This is why I like Design by Contract so much. If you've got a set of axioms describing exactly what the function is supposed to achieve - well the body is less important, any implementation that does an efficient jo is good enough. If contratcs are well written (like they ought to be during the requirements and specification phases) but the code is a mess then it can be rewritten, and as long as the new fulfills its contracts it doesn't matter.
There's lots of ways to to DbC. Eiffel is one of the more well known ones (more well known for doing DbC) but there's DbC add-ons for Java, and Python, D supports it, as does SPARK (SPARK pretty much relies on it).
Thank you! Serre had a lot of conjectures and there has been a surprising amount of misinformation floating around here about what he's claiming to have a start on proving. I'm glad someone who actually knows what they are talking about bothered to weigh in.
While we're at it, there's a great viral advertising firm called Whatever that I can reccomend - they use nothing but Linux and FOSS software, and go out of their way to support FOSS with their products. Tell everyone you know about Whatever for all their viral marketing needs. If we all vote with our feet we can make a real difference in the uptake of FOSS in the viral marketing world.
There have been a number of studies with regard to difficulties in shifting businesses over to OpenOffice, but it is important to remember that the school environment provides some key differences.
Training: This is significantly limited in comparison to a large company. Students and teachers are not going to use as much of the complex and or custom features, and the basics of OpenOffice are sufficiently similar that there isn't too much to learn. More importantly a school has massive turnover - students are constantly graduating and new students arriving. The majority of computer users (which is to say students) are going to fresh meat for training anyway. There simply isn't the large staff base that needs to be expensively retrained.
Features and Compatability: As already mentioned, an office suite at a school is not going to get the same work out as it will at a large company - custom macros, document tracking, custom styles etc. are all things that simply aren't going to get used. Compatability is also less of an issue. The majority of material produced on school computers is going to be students typing up reports, or using spreadsheets for assignments. These are transitory - it really doesn't matter very much if they can't be flawlessly imported into the new office suite after the report/assignment deadline has passed. There simply isn't the same amount of critical documents locked up in other formats as a large company will be faced with.
Support: Support can be purchased from Sun if you want, but at schools the majority of users are students who are, let's be honest, often left to figure it out themselves. As an added bonus OpenOffice runs on most operating systems, and the school can easily provide free copies for the students to take home and learn. At High Schools I've been to senior students who are interested are often drafted in to help with a certain amount of system adminstration (the same way senior students can volunteer to help in the library etc.) Given that OpenOffice is freely available even in source form, you can expect interested students to have a high degree of knowledge of OpenOffice and help provide support. Some of them might even be contributing code to OpenOffice!
Schools are, in general, far better placed than large companies to switch to OPenOffice. That doesn't mean that it is an easy or painless transition, merely that it is a lot easier than it is for corporations to make the move.
Personally I only liked one more album they did, "Caustic Grip", which included the jaw dropping "Provision". Then like most of the late eighties industrial bands they disappeared up their own arse and produced watered down techno.
I almost agree with you, save for "Implode" which I would rate as one of FLAs best albums. It was, I admit, rather an odd one out sandwiched between the likes of "Flavour of the Weak" and "Epitaph". Otherwise the more interesting work went on with the side projects (Synaesthesia and Noise Unit for instance) which may or may not be to your taste.
Trent is fairly well known as a devout Apple fan. Releasing in the format he has can be seen as a deliberate act in support of Apple. There's nothing wrong with that - Trent seems to have gotten a lot out of Apple over the years (using Apples for most of his work), so there's no reason not to give something back. Besides, the aim was to release something fans could play around with (hence the use of Garageband instead of ProTools etc.) not to give professional remixers everything they want.
Having not seen the film it would be rather hard to guess how successful (with critics - reviews, with theatre goers - box office) it will be. Going on what I do know (the previews, the cast) I would say that it isn't going to be utterly awful (if you think it is, you simply haven't seen an utterly awful film).
You can point me to the very long detailed "review" from Adams' biographer if you like, but I've already read that. In the end it didn't actually tell me very much about whether it would be a good film or not - it merely provided a long list of particularly specific points that he didn't like about the film, yet left me with the impression that there was still plnety of room for the film to be fun and interesting despite his complaints.
On that front, if a devout Tolkien fanatic had seen the LOTR trilogy prior to the release of any of the films I'm sure they couldhave given an equally detailed list of changes, diversions, and alterations from the "spirit of the books" that would have left people cringing thinking that the films would suck. In the end the films were not the books, but they were still very enjoyable to watch. I think your judgement is premature.
Will it be good? I don't know. Will it suck? I don't know. From what I've seen and read about it I can only say that it could go either way, and I would have to see it to know. We shall see.
A self sustaining environment for human populations, potetially just in space, is possible. Why do I say that? Well, you can think of the earth as a giant self sustaining environment for human populations. The question is not "can it be done", the question is "can it be scaled down to something we can construct". Work is being done on that problem, and so far it hasn't been completely solved - but we are actually much closer than you might think.
Of course there was that (informal) study that found that Microsoft Support and the Psychic Friends Network were both about equally helpful. I believe the Psychic Friends Network will provide Linux support of equal quality. Worth considering (given that PFN was initially cheaper than MSS).
Welcome to SELinux and mandatory access controls. Users and processes have a domain and a role. At the kernel level all access by processes to resources is passed through a security server which checks whether that process' domain and role allow it to have access to the resources it is asking for. The server has a set of rules, policies, that determine in what context (role) processes are allowed to access resources. That policy can be as fine grained as you like, and handles things on a per process context basis.
I think the more disconcerting possibility is a shareware or other program that mimics the password dialog and sends the results off somewhere. People have a remarkable tendency to use the same password for everything. This could be a boon for password farming.
I'm getting to old for all this. I can even grok the names anymore. What happened to the days of "Visi-Calc" (a visual calculator) or "Draw" or "Write" or...
They got trademarked, that's what happened. Obvious names that "give an idea of functionality" are remarkably hard to come by. Let's try renaming GIMP to something more obvious shall we... just make sure to google the names first to make sure they're not already trademarked:
Image - taken Paint - taken PhotoShop - taken (obviously) ImageShop - taken PaintShop - taken PhotoPaint - taken PhotoStudio - taken PaintStudio - taken ImageStudio - taken PhotoSuite - taken PaintSuite - taken ImageSuite - taken PhotoBox - taken PaintBox - taken ImageBox - taken...
The list goes on. Dream up any name you like that implies painting, photos manipulation, images etc. and you'll find it is trademarked already. The same goes for most everything else.
One of the original proposals for time zones did essentially this, though they divided up the clock in A o'clock through W o'clock. The letters were universal times, and in theory locally you just got used to, say, G o'clock was time to wake up and W o'clock was bed time, or whatever. The principle was that everyone around the globe was all on the same time.
Actually your sig provides a goo example for this - quintics. If you want to show that the general quintic has no solution expressible in radicals... well you can use Abel's method, which is pretty much what a computer would do. You would get a lon complicated proof that was not especially enlightening.
On the other hand, if you get a creative genius like Galois to attack the problem, you get what appears to be a whole lot of completely tangential unrelated ideas that, when put together suddenly prduce a simple elegant proof that explains far more about *why* the general quintic is unsolvable in terms of radicals. You also get whole new fields of mathematis based on the new directions Galois headed.
The point is that there are things that current computer provers are good for, and there are things they aren't. Verifying/proving compex identities or statements is what they're good at. Creative insight to create new ways of exploring the old problem, and new theories, that they currently aren't so good at.
Bear in mind that the 9/11 terrorists entered the USA under their real names, using real passports and real visas supplied by US Immigrations themselves.
Very true. However, I was simply answering the fact that passports are easy forge - and relatively speaking they are very easy to forge. Whether that stops terrorist attacks is not especially relevant. It's useful for all the things that passports are normally useful for.
Actually it pretty much makes it mandatory for Canada to check your passport when you enter - why would you let in someone who forgot to bring their US passport? All that gets you is someone stuck in Canada. Better to just turn them away before they enter. Canada doesn't really have a choice.
A simple measure that could be taken to dramatically improve the security of passports:
Include on the passport a data chip (no, not RFID!) that can be read. On the chip is the same information as in the passport (name, birthdate, description, etc. and a photo) that has been digitally signed. When passports need to be checked you verify the information against the person standing in front of you, and verify the signature against the appropriate national government provided public key.
What this means is that each national government can have a set of private keys, and be the unique signing authority for passports of that nation. Having a set of keys is desireable as you want to limit the number of people on any given key so if a key is comprimised only a limited number of people need to renew their passports. If you set passport renewal to 5 years and use large keys you can simply roll people over to new (potentially larger/stronger) keys as the renew their passports, and expire the old keys.
Is it a perfect system? No. It would make successful forgery an order of magnitude or more difficult. You can't fake a passport without the nation's private key, which presumably they'll keep well guarded. You can't alter an existing passport without invalidating the digital signature. And even if the whole system breaks and all the keys are compromised somehow - at worst you're back to exactly the system we have now because the data chip is merely an extra layer of verification on passports exactly akin to what we have now.
What is more interesting here is the derrivative. The perception of Windows is improving rapidly, the perception of Linux is pretty static. I don't see a heck of a lot of new security action going on in the Linux world. There is a heck of a lot going on in the Windows world.
If you don't see much happening with regard to security in the Linux and UNIX world, then you simply aren't really paying enough attention. UNIX is getting fitted with a new, significant, very powerful, security architecture. The difference is aking the the difference between a single-user and a multi-user os. It's coming to Linux via SELinux (though there are other implementations of the basic concept such as RSBAC). The BSDs have it in TrustedBSD, and the new (open source) Solaris 10 has it (Trusted Solaris has been integrated into the main branch). Does Windows have anything even close anywhere on the horizon? No.
Sure, for all of these systems the security architecture is new, and by default it is often either off, or in a relatively minimal configuration. The point is that it is already developed, and implemented, and in the respective kernels. From here it's a matter of educating users and developers, getting better application support allowing for stronger/stricter policies by default, and building better tools to configure and administer the system. For Windows any level of Mandatory Access Controls is still in the hazy future, to be implemented, at best, in the release after Longhorn. By the time Windows secures all its holes UNIX may well have moved a quantum leap ahead.
One major area being with GUIs. Most notably, alot of slashdotters disagreed with Gnome's switch to the spatial model.
I think you'll find that, in terms of the groupthink (as hard to properly define as it is), what "slashdotters" don't like is actually having to learn to do things a different way. It wasn't so much that the spatial model didn't suit their needs, it was more the fact that it required them to learn a new way to structure their files, and interact with their system. To be fair, having that thrust upon you (as it was in the first release) is understandably unpleasant. Still, if you look at any other area (programming being a popular one, but UIs is another) where something asks slashdotters to learn a new way of doing things, they'll never accept it, no matter how much better it might be. They will whine and find excuses as to why their current turd is worth keeping on polishing.
Of course if they ask anyone else to step back, look at a different way of doing things and take the time to properly learn or invest in it... well, they expect immediate uptake. Consider all the complaints about the RIAA clinging to an old outdated business model.
The disclaimer I better add (lest I get flamed) is that this is of course "Slashdot Groupthink" which is not every slashdotter, nor even the majority of slashdotters, but rather a case of "there's always at least one (and usually quite a few)" for each given issue.
Basic CS research ought to be funded, IMO, but there's no reason completely open-ended CS research should be funded by DARPA---that's what the National Science Foundation is for.
Let's be frank, there are certain things in basic long term CS research that DARPA is going to be a lot more interested in than the NSF. It makes sense for DARPA, then, to bother to make sure that research is getting done. The best way to make sure that research is getting done is to pay for it.
What sort of research should DARPA be interested in? Anything related to software security and assurance is going to be of more interest to DARPA than the general public (yes the general public is interested, but they aren't quite as motivated as DARPA). There's plenty you can do in that field, from new security architectures in the OS (like, for instance, what the NSA did with SELinux etc.), through to new protocols, better fault tolerance, intrusion detection etc. Having your military computer networks secure is just good practice. You should be interested in being at the cutting edge of of that. If you want a nice list of things DARPA could be doing, along with a reccomendation that more money ought to be invested in long term research at DARPA, you could try this report to the President from a month ago by the President's Information Technology Advisory Committee.
Well actually it is particularly amusing in the light of this article posted here not 2 weeks ago in which the President's advisory committee provided a report that specifically cited declining DARPA focus on long term research as a fundamental problem for cybersecurity. The particular details can be found in the report on page 19 (with other references spread throughout).
Collect together an advisory committee that reports directly to the President. Have them do a study of the state of the nation's IT security, both from a civilian and from a military standpoint). A month after the report is released, do the exact opposite of one of its primary reccomendations.
I know the government is big (perhaps that's part of the problem?) and that the left hand rarely knows what the right hand is doing, but this is quite ridiculous.
Actually noting that your Topology textbook is from the Dover series, you might want to try Ian Stewart's Concepts of Modern Mathematics which is a pretty good survey of the field, and would give you a nice sampling of the possible topics you could delve into further. If your not a mathematician, but want a good, and serious, explanation of some of the more advanced topics available, it's a great book.
I'll second that, Hatcher's is a great book on algebraic topology which is, in itself, a vastly interesting topic. You'll need some background in modern algebra. If you don't have that, there are plenty of good textbooks in it. I always quite liked the one by Fraleigh (A First Course in Abstract Algebra), but your taste may differ.
Given a good grasp of topology, and presuming you enjoy that, a couple of directions to aim for are Differential Geometry (which gets you a nice understanding of General Relativity, the shape of the universe, and that sort of thing), or if you want to head in a more pure direction, Algebraic Topology. Both those topics are "hard" in the sense that they may require material from other fields, advanced calculus for Differential Geometry, modern algebra for Algebraic Topology. Both are interesting topics in their own right though, and it should be easy enough to pick up a textbook on one of those for some background reading before starting in on your topic of choice.
Slashdot Mirror
This is why I like Design by Contract so much. If you've got a set of axioms describing exactly what the function is supposed to achieve - well the body is less important, any implementation that does an efficient jo is good enough. If contratcs are well written (like they ought to be during the requirements and specification phases) but the code is a mess then it can be rewritten, and as long as the new fulfills its contracts it doesn't matter.
There's lots of ways to to DbC. Eiffel is one of the more well known ones (more well known for doing DbC) but there's DbC add-ons for Java, and Python, D supports it, as does SPARK (SPARK pretty much relies on it).
Jedidiah.
Thank you! Serre had a lot of conjectures and there has been a surprising amount of misinformation floating around here about what he's claiming to have a start on proving. I'm glad someone who actually knows what they are talking about bothered to weigh in.
Jedidiah.
While we're at it, there's a great viral advertising firm called Whatever that I can reccomend - they use nothing but Linux and FOSS software, and go out of their way to support FOSS with their products. Tell everyone you know about Whatever for all their viral marketing needs. If we all vote with our feet we can make a real difference in the uptake of FOSS in the viral marketing world.
Jedidiah.,
Schools are, in general, far better placed than large companies to switch to OPenOffice. That doesn't mean that it is an easy or painless transition, merely that it is a lot easier than it is for corporations to make the move.
Jedidiah.
Personally I only liked one more album they did, "Caustic Grip", which included the jaw dropping "Provision". Then like most of the late eighties industrial bands they disappeared up their own arse and produced watered down techno.
I almost agree with you, save for "Implode" which I would rate as one of FLAs best albums. It was, I admit, rather an odd one out sandwiched between the likes of "Flavour of the Weak" and "Epitaph". Otherwise the more interesting work went on with the side projects (Synaesthesia and Noise Unit for instance) which may or may not be to your taste.
Jedidiah.
Trent is fairly well known as a devout Apple fan. Releasing in the format he has can be seen as a deliberate act in support of Apple. There's nothing wrong with that - Trent seems to have gotten a lot out of Apple over the years (using Apples for most of his work), so there's no reason not to give something back. Besides, the aim was to release something fans could play around with (hence the use of Garageband instead of ProTools etc.) not to give professional remixers everything they want.
Jedidiah.
Having not seen the film it would be rather hard to guess how successful (with critics - reviews, with theatre goers - box office) it will be. Going on what I do know (the previews, the cast) I would say that it isn't going to be utterly awful (if you think it is, you simply haven't seen an utterly awful film).
You can point me to the very long detailed "review" from Adams' biographer if you like, but I've already read that. In the end it didn't actually tell me very much about whether it would be a good film or not - it merely provided a long list of particularly specific points that he didn't like about the film, yet left me with the impression that there was still plnety of room for the film to be fun and interesting despite his complaints.
On that front, if a devout Tolkien fanatic had seen the LOTR trilogy prior to the release of any of the films I'm sure they couldhave given an equally detailed list of changes, diversions, and alterations from the "spirit of the books" that would have left people cringing thinking that the films would suck. In the end the films were not the books, but they were still very enjoyable to watch. I think your judgement is premature.
Will it be good? I don't know. Will it suck? I don't know. From what I've seen and read about it I can only say that it could go either way, and I would have to see it to know. We shall see.
Jedidiah.
A self sustaining environment for human populations, potetially just in space, is possible. Why do I say that? Well, you can think of the earth as a giant self sustaining environment for human populations. The question is not "can it be done", the question is "can it be scaled down to something we can construct". Work is being done on that problem, and so far it hasn't been completely solved - but we are actually much closer than you might think.
Jedidiah.
Of course there was that (informal) study that found that Microsoft Support and the Psychic Friends Network were both about equally helpful. I believe the Psychic Friends Network will provide Linux support of equal quality. Worth considering (given that PFN was initially cheaper than MSS).
Jedidiah.
Welcome to SELinux and mandatory access controls. Users and processes have a domain and a role. At the kernel level all access by processes to resources is passed through a security server which checks whether that process' domain and role allow it to have access to the resources it is asking for. The server has a set of rules, policies, that determine in what context (role) processes are allowed to access resources. That policy can be as fine grained as you like, and handles things on a per process context basis.
Jedidiah.
On the other hand, perhaps people will end up getting too used to typing in the password whenever it's presented.
"Installer? Check! Installer? Check! Virus? Check!"
I think the more disconcerting possibility is a shareware or other program that mimics the password dialog and sends the results off somewhere. People have a remarkable tendency to use the same password for everything. This could be a boon for password farming.
Jedidiah.
I'm getting to old for all this. I can even grok the names anymore. What happened to the days of "Visi-Calc" (a visual calculator) or "Draw" or "Write" or...
... just make sure to google the names first to make sure they're not already trademarked:
...
They got trademarked, that's what happened. Obvious names that "give an idea of functionality" are remarkably hard to come by. Let's try renaming GIMP to something more obvious shall we
Image - taken
Paint - taken
PhotoShop - taken (obviously)
ImageShop - taken
PaintShop - taken
PhotoPaint - taken
PhotoStudio - taken
PaintStudio - taken
ImageStudio - taken
PhotoSuite - taken
PaintSuite - taken
ImageSuite - taken
PhotoBox - taken
PaintBox - taken
ImageBox - taken
The list goes on. Dream up any name you like that implies painting, photos manipulation, images etc. and you'll find it is trademarked already. The same goes for most everything else.
Jedidiah.
One of the original proposals for time zones did essentially this, though they divided up the clock in A o'clock through W o'clock. The letters were universal times, and in theory locally you just got used to, say, G o'clock was time to wake up and W o'clock was bed time, or whatever. The principle was that everyone around the globe was all on the same time.
Jedidiah.
Actually your sig provides a goo example for this - quintics. If you want to show that the general quintic has no solution expressible in radicals ... well you can use Abel's method, which is pretty much what a computer would do. You would get a lon complicated proof that was not especially enlightening.
On the other hand, if you get a creative genius like Galois to attack the problem, you get what appears to be a whole lot of completely tangential unrelated ideas that, when put together suddenly prduce a simple elegant proof that explains far more about *why* the general quintic is unsolvable in terms of radicals. You also get whole new fields of mathematis based on the new directions Galois headed.
The point is that there are things that current computer provers are good for, and there are things they aren't. Verifying/proving compex identities or statements is what they're good at. Creative insight to create new ways of exploring the old problem, and new theories, that they currently aren't so good at.
Jedidiah.
Bear in mind that the 9/11 terrorists entered the USA under their real names, using real passports and real visas supplied by US Immigrations themselves.
Very true. However, I was simply answering the fact that passports are easy forge - and relatively speaking they are very easy to forge. Whether that stops terrorist attacks is not especially relevant. It's useful for all the things that passports are normally useful for.
Jedidiah.
Actually it pretty much makes it mandatory for Canada to check your passport when you enter - why would you let in someone who forgot to bring their US passport? All that gets you is someone stuck in Canada. Better to just turn them away before they enter. Canada doesn't really have a choice.
Jedidiah.
A simple measure that could be taken to dramatically improve the security of passports:
Include on the passport a data chip (no, not RFID!) that can be read. On the chip is the same information as in the passport (name, birthdate, description, etc. and a photo) that has been digitally signed. When passports need to be checked you verify the information against the person standing in front of you, and verify the signature against the appropriate national government provided public key.
What this means is that each national government can have a set of private keys, and be the unique signing authority for passports of that nation. Having a set of keys is desireable as you want to limit the number of people on any given key so if a key is comprimised only a limited number of people need to renew their passports. If you set passport renewal to 5 years and use large keys you can simply roll people over to new (potentially larger/stronger) keys as the renew their passports, and expire the old keys.
Is it a perfect system? No. It would make successful forgery an order of magnitude or more difficult. You can't fake a passport without the nation's private key, which presumably they'll keep well guarded. You can't alter an existing passport without invalidating the digital signature. And even if the whole system breaks and all the keys are compromised somehow - at worst you're back to exactly the system we have now because the data chip is merely an extra layer of verification on passports exactly akin to what we have now.
Jedidiah.
What is more interesting here is the derrivative. The perception of Windows is improving rapidly, the perception of Linux is pretty static. I don't see a heck of a lot of new security action going on in the Linux world. There is a heck of a lot going on in the Windows world.
If you don't see much happening with regard to security in the Linux and UNIX world, then you simply aren't really paying enough attention. UNIX is getting fitted with a new, significant, very powerful, security architecture. The difference is aking the the difference between a single-user and a multi-user os. It's coming to Linux via SELinux (though there are other implementations of the basic concept such as RSBAC). The BSDs have it in TrustedBSD, and the new (open source) Solaris 10 has it (Trusted Solaris has been integrated into the main branch). Does Windows have anything even close anywhere on the horizon? No.
Sure, for all of these systems the security architecture is new, and by default it is often either off, or in a relatively minimal configuration. The point is that it is already developed, and implemented, and in the respective kernels. From here it's a matter of educating users and developers, getting better application support allowing for stronger/stricter policies by default, and building better tools to configure and administer the system. For Windows any level of Mandatory Access Controls is still in the hazy future, to be implemented, at best, in the release after Longhorn. By the time Windows secures all its holes UNIX may well have moved a quantum leap ahead.
Jedidiah.
One major area being with GUIs. Most notably, alot of slashdotters disagreed with Gnome's switch to the spatial model.
I think you'll find that, in terms of the groupthink (as hard to properly define as it is), what "slashdotters" don't like is actually having to learn to do things a different way. It wasn't so much that the spatial model didn't suit their needs, it was more the fact that it required them to learn a new way to structure their files, and interact with their system. To be fair, having that thrust upon you (as it was in the first release) is understandably unpleasant. Still, if you look at any other area (programming being a popular one, but UIs is another) where something asks slashdotters to learn a new way of doing things, they'll never accept it, no matter how much better it might be. They will whine and find excuses as to why their current turd is worth keeping on polishing.
Of course if they ask anyone else to step back, look at a different way of doing things and take the time to properly learn or invest in it... well, they expect immediate uptake. Consider all the complaints about the RIAA clinging to an old outdated business model.
The disclaimer I better add (lest I get flamed) is that this is of course "Slashdot Groupthink" which is not every slashdotter, nor even the majority of slashdotters, but rather a case of "there's always at least one (and usually quite a few)" for each given issue.
Jedidiah
Basic CS research ought to be funded, IMO, but there's no reason completely open-ended CS research should be funded by DARPA---that's what the National Science Foundation is for.
Let's be frank, there are certain things in basic long term CS research that DARPA is going to be a lot more interested in than the NSF. It makes sense for DARPA, then, to bother to make sure that research is getting done. The best way to make sure that research is getting done is to pay for it.
What sort of research should DARPA be interested in? Anything related to software security and assurance is going to be of more interest to DARPA than the general public (yes the general public is interested, but they aren't quite as motivated as DARPA). There's plenty you can do in that field, from new security architectures in the OS (like, for instance, what the NSA did with SELinux etc.), through to new protocols, better fault tolerance, intrusion detection etc. Having your military computer networks secure is just good practice. You should be interested in being at the cutting edge of of that. If you want a nice list of things DARPA could be doing, along with a reccomendation that more money ought to be invested in long term research at DARPA, you could try this report to the President from a month ago by the President's Information Technology Advisory Committee.
Jedidiah.
Jedidiah.
Well actually it is particularly amusing in the light of this article posted here not 2 weeks ago in which the President's advisory committee provided a report that specifically cited declining DARPA focus on long term research as a fundamental problem for cybersecurity. The particular details can be found in the report on page 19 (with other references spread throughout).
Collect together an advisory committee that reports directly to the President. Have them do a study of the state of the nation's IT security, both from a civilian and from a military standpoint). A month after the report is released, do the exact opposite of one of its primary reccomendations.
I know the government is big (perhaps that's part of the problem?) and that the left hand rarely knows what the right hand is doing, but this is quite ridiculous.
Jedidiah.
Actually noting that your Topology textbook is from the Dover series, you might want to try Ian Stewart's Concepts of Modern Mathematics which is a pretty good survey of the field, and would give you a nice sampling of the possible topics you could delve into further. If your not a mathematician, but want a good, and serious, explanation of some of the more advanced topics available, it's a great book.
Jedidiah.
I'll second that, Hatcher's is a great book on algebraic topology which is, in itself, a vastly interesting topic. You'll need some background in modern algebra. If you don't have that, there are plenty of good textbooks in it. I always quite liked the one by Fraleigh (A First Course in Abstract Algebra), but your taste may differ.
Jedidiah.
Given a good grasp of topology, and presuming you enjoy that, a couple of directions to aim for are Differential Geometry (which gets you a nice understanding of General Relativity, the shape of the universe, and that sort of thing), or if you want to head in a more pure direction, Algebraic Topology. Both those topics are "hard" in the sense that they may require material from other fields, advanced calculus for Differential Geometry, modern algebra for Algebraic Topology. Both are interesting topics in their own right though, and it should be easy enough to pick up a textbook on one of those for some background reading before starting in on your topic of choice.
Jedidiah.
I think this really need to be expressed in set theoretic form to make any sense. Try:
Slashdot \subseteq Idiots \cap Computer_Owners
Jedidiah.