There is, in fact, an entire study of how to draw order from the seeming chaos that is a encrypted message. Various types of spectrums can be taken. The book "Between Silk and Cyanide" uses this study as a backdrop in such a way as to give a good feel for it. In some sense, it isn't so different than a crossword. There is both a lot of art, and a lot of science in it.
It is more difficult to develop a sense for this when considering non-text input. It is more difficult to see how it could work when using modern cyphers. But the principles are the same.
Any particular type of message will have some standard formatting information, some patterns that the cryptanalyst can look for. The modern, popular, RC4 stream cipher does have a perceptable bias in the stream of numbers produced. So hope is not lost with modern cyphers.
Actually, having line of sight to many cells is the primary reason that the FCC declares them illegal for use at altitude. Or rather, the licenses issued are for ground operation only. Basically, each cell likes to allocate frequencies independently.
A certain amount of overlap is allowable, and does provide seemless handoff. However, when you are halfway between cells, you tend to use bandwith from both. Minimizing unecessary dual use is one reason that cell phones allways have multiply power settings.
There are also cases where too many cells in range will confuse the cell phone itself. I've heard stories of people overlooking a city and finding that they need to go down into the basement to make their cell phone work. Being able to handle 2 or 3 cells of overlap is very different than being able to hand 7-8.
TeX, not a published standard? Maybe not a committe standard, but the TeXbook and the source code define it rather well. It has changed far less, and is far more well defined than, say HTML. This is especially true if you focuse of plain TeX, not LaTeX.
Any standard can die when all users abandon it. Any set of tools can be lost to the land of unmaintained software. Being the product of committe, and having a traditionally styled standards document does not provide these protections. It increases the odds of widespread and long lasting acceptence. But being free and filling a well defined nice can cause the same acceptence. TeX is a case and point.
Look at java.lang.Math. Effectivly the class is just a namespace holder for a bunch of static methods. The result is functionaly equivalent to bare methods in C++.
Or at least, they were never meant to be processed in real time. Maybe machines have gotten fast enough to do this now, but it used to be very much not a real time display system.
The other problem with TeX is that it is really intended for printed pages. People are used to systems which pander more to screen resolution than TeX does.
Overall, I don't think that TeX, as currently implemented, is likely to replace html. One could come up with a totally different backend with the same language, but that would be a real project.
On the other hand, dvi, perhaps with a standard bundling method to include images and other figures would make a really nice open replacement for pdf. And the underlying \box model of the TeX engine totally rocks. It even made it into the swing libraries as a layout manager. Is the only layout manager that is easy to use:-)
If donations to the theoretical non-profit where paid to Theo as income, then of course it would be taxed as income. On the other hand, the money could go into hardware, connectivity, or other 'buisiness' expenses paid by the non-profit. In this case, I suspect that it would still be tax free. I think that companies are generally taxed on profits, not revenue. I'd expect non-profits, if taxed at all, to be taxed in the same way.
Also, I'm no tax lawyer, but I'd expect that total gift income (and certainly any inocome from CD sales) is taxable once the total amount is sufficiently large. If most of that money was going to support the project, there could be tax savings in creating an organization.
I don't really like it
how Java forces everything to be OO.
Say that to a smalltalk programmer. Java forces nothing of the sort. Besides string handling, it does force fairly strong typing (except for the ability to cast, which I'd trade in a minute for good Standard ML style dynamic strong typing, or even a templating system). But that isn't the same as forcing OO. If you have method overloading, and if operators are mapped onto methods, it becomes perfectly OO to have overloaded operators.
If you want constrained programming, go try to do some heavy I/O in a pure functional language.
All household DVD players and DVD roms are region coded by the factory. Early DVD roms relied on the software doing the checking, but the recent models have hardware (well, to be honest, firmware) protections.
This is our good friends the RIAA in action. However, these technologies are historically unreliable once in the field. So if somebody were to do a search they might find various methods for embarassing the designers of these methods.
Actually, I think that the abovementioned CA law, or perhaps later court cases, require that the company make you aware of this.
It was formally attached to the end of my employment agreement. They required me to initial that I had read it. It my experience, they only do this when they are trying to cover their asses. The agreement itself almost seemed to contradict the statute, but it was made very clear that the statute had priority if push came to shove.
Proof of ownership? You mean those certificates of Authenticity don't work the other way? All those holograms and strange inks, and it's authenticity doesn't mean anything?
Somehow, I'm not surprised that it exists to protect Microsoft, instead of me. But you'd think that such things would cut both ways.
If it is pirated, you obviously haven't agreed to any such licenses. If it isn't pirated, there is no reason for concern. Either way, no search warrant should be issued. Or something like that:-)
While I don't know, I rather suspect that those folks at MAPS would be
happy to listen to your story, and perhaps contact those folks for you. Since emailing seems to be part of their buisiness, they probably can and will respond to this.
If not, they'll end up listed, and you can hope that your ISP subscribes to MAPS.
Also, if they have a TRUSTe certification, you can contact that group. TRUSTe is pretty rubber stampish. All it says is that you have a privacy policy and follow it. But it does say that much. And many internet sites won't work with you if you don't have it.
The beauty of open source is that once done right it is far more likely to stick around. Build an OS that is good enough and open source, and it is likely to stay open source.
As opposed to the closed source world, where companies go out of buisiness, stable versions are replaced with less stable versions, and prices can increase.
The result is that the more people who need a product, and the more they need it to be stable over time, the more it makes sense to open source it. The greatest open source projects are the most ubiquitous. Gui cd players are all well and good, but 'less' is universal, and without competition.
Sure it would. But first you'd have to clean up the bugs:-P Then you could build inport/export filters, or whatever else you like that MS hasn't already done. Or done well.
It is really quite scary that ID creates a geometry engine as flexible and stable as Quake - while Microsoft ships a word processor that crashes whenever I try to use footnotes, and which isn't a lot more useable than the Quark Express I used back when the Mac was king of desktop publishing.
Course, you'll note that Quark Express import filters are hard to find. Seems that anybody with access to the quark file format signs an agreement to not produce any filter to import the format to a different program. Heard that years ago, don't know if it is still true. Some things never change.
Copy collection? Link count? Big O notation? Splay trees? Strong typechecking? Proof carrying code? Formal models? What are these? What certification covers them? What use could they possibly be on a resume? I mean, how would you list such interests?
:-P
There is actually a serious question in that though. In my experience, recruiters don't tend to take the statement "well, I'm sort of a whatever language you happen to need kind of guy" very well. Even suggestion that you are really good at, say, optimization and performance fixes, is a difficult thing for some recruiter types to grasp.
Also, in my experience, people who can talk about such things are few and far between. That statement intentionally cuts across both nationality and experience. Well, I guess that they tend to come in clumps, but the clumps are still few and far between.
Infinity 'squared' is just infinity again - the same infinity. Infinity 'factorial' (the power set = the set of all subsets) on the other hand is a different infinity.
For those who like such things: For any set, take any function which maps it into it's power set (transforms each element of the set into a subset of the set). For this function, consider the set of all points which are not within the set which they map to. This set might be empty, but it is defined for the function. Whatever the subset, no point can be mapped to it by the function in question (think about it!). So there is an element of the power set which is not mapped to by the function. No matter what the function is. No matter what the set is.
What is the violation exactly? From the article I gather that Sun is taking some device driver code available to them only under the GPL, and then recompiling and redistributing it for Solaris.
If they are not redistributing the source code for the module, then they are not acting within the bounds of the GPL, by my reading of the document. As I undestand the GPL, it says that to redistribute derived works you must redistribute the source. If some of the header files required to build the version modified for Solaris are not freely distributable, Sun is perhaps in a grey area.
Or is the problem that the resulting device driver is seen as part of the os? The OS is clearly not open. Does the resulting assembly become a derived work, and hence be non-redistributable without source. Should it?
You are correct. I originally misread, and followed the previous commenter's lead. The goal of the system is to reduce the copying of content more than of applications. The system requires both trusted applications and a trusted OS to matter. I don't foresee any system based on this lasting more than a month or two, depending if there is any content worth stealing this way.
It may be a play to gain more of the acceptance that the MO disks have found so hard to receive. "Please the RIAA and the consumer will follow", or some such.
When did you last try to copy a commercial program from one system to another? Quite a few really require re-installing the software.
Now, if they required online registration as part of install (before activation) and if this linked your CD-key to your hard drive identifier... then we'd be in buissiness.
Actually though, the hard drive is a pretty lousy place for this. Would be comparativly easy to patch the OS to change it (see the standard work-around for software region checking of DVDs).
CPU would be harder to patch universally.
Fortunatly though, Intel got burnt trying to add unique chip ids for privacy reasons. No software vendor will require ID'd processors untill they are ubiquitious (can't risk loosing market share). No hardware vendor wants the privacy loss uproar.
Be interesting to see how long this 'hard drive id' idea lasts.
There are different levels and sublevels of knowledge about the system. Dedicated data entry people will learn the screens that they enter into by heart. All they care is that they can keep their hands on the keyboard as they do their job. They may only known one path through the program, but they know it well. This makes for different design criteria than most desktop apps.
The real art is a seemless merging of the levels. I think that the way to do this is do have some inner theme which 'does the work' and becomes apparent over time. You find yourself slowly unearthing the model that the programmer used, and as this happens you quickly gain control over the program.
TeX, (especially plain TeX), is a good example of this. First you learn to enter equations and text, using a fairly simple syntax for markup and watch them come out prettified. Then, at some point, as you try more things, and lookup more tutorial examples, you start to become aware of the \box concept. Eventually, you find that you can do anything with the right combination of \hbox and \vbox. In fact, TeX is mostly written in terms of them.
Suddenly, you can read most Plain TeX macro packages, and a good bit of the TeX Plain Format source. I was never a fan of LaTeX, because it adds a lot of other conventions and such to learn - put another way, the LaTeX learning curve has a longer flat part at the beginning, but seems a harder mountain to climb past that point.
The shape of the learning curve is, in my mind, the ultimate parameter for the software architect to get right.
Problem is airflow. Too fuzzy and you stifle it. Need more than surface area for good thermal dispation. You need to be thermally 'near' the hotspot (you need to be hot), and you need lots of airflow.
Maybe if you sandwiched the fuzz layer between the chip itself and some other surface, and then forced air - or freon - through the center of the sandwich....
I didn't check the legacy sections. I think that anything there would have showed up in the task manager. I think that they are only used when 16-bit compatibility mode is activated - when a 16 bit program is started. Such programs show up in the task manager under ntvdm or whatever it is. I tried killing _everything_. Nothing I could kill would take the dialog box with it. The only process to take any CPU time processing the click was csrss.
The number of legacy sections is unfortunate, but the the lack of documentation for the non-legacy sections are my real complaint here. Linux kernel recompilation and module configuration isn't entirely transparent, but if you understand it, and understand rc scripts you can track down any such error. Sendmail might not work without a bit more education, but at least you have tools to find what is complaining.
The problem is convincing/training the users to use unix as desktop/laptop machines. Also, they've become quite used to Exchange for not just email, but calendering and global todolisting also. Exchange seems just barely integrated enough of a groupware to be hard to pry from their hands.
Hence the exchange war-stories.
Any integrated groupware opensource projects out there?
Slashdot Mirror
It is more difficult to develop a sense for this when considering non-text input. It is more difficult to see how it could work when using modern cyphers. But the principles are the same.
Any particular type of message will have some standard formatting information, some patterns that the cryptanalyst can look for. The modern, popular, RC4 stream cipher does have a perceptable bias in the stream of numbers produced. So hope is not lost with modern cyphers.
A certain amount of overlap is allowable, and does provide seemless handoff. However, when you are halfway between cells, you tend to use bandwith from both. Minimizing unecessary dual use is one reason that cell phones allways have multiply power settings.
There are also cases where too many cells in range will confuse the cell phone itself. I've heard stories of people overlooking a city and finding that they need to go down into the basement to make their cell phone work. Being able to handle 2 or 3 cells of overlap is very different than being able to hand 7-8.
Any standard can die when all users abandon it. Any set of tools can be lost to the land of unmaintained software. Being the product of committe, and having a traditionally styled standards document does not provide these protections. It increases the odds of widespread and long lasting acceptence. But being free and filling a well defined nice can cause the same acceptence. TeX is a case and point.
Look at java.lang.Math. Effectivly the class is just a namespace holder for a bunch of static methods. The result is functionaly equivalent to bare methods in C++.
The other problem with TeX is that it is really intended for printed pages. People are used to systems which pander more to screen resolution than TeX does.
Overall, I don't think that TeX, as currently implemented, is likely to replace html. One could come up with a totally different backend with the same language, but that would be a real project.
On the other hand, dvi, perhaps with a standard bundling method to include images and other figures would make a really nice open replacement for pdf. And the underlying \box model of the TeX engine totally rocks. It even made it into the swing libraries as a layout manager. Is the only layout manager that is easy to use :-)
Also, I'm no tax lawyer, but I'd expect that total gift income (and certainly any inocome from CD sales) is taxable once the total amount is sufficiently large. If most of that money was going to support the project, there could be tax savings in creating an organization.
Good question. But I know I read a ragged paperback copy some years ago. Maybe them calling it a new book is a social engineering trick.
Say that to a smalltalk programmer. Java forces nothing of the sort. Besides string handling, it does force fairly strong typing (except for the ability to cast, which I'd trade in a minute for good Standard ML style dynamic strong typing, or even a templating system). But that isn't the same as forcing OO. If you have method overloading, and if operators are mapped onto methods, it becomes perfectly OO to have overloaded operators.
If you want constrained programming, go try to do some heavy I/O in a pure functional language.
This is our good friends the RIAA in action. However, these technologies are historically unreliable once in the field. So if somebody were to do a search they might find various methods for embarassing the designers of these methods.
It was formally attached to the end of my employment agreement. They required me to initial that I had read it. It my experience, they only do this when they are trying to cover their asses. The agreement itself almost seemed to contradict the statute, but it was made very clear that the statute had priority if push came to shove.
Somehow, I'm not surprised that it exists to protect Microsoft, instead of me. But you'd think that such things would cut both ways.
If it is pirated, you obviously haven't agreed to any such licenses. If it isn't pirated, there is no reason for concern. Either way, no search warrant should be issued. Or something like that :-)
If not, they'll end up listed, and you can hope that your ISP subscribes to MAPS.
Also, if they have a TRUSTe certification, you can contact that group. TRUSTe is pretty rubber stampish. All it says is that you have a privacy policy and follow it. But it does say that much. And many internet sites won't work with you if you don't have it.
Growth != sales
As opposed to the closed source world, where companies go out of buisiness, stable versions are replaced with less stable versions, and prices can increase.
The result is that the more people who need a product, and the more they need it to be stable over time, the more it makes sense to open source it. The greatest open source projects are the most ubiquitous. Gui cd players are all well and good, but 'less' is universal, and without competition.
It is really quite scary that ID creates a geometry engine as flexible and stable as Quake - while Microsoft ships a word processor that crashes whenever I try to use footnotes, and which isn't a lot more useable than the Quark Express I used back when the Mac was king of desktop publishing.
Course, you'll note that Quark Express import filters are hard to find. Seems that anybody with access to the quark file format signs an agreement to not produce any filter to import the format to a different program. Heard that years ago, don't know if it is still true. Some things never change.
There is actually a serious question in that though. In my experience, recruiters don't tend to take the statement "well, I'm sort of a whatever language you happen to need kind of guy" very well. Even suggestion that you are really good at, say, optimization and performance fixes, is a difficult thing for some recruiter types to grasp.
Also, in my experience, people who can talk about such things are few and far between. That statement intentionally cuts across both nationality and experience. Well, I guess that they tend to come in clumps, but the clumps are still few and far between.
For those who like such things: For any set, take any function which maps it into it's power set (transforms each element of the set into a subset of the set). For this function, consider the set of all points which are not within the set which they map to. This set might be empty, but it is defined for the function. Whatever the subset, no point can be mapped to it by the function in question (think about it!). So there is an element of the power set which is not mapped to by the function. No matter what the function is. No matter what the set is.
If they are not redistributing the source code for the module, then they are not acting within the bounds of the GPL, by my reading of the document. As I undestand the GPL, it says that to redistribute derived works you must redistribute the source. If some of the header files required to build the version modified for Solaris are not freely distributable, Sun is perhaps in a grey area.
Or is the problem that the resulting device driver is seen as part of the os? The OS is clearly not open. Does the resulting assembly become a derived work, and hence be non-redistributable without source. Should it?
It may be a play to gain more of the acceptance that the MO disks have found so hard to receive. "Please the RIAA and the consumer will follow", or some such.
Now, if they required online registration as part of install (before activation) and if this linked your CD-key to your hard drive identifier... then we'd be in buissiness.
Actually though, the hard drive is a pretty lousy place for this. Would be comparativly easy to patch the OS to change it (see the standard work-around for software region checking of DVDs).
CPU would be harder to patch universally.
Fortunatly though, Intel got burnt trying to add unique chip ids for privacy reasons. No software vendor will require ID'd processors untill they are ubiquitious (can't risk loosing market share). No hardware vendor wants the privacy loss uproar.
Be interesting to see how long this 'hard drive id' idea lasts.
The real art is a seemless merging of the levels. I think that the way to do this is do have some inner theme which 'does the work' and becomes apparent over time. You find yourself slowly unearthing the model that the programmer used, and as this happens you quickly gain control over the program.
TeX, (especially plain TeX), is a good example of this. First you learn to enter equations and text, using a fairly simple syntax for markup and watch them come out prettified. Then, at some point, as you try more things, and lookup more tutorial examples, you start to become aware of the \box concept. Eventually, you find that you can do anything with the right combination of \hbox and \vbox. In fact, TeX is mostly written in terms of them.
Suddenly, you can read most Plain TeX macro packages, and a good bit of the TeX Plain Format source. I was never a fan of LaTeX, because it adds a lot of other conventions and such to learn - put another way, the LaTeX learning curve has a longer flat part at the beginning, but seems a harder mountain to climb past that point.
The shape of the learning curve is, in my mind, the ultimate parameter for the software architect to get right.
Maybe if you sandwiched the fuzz layer between the chip itself and some other surface, and then forced air - or freon - through the center of the sandwich....
The number of legacy sections is unfortunate, but the the lack of documentation for the non-legacy sections are my real complaint here. Linux kernel recompilation and module configuration isn't entirely transparent, but if you understand it, and understand rc scripts you can track down any such error. Sendmail might not work without a bit more education, but at least you have tools to find what is complaining.
Hence the exchange war-stories.
Any integrated groupware opensource projects out there?