It seems to that there would be two ways. The first way could be to just not scan for programs that look for the hole the FBI uses. Or, it could look to see if the attacking program is the FBI program, and if so let it in.
If the take the first way, then any program can exploit that hole. This means that if we see a hole that McAfee doesn't cover for an exceptionally long time, then it is possibly the one the FBI uses. But more imporantly, people will get peaved at the when McAfee fails to protect the from virii.
If McAfee checks if the attacking program is from the FBI (say by look for a specific string or signature), then it should be reasonably easy for any competent hacker to figure out how a hostile 3rd party program could take advantage of the FBI loophole. Of course, said hacker should preferably be a respectable academic, and preferably a foreign national with no imeddiate travel plans to the US.
But, it seems that if McAfee really is cooperating with the FBI, then in the not too distant future their software will be torn the threads, and perhaps they will loose market share.
Of course, we don't know that Norton and other haven't done the same thing, but with less publicity. I wonder how feasible an open source virus scanner would be.
Of course, in this day and age, do you really need a competent virus scanner, instead of just good OS security? I mean, how often are unix boxes compromised in by programs with no local access in ways that a good firewall would have prevented? Although, we also would have to worry about no exploitable holes that a program run by a user reading email could take advantage of...
Well, boot sector and related virii seem to be pretty much dead, especially if you use linux, netbsd, etc.
I generally don't care (and never have) to watch sports. Even when I actively played baseball or tennis, I wasn't a big fan of watching sports.
As to Buffy. Well, that weeks episode wasn't just any Buffy, it was "Buffy: The Musical". I'm sorry. I know I'm pathetic. But I just couldn't resist trying to watch them sing.
OK, now where is that branding iron. Lets get this over with. Just brand me pathetic and be done with it.
Alternatively, Bill could afford a lot more right now. If his wealth is all going to chaaardity, and he can't possibly spend a fraction of it in his lifetime, what's he waiting for?
Keep in mind, anytime Billy G. spends more than a few million on something, he has to give up MS stock to do it. And giving up MS stock means that he is giving up control of his company. When he gave $18 billion of MS stock to the Bill and Melinda Gates foundation, that meant that he was making them a very large stock holder in MS. I expect that there is probably some bit in the contracts that say that Billy G. still controls the votes of those particular shares, or something like that, so he still retains that fraction of his control.
Well, it is a hard to say specifically what to do, but you should start by turning off unneeded services. One of the easiest ways to do this is to uninstall unneeded software, such as do you really need an FTP server running? NFS server? etc. Another way, is to set port security better. For instance, people coming to your machine via @home should not be allowed to connect to your NFS ports, among others.
Also, you really need to make sure you keep up to date with the patches, and it wouldn't hurt to run the bastile hardening scripts. In my experience, debian is much better from a security perspective, for people who don't know much about security.
If doing that is too hard, one of the easiest things you can do is to just use a seperate NAT (an OpenBSD box with 2 ethernics, or a Linksys router, etc). This will stop many problems dead in their tracks, but it can also reak havoc with games, unless the gaming machine is designated the DMZ (then since that machine is just for games, it shouldn't be as hard to rebuild if it is NT, or secure if it is linux, since a game box definately doesn't need an NFS or FTP server, among other things.
Besides, if the holes are never exposed publicly, then there will be little insentive
As far as I know, there aren't any word processors compatible with Word written in Lisp. I think that it has been quite some time since a word processor was written in word, which doesn't help.
As to games, I don't know of any games written front to back in Lisp. I do know that several games have used Lisp for the intelligence. Abuse was one example of this. Crash Bandicoot was another.
Not a game, but the software used for the modelling and animation of games like Mario 64, Zelda 64, FF7 at least (possibly 8 and 9, don't know when the moved to Maya), were done in a program name N-World (now Mirai), which is written in Lisp.
Now, could you write an MPEG decoder or MP3 player in Lisp? Definately. Would they be real time? I don't know. However, if you write something in lisp and it isn't fast enough, if you are sure you are using the most efficient algorithms, then it is quite acceptable to take the piece of code and rewrite it in assembly. This is in effect what happens when MPEG and MP3 decoders are optimized for things like MMX, Altivec, SSE, 3dnow, etc.
Lisp doesn't presume to force everything down your throat. Things like graphics toolkits are meant to be supplied seperately. So, numerous lisp and scheme implementations have bindings for things like X11, GTK, MFC, and probably soon Aqua. There are also bindings to OpenGL, and scene graph tools written in Lisp. And some of the most pioneering imaging software was originally written in Lisp. Just look at the amazing work done on the Symbolics machines (which ran lisp natively).
For commercial uses of Lisp, check out http://www.franz.com/success/index.lhtml. And also try searching google sometime for other places it pops up.
The real lost art is the drive to dive down in and optimize. The "secrets" in those books tend to be specific to a certain chip, so how you optimized for a 386 isn't really the same as how you optimize for a Pentium, which is different from how you optimize for a Pentium 4, let alone a StrongARM.
A large reason is that it isn't needed as much for getting higher performance. Just wait 6 months, and you have a lot more power at hand. It would have taken you 6 months to fully hand optimize the engine to begin with, assuming that you didn't need to totally rework your OO model to accomodate the optimizations.
The biggest things people can do is to figure out how to analyze code for performance. This include making sure that you are using good algorithms at a high level (say C++), and making sure you have profile were the bottle necks are (no point optimizing in assembly something that only runs at the start up of the program). Then learn to count CPU cycles and how to use even lower level timing and profiling tools.
It really can be a lot of fun, but it also is something that should be left to the very last step. So, if the game isn't worth playing, don't bother optimizing.
I wish I'd been around the computer scene back when the Pixel Planes project was in full swing, and the Pixar Image Computer was hot, and Symbolics machines were to be the wave of the future. Nowdays, it is hard to find such machines, let alone people to help you get them working again. I'm still fighting with getting NeWS working on my Sun3, and that is no where near as rare as the other systems mentioned above.
How did you steer with the treadmill? I have this idea of using some method to track when the player turns, and then rotate the orientation of the treadmill under them. But, I don't have easy access to a proper turn table for this, though if I was serious, my grandfather could probably help me retro fit or build one.
Until someone comes up with a good way to use a tread mill and steer by walking in a Cave, I think that CAVES are going to be less than ideal for most entertainment purposes.
Now, sit down in a cockpit based games are another issue. Those are possibly even cheaper to put together sims for, and they map well to how the game would be played.
Also, PowerGloves are crap. Better to build your own. It isn't that hard to make superior finger bend sensors out of fiber optics, and you just attach a tracker of some sort (hack the tracker off a powerglove if you are really desperate, or use a Fasttrack if you have the money, or do something in between). But, gloves also kinda stink because of lack of tactile feedback. So, for shoots, just track some sort of toy gun instead.
Ahem. If the best techonologies won, we wouldn't be running Nextstep on Alpha, we would be running Genera on Alpha. Oh wait. We can do that. Now, if only Symbolics would drop the price of OpenGenera, and if only they hadn't dumped the S-Graphics unit. Nichimen/WingedEdge/izware/current name of the month seem to be making a real mess of what should be some kick butt products.
Excel seems to be able to scale for one user up to around 70megs. Beyond that, it is better to just use it as a querying and analysis tool front ended to an Access/ODBC/ADO database.
However, what was also interesting was that the same thing happens when using sofware only playback on other machines. I wonder how they do that... Probably some Win32 API call.
In 1996 it was pretty clear to me that ATX was where everything was going, although I went with AT anyway since on the particular machine I bought it saved me over $150, which was a lot to me at the time.
Now, 5 years later, I'm trying to purge out AT gear, but it will be another few years before I'm completely rid of it. It is rather nice in it's simplicity though. I'm kinda sad at the change.
ATX doesn't seem to have a major replacement yet, but people keep making different extensions to it which can cause confusion.
When it comes to sound, professional gear is called for, even if you are just doing it as a hobby. Seriously, try using a cheap 4-track versus an ADAT someday, and you will quickly see what I mean. If you don't want to use total junk, then being a hobbiest takes some serious cash, and anything that can be done to reduce the amount is helpfull. Thankfully, most sound gear and musical instruments don't go out of date quickly, so there is a healthy used market for the gear.
I was assuming that little billy had his own computer. Since 486s tend to be free for the taking, and a 486 would run linux quite nicely, it is a reasonable assumption. I had more than one of my own computers before I could afford my own ISP.
AOL users on Linux would definately be a minority. However, AOL does have a very impressive world wide network, which makes them very appealing for people who travel a lot. AOL users get scorned quite a bit, but I've met a few who would make the average/.er shiver with their computer knowledge.
Plus, as mentioned elsewhere, lots of kids are stuck using the family ISP, and Mom just refuses to switch. These kids too could now use linux.
Is this pipe interface documented anywhere public? I have access to mathematica at school, and hacking new frontends up for it could be fun. I have a few ideas...
Yes, but how many people volunteer for the military that already know how to admin NT? I'm thinking that most admins in the military were trained by the military.
Perhaps the reason it won't die is because to those of us untrained people who through a network together using hubs rather than switches, it looks to be true. When the network load meter gets to about 60%, the col. LED stays on almost permenantly.
I'm an working at piecing together the gear to run FDDI. It is a 100mbit token ring design. It probably isn't quite as good as full duplex fast ethernet fully switched, but it is cheaper, especially when looking at hooking up older machines from SGI and Sun.
Slashdot Mirror
Yes, you would be buying records, then copying them onto your 24bit/96khz digital audio player.
Actually, a logrithmic encoding scheme would be better, but that would mean custom DACs.
It seems to that there would be two ways. The first way could be to just not scan for programs that look for the hole the FBI uses. Or, it could look to see if the attacking program is the FBI program, and if so let it in.
If the take the first way, then any program can exploit that hole. This means that if we see a hole that McAfee doesn't cover for an exceptionally long time, then it is possibly the one the FBI uses. But more imporantly, people will get peaved at the when McAfee fails to protect the from virii.
If McAfee checks if the attacking program is from the FBI (say by look for a specific string or signature), then it should be reasonably easy for any competent hacker to figure out how a hostile 3rd party program could take advantage of the FBI loophole. Of course, said hacker should preferably be a respectable academic, and preferably a foreign national with no imeddiate travel plans to the US.
But, it seems that if McAfee really is cooperating with the FBI, then in the not too distant future their software will be torn the threads, and perhaps they will loose market share.
Of course, we don't know that Norton and other haven't done the same thing, but with less publicity. I wonder how feasible an open source virus scanner would be.
Of course, in this day and age, do you really need a competent virus scanner, instead of just good OS security? I mean, how often are unix boxes compromised in by programs with no local access in ways that a good firewall would have prevented? Although, we also would have to worry about no exploitable holes that a program run by a user reading email could take advantage of...
Well, boot sector and related virii seem to be pretty much dead, especially if you use linux, netbsd, etc.
I generally don't care (and never have) to watch sports. Even when I actively played baseball or tennis, I wasn't a big fan of watching sports.
As to Buffy. Well, that weeks episode wasn't just any Buffy, it was "Buffy: The Musical". I'm sorry. I know I'm pathetic. But I just couldn't resist trying to watch them sing.
OK, now where is that branding iron. Lets get this over with. Just brand me pathetic and be done with it.
Keep in mind, anytime Billy G. spends more than a few million on something, he has to give up MS stock to do it. And giving up MS stock means that he is giving up control of his company. When he gave $18 billion of MS stock to the Bill and Melinda Gates foundation, that meant that he was making them a very large stock holder in MS. I expect that there is probably some bit in the contracts that say that Billy G. still controls the votes of those particular shares, or something like that, so he still retains that fraction of his control.
Well, it is a hard to say specifically what to do, but you should start by turning off unneeded services. One of the easiest ways to do this is to uninstall unneeded software, such as do you really need an FTP server running? NFS server? etc. Another way, is to set port security better. For instance, people coming to your machine via @home should not be allowed to connect to your NFS ports, among others.
Also, you really need to make sure you keep up to date with the patches, and it wouldn't hurt to run the bastile hardening scripts. In my experience, debian is much better from a security perspective, for people who don't know much about security.
If doing that is too hard, one of the easiest things you can do is to just use a seperate NAT (an OpenBSD box with 2 ethernics, or a Linksys router, etc). This will stop many problems dead in their tracks, but it can also reak havoc with games, unless the gaming machine is designated the DMZ (then since that machine is just for games, it shouldn't be as hard to rebuild if it is NT, or secure if it is linux, since a game box definately doesn't need an NFS or FTP server, among other things.
Besides, if the holes are never exposed publicly, then there will be little insentive
Hockey pre-empted Buffy in my area. And I had a tape in the VCR and all too.
As far as I know, there aren't any word processors compatible with Word written in Lisp. I think that it has been quite some time since a word processor was written in word, which doesn't help.
As to games, I don't know of any games written front to back in Lisp. I do know that several games have used Lisp for the intelligence. Abuse was one example of this. Crash Bandicoot was another.
Not a game, but the software used for the modelling and animation of games like Mario 64, Zelda 64, FF7 at least (possibly 8 and 9, don't know when the moved to Maya), were done in a program name N-World (now Mirai), which is written in Lisp.
Now, could you write an MPEG decoder or MP3 player in Lisp? Definately. Would they be real time? I don't know. However, if you write something in lisp and it isn't fast enough, if you are sure you are using the most efficient algorithms, then it is quite acceptable to take the piece of code and rewrite it in assembly. This is in effect what happens when MPEG and MP3 decoders are optimized for things like MMX, Altivec, SSE, 3dnow, etc.
Lisp doesn't presume to force everything down your throat. Things like graphics toolkits are meant to be supplied seperately. So, numerous lisp and scheme implementations have bindings for things like X11, GTK, MFC, and probably soon Aqua. There are also bindings to OpenGL, and scene graph tools written in Lisp. And some of the most pioneering imaging software was originally written in Lisp. Just look at the amazing work done on the Symbolics machines (which ran lisp natively).
For commercial uses of Lisp, check out http://www.franz.com/success/index.lhtml. And also try searching google sometime for other places it pops up.
The real lost art is the drive to dive down in and optimize. The "secrets" in those books tend to be specific to a certain chip, so how you optimized for a 386 isn't really the same as how you optimize for a Pentium, which is different from how you optimize for a Pentium 4, let alone a StrongARM.
A large reason is that it isn't needed as much for getting higher performance. Just wait 6 months, and you have a lot more power at hand. It would have taken you 6 months to fully hand optimize the engine to begin with, assuming that you didn't need to totally rework your OO model to accomodate the optimizations.
The biggest things people can do is to figure out how to analyze code for performance. This include making sure that you are using good algorithms at a high level (say C++), and making sure you have profile were the bottle necks are (no point optimizing in assembly something that only runs at the start up of the program). Then learn to count CPU cycles and how to use even lower level timing and profiling tools.
It really can be a lot of fun, but it also is something that should be left to the very last step. So, if the game isn't worth playing, don't bother optimizing.
I wish I'd been around the computer scene back when the Pixel Planes project was in full swing, and the Pixar Image Computer was hot, and Symbolics machines were to be the wave of the future. Nowdays, it is hard to find such machines, let alone people to help you get them working again. I'm still fighting with getting NeWS working on my Sun3, and that is no where near as rare as the other systems mentioned above.
How did you steer with the treadmill? I have this idea of using some method to track when the player turns, and then rotate the orientation of the treadmill under them. But, I don't have easy access to a proper turn table for this, though if I was serious, my grandfather could probably help me retro fit or build one.
Until someone comes up with a good way to use a tread mill and steer by walking in a Cave, I think that CAVES are going to be less than ideal for most entertainment purposes.
Now, sit down in a cockpit based games are another issue. Those are possibly even cheaper to put together sims for, and they map well to how the game would be played.
Also, PowerGloves are crap. Better to build your own. It isn't that hard to make superior finger bend sensors out of fiber optics, and you just attach a tracker of some sort (hack the tracker off a powerglove if you are really desperate, or use a Fasttrack if you have the money, or do something in between). But, gloves also kinda stink because of lack of tactile feedback. So, for shoots, just track some sort of toy gun instead.
Ahem. If the best techonologies won, we wouldn't be running Nextstep on Alpha, we would be running Genera on Alpha. Oh wait. We can do that. Now, if only Symbolics would drop the price of OpenGenera, and if only they hadn't dumped the S-Graphics unit. Nichimen/WingedEdge/izware/current name of the month seem to be making a real mess of what should be some kick butt products.
Excel seems to be able to scale for one user up to around 70megs. Beyond that, it is better to just use it as a querying and analysis tool front ended to an Access/ODBC/ADO database.
Wanna put some money on that?
Just treat each sheet in an excel file as a new table, and then normalize as usual. Pretty ugly, and I don't really know what the point is though.
I've noticed that with my DXR2 playback card.
However, what was also interesting was that the same thing happens when using sofware only playback on other machines. I wonder how they do that... Probably some Win32 API call.
In 1996 it was pretty clear to me that ATX was where everything was going, although I went with AT anyway since on the particular machine I bought it saved me over $150, which was a lot to me at the time.
Now, 5 years later, I'm trying to purge out AT gear, but it will be another few years before I'm completely rid of it. It is rather nice in it's simplicity though. I'm kinda sad at the change.
ATX doesn't seem to have a major replacement yet, but people keep making different extensions to it which can cause confusion.
When it comes to sound, professional gear is called for, even if you are just doing it as a hobby. Seriously, try using a cheap 4-track versus an ADAT someday, and you will quickly see what I mean. If you don't want to use total junk, then being a hobbiest takes some serious cash, and anything that can be done to reduce the amount is helpfull. Thankfully, most sound gear and musical instruments don't go out of date quickly, so there is a healthy used market for the gear.
I was assuming that little billy had his own computer. Since 486s tend to be free for the taking, and a 486 would run linux quite nicely, it is a reasonable assumption. I had more than one of my own computers before I could afford my own ISP.
AOL users on Linux would definately be a minority. However, AOL does have a very impressive world wide network, which makes them very appealing for people who travel a lot. AOL users get scorned quite a bit, but I've met a few who would make the average /.er shiver with their computer knowledge.
Plus, as mentioned elsewhere, lots of kids are stuck using the family ISP, and Mom just refuses to switch. These kids too could now use linux.
Somehow I would have thought it would cost more. Either that, or you have a particularly expensive house.
Is this pipe interface documented anywhere public? I have access to mathematica at school, and hacking new frontends up for it could be fun. I have a few ideas...
>You really think all CDs of performances of >Bach's work are copyright free?
They are copyright free if I record myself or a friend (who signs a waiver) playing Bach from a pre 1925 arrangement.
Yes, but how many people volunteer for the military that already know how to admin NT? I'm thinking that most admins in the military were trained by the military.
Perhaps the reason it won't die is because to those of us untrained people who through a network together using hubs rather than switches, it looks to be true. When the network load meter gets to about 60%, the col. LED stays on almost permenantly.
I'm an working at piecing together the gear to run FDDI. It is a 100mbit token ring design. It probably isn't quite as good as full duplex fast ethernet fully switched, but it is cheaper, especially when looking at hooking up older machines from SGI and Sun.
So, the m68k hits a wall and dies? That doesn't seem to be the result on the Sun3s I've seen stressed.
Also, I doubt that is the case on VAX machines (VAX being the CISCiest chip ever).