Such boasts are obviously proved empty by full read write access from a boot disk
Any non encrypted filesystem would have all access controls subverted when mounted under a foreign operating system. Claiming anything else is absurdity. Do you think that ext2 holds the chmod based security when I mount it up under a box that it wasnt created under? You do realize that any user with root on any Linux box could simply reset all the attributes meant to keep users off/let them in under a different box?
Why would you hold MS to fault for something that is unversial and by its nature fundamental?
Point being though, if you want to discourage users from seeing the source barcode meets the GPL but effectively stops users from seeing/changing the code.
They are getting a lot of heat because they are the big target right now. The Big One. The Video Game industry is bigger than the movies, and soon bigger than TV and even music.
Anything can be 'open-sourced' - it just takes an understanding of how it will be used and the implications of that use; a good developer will design for flexibility, rather than embedding key logic that may need to change inside the program.
In the case of a tax system those configuration files would be tax code. And that tax code is in fact open source.
The bottom line is though, even without the configuration files, you'd be able to engineer enough of how the system works - limits, test cases, even bugs! - that you'd have an unfair advantage. In the OSS world the common goal is *good software*. In the cases I give the goal would be to gain an advantage over the next guy. The motivations are different and the process falls apart.
Opening the source does not NECESSARILY allow people to take advantage of it.
Yes, actually, it does. As long as I know the rules, I can subvert them, test them, and gain unfair advantage. For example, in you crappy search engine, I'd advice my client to create millions and millions and milliosn of pages all linking back to the One True Page. IF you really rated pages randomly my client would get much better results than just having one page.
The point here is that, even if you know the algorithm, there are many factors that you either (1) may not know, or (2) cannot control. A well designed algorithm will be robust in the face of such things.
Except the cases I give your theory falls apart. In the tax scanerio you *know* all the factors - its your data. And you can control all the inputs. Meaning, you can drastically alter results based on the input you provide.
In the real world - the examples I provide - the systems are not and cannot be cryptographically secure, like the systems you pretend to compare them to. These systems are by their nature based on a strict set of rules - laws! - and must conform to them. Random audits are not effective, and have actually been banned by courts in many states. Audits have to based on probable cause - and that probable cause is litigated reguruarly.
The point is in the real world open sourcing many types of systems would lead to the complete ineffectiviness of the system.
You have this false test of when an algorithm is poorly designed or not. And the main thing you claim makes a nice algorithm is one that is based on some type of random seed. In fact in the real world this is rare. Real people will not accept the fact their account is being "randomly audited".
Of course, given that they won the antitrust case by paying off the Federal government after Bush got elected, this shouldn't surprise anyone...
Actually, you should get your facts straight. Bush campaigned repeatedly on the promise of ending the MS anti-trust trial as soon as possible via settlement.
Believe it or not, there are some applications that CANNOT BE EFFECTIVELY OPEN SOURCED.
Open sourcing a search engine would 100% guarantee absolute junk for results.
If I had access to the code behind PageRank, I could guarantee my clients get excellent pacement. Same with other people. Honest people who just put up a website/page would be left in the dust by spammers.
Other examples where obscurity is the ONLY security:
1. Code that states/federal revenue services use to flag accounts for audits. This code, in the public, would instantly destroy the revenue stream of the government. Accountants with programming skills could determine *exactly* what limits they could test and get away with. Every return filed would be for the maximum amount that the code would allow without triggering and audit.
2. Fraud detection code used by credit companies, service providers, etc. Armed with this code crackers would have free reign over credit cards and online payment systems. Exact patterns of usuage could be setup to guarantee that fraud flags would not be triggered.
3. Code that determines which passengers get flagged for pre-flight searches. Armed with this information criminals could fashion profiles that guarantee they will not be probed in-depth.
These four examples destroy your silly notion. Open Source is not a magic pill. A truly open source version of Google would be a useless tool within a matter of weeks, if not days.
I think the strong court we have is overall a VERY good thing.
No, its a VERY bad thing.
It has lead down a path to a point where no one has any idea what any given judge will rule on any given day. Activism has taken things to the point now that there is positively no consistency in law and no coherence at all.
Any law however well intentioned that subverts the liberties of the people of the united states is a bad law and any judge worth his salt would deny the party using this law to subvert said liberties to prevail should be and eventually will be found to be in error by his peers.
That's absurd, and the system you describe is anarchy.
If Congress legally passes a law that is Constitutional and within the scope of their legal right to legislate, and a judge deny's the application of that law because it "subverts the liberties of the people" you have in effect no government at all.
When a law has been duly passed by Congress, and signed by the President, and upheld by the courts as legal and constitutional judges and law enforcement agencies are duty bound to enforce it. Just because it is "Bad" or "subverts" liberties does't meant judges can ignore it.
If you advocate otherwise you advocate for the fall of the American system.
And I, and millions of others, have written letters to congresspersons, only to get predigested letters back stating "everything is ok, please vote for me".
And what did you do with that letter? Follow-up or toss it? Have you actually talked to your Congress person? If you want to influence your rep's opinion, here are some tips: talk to him/her. Your House Rep. Not your senator. Get them on the phone or better yet do it in person when they are in state. It's not that hard. I've spoken on the phone with my rep at least 4 times since his first election. Straightforward really. Call and ask to speak with him/her. Tell them you are a voter, and give them your street address/village/area of town. If they can't talk ask them for a callback time and date when the rep. will be accountable for calling. Get the person on the phone's name. Call back a few hours later and confirm that they have all the details right - phone number, your name, address, etc. Give them the topic vaguely but honestly and without editorial - "I would like to discuss the Congressman's stances on copyrights" is better than "I demand that the Congressperson stop pandering to big business and the hegemony of the elite few".
I could garountee you that if you sat me down infront of the senate and asked me to explain to them why the DMCA is bad I could convince them within an hour as could just about any well educated technically inclined individual could.
No, I don't think so. There are other opinion, and frankly, most people don't give a damn about the things you'll tell them. For a lot of people, the issue is this: should a product designed to be uncopyable be protected by the law as though it was really uncopyable? Anything you can say or I can say isn't going to change that. But the DMCA isn't the issue.
On its face, the DMCA isn't vastly unconstitutional. It may in fact violate the Constitution in my opinion, but based on how courts have ruled before, based on how things have been going in terms of copyright and law in general, it's not a law that is unconstitutional on its face. So then the question is, once an individual judge has ruled the law constitional in his mind, is better that:
The judge disregard his finding, and other courts finding, and rule in favor of the defendant despite the law - effectively saying that the law is legal and valid but just not going to be enforced in this one case
OR
The judge rule in favor of the plantiff and highlight the exact reasons in the law that the defendant is guilty
I prepose the second is better, the "good" ruling. The first ruling gets that one person off, but does nothing to further the cause getting the law I disagree with repealed.
The measure of a goverments success, in any incarnation, is it's ability to solve conflicts between people. A good decision would solve the majority of conflicts, while a bad decision would solve the minorty of them and a really bad decision would cause even more conflicts.
See, I suggest that is stupid, and is actually the reason why things aren't going that well. A court should rule based on a few narrow things: the facts as proven in court, the law as presented by Congress, and case law provided by other Courts. All this other stuff you mention is silly. And what it leads to is people feeling that the law is arbitrary. Theoretically, if you present 100 judges with the same law and 100 identical defendents all charged for comitting the same act that they did in fact commit you should get 100 rulings that are identical. Now, you'd get a 100 different rulings, many that conflict. Because the law is ruled on without consistency.
All this blather about corporations, power, money, corruption is useful and fun to talk about if you are Congress. If you are the courts, it's all inane drivel irrelevant to the stuff at hand.
When defendants act as their own laywer in court they end up sounding like you. Talking about society this and power structure that. In the end they forget that the role of courts is not to make a just society, not to determine good or evil, but rather, to take a specific instance of an act, compare it to a specific instance of law, and decide how they relate.
Courts aren't in the business of restoring peoples faith in hte law. And they are not in the business in picking which laws are good or bad. Courts are in the business of ruling on how laws should be applied, which laws conflict with each other, and which laws are not enforceable.
A good decision here could go a way to help restoring people's faith in the law.
In my book, a good decision is a ruling rooted only in the law. In a lot of/. opinion, a good opinion is one that is what they want to hear to promote their agenda.
Of course a bad decision will confirm everybody's worst fears.
The worst fear beaing that the laws are made by and for those with money? If you think that is the case then your beef is with Congress.
Let me ask you this, to kinda of solidfy my kind of meandering point:
If Congress passes a law with the purpose of enriching the powerful and wealthy at the expense of the little guy, would a court ruling that enforces this law and bilks the little guy out of money bve a good ruling or a bad ruling?
Around here, the expectation is that it would be a bad ruling, regardless of how the ruling adheres to the written law. In my book, that ruling would be a good ruling.
My hope with this whole DMCA case is that the court follows the word of the law exactly. And in the future, I hope courts fully enforce every aspect of the DMCA, so that in the future, I can go to my congress person and use it as Exhibit A. I've gone to my people in Congress, and talked to them, and you know what? In every case they want examples of how the DMCA has been abused and how courts ruled.
The bottom line here is that a good and a bad ruling are really interesting questions.
The beta code was released to beta testers who paid for the privelage. DR-DOS's case was pretty weak, but MS basically made it go away to clear the anti-monopoly slate.
Obviously you are not a programmer. None of the features you are talking about are given to you by using Windows. They are all features of the ATM application and, again, could easily be done in 640k or less.
I am a programmer. And I've done a large amount of systems programming.
These things require complex interactions and transactions. It probably could be done in 640k - but not quickly! That's the point of using Windows - do these things quickly. All you are talking about is a few extra menu items which, under the hood, do the same thing. From the user standpoint there is additional functionality, but from an application standpoint there is not.
An absolute joke. You first claimed all you needed was a simple binary response - yes or no. But in fact what I've described requires complex interactions. Changing a pin, getting a statement, applying payments in realtime - it all takes complex interactions you clearly don't understand.
BUt really, all these atms need is one application and some loading code. They do not need a full operating system just to do what they do.
In that case no one needs an OS. Just some loading code and an application.
The point of using Windows was speed of development and cost. I am not supporting its use, and I wouldn't use it if it were me making that choice. But clearly you have no idea what type of information shapes decisions.
Hiring a systems engineer and developer to develop and ATM from scratch, writing some "loading code" and "application code" would be expensive. Development of everything is essentially from scratch - hardware drivers, transport layers, presentation layers, etc. Developing similiar functionality in the user layer of Windows takes a fraction of the time. I've done it both ways - and yes, Windows (or Linux, or any OS-based operation) takes a very small fraction of the time.
Furthermore, it IS good enough to use Windows. Why? How do I know? Because there a world of competitors and alternative products yet many banks still rely on Windows. Therefore by logical implication they - the banks who make the financial decisions - believe it is worth the trouble. And that's the bottom line. They've done the cost analysis, they've done the research, their decision is clear.
Your way is stupid, really. ATMs like you describe suck. And whether its Windows or QNX or WindRiver or Linux the OS model is here to stay. Commodity software development is vastly cheaper than custom designed hardware/software packages that you advocate.
Look, goddamnit, it's an ATM. It only has to count to 200, add and subtract, send a 12 digit number and a four digit number and get a binary response. If you are not a spiffy enough programmer to do that in 640k, you do not deserve to write for an ATM machine. KISS is an important principle especially when it comes to security.
Except, no, that's not good enough. If all things being equal, that's all your bank provided at the ATM, I'd stick with my credit union.
Things my credit union ATM's do:
Allow me to setup a preference so that on the pin entry screen I can press an alternate button and get a predetermined amount of cash. This saves a lot of time. Insert card. Enter six digit pin. Press cash button instead of Enter. Bamo. Cash. Card out. I drive away.
Allow me to transfer money between accounts.
Allow me to pay my mortgage and/or personal loan from any of my accounts.
Get a "mini-statement" of the month's activity to date.
Change my PIN.
Make deposits.
You simplify things to a point of silliness. Your vision is stuck in the late 80's. That's what ATMs used to do. Banks want people to really use ATMs. They want to offload expensive branches and employees into nifty banking centers called ATMs. It needs to be more than a "money machine". It needs to be banking center.
You program your ATM, and see what banks will buy it. Meanwhile while people like you are barking that so-and-so don't deserve to be programming companies like Diebold are eating up the marketplace. Think you can do better? Provide all the connectivity, functionality and still sell a more secure product? Then do it. No jabbering, no equovocations. Get out and do it. Make it happen. It's simple according to you.
After you've worked on it a few years come back to me and we can talk. It's not simple. That's why places like Diebold and other banks use Windows. Because using XPE is easy. It's fast and cheap to get a box out the door. Perfect? Nope. Good enough? Looks that way.
So you stick to KISS. But I won't be investing any banks that buy your KISS'd ATM. And I won't be doing business with them. Frankly, ATMs that can only dispense cash suck.
And how do you know that engineers didnt speak up?
It's obvious you've never had to deal with the amazing complexities of working in a large environment. Decisions do not emnate from where you might always expect. Every choice has political ramifications. Every choice has reprecutions.
Choices with technical errors are often are less expensive than other choices without those technical errors.
You have no idea what the process involved here was.
In fairness and respectability, and not to break your mind set of conspiracy, neopotism, fascism and republicanism, but:
The voting machines in question are not networked. Things don't happen in a networked way with them - basically data transfer is done manually.
In fairness then, if you are looking for a platform for a non-networked kiosk, XPE isnt probably that bad of choice. Clearly better exisit, but in terms of cost and time of development it doesn't really seem that it's a hideous choice.
Slashdot Mirror
Such boasts are obviously proved empty by full read write access from a boot disk
Any non encrypted filesystem would have all access controls subverted when mounted under a foreign operating system. Claiming anything else is absurdity. Do you think that ext2 holds the chmod based security when I mount it up under a box that it wasnt created under? You do realize that any user with root on any Linux box could simply reset all the attributes meant to keep users off/let them in under a different box?
Why would you hold MS to fault for something that is unversial and by its nature fundamental?
"Customarily" refers largely to how it has been done. Coding began as holes in paper. That was how programs were exchanged and "run".
So the custom really is paper. It goes back to the beginning of coding and continue to this day. It is a customary mode of code interchange.
Point being though, if you want to discourage users from seeing the source barcode meets the GPL but effectively stops users from seeing/changing the code.
Because as recent events show, video games are bigger than TV for today's youth, and bigger than movies, and bigger than anything.
People forget, but, Electronic Arts is a $6B company. That's really big people.
The TV rating folks over at Nielsen have just reported a big drop in male viewership of all media outlets.
They are getting a lot of heat because they are the big target right now. The Big One. The Video Game industry is bigger than the movies, and soon bigger than TV and even music.
What if it is in barcoded format?
Well then.. you've defeated the purprose of open source... i mean, if we cant tweak the algorithm, then, whats the point of having everything else?
Havent started getting those e-mails that look like articles with lots of text in but end with a bunch of spam at the bottom?
They slip through spam assassin pretty often for me. You think they came along just because? Or because of testing against anti-spam systems?
They dont even need the source code, just a working copy to test against.
Anything can be 'open-sourced' - it just takes an understanding of how it will be used and the implications of that use; a good developer will design for flexibility, rather than embedding key logic that may need to change inside the program.
In the case of a tax system those configuration files would be tax code. And that tax code is in fact open source.
The bottom line is though, even without the configuration files, you'd be able to engineer enough of how the system works - limits, test cases, even bugs! - that you'd have an unfair advantage. In the OSS world the common goal is *good software*. In the cases I give the goal would be to gain an advantage over the next guy. The motivations are different and the process falls apart.
Opening the source does not NECESSARILY allow people to take advantage of it.
Yes, actually, it does. As long as I know the rules, I can subvert them, test them, and gain unfair advantage. For example, in you crappy search engine, I'd advice my client to create millions and millions and milliosn of pages all linking back to the One True Page. IF you really rated pages randomly my client would get much better results than just having one page.
The point here is that, even if you know the algorithm, there are many factors that you either (1) may not know, or (2) cannot control. A well designed algorithm will be robust in the face of such things.
Except the cases I give your theory falls apart. In the tax scanerio you *know* all the factors - its your data. And you can control all the inputs. Meaning, you can drastically alter results based on the input you provide.
In the real world - the examples I provide - the systems are not and cannot be cryptographically secure, like the systems you pretend to compare them to. These systems are by their nature based on a strict set of rules - laws! - and must conform to them. Random audits are not effective, and have actually been banned by courts in many states. Audits have to based on probable cause - and that probable cause is litigated reguruarly.
The point is in the real world open sourcing many types of systems would lead to the complete ineffectiviness of the system.
You have this false test of when an algorithm is poorly designed or not. And the main thing you claim makes a nice algorithm is one that is based on some type of random seed. In fact in the real world this is rare. Real people will not accept the fact their account is being "randomly audited".
It has nothing to do with shoddy... its has to with the fact that if you have the code you know the rules.. and can subvert the rules..
Some systems can only be secure through obscurity.
Or it could be that to give discounts to other places they have to jack prices in the US up.
It's actually very socialist...
Of course, given that they won the antitrust case by paying off the Federal government after Bush got elected, this shouldn't surprise anyone...
Actually, you should get your facts straight. Bush campaigned repeatedly on the promise of ending the MS anti-trust trial as soon as possible via settlement.
Nope.
Believe it or not, there are some applications that CANNOT BE EFFECTIVELY OPEN SOURCED.
Open sourcing a search engine would 100% guarantee absolute junk for results.
If I had access to the code behind PageRank, I could guarantee my clients get excellent pacement. Same with other people. Honest people who just put up a website/page would be left in the dust by spammers.
Other examples where obscurity is the ONLY security:
1. Code that states/federal revenue services use to flag accounts for audits. This code, in the public, would instantly destroy the revenue stream of the government. Accountants with programming skills could determine *exactly* what limits they could test and get away with. Every return filed would be for the maximum amount that the code would allow without triggering and audit.
2. Fraud detection code used by credit companies, service providers, etc. Armed with this code crackers would have free reign over credit cards and online payment systems. Exact patterns of usuage could be setup to guarantee that fraud flags would not be triggered.
3. Code that determines which passengers get flagged for pre-flight searches. Armed with this information criminals could fashion profiles that guarantee they will not be probed in-depth.
These four examples destroy your silly notion. Open Source is not a magic pill. A truly open source version of Google would be a useless tool within a matter of weeks, if not days.
I think the strong court we have is overall a VERY good thing.
No, its a VERY bad thing.
It has lead down a path to a point where no one has any idea what any given judge will rule on any given day. Activism has taken things to the point now that there is positively no consistency in law and no coherence at all.
Any law however well intentioned that subverts the liberties of the people of the united states is a bad law and any judge worth his salt would deny the party using this law to subvert said liberties to prevail should be and eventually will be found to be in error by his peers.
That's absurd, and the system you describe is anarchy.
If Congress legally passes a law that is Constitutional and within the scope of their legal right to legislate, and a judge deny's the application of that law because it "subverts the liberties of the people" you have in effect no government at all.
When a law has been duly passed by Congress, and signed by the President, and upheld by the courts as legal and constitutional judges and law enforcement agencies are duty bound to enforce it. Just because it is "Bad" or "subverts" liberties does't meant judges can ignore it.
If you advocate otherwise you advocate for the fall of the American system.
Ohh that's brilliant.
This is why Congresspeople don't really like to talk to the public one on one.
I could garountee you that if you sat me down infront of the senate and asked me to explain to them why the DMCA is bad I could convince them within an hour as could just about any well educated technically inclined individual could.
No, I don't think so. There are other opinion, and frankly, most people don't give a damn about the things you'll tell them. For a lot of people, the issue is this: should a product designed to be uncopyable be protected by the law as though it was really uncopyable? Anything you can say or I can say isn't going to change that. But the DMCA isn't the issue.
On its face, the DMCA isn't vastly unconstitutional. It may in fact violate the Constitution in my opinion, but based on how courts have ruled before, based on how things have been going in terms of copyright and law in general, it's not a law that is unconstitutional on its face. So then the question is, once an individual judge has ruled the law constitional in his mind, is better that:
The judge disregard his finding, and other courts finding, and rule in favor of the defendant despite the law - effectively saying that the law is legal and valid but just not going to be enforced in this one case OR
The judge rule in favor of the plantiff and highlight the exact reasons in the law that the defendant is guilty
I prepose the second is better, the "good" ruling. The first ruling gets that one person off, but does nothing to further the cause getting the law I disagree with repealed.
The measure of a goverments success, in any incarnation, is it's ability to solve conflicts between people. A good decision would solve the majority of conflicts, while a bad decision would solve the minorty of them and a really bad decision would cause even more conflicts.
See, I suggest that is stupid, and is actually the reason why things aren't going that well. A court should rule based on a few narrow things: the facts as proven in court, the law as presented by Congress, and case law provided by other Courts. All this other stuff you mention is silly. And what it leads to is people feeling that the law is arbitrary. Theoretically, if you present 100 judges with the same law and 100 identical defendents all charged for comitting the same act that they did in fact commit you should get 100 rulings that are identical. Now, you'd get a 100 different rulings, many that conflict. Because the law is ruled on without consistency.
All this blather about corporations, power, money, corruption is useful and fun to talk about if you are Congress. If you are the courts, it's all inane drivel irrelevant to the stuff at hand.
When defendants act as their own laywer in court they end up sounding like you. Talking about society this and power structure that. In the end they forget that the role of courts is not to make a just society, not to determine good or evil, but rather, to take a specific instance of an act, compare it to a specific instance of law, and decide how they relate.
Courts aren't in the business of restoring peoples faith in hte law. And they are not in the business in picking which laws are good or bad. Courts are in the business of ruling on how laws should be applied, which laws conflict with each other, and which laws are not enforceable.
/. opinion, a good opinion is one that is what they want to hear to promote their agenda.
A good decision here could go a way to help restoring people's faith in the law.
In my book, a good decision is a ruling rooted only in the law. In a lot of
Of course a bad decision will confirm everybody's worst fears.
The worst fear beaing that the laws are made by and for those with money? If you think that is the case then your beef is with Congress.
Let me ask you this, to kinda of solidfy my kind of meandering point:
If Congress passes a law with the purpose of enriching the powerful and wealthy at the expense of the little guy, would a court ruling that enforces this law and bilks the little guy out of money bve a good ruling or a bad ruling?
Around here, the expectation is that it would be a bad ruling, regardless of how the ruling adheres to the written law. In my book, that ruling would be a good ruling.
My hope with this whole DMCA case is that the court follows the word of the law exactly. And in the future, I hope courts fully enforce every aspect of the DMCA, so that in the future, I can go to my congress person and use it as Exhibit A. I've gone to my people in Congress, and talked to them, and you know what? In every case they want examples of how the DMCA has been abused and how courts ruled.
The bottom line here is that a good and a bad ruling are really interesting questions.
The beta code was released to beta testers who paid for the privelage. DR-DOS's case was pretty weak, but MS basically made it go away to clear the anti-monopoly slate.
Obviously you are not a programmer. None of the features you are talking about are given to you by using Windows. They are all features of the ATM application and, again, could easily be done in 640k or less.
I am a programmer. And I've done a large amount of systems programming.
These things require complex interactions and transactions. It probably could be done in 640k - but not quickly! That's the point of using Windows - do these things quickly.
All you are talking about is a few extra menu items which, under the hood, do the same thing. From the user standpoint there is additional functionality, but from an application standpoint there is not.
An absolute joke. You first claimed all you needed was a simple binary response - yes or no. But in fact what I've described requires complex interactions. Changing a pin, getting a statement, applying payments in realtime - it all takes complex interactions you clearly don't understand.
BUt really, all these atms need is one application and some loading code. They do not need a full operating system just to do what they do.
In that case no one needs an OS. Just some loading code and an application.
The point of using Windows was speed of development and cost. I am not supporting its use, and I wouldn't use it if it were me making that choice. But clearly you have no idea what type of information shapes decisions.
Hiring a systems engineer and developer to develop and ATM from scratch, writing some "loading code" and "application code" would be expensive. Development of everything is essentially from scratch - hardware drivers, transport layers, presentation layers, etc. Developing similiar functionality in the user layer of Windows takes a fraction of the time. I've done it both ways - and yes, Windows (or Linux, or any OS-based operation) takes a very small fraction of the time.
Furthermore, it IS good enough to use Windows. Why? How do I know? Because there a world of competitors and alternative products yet many banks still rely on Windows. Therefore by logical implication they - the banks who make the financial decisions - believe it is worth the trouble. And that's the bottom line. They've done the cost analysis, they've done the research, their decision is clear.
Your way is stupid, really. ATMs like you describe suck. And whether its Windows or QNX or WindRiver or Linux the OS model is here to stay. Commodity software development is vastly cheaper than custom designed hardware/software packages that you advocate.
Ahh yes, well, aren't you spiffy...
Look, goddamnit, it's an ATM. It only has to count to 200, add and subtract, send a 12 digit number and a four digit number and get a binary response. If you are not a spiffy enough programmer to do that in 640k, you do not deserve to write for an ATM machine. KISS is an important principle especially when it comes to security.
Except, no, that's not good enough. If all things being equal, that's all your bank provided at the ATM, I'd stick with my credit union.
Things my credit union ATM's do:
Allow me to setup a preference so that on the pin entry screen I can press an alternate button and get a predetermined amount of cash. This saves a lot of time. Insert card. Enter six digit pin. Press cash button instead of Enter. Bamo. Cash. Card out. I drive away.
Allow me to transfer money between accounts.
Allow me to pay my mortgage and/or personal loan from any of my accounts.
Get a "mini-statement" of the month's activity to date.
Change my PIN.
Make deposits.
You simplify things to a point of silliness. Your vision is stuck in the late 80's. That's what ATMs used to do. Banks want people to really use ATMs. They want to offload expensive branches and employees into nifty banking centers called ATMs. It needs to be more than a "money machine". It needs to be banking center.
You program your ATM, and see what banks will buy it. Meanwhile while people like you are barking that so-and-so don't deserve to be programming companies like Diebold are eating up the marketplace. Think you can do better? Provide all the connectivity, functionality and still sell a more secure product? Then do it. No jabbering, no equovocations. Get out and do it. Make it happen. It's simple according to you.
After you've worked on it a few years come back to me and we can talk. It's not simple. That's why places like Diebold and other banks use Windows. Because using XPE is easy. It's fast and cheap to get a box out the door. Perfect? Nope. Good enough? Looks that way.
So you stick to KISS. But I won't be investing any banks that buy your KISS'd ATM. And I won't be doing business with them. Frankly, ATMs that can only dispense cash suck.
And how do you know that engineers didnt speak up?
It's obvious you've never had to deal with the amazing complexities of working in a large environment. Decisions do not emnate from where you might always expect. Every choice has political ramifications. Every choice has reprecutions.
Choices with technical errors are often are less expensive than other choices without those technical errors.
You have no idea what the process involved here was.
In fairness and respectability, and not to break your mind set of conspiracy, neopotism, fascism and republicanism, but:
The voting machines in question are not networked. Things don't happen in a networked way with them - basically data transfer is done manually.
In fairness then, if you are looking for a platform for a non-networked kiosk, XPE isnt probably that bad of choice. Clearly better exisit, but in terms of cost and time of development it doesn't really seem that it's a hideous choice.
Or they could be reforming the problems that made SVG unlikley to suceed.
Time will tell, I suppose.
They added error messages on Windows (3.1?) so that people would get scared when they would run Windows on another DOS than MS-DOS.
It was never on release code, only on beta code and only against DR-DOS. It was a non-fatal error, and they settled for about $20M.
It's not really a good example to prove the point.