Actually I interned for 3 years doing embedded avionics for things like fighter jets, the space shuttle, and commercial airliners. After finishing school I got laid off, since I was working there part time my senior year on the Electronic Flight Bag project for Boeing. Then I started contract work, quit that when the network security startup came around, we sold the company, and I went back to embedded work as a contractor on the A380 super jumbo for airbus. Then I quit that job because the of poor planning, too many hours, and no respite in sight after a year of working my ass off. Now I've got a full-time gig doing maritime control systems, at a base just under 6 figures with benefits pushing it over.;)
My current job is awesome, 39-40 hr weeks, low stress, very manageable projects and time for grad school in the evenings after I establish residency here for in-state tuition. Oh and my boss wants to move towards embedded code generation using modeling tools like MATLAB & Simulink. This all makes for a very happy programmer =)
I wonder if the democrat controlled congress has anything to do with this? I know Microsoft and others are constantly lobbying to import more slaves^H^H^H^H^H^H H1-B visa workers to get around paying the wages necessary for skilled domestic labor.
I think you missed the part about not having room for an interpreter in the micros I work with. Some have as little as 32K ROM and 2K of RAM. When working with some 8 bit chips the resources are even more limited and its stops making sense to use C Code. Saying that is not the case is ignoring industry, you know anything in the controls/automation field.
I consider Python a real language, I do know that it can be compiled. I have no interest in learning web programming, I simply do not have time to focus in that direction. I've worked with many scripting languages. I was not implying that web programming is not a real discipline in CS, just that calling those of us which work with tools that generate machine code dinosaurs is arrogant.
PS - I made low 6 figures last year, some from stock options from a successful startup, most from a healthy hourly rate doing contract work with a bit of overtime.
I resent that remark. I'm an embedded developer, I work a lot with microcontrollers, usually with less than a MB of RAM and Flash. My skills are quite rare in the market these days because of attitudes like that. I'm not going to brag about my salary, but since my current employer spent 4 yrs trying to fill my position, its quite good.
I'm not a 50 yr old dinosaur either, I'm 26. I happen to use a lot of python in my day to day work to do little side tasks faster. Languages are just tools, however the only tools I can use for the firmware I write involve C code and Assembly. However I will be generating code for the next project using MATLab and Simulink, but if I did not know how to work with the output, all those ultra advanced tools would be useless.
It is not yet a treatment, only experimental compound used on lab animals. As is, if the compound were modified and approved for use on humans it would not be immediately useful against Parkinson's. Eventually what is learned by further studying the processes involved the article asserts it may one day lead to treatments for Parkinson's.
The point I made was not that copyright infringement is theft, just that the parent poster completely missed the issue and posted the same thing that gets posted in every one of these stories. We get it, copyright infringement is not theft. Unfortunately had the poster possibly RTFA he would have realized that profiting from a copyrighted work and not paying the licensing fees for distribution is Stealing/Theft/Robbery whatever you want to call it. You've now taken money from consumers and stolen the RIAA's share of it. This is not about the music, its about the money.
That's why they said that allofmp3.com was robbing them of sales. They are contending that the sales which are made by that site directly correlate to purchases not made through RIAA sanctioned channels. Of course their assumption is that without sites like allofmp3.com they would be making those sales, therefore it is stealing. Whether those sales would then go to official channels is a moot point.
I know the copyright infringement != theft line is common around here, but please respond intelligently to issues raised by someone on the opposing side of an issue. This may prevent you from being immediately dismissed as a closed-minded zealot. The difference between the is when the copyright is infringed and the music is sold for a profit, that is stealing or theft of profits on the copyright through infringement. When the music is illegally downloaded by an individual it is copyright infringement, not theft, since it is difficult to say whether that one individual would have paid for the music otherwise.
IANAL, so I may be completely wrong, but that is my understanding of the issue. There's a huge difference between the college student who downloads a couple songs here and there and shares with friends, all 100,000,000 of them, and a company which is distributing the copyrighted material for a profit. Even a broken clock is right twice a day.
They use them at the edge of the parking lots, like an electric fence, to keep homeless people from stealing the carts. It reduces the amount of shopping carts they need to replace each year and saves the store money.
This is for emergency response to strokes, heart attack, etc to prevent healthy cells from being destroyed by the enzymes which cause apoptosis. It would follow common sense that the inhibitor would be out of the system within a day or so after treatment with the drug is ended.
In conclusion, RTFA.
And I thought my friend who works as a vulnerability researcher was overly paranoid when he had a browser profile only for bank transactions.
This class of attack has been known for years, its only recently that its come into the spotlight, and been named and quantified. Or maybe its just a slashvertisement since CheckPoint is the first to "Patch" it. When everyone was vulnerable, why make the sheeple scared, scared users don't make online transactions. Now its "Fixed," but only the people at CheckPoint have fixed it.
I wonder how many other design level vulnerabilities there are which don't get press because no one has a solution they can market yet.
I've seen both good and bad managers in that regard. Currently my only manager is the head of engineering and has been at this company for over a decade. He has a vested interest in seeing the company succeed in the long term. This is what I would consider a sane environment for software development. Long term sustainability of products is a priority here. Since I work in the embedded world, once deployed, the next upgrade may be 5-6 years down the line. While I realize this is not typical, rational people making decisions based on long term effects, its my preferred environment to work in.
However, at places which have not learned the most fundamental rule of the industry, I make a point of ridiculing anyone who says "We'll do it right later" in a meeting. Publicly shaming people in front of their peers for trying to take a short cut, which compromises quality, future profitability, and actually takes longer in the long run is FUN! In the real world there is no later. Once it is working and released, there is no reason to go back and fix it. The risk involved in that effort, introducing new bugs, and the lack of rewards, it already works, make it impossible to justify touching the code before the next feature is needed or bug is identified. Luckily I'm skilled enough that I can refuse to do things in a way which is known to be wrong, and if it comes down to it, I can leave and easily find a job for an employer thats tired of lazy short-sighted developers or managers sabotaging their future efforts.
Tying back into the original discussion, open source projects need to start rejecting undocumented code. This would allow a greater range of skilled developers, who do not have months of time to ramp up Full Time on their project, to fix non-trivial bugs. It also makes the code more transparent for a security audit by source review. Sure there are automated tools which can catch trivial problems, but there's no substitute to having a skilled professional review the code looking for design flaws as opposed to coding errors. The automated tools only look for certain classes of exploits like buffer overflows and priviledge escalation. It will not stop someone, say funded by organized crime, from spending months looking for said design flaw to find a way to compromise the host machine.
The fundamental difference between many closed source projects and open source projects is the mentality of the developers. In the open source world no one is forcing them to do code reviews. There are not any requirements as to the process which should be used in development and testing. While many projects have these standards, lumping in some newbie grad student's proof of concept program, which got extended in a couple people's free time, with projects like the Linux Kernel which have rigorous standards and process is not creating a desirable picture for quality projects. Then comparing the aggregate of many projects from a wide variety of developers using just as many standards, to that of a single corporation with a single set of standards is comparing apples and oranges.
The reason people have been comparing Linux, in the broader sense, to Microsoft and saying "look at how bad MS is doing," is because for the amount of revenue Microsoft has generated over the years, they have no excuse for problems that they have had in regard to security over the last few years. I'm not saying this to belittle linux or imply that Linux is a hobby operating system. But to say that every linux compatible product is developed using professional standards is incorrect. The fact that a system which is comprised of both professional and amateur free software is even in the same league let alone surpassing a corporation with Microsoft's market cap, is very bad reflection on Microsoft.
To blame Microsoft's failings on its size is doing it an injustice. The blame lies in it's corporate culture, lack of flexibility, bad communication between teams, etc... These are all things that a company like Microsoft sh
How many times do you think a manager needs to get burned by having unmaintainable code before they wise up and budget the time?
Sustainability should be brought up during the review process when the code is first baselined, otherwise it should never be allowed in source control. Doing things right the first time saves time later, no matter whether its the original developer or some poor schmuck they pulled off the street. Once you put away the code for Product X to start working on Product Y for 2 years, and you come back to Product X, if your code is not maintainable even the original developer will waste weeks or months ramping back up.
I believe the important distinction here is that the consoles were modded and still able to execute signed code, copied games. When you modify a console to run unsigned code it is not a DMCA violation since you are not circumventing copyright protection.
I'm unfamiliar with the XBOX 360 modding specifics, but with the original XBOX there were 2 types of BIOS that could be obtained. The first is a completely legal linux-loader BIOS, the popular one being the Cromwell BIOS. The second being a not exactly legal 3rd party BIOS obtained from the usual places (if you don't know, I'm not going to tell you), which can execute both signed and unsigned code. This can be quite handy for legally(?) backing up games you have licenses for on your hard drive and running programs like XBMC. IANAL, so I don't know if backing up a game you own constitutes circumventing a copyright, so YMMV.
This guy was distributing pirated versions of games, backed up games without having a license (game disc), and the means to use them. All of which is highly illegal. Its a shame that someone bright enough to pull it off was not bright enough to know better than to advertise it publically.
While you may need more programmers, you desperately need a technical writer. Reading your project main page made my brain hurt. Its okay to end a sentence. One thing you may want to try is reading it aloud. If you need another breath before you finish the sentence, then the sentence is too probably too long.
You may want to check out Gentoo Catalyst. Before I get flamed about having to watch code compile etc... Let me just say that this tool allows you to completely customize your install stages, and you can make a stage 3 installer which only uses binary packages.
For a large scale roll out this tool is ideal. It is not easy or idiot proof on the admin side, but once you make your custom ISOs, it will be painless on the user side.
The goal of the catalyst project is to provide a single multi-faceted tool that can reliably build all aspects of a Gentoo Linux release: stage tarballs, GRP package sets, and install CDs.
Our specific development goals for catalyst include the following: ensuring it provides high-quality builds of Gentoo Linux, and for the tool to be easy to use, customize, extend and maintain. The catalyst tool is intended to be used by those who wish to create their own customized versions of Gentoo Linux, or their own customized LiveCDs. Our goal is to make catalyst a powerful tool that's a pleasure to use, and to ensure that the code we write is maintainable and of high-quality.
Maybe you just need to look harder for the message. Go watch "The Number 23" and tell me if you haven't been trying to contact yourself from the future.
Most embedded systems run some sort of ARM variant
You're forgetting embedded PowerPC chips, and entire families of 8 bit and 16 bit chips which are still in widespread use throughout the embedded world. You think it takes a 32 bit chip to drive a discman? What about dumb cellphones, I can garuantee that almost none of those would use a 32 bit chip, the almost is because invariably some products end up over-engineered.
The word PC is vastly over-used since it used to mean IBM PC Compatible (x86). This ultra mobile ITX board may be a mini computer on a small board, but the fact is that custom motherboards are made for many devices. With this you don't even get to choose what peripherals you get with the system, its common to have many peripherals over an I2C bus, SPI bus, or directly wired to pins for System on a Chip archs. And for the record linux will NOT run on many of these devices, mostly due to memory considerations, but as a form factor embedded motherboard with low power dissipation and many power saving modes you will find that they are MANY MANY more options than just ARM, x86, or PPC.
Most serious embedded developers wouldn't be looking at these for small tasks, nor for a sub-notebook. This is a handheld board, and could potentially be used in larger devices, but at that point you might as well get the cheaper picoITX board. This is a product with a fairly limited niche, but its definitely cool for what it is, if thats what you need.
That being said its been a couple years since I worked on an architecture with more than 1 MB of RAM so YMMV.
I meant symbol table in regards to the compiler's parsing of the code using the grammar and representing the various expressions in a tree format. The exact terms escape me at the moment.
We use C mainly because the C++ compilers available for the architecture suck when compared to the C compilers for code size and speed. And when you only have 2K RAM and 32K ROM, efficient compilers are very important.
I've seen C++ used on project for embedded x86, but that's the only time I've seen it used. And for Embedded x86 they will go ahead and make the jump to C#.NET if its a Windows platform, just because good C and Assembly programmers are scarce these days and.NET is managed which can prevent a lot of problems if you have the resources in the device to run it. It also causes lots of problems, but thats because its MS. As an added bonus, if you work for a Fortune 100 company with Enterprise Support for XP Embedded you get to yell at the developer over the phone who "Designed it not to work that way" over at MS ^_^
Does the use of const int vs #define compile any differently? I work mostly with small embedded projects where every byte counts, though I've always trusted the compilers to handle the optimization and the standard seems to be #defines rather than const int across several companies I've worked at. Though IIRC #defines are resolved by the precompiler rather than the compiler which may result in smaller executable size since there would be less overall symbols in the symbol table. Since it may be that all constants with a value of 5 are resolved to the same one byte symbol, while if you have several constant int's with a value of 5 they each get their own word in the symbol table, depending on how smart your compiler is. That could be a legacy habit carried forward from older compilers though.
The fact you can even use assert()'s means you work in a vastly different arena than myself, I'm lucky if I get more than 3 breakpoints at a time. So I do input sanitization similar to described in the article.
Excellent analysis of the current situation regarding legality of the code. You left out that the XBOX lacks the processing power to handle 720p quality video (Xvid compressed), even though it is capable of output at that resolution. This among other things which have been accumulating just go to demonstrate that XBMC has outgrown the XBOX making a port desirable to best utilize the effort spent so far in development. A development effort which has been on going since the turn of the century. Waste not, Want not.
Slashdot Mirror
Actually I interned for 3 years doing embedded avionics for things like fighter jets, the space shuttle, and commercial airliners. After finishing school I got laid off, since I was working there part time my senior year on the Electronic Flight Bag project for Boeing. Then I started contract work, quit that when the network security startup came around, we sold the company, and I went back to embedded work as a contractor on the A380 super jumbo for airbus. Then I quit that job because the of poor planning, too many hours, and no respite in sight after a year of working my ass off. Now I've got a full-time gig doing maritime control systems, at a base just under 6 figures with benefits pushing it over. ;)
My current job is awesome, 39-40 hr weeks, low stress, very manageable projects and time for grad school in the evenings after I establish residency here for in-state tuition. Oh and my boss wants to move towards embedded code generation using modeling tools like MATLAB & Simulink. This all makes for a very happy programmer =)
I call H1-BS.
I wonder if the democrat controlled congress has anything to do with this? I know Microsoft and others are constantly lobbying to import more slaves^H^H^H^H^H^H H1-B visa workers to get around paying the wages necessary for skilled domestic labor.
I think you missed the part about not having room for an interpreter in the micros I work with. Some have as little as 32K ROM and 2K of RAM. When working with some 8 bit chips the resources are even more limited and its stops making sense to use C Code. Saying that is not the case is ignoring industry, you know anything in the controls/automation field. I consider Python a real language, I do know that it can be compiled. I have no interest in learning web programming, I simply do not have time to focus in that direction. I've worked with many scripting languages. I was not implying that web programming is not a real discipline in CS, just that calling those of us which work with tools that generate machine code dinosaurs is arrogant. PS - I made low 6 figures last year, some from stock options from a successful startup, most from a healthy hourly rate doing contract work with a bit of overtime.
I resent that remark. I'm an embedded developer, I work a lot with microcontrollers, usually with less than a MB of RAM and Flash. My skills are quite rare in the market these days because of attitudes like that. I'm not going to brag about my salary, but since my current employer spent 4 yrs trying to fill my position, its quite good.
I'm not a 50 yr old dinosaur either, I'm 26. I happen to use a lot of python in my day to day work to do little side tasks faster. Languages are just tools, however the only tools I can use for the firmware I write involve C code and Assembly. However I will be generating code for the next project using MATLab and Simulink, but if I did not know how to work with the output, all those ultra advanced tools would be useless.
It is not yet a treatment, only experimental compound used on lab animals. As is, if the compound were modified and approved for use on humans it would not be immediately useful against Parkinson's. Eventually what is learned by further studying the processes involved the article asserts it may one day lead to treatments for Parkinson's.
Nothing to see here, move along.
I am not a bullet point. I think the word you were looking for was non-sequitur. *I am a spelling nazi though*
The point I made was not that copyright infringement is theft, just that the parent poster completely missed the issue and posted the same thing that gets posted in every one of these stories. We get it, copyright infringement is not theft. Unfortunately had the poster possibly RTFA he would have realized that profiting from a copyrighted work and not paying the licensing fees for distribution is Stealing/Theft/Robbery whatever you want to call it. You've now taken money from consumers and stolen the RIAA's share of it. This is not about the music, its about the money.
I know the copyright infringement != theft line is common around here, but please respond intelligently to issues raised by someone on the opposing side of an issue. This may prevent you from being immediately dismissed as a closed-minded zealot. The difference between the is when the copyright is infringed and the music is sold for a profit, that is stealing or theft of profits on the copyright through infringement. When the music is illegally downloaded by an individual it is copyright infringement, not theft, since it is difficult to say whether that one individual would have paid for the music otherwise.
IANAL, so I may be completely wrong, but that is my understanding of the issue. There's a huge difference between the college student who downloads a couple songs here and there and shares with friends, all 100,000,000 of them, and a company which is distributing the copyrighted material for a profit. Even a broken clock is right twice a day.
They use them at the edge of the parking lots, like an electric fence, to keep homeless people from stealing the carts. It reduces the amount of shopping carts they need to replace each year and saves the store money.
This is for emergency response to strokes, heart attack, etc to prevent healthy cells from being destroyed by the enzymes which cause apoptosis. It would follow common sense that the inhibitor would be out of the system within a day or so after treatment with the drug is ended. In conclusion, RTFA.
And I thought my friend who works as a vulnerability researcher was overly paranoid when he had a browser profile only for bank transactions.
This class of attack has been known for years, its only recently that its come into the spotlight, and been named and quantified. Or maybe its just a slashvertisement since CheckPoint is the first to "Patch" it. When everyone was vulnerable, why make the sheeple scared, scared users don't make online transactions. Now its "Fixed," but only the people at CheckPoint have fixed it.
I wonder how many other design level vulnerabilities there are which don't get press because no one has a solution they can market yet.
I've seen both good and bad managers in that regard. Currently my only manager is the head of engineering and has been at this company for over a decade. He has a vested interest in seeing the company succeed in the long term. This is what I would consider a sane environment for software development. Long term sustainability of products is a priority here. Since I work in the embedded world, once deployed, the next upgrade may be 5-6 years down the line. While I realize this is not typical, rational people making decisions based on long term effects, its my preferred environment to work in.
However, at places which have not learned the most fundamental rule of the industry, I make a point of ridiculing anyone who says "We'll do it right later" in a meeting. Publicly shaming people in front of their peers for trying to take a short cut, which compromises quality, future profitability, and actually takes longer in the long run is FUN! In the real world there is no later. Once it is working and released, there is no reason to go back and fix it. The risk involved in that effort, introducing new bugs, and the lack of rewards, it already works, make it impossible to justify touching the code before the next feature is needed or bug is identified. Luckily I'm skilled enough that I can refuse to do things in a way which is known to be wrong, and if it comes down to it, I can leave and easily find a job for an employer thats tired of lazy short-sighted developers or managers sabotaging their future efforts.
Tying back into the original discussion, open source projects need to start rejecting undocumented code. This would allow a greater range of skilled developers, who do not have months of time to ramp up Full Time on their project, to fix non-trivial bugs. It also makes the code more transparent for a security audit by source review. Sure there are automated tools which can catch trivial problems, but there's no substitute to having a skilled professional review the code looking for design flaws as opposed to coding errors. The automated tools only look for certain classes of exploits like buffer overflows and priviledge escalation. It will not stop someone, say funded by organized crime, from spending months looking for said design flaw to find a way to compromise the host machine.
The fundamental difference between many closed source projects and open source projects is the mentality of the developers. In the open source world no one is forcing them to do code reviews. There are not any requirements as to the process which should be used in development and testing. While many projects have these standards, lumping in some newbie grad student's proof of concept program, which got extended in a couple people's free time, with projects like the Linux Kernel which have rigorous standards and process is not creating a desirable picture for quality projects. Then comparing the aggregate of many projects from a wide variety of developers using just as many standards, to that of a single corporation with a single set of standards is comparing apples and oranges.
The reason people have been comparing Linux, in the broader sense, to Microsoft and saying "look at how bad MS is doing," is because for the amount of revenue Microsoft has generated over the years, they have no excuse for problems that they have had in regard to security over the last few years. I'm not saying this to belittle linux or imply that Linux is a hobby operating system. But to say that every linux compatible product is developed using professional standards is incorrect. The fact that a system which is comprised of both professional and amateur free software is even in the same league let alone surpassing a corporation with Microsoft's market cap, is very bad reflection on Microsoft.
To blame Microsoft's failings on its size is doing it an injustice. The blame lies in it's corporate culture, lack of flexibility, bad communication between teams, etc... These are all things that a company like Microsoft sh
How many times do you think a manager needs to get burned by having unmaintainable code before they wise up and budget the time?
Sustainability should be brought up during the review process when the code is first baselined, otherwise it should never be allowed in source control. Doing things right the first time saves time later, no matter whether its the original developer or some poor schmuck they pulled off the street. Once you put away the code for Product X to start working on Product Y for 2 years, and you come back to Product X, if your code is not maintainable even the original developer will waste weeks or months ramping back up.
If by "hard to find" you mean won't be in stores for two weeks, I'm right there with you ;)
I believe the important distinction here is that the consoles were modded and still able to execute signed code, copied games. When you modify a console to run unsigned code it is not a DMCA violation since you are not circumventing copyright protection.
I'm unfamiliar with the XBOX 360 modding specifics, but with the original XBOX there were 2 types of BIOS that could be obtained. The first is a completely legal linux-loader BIOS, the popular one being the Cromwell BIOS. The second being a not exactly legal 3rd party BIOS obtained from the usual places (if you don't know, I'm not going to tell you), which can execute both signed and unsigned code. This can be quite handy for legally(?) backing up games you have licenses for on your hard drive and running programs like XBMC. IANAL, so I don't know if backing up a game you own constitutes circumventing a copyright, so YMMV.
This guy was distributing pirated versions of games, backed up games without having a license (game disc), and the means to use them. All of which is highly illegal. Its a shame that someone bright enough to pull it off was not bright enough to know better than to advertise it publically.
While you may need more programmers, you desperately need a technical writer. Reading your project main page made my brain hurt. Its okay to end a sentence. One thing you may want to try is reading it aloud. If you need another breath before you finish the sentence, then the sentence is too probably too long.
For a large scale roll out this tool is ideal. It is not easy or idiot proof on the admin side, but once you make your custom ISOs, it will be painless on the user side.
From their site: http://www.gentoo.org/proj/en/releng/catalyst/
Let me be the first to combine those patents into breathing while walking.
Maybe you just need to look harder for the message. Go watch "The Number 23" and tell me if you haven't been trying to contact yourself from the future.
The word PC is vastly over-used since it used to mean IBM PC Compatible (x86). This ultra mobile ITX board may be a mini computer on a small board, but the fact is that custom motherboards are made for many devices. With this you don't even get to choose what peripherals you get with the system, its common to have many peripherals over an I2C bus, SPI bus, or directly wired to pins for System on a Chip archs. And for the record linux will NOT run on many of these devices, mostly due to memory considerations, but as a form factor embedded motherboard with low power dissipation and many power saving modes you will find that they are MANY MANY more options than just ARM, x86, or PPC.
Most serious embedded developers wouldn't be looking at these for small tasks, nor for a sub-notebook. This is a handheld board, and could potentially be used in larger devices, but at that point you might as well get the cheaper picoITX board. This is a product with a fairly limited niche, but its definitely cool for what it is, if thats what you need.
That being said its been a couple years since I worked on an architecture with more than 1 MB of RAM so YMMV.
The real question is: "Does it run Beryl?"
I meant symbol table in regards to the compiler's parsing of the code using the grammar and representing the various expressions in a tree format. The exact terms escape me at the moment.
.NET is managed which can prevent a lot of problems if you have the resources in the device to run it. It also causes lots of problems, but thats because its MS. As an added bonus, if you work for a Fortune 100 company with Enterprise Support for XP Embedded you get to yell at the developer over the phone who "Designed it not to work that way" over at MS ^_^
We use C mainly because the C++ compilers available for the architecture suck when compared to the C compilers for code size and speed. And when you only have 2K RAM and 32K ROM, efficient compilers are very important.
I've seen C++ used on project for embedded x86, but that's the only time I've seen it used. And for Embedded x86 they will go ahead and make the jump to C#.NET if its a Windows platform, just because good C and Assembly programmers are scarce these days and
Does the use of const int vs #define compile any differently? I work mostly with small embedded projects where every byte counts, though I've always trusted the compilers to handle the optimization and the standard seems to be #defines rather than const int across several companies I've worked at. Though IIRC #defines are resolved by the precompiler rather than the compiler which may result in smaller executable size since there would be less overall symbols in the symbol table. Since it may be that all constants with a value of 5 are resolved to the same one byte symbol, while if you have several constant int's with a value of 5 they each get their own word in the symbol table, depending on how smart your compiler is. That could be a legacy habit carried forward from older compilers though.
The fact you can even use assert()'s means you work in a vastly different arena than myself, I'm lucky if I get more than 3 breakpoints at a time. So I do input sanitization similar to described in the article.
Excellent analysis of the current situation regarding legality of the code. You left out that the XBOX lacks the processing power to handle 720p quality video (Xvid compressed), even though it is capable of output at that resolution. This among other things which have been accumulating just go to demonstrate that XBMC has outgrown the XBOX making a port desirable to best utilize the effort spent so far in development. A development effort which has been on going since the turn of the century. Waste not, Want not.