Another example is from Call of Duty there was a flaw with running MSN Messenger that would allow you to see through walls and other textured areas of the levels.
Okay, for any game designers who have been hiding under a rock for the past decade (which sadly seems like a lot of them) here's online game design rule #1:
If the player's computer knows something, expect that the player knows it too. If you design the game so that the player's computer knows something before the player should then you are practically begging cheaters to ruin the game.
And rule #2 is probably that anything which depends heavily on the player's dexterity begs for cheaters as well. "Aimbots" is the cited example -- cheat programs that aim for you. You can't actually prevent this. Code integrity checks? Fine, intercept the driver. Driver integrity checks? Fine, run it in a virtual machine and run the bot outside of the vm.
Seriously, complaining about this and calling folks cheaters is like dropping $20 on the street and complaining about thieves when you go back and find it gone. Of course its gone. Duh.
he has every right to refer you to ask again after class or in recitation or in office hours.
Works for me. Just don't get bent out of shape that I interrupted long enough to say, "Hey, at least one of your students hasn't understood what you're saying." Remember, your students are there to learn. If they didn't want to learn they wouldn't have shown up for the lecture. If your planned lecture fails to teach them then you need a new plan and shame on you for shooting the messenger.
As a comparative measure, you weren't graded on equal footing because you had extra time. Welcome to real life.
I should mention that all of that happened more than a decade ago. I've been a part of "real life" for quite some time now. Guess what? Half the work has no fixed deadline and for the half that does 99% of the deadlines are soft. When something has to be pushed back a day or two days or a week because that's what it takes to have a working product, its simply not a problem.
But again, you miss the point: real life comparisons are irrelevant. The student's grade has one purpose and one purpose only: Its a measure of the student's mastery of the subject.
See, here's the thing: when students walk in to your course they have at least a modicum of respect for you. Some have more, some have less but to every student you're the professor. They don't start out with contempt; that has to be earned.
When you make the grade about timeliness or the neatness or whether the student looked at you crosseyed then you have lost sight of the purpose of assigning a grade. You earn the disrespect that the students show you.
There's a correlary to my experience with the problem professor. A couple years later I was inspired by one of the problems offered in a similar course. I delved into it far beyond the assignment. I must have burned dozens of hours experiementing and searching out datasets I could test against. I ended up turning in the whole expansive thing weeks late. The professor's response? An A for the project, an A for the course.
I made sure to effusively thank him for giving me the latitude and I have tremendous respect for him to this day. I still remember and use the skills he taught me. He was one of the professors that made a difference.
Is he supposed to stop lecture and debate every point with anyone who has a question, even if derails the course syllabus?
Well, yes. That's the whole point of having a professor and a classroom -- I can ask him to discuss particular points more exhaustively until I understand. That includes a degree of debate ("But why isn't this answer correct instead?").
If he's just going to stand there and lecture then why should I bother showing up for class? What have I gained that I couldn't have gotten from reading the book?
I had probably my most infuriating experience in college when I turned in a difficult programming assignment a day late and got a C. It worked perfectly. Classmates who turned in programs on time which compiled but crashed immediately received B's. That professor deserved a reminder that he was there for my convenience, not the other way around. He made it a question of authority instead of a question of learning.
As far as I'm concerned a student's grade leaving a class should accurately reflect his knowledge of the subject. If any choice the professor makes impedes that process then the professor has failed at that part of his job.
Got news for you: Morale is not a rational thing. Its a question of hope. Folks like to believe that good things are in store for them if they work hard and do a good job. A layoff is the single most capricious thing you can do to an employee. Its deadly to morale in any form.
But that misses the point of my original post which is this: you don't lay off the founder. It just isn't done. The founder is part of the corporate story, the history you tell the customers. You know, marketing. Sacking him (or laying him off) looks incredibly bad; its a serious breach in the corporate continuity.
No single employee costs so much that you can't keep him on the payroll. You only ask the founder to leave if he's getting in the way and after a few private discussions he's still getting in the way and distracting the other employees from the corporate focus. That's not a lay-off. That's fired for cause.
IMHO, the CEO of Mandriva is lying though his teeth when he says, "Its just a lay-off."
I'm sorry but the founder is not laid off. He quits if he tires of the company's direction or he's fired if he becomes an obstacle but he's not laid off. It's a question of morale: If the founder himself is of so little value that he can be laid off then every other employee is worthless too. When your employer shows they don't value your presence its past time to jump ship.
The real question is: When are you going to release a set of patches for Linux 2.6?
The openwall patches for 2.4 do the following three really useful things. Hardware compatibility is pushing me to 2.6 but I'd sure like to have the patches:
Non-executable stack (defeats most buffer-overflow attacks) Restricted links and fifos in/tmp Restricted/proc
Pretty much anything on the subject will do. Stuff published by O'Reilly is generally good or at least ok. The point is not to gain deep insight -- that's a moving target that comes from experience. The point is get a quick overview of the breadth of the field.
As with most things involving deep technical expertise, you don't choose the career so much as the career chooses you. Here's how it goes for network security:
You work as a junior network administrator. You get interested in the security aspects. You find you have a knack for it and tend to spend any unassigned manhours scanning logs for connection attempts and looking up the ports to see what the originator was attempting. Your boss notices that you have a knack for it and lets you spend more time working on it. You start reading the available literature to gain more insight. A job comes along where they're looking for a network security specialist instead of a general network admin. You apply and get the job. With all of your work-hours spent on network security your rate of learning increases. You run in to a few unusual situations and start to consult with experts on the 'net. etc. At some point you cross a line. Now you are one the experts and folks consult with you.
You'll notice there is no coursework listed anywhere in there. It wasn't an oversight. Coursework provides a decent overview for folks who don't have the knack. It lets them get by without being completely ignorant. Someone with the knack, someone who should consider network security as a career path, will get the same results by spending an evening with a book.
Which would make it pretty easy for me to figure out which of my sysadmins screwed up the server, wouldn't it? (Hint: You.)
Alternate methods
on
Sudo vs. Root
·
· Score: 3, Informative
I ran in to these kinds of issues back in the Solaris 2.2 days and came up with a different solution.
Solaris' problems were even more acute. Sudo was a download; it didn't come with the system. If you changed root's shell from the minimal Bourne shell the boot scripts would malfunction. More, root's home directory was "/". So setting up a personalized environment where you could use root access effectively was a pain.
The solution I came up with was a second root account. I just added another name with uid 0 using a seperate password, a seperate home directory and the ksh shell. Then I randomized the main root password, stored it away and promptly forgot it. I'd only need it for fsck on boot.
Later when I was in charge of multiple system administrators I gave each one their own root account. This let them set up their environment in a way that worked for them, it showed me who was using root commands when and it logged their commands to individual.bash_historys so I could see who screwed up.
It also means that like with sudo when a sysadmin leaves I don't have to change all the passwords. I just delete their account.
I still use sudo for folks who I don't expect to do much as root, but the sysadmins get their own root account.
They're going about it all wrong. If they want.xxx to fly, they should require the ICANN to create one and pass a law affirming that if a web site is only accessible via its.xxx name then the site operator is deemed to have taken adequate care to prevent access by minors.
Then let the individual site operators decide whether they want the liability shield. Guess what? They want it. And if that means they elementary schools will have an easy time blocking access I guarantee they won't shed a tear.
I use an old IBM keyboard with a solid click to the keys. The tactile feedback lets me know that the key has been pressed so I can release pressure before the key bottoms out.
Whenever I try to use the more recent el-cheapo keyboards I find that I constantly bottom the keys forcing my fingers to a full stop while the muscle is still taut. The impact on my hands is uncomfortable.
What throws me is going straight from a first-person-shooter to driving my car. For the first few minutes my reflexes feel all wrong. I keep wanting to move like I do in the game and realize just before the action kicks in that I can't do that.
Also included: a standby spare so that when that much heat in that small a package burns itself out in six months you don't have to wait for a replacement. Supermicro is generally pretty good but packing a kilowatt of consumption into a 1U package is about as smart as running your home PC in the oven set on bake.
Maybe I'm just sore because I've spent the last few weeks identifying the bad ram in last year's opteron rackmounts from Penguin. 2 gig ECC dimms and I'm seeing a 40% failure rate from multiple manufacturers. They stacked the damn chips one on top of another. There's no where for the heat to go. Of course they're going to fail.
I programmed one of my servers to notice if I'm logged on to the computer at work. If I am then it relaxes the temperature setpoint. When it sees me log off it returns the setpoint to normal so that the house is comfortable when I get home.
I don't have to do anything special; it simply notices whether I'm online at work or not and reacts accordingly.
I've broken the DRM on my ebooks and put them in a secure location on web site but my blackberry isn't fast enough. It can't hold the entire book and it takes a minute or more to download the next chapter. Imagine the page in a book refusing to turn for 60 seconds. Its just too darn slow!
I have AT&T/Cingular. I've tried Verizon and Nextel and they seem faster but still not fast enough.
Other than that, I'm happy. I'd much rather read the ebook. So I guess its really three things:
1. No DRM! I won't buy your book unless I know I can read it the way I feel like reading it. If I know I can crack your DRM I'll buy it but then why would you bother? If I can't crack your DRM I'll just wait for someone to scan it to a.txt and then I'll read it. You won't get paid then but that's your problem.
Ironic, isn't it. You treat me like you suspect I'll steal your book and - cause and effect - I do exactly that as a result.
2. Convertable format. Whatever format you give it to me in, its wrong. You can give it to me in exactly the format I specify and next week I'll decide I like something else. Don't worry though, I'll convert it to a format I like. Just make sure one of your formats is a.txt or a.rtf that's an easy source to convert from.
I should mention that BAEN Books (http://www.webscription.net/) passes both of these requirements with flying colors. Kudos to a publisher that gets it. I've spent hundreds of dollars there.
3. A blackberryish device that's fast enough to keep up. I don't want a special bookreader that I have to download books to. I want my phone/email/pda device that I take with me everywhere to serve as a reader. I want it to connect to my home server and fetch whatever book I feel like reading and I want it to happen fast.
Its not a hybrid drive, but the clever one I saw was a hard disk-backed 2 gig "solid state" drive. It was a board in a 1U rackmount box with 2 gigs of regular ram, a battery and a 2 gig hard disk.
All reads and writes are to the ram. The hard disk isn't even running. When power is lost for more than 15 seconds, the hard disk spins up and the contents of the ram are written out to disk. When power comes back on, the drive spins up, the contents of the disk are loaded back in to ram and operation resumes.
I got some liquid latex from Home Depot for a similar problem. They talk about using it to coat tool handles, but it works great for coating bare electronics too.
I had all kinds of trouble using Vonage yesterday morning. Voicemails were not forwarded to my email and logins to their web page timed out. But that was a Vonage problem not an ISP problem -- I tried from different providers with the same result.
Can't all of these web sites exclude their material from Google by using the industry standard robots.txt file? I know that doesn't apply to the book search, but it certainly applies to Perfect 10's web sites. If so, why is there any legal challenge to Google's web site search functions?
1 Quad Xeon P3, Debian Linux, Sendmail server 1 Dual P3, Debian Linux, Web server 1 Athlon, Debian Linux, DNS and misc. server 1 Celeron, Debian Linux -- sentimental reasons, its in a custom wood-grain AT tower case and nothing later will fit in an AT case. 2 Single P3s, Debian Linux, firewall/routers for my two broadband links (1 Cox Business Cable @ 5mbps + 17 IPs, 1 Verizon Residential Fios @ 15 mbps). I used to do it with a single machine but Linux 2.4's policy-based routing gets some of the corner-cases wrong. 1 Sparc Ultra 5, Solaris, hosting for a friend. 1 Pentium Pro, Debian Linux, legacy web server whose software malfunctions on more current machines 2 3-com managed switches 1 Vonage phone adapter
Powered sometimes: 1 P4 2.8ghz, Windows 2000 (games and console) 1 Dual P3, used for bench-testing hard drives at the moment
Haven't been powered in a while: 1 Eight 85 mhz Sparcserver 1000 maxed out 1 Quad 40 mhz Sparcserver 670 with 4 Seagate 8-inch IPI hard drives running Solaris 2.5.1. 3 Sparc 20s 1 Sparc IPX 1 Sparc IPC 2 Old macs 2 Commodore 64s 1 Atari 800 Various Cisco routers Various game consoles
And positively no garden gnomes. What kind of a freak has garden gnomes?
Slashdot Mirror
Another example is from Call of Duty there was a flaw with running MSN Messenger that would allow you to see through walls and other textured areas of the levels.
Okay, for any game designers who have been hiding under a rock for the past decade (which sadly seems like a lot of them) here's online game design rule #1:
If the player's computer knows something, expect that the player knows it too. If you design the game so that the player's computer knows something before the player should then you are practically begging cheaters to ruin the game.
And rule #2 is probably that anything which depends heavily on the player's dexterity begs for cheaters as well. "Aimbots" is the cited example -- cheat programs that aim for you. You can't actually prevent this. Code integrity checks? Fine, intercept the driver. Driver integrity checks? Fine, run it in a virtual machine and run the bot outside of the vm.
Seriously, complaining about this and calling folks cheaters is like dropping $20 on the street and complaining about thieves when you go back and find it gone. Of course its gone. Duh.
he has every right to refer you to ask again after class or in recitation or in office hours.
Works for me. Just don't get bent out of shape that I interrupted long enough to say, "Hey, at least one of your students hasn't understood what you're saying." Remember, your students are there to learn. If they didn't want to learn they wouldn't have shown up for the lecture. If your planned lecture fails to teach them then you need a new plan and shame on you for shooting the messenger.
As a comparative measure, you weren't graded on equal footing because you had extra time. Welcome to real life.
I should mention that all of that happened more than a decade ago. I've been a part of "real life" for quite some time now. Guess what? Half the work has no fixed deadline and for the half that does 99% of the deadlines are soft. When something has to be pushed back a day or two days or a week because that's what it takes to have a working product, its simply not a problem.
But again, you miss the point: real life comparisons are irrelevant. The student's grade has one purpose and one purpose only: Its a measure of the student's mastery of the subject.
See, here's the thing: when students walk in to your course they have at least a modicum of respect for you. Some have more, some have less but to every student you're the professor. They don't start out with contempt; that has to be earned.
When you make the grade about timeliness or the neatness or whether the student looked at you crosseyed then you have lost sight of the purpose of assigning a grade. You earn the disrespect that the students show you.
There's a correlary to my experience with the problem professor. A couple years later I was inspired by one of the problems offered in a similar course. I delved into it far beyond the assignment. I must have burned dozens of hours experiementing and searching out datasets I could test against. I ended up turning in the whole expansive thing weeks late. The professor's response? An A for the project, an A for the course.
I made sure to effusively thank him for giving me the latitude and I have tremendous respect for him to this day. I still remember and use the skills he taught me. He was one of the professors that made a difference.
Is he supposed to stop lecture and debate every point with anyone who has a question, even if derails the course syllabus?
Well, yes. That's the whole point of having a professor and a classroom -- I can ask him to discuss particular points more exhaustively until I understand. That includes a degree of debate ("But why isn't this answer correct instead?").
If he's just going to stand there and lecture then why should I bother showing up for class? What have I gained that I couldn't have gotten from reading the book?
I had probably my most infuriating experience in college when I turned in a difficult programming assignment a day late and got a C. It worked perfectly. Classmates who turned in programs on time which compiled but crashed immediately received B's. That professor deserved a reminder that he was there for my convenience, not the other way around. He made it a question of authority instead of a question of learning.
As far as I'm concerned a student's grade leaving a class should accurately reflect his knowledge of the subject. If any choice the professor makes impedes that process then the professor has failed at that part of his job.
The non-executable stack is in 2.6.xx already.
/proc with an LSM too.
Then why does the stacktest.c program from openwall succeed in simulating a buffer overflow in SuSE Enterprise 9 with kernel 2.6.15.6?
You can restrict
Yeah? Which?
Got news for you: Morale is not a rational thing. Its a question of hope. Folks like to believe that good things are in store for them if they work hard and do a good job. A layoff is the single most capricious thing you can do to an employee. Its deadly to morale in any form.
But that misses the point of my original post which is this: you don't lay off the founder. It just isn't done. The founder is part of the corporate story, the history you tell the customers. You know, marketing. Sacking him (or laying him off) looks incredibly bad; its a serious breach in the corporate continuity.
No single employee costs so much that you can't keep him on the payroll. You only ask the founder to leave if he's getting in the way and after a few private discussions he's still getting in the way and distracting the other employees from the corporate focus. That's not a lay-off. That's fired for cause.
IMHO, the CEO of Mandriva is lying though his teeth when he says, "Its just a lay-off."
Gael was not fired. He was laid off.
I'm sorry but the founder is not laid off. He quits if he tires of the company's direction or he's fired if he becomes an obstacle but he's not laid off. It's a question of morale: If the founder himself is of so little value that he can be laid off then every other employee is worthless too. When your employer shows they don't value your presence its past time to jump ship.
The real question is: When are you going to release a set of patches for Linux 2.6?
/tmp /proc
The openwall patches for 2.4 do the following three really useful things. Hardware compatibility is pushing me to 2.6 but I'd sure like to have the patches:
Non-executable stack (defeats most buffer-overflow attacks)
Restricted links and fifos in
Restricted
what are some of the books?
Pretty much anything on the subject will do. Stuff published by O'Reilly is generally good or at least ok. The point is not to gain deep insight -- that's a moving target that comes from experience. The point is get a quick overview of the breadth of the field.
As with most things involving deep technical expertise, you don't choose the career so much as the career chooses you. Here's how it goes for network security:
You work as a junior network administrator.
You get interested in the security aspects.
You find you have a knack for it and tend to spend any unassigned manhours scanning logs for connection attempts and looking up the ports to see what the originator was attempting.
Your boss notices that you have a knack for it and lets you spend more time working on it.
You start reading the available literature to gain more insight.
A job comes along where they're looking for a network security specialist instead of a general network admin. You apply and get the job.
With all of your work-hours spent on network security your rate of learning increases.
You run in to a few unusual situations and start to consult with experts on the 'net.
etc.
At some point you cross a line. Now you are one the experts and folks consult with you.
You'll notice there is no coursework listed anywhere in there. It wasn't an oversight. Coursework provides a decent overview for folks who don't have the knack. It lets them get by without being completely ignorant. Someone with the knack, someone who should consider network security as a career path, will get the same results by spending an evening with a book.
I shoulda gotten a patent. Then I could sue all those independent inventors. :P
Which would make it pretty easy for me to figure out which of my sysadmins screwed up the server, wouldn't it? (Hint: You.)
I ran in to these kinds of issues back in the Solaris 2.2 days and came up with a different solution.
.bash_historys so I could see who screwed up.
Solaris' problems were even more acute. Sudo was a download; it didn't come with the system. If you changed root's shell from the minimal Bourne shell the boot scripts would malfunction. More, root's home directory was "/". So setting up a personalized environment where you could use root access effectively was a pain.
The solution I came up with was a second root account. I just added another name with uid 0 using a seperate password, a seperate home directory and the ksh shell. Then I randomized the main root password, stored it away and promptly forgot it. I'd only need it for fsck on boot.
Later when I was in charge of multiple system administrators I gave each one their own root account. This let them set up their environment in a way that worked for them, it showed me who was using root commands when and it logged their commands to individual
It also means that like with sudo when a sysadmin leaves I don't have to change all the passwords. I just delete their account.
I still use sudo for folks who I don't expect to do much as root, but the sysadmins get their own root account.
They're going about it all wrong. If they want .xxx to fly, they should require the ICANN to create one and pass a law affirming that if a web site is only accessible via its .xxx name then the site operator is deemed to have taken adequate care to prevent access by minors.
Then let the individual site operators decide whether they want the liability shield. Guess what? They want it. And if that means they elementary schools will have an easy time blocking access I guarantee they won't shed a tear.
I use an old IBM keyboard with a solid click to the keys. The tactile feedback lets me know that the key has been pressed so I can release pressure before the key bottoms out.
Whenever I try to use the more recent el-cheapo keyboards I find that I constantly bottom the keys forcing my fingers to a full stop while the muscle is still taut. The impact on my hands is uncomfortable.
What throws me is going straight from a first-person-shooter to driving my car. For the first few minutes my reflexes feel all wrong. I keep wanting to move like I do in the game and realize just before the action kicks in that I can't do that.
Also included: a standby spare so that when that much heat in that small a package burns itself out in six months you don't have to wait for a replacement. Supermicro is generally pretty good but packing a kilowatt of consumption into a 1U package is about as smart as running your home PC in the oven set on bake.
Maybe I'm just sore because I've spent the last few weeks identifying the bad ram in last year's opteron rackmounts from Penguin. 2 gig ECC dimms and I'm seeing a 40% failure rate from multiple manufacturers. They stacked the damn chips one on top of another. There's no where for the heat to go. Of course they're going to fail.
I programmed one of my servers to notice if I'm logged on to the computer at work. If I am then it relaxes the temperature setpoint. When it sees me log off it returns the setpoint to normal so that the house is comfortable when I get home.
I don't have to do anything special; it simply notices whether I'm online at work or not and reacts accordingly.
I've broken the DRM on my ebooks and put them in a secure location on web site but my blackberry isn't fast enough. It can't hold the entire book and it takes a minute or more to download the next chapter. Imagine the page in a book refusing to turn for 60 seconds. Its just too darn slow!
.txt and then I'll read it. You won't get paid then but that's your problem.
.txt or a .rtf that's an easy source to convert from.
I have AT&T/Cingular. I've tried Verizon and Nextel and they seem faster but still not fast enough.
Other than that, I'm happy. I'd much rather read the ebook. So I guess its really three things:
1. No DRM! I won't buy your book unless I know I can read it the way I feel like reading it. If I know I can crack your DRM I'll buy it but then why would you bother? If I can't crack your DRM I'll just wait for someone to scan it to a
Ironic, isn't it. You treat me like you suspect I'll steal your book and - cause and effect - I do exactly that as a result.
2. Convertable format. Whatever format you give it to me in, its wrong. You can give it to me in exactly the format I specify and next week I'll decide I like something else. Don't worry though, I'll convert it to a format I like. Just make sure one of your formats is a
I should mention that BAEN Books (http://www.webscription.net/) passes both of these requirements with flying colors. Kudos to a publisher that gets it. I've spent hundreds of dollars there.
3. A blackberryish device that's fast enough to keep up. I don't want a special bookreader that I have to download books to. I want my phone/email/pda device that I take with me everywhere to serve as a reader. I want it to connect to my home server and fetch whatever book I feel like reading and I want it to happen fast.
At my home depot it was in these thin yellow cans in the "secure" tools section. I forget the brand name but they had blue, yellow and red.
Its not a hybrid drive, but the clever one I saw was a hard disk-backed 2 gig "solid state" drive. It was a board in a 1U rackmount box with 2 gigs of regular ram, a battery and a 2 gig hard disk.
All reads and writes are to the ram. The hard disk isn't even running. When power is lost for more than 15 seconds, the hard disk spins up and the contents of the ram are written out to disk. When power comes back on, the drive spins up, the contents of the disk are loaded back in to ram and operation resumes.
I got some liquid latex from Home Depot for a similar problem. They talk about using it to coat tool handles, but it works great for coating bare electronics too.
I had all kinds of trouble using Vonage yesterday morning. Voicemails were not forwarded to my email and logins to their web page timed out. But that was a Vonage problem not an ISP problem -- I tried from different providers with the same result.
Okay, so they're removed from the infringing site and one crawl later they're removed from google. What's the problem? Why is it in court?
Can't all of these web sites exclude their material from Google by using the industry standard robots.txt file? I know that doesn't apply to the book search, but it certainly applies to Perfect 10's web sites. If so, why is there any legal challenge to Google's web site search functions?
Normally powered on:
1 Quad Xeon P3, Debian Linux, Sendmail server
1 Dual P3, Debian Linux, Web server
1 Athlon, Debian Linux, DNS and misc. server
1 Celeron, Debian Linux -- sentimental reasons, its in a custom wood-grain AT tower case and nothing later will fit in an AT case.
2 Single P3s, Debian Linux, firewall/routers for my two broadband links (1 Cox Business Cable @ 5mbps + 17 IPs, 1 Verizon Residential Fios @ 15 mbps). I used to do it with a single machine but Linux 2.4's policy-based routing gets some of the corner-cases wrong.
1 Sparc Ultra 5, Solaris, hosting for a friend.
1 Pentium Pro, Debian Linux, legacy web server whose software malfunctions on more current machines
2 3-com managed switches
1 Vonage phone adapter
Powered sometimes:
1 P4 2.8ghz, Windows 2000 (games and console)
1 Dual P3, used for bench-testing hard drives at the moment
Haven't been powered in a while:
1 Eight 85 mhz Sparcserver 1000 maxed out
1 Quad 40 mhz Sparcserver 670 with 4 Seagate 8-inch IPI hard drives running Solaris 2.5.1.
3 Sparc 20s
1 Sparc IPX
1 Sparc IPC
2 Old macs
2 Commodore 64s
1 Atari 800
Various Cisco routers
Various game consoles
And positively no garden gnomes. What kind of a freak has garden gnomes?