you know, you can justify not sending some stuff out for open bid. Sounds like you could have had a chance at getting permission to do a sole source.
It's hard these days to do a sole source procurement - and I think it should be. Doesn't always make things go the way I want them to, but I think the rule is a good one.
Working for a computer and IT services vendor to the Fed. Govt. I was always mystified by the procurement procedures. As you put it so well, I did learn the key was in writing the specs. I also learned part of the key was helping my Govt. counterparts write the specs.
That's a dangerous game if your company has any financial interest in the contract - but you're right, it's all in the statement of work or RFQ. I wrote a SOW for a cell phone contract once that required coverage in 35 little towns - about a third of them didn't even have post offices. My preferred provider lost the bid but so far the company that won the bid has provided pretty good service.
You know, to be honest I didn't think of that. The end-of-year feeding frenzy around here means they call you at noon and want you to spend a quarter of a million bucks before you go home for the day - with vendor sources and full justification as to why you need the hardware they wouldn't let you have all year long.
I work for an agency under DoD. I had some end of year fundage to spend so I decided to buy some printers and networkable scanners - about $100k worth of gadgets. I'd requested HP hardware because that's pretty much what the infrastructure here is geared to support.
Because of the size of the contract award the thing went out for open bid - and I was contacted by another printer manufacturer. I won't tell you their name, but their initials are L-e-x-m-a-r-k.
Strongarm tactics ensue. First the local contracting office asked me to define printing and scanning requirements as the Other Printer Company believes they can meet my requirements at a lower cost - but we won't mention the fact that all the supplies I have in stock are from the Printer Company I Wanted To Use and adding another hardware vendor would be a logistical nightmare.
So - starting with the network scanners I start looking at hardware specs. The Other Printer Company says they can meet my requirements, but since a digital sender is an input device as opposed to an output device, I would have to get the new hardware certifiied by the network spies and I don't have time to do that, so for that part of the procurement I got the hardware I requested.
The printers were another matter. Once you've specified dpi, print speed and networking capabilities you've pretty much got to go with whoever brings the lowest bid - so the Other Printer Company won that.
During the acquisition process I felt like I was being strongarmed by the Other Printer Company and since I couldn't give a good reason not to use their hardware I have to use it. If I'd have had a week instead of a day to process the procurement I probably could have.
I have learned that I need to fine-tune my hardware requirements to keep it from happening again - but manufacturers can and will sue the government for buying from somebody else.
From TFA...
In particular, it said that: "for equivalent open source and Microsoft Windows systems, the open source system will require less memory and a slower processor speed for the same functionality." It points out that Windows needs a hardware refresh every three to four years, while a Linux box might need a new platform as infrequently as every six to eight years - half as often.
As I said in the subject line, rubbish.
I have three identical machines in my home office - all three are Dell Precision Workstation 360 machines with 2.8GHz processors, 1GB o'memory and identical 36GB 10k rpm U320 SCSI hard drives. Two of them run Windows XP with all patches and one runs CentOS 4.4 - again, with all patches. Window manager is Gnome - which for me is a good balance between the bloatware that is KDE and my favorite but frequently irritating window manager, XFCE4.
The Linux box is *not* faster than the Windows boxes. Granted, the Windows machines mainly just surf the web or play music or store pictures and the Linux box acts as a mail/web server for all of three users, but the machine boots slower, applications load slower and the whole machine is less responsive.
Gnome will bring up a desktop about five seconds slower than XFCE will, which is pretty darned speedy - but counting the time it takes me to boot both machines, log on (and type startx on the Linux box) I have a working desktop quicker in Windows than I do in Linux.
Now I will say that while I do have a working desktop on the Windows box there are still services starting in the background, but on identical hardware all applications are more responsive on the Redmond machines than on the CentOS box.
I guess the Linux box would last 8 years - if I never patched it.
1) No high-end AMD systems (their instruction set is better supported on Linux)
2) High-end systems do not provide an option for high-end nVidia cards (their drivers work better on linux)
3) All high-end systems require purchase of MS Windows Last fall I bought three off-lease Dell Precision Workstations from Dell Financial Services All were 2.8GHz boxes with 1GB RAM, 36G SCSI hard drives and Nvidia Quadro graphics cards. Paid ~$400 for each of them, loaded CentOS 4 on one and Windows XP on the other two.
Even with the included Windows XP CD I had to download and install SCSI drivers before XP would install - CentOS just worked. I did have to install Nvidia's driver on all three boxes to get to my native 1680x1050 resolution and I have to reinstall the Linux driver with every kernel upgrade, but Nvidia's got a shell script that makes that a no-brainer. I think the driver install in Linux is easier than the one in Windows.
This sounds very exciting, but it left me wonder that if there is a simple answer to fixing cancer, why hasn't nature evolved a way to apply it on its own, the same way out bodies raise temperature to fight infections, or eject bowl content when poisoned. Cancer, after all, is quite common. Just my opinion but it's because cancer is of itself a failure of the host immune system.
"Cancer" is a catchall term that describes more than 100 different diseases that share some or all of the same characteristics - namely cellular malformation that's not corrected by the host immune system, rapid cellular division and a particularly interesting trait called angiogenesis - the ability to create its own blood supply. You're born with all the blood vessels you'll ever have - and the only time you grow new ones if if you have cancer or are pregnant.
Interestingly, Thalidomide (yeah, the sleeping pill that caused all the birth defects in the early '60s) is currently seeing a fair bit of use as an antiangiogenetic agent - the same drug that causes hideous birth defects by suppressing blood vessel formation is now having limited success as an anticancer agent. I have a couple of internet friends doing Thalidomide right now - they have to sign a statement saying they understand the risk, that they will take steps to keep from becoming pregnant and the drug is packaged with scary pictures of a baby in a red circle with a slash through it.
If this treatment or some other new treatment did in fact cure many cancers inexpensively, it would be a nightmare come true for the chemotherapy industry.
How big is that industry, something like $40 billion a year? Wow...
The vast majority of chemotherapy costs are *not* drug costs.
Case in point: My spousal unit has Stage IV breast cancer and has had for almost eight years. She's currently on her sixth course (not round, course) of chemotherapy and is doing pretty well, thankyouverymuch;-)
Let's talk about drug costs. Doxorubicin (trade name Adriamycin) is generally accepted first line chemotherapy for breast cancer in combination with another drug, Cytoxan. In some parts of the world they use Epirubicin instead, but I digress.
Wholesale drug cost for a single dose of Adriamycin is about $300. Considering the drug is toxic as hell and requires special handling and disposal protocol at the manufacturer level that seems reasonable. Cost to administer? Seven thousand bucks - I can show you the bill. Since it's administered as an IV push over a couple hours it takes a little more work than other drugs but in a healthcare system that charges $180 to start an IV (doesn't matter that the spousal unit has a mediport in her chest) or $40 for a liter of normal saline for that IV it's not difficult to see where the markup is.
Don't get me wrong, I have no real love for drug companies - the two most widely prescribed post-chemotherapy antinausea meds are Zofran and Kytril. Zoftan has an average wholesale price (AWP) of $26.25 _per_tablet_. Kytril is even better, at $59.67 per pill. They normally give you Kytril if Zofran doesn't work. You pay or you puke for a couple days - your choice.
The drug that's keeping the spousal unit alive right now is called Herceptin and costs about $48k per year. Reasonably new monoclonal antibody made from the ovaries of Chinese hamsters. Guess $1000 per dose is reasonable since you probably gotta grind up a lot of hamsters to get enough to be useful.
But again, it costs much more to give the IV than the drug costs.
I got sued by our local cancer center because my insurance company decided to play games one month. Since the patient is responsible for medical charges in order for the hospital to collect they have to sue you and name the insurance company as a codefendant. One month of treatment - which included two rounds of chemotherapy, two 15 minute doctor visits, probably eight blood tests and ten days worth of a drug called Neupogen they used to give you to stimulate white blood count?
Glad you asked. $39,000. Thirty Nine Thousand Dollars for one month's treatment. Drug costs? Less than 20% of that. Fortunately it only cost me $150 to a friend who's an attorney to write a letter to the insurance company to make it all go away.
One more and then I'll STFU. I hate insurance companies too, even though the spousal unit would probably be dead without them. The aforementioned Neupogen? They give it to stimulate white blood cell production - we decided to give it at home. Got the oncology nurses to give me a class on giving injections and even got to practice on the spousal unit. Our oncologist writes the prescription and I take it to the corner pharmacy. Pharmacy calls two days later and says prescription is in, your insurance company won't cover it and please bring in a check for $2800.
Turns out my insurance company at the time was really an insurance broker and the prescription and major medical components got in a big argument over who was gonna get to pay for this. Major medical lost - and rather than give the shot at home we had to go in for an office visit every day from day 4 through about day 10 of each treatment cycle to get the shot.
There are so many holes in the system it's tough to pinpoint any one problem, but I've rambled long enough.
...I don't know what the hell it is with Windows-only admins and rebooting. The kind of instability that required reboots all the time was reduced drastically with Win2k and win2k3, yet that insatiable urge to reboot first and ask questions later still plauges my Windows-only counterparts.
It's still a good test when interviewing prospective Exchange/MS SQL/Windows admins. Ask them how to troubleshoot a fairly common scenario and if they mention the word reboot they go into the 'do not hire' pile.
True, but we've pretty much given up on the idea that the chip can be sniffed without the bearer's knowledge. An antenna that would provide the required gain isn't gonna be real easily hidden.
Here's a really short article that shows what you're up against with a picture of an antenna that'll pick up passive RFID at 50 feet -
The fly in the ointment is the electromagnetic field required to power the chip. To double the distance between RFID reader and chip requires about eight times the power. So - just for grins let's say we've got a best-case read of half a meter at the FCC limit of one watt RFID receiver power. It'd take 8 watts of power to activate the chip from one meter away, 64 watts to power it two meters away and 512 transmitter watts to activate the chip from four meters away. You're gonna need a big antenna or lotsa juice - or both;-)
...Incidentally, the ten-inch range you mention -- presumably that using a standard reader/writer without any kind of high-power tranmitter or any other Handwavium-powered device to boost the range? I would guess that, inverse cube root law not withstanding, the range could be increased somewhat with applied effort.
I'm sure you could, given enough effort - and you're correct, I'm using off-the-shelf components.
Given enough effort I'm sure one can read these things from a much greater distance; but as the distance increases the portability of the reader decreases significantly;-)
All respect, but the part you're missing with the whole passive RFID thing is that yes, the signal can be picked up given enough antenna gain but you've got to induce enough current in the RFID tag's antenna to power the chip before it will transmit anything. Passive RFID signals don't always exist, they only exist when the RFID tag is energized by an RFID reader.
For an isotropic (directionless) transmitter/receiver pair, the power requirement is distance^4. That is not exponential.
As a matter of fact it is - the most basic definition being "of or involving exponents";-)
You're correct - boosting receive gain will increase range but you've still gotta induce enough current in the chip's antenna to get it to transmit in the first place.
While you may "do" it for a living, it sounds like you don't hack it for a living. It takes a whole different mindset to look for vulnerabilities to exploit.
This is true. Given a directional antenna and enough power you can read most anything. Truth be told I hadn't considered the 'American detector'. Interesting thought.
Actually I'm neither - and couldn't read your link. Pesky 404s anyway - but UHF RFID isn't what's being fielded here;-)
In practice your 1 meter degrades to considerably less than half that distance under suboptimal conditions - at least in my experience. I get reliable reads out to about ten inches.
I do this stuff (among other things) for a living.
There are two types of RFID tags - active (carries its own power supply) and passive (powered by the magnetic field generated by an RFID reader). The best active tags can be read a couple hundred feet away - that's what you use to go speeding through toll booths and such.
Passive tags (like the one in the passport) can only be read a few inches away and someone with even a basic knowledge of physics knows that the power requirement to maintain an adequate magnetic field increases exponentially with distance.
Also, when using the FUD Technologies Nuclear Long Distance Handheld Omnidirectional RFID Reader® one has to remember that tags operating on the same frequency will tend to interfere with each other, reducing the chance of getting a good read.
My suggestion is to take the tinfoil off one's head, wrap their passport in it and go about their business;-)
It's just a hassle to deal with windows erasing the MBR.
Next time install grub on the/boot partition instead of the MBR, use dd to make an image of that partition's boot sector, put a copy of that image in the root of c: and call it with a Windows bootloader. Easy as pie.
Then you can install Linux or Windows without trashing the other OS - there are a whole pile of HOWTOs available on the web to help you do this.
Slashdot Mirror
It's hard these days to do a sole source procurement - and I think it should be. Doesn't always make things go the way I want them to, but I think the rule is a good one.
That's a dangerous game if your company has any financial interest in the contract - but you're right, it's all in the statement of work or RFQ. I wrote a SOW for a cell phone contract once that required coverage in 35 little towns - about a third of them didn't even have post offices. My preferred provider lost the bid but so far the company that won the bid has provided pretty good service.
You know, to be honest I didn't think of that. The end-of-year feeding frenzy around here means they call you at noon and want you to spend a quarter of a million bucks before you go home for the day - with vendor sources and full justification as to why you need the hardware they wouldn't let you have all year long.
I work for an agency under DoD. I had some end of year fundage to spend so I decided to buy some printers and networkable scanners - about $100k worth of gadgets. I'd requested HP hardware because that's pretty much what the infrastructure here is geared to support.
Because of the size of the contract award the thing went out for open bid - and I was contacted by another printer manufacturer. I won't tell you their name, but their initials are L-e-x-m-a-r-k.
Strongarm tactics ensue. First the local contracting office asked me to define printing and scanning requirements as the Other Printer Company believes they can meet my requirements at a lower cost - but we won't mention the fact that all the supplies I have in stock are from the Printer Company I Wanted To Use and adding another hardware vendor would be a logistical nightmare.
So - starting with the network scanners I start looking at hardware specs. The Other Printer Company says they can meet my requirements, but since a digital sender is an input device as opposed to an output device, I would have to get the new hardware certifiied by the network spies and I don't have time to do that, so for that part of the procurement I got the hardware I requested.
The printers were another matter. Once you've specified dpi, print speed and networking capabilities you've pretty much got to go with whoever brings the lowest bid - so the Other Printer Company won that.
During the acquisition process I felt like I was being strongarmed by the Other Printer Company and since I couldn't give a good reason not to use their hardware I have to use it. If I'd have had a week instead of a day to process the procurement I probably could have.
I have learned that I need to fine-tune my hardware requirements to keep it from happening again - but manufacturers can and will sue the government for buying from somebody else.
As I said in the subject line, rubbish.
I have three identical machines in my home office - all three are Dell Precision Workstation 360 machines with 2.8GHz processors, 1GB o'memory and identical 36GB 10k rpm U320 SCSI hard drives. Two of them run Windows XP with all patches and one runs CentOS 4.4 - again, with all patches. Window manager is Gnome - which for me is a good balance between the bloatware that is KDE and my favorite but frequently irritating window manager, XFCE4.
The Linux box is *not* faster than the Windows boxes. Granted, the Windows machines mainly just surf the web or play music or store pictures and the Linux box acts as a mail/web server for all of three users, but the machine boots slower, applications load slower and the whole machine is less responsive.
Gnome will bring up a desktop about five seconds slower than XFCE will, which is pretty darned speedy - but counting the time it takes me to boot both machines, log on (and type startx on the Linux box) I have a working desktop quicker in Windows than I do in Linux.
Now I will say that while I do have a working desktop on the Windows box there are still services starting in the background, but on identical hardware all applications are more responsive on the Redmond machines than on the CentOS box.
I guess the Linux box would last 8 years - if I never patched it.
I work for an agency under DoD. Vista, OFfice 2007 and IE7 are also verboten here.
Government geels in decisiomaking capacity don't get into beta programs and it's easier to ban a software package than learn how to support it.
2) High-end systems do not provide an option for high-end nVidia cards (their drivers work better on linux)
3) All high-end systems require purchase of MS Windows Last fall I bought three off-lease Dell Precision Workstations from Dell Financial Services All were 2.8GHz boxes with 1GB RAM, 36G SCSI hard drives and Nvidia Quadro graphics cards. Paid ~$400 for each of them, loaded CentOS 4 on one and Windows XP on the other two.
Even with the included Windows XP CD I had to download and install SCSI drivers before XP would install - CentOS just worked. I did have to install Nvidia's driver on all three boxes to get to my native 1680x1050 resolution and I have to reinstall the Linux driver with every kernel upgrade, but Nvidia's got a shell script that makes that a no-brainer. I think the driver install in Linux is easier than the one in Windows.
An oversimplification. My apologies.
"Cancer" is a catchall term that describes more than 100 different diseases that share some or all of the same characteristics - namely cellular malformation that's not corrected by the host immune system, rapid cellular division and a particularly interesting trait called angiogenesis - the ability to create its own blood supply. You're born with all the blood vessels you'll ever have - and the only time you grow new ones if if you have cancer or are pregnant.
Interestingly, Thalidomide (yeah, the sleeping pill that caused all the birth defects in the early '60s) is currently seeing a fair bit of use as an antiangiogenetic agent - the same drug that causes hideous birth defects by suppressing blood vessel formation is now having limited success as an anticancer agent. I have a couple of internet friends doing Thalidomide right now - they have to sign a statement saying they understand the risk, that they will take steps to keep from becoming pregnant and the drug is packaged with scary pictures of a baby in a red circle with a slash through it.
But I digress.
The vast majority of chemotherapy costs are *not* drug costs.
Case in point: My spousal unit has Stage IV breast cancer and has had for almost eight years. She's currently on her sixth course (not round, course) of chemotherapy and is doing pretty well, thankyouverymuch ;-)
Let's talk about drug costs. Doxorubicin (trade name Adriamycin) is generally accepted first line chemotherapy for breast cancer in combination with another drug, Cytoxan. In some parts of the world they use Epirubicin instead, but I digress.
Wholesale drug cost for a single dose of Adriamycin is about $300. Considering the drug is toxic as hell and requires special handling and disposal protocol at the manufacturer level that seems reasonable. Cost to administer? Seven thousand bucks - I can show you the bill. Since it's administered as an IV push over a couple hours it takes a little more work than other drugs but in a healthcare system that charges $180 to start an IV (doesn't matter that the spousal unit has a mediport in her chest) or $40 for a liter of normal saline for that IV it's not difficult to see where the markup is.
Don't get me wrong, I have no real love for drug companies - the two most widely prescribed post-chemotherapy antinausea meds are Zofran and Kytril. Zoftan has an average wholesale price (AWP) of $26.25 _per_tablet_. Kytril is even better, at $59.67 per pill. They normally give you Kytril if Zofran doesn't work. You pay or you puke for a couple days - your choice.
The drug that's keeping the spousal unit alive right now is called Herceptin and costs about $48k per year. Reasonably new monoclonal antibody made from the ovaries of Chinese hamsters. Guess $1000 per dose is reasonable since you probably gotta grind up a lot of hamsters to get enough to be useful.
But again, it costs much more to give the IV than the drug costs.
I got sued by our local cancer center because my insurance company decided to play games one month. Since the patient is responsible for medical charges in order for the hospital to collect they have to sue you and name the insurance company as a codefendant. One month of treatment - which included two rounds of chemotherapy, two 15 minute doctor visits, probably eight blood tests and ten days worth of a drug called Neupogen they used to give you to stimulate white blood count?
Glad you asked. $39,000. Thirty Nine Thousand Dollars for one month's treatment. Drug costs? Less than 20% of that. Fortunately it only cost me $150 to a friend who's an attorney to write a letter to the insurance company to make it all go away.
One more and then I'll STFU. I hate insurance companies too, even though the spousal unit would probably be dead without them. The aforementioned Neupogen? They give it to stimulate white blood cell production - we decided to give it at home. Got the oncology nurses to give me a class on giving injections and even got to practice on the spousal unit. Our oncologist writes the prescription and I take it to the corner pharmacy. Pharmacy calls two days later and says prescription is in, your insurance company won't cover it and please bring in a check for $2800.
Turns out my insurance company at the time was really an insurance broker and the prescription and major medical components got in a big argument over who was gonna get to pay for this. Major medical lost - and rather than give the shot at home we had to go in for an office visit every day from day 4 through about day 10 of each treatment cycle to get the shot.
There are so many holes in the system it's tough to pinpoint any one problem, but I've rambled long enough.
cheers -
...I don't know what the hell it is with Windows-only admins and rebooting. The kind of instability that required reboots all the time was reduced drastically with Win2k and win2k3, yet that insatiable urge to reboot first and ask questions later still plauges my Windows-only counterparts.It's still a good test when interviewing prospective Exchange/MS SQL/Windows admins. Ask them how to troubleshoot a fairly common scenario and if they mention the word reboot they go into the 'do not hire' pile.
I know mine are but I thought it was just part of being middle-aged.
Actually I didn't *see* the slash. I'd expect even the aforementioned script kiddie could post a valid link.
Okay, you win. UHF RFID can be read 4-8 meters away. That's still not what's in the passports.
http://en.wikipedia.org/wiki/Oliver_Sipple
Here's a really short article that shows what you're up against with a picture of an antenna that'll pick up passive RFID at 50 feet -
http://www.dvorak.org/blog/?p=2437
The fly in the ointment is the electromagnetic field required to power the chip. To double the distance between RFID reader and chip requires about eight times the power. So - just for grins let's say we've got a best-case read of half a meter at the FCC limit of one watt RFID receiver power. It'd take 8 watts of power to activate the chip from one meter away, 64 watts to power it two meters away and 512 transmitter watts to activate the chip from four meters away. You're gonna need a big antenna or lotsa juice - or both ;-)
Given enough effort I'm sure one can read these things from a much greater distance; but as the distance increases the portability of the reader decreases significantly ;-)
All respect, but the part you're missing with the whole passive RFID thing is that yes, the signal can be picked up given enough antenna gain but you've got to induce enough current in the RFID tag's antenna to power the chip before it will transmit anything. Passive RFID signals don't always exist, they only exist when the RFID tag is energized by an RFID reader.
You're correct - boosting receive gain will increase range but you've still gotta induce enough current in the chip's antenna to get it to transmit in the first place.
Actually I'm neither - and couldn't read your link. Pesky 404s anyway - but UHF RFID isn't what's being fielded here ;-)
In practice your 1 meter degrades to considerably less than half that distance under suboptimal conditions - at least in my experience. I get reliable reads out to about ten inches.
I do this stuff (among other things) for a living.
;-)
There are two types of RFID tags - active (carries its own power supply) and passive (powered by the magnetic field generated by an RFID reader). The best active tags can be read a couple hundred feet away - that's what you use to go speeding through toll booths and such.
Passive tags (like the one in the passport) can only be read a few inches away and someone with even a basic knowledge of physics knows that the power requirement to maintain an adequate magnetic field increases exponentially with distance.
Also, when using the FUD Technologies Nuclear Long Distance Handheld Omnidirectional RFID Reader® one has to remember that tags operating on the same frequency will tend to interfere with each other, reducing the chance of getting a good read.
My suggestion is to take the tinfoil off one's head, wrap their passport in it and go about their business
Then you can install Linux or Windows without trashing the other OS - there are a whole pile of HOWTOs available on the web to help you do this.