I've read one account of a person who regularly gargled with that stuff, until one day he cracked a tooth...
Having said that, I worked in a place where liquid nitrogen (LN2) was used to cool the drinks during parties (it was dumped into a large volume of water which contained the bottles; the drinks would freeze if you put the bottles directly into the nitrogen).
Actually it was kinda cool: I was working in an office almost directly below three huge LN2 tanks, containing about 160,000 liters of the stuff. You remember that scene from James Bond where the programmer screams "I'm *invincible*" before being hit by a wave of LN2 and shattering? I always feared that would happen to me if I ever made a deadline;-)
Anyway, that was my excuse for being chronically late;-)
do you need rom revision of 3 or above to run amiga os 3 or above?
Yes. Most of what constitutes AmigaOS 3.0 is in those roms, and the diskloaded parts require it. As you say, it is possible to load ROMs from disk using a bit of trickery.
While you can decode JPEG compression, it's butt slow on a 68030
You hear no argument from me there;-) What I meant was that the software implementation is simple, not that the algorithm runs quickly.
I honestly don't know if 24bit graphics cards required OS 3.0.... I did just fine with Amiga OS 1.3 and 2.x..
Perhaps you are right - it's been a while since I actually used an Amiga. Unfortunately...
We were a department of five, providing vital services for a company of about 40, which in turn was providing services to a much larger company. The (smaller) company was fucking us over in all directions, and finally enough was enough.
It is important to realize that we were effectively irreplacable (unique job-specific skills, nothing to do with computers). Or so we thought...
Three of us (not including myself) went ahead and set up a company, and offered our services to the larger company directly. The smaller company then started on a campaign of threats, allegations, lies, and FUD that would make Microsoft blush. The larger company used us as a lever for negotiating a better contract with the smaller company, then unceremoniously dumped us.
So would I do it again? Hell, yes. In fact I would do it sooner, and with less restraint. This is important to realize: if we had realized what was coming we would have been less galant towards our former boss (not keeping the systems going while we were setting up our new company, for example - the price would have been high, but it would also have put tremendous pressure on our boss). And we wouldn't have believed the (verbal) assurances the larger company gave us regarding our soon-to-be contract with them.
The story is far more complicated than this little message (I could write a book about that period), but the general idea I think is clear: we were in a bad situation, we fought, we lost, and we have no regrets.
Some lessons you may want to remember:
- Your former colleagues may suddenly turn into your worst enemies. They'll lie to you. They'll try to make you fail in all ways that count. And they may pretend they are still your friend while they are at it.
- Individual members of your group may be bribed by your former boss to come back into the fold, thereby bringing back all that irreplacable knowledge.
Are you ready to fight? Can you afford to lose? If so, go for it.
When Amiga/PPC hardware started appearing, I was keen to do some portage of unix-ish type stuff - except the PPC dev toolchain was so woeful it made me want to cry.
I remember that. I bought a PPC board with the specific intention to do some cool ports, but it just couldn't be done. The PPC stuff just never worked, and it was plain impossible to find out what was broken: the hardware itself, the communication software, the PPC kernel, the PPC compiler,... And it wasn't just my own software: there has not been a single piece of software that ran stable, everything would crash sooner or later.
A major problem was that any IO had to go through the 68K, and context switches were hideously slow. Thus the utility of the PPC was severely limited to specific (long-running) tasks. Quickly building some PPC support into existing applications was impossible.
And there were two kernels for it (the "official" PPC kernel and another one which was slightly different), which could not run at the same time and which were not compatible...
The whole PPC thing stank from day one. Too bad, it could have been so good...
Back then I was using AWeb, in 24-bit color, using a Picasso IV graphics card. There were never any issues with images being dithered or anything - it just worked in 24-bits.
Most of the graphics cards require at least OS 3.0. I'd guess that most (both?;-) ) remaining Amiga users are at least using that version.
As for JPEG decoding, that's a trivial exercise: just use the excellent datatypes system and load any image format as if it were native. There are plenty JPEG datatypes (PPC, 68060, no math copro, etc.) that will handle the decoding for you.
Like you I was and still am a big fan of the Amiga. It was a clean, elegant, responsive, *pleasant* system, and I miss it dearly. No doubt some people here on/. will rant about it being obsolete, but there has never been a finer OS.
AmigaOS has absolutely nothing to do with UNIX (and the mere fact that it multitasks doesn't change that one bit).
Furthermore, while I appreciate the humor in "cooperative memory protection", it doesn't really mean much now does it?;-)
As for how you should do it, that depends on how Mozilla talks to the underlying OS. Does it have a decent abstraction layer? If so, a port may be feasible. Mind, I wouldn't want to tangle with BOOPSI ("Basic Object Oriented Programming System for Intuition") again, with its weird "you can clip anytime except when painting" rule...
Another question is if it uses fork() for anything more than loading external processes. AmigaOS does not and cannot support this call (although obviously it has its own call for starting programs).
(1) Schedule realistically, so that tasks can be completed without overtime. This may mean some things just cannot be done in the desired time period. Learn to accept that.
(2) Hire and retain sufficient staff, so that the work can be shared between multiple people. This may mean that some of the time the company will be overstaffed. Accept that too.
Obviously both these suggestions come with a pricetag, but lost missions aren't free either...
I work in a company doing this sort of thing. I regularly see requests for people that have knowledge of electronics, antena's, communication protocols, and software. Usually it is required to be a european citizen or have a work permit.
Out of curiosity, have you ever seen the libpng code and API? They do not really inspire confidence - sure, the person who can juggle all those #defines must be really smart, but I'd rather he write code that people can actually understand.
It sounds like you either have a delusionally high estimation of your own abilities relative to those of your library-writing peers
I know what my own skills are worth, but I know nothing about the person who wrote the library. He may be an absolute code guru, or he may be a student who wrote this for a class-project about platform-independency. Unless I spend a significant amount of time analysing the code I won't know - and I do that I might as well write it myself...
I think you need libpng and libz (instead of libsdl). One risk is that libpng _will_ abort if you feed it a malformed png. Doing the 8-byte header check seems to decrease the risk of this actually happening, but I'm not sure if it eliminates the possibility. The code is simply too cluttered for me to analyse, thanks to an absolute overload on #defines.
As for documentation: like so much UNIX documentation it goes into great detail about weird, abnormal cases, but completely ignores the common case where you have a PNG and want a bitmap representation of it without any fuss.
For what its worth, in my case I ended up reusing the library - I considered writing my own both too risky and too much work.
Why exactly does Nintendo believe something illegal is going on? There are no laws that restrict the citizens of Europe from buying in Japan, nor are there any laws that restrict Japanese shops from selling to European clients. If I wanted to order a dozen copies of a Japanese game there is absolutely no problem as far as the law is concerned. Moreover, I have the right to resell those copies in the local market.
Nintendo, on the other hand, is trying to control the distribution channel. I suspect that might just be illegal; it certainly sounds rather monopolistic. Maybe the commission should look into this again, although with their glacial speed we could expect a decision no sooner than 2015.
- Security around spacecraft security is very tight, especially on a launch site. It is hard to get to them, and they are inspected constantly (not for stowaways, but one would be detected quickly enough).
- Spacecraft tend to be rather small, and filled with equipment. Certainly Mars Express (the european spacecraft) is far too small to contain a human being. I have not seen the american spacecraft but I'd guess they are not much bigger.
- The weight of the spacecraft is known with high accuracy, and verified before launch.
- The center of gravity of the spacecraft is known. Changing it (by tagging on extra weight) will cause maneuvring to fail, sending the thing to the wrong location.
- The trip takes a significantly long time (many months). You'd be long dead by the time you arrived (from lack of oxygen, radiation, etc.).
- The launch may very well kill you: not every launcher is human-rated, and some produce vibrations strong enough to kill a human passenger.
Finally, I don't want to discourage you but as far as I know noone has ever been able to make a picture of one of the moon landing sites proving there was something there. Your grave would most likely suffer a similar fate.
In my experience, good engineers do tend to look at possible bad outcomes. Its the management that shuffles their views under the carpet.
I've had plenty of these discussions myself: I see a _potential_ problem and report it, and some dimwit asks me if it is likely to happen (which is the wrong question - it should be, "can we afford the consequences if it happens?"). My answer is usually "it is not likely, but *if* it happens we are in deep shit". At this point the manager in question has to make a very tough choice, based on limited information and knowledge: should a solution be developed (possibly at great cost), or should the problem be ignored (also possibly at great cost)? I don't envy the people making these decisions, and the record shows they get it wrong on a regular basis.
As far as I can see, in both shuttle accidents, the fatal problem was reported by engineers and subsequently ignored by management.
As a purely theoretical question, ask yourself this: how much money would you be willing to spend to verify that the shuttle, while it was in orbit, was undamaged? Because that is what this comes down to: risk was balanced against budget and budget won.
And on a similar note, I'm curious to learn how many potential problems are reported for each shuttle flight, and how many of these actually turn into real problems. I wouldn't be surprised if that ratio were actually quite high.
Reusables basically provide a service that is not much in demand: returning goods from orbit down to Earth. If you take that away you are left with a very simple requirement: returning astronauts. There is a good, tried and tested way of doing that: capsules. A capsule is cheap to build and cheap to launch.
When you have capsules, you do not need to to launch wings, control mechanisms, and all the other bits that make up an airplane into orbit. That saves a huge amount of weight. The saved weight can be "spent" in two ways: cheaper rockets and larger payloads.
The cost advantage of cheaper rockets (i.e. Soyuz) over a shuttle is obvious (something like a factor 20 or so).
The cost advantage of larger payloads is also obvious: all of your energy is used lifting stuff that actually needs to be in space, rather than used to return safely to Earth. A shuttle launch assembly minus the shuttle (instead imagine a huge cargo container in its place) would have phenomenal lifting power; in fact, the Russian equivalent of this, Energia, could have launched the entire weight of ISS in just three launches! That would have saved a few hundred billion dollars.
So stop thinking of the shuttle as a "cool space plane", and instead consider it to be a highly over-engineered, *heavy* method of returning to Earth.
So what is the problem with this? I guess it has to do with NASA being afraid to lose face, which seems inevitable when they give up on the shuttle and return to expendable launchers.
Here's what I believe the various space agencies should do to replace the shuttle:
- Develop an expendable launcher that can lift *really heavy* items into a useful orbit (which can be ISS orbit, geostationary, or some escape orbit). Russia's Energia would be a good starting point, as would the shuttle launch stack (they are related anyway). This would be the heavy workhorse for orbital construction.
- Develop an expendable launcher that can lift people to that same orbit. Put a capsule on it in which people can travel comfortably and safely during the entire trip - i.e. it is fine to overengineer a bit, since it will one day safe lives. Use that just for people.
- Want to do something in orbit? Put your equipment on a big launcher (together with a lot of other stuff), and put your people on a small, safe launcher. The big launcher delivers to ISS (or some other orbital facility - constructing them will be cheap with this model), where the resident astronauts can install and operate it.
Sorry, but criticism does not equal hatred. I offer criticism, but not hatred. For that matter, so did the French government when it rejected the war in Iraq: no hatred, just criticism.
As for France and Austria, they were extremely foolish (in the case of France) and irresponsible (in the case of Austria) to elect Le Pen (almost) and Haider. And that too is criticism, rather than hatred.
Many of us _hate_ you? Whatever gave you that idea?
It is certainly true many Europeans are not major fans of your current leadership, but frankly most of us hope you will eventually regain your senses and vote someone more sensible into the white house.
I'm disturbed that you felt you needed to post this. I didn't realize American perception of Europe can be so wrong. What do you think of when someone mentions Europe?
Aside from his poor taste in word choice, the author makes the flawed fundamental assumption that all works will employ DRM in the strictest terms. This, I'm quite sure, will not be the case.
Have you seen any evidence of restraint among corporations when it comes to protecting their "property"?
Crimes against humanity? So, we have some kind of inalienable right to listen to the latest Britney Spears blather, and DRM infringes upon this right?
Strange as this may sound now, *yes*. I'll be the first to admit that losing Britney is not a particularly great loss, but in general we are defined by our culture. It is something we leave behind for later generations. Why is jazz, for example, considered a worthwhile legacy from an earlier age? What would we know about it now if it had all been allowed to die under a technological restriction scheme?
The point I'm trying to make is this: if we allow DRM to be implemented, *nothing* from our age will remain. Are you willing to say that nothing that has been produced in our lifetime is worth keeping? That our (and for our, maybe you should read "western") culture should effectively stop around the year 2000, with no records being kept after that point (other than a few moneymakers the big corporations are willing to sell to us)?
Put in this terms, I believe the phrase "crime against humanity" is justified.
I have been doing my own little battle with Agilent for several years. I want to have a Linux driver for their GPIB cards, but they are utterly unwilling to provide it.
Recently this went so far that I had a very candid talk with one of their sales people. I made it clear I would move to a different manufacturer if they would not provide drivers or some means for me to write them. He would ask within their organisation. Several days later I received an email titled "Solution to your GPIB driver problem". To my astonishment, it gave contact details for several other manufacturers of GPIB cards!
So, isn't this weird? They'd rather lose our custom than provide us with a driver, or sufficient details to write our own! And given the amount of stuff we are buying from them that's a pretty big decision.
So now we are talking with National Instruments, and they are very pleased to have us as customers. Moreover, they have Linux drivers for their GPIB cards. I haven't seen these drivers yet, and I sure hope they will be of acceptable quality.
I'm interested in hearing what experiences other people here have had with GPIB cards under Linux - especially those of NI, which we are about to buy. When you mention lousy Linux support, does that include their GPIB cards?
You are correct. ESA has made it a hard requirement that no american-sourced parts will find its way into Galileo. That's because they come with too many strings attached; limitations on who can see the spacecraft, who can launch them, where they can go in orbit, etc.
Your understanding of the consequences of the halting problem is incomplete. It is not a proof that it is impossible to determine of any given program whether or not it stops in finite time. It is merely a proof that there exists a class of programs for this determination cannot be made. However, there are also many programs for which it can easily be determined whether or not it stops in finite time, and the same thing is true for decompilation.
Furthermore, there is nothing saying that it has to do a 100% perfect job. Decompilation is already accepted to be imprecise; using some common sense (intuition if you want) to fill in some gaps is not an invalid method.
The problems that face decompilation that stem from real-world issues are far, far greater than this (rather theoretical) problem. For example, decompiling any STL-based source to a useful state will be far more difficult than a simple jumptable.
Slashdot Mirror
Having said that, I worked in a place where liquid nitrogen (LN2) was used to cool the drinks during parties (it was dumped into a large volume of water which contained the bottles; the drinks would freeze if you put the bottles directly into the nitrogen).
Actually it was kinda cool: I was working in an office almost directly below three huge LN2 tanks, containing about 160,000 liters of the stuff. You remember that scene from James Bond where the programmer screams "I'm *invincible*" before being hit by a wave of LN2 and shattering? I always feared that would happen to me if I ever made a deadline ;-)
Anyway, that was my excuse for being chronically late ;-)
Creating jobs is massively important to winning elections, so governments would be foolish to contract software development out to cheaper countries.
Looks like a bunch of vaporware to me... ;-)
Thag the Warrior held the patent on civilisation, and his descendents kept extending the duration of the patent...
Yes. Most of what constitutes AmigaOS 3.0 is in those roms, and the diskloaded parts require it. As you say, it is possible to load ROMs from disk using a bit of trickery.
While you can decode JPEG compression, it's butt slow on a 68030
You hear no argument from me there ;-) What I meant was that the software implementation is simple, not that the algorithm runs quickly.
I honestly don't know if 24bit graphics cards required OS 3.0.... I did just fine with Amiga OS 1.3 and 2.x..
Perhaps you are right - it's been a while since I actually used an Amiga. Unfortunately...
It is important to realize that we were effectively irreplacable (unique job-specific skills, nothing to do with computers). Or so we thought...
Three of us (not including myself) went ahead and set up a company, and offered our services to the larger company directly. The smaller company then started on a campaign of threats, allegations, lies, and FUD that would make Microsoft blush. The larger company used us as a lever for negotiating a better contract with the smaller company, then unceremoniously dumped us.
So would I do it again? Hell, yes. In fact I would do it sooner, and with less restraint. This is important to realize: if we had realized what was coming we would have been less galant towards our former boss (not keeping the systems going while we were setting up our new company, for example - the price would have been high, but it would also have put tremendous pressure on our boss). And we wouldn't have believed the (verbal) assurances the larger company gave us regarding our soon-to-be contract with them.
The story is far more complicated than this little message (I could write a book about that period), but the general idea I think is clear: we were in a bad situation, we fought, we lost, and we have no regrets.
Some lessons you may want to remember:
- Your former colleagues may suddenly turn into your worst enemies. They'll lie to you. They'll try to make you fail in all ways that count. And they may pretend they are still your friend while they are at it.
- Individual members of your group may be bribed by your former boss to come back into the fold, thereby bringing back all that irreplacable knowledge.
Are you ready to fight? Can you afford to lose? If so, go for it.
I remember that. I bought a PPC board with the specific intention to do some cool ports, but it just couldn't be done. The PPC stuff just never worked, and it was plain impossible to find out what was broken: the hardware itself, the communication software, the PPC kernel, the PPC compiler, ... And it wasn't just my own software: there has not been a single piece of software that ran stable, everything would crash sooner or later.
A major problem was that any IO had to go through the 68K, and context switches were hideously slow. Thus the utility of the PPC was severely limited to specific (long-running) tasks. Quickly building some PPC support into existing applications was impossible.
And there were two kernels for it (the "official" PPC kernel and another one which was slightly different), which could not run at the same time and which were not compatible...
The whole PPC thing stank from day one. Too bad, it could have been so good...
Most of the graphics cards require at least OS 3.0. I'd guess that most (both? ;-) ) remaining Amiga users are at least using that version.
As for JPEG decoding, that's a trivial exercise: just use the excellent datatypes system and load any image format as if it were native. There are plenty JPEG datatypes (PPC, 68060, no math copro, etc.) that will handle the decoding for you.
Like you I was and still am a big fan of the Amiga. It was a clean, elegant, responsive, *pleasant* system, and I miss it dearly. No doubt some people here on /. will rant about it being obsolete, but there has never been a finer OS.
Furthermore, while I appreciate the humor in "cooperative memory protection", it doesn't really mean much now does it? ;-)
As for how you should do it, that depends on how Mozilla talks to the underlying OS. Does it have a decent abstraction layer? If so, a port may be feasible. Mind, I wouldn't want to tangle with BOOPSI ("Basic Object Oriented Programming System for Intuition") again, with its weird "you can clip anytime except when painting" rule...
Another question is if it uses fork() for anything more than loading external processes. AmigaOS does not and cannot support this call (although obviously it has its own call for starting programs).
(1) Schedule realistically, so that tasks can be completed without overtime. This may mean some things just cannot be done in the desired time period. Learn to accept that.
(2) Hire and retain sufficient staff, so that the work can be shared between multiple people. This may mean that some of the time the company will be overstaffed. Accept that too.
Obviously both these suggestions come with a pricetag, but lost missions aren't free either...
There will not be any american components in Galileo for precisely the reason you state. We are not entirely stupid here in Europe you know...
Check it out: Terma; click on jobs.
It sounds like you either have a delusionally high estimation of your own abilities relative to those of your library-writing peers
I know what my own skills are worth, but I know nothing about the person who wrote the library. He may be an absolute code guru, or he may be a student who wrote this for a class-project about platform-independency. Unless I spend a significant amount of time analysing the code I won't know - and I do that I might as well write it myself...
As for documentation: like so much UNIX documentation it goes into great detail about weird, abnormal cases, but completely ignores the common case where you have a PNG and want a bitmap representation of it without any fuss.
For what its worth, in my case I ended up reusing the library - I considered writing my own both too risky and too much work.
Nintendo, on the other hand, is trying to control the distribution channel. I suspect that might just be illegal; it certainly sounds rather monopolistic. Maybe the commission should look into this again, although with their glacial speed we could expect a decision no sooner than 2015.
- Security around spacecraft security is very tight, especially on a launch site. It is hard to get to them, and they are inspected constantly (not for stowaways, but one would be detected quickly enough).
- Spacecraft tend to be rather small, and filled with equipment. Certainly Mars Express (the european spacecraft) is far too small to contain a human being. I have not seen the american spacecraft but I'd guess they are not much bigger.
- The weight of the spacecraft is known with high accuracy, and verified before launch.
- The center of gravity of the spacecraft is known. Changing it (by tagging on extra weight) will cause maneuvring to fail, sending the thing to the wrong location.
- The trip takes a significantly long time (many months). You'd be long dead by the time you arrived (from lack of oxygen, radiation, etc.).
- The launch may very well kill you: not every launcher is human-rated, and some produce vibrations strong enough to kill a human passenger.
Finally, I don't want to discourage you but as far as I know noone has ever been able to make a picture of one of the moon landing sites proving there was something there. Your grave would most likely suffer a similar fate.
I've had plenty of these discussions myself: I see a _potential_ problem and report it, and some dimwit asks me if it is likely to happen (which is the wrong question - it should be, "can we afford the consequences if it happens?"). My answer is usually "it is not likely, but *if* it happens we are in deep shit". At this point the manager in question has to make a very tough choice, based on limited information and knowledge: should a solution be developed (possibly at great cost), or should the problem be ignored (also possibly at great cost)? I don't envy the people making these decisions, and the record shows they get it wrong on a regular basis.
As far as I can see, in both shuttle accidents, the fatal problem was reported by engineers and subsequently ignored by management.
As a purely theoretical question, ask yourself this: how much money would you be willing to spend to verify that the shuttle, while it was in orbit, was undamaged? Because that is what this comes down to: risk was balanced against budget and budget won.
And on a similar note, I'm curious to learn how many potential problems are reported for each shuttle flight, and how many of these actually turn into real problems. I wouldn't be surprised if that ratio were actually quite high.
When you have capsules, you do not need to to launch wings, control mechanisms, and all the other bits that make up an airplane into orbit. That saves a huge amount of weight. The saved weight can be "spent" in two ways: cheaper rockets and larger payloads.
The cost advantage of cheaper rockets (i.e. Soyuz) over a shuttle is obvious (something like a factor 20 or so).
The cost advantage of larger payloads is also obvious: all of your energy is used lifting stuff that actually needs to be in space, rather than used to return safely to Earth. A shuttle launch assembly minus the shuttle (instead imagine a huge cargo container in its place) would have phenomenal lifting power; in fact, the Russian equivalent of this, Energia, could have launched the entire weight of ISS in just three launches! That would have saved a few hundred billion dollars.
So stop thinking of the shuttle as a "cool space plane", and instead consider it to be a highly over-engineered, *heavy* method of returning to Earth.
So what is the problem with this? I guess it has to do with NASA being afraid to lose face, which seems inevitable when they give up on the shuttle and return to expendable launchers.
Here's what I believe the various space agencies should do to replace the shuttle:
- Develop an expendable launcher that can lift *really heavy* items into a useful orbit (which can be ISS orbit, geostationary, or some escape orbit). Russia's Energia would be a good starting point, as would the shuttle launch stack (they are related anyway). This would be the heavy workhorse for orbital construction.
- Develop an expendable launcher that can lift people to that same orbit. Put a capsule on it in which people can travel comfortably and safely during the entire trip - i.e. it is fine to overengineer a bit, since it will one day safe lives. Use that just for people.
- Want to do something in orbit? Put your equipment on a big launcher (together with a lot of other stuff), and put your people on a small, safe launcher. The big launcher delivers to ISS (or some other orbital facility - constructing them will be cheap with this model), where the resident astronauts can install and operate it.
As for France and Austria, they were extremely foolish (in the case of France) and irresponsible (in the case of Austria) to elect Le Pen (almost) and Haider. And that too is criticism, rather than hatred.
It is certainly true many Europeans are not major fans of your current leadership, but frankly most of us hope you will eventually regain your senses and vote someone more sensible into the white house.
I'm disturbed that you felt you needed to post this. I didn't realize American perception of Europe can be so wrong. What do you think of when someone mentions Europe?
It is not an american spacecraft. NASA is not involved.
Have you seen any evidence of restraint among corporations when it comes to protecting their "property"?
Crimes against humanity? So, we have some kind of inalienable right to listen to the latest Britney Spears blather, and DRM infringes upon this right?
Strange as this may sound now, *yes*. I'll be the first to admit that losing Britney is not a particularly great loss, but in general we are defined by our culture. It is something we leave behind for later generations. Why is jazz, for example, considered a worthwhile legacy from an earlier age? What would we know about it now if it had all been allowed to die under a technological restriction scheme?
The point I'm trying to make is this: if we allow DRM to be implemented, *nothing* from our age will remain. Are you willing to say that nothing that has been produced in our lifetime is worth keeping? That our (and for our, maybe you should read "western") culture should effectively stop around the year 2000, with no records being kept after that point (other than a few moneymakers the big corporations are willing to sell to us)?
Put in this terms, I believe the phrase "crime against humanity" is justified.
Recently this went so far that I had a very candid talk with one of their sales people. I made it clear I would move to a different manufacturer if they would not provide drivers or some means for me to write them. He would ask within their organisation. Several days later I received an email titled "Solution to your GPIB driver problem". To my astonishment, it gave contact details for several other manufacturers of GPIB cards!
So, isn't this weird? They'd rather lose our custom than provide us with a driver, or sufficient details to write our own! And given the amount of stuff we are buying from them that's a pretty big decision.
So now we are talking with National Instruments, and they are very pleased to have us as customers. Moreover, they have Linux drivers for their GPIB cards. I haven't seen these drivers yet, and I sure hope they will be of acceptable quality.
I'm interested in hearing what experiences other people here have had with GPIB cards under Linux - especially those of NI, which we are about to buy. When you mention lousy Linux support, does that include their GPIB cards?
You are correct. ESA has made it a hard requirement that no american-sourced parts will find its way into Galileo. That's because they come with too many strings attached; limitations on who can see the spacecraft, who can launch them, where they can go in orbit, etc.
Furthermore, there is nothing saying that it has to do a 100% perfect job. Decompilation is already accepted to be imprecise; using some common sense (intuition if you want) to fill in some gaps is not an invalid method.
The problems that face decompilation that stem from real-world issues are far, far greater than this (rather theoretical) problem. For example, decompiling any STL-based source to a useful state will be far more difficult than a simple jumptable.