way to create VX a secret (for example), there is a greatly reduced chance of it being used. I don't have to worry about keeping chemical warfare gear and toxin antidotes on my person at all times.
Oklahoma City and Sept. 11 were extremely localized in scope. A small geographic area was affected. I doubt that the synthesis of VX is much more complicated if you know the secret than these acts. Thank God Iraq couldn't find the way to make VX (they abandoned the program in '1989).
If they did, we would have a far greater problem on our hands; a few well-placed aerosol cans could depopulate a city. And how do you protect from that? Nuke Iraq, Afghanistan, Pakistan, and every other country with ties to al-queda? How many innocents would that claim? Would it solve anything?
Sorry-- but the way to solve the problem is to prevent it from happening. Don't spread the information by publishing it. Destroy all records of the existence of the information, and of the information itself. Inform and 'convert' all who know and can reproduce the information of the real (and dire) dangers that can arise should the secret get out. And, barring that-- and for all of the coldness involved, kill or otherwise silence the people involved who can reproduce or easily re-research even parts of the information from their memory.
One thing is a guaranteed fact: If 'sensitive' information is easily available, it will eventually be misused. And many bits of information, when misused, end catastrophically. And it'll always be easier to destroy than create; entropy doesn't help to order/organize anything.
There is no acceptable choice-- either:
murder a few brilliant scientists who is guilty only of finding (even unintentionally) a dangerous bit of information.
or
watch thousands or millions of other perfectly innocent people murdered because of the information's misuse.
It's a hard choice, but a good government tries to serve the greatest number of people; not its most talented people.
Maybe we should try to build a world where people aren't trying to kill us.
If you want to go that route, Stalin and Mao had the right idea. Just kill all of your ideological and social opponents.
So did George Orwell-- the "Thought Police" is a work of sheer brilliance.
So how many people have to die or be 'conditioned' before that becomes a real possibility? Our focus is not the problem; we've had the same kind of societal problems (if not the exact same problems) for all of recorded history.
Most of us prefer to have our own thoughts, ideals, and moralities. And, unsurprisingly, we often disagree. There are always a few people who believe so strongly that they will die for their belief-- they'll lie to themselves, kill, mangle, and destroy for that belief.
Of course, cluster bombs have a treaty banning their use as an anti-personnell weapon. Against unarmored vehicles, fine... but not troops.
For the most part, using any aircraft delivered weapon as an antipersonnel measure is considered a waste of resources. (Exceptions being large distances, or where the Media might get footage of 'friendly' soldiers taking fire.)
If you want a good example of an anti-personell weapon, it's simple non-explosive small arms. Any recently (15 years) designed NATO aircraft ordinance is simply not intended for anti-personnell tasks. It's just not how we fight wars anymore. There is no need to kill off the opposing army-- just cut off their supply and command structures and leave the army hungry, thirsty, confused, and with low amunition. Armies are made of humans-- when getting the next meal is more difficult than fighting the 'enemy', the war is lost.
And, one other point: Only CISC uses microcode (as using dedicated logic to decode is even more complex). RISC uses pure dedicated logic to decode. Using a microcode decoder does reduce complexity of a CISC design (a dedicated logic decoder is trulyl monstrous in a CISC design). RISC is able to get away with using dedicated decoding because they have so few (and considerably more simple) instructions.
And microcode has made huge leaps in efficiency lately -- current designs have microcode that executes close to 95% as fast as pure dedicated logic (so it's only about 5% slower to decode).
And for a link (I didn't try hard) you can do a quick 1-min google search...
But here's some I turned up:
http://www.geocities.com/imac_driver/cpu.html (not SPEC, but it does compare FLOPS (Floating Point Ops/sec) and has links to other comparisons.
Well, I do want to try & stay away from poorly designed games;) I mean-- if they don't bother to get a somewhat trivial issue like that right, it makes me wonder what other things they fudged.
Let me get this out of the way: I design microprocessors (but mine are more specialized than the generic CPU's pumped out by Intel & AMD). In every case it's much easier, and less costly to use a RISC design. And the RISC design crunches numbers just as quickly.
I'll admit this: I neglected cache memory and cache control circuitry. This takes up about 1/3 of the design, and modifies things a bit-- but the relative size of the execution pipelines to the instruction decode is still roughly 50/50.
Instruction decode also requires the following (and possibly more) sections (That you appear to have missed):
PreDecode Array/CTL
Scan/Align
Instruction Control Unit
Branch Prediction
Floating Point Control
Floating Point Scheduler
And, the Athlon can not decode 3 instructions at the same time. It can dispatch 3 instructions to the integer RISC pipelines at the same time. This is an important distinction, as 1 CISC instruction is equivalent of 1 to 10 (or more) RISC instructions. Each RISC instruction takes 1 clock tick to execute. But a CISC operation can take anywhere from one to several hundred clock cycles to execute. Overall, it takes the same amount of time-- its' just that the RISC does it in several small steps to 1 huge CISC step.
In other words, it decodes 1 x86 instruction, and then can dispatch up to 3 of the 'RISC ops' that are required to do the same task as 1 x86 instruction. (But remember-- 1 x86 instruction may be a lot more than just 3 'RISC ops').
Another thing to remember is that the size of the instruction decode is relatively static-- but the execution stage has grown progressively larger (by adding multiple identical pipelines). Only now when there are 5+ pipelines is the execution stage approaching the size of the decoder.
With fewer pipelines (such as '486, which has 1 FP and 1 integer, which aren't interchangeable), the decode was much larger than the execution logic.
In contrast, a RISC design (single-pipeline) takes up ~15% of the total core size. Superscalar (multiple identical pipelines), and adding cache can reduce the decode to 2-3% of the total die.
Where a CISC design (single pipeline) the decode is ~75% (this is textbook data; from an x86 architecture text). Superscaler, cached designs reduce the decode to ~33% of the total die.
So there's a RISC decode (2-3% of total die), vs. CISC decode (33% of total die)
Don't try to tell me that the difference is small or insignificant. The PowerPC G4 (which SPEC's faster than an Athlon of twice its clock speed) has less than half the transistors of the Athlon. And it requires less than 1/3 of the power.
Decoded MP3 is still missing ~90% of the original sound. It's just that human ears have physical limitations that make this possible. (Things like the eardrum moves 'in' faster than it moves 'out')
Transcoding is trying to force Vorbis to work with its hands tied; a decoded wav is still missing 90% of the original signal.
The result is somewhat like a cook-off.
The task: Create a low-fat chicken parmesean recipie that tases as good as a specific high-fat chicken parmesean recipie
(Wav->MP3) Chef 'A' uses his methods to prepare a low-fat chicken parmesean.
(Wav->OGG) Chef 'B' uses his (different) methods to prepare a low-fat chicken parmesean.
In both cases, either chef may choose from the available ingredients of his/her choice to make the best-tasting meal. In this case, chef B wins easily.
But MP3->OGG is like:
Forcing Chef 'B' to take the meal Chef 'A' created, 'un-cook' [decode] it, use the ingredients Chef 'A' used, and create a better meal.
Chef 'B' has to 'start' with the amounts Chef 'A' finished with.
Chef 'B' would have used an entirely different recipie to create his low-fat dish; but since he's limited to what Chef 'A' 'served', he has to make do.
It doesn't matter if Chef 'B' is a much better cook. He's not able to use the ingredients of his choosing. Chef 'B' is essentially limited to modifying Chef 'A's recipie, and therefore he's not able to achieve his potential.
That's what people said when there were mac clones. Mac clones are what almost killed Apple. Apple is a computer company-- only a tiny fraction of their total income is from software, and most of their software profits come from QuickTime-- not OS X.
And dreaming that Apple's market share would grow is just that -- dreaming. Microsoft's strength is Windows-compatibility. And don't forget that since Microsoft owns a significant percentage of Apple, they still get to have their say.
And, finally -- why did Apple sell so much of itself to Microsoft? To get Office ported to OS X quickly. Apple isn't stupid. They know to be successful in the mass market, they have to have Microsoft Office, because that's what everybody uses at work (or school, etc).
And if you think Apple cares about your $129, you're kidding yourself. They want the $800-3000 profit they recieve from selling hardware. They can stay in business (and are doing better than many of the clones makers, like Gateway). Apple cannot stay in business if there are clones running Macintosh.
Slashdot Mirror
Honest query: I've thought of this myself, and wonder what problems there may be to implementing a UI toolkit (such as GTK or QT) so that it uses GLX to render the widgets, rather than the current method. (Since making a window using GLX is fairly close to a no-brainer; but how to take an app like GIMP or Konq and putting the entire app onto the GLX extension is a mystry to me...)
X is a big dumb slow ram hog
So is Windows' GDI32.exe. So is OS X's Quartz. If you want to believe either is a memory featherweight, I got a few investment ideas for you...
impossible to configure without a lot of help
The exact same can be said of both Windows and Macintosh -- and PicoGUI. Configuring hardware is not, nor has it ever been, easy. The only reason OSX or Windows seem so 'easy' to configure is that Apple or Microsoft wrote programs to provide "a lot of help". X doesn't have anywhere near the amount of automation built-into it. PicoGUI doen't auto-configure itself either; with a few exceptions, the only way you'll get PicoGUI to work is through X.
The thing here is that X is just a graphics subsystem. It does not contain configuration utilities, or its own widget sets. Strip off the (many, many) shiny "computer administration for dummies" configuration programs, and you've got the ease-of-use of Linux. (Actually, linux is probably much easier to configure manually than GDI32 or Quartz would be.) It isn't a fault of X -- Nobody's getting paid $90k a year to write these config utilities. Font configuration problems? What?!? You mean you have to do it yourself!!! Oh no!!!
Look - on some level, there will always be a need for people who understand how to configure GUI systems by hand, manually -- somebody has to know how to do that if they are to write a 'config for dummies' applet. Don't blame your refusal to learn on X. If you find the current system inadequate, then fix it yourself. Around the time you figure out how to make a spiffy auto-configure, you'll most likely have gotten past the need for such a utility, and won't be as motivated to 'fix' a system that was never really broken -- just misunderstood. (Of course, it helps that when people try to explain how confusing X is, and then try to straighten things out, usually end up spouting off incorrect information -- they're confused because they never knew the facts to begin with, and others take it as the truth, and the cycle repeats, only worse each time, until X has a terribly bad wrap.)
Something smaller faster and more elegant And you compare X to GDI32 or Quartz Extreme? They aren't smaller, or faster... It's all in the graphics driver itself there -- not the GUI system. All of this eye candy people crave requires a great deal of resources -- Anti-Aliasing, Themes, gradients, pixmaps -- all of them take significant resources, and ever-increasing numbers of processor cycles to execute. Hell -- Microsoft's own 'theme management' for WinXP is several times slower & larger than many of the alternate themeing systems (such as StarDock's WindowBlinds) Quartz offloads everything that it can to the video card-- earlier releases (before 10.2) used the same 'primitive' drawing techniques that X uses, with similar performance.
As far as elegance goes -- Elegant to use, or elegant to program? Win32 widgets are a nightmare to program. I've not had a chance to do OS X, but I understand they are elegant. And X widgets (at least QT and GTK) are quite elegant to program.
As far as 'elegant looking' -- blame the writers of the applications.
Most commercial apps have graphical design teams who specialize in GUI design placing the widgets. Linux apps have a programmer placing them in a way that seems logical to him/her, and attempts to match it as best as possible to various GUI standards (such as Apple's or KDE's).
Standardization on the GUI is something that has met rather vehemant opposition -- look at the KDE-GNOME wars!!!
But X itself is not the problem. X only provides the ability to have a keyboard, mouse, and graphics. Everything else is higher-level, and not a part of X.
The reason is that the native widgets /should/ use less system resources...
If only there was a theme that used the OS native widgets, without the ugly 'classic' icons...
Ogg would require a very expensive general-purpose DSP, a discrete DAC chip, and glue circuitry in addition to paying the normal patent license fees (no, MP3 is not going anywhere). That would increase the cost by more than $50 just for the hardware, and several dollars more to amortize the cost of developing a software decoder for the DSP.
Do you have any clue at all how ludicrous this statement is? I design chips, and integrate them into systems. This statement shows a clear lack of understanding of both the way hardware is built, and of the requirements for decoding OGG, WMA, and MP3.
Here's the run-down:
Ogg would require a very expensive general-purpose DSP
First, decoding MP3 or WMA requires a very expensive, DSP as well. Supporting anything other than MP3 requires a general-purpose DSP; the only reason MP3 does not require a general-purpose DSP is that there are expensive, single-purpose MP3 decoding chips on the market. But the instant you throw in anything but MP3, you have to use a generic processor. There are no single-chip WMA decoders, nor single-chip realaudio, or AAC decoders.
A discrete DAC chip: This is required in every case: MP3, WMA, or anything else. The raw, decoded PCM that comes out of the decoder is then piped into a dedicated DAC to convert it to an analog signal. There are several reasons to do this, noise and distortion top the list. Besides - a 16-bit DAC chip is -- quite literally, at volume, $0.02.) And it's already required for the MP3 or WMA decoding anyway -- so that isn't adding anything to the cost.
glue circuitry: This is again, standard practice for all players, and is a trivial (and in most cases, zero-cost in time, effort, and money) problem.
But all this is a moot point: Most of the tasks required to decode MP3, OGG, WMA, AAC, or any similar codec are identical. The tasks are performed by a single low-cost chip dedicated to the task. (FWIW, I did a cost-analysis of an Ogg Vorbis decoding chip; it would take one engineer about 5 weeks to design. From there it's the initial fab cost that is the problem... the chip itself would be a couple of pennies to make per unit.)
Two of the central operations of both MP3 and OGG Vorbis are the Discrete Cosine Transform, and Huffman encoding. There are many, many makes and models of DCT and iDCT chips on the market, as well as huffman decoding chips. Nearly 99% of the hardware cost in decoding MP3 and OGG chips can be done by these two chips; and it works equally well for both MP3 and Vorbis. The microcontroller used will no doubt be more expensive in an OGG decoding solution when compared to an MP3-only decoder-- but the same can be said of WMA.
The more modern MP3 players (unless you count the $60/16MB MP3-only POS models) are all small computers. The microcontroller sends the commands off to the iDCT and huffman decoder chips, which do most of the work. OGG requires 2-3k more memory than MP3-- a trivial hardware cost of less than $0.01/unit.
And, finally, there are currently-existing MP3 players that were designed for MP3 (but also play WMA or RealAudio), that have had vorbis ported to it. (One prime example is the ill-fated Iomega HipZip, which was a PocketZIP/click drive + MP3 player. The vorbis porting was done under NDA, and never released after Iomega killed the HipZip altogether) Since Archos released specs for its jukebox, there is a vorbis development effort there as well.
In nearly every new MP3 player, the player itself requires no additional hardware. It's all in the firmware program. And with a pre-existing, 0-cost integer decoder, the development cost would be minimal. In all reality, it's no more difficult to make the hardware in an 'MP3' player play Vorbis than it is to support WMA or Real.
Lastly, MS is the only one who's at least claiming that they support multiple channel audio support, mathematically lossless compression scheme (at 2-3:1)
Um? Ogg Vorbis has supported multi-channel audio for years (yes, years. It has always been able to support 256 distinct audio channels) You just don't see too many 5.1 OGG files because, well... you don't see many 5.1 albums in the store, do ya? There is working support for 5.1 (and maybe 6.1 by now) Ogg Vorbis, and has been for quite a while now.
And for the compression claim: FLAC and Monkey's Audio both achieve similar compression scores -- and more importantly, it actually exists. This high-compression lossless WMA is still more vaporware from the company that invented vaporware.
There are plenty of sites who have independantly reviewed Vorbis vs WMA vs MP3Pro in double-blind tests. MP3Pro generally comes out on top, with Vorbis slightly behind. WMA is way behind, with AAC between WMA and Vorbis.k And WMA comes out better than traditional MP3.
And, finally: The techniquies that MP3, WMA, and Vorbis use to compress audio are all quantative: It can be mathematically proven which is 'best' for a given signal. The devil in the details is that Human ears just plain suck, and can easily find a more accurate reproduction to sound 'worse'.
I espescially liked the snobby, elitist "pimple-faced linux geeks listening through their awesome *cough* Logitech speakers" I really wonder how many stereotypes are in this one statement. Pimple faced linux geeks? Odd considering the majority of Vorbis software is for Windows. And that its primary supporters are those with college degrees in math, science, and engineering-- Professions well above the median income bracket. Implications of crappy Logitech speakers? While not the best, double-blind tests have shown many of their models to be far better than any shelf system, and at least as good as most of the so-called 'home-theatre' systems. After you get past that price range, the room you're listening in has as much of an effect as the electronics do. (And power line noise, etc. etc...) And, no, I don't have Logitech speakers, but I DO read reviews. I've heard them myself. They aren't at all bad-- I'd put them against any teenie-bopper's $5k "Thud" car audio in a heartbeat.
But, what do you expect from a world where the best and brightest are outcasts, and the druggie popularity-whore is lauded?
It basically comes down to CPU power. Most of the embedded chips do not have enough processing power to support OGG decoding.
This is quite false. In fact, many of the required decoding operations (IDCT being the primary one) are so close to identical it hardly matters. Moreover, decoding of Vorbis is nearly as efficient as MP3 decoding (give it the same kind of optimization time MP3 has had, and we can revisit the statement.) Since WMA is closed, I haven't seen any good papers on WMA decoder efficiency, but I think there's little reason to think it behaves much different to MP3 or Vorbis.
And there's one key point that makes the CPU power argument fall flat on its face: Encoding into either MP3 or WMA, requires massive (at least to an embedded system) resources; espescially when compared to the amount of processing power needed to decode the same piece. The difference is several orders of magnitude.
If the thing can record and encode to MP3, (espescially in real-time), Vorbis decoding is a no-brainer. Espescially with a fully-integerized decoder freely available.
Frankly, I think the thing should have the following decoders (in order of 'significance')
MP3 -- The de facto standard.
AAC -- The 'new' MPEG standard (MPEG4), which not only encompasses MPEG4, but also RealAudio8+
WMA -- Pandering to Microsoft hasn't been bad for Creative in the past, and since there are millions who just use WMA because it's built-in to Windows, it makes sense to support it.
Vorbis -- The hacker favorite, but also lacks the licencing fees required for MP3, AAC, or WMA.
MP3Pro -- Less popular than Vorbis, but it has the blessing of Thompson and Frauhaufer.
Ideally, it should be able to have decoder 'plugins' that you download, and are stored on the disk with the music (ie. not in firmware).
AS/400s dont use apples regular PowerPC chips moron.
They use PowerPC/400 chips with proprietary cores and instructions.
This is like saying that the Pentium4 is a cpu with proprietary cores and instructions. The P4 has SSE2 instructions, where earlier Pentiums (and any AMD CPU) do not have them. Until the Palomino core (AthlonXP & MP), AMD had no support for Intel's SSE instructions (which Intel introduced with the P3). On the other hand, the Athlon is also a proprietary core, with its own proprietary instructions (3DNow!), that are not available on any Pentium.
PowerPC is like saying x86 or SPARC - a family of compatible processors, with a set of identical instructions.
The difference between the PowerPC/400 and the PowerPC "G4" that Apples uses is that the G4 (not the PPC/400) has the proprietary instructions & core (Altivec is a Motorola-proprietary tech).
More likely they'd just come out with new Macs that are functionally identical to the current machines, but with an Intel or AMD chip inside instead
By functionally identical, I assume you mean that the user doesn't see the difference?
And, of course, it's pretty obvious that Apple may use an AMD or Intel chip-- but it won't be x86. More like Itanium2/3 or a modified Opteron (that is not x86 compatible).
But there's one thing that many fail to see: Unlike x86, with an Apple, you can't just pop out the CPU and replace it with a faster one. The cases aren't exactly 'standard' either. If they switched to an Intel or AMD chip, the Apple motherboards likely wouldn't allow for CPU upgrades at all.
And, of course, for the overclocking crowd (whose only desire seems to be to run OS X on their cheapo hardware): Forget about overclocking anything from Apple if they go Intel/AMD. Apple values stability far more than speed-- they won't push any of their hardware past its design, nor will they allow it to happen. Locking the clock speed to the motherboard/cpu is how Apple can ensure that shady dealers won't try to sell an overclocked (and unstable) Apple, while claiming that it's not overclocked. The whole 'unlocking Athlons' is a moot point: x86 motherboards are capable of multiple clock speeds; primarily so it can handle CPU's of multiple speeds (and many are designed to allow overclocking of the CPU). But in an Apple design- expect to see the motherboard and CPU speeds permanently locked (Easy to do, espescially if the clock generator is part of the chipset, and requires 100's of surface-mount de&re soldering to replace; a task that is nearly certain to ruin the motherboard)
But since it is Unis and the military who were conected, it meant students and military people got free beer net.
You make it sound like there were other ways to access the internet. There WAS NO 'internet' in 1969. The military ARPANET WAS IT. Service(wo)men didn't have access, nor did students. It was years later that either were granted access to use ARPANET at all. (For that matter, it was decades later that either had any real access to computers; even then they were restricted to VT-100 and other 'dummy' terminals.)
It consisted of newsgroups, e-mail and FTP. There were different, conflicting methods of addressing email and other name resolution. The modern versions of these standards came not from cooperation, but from one method becoming the 'defacto' standard because of its popularity.
It's easily arguable that the only reason the tech was 'free' is because software didn't qualify for copyrights and patents at the time.
The first newsgroups were moderated and censored before anything could be passed on. It was a military network, and subject to military rules and security. Only individuals (nearly always scientists and engineers) with security clearance were permitted to use it at first. It allowed "Manhattan Project" like collaboration in a decentralized form; no single bomb could kill all of the country's best scientists at once. It allowed a higher level of dissemenation for 'sensitive' information that normally would have been classified and distributed on a "need to know" basis.
It wasn't until much later that the military released control, and the internet became 'free'. There was no 'Web', no HTTP-- It wasn't until '83 that telnet began. All servers had to be approved by the US military.
still it had a tradition of free beer and free speech
There's a significant difference between "a tradition of freedom", and being guaranteed it. If you give a slave "a tradition of freedom" s/he's still a slave-- their freedom is not secure, and can be revoked at any time. Such was the case with DARPAnet. The free speech didn't exist until years after the network existed. The "Free Beer" (in the form of software) existed because non-free software (and espescially software patents) were either completely or nearly non-existent.
The internet was meant to be free you troll
That's a patently false joke. The internet was meant to be a massively distributed fail-safe way to transfer research information, so that in the advent of a nuclear war, the remaining research institutions could collaborate with each other (and develop tech to 'win the war'). Distributing porn probably ranked second on the list.
The internet was originally DARPAnet - A U.S. military research network. It linked universities whose research was funded by the military.
The internet was meant to research warfare and weapons. Being 'free' had nothing to do with it.
The difference is that the MPAA has a leagal (and, in the U.S. a constitutional) right to protect their content; to prevent its copying. The MPAA paid for the research and development of CSS. (although they didn't get their money's worth). The MPAA invested millions to make DVD's a widely-accepted standard. The MPAA recoups the investments a few pennies at a time for every DVD sold. Consumers pay for it in the end-- but they get a real product in return. There's nothing forcing us to buy the movies. We aren't required to pay for DVD's encryption if we don't use it.
Spammers & spam is a totally different story. Consumers (and businesses) are involuntarily forced to shoulder nearly all costs (free netzero here, a hacked server there; no bills, no money to pay), and recieve absolutely nothing -- no product, no service -- in return.
The only other form of involuntary, forced payment is taxes and fines-- but the payment provides products and services (education, roads, police...). While taxes & fines take, they give back.
Secondly, micro-pay. We see this tossed around a lot. Now, I don't like the idea of paying for email, but perhaps that's what it will take to remove spammers. Could someone create a "premium" email service which would require that senders pay
A big reason why people don't like micropay is that it requires personal information (how else to pay?). The other problem, of course, is the idea of paying for the service (the same one as outlawing weapons): Only legitimate, leagal users would respect the system.
However, since spammers already use forged headers and hacked servers -- not to mention the con jobs involved -- there is no reason why they would send via a cost-per-message service.
In addition, there's the other, real problem: What's to stop spammers from creating their own servers to (Which they already do) to bypass the fees and restrictions involved?
Think of it like waste water (sewage):
There is the leagal, lawful way that costs money, and for which there is a bill. In some cases (chemical plants, hospitals, food processing plants) the sewage is monitored, and 'dirtier' water costs more.
There's also the storm drain, which is free and untraceable.
Spammers often create their own software; perfectly secure, patched software only means they won't be using my server. It doesn't mean they can't roll their own software that flat-out breaks e-mail addressing standards: Leave incorrect routing, source IP address, invalid everything to make it untraceable. Our e-mail system is designed to deliver mail -- not to be traceable. Even if everything else is bogus/forged, as long as the message contains a valid destination, the message will be delivered. Creating a system that ensures the message arrives at the recipient is easy; having a valid, traceable trail is much harder-- espescially when everything can be forged.
The only real method I can think of is a forced abandonment and change of the entire e-mail system to a completely closed, centralized, audited, encrypted, and controlled email system, with an unforgeable key for each user-- and that would only reduce spamming until the system is compromised.
And how do you force everyone to abandon an old, free email system for a new, metered system of dubious security/privacy & de-spamifiedness?
So, making it leagal to target spammers (both the spamming bodies, and the organizations/people they send spam for.) is the only fair way to combat it-- if advertisers actually have a financial penalty per victim, it won't be the rampant problem we now have. And if it's more expensive than alternatives, then advertisers will choose a cheaper method.
Technology is expensive and even worse than our imperfect leagal system. Spam-blocking often also blocks valid emails (like when I opt-in to a list that tells me a patch is available for my software). Using a techical solution still forces the consumer to shoulder the full cost of blocking spam. It doesn't cost spammers anything if the spam is blocked. They don't pay for the spam-blocking software. They don't have to invest time and talent in implementing it.
When I see ads on TV, in print, or hear them on the radio-- there is a real, desireable product (media, news, etc.) delivered in exchange for the chance to advertise to me. I accept the intrusion of advertisement in exchange for the products and/or services I recieve between commercials.
At least NetZero gives me free internet access in exchange for having ads pushed to my desktop.
However, spam provides me with absolutely nothing-- no product, no service. What's more-- I am the one shouldering nearly all the cost! I have to take the time to download (even parts of) them. I pay for the bandwidth. I pay the price for the email box; either directly (ISP) or banner ads (Hotmail). And I have to go to the effort of identifying and deleting them. The spammer's only cost is time.
Advertising is a heavily regulated & legislated industry for a good reason. Legislating spam is the continuing efforts of lawmakers to regulate the advertising industry. Thanks to legislation, advertisers cannot:
Drive down the street with a thunderous PA system blaring their piece at maximum volume (ice cream trucks can only play music-- but they can't have any lyrics or other non-instrumental content)
Solicit wares door-to-door (without a licence to identify them and provide leagal accountability)
Telemarketers can only call during certain hours-- and they bear a great financial cost to make the calls.
And in all 3 of the above cases, there is a significant operating cost to 'reach the audience'. In the case of spam, there is no financial difference between one and one million targets, and they provide no service or good for the privelige.
Try to remember: Every law can be broken-- but every technology can be defeated. The difference is breaking a law has a penalty.
The problem is that in many cases, the counter-measure, while being released far sooner (perhaps even decades), can very easily come too late.
What good is a countermeasure / vacciene / cure to a deadly virus, if because of that little tidbit being published openly, terrorists were able to infect and kill a few hundred thousand?
It doesn't matter that the countermeasure came out sooner because of the publication. The countermeausre would not have been necessary to begin with. The disease would have remained a 'niche' disease contained to a small geographical area, and a small population. But with premature publication, the disease was easily modified into a weapon; and used before a countermeasure existed. It is still too late for the dead or dying by the time the countermeasure is developed.
If by witholding the information, lives can be saved... it's worth the extra time. Even if it saves only one life, it's worth the extra time.
It is like the close watch the nuclear countries have over non-nuclear countries (but who have fission reactors for electricity). We still have no countermeasure against a nuclear bomb in a city. But if we can keep any more bombs from being made, there is a greatly reduced chance of a bomb ever detonating in a city.
I'll say this: If nothing else, at least you can see that the Itanium != Opteron competitor. The intended markets are so *completely* different... I finally gave up in apathy from the rabid responses from Opteron fans.
I suspect they aren't so much Opteron fans as they are cheap, and prefer to put up with poorly designed, inexpensive, rushed-design hardware; they are unwilling to fork over the extra cash for a computer with a better and more elegant design and architecture. (I'm not saying that the Opteron isn't an excellent piece of technology. I'm saying that if the same technology & effort that was spent on the Opteron went to something non-x86, things would be even better)
Another hallmark is they want Apple to release OS X for their cheap x86 hardware; somehow believing that it would be at least as stable on the horiffically diverse x86 (and its much more erratic (and often terrible) quality control, to say nothing of quality) than the tightly controlled & heavily tested hardware it currently runs on.
Even when shown the facts they don't want to believe that Apple is a hardware company first, and software second-- and when Apple allowed Mac clones, it almost bankrupted them.
I think he meant go 32-way with a system that uses x86 processors, and is fully x86 compatible.
I know there are systems that are more than 32-way using x86 processors -- but they aren't x86 compatible at all.
The funny thing about PC Rags reviews is that the data and graphics often are accurate but the text and titles are not.
Yeah, but at least you could get real facts if you looked for them. I knew they were way off base around 1990 -- and I was 12 at the time. I mean - I was leagally stupid/gullible back then, and I could see the lies as plain as day. But at least I could find something useful in the tables & graphs shown (and at least more factual).
I tried PC magazine about 3 years ago. I was expecting software comparisons, hardware reviews... things like I was used to seeing.
What I got was 'We compare this shareware html editor to Microsoft FrontPage'. There was nothing to do with PC's -- it was more like "Web Page magazine" Not at all what I was expecting. And even the comparisons were far from fair (a cheap $30 shareware utility to a $500 Microsoft utility?) The funny thing was how well the $30 shareware did against Microsoft's behemoths.
But the problem was still there: It wasn't so much that they weren't objective; I knew that a decade before. But they also gave up on being a technical magazine. They didn't even provide comparisons for anything useful. All 4 'major' office suites had a release (Corel WordPerfect, Lotus SmartSuite, Microsoft Office, and StarOffice) -- but Microsoft office was the only 'review'. Absolutely nothing -- not even a mention of any kind -- of the other office packages.
Nothing comparing the speeds of AMD vs. Intel. Nothing with speed or reliability of Drives, motherboards, video cards. Nothing but web hosting and web authoring tools & services.
So, I gave up on ZD, and pretty much tossed the whole tech magazine journalism altogether; at least in the broad sense. Linux Journal and DDJ are still good (although, at least in the case of LJ, it's targeted to Linux peoples, rather than a more general magazine) here's hoping they keep their technical content.
The only problem with the argument is that the world doesn't have a well-organized research structure; concurrent development of a technology without any sharing of information is a reality. Leibnitz and Newton both developed calculus around the same time, independantly of each other. (Both methods work fine, but are *very* different.)
So, to be honest, a fair amount of the time, the exact same technology is researched at the same time, completely independant of each other; the discovery is literally made twice before anything is published. It's also paid for twice. That doesn't mean that there was waste involved.
Frankly, I can deal with slower development of some technologies if its censorship increases my personal safety. This is espescially true in the case of leathal chemical agents, or deadly (but uncommon) bioagents such as ebola or anthrax.
Yes you can; it's been done before. In many cases, keeping a secret is much easier than the problems the information's misuse can cause. Some things should be kept secret, and world is a better place for it. By keeping the way to create VX a secret (for example), there is a greatly reduced chance of it being used. I don't have to worry about keeping chemical warfare gear and toxin antidotes on my person at all times.
Oklahoma City and Sept. 11 were extremely localized in scope. A small geographic area was affected. I doubt that the synthesis of VX is much more complicated if you know the secret than these acts. Thank God Iraq couldn't find the way to make VX (they abandoned the program in '1989).
If they did, we would have a far greater problem on our hands; a few well-placed aerosol cans could depopulate a city. And how do you protect from that? Nuke Iraq, Afghanistan, Pakistan, and every other country with ties to al-queda? How many innocents would that claim? Would it solve anything?
Sorry-- but the way to solve the problem is to prevent it from happening. Don't spread the information by publishing it. Destroy all records of the existence of the information, and of the information itself. Inform and 'convert' all who know and can reproduce the information of the real (and dire) dangers that can arise should the secret get out. And, barring that-- and for all of the coldness involved, kill or otherwise silence the people involved who can reproduce or easily re-research even parts of the information from their memory.
One thing is a guaranteed fact: If 'sensitive' information is easily available, it will eventually be misused. And many bits of information, when misused, end catastrophically. And it'll always be easier to destroy than create; entropy doesn't help to order/organize anything.
There is no acceptable choice-- either:
murder a few brilliant scientists who is guilty only of finding (even unintentionally) a dangerous bit of information.
or
watch thousands or millions of other perfectly innocent people murdered because of the information's misuse.
It's a hard choice, but a good government tries to serve the greatest number of people; not its most talented people.
Maybe we should try to build a world where people aren't trying to kill us.
If you want to go that route, Stalin and Mao had the right idea. Just kill all of your ideological and social opponents.
So did George Orwell-- the "Thought Police" is a work of sheer brilliance.
So how many people have to die or be 'conditioned' before that becomes a real possibility? Our focus is not the problem; we've had the same kind of societal problems (if not the exact same problems) for all of recorded history.
Most of us prefer to have our own thoughts, ideals, and moralities. And, unsurprisingly, we often disagree. There are always a few people who believe so strongly that they will die for their belief-- they'll lie to themselves, kill, mangle, and destroy for that belief.
Of course, cluster bombs have a treaty banning their use as an anti-personnell weapon. Against unarmored vehicles, fine... but not troops.
For the most part, using any aircraft delivered weapon as an antipersonnel measure is considered a waste of resources. (Exceptions being large distances, or where the Media might get footage of 'friendly' soldiers taking fire.)
If you want a good example of an anti-personell weapon, it's simple non-explosive small arms. Any recently (15 years) designed NATO aircraft ordinance is simply not intended for anti-personnell tasks. It's just not how we fight wars anymore. There is no need to kill off the opposing army-- just cut off their supply and command structures and leave the army hungry, thirsty, confused, and with low amunition. Armies are made of humans-- when getting the next meal is more difficult than fighting the 'enemy', the war is lost.
And, one other point: Only CISC uses microcode (as using dedicated logic to decode is even more complex). RISC uses pure dedicated logic to decode. Using a microcode decoder does reduce complexity of a CISC design (a dedicated logic decoder is trulyl monstrous in a CISC design). RISC is able to get away with using dedicated decoding because they have so few (and considerably more simple) instructions.
And microcode has made huge leaps in efficiency lately -- current designs have microcode that executes close to 95% as fast as pure dedicated logic (so it's only about 5% slower to decode).
And for a link (I didn't try hard) you can do a quick 1-min google search...
But here's some I turned up:
http://www.geocities.com/imac_driver/cpu.html (not SPEC, but it does compare FLOPS (Floating Point Ops/sec) and has links to other comparisons.
Well, I do want to try & stay away from poorly designed games ;) I mean-- if they don't bother to get a somewhat trivial issue like that right, it makes me wonder what other things they fudged.
Let me get this out of the way: I design microprocessors (but mine are more specialized than the generic CPU's pumped out by Intel & AMD). In every case it's much easier, and less costly to use a RISC design. And the RISC design crunches numbers just as quickly.
I'll admit this: I neglected cache memory and cache control circuitry. This takes up about 1/3 of the design, and modifies things a bit-- but the relative size of the execution pipelines to the instruction decode is still roughly 50/50.
Instruction decode also requires the following (and possibly more) sections (That you appear to have missed):
PreDecode Array/CTL
Scan/Align
Instruction Control Unit
Branch Prediction
Floating Point Control
Floating Point Scheduler
And, the Athlon can not decode 3 instructions at the same time. It can dispatch 3 instructions to the integer RISC pipelines at the same time. This is an important distinction, as 1 CISC instruction is equivalent of 1 to 10 (or more) RISC instructions. Each RISC instruction takes 1 clock tick to execute. But a CISC operation can take anywhere from one to several hundred clock cycles to execute. Overall, it takes the same amount of time-- its' just that the RISC does it in several small steps to 1 huge CISC step.
In other words, it decodes 1 x86 instruction, and then can dispatch up to 3 of the 'RISC ops' that are required to do the same task as 1 x86 instruction. (But remember-- 1 x86 instruction may be a lot more than just 3 'RISC ops').
Another thing to remember is that the size of the instruction decode is relatively static-- but the execution stage has grown progressively larger (by adding multiple identical pipelines). Only now when there are 5+ pipelines is the execution stage approaching the size of the decoder.
With fewer pipelines (such as '486, which has 1 FP and 1 integer, which aren't interchangeable), the decode was much larger than the execution logic.
In contrast, a RISC design (single-pipeline) takes up ~15% of the total core size. Superscalar (multiple identical pipelines), and adding cache can reduce the decode to 2-3% of the total die.
Where a CISC design (single pipeline) the decode is ~75% (this is textbook data; from an x86 architecture text). Superscaler, cached designs reduce the decode to ~33% of the total die.
So there's a RISC decode (2-3% of total die), vs. CISC decode (33% of total die)
Don't try to tell me that the difference is small or insignificant. The PowerPC G4 (which SPEC's faster than an Athlon of twice its clock speed) has less than half the transistors of the Athlon. And it requires less than 1/3 of the power.
And what masterpieces of good design are these?
Decoded MP3 is still missing ~90% of the original sound. It's just that human ears have physical limitations that make this possible. (Things like the eardrum moves 'in' faster than it moves 'out')
Transcoding is trying to force Vorbis to work with its hands tied; a decoded wav is still missing 90% of the original signal.
The result is somewhat like a cook-off.
The task: Create a low-fat chicken parmesean recipie that tases as good as a specific high-fat chicken parmesean recipie
(Wav->MP3) Chef 'A' uses his methods to prepare a low-fat chicken parmesean.
(Wav->OGG) Chef 'B' uses his (different) methods to prepare a low-fat chicken parmesean.
In both cases, either chef may choose from the available ingredients of his/her choice to make the best-tasting meal. In this case, chef B wins easily.
But MP3->OGG is like:
Forcing Chef 'B' to take the meal Chef 'A' created, 'un-cook' [decode] it, use the ingredients Chef 'A' used, and create a better meal.
Chef 'B' has to 'start' with the amounts Chef 'A' finished with.
Chef 'B' would have used an entirely different recipie to create his low-fat dish; but since he's limited to what Chef 'A' 'served', he has to make do.
It doesn't matter if Chef 'B' is a much better cook. He's not able to use the ingredients of his choosing. Chef 'B' is essentially limited to modifying Chef 'A's recipie, and therefore he's not able to achieve his potential.
That's what people said when there were mac clones. Mac clones are what almost killed Apple. Apple is a computer company-- only a tiny fraction of their total income is from software, and most of their software profits come from QuickTime-- not OS X.
And dreaming that Apple's market share would grow is just that -- dreaming. Microsoft's strength is Windows-compatibility. And don't forget that since Microsoft owns a significant percentage of Apple, they still get to have their say.
And, finally -- why did Apple sell so much of itself to Microsoft? To get Office ported to OS X quickly. Apple isn't stupid. They know to be successful in the mass market, they have to have Microsoft Office, because that's what everybody uses at work (or school, etc).
And if you think Apple cares about your $129, you're kidding yourself. They want the $800-3000 profit they recieve from selling hardware. They can stay in business (and are doing better than many of the clones makers, like Gateway). Apple cannot stay in business if there are clones running Macintosh.