It's amazing how many people want to nitpick about various flaws in computers.
There is exactly one flaw that renders the entire thing moot: Computers do exactly what you tell them to do.
This not only includes lying, it includes lying about lying, and lying about the lying it's lying about, and lying about that lying, and then lying about that lie. Under whatever circumstances it is told to lie under.
It is theoretically impossible to confirm that a voting computer operates the way it is required by law to operate. Not just practically impossible, theoretically impossible.
Yes, they test them, but it is impossible to know that it hasn't been programmed to give the wrong result only when X happens, where X could be 'Used continually for three hours' or 'When a specific ballot is entered'.
They can look at the code, but they have no way to know that the code the computer presents them as the code is, in fact, the code the computer is actually running. It's actually impossible to do this without a detailed inspection of every byte of code in the computer by removing all all storage devices and looking at them in a trusted computer, which not only do they not do, but couldn't do, as parts of the executing code are in the BIOS and other firmware in computer.
Considering that there's several billion bytes of code in Windows alone, which is the base system they use, and that the amount of work required is exponential based on the file size, in it is not possible to even attempt this for any reasonable price.
Just look at the amount of time and money spent developing air-traffic control systems if you don't believe me, and then realize a) voting machines are running off-the-shelf OSes instead of, basically, no OS in ATC systems, b) they don't have to catch deliberate 'mistakes' in ATC systems to alter the vote, c) ATC systems are monitored while used by trained professionals, unlike voting machines, which are operated, duh, in secret by random people, d) have to have physically secure boxes, unlike ATC systems which are trivially to 'hack' if you can physically reach them, e) voting machines either have to be portable or we have to build billions of dollars of buildings to contain them permanently, with staff.
A real electronic voting machine system would cost more than ten ATC systems, and ATC systems can run to hundreds of millions of dollars. We're literally talking about a billion dollar development effort here, maybe one hundred million dollars worth of machines, and hundreds of millions of dollars a year in maintenance to do correctly and securely. And that's just if we do it once for the whole country.
Or we can do it shittily and have companies just slapping things together and misleading and bribing (with votes, maybe?) the legislature to use them. Seems to be working so far, if by 'working' you mean 'not working'.
Of course, there's the other option that computer scientists have been screaming about since the start: Have the computer print a OCRable ballot, and have computers and humans count the ballot later. Which would cost a good deal less than the machines they're building now, it would basically be a cash register.
"oh by the way, the last 800,000 computers we shipped had different hardware than reported in our disclosures."
Which was my damn point. Either Apple already lied about the hardware, and thus, are already in trouble, or they didn't lie and are fine. If you give someone something for free, it's not revenue.
And they don't have to 'restate' their previous hardware as n-compatible just because they, or someone else, created some software to make it so. It wasn't n-compatible, and, as far as Apple cares, it's still not.
The idea that it's illegal to give away something that enables a previously announced feature on hardware is completely batshit insane, and Apple is consequently a) batshit insane itself, b) lying to make money, c) lying to look like it's complying with every single law, even ones that don't actually exist, or d) lying to make SOx look like a bad law, which companies have been doing ever since it was passed.
Or some combination thereof. Of course, all this is from some guy's blog, so it's probably crap to start with.
To repeat: A lot of companies have taken issue with SOx restricting their unethical activities, and as such have started blaming all sorts of crap on it. SOx does not require any behavior from companies except in their accounting books, period. It doesn't require them to do anything at all in their interactions with customers, it just requires them to record the interactions, in their books, in certain specific ways. It is legal under SOx to give stuff away, it is legal to charge for it, it is legal for Apple to pay people to upgrade to 802.11n, it is legal for Apple to give people blowjobs to upgrade. (Well, under SOx, that is. Prostitution laws probably forbid it.)
The fact that experts disagree on its exact implications means that there is a huge nebulous grey area where it is quite conceivable that their counsel advised them to avoid stepping in because, whether it was clearly required or not, it created substantial risk of liability which could be avoided. Particularly when regulators are presently taking a microscope to Apple over compliance with accounting rules, an excess of caution isn't really surprising at all.
The law is only 'unclear' because of loopholes that may or may not exist, and haven't been tried in court.
But I have to point out this is accounting. It's not taxes, it's intended to inform people of the state of publically-traded companies. Any company attempting to find damn loopholes in that needs to stop. Just tell people the actual truth about the state of assets and liabilities and quit trying to fuck with the numbers all the time.
The accounting practices that SOX made illegal are the ones that corporate executives use to defraud investors and stockholders out of billions. It's how to hide the fact a ship is sinking until you can escape.
All this whining about SOX, and possible even what Apple's doing, is an attempt to get people upset at a law requiring companies to be honest and not list possible income from the next quarter as actual income this quarter.
It doesn't require anyone to pay for anything. If Apple had promised an upgrade to 802.11n for 5 dollars, or promised a free upgrade later, it would have required them to not actually count money from that until, you know, it happened. It certainly doesn't have anything to do with what's going on now, because 802.11n functionality was not one of the reason that people purchased the machine.
If they promised a future feature then it depends on the wording. ie was it promised as a defered feature or as an optional future upgrade. If you never promise an upgrade, and the product - as sold - was "feature complete" with no promises of future support etc, then the act probably does not come into play.
They didn't promise it at all. No one knew it could do this, or, even if they had figured it out, it was an undocumented feature that would require hacked firmware.
Apple is fucking with people. The law doesn't make them charge anything for things they do for free. It makes them account for things they promise to do when they actually do them, not before.
As Apple didn't even vaguely promise to do this, they don't have to account for anything. Even if they had promised it, charging now isn't going to fix anything, because they screwed up when they promised it and counted it as real in the previous quarters, so they're already in the wrong. In fact, if they had promised to do it, charging for the upgrade would, quite obviously, be fraud.
This claim of theirs doesn't even make any logical sense.
The product was complete. It was actually manufactured, it was sold in stores, people used it for the claimed purpose, everyone was happy, no one even knew it could do 802.11n.
Apple's fucking with people here. There's no interpetation of the law that says allowing customers to add features to a product after the sale is not legal.
Deferring the income would be selling it as a 802.11n device that will be turned on later.
Magically converting a device that no one knew was 802.11n to 802.11n is not 'deferring' anyway, anymore than cars have 'deferred' upgrades when the car dealership randomly gives them a free cup holder at their 30,000 mile oil change.
The law is designed to stop companies from selling things that don't exist yet, and accounting for them now, before they've actually made them. I.e, selling an empty lot and a contract to build a building on it counts as an empty lot now and a building when it's built, you can't count the building now.
I don't know if that example is strictly true, but that's the theory going on, because companies would use silliness like this to disclose things whenever they were convenient, and sometimes they'd even plan to back out of said contracts in the first place, and do them entirely to make their balance sheet look good.
I don't know what kind of crack Apple is on, or if this is just a scam, but no law requires anyone to charge for free, unadvertised upgrades, whether on hardware or software.
Now, what this could have done is magically change the 'value' of their existing inventory, but paradoxically, only because they're charging for a feature. I.e, the value of their inventory that can upgrade to 802.11n are original value + (estimated percentage of people who will upgrade * $4.99). That might screw up their accounting, but they did that to themselves.
You need to know how to how to signal a turn, how to parallel park, what traffic signs mean. And you have to prove to the state that you know it and can operate a car competently before you're allowed on the roads.
Really? Where do you live? Here in the US, we don't have anything like that, although we do make you get 'car insurance'. And legally you have to be able to see or be over 65.;)
And you've really it on the crux of the issue. Maintenance is IT's job. If they don't ever change the oil in their computer, or whatever, well, IT should be doing that, or even better, have written tools that do it automatically.
Operating the computer is their job. That, like a car, includes trivial amounts of maintenance, like cleaning junk out of the inside of the car, using windshield wipers when there's junk on the window, putting in gasoline, at least realizing when you have a flat or other problem and calling someone, etc.
However, too often people have such limited computer knowledge they drive around with a flat for two years, or don't know how to use the windshield wipers and crash repeatedly into things, or have a car that contains solid junk-food wrappers everywhere but where they are physically sitting.
No, all doors shouldn't, and that's a semi-amazing example I picked randomly.;)
Elevator doors and other sliding doors, for a really obvious example, don't open outward. Roll up doors either.
Maintenance doors to operate on the elevator and other equipment closets can open inward. Loading bay doors that don't roll up almost always often open inward. Non-sliding doors on balconies could have both door that opens inward and a screen door that opens outward.
You're thinking 'Everyone knows that', or 'But those are exceptions' but that's really my damn point. Management said 'Follow the fire code', plant did so, and made everything else operate logically.
If management treated plant like IT, it'd be 'make all doors open outward', and plant would be in the fun position of trying to figure out how to make two eight-feet-high four-feet-wide doors that swing inward for loading swing outward instead, presumably straight through the back of the truck.
Of course, if plant was in that position, they could actually explain why that was physically illogical and management might listen, as opposed to IT trying to explain why writing a mission-critical app that has to run on remote computers in VB and MS Access is not really a great idea.
Management should tell IT what they want to happen, which is policy, and within some logical constraints like a budget and calendar, not how to make it happen, which they do way too often.
It's basically the same thing as the Flash debate, or the embedded Java debate. Or Javascript. There are idiots out there who misuse it, making menus and stupid hover buttons out of them, so there are reactive idiots who fight for some hypothetical 'purity', despite HTML email, Flash embedding, Java embedding, and Javascript being well-defined standards.
From those people, I'd like an explanation of how yootube would have worked without Flash. Or how Google Maps would work without Javascript.
I hate stupid uses of heavy and pointless extras when the basics would have worked, too. However, I, and I suspect you, understand the important of the word 'pointless' in there.
Ah, okay, I just picked a sample one I had somewhere because I, like most human beings, just pick whatever DOCTYPE our HTML editor gives us for the version of HTML we choose.;)
And, yes, the entire thing is a mess. The point of paths was a noble one, but the end result is that absolutely nothing uses them except validators, and validators should be able to go and look up 'The w3c's standard for HTML 4.01' without have an explicit path.
It's not so much as not knowing things. After all, you can operate a car quite well in today's society without knowing anything about cars, as long as you understand the 'gasoline' thing. That's what oil change places are for. You'll end up paying more, but whatever.
IT professions, if computers were cars, would have to keep patiently explaining about the gasoline concept to some users, and have to stop other users who keep pouring maple syrup in their gas tanks, while others insisted on driving around with the hood open so the engine would stay cool.
It's not helped that IT is often micromanaged by people who know nothing about IT. You don't see that in other departments, upper management doesn't tell marketing that all advertisements will be printed on off-white paper, or tell plant that all doors should open in a certain direction.
Yet upper management sees nothing wrong with dictating exactly what tools can be used in what circumstances. That, for example, car windshields should be cleaned right-to-left.
Management often has no idea of the difference, in IT, between 'policy' decisions, which they certainly can, and must make, and 'how to implement policy' decisions, which they really shouldn't. Just ask all the people who are secretly using SAMBA because policy, instead of saying 'We must use integrated Windows file sharing because we have random people come in and hook to our network', dictated, instead, 'We will use Windows file servers' and gave IT crappy computers to implement it with, which they secretly put Linux on.
IT would be a lot less annoying if they didn't have to put up with management who didn't know the difference between decisions they must make, decisions they can make, and decisions they shouldn't make. They don't know what they know, and they don't know what they don't know.
And the same applies to users, who often don't know exactly how competent or incompetent they are. It's almost as annoying to have to walk someone through some simple thing because they're scared they might break something as it is to fix people who actually do break things. I've heard of users who were afraid to navigate through Windows Explorer to find something on the local network. They were certainly capable of doing it, but were deathly scared they might break something.
Amen. I'd rather go to House 10 times, and have him mock and berate me nine of them and find some obscure thing wrong with me the tenth, than got to a normal doctor and have him not pay attention.
OTOH, doctors like House probably don't actually exist.
...who the hell blames 'overambitious deadlines, changing requirements, and design compromises' on users? Everyone I've ever met blames them, quite rightly, on management. Or in companies developing applications to sale, on marketing.
I can just see it now:
'Bob, we've got to ship Thursday.'
'What? We haven't tracked down that crash-during-export bug! Damn users!'
*blank stare*
'Um, Bob? What users? No one's using the program yet, it hasn't shipped.'
'Oh, right. Damn marketing for promising random ship-dates without consulting with us!'
It's not any different, except the w3c is run by intelligent people, and Netscape, apparently, is not.
I've always thought the full paths were a bit stupid too, and they should have some sort of shortcut standard, one that says "Use w3c's HTML4.0 standard", and the web browser knows how to contruct a path to find w3c standards. That way, when "Use netscape's RSS0.91" standard stopped working, web browsers could have a trivial update, or their config could even be changed manually, to tell them where to find netscape's standards.
Granted, they already have something like this in the DOCTYPE, that's what '-//W3C//DTD HTML 4.01//EN' is, but then they blow it by then including the path after that. The parser should, instead, have to look at W3C and go 'Hey, I know where that is, that's w3c.org' and construct a standardized path using 'DTD HTML 4.01', like 'http://w3c.org/doctype/HTML4.01.dtd'. (And I just realized that string mysteriously doesn't include 'strict' or whatever in it, so now I'm slightly confused as to what good it's for.)
That way, when something happened to a server, the standard can be trivally updated to say 'W3C now means this domain, instead of w3c.org', and every damn page in existence doesn't have to change. Mandate that every parser should expose these locations to be reconfigured manually if needed, although obviously some sort of automatic updating is a good idea. (Notice, in general, application software doesn't need to be updated, because application software doesn't try to download the stuff in the first place.)
Now someone's going to host the DTD at some random place, and everyone will manually update everything to load the wrong URL when someone asks for "http://netscape.com/publish/formats/rss-0.91.dtd" then Netscape will move it back, and some applications will change back, and some won't, and it will be a big mess.
I understand the point of paths, in that, in theory, everyone can produce their own format and publish their own DTD. This has not, and probably is not, going to happen, and at this point all browsers interpet DOCTYPE strings as unparsable strings, and the only ones who actually read the things are the validators.
You want to keep engines away from content, keep them away from the whole page, which is doable in two different way using the robot exclusion standards.
The idea that search engines would support a tag to exclude parts of pages as 'non-official content', and yet index other parts of the same page, is a bit silly. Why would they want to do that? When they send users to pages, they send them to whole pages, so they want to know about the whole page. If a ton of spam is screwing up Google's relevancy for a page, then, duh, Google's right, the page is full of crap and not as relevant as another page that isn't full of crap.
The person I originally replied to wants to have his cake and eat it to, where he can tell Google and other search engines to pay attention to certain parts of a page, and not other parts. If Google was willing to do that, why would they spend so much time fighting 'cloaking', where different pages are sent to search engines vs. users?
Um, what, exactly, are you trying to accomplish? We already have rel="nofollow".
Now, if you're arguing that there should be away to make sections of code as user contributed, and, thus, not to be indexed, that a) doesn't make any sense, and b) is already possible.
It doesn't make any sense because having random areas of gibberish or non-related content doesn't actually hurt you in search engines, except to the extent people will stop linking to you.
And it's already possible via trivial javascript to hide that part of the page from search engines, if you actually had a reason to do that.
I'm not sure the idea makes a lot of sense, anyway. Google doesn't want to link to pages that are full of spam, so coming up with some magical tag that says 'This could be crap' isn't really going to work. They want to link to useful things, regardless of who created the content.
Once upon a time, I was in college. We had rooms of identical computers in the labs, and two different types of printers. We also had the library, with computers and one of those printer types.
What did this translate to? If you did the work the library, and printed it in some of the labs, your formatting would be off. In others you'd have no problem.
In those computer labs, during classes that had and things to print and turn in, there'd always be someone who walked in with the document to print and spent a hurried five minutes fixing their paragraphs not to changes pages in the middle, because they did it in the library. And in the other labs, about half the time you'd run into the same problem, because they did it in the other lab.
Some of the students figured out you could switch your printer to another room's printer and print them correctly...assuming you could find an empty room to print in, as teachers started getting upset at people coming into their class to collect printouts. There usually was one, the school tried to keep at least one lab open, but it was a crapshot if they had the right kind of printer.
Then people started forgetting to change printers back, so people who'd figured out what was going on would prepare a document in a single room, and print it there, only to discover their printer was set elsewhere, and then switch it back, only to discover their formatting was off. It didn't help that this process would sometimes be interrupted by a random angry teacher, who, pissed that people were printing in their classroom, had snatched their misprint off the printer and tracked down the person who'd printed it. (The name and class were usually on there.)
It was total and complete chaos the entire time I went there, because of dumbass Microsoft and their brilliant formatting-changes-with-the-printer idea. The official policy was 'Check the printer every day before you start, and only work on formatting in the lab you're going to print in', but college students and rules do not go together.
Me? I used Open Office on my laptop, 'printed' off a PDF, stuck it online, and printed from within IE, after making sure the printer was right. (IIRC, people usually changed it within Word, so everything else still printed to the right printer.) If I needed to mail a Word document, instead of turn one in, I'd export to Word format, stick it up online, view it within IE to make sure it was okay, and mail it from my laptop.
I did, at one point, manage to get my laptop logging onto their network so I could use the printers, but it was a huge hassle and I soon gave up on that.
The rest of us who have modern secure HTML renderers in our web browsers, and end up reading a dozen bulletins a day, will continue to receive them in HTML.
For example, I read my RSS feeds using rssfwd.com, and they come in as HTML, because they were originally web pages. Likewise, I get various newsletters as HTML, because otherwise I'd just end up clicking a link and reading them in a web browser, and I don't see quite what that would accomplish, especially considering both engines are Mozilla.
I mean, I understand if people want them like that, which is why most places I've seen have the option of either way. But that doesn't mean HTML mail is Wrong, just that it should be an option for automatically mailed things.
What is Wrong are people with stupid mailers that send pointless HTML. Yeah, thanks, I really needed you to specify the font size, it's not like it's my computer or anything.
But, frankly, people who annoy me more are those who insist on responding to the top of all their email, thus rendering discussions almost impossible to follow even if everyone else does that.
Then they'll run the entire OS under a debugger, or under a virtualizer and debugger. Or just decode it off the disk.
Any machine can emulate any other machine, period. It is theoretically impossible to keep decoded data from the hands of someone who has all the secrets. I like how various industries are apparently unaware that computer science is a science and knows that that such concepts are flatly impossible. As long as you can physically read all the data used to decode something, and the thing you want to decode, you can crack it.
Protecting memory won't cut it, the data's on the disk. Protecting the files on the disk won't cut it, because they can be read in Linux. Encrypting the files won't cut it, because the Windows decryption key can be read in Linux. And so on and so on. It's a theoretical impossiblity, which seems to have escaped notice of the people trying to implement it, who, I guess, just keep thinking they're implementing crappy systems.
It's akin to designing a perpetual motion machine...it always looks like there's some fatal flaw you can fix to get it to work, but, sadly, no, there isn't. It cannot work in theory, and thus will always fail in practice.
Now, it might be possible to have complete software protection if some of the secrets are locked into silicon. With such hardware protection, people might be required to do hardware hacking to crack it, although I'm not entirely sure of that. It always can be done with hardware hacking, even if the hardware hack required is to disassemble a decrypting plasma display and pipe the decoded output to another computer.
All of your conspiracies revolve around much larger elements than preventing someone from speaking. Do you even realize the culture we exist in? It's so much easier to discredit than silence.
You both discredit them and you stop their work from be distributed in any easy manner. Um, duh.
As for the CIA, I love how your explanations escalate to cover the large leaps in logic. Now the CIA has to be involved to forge the document. If you thought about what you were saying you would realize there are easier methods to accomplish the same thing.
At no point did I even vaguely hint the CIA would be involved. I said the CIA has illegally murdered people, so saying the government would have to commit the horrible crime of forgery is not really a stretch. If it is too much of a stretch for you, I offered the alternative of burying them in legal fees for years until they are forced to sell the copyright.
Your assertions are ridiculous, and you make assumptions that are a little to far on the side of paranoid for me to take seriously.
And, frankly, you're too dumb for me to take seriously, because you sit and claim that what I describe is impossible, when the entire point of the article is 'Assume a universe with perfect DRM. Is that okay?' and I pointed out merely why it wouldn't be okay. Namely, because perfect DRM must require checking it with the government, and allowing the government to physically disallow viewing and copying of all information is a very very bad idea.
It's like someone saying 'Would everyone having access to easy-to-operate personal jetpacks be a good idea or a bad idea?', and I point out they'd rewrite various assumptions on privacy and trespassing, and you point out that that such jetpacks are unlikely. It's like you have no concept of a hypothetical question of the form 'Obviously, X is currently implemented horrible, but if there was nothing wrong with the implementation, would it be a good idea?'.
Rather obviously we won't have the DRM that I, or everyone else in the damn discussion except you, are talking about at any point in time. The question as asked is intended to spur the level of discussion past 'DRM doesn't let me do what I legally am allowed to' to 'DRM is inherently a bad idea even if it worked'.
Although I suspect a hypothetical question that would have been more useful would be 'Is it theoretically possible to write DRM that is perfectly in accordance with the law?', because, as I pointed out somewhere else in this discussion, that would require mind-reading abilities.
Next on Mythbusters: Is it possible to hand someone content on a computer-readable medium and the software tools to view it, without them being able to copy it?
If hackers wish, they can also next time be more secretive about which player they're using to get the volume keys for decryption with, internally to their warez group, to significantly slow down software player bans.
There's the failure of the system, right there. At some point in time, if the HD-DVD actually start revoking all found keys, they are going to find a key and not reveal they've cracked it.
They have, after all, experts at reverse engineering thanks to their copy-protection cracking. Tracking down a key is nothing. Once the studios start revoking all the keys they find, they simply won't tell anyone the keys outside of trusted people.
At that point, their methods will have to change a little, from their current 'Rip, encode, and upload to our distribution hubs' to 'rip a copy of all the disks you have, but instead of trying to encode them, send them to us and we'll decrypt them first and then encode them for distribution.' policy.
Anyone who thinks that's a lot of bandwidth have no idea how much bandwidth and storage space cracking groups have at their disposal.
OTOH, possibly people who have access to their distribution hubs are already trusted enough to hold the key, especially, if they can crack a new key every month or so and so have a fairly high level of acceptable losses. Keep the key as 'need to know', to people who are actually doing the ripping so they'd be unlikely to incriminate themselves by telling anyone, but when it gets out, don't worry about it...release it to the world and hand all the trusted people a new key.
Slashdot Mirror
Yes, exactly.
It's amazing how many people want to nitpick about various flaws in computers.
There is exactly one flaw that renders the entire thing moot: Computers do exactly what you tell them to do.
This not only includes lying, it includes lying about lying, and lying about the lying it's lying about, and lying about that lying, and then lying about that lie. Under whatever circumstances it is told to lie under.
It is theoretically impossible to confirm that a voting computer operates the way it is required by law to operate. Not just practically impossible, theoretically impossible.
Yes, they test them, but it is impossible to know that it hasn't been programmed to give the wrong result only when X happens, where X could be 'Used continually for three hours' or 'When a specific ballot is entered'.
They can look at the code, but they have no way to know that the code the computer presents them as the code is, in fact, the code the computer is actually running. It's actually impossible to do this without a detailed inspection of every byte of code in the computer by removing all all storage devices and looking at them in a trusted computer, which not only do they not do, but couldn't do, as parts of the executing code are in the BIOS and other firmware in computer.
Considering that there's several billion bytes of code in Windows alone, which is the base system they use, and that the amount of work required is exponential based on the file size, in it is not possible to even attempt this for any reasonable price.
Just look at the amount of time and money spent developing air-traffic control systems if you don't believe me, and then realize a) voting machines are running off-the-shelf OSes instead of, basically, no OS in ATC systems, b) they don't have to catch deliberate 'mistakes' in ATC systems to alter the vote, c) ATC systems are monitored while used by trained professionals, unlike voting machines, which are operated, duh, in secret by random people, d) have to have physically secure boxes, unlike ATC systems which are trivially to 'hack' if you can physically reach them, e) voting machines either have to be portable or we have to build billions of dollars of buildings to contain them permanently, with staff.
A real electronic voting machine system would cost more than ten ATC systems, and ATC systems can run to hundreds of millions of dollars. We're literally talking about a billion dollar development effort here, maybe one hundred million dollars worth of machines, and hundreds of millions of dollars a year in maintenance to do correctly and securely. And that's just if we do it once for the whole country.
Or we can do it shittily and have companies just slapping things together and misleading and bribing (with votes, maybe?) the legislature to use them. Seems to be working so far, if by 'working' you mean 'not working'.
Of course, there's the other option that computer scientists have been screaming about since the start: Have the computer print a OCRable ballot, and have computers and humans count the ballot later. Which would cost a good deal less than the machines they're building now, it would basically be a cash register.
"oh by the way, the last 800,000 computers we shipped had different hardware than reported in our disclosures."
Which was my damn point. Either Apple already lied about the hardware, and thus, are already in trouble, or they didn't lie and are fine. If you give someone something for free, it's not revenue.
And they don't have to 'restate' their previous hardware as n-compatible just because they, or someone else, created some software to make it so. It wasn't n-compatible, and, as far as Apple cares, it's still not.
The idea that it's illegal to give away something that enables a previously announced feature on hardware is completely batshit insane, and Apple is consequently a) batshit insane itself, b) lying to make money, c) lying to look like it's complying with every single law, even ones that don't actually exist, or d) lying to make SOx look like a bad law, which companies have been doing ever since it was passed.
Or some combination thereof. Of course, all this is from some guy's blog, so it's probably crap to start with.
To repeat: A lot of companies have taken issue with SOx restricting their unethical activities, and as such have started blaming all sorts of crap on it. SOx does not require any behavior from companies except in their accounting books, period. It doesn't require them to do anything at all in their interactions with customers, it just requires them to record the interactions, in their books, in certain specific ways. It is legal under SOx to give stuff away, it is legal to charge for it, it is legal for Apple to pay people to upgrade to 802.11n, it is legal for Apple to give people blowjobs to upgrade. (Well, under SOx, that is. Prostitution laws probably forbid it.)
The fact that experts disagree on its exact implications means that there is a huge nebulous grey area where it is quite conceivable that their counsel advised them to avoid stepping in because, whether it was clearly required or not, it created substantial risk of liability which could be avoided. Particularly when regulators are presently taking a microscope to Apple over compliance with accounting rules, an excess of caution isn't really surprising at all.
The law is only 'unclear' because of loopholes that may or may not exist, and haven't been tried in court.
But I have to point out this is accounting. It's not taxes, it's intended to inform people of the state of publically-traded companies. Any company attempting to find damn loopholes in that needs to stop. Just tell people the actual truth about the state of assets and liabilities and quit trying to fuck with the numbers all the time.
The accounting practices that SOX made illegal are the ones that corporate executives use to defraud investors and stockholders out of billions. It's how to hide the fact a ship is sinking until you can escape.
All this whining about SOX, and possible even what Apple's doing, is an attempt to get people upset at a law requiring companies to be honest and not list possible income from the next quarter as actual income this quarter.
It doesn't require anyone to pay for anything. If Apple had promised an upgrade to 802.11n for 5 dollars, or promised a free upgrade later, it would have required them to not actually count money from that until, you know, it happened. It certainly doesn't have anything to do with what's going on now, because 802.11n functionality was not one of the reason that people purchased the machine.
If they promised a future feature then it depends on the wording. ie was it promised as a defered feature or as an optional future upgrade. If you never promise an upgrade, and the product - as sold - was "feature complete" with no promises of future support etc, then the act probably does not come into play.
They didn't promise it at all. No one knew it could do this, or, even if they had figured it out, it was an undocumented feature that would require hacked firmware.
Apple is fucking with people. The law doesn't make them charge anything for things they do for free. It makes them account for things they promise to do when they actually do them, not before.
As Apple didn't even vaguely promise to do this, they don't have to account for anything. Even if they had promised it, charging now isn't going to fix anything, because they screwed up when they promised it and counted it as real in the previous quarters, so they're already in the wrong. In fact, if they had promised to do it, charging for the upgrade would, quite obviously, be fraud.
This claim of theirs doesn't even make any logical sense.
The product was complete. It was actually manufactured, it was sold in stores, people used it for the claimed purpose, everyone was happy, no one even knew it could do 802.11n.
Apple's fucking with people here. There's no interpetation of the law that says allowing customers to add features to a product after the sale is not legal.
That makes no sense at all.
Deferring the income would be selling it as a 802.11n device that will be turned on later.
Magically converting a device that no one knew was 802.11n to 802.11n is not 'deferring' anyway, anymore than cars have 'deferred' upgrades when the car dealership randomly gives them a free cup holder at their 30,000 mile oil change.
The law is designed to stop companies from selling things that don't exist yet, and accounting for them now, before they've actually made them. I.e, selling an empty lot and a contract to build a building on it counts as an empty lot now and a building when it's built, you can't count the building now.
I don't know if that example is strictly true, but that's the theory going on, because companies would use silliness like this to disclose things whenever they were convenient, and sometimes they'd even plan to back out of said contracts in the first place, and do them entirely to make their balance sheet look good.
I don't know what kind of crack Apple is on, or if this is just a scam, but no law requires anyone to charge for free, unadvertised upgrades, whether on hardware or software.
Now, what this could have done is magically change the 'value' of their existing inventory, but paradoxically, only because they're charging for a feature. I.e, the value of their inventory that can upgrade to 802.11n are original value + (estimated percentage of people who will upgrade * $4.99). That might screw up their accounting, but they did that to themselves.
You need to know how to how to signal a turn, how to parallel park, what traffic signs mean. And you have to prove to the state that you know it and can operate a car competently before you're allowed on the roads.
Really? Where do you live? Here in the US, we don't have anything like that, although we do make you get 'car insurance'. And legally you have to be able to see or be over 65. ;)
And you've really it on the crux of the issue. Maintenance is IT's job. If they don't ever change the oil in their computer, or whatever, well, IT should be doing that, or even better, have written tools that do it automatically.
Operating the computer is their job. That, like a car, includes trivial amounts of maintenance, like cleaning junk out of the inside of the car, using windshield wipers when there's junk on the window, putting in gasoline, at least realizing when you have a flat or other problem and calling someone, etc.
However, too often people have such limited computer knowledge they drive around with a flat for two years, or don't know how to use the windshield wipers and crash repeatedly into things, or have a car that contains solid junk-food wrappers everywhere but where they are physically sitting.
No, all doors shouldn't, and that's a semi-amazing example I picked randomly. ;)
Elevator doors and other sliding doors, for a really obvious example, don't open outward. Roll up doors either.
Maintenance doors to operate on the elevator and other equipment closets can open inward. Loading bay doors that don't roll up almost always often open inward. Non-sliding doors on balconies could have both door that opens inward and a screen door that opens outward.
You're thinking 'Everyone knows that', or 'But those are exceptions' but that's really my damn point. Management said 'Follow the fire code', plant did so, and made everything else operate logically.
If management treated plant like IT, it'd be 'make all doors open outward', and plant would be in the fun position of trying to figure out how to make two eight-feet-high four-feet-wide doors that swing inward for loading swing outward instead, presumably straight through the back of the truck.
Of course, if plant was in that position, they could actually explain why that was physically illogical and management might listen, as opposed to IT trying to explain why writing a mission-critical app that has to run on remote computers in VB and MS Access is not really a great idea.
Management should tell IT what they want to happen, which is policy, and within some logical constraints like a budget and calendar, not how to make it happen, which they do way too often.
It's basically the same thing as the Flash debate, or the embedded Java debate. Or Javascript. There are idiots out there who misuse it, making menus and stupid hover buttons out of them, so there are reactive idiots who fight for some hypothetical 'purity', despite HTML email, Flash embedding, Java embedding, and Javascript being well-defined standards.
From those people, I'd like an explanation of how yootube would have worked without Flash. Or how Google Maps would work without Javascript.
I hate stupid uses of heavy and pointless extras when the basics would have worked, too. However, I, and I suspect you, understand the important of the word 'pointless' in there.
Ah, okay, I just picked a sample one I had somewhere because I, like most human beings, just pick whatever DOCTYPE our HTML editor gives us for the version of HTML we choose. ;)
And, yes, the entire thing is a mess. The point of paths was a noble one, but the end result is that absolutely nothing uses them except validators, and validators should be able to go and look up 'The w3c's standard for HTML 4.01' without have an explicit path.
If the w3c collapses web standards are screwed anyway.
It's not so much as not knowing things. After all, you can operate a car quite well in today's society without knowing anything about cars, as long as you understand the 'gasoline' thing. That's what oil change places are for. You'll end up paying more, but whatever.
IT professions, if computers were cars, would have to keep patiently explaining about the gasoline concept to some users, and have to stop other users who keep pouring maple syrup in their gas tanks, while others insisted on driving around with the hood open so the engine would stay cool.
It's not helped that IT is often micromanaged by people who know nothing about IT. You don't see that in other departments, upper management doesn't tell marketing that all advertisements will be printed on off-white paper, or tell plant that all doors should open in a certain direction.
Yet upper management sees nothing wrong with dictating exactly what tools can be used in what circumstances. That, for example, car windshields should be cleaned right-to-left.
Management often has no idea of the difference, in IT, between 'policy' decisions, which they certainly can, and must make, and 'how to implement policy' decisions, which they really shouldn't. Just ask all the people who are secretly using SAMBA because policy, instead of saying 'We must use integrated Windows file sharing because we have random people come in and hook to our network', dictated, instead, 'We will use Windows file servers' and gave IT crappy computers to implement it with, which they secretly put Linux on.
IT would be a lot less annoying if they didn't have to put up with management who didn't know the difference between decisions they must make, decisions they can make, and decisions they shouldn't make. They don't know what they know, and they don't know what they don't know.
And the same applies to users, who often don't know exactly how competent or incompetent they are. It's almost as annoying to have to walk someone through some simple thing because they're scared they might break something as it is to fix people who actually do break things. I've heard of users who were afraid to navigate through Windows Explorer to find something on the local network. They were certainly capable of doing it, but were deathly scared they might break something.
Amen. I'd rather go to House 10 times, and have him mock and berate me nine of them and find some obscure thing wrong with me the tenth, than got to a normal doctor and have him not pay attention.
OTOH, doctors like House probably don't actually exist.
...who the hell blames 'overambitious deadlines, changing requirements, and design compromises' on users? Everyone I've ever met blames them, quite rightly, on management. Or in companies developing applications to sale, on marketing.
I can just see it now:
'Bob, we've got to ship Thursday.'
'What? We haven't tracked down that crash-during-export bug! Damn users!'
*blank stare*
'Um, Bob? What users? No one's using the program yet, it hasn't shipped.'
'Oh, right. Damn marketing for promising random ship-dates without consulting with us!'
It's not any different, except the w3c is run by intelligent people, and Netscape, apparently, is not.
I've always thought the full paths were a bit stupid too, and they should have some sort of shortcut standard, one that says "Use w3c's HTML4.0 standard", and the web browser knows how to contruct a path to find w3c standards. That way, when "Use netscape's RSS0.91" standard stopped working, web browsers could have a trivial update, or their config could even be changed manually, to tell them where to find netscape's standards.
Granted, they already have something like this in the DOCTYPE, that's what '-//W3C//DTD HTML 4.01//EN' is, but then they blow it by then including the path after that. The parser should, instead, have to look at W3C and go 'Hey, I know where that is, that's w3c.org' and construct a standardized path using 'DTD HTML 4.01', like 'http://w3c.org/doctype/HTML4.01.dtd'. (And I just realized that string mysteriously doesn't include 'strict' or whatever in it, so now I'm slightly confused as to what good it's for.)
That way, when something happened to a server, the standard can be trivally updated to say 'W3C now means this domain, instead of w3c.org', and every damn page in existence doesn't have to change. Mandate that every parser should expose these locations to be reconfigured manually if needed, although obviously some sort of automatic updating is a good idea. (Notice, in general, application software doesn't need to be updated, because application software doesn't try to download the stuff in the first place.)
Now someone's going to host the DTD at some random place, and everyone will manually update everything to load the wrong URL when someone asks for "http://netscape.com/publish/formats/rss-0.91.dtd" then Netscape will move it back, and some applications will change back, and some won't, and it will be a big mess.
I understand the point of paths, in that, in theory, everyone can produce their own format and publish their own DTD. This has not, and probably is not, going to happen, and at this point all browsers interpet DOCTYPE strings as unparsable strings, and the only ones who actually read the things are the validators.
Well, yeah, the whole thing is a bit silly.
You want to keep engines away from content, keep them away from the whole page, which is doable in two different way using the robot exclusion standards.
The idea that search engines would support a tag to exclude parts of pages as 'non-official content', and yet index other parts of the same page, is a bit silly. Why would they want to do that? When they send users to pages, they send them to whole pages, so they want to know about the whole page. If a ton of spam is screwing up Google's relevancy for a page, then, duh, Google's right, the page is full of crap and not as relevant as another page that isn't full of crap.
The person I originally replied to wants to have his cake and eat it to, where he can tell Google and other search engines to pay attention to certain parts of a page, and not other parts. If Google was willing to do that, why would they spend so much time fighting 'cloaking', where different pages are sent to search engines vs. users?
Um, what, exactly, are you trying to accomplish? We already have rel="nofollow".
Now, if you're arguing that there should be away to make sections of code as user contributed, and, thus, not to be indexed, that a) doesn't make any sense, and b) is already possible.
It doesn't make any sense because having random areas of gibberish or non-related content doesn't actually hurt you in search engines, except to the extent people will stop linking to you.
And it's already possible via trivial javascript to hide that part of the page from search engines, if you actually had a reason to do that.
I'm not sure the idea makes a lot of sense, anyway. Google doesn't want to link to pages that are full of spam, so coming up with some magical tag that says 'This could be crap' isn't really going to work. They want to link to useful things, regardless of who created the content.
The greatest gotcha ever.
Once upon a time, I was in college. We had rooms of identical computers in the labs, and two different types of printers. We also had the library, with computers and one of those printer types.
What did this translate to? If you did the work the library, and printed it in some of the labs, your formatting would be off. In others you'd have no problem.
In those computer labs, during classes that had and things to print and turn in, there'd always be someone who walked in with the document to print and spent a hurried five minutes fixing their paragraphs not to changes pages in the middle, because they did it in the library. And in the other labs, about half the time you'd run into the same problem, because they did it in the other lab.
Some of the students figured out you could switch your printer to another room's printer and print them correctly...assuming you could find an empty room to print in, as teachers started getting upset at people coming into their class to collect printouts. There usually was one, the school tried to keep at least one lab open, but it was a crapshot if they had the right kind of printer.
Then people started forgetting to change printers back, so people who'd figured out what was going on would prepare a document in a single room, and print it there, only to discover their printer was set elsewhere, and then switch it back, only to discover their formatting was off. It didn't help that this process would sometimes be interrupted by a random angry teacher, who, pissed that people were printing in their classroom, had snatched their misprint off the printer and tracked down the person who'd printed it. (The name and class were usually on there.)
It was total and complete chaos the entire time I went there, because of dumbass Microsoft and their brilliant formatting-changes-with-the-printer idea. The official policy was 'Check the printer every day before you start, and only work on formatting in the lab you're going to print in', but college students and rules do not go together.
Me? I used Open Office on my laptop, 'printed' off a PDF, stuck it online, and printed from within IE, after making sure the printer was right. (IIRC, people usually changed it within Word, so everything else still printed to the right printer.) If I needed to mail a Word document, instead of turn one in, I'd export to Word format, stick it up online, view it within IE to make sure it was okay, and mail it from my laptop.
I did, at one point, manage to get my laptop logging onto their network so I could use the printers, but it was a huge hassle and I soon gave up on that.
Yes, they can send you that.
The rest of us who have modern secure HTML renderers in our web browsers, and end up reading a dozen bulletins a day, will continue to receive them in HTML.
They aren't trite when it's useful stuff.
For example, I read my RSS feeds using rssfwd.com, and they come in as HTML, because they were originally web pages. Likewise, I get various newsletters as HTML, because otherwise I'd just end up clicking a link and reading them in a web browser, and I don't see quite what that would accomplish, especially considering both engines are Mozilla.
I mean, I understand if people want them like that, which is why most places I've seen have the option of either way. But that doesn't mean HTML mail is Wrong, just that it should be an option for automatically mailed things.
What is Wrong are people with stupid mailers that send pointless HTML. Yeah, thanks, I really needed you to specify the font size, it's not like it's my computer or anything.
But, frankly, people who annoy me more are those who insist on responding to the top of all their email, thus rendering discussions almost impossible to follow even if everyone else does that.
Then they'll run the entire OS under a debugger, or under a virtualizer and debugger. Or just decode it off the disk.
Any machine can emulate any other machine, period. It is theoretically impossible to keep decoded data from the hands of someone who has all the secrets. I like how various industries are apparently unaware that computer science is a science and knows that that such concepts are flatly impossible. As long as you can physically read all the data used to decode something, and the thing you want to decode, you can crack it.
Protecting memory won't cut it, the data's on the disk. Protecting the files on the disk won't cut it, because they can be read in Linux. Encrypting the files won't cut it, because the Windows decryption key can be read in Linux. And so on and so on. It's a theoretical impossiblity, which seems to have escaped notice of the people trying to implement it, who, I guess, just keep thinking they're implementing crappy systems.
It's akin to designing a perpetual motion machine...it always looks like there's some fatal flaw you can fix to get it to work, but, sadly, no, there isn't. It cannot work in theory, and thus will always fail in practice.
Now, it might be possible to have complete software protection if some of the secrets are locked into silicon. With such hardware protection, people might be required to do hardware hacking to crack it, although I'm not entirely sure of that. It always can be done with hardware hacking, even if the hardware hack required is to disassemble a decrypting plasma display and pipe the decoded output to another computer.
All of your conspiracies revolve around much larger elements than preventing someone from speaking. Do you even realize the culture we exist in? It's so much easier to discredit than silence.
You both discredit them and you stop their work from be distributed in any easy manner. Um, duh.
As for the CIA, I love how your explanations escalate to cover the large leaps in logic. Now the CIA has to be involved to forge the document. If you thought about what you were saying you would realize there are easier methods to accomplish the same thing.
At no point did I even vaguely hint the CIA would be involved. I said the CIA has illegally murdered people, so saying the government would have to commit the horrible crime of forgery is not really a stretch. If it is too much of a stretch for you, I offered the alternative of burying them in legal fees for years until they are forced to sell the copyright.
Your assertions are ridiculous, and you make assumptions that are a little to far on the side of paranoid for me to take seriously.
And, frankly, you're too dumb for me to take seriously, because you sit and claim that what I describe is impossible, when the entire point of the article is 'Assume a universe with perfect DRM. Is that okay?' and I pointed out merely why it wouldn't be okay. Namely, because perfect DRM must require checking it with the government, and allowing the government to physically disallow viewing and copying of all information is a very very bad idea.
It's like someone saying 'Would everyone having access to easy-to-operate personal jetpacks be a good idea or a bad idea?', and I point out they'd rewrite various assumptions on privacy and trespassing, and you point out that that such jetpacks are unlikely. It's like you have no concept of a hypothetical question of the form 'Obviously, X is currently implemented horrible, but if there was nothing wrong with the implementation, would it be a good idea?'.
Rather obviously we won't have the DRM that I, or everyone else in the damn discussion except you, are talking about at any point in time. The question as asked is intended to spur the level of discussion past 'DRM doesn't let me do what I legally am allowed to' to 'DRM is inherently a bad idea even if it worked'.
Although I suspect a hypothetical question that would have been more useful would be 'Is it theoretically possible to write DRM that is perfectly in accordance with the law?', because, as I pointed out somewhere else in this discussion, that would require mind-reading abilities.
Next on Mythbusters: Is it possible to hand someone content on a computer-readable medium and the software tools to view it, without them being able to copy it?
The conclusion: No way in hell.
If hackers wish, they can also next time be more secretive about which player they're using to get the volume keys for decryption with, internally to their warez group, to significantly slow down software player bans.
There's the failure of the system, right there. At some point in time, if the HD-DVD actually start revoking all found keys, they are going to find a key and not reveal they've cracked it.
They have, after all, experts at reverse engineering thanks to their copy-protection cracking. Tracking down a key is nothing. Once the studios start revoking all the keys they find, they simply won't tell anyone the keys outside of trusted people.
At that point, their methods will have to change a little, from their current 'Rip, encode, and upload to our distribution hubs' to 'rip a copy of all the disks you have, but instead of trying to encode them, send them to us and we'll decrypt them first and then encode them for distribution.' policy.
Anyone who thinks that's a lot of bandwidth have no idea how much bandwidth and storage space cracking groups have at their disposal.
OTOH, possibly people who have access to their distribution hubs are already trusted enough to hold the key, especially, if they can crack a new key every month or so and so have a fairly high level of acceptable losses. Keep the key as 'need to know', to people who are actually doing the ripping so they'd be unlikely to incriminate themselves by telling anyone, but when it gets out, don't worry about it...release it to the world and hand all the trusted people a new key.