Well, the website survived just fine, but the 'MogCam' is down. Guess 'slashdotted' wasn't one of the 'extreme conditions' the vehicle was designed for:)
OK, I actually made a mistake the first time. I was trying to figure out what the bytes I cut off at the beginning were, and I realized the whole image is a loop, which was started partway through. So, I've moved the partial section from the bottom to complete the section at the top, and updated the GIF accordingly. It should now be correct, except for the noise.
There isn't any actual 'code' in this, it's just an on/off raster image just like all the other message we send out. The first 69 bytes (?!) appear to be noise. If you cut those off, remove all the line feeds, then wrap every 127 bytes, then replace all the 1s with 0xFF, 0s with 0x00, load it into Photoshop as a 127x2148 RAW image file, enlarge it to 400% for readability, and save it as a GIF, you end up with this: http://www.perlstorm.net/message.gif
Of course, making sense out of the resulting image could take a while. At the top they're counting in binary, and seem to be assigning an arbitrary symbol for each number. The symbols seem to have been chosen in an attempt to make them out even when partially garbled. Those symbols and certain pictures are then used throughout the rest of the image. Heh, and check out the naked ppl!
Regarding your drive burps, the Fasttrack clients (Kazaa, Morpheus) by default act as 'super-nodes', which maintain lists of shared files not only on themselves, but on 'nearby' peers as well. They then respond to search requests not only with their hits, but also with hits from other nearby nodes. The idea behind this is that people with fast connections and processors take some of the burden off the lesser-endowed. It seems to me that making this be the default behavior in every client sort of defeats the purpose, but they didn't ask me. I suspect that this behavior (indexing and searching file lists from other hosts) is probably at least partially responsible for your drive activity. You might try going into the options and disabling Super-Node behavior and see if that stops it. I'm too lazy to install one of the clients on this box just to check:)
Their crappy video scares me. First it targets an earbud and identifies it as a communications device. Then it targets the headphone loop and identifies an audio device. Then it targets a PDA and identifies it as such. Then it targets the guys mouth and identifies it as an 'Other Pocket'.
Just how many bodily orifices are they including when they say it gives you 15 pockets?
Second, the linux port (Debian!) for this processor is already stable, homepage: http://www.m17n.org/linux-sh/, all we'd need to know is what kind of display it will have.
For everyone complaining about the lack of a mirror, the.avi's are making the rounds on gnutella right now. The file names you want are 'cab-brasseyespecial1.avi' and 'cab-brasseyespecial2.avi', a search for 'brasseye' should bring them up.
Remember, PLEASE share them back out once you get them! I've had 73 ppl download these from me already and there's only like 3 ppl on this section of gnet sharing this. Thanks.
-Jade E.
They skipped an entire generation.
on
Water Guns
·
· Score: 3
The article went straight from squirt pistols to second generation Super Soakers, without ever mentioning the originals. The original Super Soakers weren't of the 2 reservior variety, but were a single reservior into which you pumped more and more air, not water, to build pressure.
Doesn't say much for their research.
-Jade E.
P.S. Yes, I've emailed the author about this and posted it in their forum. No response yet.
I work for Sprint, who has a pretty good training program, except that you have to do it all on your own time. You get a $5000/year training budget. If you want to take school classes, you pay up front then get reimbursed as long as you get a B. If you want to take cert prep classes, you don't even have to pay up front (thanks to a deal with a local training place, this will vary in some areas), but Sprint will only pay if you pass the exam (unfortunately, they don't pay the exam fees.) You're responsible for paying for the course if you fail. If you want to take Sprint's own internal training (named the University of Excellence), which is offered as CBT nationwide and actual classroom training in a couple of major cities (Kansas City being the main campus), the financial arrangements are the same as cert classes (You don't pay up front, but you pay later if you fail), except that the CBT courses are generally only like $20.
There's a catch, though. Whenever you submit the request form for a class (which are generally automatically approved), you sign a contract saying that you'll pay back your education costs if you leave the company within a year of the date on the check. It's not cumulative, so if you take a bunch of classes over 6 months you only need to work a year after the last one is payed for, but it's incentive to not go out and get a CCNA or something just so you can get a better job elsewhere:)
I'm a former employee of a Fortune 500 company. Specifically EDS (Electronic Data Systems), currently at #106 on Fortune's website. While employed there (A large mistake), I was put in charge of a project to create an Intranet for a technical support call center. There had been extensive discussion of this, as the knowledgebase they were using was totally unacceptable and it's use actually served to *increase* the call times they were so desperately concerned with. I, along with an employee assigned to help me, prepared an exhaustive presentation that presented in both technical and laymans terms the pros and cons of each solution we were considering (A new NT Server, NT on an existing PC, hosting on one of the existing NT servers, Linux on an existing PC, and purchasing external web hosting.)
When we made the presentation, the building manager (The highest official at that location) was blown away. He actually went so far as to pull one of the new PC's we had just received and call the IT peon (A clueless MCSE) in to hook us up with a network connection so we could get started. The project managers who worked for him, unfortunately, convinced him he was being a little hasty and needed to speak with the regional IT director (or someone of that nature) before we started. The next day we were told the entire intranet project was cancelled. Why, we asked. Because we couldn't use Linux. We asked 2 further questions: If they didn't want Linux, why couldn't we use NT? And why didn't they want Linux?
Two days later the answers from the corporate cloud collectively known as 'IT' returned. 1) We couldn't use NT because there was no way we were going to add an unsecured server to the network. Their definition of 'unsecured' was 'not set up by IT', and IT refused to help. Second, there was no way in hell any of the internal webservers was going to host the site, they were all dedicated to the sites they were already hosting.
This left only our last question, why not Linux. And the answer stunned us, not only because they deigned to answer in *multiple* sentences (Something never encountered before or since from that department), but also in it's sheer mindboggling stupidity. The answer was, roughly:
1) Nothing is secure unless IT sets it up. IT doesn't know how to set up Linux, therefore Linux isn't secure.
2) Linux would interfere with the NT domains.
3) Most importantly, EDS had a single-vendor contract with Microsoft and was in the process of replacing all it's non-MS servers with NT, supposedly including our CentreVu CMS system, among others. Using Linux for the intranet would break that contract.
I have no idea what the IT department was smoking, especially if the contract BS was actually true, but it obviously messed with their heads. And perhaps the worst part of the whole thing was their final suggestion to alleviate the knowledgebase problems: We should build the intranet site on Geocities.
I'm tired, so I'll leave you to draw your own conclusions.
Microsoft... is trying to make a central repository of personal information... stored in redundant Microsoft controlled datacenters... including credit card numbers...
So all I need to do is SE a domain admin password from *one* microsoft employee, with a *crapload* of them to choose from, and I get *millions and millions* of credit card numbers, addresses, perhaps even bank accounts?
1) This isn't new. I downloaded their IPv6 version of Fnord! 3 months ago, and it's been around longer than that (The oldest version on their FTP server is from 9/2000.)
2) This isn't 'hidden', it's linked off the Microsoft Research IPv6 homepage, which in turn is linked off of the Current Research page (Although they want you to register if you go in that way).
3) This isn't Microsoft, the OS maker, it's Microsoft Research, the R&D Lab.
4) Microsoft didn't write Fnord!, it specifically says on their page 'Fnord! is a Windows web server we found on the net and ported to run on our stack.' and 'Fnord! was apparently written by Brian Morin while a student at Worcester Polytechnic Institute.' The readme.msr goes on to give the URL of the authors website (now defunct) and a URL of a Tucows mirror where the original could still be found when they wrote the readme. (I didn't check if that mirror still works.)
5) Yes, they included the source code in their port. No, it's not for Visual C++. It's actually a Borland C++ 4.5 project, although I believe their binaries are compiled with VC++.
6) Finally, I'd like to point out that there are a lot of other interesting-sounding files available on ftp.research.microsoft.com, from the/pub,/users, and several other directories.
AMD isn't hyping this Transmeta deal any more than they have any of the other more major alliances (not alignments) they've developed. CNNfn isn't even really hyping this deal, even though they seem to immensely enjoy referring to Transmeta as an 'upstart'. In reality, the only things at work here are one CNNfn reporter's obsession with Transmeta (Seriously, try a search for 'Richard Richtmyer' at CNNfn. He's the author of this article, and 4 of his 6 articles this month so far are about Transmeta.) and the/. crowd's obsession with all things Linux. (If you don't know the connection between Transmeta and Linux, well, I can't help you.)
Anyways, nothing bad on AMD's part, they had no control over this.
Well, I just finished downloading their sample file (a 20MB quicktime (uugh) trailer), and the technology seems to work, albeit not as well as a single fast server. When the download started there were 3 participants in the 'active mesh' and I was getting 41KB/sec or so transfer. By the time it was done there were 12 participants, and I finished the download at 55KB/sec. Even after the download finished, the Swarmcast client is sitting in my tray (Yes, this is a windows box) happily transmitting at between 8 and 14KB/sec.
Thing is, I don't see why anyone would use this. In reality, the transfer rates aren't as good as a single fast dedicated server (I can easily get 75-80KB/sec on this line), so there's really no gain on the client end from using this. On the server end, yeah you're using a lot less bandwidth. This might be useful for open source projects or other products that are downloaded by knowledgable people, but you're average computer luser isn't going to want to download 6MB of Java Runtime/Swarmcast Client just so they can save your company money by getting slower-than-normal downloads. (And I shudder to think what a distributed p2p network comprised mostly of 56k modems would be like, at least at the moment it seems most of the users online are broadband.)
Anyways, pessimism aside, here's what I'd like to see:
Some indication of how much bandwidth you're 'serving'. (My estimate above is based on the stats in my RRAS console, there's no indicator anywhere that you're even sending data.)
An option to automatically shut the client down when you're done with the file. This may seem selfihs, but I still know people paying by the hour for 'net access, and until they can be sure they can start a download with this and have their connection go idle when it's done, they're not going to use it.
Some sort of statistics (Number of packets received, number served, total clients served, things like that.)
OK, I'm done for the moment. If they play this right I can see it helping out smaller outfits with knowledgable users, maybe eventually even going mainstream if they can convince people it's worth their while to install. It's definitely sparse on information in it's current form, but hey, it's a beta. It's a good idea, and I wish them luck.
Another common practice is to use partitioned or "hidden" sectors to hold specific information. That practice is undocumented and results in severe system problems that are difficult to debug. Over the years, broken implementations and tools have been released to the public, making support difficult.
It'd be nice if they bothered to add that this 'undocumented' and 'broken' practice is exactly how disk signatures and dynamic disks are implemented in Windows 2000. (And maybe previous versions of NT, I'm not sure)
The PC version (for Windows or your favorite DOS emulator) is available here (from www.gamingdepot.com, in the 'adventure' section, if you don't trust links)
Hey everbody! Come listen to my new 'Net broadcast station! Use your favorite streaming client to listen at 192.168.0.1! 24 hour 24k encodes of your favorite music!
And, as an added bonus, if you've use *my* client, every 4 hours or so the encrypted data in the music stream gets saved into a high quality, watermark-free, 196k vbr encode of some of your favorite songs! Remember, it's not piracy if they never catch you!
Imagine what your reaction would have been 5 years ago if someone had told you that, 5 years from now:
* IBM would be making free, open source software for Linux.
* Microsoft would use a RedHat package as a defense exhibit in an anti-trust trial.
* All the preparations for Y2K would be successful, and nothing would happen.
* AOL would merge with Time Warner, but only after buying CompuServe, Mirabilis, and Nullsoft.
And finally...
* Apple would be selling their own flavor of UNIX.
There's one other point you're missing. (In the US at least,) Copyrights last 20 years. Should you still have a DVD laying around 20 years after it was made, you have every right to make copies for all your friends. (Dude, look what found in my closet!! We used to LOVE this movie!) If they successfully implement this watermark without some kind of expiration date, they're breaking the constitution, which grants limited powers of patents and copyrights. Then again, if they do implement an expiration date, you're going to see a lot of computers with their clocks set to 2025.
On a different note, I'm not really worried about this at all. The new watermark is going to be dependent on hardware support. People don't want watermarks. So guess what's going to happen to the first manufacturer to enforce watermarks? That's right, nobody's going to buy their burner. And if they allow watermark detection, but don't enforce it at the hardware level, then you can damn well bet that somebody (Me, for example) is going to write a burning program that won't enforce it at the software level.
So, like I said, I'm not worried about it. Once consumer grade DVD burners catch up, I'll be able to backup my DVDs. (No, that's not a euphamism for pirating them.) And the pirates in China will still be able to make their bootleg copies, for whatever it's worth to them. All in all, this isn't really going to affect anybody, unless they go out and buy a watermark-enforcing piece of hardware. And even if they do, I bet it'll be a fun hack to remove the watermark detection from the circuit.
Tcsh isn't the only shell that's been ported to Windows. But to answer your question, that's not the way Windows works (In recent versions, at least.) All apps interface with the kernel through it's DLL calls (OK, not *ALL* apps, but any that use the GUI.). There isn't really a 'shell', even Explorer (The standard 'desktop' you see when you boot) is just another App. When you run a command prompt (Using 9x's command.com, or 2000's much-improved cmd.exe), all you're really doing is invoking an appication that provides a minimal execution environment for command line apps (Those being the ones that can't [don't] directly access the kernel.) So, any app that can provide that environment can execute those apps. This includes things that act like command interpreters, such as shell ports (I know several people that swear by bash under 2k), but can also include full blown GUI apps that provide a 'hidden' environment for a command line app, providing input to the process and capturing it's output. (Sort of like a tremendous badly implemented subprocess pipe.)
So, there's you're offtopic answer.
Now, so this entire post isn't totally offtopic, as I'm writing this the kernel.org bandwodth meter is down to 30.11, so I guess the initial rush is over. I wonder how many people download the new kernels just to say they did, but never actually install them...
I dont know exactly what (all) methods they employ to detect attacks, but the University of Arizona is already using autonomous intrusion detection boxes. I do, however, know 2 things about them for sure:
1) When they detect intrusions, their response is to telnet to the edge router for whichever line the attack is coming through, and block the IP there, for increasingly longer periods.
2) They consider it an attack if you try to FXP a file to a server inside the U when both you and the source server are outside. This is, of course, how I first became aware of them.
The netadmin I know there tells me these boxen are called 'NetRangers', and we had a lengthy theoretical talk about how scary it is for autonomous devices to have exec access to your routers, and wondering whether they're smart enough to detect a constant barrage of packets with rotating forged sources before most of the internet is blocked at the routers.
This probably reminds you of the project from UoW's HITL because it *is* the project from UoW's HITL. It'd be nice if Microvision bothered to mention that. Oh well, at least the Dev team's page at UoW mentions that Microvision are the ones developing commercial applications for it. Also has an actual picture of the prototype, which isn't nearly as high-tech as the Microvision site would lead you to believe.
Slashdot Mirror
Well, the website survived just fine, but the 'MogCam' is down. Guess 'slashdotted' wasn't one of the 'extreme conditions' the vehicle was designed for :)
OK, I actually made a mistake the first time. I was trying to figure out what the bytes I cut off at the beginning were, and I realized the whole image is a loop, which was started partway through. So, I've moved the partial section from the bottom to complete the section at the top, and updated the GIF accordingly. It should now be correct, except for the noise.
Of course, making sense out of the resulting image could take a while. At the top they're counting in binary, and seem to be assigning an arbitrary symbol for each number. The symbols seem to have been chosen in an attempt to make them out even when partially garbled. Those symbols and certain pictures are then used throughout the rest of the image. Heh, and check out the naked ppl!
Regarding your drive burps, the Fasttrack clients (Kazaa, Morpheus) by default act as 'super-nodes', which maintain lists of shared files not only on themselves, but on 'nearby' peers as well. They then respond to search requests not only with their hits, but also with hits from other nearby nodes. The idea behind this is that people with fast connections and processors take some of the burden off the lesser-endowed. It seems to me that making this be the default behavior in every client sort of defeats the purpose, but they didn't ask me. I suspect that this behavior (indexing and searching file lists from other hosts) is probably at least partially responsible for your drive activity. You might try going into the options and disabling Super-Node behavior and see if that stops it. I'm too lazy to install one of the clients on this box just to check :)
Just how many bodily orifices are they including when they say it gives you 15 pockets?
Second, the linux port (Debian!) for this processor is already stable, homepage: http://www.m17n.org/linux-sh/, all we'd need to know is what kind of display it will have.
It's Gnutella's standard port, 6346. The filenames are the ones in the top post, do a search for brasseye and they should come up immediately.
-Jade E.
Remember, PLEASE share them back out once you get them! I've had 73 ppl download these from me already and there's only like 3 ppl on this section of gnet sharing this. Thanks.
-Jade E.
Doesn't say much for their research.
-Jade E.
P.S. Yes, I've emailed the author about this and posted it in their forum. No response yet.
There's a catch, though. Whenever you submit the request form for a class (which are generally automatically approved), you sign a contract saying that you'll pay back your education costs if you leave the company within a year of the date on the check. It's not cumulative, so if you take a bunch of classes over 6 months you only need to work a year after the last one is payed for, but it's incentive to not go out and get a CCNA or something just so you can get a better job elsewhere :)
-Jade E.
When we made the presentation, the building manager (The highest official at that location) was blown away. He actually went so far as to pull one of the new PC's we had just received and call the IT peon (A clueless MCSE) in to hook us up with a network connection so we could get started. The project managers who worked for him, unfortunately, convinced him he was being a little hasty and needed to speak with the regional IT director (or someone of that nature) before we started. The next day we were told the entire intranet project was cancelled. Why, we asked. Because we couldn't use Linux. We asked 2 further questions: If they didn't want Linux, why couldn't we use NT? And why didn't they want Linux?
Two days later the answers from the corporate cloud collectively known as 'IT' returned. 1) We couldn't use NT because there was no way we were going to add an unsecured server to the network. Their definition of 'unsecured' was 'not set up by IT', and IT refused to help. Second, there was no way in hell any of the internal webservers was going to host the site, they were all dedicated to the sites they were already hosting.
This left only our last question, why not Linux. And the answer stunned us, not only because they deigned to answer in *multiple* sentences (Something never encountered before or since from that department), but also in it's sheer mindboggling stupidity. The answer was, roughly:
1) Nothing is secure unless IT sets it up. IT doesn't know how to set up Linux, therefore Linux isn't secure.
2) Linux would interfere with the NT domains.
3) Most importantly, EDS had a single-vendor contract with Microsoft and was in the process of replacing all it's non-MS servers with NT, supposedly including our CentreVu CMS system, among others. Using Linux for the intranet would break that contract.
I have no idea what the IT department was smoking, especially if the contract BS was actually true, but it obviously messed with their heads. And perhaps the worst part of the whole thing was their final suggestion to alleviate the knowledgebase problems: We should build the intranet site on Geocities.
I'm tired, so I'll leave you to draw your own conclusions.
-Jade E.
Microsoft... is trying to make a central repository of personal information... stored in redundant Microsoft controlled datacenters... including credit card numbers...
So all I need to do is SE a domain admin password from *one* microsoft employee, with a *crapload* of them to choose from, and I get *millions and millions* of credit card numbers, addresses, perhaps even bank accounts?
OK. I'm up for that.
-Jade E.
1) This isn't new. I downloaded their IPv6 version of Fnord! 3 months ago, and it's been around longer than that (The oldest version on their FTP server is from 9/2000.)
/pub, /users, and several other directories.
2) This isn't 'hidden', it's linked off the Microsoft Research IPv6 homepage, which in turn is linked off of the Current Research page (Although they want you to register if you go in that way).
3) This isn't Microsoft, the OS maker, it's Microsoft Research, the R&D Lab.
4) Microsoft didn't write Fnord!, it specifically says on their page 'Fnord! is a Windows web server we found on the net and ported to run on our stack.' and 'Fnord! was apparently written by Brian Morin while a student at Worcester Polytechnic Institute.' The readme.msr goes on to give the URL of the authors website (now defunct) and a URL of a Tucows mirror where the original could still be found when they wrote the readme. (I didn't check if that mirror still works.)
5) Yes, they included the source code in their port. No, it's not for Visual C++. It's actually a Borland C++ 4.5 project, although I believe their binaries are compiled with VC++.
6) Finally, I'd like to point out that there are a lot of other interesting-sounding files available on ftp.research.microsoft.com, from the
-Jade E.
AMD isn't hyping this Transmeta deal any more than they have any of the other more major alliances (not alignments) they've developed. CNNfn isn't even really hyping this deal, even though they seem to immensely enjoy referring to Transmeta as an 'upstart'. In reality, the only things at work here are one CNNfn reporter's obsession with Transmeta (Seriously, try a search for 'Richard Richtmyer' at CNNfn. He's the author of this article, and 4 of his 6 articles this month so far are about Transmeta.) and the /. crowd's obsession with all things Linux. (If you don't know the connection between Transmeta and Linux, well, I can't help you.)
Anyways, nothing bad on AMD's part, they had no control over this.
-Jade E.
Thing is, I don't see why anyone would use this. In reality, the transfer rates aren't as good as a single fast dedicated server (I can easily get 75-80KB/sec on this line), so there's really no gain on the client end from using this. On the server end, yeah you're using a lot less bandwidth. This might be useful for open source projects or other products that are downloaded by knowledgable people, but you're average computer luser isn't going to want to download 6MB of Java Runtime/Swarmcast Client just so they can save your company money by getting slower-than-normal downloads. (And I shudder to think what a distributed p2p network comprised mostly of 56k modems would be like, at least at the moment it seems most of the users online are broadband.)
Anyways, pessimism aside, here's what I'd like to see:
OK, I'm done for the moment. If they play this right I can see it helping out smaller outfits with knowledgable users, maybe eventually even going mainstream if they can convince people it's worth their while to install. It's definitely sparse on information in it's current form, but hey, it's a beta. It's a good idea, and I wish them luck.
-Jade E.
It'd be nice if they bothered to add that this 'undocumented' and 'broken' practice is exactly how disk signatures and dynamic disks are implemented in Windows 2000. (And maybe previous versions of NT, I'm not sure)
-Jade E.
It's a great game.
-Jade E.
Hey everbody! Come listen to my new 'Net broadcast station! Use your favorite streaming client to listen at 192.168.0.1! 24 hour 24k encodes of your favorite music!
And, as an added bonus, if you've use *my* client, every 4 hours or so the encrypted data in the music stream gets saved into a high quality, watermark-free, 196k vbr encode of some of your favorite songs! Remember, it's not piracy if they never catch you!
-Jade E.
* IBM would be making free, open source software for Linux.
* Microsoft would use a RedHat package as a defense exhibit in an anti-trust trial.
* All the preparations for Y2K would be successful, and nothing would happen.
* AOL would merge with Time Warner, but only after buying CompuServe, Mirabilis, and Nullsoft.
And finally...
* Apple would be selling their own flavor of UNIX.
-Jade E.
Slashdot story: Self-Destructing DVDs: Son of DIVX
There's not much info there, but you can also check SpectraDisc's own site.
Scary, huh?
-Jade E.
On a different note, I'm not really worried about this at all. The new watermark is going to be dependent on hardware support. People don't want watermarks. So guess what's going to happen to the first manufacturer to enforce watermarks? That's right, nobody's going to buy their burner. And if they allow watermark detection, but don't enforce it at the hardware level, then you can damn well bet that somebody (Me, for example) is going to write a burning program that won't enforce it at the software level.
So, like I said, I'm not worried about it. Once consumer grade DVD burners catch up, I'll be able to backup my DVDs. (No, that's not a euphamism for pirating them.) And the pirates in China will still be able to make their bootleg copies, for whatever it's worth to them. All in all, this isn't really going to affect anybody, unless they go out and buy a watermark-enforcing piece of hardware. And even if they do, I bet it'll be a fun hack to remove the watermark detection from the circuit.
-Jade E.
Tcsh isn't the only shell that's been ported to Windows. But to answer your question, that's not the way Windows works (In recent versions, at least.) All apps interface with the kernel through it's DLL calls (OK, not *ALL* apps, but any that use the GUI.). There isn't really a 'shell', even Explorer (The standard 'desktop' you see when you boot) is just another App. When you run a command prompt (Using 9x's command.com, or 2000's much-improved cmd.exe), all you're really doing is invoking an appication that provides a minimal execution environment for command line apps (Those being the ones that can't [don't] directly access the kernel.) So, any app that can provide that environment can execute those apps. This includes things that act like command interpreters, such as shell ports (I know several people that swear by bash under 2k), but can also include full blown GUI apps that provide a 'hidden' environment for a command line app, providing input to the process and capturing it's output. (Sort of like a tremendous badly implemented subprocess pipe.)
So, there's you're offtopic answer.
Now, so this entire post isn't totally offtopic, as I'm writing this the kernel.org bandwodth meter is down to 30.11, so I guess the initial rush is over. I wonder how many people download the new kernels just to say they did, but never actually install them...
I dont know exactly what (all) methods they employ to detect attacks, but the University of Arizona is already using autonomous intrusion detection boxes. I do, however, know 2 things about them for sure:
1) When they detect intrusions, their response is to telnet to the edge router for whichever line the attack is coming through, and block the IP there, for increasingly longer periods.
2) They consider it an attack if you try to FXP a file to a server inside the U when both you and the source server are outside. This is, of course, how I first became aware of them.
The netadmin I know there tells me these boxen are called 'NetRangers', and we had a lengthy theoretical talk about how scary it is for autonomous devices to have exec access to your routers, and wondering whether they're smart enough to detect a constant barrage of packets with rotating forged sources before most of the internet is blocked at the routers.
This probably reminds you of the project from UoW's HITL because it *is* the project from UoW's HITL. It'd be nice if Microvision bothered to mention that. Oh well, at least the Dev team's page at UoW mentions that Microvision are the ones developing commercial applications for it. Also has an actual picture of the prototype, which isn't nearly as high-tech as the Microvision site would lead you to believe.