If you're a mixing engineer counting on tape track crosstalk as a creative tool (eg, the arbitrary positioning of one track next to another yields a different mix than if you order them differently on the tape), go shoot yourself. Now. You have no reason for continued existance.
While I completely agree that mixing is an art, it is also a science... that's why mixing experts are called engineers. It's what we do -- the artistic application of scientific principles. If you start to count on random crap like crosstalk, you're either making it up as you go along, or you're trying to make people think there's black magic in what you do, when there really isn't.
And this only applies to analog tape anyway. If you actually did get crosstalk on digital tape (which you don't, ever), it would result in totally random sound. Most studios were well into the land of digital tape before HDD recording became the established norm.
Yeah, that's exactly it.. there's kind of an arms race in some kinds of music (hard rock, metal) to get progressive louder... despite the fact that ultimately results in lower effective fidelity. Same trick most TV ads employ to sound louder than the show you're watching.
And yeah, it stinks. All of these years of development to deliver a cheap medium (CD) with a 96dB S/N ratio (not to mention the advanced formats that deliver 144dB S/N or thereabouts... at least until you convert to real world analog electronics), and some bands want to use only 5-10% of that.
No, that doesn't work... at least not in C64 BASIC. "WAIT" requires three operands... what you have there is a syntax error. You can SYS 6502, but it just clears the screen.
I don't know if this worked on the PET or not, but consider... Commodore licensed BASIC from Microsoft for a one-time payment of about $10,000 back in the mid 1970s. That was the last time Microsoft was involved in any capacity. Commodore was free to change it any way possible, so as they went forward, Commodore BASIC looked less and less like Microsoft's... bugs were fixed, and I'm sure any cookies found were removed.
On the Commodore 128, though, SYS 12800,123,45,6 is a good one (I'm in there...)
Yes... you can run a program Apple doesn't control get paid for. That is, after all, the highest crime possible on the iPhone. And the reason open interpreters are not permitted. Regular iPhone applications are subject to the approvals process of the Powers That Be at Apple. But they would not be able to approve C64 BASIC, Java, or Shockwave programs (for example). And since Apple know much better than you which applications should be running on the iPhone, that would not be good.
The security argument is a straw man, if it's even being made... no BASIC program could influence anything outside of the 6510 "sandbox" of the emulator. In fact, in these things, even the I/O is emulated... you can't even write to a normal system file, you read from and write to special system files, which are managed under a 1541 disc emulator.
For anyone actually interested in a full fledged C64 emulator, there's a complete, uncrippled version of the Frodo C64 emulator available, free, on the Android apps store right now... though they need to fix the keymap to work with the DROID keyboard:-)
CDMA locking isn't like GSM locking.. there's no real lock. And Motorola seems to be offering the DROID on their web site, at full price.
The problem with moving to another carrier (well, other than the fact that other carrier is Sprint... or maybe Cricket now, too, if you're in a few cities) is that the phones have their ID codes built-in, not portable as with SIM cards. So while I could buy that DROID at Motorola, Sprint will say "oh, hey, that's a Verizon ID you have there, I'm not going to allow that phone on my network". So, while it's effectively a lock, it's server-side, not client-side. So you can't work around it as you can with unlocking on GSM.
You can hack it, if you root the phone. Maybe some other ways, too...
But if you do that, you'll most likely run afoul of the 5GB limit, unless you're a very casual tetherer. What you're really paying for with that additional $30/month is a second 5GB of data.
Somehow, all of the cell phone companies are basically together on the idea that 5GB == Unlimited if you're on a smart phone of any kind and you're not tethering... eg, they think that's all one can in practice use. I'm sure they're wrong... you could exceed that in a relatively short number of hours (5 or so) watching YouTube videos in HD on your Droid (HD on YouTube is 2Mb/s). But that's their claim.
The reason they're on different technologies is that a bunch of different techs for digital cellular evolved in the analog (AMPS) days, and there was no call by the FCC to set any kind of standards.
Nextel started out early... they bought out a new area of spectrum, and used the Motorola-proprietary iDEN system. AT&T launched with D-AMPS, a TDMA tech like GSM, but not compatible with GSM. Verizon launched with CDMA, based on a proprietary technology developed by QualComm. Sprint also went to CDMA. VoiceStream started later, initially very small, using GSM phones based on the European standard, but at the usual US frequencies. They were later acquired by the German Telecom, and thus T-Mobile started out here... still the small guys, but not uselessly small. Cingular also used GSM, and when they acquired AT&T Mobility (the wireless chunk of the AT&T business...) they EOLed AT&T's D-AMPS system (they shut the last ones down last year) and started moving customers over to GSM systems. And they bought the AT&T name, too.
There's some technical advantage to the CDMA technology, which is why all of the 3G technologies are CDMA rather than TDMA. On the other hand, there's a huge advantage to a world-wide standard... it's kind of a shame GSM didn't get an early start in the USA.
For 4G, everyone but Sprint is going to LTE... Sprint is tapping the potential market advantage of being first by pushing out CDMA/WiMax gear, starting with a trial last year in Baltimore and now in several cities (if you're in the right place), rather than wait for the LTE rollout, which should be starting in trials next year.
And it's also a 5GB/month cap on AT&T. I'm not sure if it's a "hard" cap like Verizon (eg, you pay extra for the first byte over 5GB) or a soft cap like Verizon used to have (eg, if you go over by much, or too often, they cancel your contract). But it's there. They explicitly mention 5GB now for any plan that officially allows tethering. For all others, including iPhone, they still say "unlimited", but if you really stay restricted to only the functions you're permitted (there's a long list of permitted and restricted behaviors on AT&T), you can't likely exceed 5GB.
At least Verizon is officially going to allow tethering and higher limits, for a price. As it stands, there's no way to get beyond 5GB per month on the other services... in the USA, of course. Many other countries have more reasonable limits and lower costs.
The extra charge for exchange server access is only on business accounts... don't know if it's provider provisioned or not, but it sounds like it. And makes no difference to consumer clients who simply want to access their exchange server at work.
They're basically telling you it's $30 per month per 5GB. True, a long, long time ago, Verizon claimed EvDO access was unlimited without the asterisk... but even then, it was limited. They just had a higher, secret limit, and when you exceeded it, they kicked you off. Basically, this meant a relatively few top bandwidth users were kicked off every month. It's also true that they do explicitly say something like "Unlimited (5GB)"... they're not hiding this in fine print. They claim to offer an "unlimited" data plan, perhaps without asterisk, but this is only for "media enhanced" phones: sub-smart units that do video and web but don't support applications.
T-Mobile has an "Unlimited" data plan, but they're like Verizon was... they tell you, in the fine print, that if you use it too much, they'll drop speed or kick you off entirely (without qualifying "too much"). Sprint's plan explicitly says "Unlimited" 4G (if you're one of the few in range of a Sprint WiMax connection) and 5GB at 3G speeds. They still hold the option of cutting you off of 4G if your unlimited use is too unlimited... no specific caps mentioned. AT&T's like Verizon... they have a "claims to be unlimited" plan for non-smart phones, and 5GB limits for smart phones or laptop dongles. But they're still very secretive about the 5GB cap (it's not easily found in print), though they do outlaw tethering and just about anything else you could possibly to do chew up excessive bandwidth.
Apple can put anything they like in the EULA... Apple only, or they can even claim that, if you install this, all your bases are belong to them.
Whether any of that's actually legal for them to claim, or even the whole notion of the EULA (you didn't sign anything, you were forced to put the physical product into a state of non-refundability before you could even read their license agreement), is a matter for courts to decide. Apple will never put this agreement in that position... they don't want the validity of the EULA tested in court, because they're likely to lose. So they fight on other things, those that will not threaten the EULA itself.
They stop selling boxed versions of MacOS, and only sell downloads. In order to get the download, you provide the key from your previous version (presumably, the one that came on the Apple machine you're upgrading). Apple not only ensures that only Mac users upgrade, they get the full retail price, no chance of competition, all cash to Apple, no middleman. Just like iTunes! Surprised they aren't using iTunes for this already... it's so very Apple.
1. Yeah, the software companies use the "radar detector" policy. There are several states that declare radar detectors illegal. If they catch you, they will confiscate the detector and give you a ticket. Most people just pay. If you actually pursue it in court, they will drop the case before it reaches a court that could actually test the law (this is covered by Federal Law, not state law... states do not have the ability to limit your reception of radio signals). There are few other laws like this, which stay on the books because they're fairly effective as-is, but would be thrown out if challenged. So no software company wants to see EULAs themselves challenged, but unitl they're specifically declared invalid, the presumption in court is generally that it's a binding license.
2. Apple is a big company with lots of money, and a determined management in this. They don't take on Psystar with one complaint, they think of 20 or 30 things they believe Psystar is doing illegally (sure, according to Apple), and they force them to fight this in court. If was the typical IP case, they'd really be after royalties, and they'd much rather see the case settled in private, out of court. This is why every PC company pays a 3% royalty to IBM on the PC hardware (well, at least those patents still good... I sat in on some of this back in the 1980s). They'll toss 20 or 30 patents your way, and your legal team has to respond. You'll look at these, discuss why you don't violate some of them, why others are pure drek (IBM won a patent on cut and paste between text buffers... if you do this operation in Emacs, yup, that's a violation... even given that the patent was from 1984, and Emacs did this same thing in 1979 and earlier). You can't simply claim "I don't violate these", you have to fight it out in court. Once per patent. If you get through that first stack of 20 or 30, they have another waiting. And another. So no one fights IBM.
Apple can't be as creative, but they will keep at this until they win or until Psystar goes under.
Apple also fights this in the public mind. If you think Psystar is on the verge of losing, you don't buy their computer, you get a Mac, or hopefully, just forget about MacOS entirely and run Linux or Windows. But this way, they suck legal fees out of Psystar, scare off any possible investors, and cut out their source of income.
This is old-skool Big Fish vs. Little Fish tactic, and it doesn't really matter who's right, the Big Fish is usually the winner. There was a battle over audio chips in the 90s... Creative Labs had been the reigning sound card company, but was late to deliver PCI. Synth company Ensoniq and IC startup Aureal both released much better chips for cheap, both targeting Creative Labs bread and butter -- the sound card that goes in the OEM box. Aureal even had the nerve to sell their card with cool native DSP stuff, like a 3D audio API and "surround sound" effects in stereo.
Creative Labs could have taken an engineering approach as a response, but they had money. So they bought Ensoniq and sued Aureal. After some fighting, Aureal's investors got cold feet, they lost OEM income... and finally won their case. Just in time to declare bankruptcy... and have their bones bought up by Creative Labs.
A copy in RAM is fleeting and, particularly in a modern computer, incomplete -- the whole program is likely as not to never exist entirely in RAM.
The best analogy is reading... there is a fleeting "copy" of the book or web page you're reading on your retina during the reading process, as a necessary part of using the book as intended. For a program, same thing.. there is a fleeting "copy" of the program code in RAM, L3, L2, and/or L1 cache as a necessary part of using the program as intended. Remove the book, the image is gone... yank that SATA cable, and you'll have much the same "gone" factor on any modern desktop computer.
Execute in place on flash is based on NOR Flash, which is random access, just like a DRAM. The NAND flash in flash drives cannot be directly driven by a CPU bus... it works best treated as a storage device, just like SSDs do.
Certainly, you can design a storage technology to be memory mapped... Firewire does this. But you don't have the efficiencies of random access there, even if you're good at pretending.
On the other hand, efficient VM systems do usually file map read-only sections anyway. You don't load a program entirely, you just build the memory table corresponding to the on-disc image, and if yo need certain code, take the fault and load just that block. This is inherently very short lived copies, not permanent in any way, and also not being transmitted. This smacks of legal beagles who just don't understand the nature of a computer system in operation coming up with a "Hail Mary" attack that their competition was no better equipped to defend.
Entirely incorrect. Apple shut down the cloning program because they couldn't compete.
When I visited the Apple CHRP group, in January 2007, the fastest computer they had, which they claimed was the fastest Mac ever tested, was a Motorola StarMax. This was CHRP -- the next generation stuff, which went into the PPC7xx machines from Apple, after a fashion (they changed some of the standard stuff, but it was pretty close).
You could hardly claim Motorola was a "shitty" manufacturer...and they made nearly all of the clone motherboards, as well as those for Apple at the time. This was actually much better than the PC industry, since only a tiny handful of companies (IBM, Motorola, UMAX, maybe Power Computing though I'm not sure). Anyone else was using motherboards under sublicense. That's what we did at PIOS Computer... we bought Motorola motherboards through UMAX, and built our own CPU cards (we shipped the world's first production 300MHz MacOS system, only in Europe).
Jobs killed cloning because Apple was losing share... they couldn't compete. They had crappier designs and they were used to getting many times the margin of companies working in the PC industry. Everyone else had PC industry sensibilities. Apple wouldn't have lost so much of their share of the MacOS market if the early "cloners" had been crappy companies... the big guns, like Motorola and IBM, were waiting on the CHRP systems before they did retail themselves. There was pretty much just one way to make the Motorola motherboard... CHRP systems threw the architecture wide open. And every one Apple saw was better than what they were making.
It was they typical American stupid business move: they looked at the next few quarters, not the next few decades. But that was Jobs at the helm. And sure, after falling from 6% of the market down to below 1.5%, they've rebounded... by using 100% PC hardware and still getting people to pay 2x-3x the price of a similarly equipped PC. The iPod, and even more the iPhone has been a bit of a sales booster, but that'll start to fail as the iPhone is eaten by Android over the next several years (these things do take time... even the original IBM PC took a few years to overtake Apple's 8-bit sales).
But make no mistake: Jobs shutting down open MacOS licenseing ensured the dominance of the PC (and pretty much Windows too unless MS keeps screwing up) forevermore.
Apple bumped the MacOS 7.6 up to MacOS 8.0 to end licensing, sure (which tells you just how stupid the lawyers at IBM and Motorola were, that such a thing could be so arbitrary). But this wasn't really due to the lack of money being made on MacOS -- the other Mac-compatible companies (including mine at the time, PIOS Computer AG) were certainly paying to license MacOS.
Apple had too problems. One was simple -- Steve Jobs was and is a control freak, and he wants it his way, or no way. Since he had no control over the "clones" (which were only clones in the sense of everyone using the same Motorola-made motherboards that Apple was using), this was a problem.
The larger problem was that every next-generation machine was faster than Apple's, and their market share had already fallen considerably, even while the MacOS sales base had grown some (all the way to about 6% of the desktop market). Power Computing, UMAX, and Motorola were simply making better computers. PIOS Computer was the first to production with a 300MHz product, the fastest in the day. Gigantic Apple, for whatever reasons, couldn't compete with the other MacOS companies.
It's not simply that they undercharged for the OS... they didn't. But if they had, they could have raised the licensing fees a bit, and done just fine. The problem was, Apple was used to crazy-huge markups and no direct competition with their fairly crappy hardware (early PPC stuff from Apple wasn't as bad as their 68K machine designs, but it was still pretty bad). All these other companies came along with PC industry sense of competition and price margins. Apple's final offer to keep cloning going as the insane notion of charging for MacOS by the type of CPU it was going to run on, up to $500 a copy. That would, in theory, allow Apple to compete fairly with the "clones".
Of course, it was this same logic that led Apple to keep falling, down as low as 1.5% of all desktop sales. They have recovered, largely by making the Mac just another plain old PC that they charge 2x-3x versus Dell or HP... great for the bottom line, as long as they can keep finding suckers to pay like that. Some of this has also been driven by the iPod/iPhone... users who get into those toys are brought into the Apple "reality distortion field", to where buying an Apple seems like a reasonable decision, whereas before, it would have been "what, are you nuts".
That people "require" MS Office has mostly to do with a lack of eduction... the vast majority of PC users don't know a thing about MS Office, or FOSS in general. Or, for that matter, anything that's not firmly entrenched in consumer product advertising. They know Coke and Pepsi, not so much Maine Root, Vignette, or Boylan's. They know Budweiser and Miller, not Flying Fish, Anchor, or Rogue.
On computers, they haven't even been shown that the OS is a choice -- regular folks think "Windows" == PC, and their only alternative is a Mac... also just a PC these days, but that subtlety is even lost on most Mac users. Where will they even hear about Open Office, except maybe when one of us installs it for them. And once done, prepare to support it...
MIPS was the first version of NT.. the x86 version was ported from that. Not much more than a historical footnote anyway, since they were both done by the time NT actually shipped to end users.
Windows NT 4.0 was supported under active contracts in the day (mid 1990s). It was the move to Windows 2000 (NT 5) that caused the problems. Microsoft never did much if any of the porting work anyway, it was all done at other companies. By Win2K development time (late 1990s), they decided to ask for something in excess of $25 million from NEC (MIPS), DEC (Alpha), and Motorola (PowerPC) to allow these companies to do all the work needed to let Microsoft sell Windows 2000 on those various platforms.
I was working closely with Motorola at the time. Windows NT represented less than 5% of their sales within the StarMax (PPC desktop PCs) lines, and even less from partner IBM. Plus, they were high on the crack Apple had been selling them about MacOS compatibility. So they nixed this, as did the others.
I do recall rumors of both PA-RISC and SPARC ports. I don't know if any of those were real.
There is a great big potential for games on ARM machines. Not in direct competition with the PC, but as an alternative.
Gaming is big... an estimated $57 billion in 2009. PC gaming is a big chunk, sure... about $11 billion. But that ought to tell you... PCs are not the dominant form of gaming anymore, even if they're the largest single platform, in terms of dollars.
Witness the example of the Wii... it doesn't really compare to the X-Box 360, PS3, or a PC in gaming... yet it's been remarkably successful. Why? It changed the model, created a new sector of gaming that doesn't count so much on hardware power. Look at Wii Fit... that's doing about $20 million per week, internationally.. sold over 20 million units, about $2 billion in total sales. No one would have even considered that a reasonably risk in the PC gaming world. Same with the GameBoy (ARM-based, since the Gameboy Advance).
The iPhone (ARM too) has done well... over $250 million in gaming sales from 2008-2009. Not PC big, but consider, these are $1-$5 games, largely. And this is more digital download revenue than any console platform to date. Current estimates put handheld gaming at about $12 billion by 2014... certainly on par with the PC of today... we'll bave to see if PC games sales drop or increase in the next five years. And "handheld" means nearly all ARM based. They also currently expect Apple to own about 25% of that gaming market.
Actually, NT was on the MIPS first, thank to NEC championing the ARC initiative, which defined a MIPS-based platform as an x86 alternative. After that, there were x86, Alpha, and PowerPC ports in public, and a long rumored SPARC port.
NT on the low-level did many things wright. They defined a HAL (Hardware Abstraction Layer) which, when used correctly, accounts for driver-level differences in machine architecture, even things like endian-ness.
The big problem in the non-x86 market for NT was that Microsoft didn't write these, they just enabled them. So, for example, Motorola did the NT port to PowerPC. They had to pay Microsoft millions for the opportunity. Even the MIPS version, while started at Microsoft, was spun out to NEC's control. DEC did the Alpha port... one reason why DEC's was the only version that also ran some x86 32-bit binaries (they provided a dynamic code translation mechanism, no one else did).
I would put MS's moving Windows to ARM as a self-preservation move... but maybe too late already. ARM is flourishing on open source software, and the only time anyone even mentions Windows is when ARMs are pushed up into competition with x86s. But they pretty much already own most of the rest of computing. The reason is simple: they're the IBM PC of the CPU world.
The PC didn't succeed because it was a good design... it was initially a horrible design. It won early on through a few market factors (the IBM name, the math chip's effect on the use of spreadsheets as models), but it dominated because no single entity had control.. anyone could make a PC. Today, since Advanced RISC Machines has chosen to just remain a design center, not a manufacturer, anyone can make an ARM chip. After a decade+ of that, it's awfully hard to find a significant electronic device powered by something else. There are billion of ARMS sold per year.
Windows might have a shot on ARM-based netbooks, and perhaps if ARM chips upscale to be even more into x86 territory. But it fails in all of the big market things ARM is doing well today, some of which, like pocket computers (eg, smartphones), even WinCE has largely failed. These are going to be the future of computing for many users, not PCs.
Indeed... Linux is starting to benefit from the long history of a consistent driver model.
When a new version of Windows comes out, more often than not, the driver model has changed. Microsoft may have the occasional legit reason for changing the driver model, but unless the current model is seriously broken (and after a half dozen "this breaks" revisions, you would think they finally have it right.. only, it's changed again for Win7), they could simply add additional functions... old drivers still work, new ones offer new features.
The problem here, from MS's point of view, is that drivers have been a magnificent lever in implementing MS's policies. When MS sells Windows to OEMs (Dell, Sony, HP, Toshiba, etc), they get a discounted price that'd dependent on every driver shipped in a new PC of any kind having passed MS's driver certification. This isn't really a test against buggy drivers (though that's how it's sold), but a test of toeing the line. The end result is that no one without current, certified drivers gets in any new PC. To a lesser effect, they also don't get that shiny "Made for Windows" sticker on the retail box. This is also the reason that pretty much all Windows hardware ships with old drivers.
So, there might have been no problem in running XP drivers in Vista or Win7. But when XP came out, 64-bit support was optional. In Vista, it was mandatory... can't get certification without it. So, to get all that stuff written, they force everyone to revise every driver. With Win7, they didn't entirely break the model, but they added features to some drivers.. many things HAD to be rewritten. The hardware companies put up with this because [a] Microsoft is the 8,000lbs gorilla, and [b] in doing this, they're weeding out perfectly good older hardware, so at least some hardware vendors will benefit.
But the bottom line is that supporting new versions of Linux is easy, once you get that knowledge in-house. There may be no reason to update drivers very often, and so they build a collection of good drivers. Given these don't need to be revised, the stuff that does need revision, and particularly the new stuff, can be released faster, even with fewer people on the job.
I think we're also potentially poised on a "Great Awakening" about the advantages of Open Source. That's been understood, behind the scenes, among server people and embedded device companies, but Android, for one, is going to break this wide open -- Android will be the thing that knocks out Apple and RIM in the smart phone business... that business that is currently defining the major computing platform of the future. Netbook and even desktop companies will need to learn from this.. the successful ones will.
Windows itself may even be a questionably thing in today's market. If you're talking about WinCE, it's already on ARM, and not a leading applications platform, largely due to Microsoft not pushing that for a decade or so. On the desktop, there's no binary compatibility, and at least for the present, ARM could never compete except against low-end Netbooks.
The simple fact is that ARM is everywhere else. Some years back, I designed a high-end digital radio controller using a single TI MSP430 microcontroller. It worked fine, but the memory capacity on the device (60K Flash) was a limiting factor. The replacement for that radio controller had not one but two NXP ARM microcontrollers (one 256K flash, one 32K flash)... it sold for $200 less (MSRP), and the total cost of a USB solution was less (and much faster) with the ARM. They have not entirely conquered embedded, but they're strong on everything between 8-bit and desktop-class, already.
So while I think ARM could eventually move to the desktop, I think the real interesting thing here is that the desktop is more likely moving to the ARM. When personal computing began, everyone needed a personal computer, on the desktop... there was no other way. Today, there are already a class of people doing more traditional computer stuff on their "phones"... which aren't any just phones, anymore than a PC is just a wordprocessor. The latest generation of "phone" had 500-1000MHz CPU, with >VGA class screens, 3D video acceleration, etc.
In the not too distant future (in the USA.. this is already happening in Japan), many users will consider the "pocket computer" as their primary computing device. When you get home, drop it into the charger/cradle, and it'll hook into a regular monitor, keyboard, mouse, etc. That device will be ARM based. Those who need more computation power than the "phone" can provide might easily tape the CPU power of their games console or Blu-Ray player.. no real need for that, usually, in a mobile situation. One the internet connections are fast enough, this could be done "in the cloud", too... just like the data storage the modern mobile systems (Android, Palm WebOS) are already tapping for mobile computing.
This is an ideal time to keep Windows out of the way mainstream mobile computing is headed, and eventually, mainstream computing. No, it won't replace all desktop PCs, but really, there are only so many desktop users who really have any use for a quad-processor 3GHz PC anyway. And while I have one of these, I'm looking toward the mobile device becoming more of a central device for "regular" computing, and the PC for the workstation class stuff... NOW.
A hardware firewall is just a software firewall running on some else's CPU.
Though there are a few PCs coming out with an auxilary ARM CPU... you boot up on the ARM quick, run Linux or Android or something, to do simple things, but boot the full x86 for "real work". You could put a firewall and router on the ARM, give it something to do when the "full PC" mode is activate. Hmmm.....
Slashdot Mirror
If you're a mixing engineer counting on tape track crosstalk as a creative tool (eg, the arbitrary positioning of one track next to another yields a different mix than if you order them differently on the tape), go shoot yourself. Now. You have no reason for continued existance.
While I completely agree that mixing is an art, it is also a science... that's why mixing experts are called engineers. It's what we do -- the artistic application of scientific principles. If you start to count on random crap like crosstalk, you're either making it up as you go along, or you're trying to make people think there's black magic in what you do, when there really isn't.
And this only applies to analog tape anyway. If you actually did get crosstalk on digital tape (which you don't, ever), it would result in totally random sound. Most studios were well into the land of digital tape before HDD recording became the established norm.
Yeah, that's exactly it.. there's kind of an arms race in some kinds of music (hard rock, metal) to get progressive louder... despite the fact that ultimately results in lower effective fidelity. Same trick most TV ads employ to sound louder than the show you're watching.
And yeah, it stinks. All of these years of development to deliver a cheap medium (CD) with a 96dB S/N ratio (not to mention the advanced formats that deliver 144dB S/N or thereabouts... at least until you convert to real world analog electronics), and some bands want to use only 5-10% of that.
No, that doesn't work... at least not in C64 BASIC. "WAIT" requires three operands... what you have there is a syntax error. You can SYS 6502, but it just clears the screen.
I don't know if this worked on the PET or not, but consider... Commodore licensed BASIC from Microsoft for a one-time payment of about $10,000 back in the mid 1970s. That was the last time Microsoft was involved in any capacity. Commodore was free to change it any way possible, so as they went forward, Commodore BASIC looked less and less like Microsoft's... bugs were fixed, and I'm sure any cookies found were removed.
On the Commodore 128, though, SYS 12800,123,45,6 is a good one (I'm in there...)
Yes... you can run a program Apple doesn't control get paid for. That is, after all, the highest crime possible on the iPhone. And the reason open interpreters are not permitted. Regular iPhone applications are subject to the approvals process of the Powers That Be at Apple. But they would not be able to approve C64 BASIC, Java, or Shockwave programs (for example). And since Apple know much better than you which applications should be running on the iPhone, that would not be good.
The security argument is a straw man, if it's even being made... no BASIC program could influence anything outside of the 6510 "sandbox" of the emulator. In fact, in these things, even the I/O is emulated... you can't even write to a normal system file, you read from and write to special system files, which are managed under a 1541 disc emulator.
For anyone actually interested in a full fledged C64 emulator, there's a complete, uncrippled version of the Frodo C64 emulator available, free, on the Android apps store right now... though they need to fix the keymap to work with the DROID keyboard :-)
CDMA locking isn't like GSM locking.. there's no real lock. And Motorola seems to be offering the DROID on their web site, at full price.
The problem with moving to another carrier (well, other than the fact that other carrier is Sprint... or maybe Cricket now, too, if you're in a few cities) is that the phones have their ID codes built-in, not portable as with SIM cards. So while I could buy that DROID at Motorola, Sprint will say "oh, hey, that's a Verizon ID you have there, I'm not going to allow that phone on my network". So, while it's effectively a lock, it's server-side, not client-side. So you can't work around it as you can with unlocking on GSM.
You can hack it, if you root the phone. Maybe some other ways, too...
But if you do that, you'll most likely run afoul of the 5GB limit, unless you're a very casual tetherer. What you're really paying for with that additional $30/month is a second 5GB of data.
Somehow, all of the cell phone companies are basically together on the idea that 5GB == Unlimited if you're on a smart phone of any kind and you're not tethering... eg, they think that's all one can in practice use. I'm sure they're wrong... you could exceed that in a relatively short number of hours (5 or so) watching YouTube videos in HD on your Droid (HD on YouTube is 2Mb/s). But that's their claim.
The reason they're on different technologies is that a bunch of different techs for digital cellular evolved in the analog (AMPS) days, and there was no call by the FCC to set any kind of standards.
Nextel started out early ... they bought out a new area of spectrum, and used the Motorola-proprietary iDEN system. AT&T launched with D-AMPS, a TDMA tech like GSM, but not compatible with GSM. Verizon launched with CDMA, based on a proprietary technology developed by QualComm. Sprint also went to CDMA. VoiceStream started later, initially very small, using GSM phones based on the European standard, but at the usual US frequencies. They were later acquired by the German Telecom, and thus T-Mobile started out here... still the small guys, but not uselessly small. Cingular also used GSM, and when they acquired AT&T Mobility (the wireless chunk of the AT&T business...) they EOLed AT&T's D-AMPS system (they shut the last ones down last year) and started moving customers over to GSM systems. And they bought the AT&T name, too.
There's some technical advantage to the CDMA technology, which is why all of the 3G technologies are CDMA rather than TDMA. On the other hand, there's a huge advantage to a world-wide standard... it's kind of a shame GSM didn't get an early start in the USA.
For 4G, everyone but Sprint is going to LTE... Sprint is tapping the potential market advantage of being first by pushing out CDMA/WiMax gear, starting with a trial last year in Baltimore and now in several cities (if you're in the right place), rather than wait for the LTE rollout, which should be starting in trials next year.
And it's also a 5GB/month cap on AT&T. I'm not sure if it's a "hard" cap like Verizon (eg, you pay extra for the first byte over 5GB) or a soft cap like Verizon used to have (eg, if you go over by much, or too often, they cancel your contract). But it's there. They explicitly mention 5GB now for any plan that officially allows tethering. For all others, including iPhone, they still say "unlimited", but if you really stay restricted to only the functions you're permitted (there's a long list of permitted and restricted behaviors on AT&T), you can't likely exceed 5GB.
At least Verizon is officially going to allow tethering and higher limits, for a price. As it stands, there's no way to get beyond 5GB per month on the other services... in the USA, of course. Many other countries have more reasonable limits and lower costs.
The extra charge for exchange server access is only on business accounts... don't know if it's provider provisioned or not, but it sounds like it. And makes no difference to consumer clients who simply want to access their exchange server at work.
They're basically telling you it's $30 per month per 5GB. True, a long, long time ago, Verizon claimed EvDO access was unlimited without the asterisk... but even then, it was limited. They just had a higher, secret limit, and when you exceeded it, they kicked you off. Basically, this meant a relatively few top bandwidth users were kicked off every month. It's also true that they do explicitly say something like "Unlimited (5GB)"... they're not hiding this in fine print. They claim to offer an "unlimited" data plan, perhaps without asterisk, but this is only for "media enhanced" phones: sub-smart units that do video and web but don't support applications.
T-Mobile has an "Unlimited" data plan, but they're like Verizon was... they tell you, in the fine print, that if you use it too much, they'll drop speed or kick you off entirely (without qualifying "too much"). Sprint's plan explicitly says "Unlimited" 4G (if you're one of the few in range of a Sprint WiMax connection) and 5GB at 3G speeds. They still hold the option of cutting you off of 4G if your unlimited use is too unlimited... no specific caps mentioned. AT&T's like Verizon... they have a "claims to be unlimited" plan for non-smart phones, and 5GB limits for smart phones or laptop dongles. But they're still very secretive about the 5GB cap (it's not easily found in print), though they do outlaw tethering and just about anything else you could possibly to do chew up excessive bandwidth.
Yes... the Mac costs 3x as much as a crappy PC, and 2x as much as a very good PC.
Apple can put anything they like in the EULA... Apple only, or they can even claim that, if you install this, all your bases are belong to them.
Whether any of that's actually legal for them to claim, or even the whole notion of the EULA (you didn't sign anything, you were forced to put the physical product into a state of non-refundability before you could even read their license agreement), is a matter for courts to decide. Apple will never put this agreement in that position... they don't want the validity of the EULA tested in court, because they're likely to lose. So they fight on other things, those that will not threaten the EULA itself.
Apple could fix this themselves, easily enough.
They stop selling boxed versions of MacOS, and only sell downloads. In order to get the download, you provide the key from your previous version (presumably, the one that came on the Apple machine you're upgrading). Apple not only ensures that only Mac users upgrade, they get the full retail price, no chance of competition, all cash to Apple, no middleman. Just like iTunes! Surprised they aren't using iTunes for this already... it's so very Apple.
1. Yeah, the software companies use the "radar detector" policy. There are several states that declare radar detectors illegal. If they catch you, they will confiscate the detector and give you a ticket. Most people just pay. If you actually pursue it in court, they will drop the case before it reaches a court that could actually test the law (this is covered by Federal Law, not state law... states do not have the ability to limit your reception of radio signals). There are few other laws like this, which stay on the books because they're fairly effective as-is, but would be thrown out if challenged. So no software company wants to see EULAs themselves challenged, but unitl they're specifically declared invalid, the presumption in court is generally that it's a binding license.
2. Apple is a big company with lots of money, and a determined management in this. They don't take on Psystar with one complaint, they think of 20 or 30 things they believe Psystar is doing illegally (sure, according to Apple), and they force them to fight this in court. If was the typical IP case, they'd really be after royalties, and they'd much rather see the case settled in private, out of court. This is why every PC company pays a 3% royalty to IBM on the PC hardware (well, at least those patents still good... I sat in on some of this back in the 1980s). They'll toss 20 or 30 patents your way, and your legal team has to respond. You'll look at these, discuss why you don't violate some of them, why others are pure drek (IBM won a patent on cut and paste between text buffers... if you do this operation in Emacs, yup, that's a violation... even given that the patent was from 1984, and Emacs did this same thing in 1979 and earlier). You can't simply claim "I don't violate these", you have to fight it out in court. Once per patent. If you get through that first stack of 20 or 30, they have another waiting. And another. So no one fights IBM.
Apple can't be as creative, but they will keep at this until they win or until Psystar goes under.
Apple also fights this in the public mind. If you think Psystar is on the verge of losing, you don't buy their computer, you get a Mac, or hopefully, just forget about MacOS entirely and run Linux or Windows. But this way, they suck legal fees out of Psystar, scare off any possible investors, and cut out their source of income.
This is old-skool Big Fish vs. Little Fish tactic, and it doesn't really matter who's right, the Big Fish is usually the winner. There was a battle over audio chips in the 90s... Creative Labs had been the reigning sound card company, but was late to deliver PCI. Synth company Ensoniq and IC startup Aureal both released much better chips for cheap, both targeting Creative Labs bread and butter -- the sound card that goes in the OEM box.
Aureal even had the nerve to sell their card with cool native DSP stuff, like a 3D audio API and "surround sound" effects in stereo.
Creative Labs could have taken an engineering approach as a response, but they had money. So they bought Ensoniq and sued Aureal. After some fighting, Aureal's investors got cold feet, they lost OEM income... and finally won their case. Just in time to declare bankruptcy... and have their bones bought up by Creative Labs.
The Law is such stuff!
A copy in RAM is fleeting and, particularly in a modern computer, incomplete -- the whole program is likely as not to never exist entirely in RAM.
The best analogy is reading... there is a fleeting "copy" of the book or web page you're reading on your retina during the reading process, as a necessary part of using the book as intended. For a program, same thing.. there is a fleeting "copy" of the program code in RAM, L3, L2, and/or L1 cache as a necessary part of using the program as intended. Remove the book, the image is gone... yank that SATA cable, and you'll have much the same "gone" factor on any modern desktop computer.
Execute in place on flash is based on NOR Flash, which is random access, just like a DRAM. The NAND flash in flash drives cannot be directly driven by a CPU bus... it works best treated as a storage device, just like SSDs do.
Certainly, you can design a storage technology to be memory mapped... Firewire does this. But you don't have the efficiencies of random access there, even if you're good at pretending.
On the other hand, efficient VM systems do usually file map read-only sections anyway. You don't load a program entirely, you just build the memory table corresponding to the on-disc image, and if yo need certain code, take the fault and load just that block. This is inherently very short lived copies, not permanent in any way, and also not being transmitted. This smacks of legal beagles who just don't understand the nature of a computer system in operation coming up with a "Hail Mary" attack that their competition was no better equipped to defend.
Entirely incorrect. Apple shut down the cloning program because they couldn't compete.
When I visited the Apple CHRP group, in January 2007, the fastest computer they had, which they claimed was the fastest Mac ever tested, was a Motorola StarMax. This was CHRP -- the next generation stuff, which went into the PPC7xx machines from Apple, after a fashion (they changed some of the standard stuff, but it was pretty close).
You could hardly claim Motorola was a "shitty" manufacturer...and they made nearly all of the clone motherboards, as well as those for Apple at the time. This was actually much better than the PC industry, since only a tiny handful of companies (IBM, Motorola, UMAX, maybe Power Computing though I'm not sure). Anyone else was using motherboards under sublicense. That's what we did at PIOS Computer... we bought Motorola motherboards through UMAX, and built our own CPU cards (we shipped the world's first production 300MHz MacOS system, only in Europe).
Jobs killed cloning because Apple was losing share... they couldn't compete. They had crappier designs and they were used to getting many times the margin of companies working in the PC industry. Everyone else had PC industry sensibilities. Apple wouldn't have lost so much of their share of the MacOS market if the early "cloners" had been crappy companies... the big guns, like Motorola and IBM, were waiting on the CHRP systems before they did retail themselves. There was pretty much just one way to make the Motorola motherboard... CHRP systems threw the architecture wide open. And every one Apple saw was better than what they were making.
It was they typical American stupid business move: they looked at the next few quarters, not the next few decades. But that was Jobs at the helm. And sure, after falling from 6% of the market down to below 1.5%, they've rebounded... by using 100% PC hardware and still getting people to pay 2x-3x the price of a similarly equipped PC. The iPod, and even more the iPhone has been a bit of a sales booster, but that'll start to fail as the iPhone is eaten by Android over the next several years (these things do take time... even the original IBM PC took a few years to overtake Apple's 8-bit sales).
But make no mistake: Jobs shutting down open MacOS licenseing ensured the dominance of the PC (and pretty much Windows too unless MS keeps screwing up) forevermore.
Or at least until the handhelds take over :-)
Apple bumped the MacOS 7.6 up to MacOS 8.0 to end licensing, sure (which tells you just how stupid the lawyers at IBM and Motorola were, that such a thing could be so arbitrary). But this wasn't really due to the lack of money being made on MacOS -- the other Mac-compatible companies (including mine at the time, PIOS Computer AG) were certainly paying to license MacOS.
Apple had too problems. One was simple -- Steve Jobs was and is a control freak, and he wants it his way, or no way. Since he had no control over the "clones" (which were only clones in the sense of everyone using the same Motorola-made motherboards that Apple was using), this was a problem.
The larger problem was that every next-generation machine was faster than Apple's, and their market share had already fallen considerably, even while the MacOS sales base had grown some (all the way to about 6% of the desktop market). Power Computing, UMAX, and Motorola were simply making better computers. PIOS Computer was the first to production with a 300MHz product, the fastest in the day. Gigantic Apple, for whatever reasons, couldn't compete with the other MacOS companies.
It's not simply that they undercharged for the OS... they didn't. But if they had, they could have raised the licensing fees a bit, and done just fine. The problem was, Apple was used to crazy-huge markups and no direct competition with their fairly crappy hardware (early PPC stuff from Apple wasn't as bad as their 68K machine designs, but it was still pretty bad). All these other companies came along with PC industry sense of competition and price margins. Apple's final offer to keep cloning going as the insane notion of charging for MacOS by the type of CPU it was going to run on, up to $500 a copy. That would, in theory, allow Apple to compete fairly with the "clones".
Of course, it was this same logic that led Apple to keep falling, down as low as 1.5% of all desktop sales. They have recovered, largely by making the Mac just another plain old PC that they charge 2x-3x versus Dell or HP... great for the bottom line, as long as they can keep finding suckers to pay like that. Some of this has also been driven by the iPod/iPhone... users who get into those toys are brought into the Apple "reality distortion field", to where buying an Apple seems like a reasonable decision, whereas before, it would have been "what, are you nuts".
Whoops... should have read "don't know a thing about Open Office".. you figured that out, I'm sure.
That people "require" MS Office has mostly to do with a lack of eduction... the vast majority of PC users don't know a thing about MS Office, or FOSS in general. Or, for that matter, anything that's not firmly entrenched in consumer product advertising. They know Coke and Pepsi, not so much Maine Root, Vignette, or Boylan's. They know Budweiser and Miller, not Flying Fish, Anchor, or Rogue.
On computers, they haven't even been shown that the OS is a choice -- regular folks think "Windows" == PC, and their only alternative is a Mac... also just a PC these days, but that subtlety is even lost on most Mac users. Where will they even hear about Open Office, except maybe when one of us installs it for them. And once done, prepare to support it...
MIPS was the first version of NT.. the x86 version was ported from that. Not much more than a historical footnote anyway, since they were both done by the time NT actually shipped to end users.
Windows NT 4.0 was supported under active contracts in the day (mid 1990s). It was the move to Windows 2000 (NT 5) that caused the problems. Microsoft never did much if any of the porting work anyway, it was all done at other companies. By Win2K development time (late 1990s), they decided to ask for something in excess of $25 million from NEC (MIPS), DEC (Alpha), and Motorola (PowerPC) to allow these companies to do all the work needed to let Microsoft sell Windows 2000 on those various platforms.
I was working closely with Motorola at the time. Windows NT represented less than 5% of their sales within the StarMax (PPC desktop PCs) lines, and even less from partner IBM. Plus, they were high on the crack Apple had been selling them about MacOS compatibility. So they nixed this, as did the others.
I do recall rumors of both PA-RISC and SPARC ports. I don't know if any of those were real.
There is a great big potential for games on ARM machines. Not in direct competition with the PC, but as an alternative.
Gaming is big... an estimated $57 billion in 2009. PC gaming is a big chunk, sure... about $11 billion. But that ought to tell you... PCs are not the dominant form of gaming anymore, even if they're the largest single platform, in terms of dollars.
Witness the example of the Wii... it doesn't really compare to the X-Box 360, PS3, or a PC in gaming... yet it's been remarkably successful. Why? It changed the model, created a new sector of gaming that doesn't count so much on hardware power. Look at Wii Fit... that's doing about $20 million per week, internationally.. sold over 20 million units, about $2 billion in total sales. No one would have even considered that a reasonably risk in the PC gaming world. Same with the GameBoy (ARM-based, since the Gameboy Advance).
The iPhone (ARM too) has done well... over $250 million in gaming sales from 2008-2009. Not PC big, but consider, these are $1-$5 games, largely. And this is more digital download revenue than any console platform to date. Current estimates put handheld gaming at about $12 billion by 2014... certainly on par with the PC of today... we'll bave to see if PC games sales drop or increase in the next five years. And "handheld" means nearly all ARM based. They also currently expect Apple to own about 25% of that gaming market.
Actually, NT was on the MIPS first, thank to NEC championing the ARC initiative, which defined a MIPS-based platform as an x86 alternative. After that, there were x86, Alpha, and PowerPC ports in public, and a long rumored SPARC port.
NT on the low-level did many things wright. They defined a HAL (Hardware Abstraction Layer) which, when used correctly, accounts for driver-level differences in machine architecture, even things like endian-ness.
The big problem in the non-x86 market for NT was that Microsoft didn't write these, they just enabled them. So, for example, Motorola did the NT port to PowerPC. They had to pay Microsoft millions for the opportunity. Even the MIPS version, while started at Microsoft, was spun out to NEC's control. DEC did the Alpha port... one reason why DEC's was the only version that also ran some x86 32-bit binaries (they provided a dynamic code translation mechanism, no one else did).
I would put MS's moving Windows to ARM as a self-preservation move... but maybe too late already. ARM is flourishing on open source software, and the only time anyone even mentions Windows is when ARMs are pushed up into competition with x86s. But they pretty much already own most of the rest of computing. The reason is simple: they're the IBM PC of the CPU world.
The PC didn't succeed because it was a good design... it was initially a horrible design. It won early on through a few market factors (the IBM name, the math chip's effect on the use of spreadsheets as models), but it dominated because no single entity had control.. anyone could make a PC. Today, since Advanced RISC Machines has chosen to just remain a design center, not a manufacturer, anyone can make an ARM chip. After a decade+ of that, it's awfully hard to find a significant electronic device powered by something else. There are billion of ARMS sold per year.
Windows might have a shot on ARM-based netbooks, and perhaps if ARM chips upscale to be even more into x86 territory. But it fails in all of the big market things ARM is doing well today, some of which, like pocket computers (eg, smartphones), even WinCE has largely failed. These are going to be the future of computing for many users, not PCs.
Indeed... Linux is starting to benefit from the long history of a consistent driver model.
When a new version of Windows comes out, more often than not, the driver model has changed. Microsoft may have the occasional legit reason for changing the driver model, but unless the current model is seriously broken (and after a half dozen "this breaks" revisions, you would think they finally have it right.. only, it's changed again for Win7), they could simply add additional functions... old drivers still work, new ones offer new features.
The problem here, from MS's point of view, is that drivers have been a magnificent lever in implementing MS's policies. When MS sells Windows to OEMs (Dell, Sony, HP, Toshiba, etc), they get a discounted price that'd dependent on every driver shipped in a new PC of any kind having passed MS's driver certification. This isn't really a test against buggy drivers (though that's how it's sold), but a test of toeing the line. The end result is that no one without current, certified drivers gets in any new PC. To a lesser effect, they also don't get that shiny "Made for Windows" sticker on the retail box. This is also the reason that pretty much all Windows hardware ships with old drivers.
So, there might have been no problem in running XP drivers in Vista or Win7. But when XP came out, 64-bit support was optional. In Vista, it was mandatory... can't get certification without it. So, to get all that stuff written, they force everyone to revise every driver. With Win7, they didn't entirely break the model, but they added features to some drivers.. many things HAD to be rewritten. The hardware companies put up with this because [a] Microsoft is the 8,000lbs gorilla, and [b] in doing this, they're weeding out perfectly good older hardware, so at least some hardware vendors will benefit.
But the bottom line is that supporting new versions of Linux is easy, once you get that knowledge in-house. There may be no reason to update drivers very often, and so they build a collection of good drivers. Given these don't need to be revised, the stuff that does need revision, and particularly the new stuff, can be released faster, even with fewer people on the job.
I think we're also potentially poised on a "Great Awakening" about the advantages of Open Source. That's been understood, behind the scenes, among server people and embedded device companies, but Android, for one, is going to break this wide open -- Android will be the thing that knocks out Apple and RIM in the smart phone business... that business that is currently defining the major computing platform of the future. Netbook and even desktop companies will need to learn from this.. the successful ones will.
Windows itself may even be a questionably thing in today's market. If you're talking about WinCE, it's already on ARM, and not a leading applications platform, largely due to Microsoft not pushing that for a decade or so. On the desktop, there's no binary compatibility, and at least for the present, ARM could never compete except against low-end Netbooks.
The simple fact is that ARM is everywhere else. Some years back, I designed a high-end digital radio controller using a single TI MSP430 microcontroller. It worked fine, but the memory capacity on the device (60K Flash) was a limiting factor. The replacement for that radio controller had not one but two NXP ARM microcontrollers (one 256K flash, one 32K flash)... it sold for $200 less (MSRP), and the total cost of a USB solution was less (and much faster) with the ARM. They have not entirely conquered embedded, but they're strong on everything between 8-bit and desktop-class, already.
So while I think ARM could eventually move to the desktop, I think the real interesting thing here is that the desktop is more likely moving to the ARM. When personal computing began, everyone needed a personal computer, on the desktop... there was no other way. Today, there are already a class of people doing more traditional computer stuff on their "phones"... which aren't any just phones, anymore than a PC is just a wordprocessor. The latest generation of "phone" had 500-1000MHz CPU, with >VGA class screens, 3D video acceleration, etc.
In the not too distant future (in the USA.. this is already happening in Japan), many users will consider the "pocket computer" as their primary computing device. When you get home, drop it into the charger/cradle, and it'll hook into a regular monitor, keyboard, mouse, etc. That device will be ARM based. Those who need more computation power than the "phone" can provide might easily tape the CPU power of their games console or Blu-Ray player.. no real need for that, usually, in a mobile situation. One the internet connections are fast enough, this could be done "in the cloud", too... just like the data storage the modern mobile systems (Android, Palm WebOS) are already tapping for mobile computing.
This is an ideal time to keep Windows out of the way mainstream mobile computing is headed, and eventually, mainstream computing. No, it won't replace all desktop PCs, but really, there are only so many desktop users who really have any use for a quad-processor 3GHz PC anyway. And while I have one of these, I'm looking toward the mobile device becoming more of a central device for "regular" computing, and the PC for the workstation class stuff... NOW.
A hardware firewall is just a software firewall running on some else's CPU.
Though there are a few PCs coming out with an auxilary ARM CPU... you boot up on the ARM quick, run Linux or Android or something, to do simple things, but boot the full x86 for "real work". You could put a firewall and router on the ARM, give it something to do when the "full PC" mode is activate. Hmmm.....