Firstly, the most readable fonts are the simplest ones: no frills. Right now, that's Arial, but fonts even simpler could be designed. And for such no-frills fonts, everything can easily be represented by equations. It should also be noted that perfect symmetry, wherever possible, as in the letter o, will always scale better than asymmetry. Without anti-aliasing, a perfect circle on your computer screen will scale down to 3 pixels by 3 pixels, and still be recognizable as a perfect circle. Such a font of perfect symmetry -- i.e., 45 degree angles, perfect circles or semi-circles -- could easily and quickly be designed. As for the curved effect of italic fonts you mentioned, this could also be represented mathamatically. This would produce a font that would be the most readable.
Secondly, as for Courrier, that's basically type-writer font, and has been around as long as type-writers. Thus, courrier fonts can't be copyrighted.
Thirdly, as for Times New Roman, that's also been around forever; its the font that was used in newspapers and journals in the early 1900's.
These are the only fonts I've claimed that anyone really needs, aside from the symbols in Greek/Roman.
As for fonts smaller than size 9/10, I have a 1600x1200 computer screen with full-scene anti-aliasing 4x4 enabled and forced in all applications, via a GeForce2 GTS. I have yet to find one font which is readable at size 8 or below.
Btw, the thought of people "owning the copyrights" to basic every-day fonts like System, Courrier, Times, and Arial is a disturbing one. Talk about carving the language up among various proprietary interests.
Alright, its obvious your thinking about fonts in a rather sloppy way. Don't think of a font as a piece of artwork; think of it as an equation, or set of equations, in which manipulating a few variables (i.e., tilt, width, etc) can change the appearance of the font.
All English characters can be described by relatively simple equations. I'll go through a few of them:
X and V: some easy ones y1 = |mx + b| for a = x = b y2 = -|mx + b| for a = x = b thickiness (in pixels) = point-size * t where t is a user-specified value (the absolute values insure that the lines will be in opposite directions)
Changing m will change how sharply the angles are at eachother, while changing b affects how centered the font is, and can change a font from a X to a V.
O: the easiest one x^2/a^2 + y^2/c^2 = 1
Pretty self-explanatory there.
And so on and so forth. I'm not going to go through every example, but all of the characters of the English language are describable by relatively simple mathematics. The more complex symnbols of the Greek language such as alpha and zeta require more complicated equations. Never-the-less, almost any character (from a Eurocentric point of view) can be expressed at its most basic level by a set of mathematical formuli, where modifying the variables changes the exact implementation of the character.
In other words, duh, you don't remake the font for each point-size you can possibly need. You use an equation which will define it perfectly for every possible point size. Alternatively, one can simply scan in fonts and have a program create the formula's to represent them, then modify the parameters of those formula's to produce the desired effects.
A simple program could even be created to allow representation of what fonts are at their most basic level (i.e., what is an a, at its most basic level) and then allow the user to manipulate the variables to produce various effects.
Producing bold, italic, underline, effects is very easy once the basic font design is created; just tilt the font for italics, increase thickness for bold, and underline for underline.
There are a billion people across the Pacific who use fonts that have upwards of 10,000 different characters. I find Japanese on non-commercial unixes simply unbearable, the fonts are so bad. While presumably the Chinese government is willing to design it's own fonts for its version of Linux, that does no good for the Japanese or Korean markets, and little for Chinese markets that do not use PRC characters.
And no, the fonts you mention are not damn easy to create.
I'll admit, I AM Eurocentric. I don't give a flying hoot about Chinese fonts or Japanese fonts (though Hieroglphics fonts would be pretty cool).
All I want is English fonts (and symbols), and, lets face it, English fonts pretty easy to create. Letters in the English language all have a symmetry to them. A U is just two lines connected by a semi-circle. An A is just two lines at an angle to eachother with a horizontal line connecting them. A B is just a line with two semi-circles meeting at a point, extending to the edge. A V is just two lines at an angle to eachother. A T is just a vertical line and a horizontal line. Y is just a v with a vertical line below it. X is two lines intersecting eachother. An S is just two connected semi-circles. An i is just a line with a dot above it. So, please, do not tell me it is hard to create fonts in the English language. I could create mathematical formuli to represent these characters on a TI-84. All simple (i.e., no frills) fonts are basically variations on the parameters of these mathematics, while complicated ones usually add a little extra.
Personally, I prefer very simple fonts like Arial (which is basically no frills, just simple mathematics) and Courrier; the most complicated font I like is Times New Roman.
That's English fonts. I'll admit, it does become very complicated when one is doing Symbols (i.e., Greek characters like zeta). But, never-the-less, it is still just mathematics, with a few different functions describing what the character is at its most rudimentary level; parameters could be used to modify the exact implentation of the font; i.e., how big should the loop on the gamma or alpha be could be manipulated by a variable.
The reason why people think that fonts are so hard is because they're not thinking about them in a rigorous mathematical way. They're thinking about them in a loose, sloppy way.
The only way your argument could be considered logical is if the OSS way of doing things was better. Unfortunately that is not the case, so your entire argument simply falls apart.
Actually, the OSS way is the better way of doing things: its produced better results. Duh. Please don't try to tell me that WinXP is as secure and stable as Debian, Slackware, OpenBSD, or FreeBSD. Its not. Period.
That's the results. And you can't argue with that: OSS has simply produced superior results so far. In other words, the evidence supports my theory. The only thing you have is your theories, which aren't backed up by reality.
Even in theory, the OSS way of doing things is better. With OSS, you have the "many eyes" advantage, in that anyone can look for problems and fix them; you can also get expert help from organizations like the RSA.
You are the one who is veering from the real issues here. You still haven't responded to the widely acknowledged fact that security/stability is much better on a properly configured (which is basically out of the box, for OpenBSD and Debian) Linux/BSD box. Nor have you responded to my point that BSD/Linux organizations can get help from the RSA as well. MS isn't the only one who can work with the RSA.
With Linux and BSD, problems -- be they performance, security, or stability -- are openly acknowledged and usually quickly fixed. No press-spin is done to try to make it look better.
With MS, problems are not acknowledged and they'll sue you via the DMCA or any other applicable law if you point out problems with their software. Additionally, fixes come out at a very slow pace (though not quite as slow as fixes come out for Solaris).
So part of it comes down to the fact that you can't trust any corporation. Corporations are there to take money from your pocket and put it into theirs: not to make good products. They'll use any methods they deem will more likely than not give them more money, be those methods legal or illegal, whether they benefit their customers or not (refer here to MS' forced upgrade scheme).
We don't need a zillion fucking fonts. Get over yourself. Most of these great fonts you people come up with are stuff we NEVER use. I've never used anything other than Times (New Roman), Arial, and Courrier (New). We, the Linux community, or most of us anyways, don't need new fonts from scratch. We need standard fonts which you see in every newspaper, article, letter, or webpage. That is: Times New Roman, Courrier New, and Arial. That's it. We don't need "font families". We need THREE fonts, which are standard and pretty damn easy to create.
In case you still haven't gotten it, NO we don't need those fancy unreadable cursive fonts; NO, we don't need those fancy unreadable English fonts. Well we need is to get THREE freaking fonts to be TrueTyped and antialiased nicely, so they scale well: Times, Courrier, and Arial. These fonts have been around for ages, so any copyrights or patents or whatever on them ran out. Courrier is basically typewriter font; typewriters have been arround for ages, so that font's available to us. Times New Roman is basically newspaper/article font, which has also been around for ages: again, no copyrights/patents apply anymore. Arial is a no-frills wont with no fancy angles, so its pretty easy to create from scratch.
By the way, you fail to mention that many Linux vendors -- i.e., RedHat, TurboLinux -- also have relationships with RSA (type in Linux to see), thus nullifying any potential advantage MS would have over them via RSA.
You don't have a point, because MS' products have a legacy for being ridden with security and stability problems, not to mention other annoying bugs. Whatever organizations MS uses to help it check for problems obviously aren't doing their job very well.
Even if these organizations are good, they obviously aren't doing enough for MS. Maybe RSA is the best at what they do, but they certainly haven't made MS' products the most stable or secure. MJ may be the best basketball player, but he didn't make the Wizards the best basketball team.
Please, your telling me it'd take years for a Linux group to design replacements for industry standards, like Courrier, Times, and Arial? And don't be mistaken, those are really the only fonts commonly used -- like 99% of the time.
Asking for donations -- even explicitly -- is fine.
Spamming is not.
By spamming, they are forcibly transferring the cost of them asking for donations from them to us.
This is wrong.
But, at least their messages are text-based, and not huge graphics eating up your bandwidth.
I understand that these people need money to run their service. However, that doesn't justify their tactics.
Similarly, I understand that the people who run charities to help the poor need money to do that...that doesn't justify them phoning me with a taped recording.
All that's relevant is the results, and MS has a crappy record.
It takes them ages to release bug fixes, and they're hardly upfront about security vulnerabilities/problems.
Maybe OSS isn't the only way. However, its certainly produced great results so far with minimal costs. Compare that to MS -- horrible results, with outrageous costs, and their products are ridiculously pricey.
What I was saying is that the actual design of the F-117 needs to remain secret.
So does the software used to create it. The worry is not that someone will find a weakness in our stealth technology (be it the F-117, B-2, or F-22), because there aren't weaknesses in it.
The worry is that a Saddam Hussein could use the software we used to create the F-117 to create a stealth jet of his own.
Its too much to hope for that two evil forces will oppose eachother.
No, evil forces unite to crush good forces.
Just look at our Senate and Congress. They might not be able to agree on tax cuts or balancing the budget, the they all unanimously agree that they deserve a fucking pay raise, despite doing their job worse than last year.
Thank you, dear RIAA, for informing me of that site which I had no idea even existed.
Now, I think I'll go and download some Christina Aguilera music. I don't particularly like her (more of a B. Spears person), but since its free, I'll take it.
Seriously, ISPs have no business blocking web sites, or otherwise censoring the net. They are there to connect people to the internet, not to block them off from parts of it that special interests think we shouldn't see.
Your argument sure sounds good, but OSS/FS/public domain libraries aren't any better than those provided by MS when you find a bug. Maybe you're a programming genius who can immediately immerse yourself in an entirely new code module and just *know* where the bug is, but for us mortals it's just not practical to go bug hunting every time we find one in someone else's "good outside service."
Your argument sure sounds good, but you ignore a few key points: documentation and communication.
Documentation. Source code isn't just plopped out there. Alot of times, the specific parts are well-documented.
Communication. If you know what kind of vulnerability your looking for, or what kind of stability problem you need to fix, or where you want to improve performance, you can e-mail the developer with your inquiries, asking him/her what the relevant parts of the source-code are.
Make products buggy as hell, then get people to upgrade and pay them for it by releasing new versions which have fixed the old bugs, but introduced new bugs. Repeat ad infinetum.
In parallel, also make sure to develop file formats and "standards" which aren't backwards compatable and don't work with any other OS', so as to lock people into MS products and force costly upgrades.
This is something that really needs to be evaluated carefully. At least I have found that for many of the tasks that I need to do using a computer the open source software either doesn't exist, or is so far behind the other alternatives that deploying would involve more money. For example, gnu/Linux and a variety of GUI frameworks for Linux lack a mature accessibility framework. The only mature speech recognition engine for Linux is not open source at this time. The Sphinx product is not yet usable. ADA compliance is one area in which I feel that mandating a complete open source shop would run into problems.
Then perhaps, in such cases, the government should buy out the proprietary product that is the best and OSS / FS it. Or perhaps they should have a contract with the company that develops the software saying, "we'll pay you XXX dollars a year" provided you develop such a product and release it as OSS / FS / public domain.
Alternatively, we could create laws which say that OSS / FS is to be used in all cases where it is a viable alternative. I would also support such a law.
Microsoft officials said it makes sense for the operating system to provide cryptographic services to any application that needs it, instead of each application having to include its own cryptographic technology
Yes, indeed, it does make sense for the OS to provide such a service to any program that wants to use it, so long as that's a GOOD service.
In general, it makes sense to provide everything from outside the program, and just have the program call on outside services. However, that means you need to make the outside services good, and it means that those writing programs don't just string together a bunch of requests (i.e., draw this, check that calls) but also work on looking for fixes to the common outside service, which would be shared by many programs.
In other words, this approach only makes sense when the outside services are OSS / FS / public domain, which means that developers of programs can check their integrity and submit improvements. Otherwise, its just a big black hole for developers: should I trust this cryptographic routine, or shouldn't I? One never knows with proprietary routines. One can check, and improve such routines provided OSS / FS.
Then perhap the problem is in the IRS' inadequate auditing processes.
If disclosure of the process would result in people scamming the IRS, then it can't be that great, which I had a hint of anyways. The IRS doesn't audit famous people like Jesse Jackson, who's certainly deserving of an audit. B/c they're afraid. And because its easier and cheaper to squeeze money out of a middle class person, whether they filled out their forms right or not. Regarding the IRS, people unconstitutionally lose their presumption of innocence: the IRS presumes you guilty until you can prove otherwise, and you lose money because of it.
Also, and I was thinking this before, don't you think we have the right to know what criteria the IRS uses for auditing people? I.e., I wouldn't be surprised if criteria like the following exists:
ONLY audit the weak and defenseless.
Leave the powerful like Jesse Jackson alone, because that's too much trouble and bad press, and we're afraid to be labeled racists.
I took issue with your example because I don't think this law is really about holding the government accountable.
Having the software they use be open certainly increases accountability, or at least the possibility for it.
State governments usually have built-in means to audit security and spending.
And they've done SO WELL, haven't they? Why, these spending audits stopped us from getting into trillions of dollars of debt. And they also stopped insane terrorists from crashing into the Wolrld Trade Center. Oh wait, they didn't.
You keep assuming that specifics of the audit process need be in the code. Why can't the specifics all be general values, which are supplied by a configuration file?
Or just have the program interpret a script in a file, and then act on it.
Regarding the post office and air traffic control, there may not be an OSS/FS solution, but the gov't could easily buy out the proprietary solution and make it such.
Regarding your confusing about my comments on Michael and O'Reilly. What I meant to say was the Michael did not say that O'Reilly had a hidden agenda; rather, Michael said that perhaps O'Reilly has a hidden agenda.
I disagree. We should be able to check every minute thing the government's doing. Your standard regarding general disclosure is nice, but who's to say what "for the most part" is?
Food for thought: Had the stockholders of Enron and Global Crossings been able to micromanage and scrutinize every move made by those companies, they might not have been screwed out of their life savings.
Slashdot Mirror
Firstly, the most readable fonts are the simplest ones: no frills. Right now, that's Arial, but fonts even simpler could be designed. And for such no-frills fonts, everything can easily be represented by equations. It should also be noted that perfect symmetry, wherever possible, as in the letter o, will always scale better than asymmetry. Without anti-aliasing, a perfect circle on your computer screen will scale down to 3 pixels by 3 pixels, and still be recognizable as a perfect circle. Such a font of perfect symmetry -- i.e., 45 degree angles, perfect circles or semi-circles -- could easily and quickly be designed. As for the curved effect of italic fonts you mentioned, this could also be represented mathamatically. This would produce a font that would be the most readable.
Secondly, as for Courrier, that's basically type-writer font, and has been around as long as type-writers. Thus, courrier fonts can't be copyrighted.
Thirdly, as for Times New Roman, that's also been around forever; its the font that was used in newspapers and journals in the early 1900's.
These are the only fonts I've claimed that anyone really needs, aside from the symbols in Greek/Roman.
As for fonts smaller than size 9/10, I have a 1600x1200 computer screen with full-scene anti-aliasing 4x4 enabled and forced in all applications, via a GeForce2 GTS. I have yet to find one font which is readable at size 8 or below.
Btw, the thought of people "owning the copyrights" to basic every-day fonts like System, Courrier, Times, and Arial is a disturbing one. Talk about carving the language up among various proprietary interests.
--All your letters belong to me!
What about encryption techniques designed specifically to foil quantum computers?
There must be some way which encryption can be made to work so that they can't break it in polynomial time.
I don't get it. Quantum computers could easily break 128-bit encryption, the same encryption which would take todays supercomputers eons to break.
Why not just use encryption that's high enough that it would take even quantum computers eons to break? (i.e., 1giga-bit encryption)?
Alright, its obvious your thinking about fonts in a rather sloppy way. Don't think of a font as a piece of artwork; think of it as an equation, or set of equations, in which manipulating a few variables (i.e., tilt, width, etc) can change the appearance of the font.
All English characters can be described by relatively simple equations. I'll go through a few of them:
X and V: some easy ones
y1 = |mx + b| for a = x = b
y2 = -|mx + b| for a = x = b
thickiness (in pixels) = point-size * t
where t is a user-specified value
(the absolute values insure that the lines will be in opposite directions)
Changing m will change how sharply the angles are at eachother, while changing b affects how centered the font is, and can change a font from a X to a V.
O: the easiest one
x^2/a^2 + y^2/c^2 = 1
Pretty self-explanatory there.
And so on and so forth. I'm not going to go through every example, but all of the characters of the English language are describable by relatively simple mathematics. The more complex symnbols of the Greek language such as alpha and zeta require more complicated equations. Never-the-less, almost any character (from a Eurocentric point of view) can be expressed at its most basic level by a set of mathematical formuli, where modifying the variables changes the exact implementation of the character.
In other words, duh, you don't remake the font for each point-size you can possibly need. You use an equation which will define it perfectly for every possible point size. Alternatively, one can simply scan in fonts and have a program create the formula's to represent them, then modify the parameters of those formula's to produce the desired effects.
A simple program could even be created to allow representation of what fonts are at their most basic level (i.e., what is an a, at its most basic level) and then allow the user to manipulate the variables to produce various effects.
Producing bold, italic, underline, effects is very easy once the basic font design is created; just tilt the font for italics, increase thickness for bold, and underline for underline.
There are a billion people across the Pacific who use fonts that have upwards of 10,000 different characters. I find Japanese on non-commercial unixes simply unbearable, the fonts are so bad. While presumably the Chinese government is willing to design it's own fonts for its version of Linux, that does no good for the Japanese or Korean markets, and little for Chinese markets that do not use PRC characters.
And no, the fonts you mention are not damn easy to create.
I'll admit, I AM Eurocentric. I don't give a flying hoot about Chinese fonts or Japanese fonts (though Hieroglphics fonts would be pretty cool).
All I want is English fonts (and symbols), and, lets face it, English fonts pretty easy to create. Letters in the English language all have a symmetry to them. A U is just two lines connected by a semi-circle. An A is just two lines at an angle to eachother with a horizontal line connecting them. A B is just a line with two semi-circles meeting at a point, extending to the edge. A V is just two lines at an angle to eachother. A T is just a vertical line and a horizontal line. Y is just a v with a vertical line below it. X is two lines intersecting eachother. An S is just two connected semi-circles. An i is just a line with a dot above it. So, please, do not tell me it is hard to create fonts in the English language. I could create mathematical formuli to represent these characters on a TI-84. All simple (i.e., no frills) fonts are basically variations on the parameters of these mathematics, while complicated ones usually add a little extra.
Personally, I prefer very simple fonts like Arial (which is basically no frills, just simple mathematics) and Courrier; the most complicated font I like is Times New Roman.
That's English fonts. I'll admit, it does become very complicated when one is doing Symbols (i.e., Greek characters like zeta). But, never-the-less, it is still just mathematics, with a few different functions describing what the character is at its most rudimentary level; parameters could be used to modify the exact implentation of the font; i.e., how big should the loop on the gamma or alpha be could be manipulated by a variable.
The reason why people think that fonts are so hard is because they're not thinking about them in a rigorous mathematical way. They're thinking about them in a loose, sloppy way.
Think about an O as x^2/a^2 + y^2/b^2 = r^2.
God, why not just use GigaFlops like real scientists do? These are numbers which actually mean something.
The only way your argument could be considered logical is if the OSS way of doing things was better. Unfortunately that is not the case, so your entire argument simply falls apart.
Actually, the OSS way is the better way of doing things: its produced better results. Duh. Please don't try to tell me that WinXP is as secure and stable as Debian, Slackware, OpenBSD, or FreeBSD. Its not. Period.
That's the results. And you can't argue with that: OSS has simply produced superior results so far. In other words, the evidence supports my theory. The only thing you have is your theories, which aren't backed up by reality.
Even in theory, the OSS way of doing things is better. With OSS, you have the "many eyes" advantage, in that anyone can look for problems and fix them; you can also get expert help from organizations like the RSA.
You are the one who is veering from the real issues here. You still haven't responded to the widely acknowledged fact that security/stability is much better on a properly configured (which is basically out of the box, for OpenBSD and Debian) Linux/BSD box. Nor have you responded to my point that BSD/Linux organizations can get help from the RSA as well. MS isn't the only one who can work with the RSA.
With Linux and BSD, problems -- be they performance, security, or stability -- are openly acknowledged and usually quickly fixed. No press-spin is done to try to make it look better.
With MS, problems are not acknowledged and they'll sue you via the DMCA or any other applicable law if you point out problems with their software. Additionally, fixes come out at a very slow pace (though not quite as slow as fixes come out for Solaris).
So part of it comes down to the fact that you can't trust any corporation. Corporations are there to take money from your pocket and put it into theirs: not to make good products. They'll use any methods they deem will more likely than not give them more money, be those methods legal or illegal, whether they benefit their customers or not (refer here to MS' forced upgrade scheme).
God, some people are fucking DENSE.
We don't need a zillion fucking fonts. Get over yourself. Most of these great fonts you people come up with are stuff we NEVER use. I've never used anything other than Times (New Roman), Arial, and Courrier (New). We, the Linux community, or most of us anyways, don't need new fonts from scratch. We need standard fonts which you see in every newspaper, article, letter, or webpage. That is: Times New Roman, Courrier New, and Arial. That's it. We don't need "font families". We need THREE fonts, which are standard and pretty damn easy to create.
In case you still haven't gotten it, NO we don't need those fancy unreadable cursive fonts; NO, we don't need those fancy unreadable English fonts. Well we need is to get THREE freaking fonts to be TrueTyped and antialiased nicely, so they scale well: Times, Courrier, and Arial. These fonts have been around for ages, so any copyrights or patents or whatever on them ran out. Courrier is basically typewriter font; typewriters have been arround for ages, so that font's available to us. Times New Roman is basically newspaper/article font, which has also been around for ages: again, no copyrights/patents apply anymore. Arial is a no-frills wont with no fancy angles, so its pretty easy to create from scratch.
By the way, you fail to mention that many Linux vendors -- i.e., RedHat, TurboLinux -- also have relationships with RSA (type in Linux to see), thus nullifying any potential advantage MS would have over them via RSA.
You don't have a point, because MS' products have a legacy for being ridden with security and stability problems, not to mention other annoying bugs. Whatever organizations MS uses to help it check for problems obviously aren't doing their job very well.
Even if these organizations are good, they obviously aren't doing enough for MS. Maybe RSA is the best at what they do, but they certainly haven't made MS' products the most stable or secure. MJ may be the best basketball player, but he didn't make the Wizards the best basketball team.
Please try thinking next time.
Please, your telling me it'd take years for a Linux group to design replacements for industry standards, like Courrier, Times, and Arial? And don't be mistaken, those are really the only fonts commonly used -- like 99% of the time.
Asking for donations -- even explicitly -- is fine.
Spamming is not.
By spamming, they are forcibly transferring the cost of them asking for donations from them to us.
This is wrong.
But, at least their messages are text-based, and not huge graphics eating up your bandwidth.
I understand that these people need money to run their service. However, that doesn't justify their tactics.
Similarly, I understand that the people who run charities to help the poor need money to do that...that doesn't justify them phoning me with a taped recording.
Your question is irrelevant.
All that's relevant is the results, and MS has a crappy record.
It takes them ages to release bug fixes, and they're hardly upfront about security vulnerabilities/problems.
Maybe OSS isn't the only way. However, its certainly produced great results so far with minimal costs. Compare that to MS -- horrible results, with outrageous costs, and their products are ridiculously pricey.
What I was saying is that the actual design of the F-117 needs to remain secret.
So does the software used to create it. The worry is not that someone will find a weakness in our stealth technology (be it the F-117, B-2, or F-22), because there aren't weaknesses in it.
The worry is that a Saddam Hussein could use the software we used to create the F-117 to create a stealth jet of his own.
That is, assuming the ISP's don't back down.
Its too much to hope for that two evil forces will oppose eachother.
No, evil forces unite to crush good forces.
Just look at our Senate and Congress. They might not be able to agree on tax cuts or balancing the budget, the they all unanimously agree that they deserve a fucking pay raise, despite doing their job worse than last year.
Thank you, dear RIAA, for informing me of that site which I had no idea even existed.
Now, I think I'll go and download some Christina Aguilera music. I don't particularly like her (more of a B. Spears person), but since its free, I'll take it.
Seriously, ISPs have no business blocking web sites, or otherwise censoring the net. They are there to connect people to the internet, not to block them off from parts of it that special interests think we shouldn't see.
Your argument sure sounds good, but OSS/FS/public domain libraries aren't any better than those provided by MS when you find a bug. Maybe you're a programming genius who can immediately immerse yourself in an entirely new code module and just *know* where the bug is, but for us mortals it's just not practical to go bug hunting every time we find one in someone else's "good outside service."
Your argument sure sounds good, but you ignore a few key points: documentation and communication.
Documentation. Source code isn't just plopped out there. Alot of times, the specific parts are well-documented.
Communication. If you know what kind of vulnerability your looking for, or what kind of stability problem you need to fix, or where you want to improve performance, you can e-mail the developer with your inquiries, asking him/her what the relevant parts of the source-code are.
Make products buggy as hell, then get people to upgrade and pay them for it by releasing new versions which have fixed the old bugs, but introduced new bugs. Repeat ad infinetum.
In parallel, also make sure to develop file formats and "standards" which aren't backwards compatable and don't work with any other OS', so as to lock people into MS products and force costly upgrades.
Bwuhahahaha.
This is something that really needs to be evaluated carefully. At least I have found that for many of the tasks that I need to do using a computer the open source software either doesn't exist, or is so far behind the other alternatives that deploying would involve more money. For example, gnu/Linux and a variety of GUI frameworks for Linux lack a mature accessibility framework. The only mature speech recognition engine for Linux is not open source at this time. The Sphinx product is not yet usable. ADA compliance is one area in which I feel that mandating a complete open source shop would run into problems.
Then perhaps, in such cases, the government should buy out the proprietary product that is the best and OSS / FS it. Or perhaps they should have a contract with the company that develops the software saying, "we'll pay you XXX dollars a year" provided you develop such a product and release it as OSS / FS / public domain.
Alternatively, we could create laws which say that OSS / FS is to be used in all cases where it is a viable alternative. I would also support such a law.
Microsoft officials said it makes sense for the operating system to provide cryptographic services to any application that needs it, instead of each application having to include its own cryptographic technology
Yes, indeed, it does make sense for the OS to provide such a service to any program that wants to use it, so long as that's a GOOD service.
In general, it makes sense to provide everything from outside the program, and just have the program call on outside services. However, that means you need to make the outside services good, and it means that those writing programs don't just string together a bunch of requests (i.e., draw this, check that calls) but also work on looking for fixes to the common outside service, which would be shared by many programs.
In other words, this approach only makes sense when the outside services are OSS / FS / public domain, which means that developers of programs can check their integrity and submit improvements. Otherwise, its just a big black hole for developers: should I trust this cryptographic routine, or shouldn't I? One never knows with proprietary routines. One can check, and improve such routines provided OSS / FS.
Then perhap the problem is in the IRS' inadequate auditing processes.
If disclosure of the process would result in people scamming the IRS, then it can't be that great, which I had a hint of anyways. The IRS doesn't audit famous people like Jesse Jackson, who's certainly deserving of an audit. B/c they're afraid. And because its easier and cheaper to squeeze money out of a middle class person, whether they filled out their forms right or not. Regarding the IRS, people unconstitutionally lose their presumption of innocence: the IRS presumes you guilty until you can prove otherwise, and you lose money because of it.
Also, and I was thinking this before, don't you think we have the right to know what criteria the IRS uses for auditing people? I.e., I wouldn't be surprised if criteria like the following exists:
ONLY audit the weak and defenseless.
Leave the powerful like Jesse Jackson alone, because that's too much trouble and bad press, and we're afraid to be labeled racists.
I took issue with your example because I don't think this law is really about holding the government accountable.
Having the software they use be open certainly increases accountability, or at least the possibility for it.
State governments usually have built-in means to audit security and spending.
And they've done SO WELL, haven't they? Why, these spending audits stopped us from getting into trillions of dollars of debt. And they also stopped insane terrorists from crashing into the Wolrld Trade Center. Oh wait, they didn't.
You keep assuming that specifics of the audit process need be in the code. Why can't the specifics all be general values, which are supplied by a configuration file?
Or just have the program interpret a script in a file, and then act on it.
Regarding the post office and air traffic control, there may not be an OSS/FS solution, but the gov't could easily buy out the proprietary solution and make it such.
Regarding your confusing about my comments on Michael and O'Reilly. What I meant to say was the Michael did not say that O'Reilly had a hidden agenda; rather, Michael said that perhaps O'Reilly has a hidden agenda.
I disagree. We should be able to check every minute thing the government's doing. Your standard regarding general disclosure is nice, but who's to say what "for the most part" is?
Food for thought: Had the stockholders of Enron and Global Crossings been able to micromanage and scrutinize every move made by those companies, they might not have been screwed out of their life savings.