I back up my data to a dvd-rw and a dvd-r. This way I am sure that I am using two different types of disks. Since the actual manufacturer of disks is generally not known until you use some special program to read the disk, I don't trust different names on the disk to mean different chemical formulations.
I also have a backup linux box that use to mirror all of my data.
I also have a fireproof safe that I put backup hard drives in.
I store the hard drives in army ammo cans, as they are the only truly
airtight containers that I have found. If there is a fire, the corrossive gases could damage the hard drive, as fireproof safes are not airtight. I have also installed Pemco smoke-seal around the door of the safe. It expands greatly at 300 degrees F, and is designed to make the safe airtight.
In my experience, all ziplock bags 'breathe' slightly. As air pressure goes up and down, air leaks into the ziplock bag. After a few months, the ziplock bag is blown up like a pillow. I have tried several brands, including the freezer models.
I have a modest bunch of computers in my office.
Each one has a separate UPS. The 500-900VA ones are pretty cheap.
Each computer powers down the UPS when the UPS gets low on power.
I also have a separate UPS for my screen, because it used be a 19" CRT and suck
power. You want to only plug the stuff that really needs uninterruptable power.
If you use a power strip and a cheaper UPS, make sure the power strip doesn't
have a surge supressor built into it, as the 'modified sine waves' will toast
the surge supressor.
I also have some line-interactive sine synthesizing 1000va UPS's that I bought
surplus. Not the cheapo square wave or 'modified sine wave' UPS's.
All reasonably sized UPS that I have seen 1000-3000VA run on 48v. That means
four 12v batteries. They tend to run a bit hot and loud. Not a problem in a
datacenter, but perhaps overkill for a home office.
If anyone wants a 1000va sine synthesizing UPS that has failed for some reason,
I have one for sale. No schematic. Made by powerware.
I was the real software lead, despite what is on their web page. After I tried to explain that Z course was hard (follow a Z gps path, but with 2 trash cans placed by darpa that had to be avoided). I wrote soem basic gps parsers, but they decided not to use them.
The technical guy in charge, Pete, tried to get all the code from the net, and found some visual basic gps following code. They tried to update their servo's, but flashed a bad program, and as a result, at the site visit, the car didn't move a millimeter under its own power.
They also had (as far as I know) no sensors hooked up to the software to avoid the trashcans. Perhaps they wanted to run them over?
I tried to explain that the software was what kept anyone else from winning last year, as the hardware was likely up to the task. They didn't seem to understand that.
I told them I was withdrawing from the team, due to differences of opinion (like I wanted to test the software, but they thought that getting it to compile was some form of testing). I had unit tests for all of my code, but the test data I had from the GPS only had one satellite. (You need 3 satellites for lat & long, and you need 4 for elevation.)
I used to be a contractor at a government lab.
Around 2002 I ordered about 24 dell precision workstations running linux. These were high end machines, about half had dual processors, about half has 20" LCD screens ($4000 each).
Every one had the video card fail. It was a german company's NVIDA card. They all failed the same way. Dell replaced them as they failed with refurburshied cards. About 50% failed more than once. We had a motherboard fail after about a year. I seem to recall one hard drive failed.
We didn't have bluescreens as we were running linux. We did get compromised because the sysadmin didn't run a firewall but wanted to use tcp wrappers. I got a new sysadmin who put a firewall on each machine and there were no more problems.
I would guess that 5% hardware failure rate per year is pretty common, based on my 24 dell high end machines.
I have seen true optimism. I was the main programmer for another DARPA grand challange team. The main architect for the vehicle had no clue that writing the software needed to finish the course would be difficult. I wrote a GPS NMEA (datastream) parser as well as a RDDF (the DARPA supplied route file) parser. Even though my code worked fine, he grabbed and/or wrote the GPS code in Visual Basic.
When the site visit happened, the car could not move at all due to some poorly flashed code in a controller. Of course they had no spare controller. I had hoped to see the vehicle move, because the code simply seeked to the next waypoint. The DARPA guys were very polite, even though they were expecting to see the vehicle run the course.
For the site visit, you have to do a Z shaped route, and DARPA gets to put two trash cans on the route and the car has to avoid them. This vehicle has some sensors installed the week before, but I don't think they were hooked up to anything. I am sure the vehicle would have crashed into the trash cans, if it could follow a GPS path.
Since the team didn't seem to understand software (and system) design, testing, or small details like that, I decided to leave the team. They still think they will win, if not in 2005, in 2006.
I was picking my truck from an expensive repair
($6000 new engine ruined by a mechanic (their insurance paid for most of it)), and I paid with my
credit card. The clerk said I had exceeded my limit. I said no way.
I called the credit card company, and they said also said I exceeded my limit. They told me that on Friday (it was currently Monday) someone had purchased $4000 of computer stuff from a company I had never heard of. They asked if it was my purchase. I said no. They asked if another purchase on Friday at the same computer store for $3000 was mine. I said no. They asked about another purchase for $2000. I said I have never heard of that computer store. The card company seemed a bit confused.
I said that the transactions were fraud. When the heard the f-word, they quickly transferred me to the fraud department. They said they had to check something about the 3 transactions. They said something like "the merchant didn't do something, so we don't have to pay". I suspect that the stuff was to be shipped somewhere other than my address and the merchant didn't ask for the security code on the back of the card, but the credit card company wouldn't say.
They said they have just cancelled my card, and they are mailing me a new card. I said swell, but I have pick up my car. They said they would temporarily raise the credit limit by $6000, and while I was on the phone the clerk should run the card, then they would cancel the card. I told the clerk this (they had been listening since I was using their phone), and they ran the card which accepted the charge. The card was cancelled just after the transasction.
I never found out how the card number was stolen.
I did go to Bank of America had them take a picture of me and emboss it on the card. This is great when bying something in person, and completely useless over the phone or internet.
Slashdot Mirror
I back up my data to a dvd-rw and a dvd-r. This way I am sure that I am using two different types of disks. Since the actual manufacturer of disks is generally not known until you use some special program to read the disk, I don't trust different names on the disk to mean different chemical formulations.
I also have a backup linux box that use to mirror all of my data.
I also have a fireproof safe that I put backup hard drives in. I store the hard drives in army ammo cans, as they are the only truly airtight containers that I have found. If there is a fire, the corrossive gases could damage the hard drive, as fireproof safes are not airtight. I have also installed Pemco smoke-seal around the door of the safe. It expands greatly at 300 degrees F, and is designed to make the safe airtight.
In my experience, all ziplock bags 'breathe' slightly. As air pressure goes up and down, air leaks into the ziplock bag. After a few months, the ziplock bag is blown up like a pillow. I have tried several brands, including the freezer models.
I have a modest bunch of computers in my office. Each one has a separate UPS. The 500-900VA ones are pretty cheap. Each computer powers down the UPS when the UPS gets low on power. I also have a separate UPS for my screen, because it used be a 19" CRT and suck power. You want to only plug the stuff that really needs uninterruptable power. If you use a power strip and a cheaper UPS, make sure the power strip doesn't have a surge supressor built into it, as the 'modified sine waves' will toast the surge supressor. I also have some line-interactive sine synthesizing 1000va UPS's that I bought surplus. Not the cheapo square wave or 'modified sine wave' UPS's. All reasonably sized UPS that I have seen 1000-3000VA run on 48v. That means four 12v batteries. They tend to run a bit hot and loud. Not a problem in a datacenter, but perhaps overkill for a home office. If anyone wants a 1000va sine synthesizing UPS that has failed for some reason, I have one for sale. No schematic. Made by powerware.
Team Robomonster was cut.
I was the real software lead, despite what is on their web page. After I tried to explain that Z course was hard (follow a Z gps path, but with 2 trash cans placed by darpa that had to be avoided). I wrote soem basic gps parsers, but they decided not to use them.
The technical guy in charge, Pete, tried to get all the code from the net, and found some visual basic gps following code. They tried to update their servo's, but flashed a bad program, and as a result, at the site visit, the car didn't move a millimeter under its own power.
They also had (as far as I know) no sensors hooked up to the software to avoid the trashcans. Perhaps they wanted to run them over?
I tried to explain that the software was what kept anyone else from winning last year, as the hardware was likely up to the task. They didn't seem to understand that.
I told them I was withdrawing from the team, due to differences of opinion (like I wanted to test the software, but they thought that getting it to compile was some form of testing). I had unit tests for all of my code, but the test data I had from the GPS only had one satellite. (You need 3 satellites for lat & long, and you need 4 for elevation.)
I used to be a contractor at a government lab. Around 2002 I ordered about 24 dell precision workstations running linux. These were high end machines, about half had dual processors, about half has 20" LCD screens ($4000 each). Every one had the video card fail. It was a german company's NVIDA card. They all failed the same way. Dell replaced them as they failed with refurburshied cards. About 50% failed more than once. We had a motherboard fail after about a year. I seem to recall one hard drive failed. We didn't have bluescreens as we were running linux. We did get compromised because the sysadmin didn't run a firewall but wanted to use tcp wrappers. I got a new sysadmin who put a firewall on each machine and there were no more problems. I would guess that 5% hardware failure rate per year is pretty common, based on my 24 dell high end machines.
I have seen true optimism. I was the main programmer for another DARPA grand challange team. The main architect for the vehicle had no clue that writing the software needed to finish the course would be difficult. I wrote a GPS NMEA (datastream) parser as well as a RDDF (the DARPA supplied route file) parser. Even though my code worked fine, he grabbed and/or wrote the GPS code in Visual Basic.
When the site visit happened, the car could not move at all due to some poorly flashed code in a controller. Of course they had no spare controller. I had hoped to see the vehicle move, because the code simply seeked to the next waypoint. The DARPA guys were very polite, even though they were expecting to see the vehicle run the course.
For the site visit, you have to do a Z shaped route, and DARPA gets to put two trash cans on the route and the car has to avoid them. This vehicle has some sensors installed the week before, but I don't think they were hooked up to anything. I am sure the vehicle would have crashed into the trash cans, if it could follow a GPS path.
Since the team didn't seem to understand software
(and system) design, testing, or small details like that, I decided to leave the team. They still think they will win, if not in 2005, in 2006.
I was picking my truck from an expensive repair ($6000 new engine ruined by a mechanic (their insurance paid for most of it)), and I paid with my credit card. The clerk said I had exceeded my limit. I said no way.
I called the credit card company, and they said also said I exceeded my limit. They told me that on Friday (it was currently Monday) someone had purchased $4000 of computer stuff from a company I had never heard of. They asked if it was my purchase. I said no. They asked if another purchase on Friday at the same computer store for $3000 was mine. I said no. They asked about another purchase for $2000. I said I have never heard of that computer store. The card company seemed a bit confused.
I said that the transactions were fraud. When the heard the f-word, they quickly transferred me to the fraud department. They said they had to check something about the 3 transactions. They said something like "the merchant didn't do something, so we don't have to pay". I suspect that the stuff was to be shipped somewhere other than my address and the merchant didn't ask for the security code on the back of the card, but the credit card company wouldn't say.
They said they have just cancelled my card, and they are mailing me a new card. I said swell, but I have pick up my car. They said they would temporarily raise the credit limit by $6000, and while I was on the phone the clerk should run the card, then they would cancel the card. I told the clerk this (they had been listening since I was using their phone), and they ran the card which accepted the charge. The card was cancelled just after the transasction.
I never found out how the card number was stolen. I did go to Bank of America had them take a picture of me and emboss it on the card. This is great when bying something in person, and completely useless over the phone or internet.