This is why it's hard to subscribe to these huge conspiracy theories about anything involving the Federal Government. I mean, if you want to successfully lead a conspiracy, you have to be competent and you have to cover your tracks.
This latest example of bumbling incompetence shows us that you cannot trust the Feds to do either.
That's why we should fear the Feds when they want to help us: considering their track records at taking care of their own problems, only a suicidal madman would trust them to manage other people's lives.
So lemme get this straight. The employee objects to the Googling that reveals his past problems with another Federal agency.
That Google search probabky took 5 minutes and actually saved the taxpayers money. Had the Air Force archive not been available on the Web, Mullins's employer would have issued the proper redtape-littered procedure to unearth Mullins's file from Federal archives. Such a thing involves weeks, if not months, of inter-agency sleuthing, and ends up costing a pretty penny. For the taxpayer, of course.
But such a search would have been so much more official, so much less oh-noes-high-tech-got-me, so Mullins would not have objected, right?
The Pentagon is a public employer. Its agents are susceptible to public scrutiny when it comes to how their salary is justified. As such, it's normal that its employment archives are conveniently available. Don't like it and still want to be a Federal employee? Get a Masters in maths and get hired by the NSA. They don't plaster the web with their employment records, for some reason.
Otherwise, get into the private sector and deal with the longer hours and the sucky retirement and heath plans, like the rest of us.
Kudos to the Weather Forecast Office in Indiana for getting rid of cheating, lying, dishonest employees.
Are you trying to extract flow information from non-instrumented valves? If so, you need to either replace the valve with something modern that outputs its data cleanly, or just insert flowmeters in series with the values to be monitored.
Either way, get something clean and professional, don't do a DIY job that's hard to maintain.
Also, once you have an instrumented device, I'd not recommend putting a homebrew micro-controller system on a factory flow, especially a glass factory, which abounds in abrasive dust. Any consumer cooling fan, for example, would quickly see its bearings ground and clogged by silicate dust. Get a real industrial device with sealed bearing, etc.
Devices that can widthstand factory floor abuse, high temperature and high dust levels are expensive, but there is a good reason.
I said:
The way normal people would do it is use an X11 graphic application
You replied: Ah, it appears that a Slashdot-crazed unix fanatic could be sillier!
Silly? Why? Please explain.
I didn't say X11 was THE way to do it. But I've seen large deployment of X11-based graphical applications running on Unix servers (AIX to be precise) that were accessed first from a variety of thin clients (X terminals among others) and later from Windows machines running X11 display servers. The whole thing was reasonably fast. They worked well enough that management was in no hurry to recode the apps to make them browser-based.
These days, when I am spec'ing new distributed GUI-based applications, I don't recommend writing against the X11 stack, but your mileage may vary.
From the ComputerWorld article: Deal and an IT employee, who spoke to Computerworld on the condition of anonymity, said part of the problem with the HealthConnect system is that the Citrix Application Delivery infrastructure implemented by Kasier just can't handle the load of the Epic Systems.
"We're the largest Citrix deployment in the world," Deal said. "We're using it in a way that's quite different from the way most organizations are using it. A lot of users use it to allow remote users to connect to the network. But we actually use it from inside the network. For every user who connects to HealthConnect, they connect via Citrix, and we're running into monumental problems in scaling the Citrix servers."
So instead of deploying the app on N thousands Windows desktop, they deciced to use Citrix to remotely connect to a pool of servers. The Citrix server and the Windows machine at the other end could not stand the load. Big surprise.
The way normal people would do it is use an X11 graphic application (X11 is available for Win32), or use a Java webstart client, or even do everything within a browser, or... But there are many, many way to architect a distributed app these days.
The ONE thing you shouldn't do is deploy lots of Windows servers, use the half-baked ICA protocol, and expect everything to be peachy.
Remember, CIO boys and girls: Uncle Bill's broken OS just cost lil' Cliff Dodd his job. Don't be the next one. Keep Win32 where it belongs, outside the server room.
Not sure about the 128 SIMDs. The PS3's cell processor has 7 so-called SPUs (Synergistic Processor Units), basically coprocessors that work on a 256k separate internal storage. If your code is embarassingly parallel, you can greatly benefit from this architecture.
The hard-coded floating point unit is single-precision indeed, and doesn't implement the full IEEE floating point spec, which raises eyebrows in the scientific double-precision junky crowd. The individual registers are 128-bit, although they are organized as 4 32-bit words.
Do data centers really use as much power cooling the server farms as running them?
A bit more actually. For every watt of energy burned by a server in a DC, you need another 1 to 2 watts to handle the air, run the AC units, run the pumps (for chilled water systems), etc. A very efficient DC has a Power Utilization Ratio of 2.5 - That is, it consumes a total power of "only" 2.5 W for each watt of computer power inside. In other words, for each watt of computer power, it uses another 1.5 W for ventilation and cooling.Power now costs about 60% as much as server expenditures in a DC.
There are already several offers on the market for spraying CPU with chilled inert liquids. Oil is just another way to do it.
Slashdot Mirror
I stick to my version of the truth.
This is why it's hard to subscribe to these huge conspiracy theories about anything involving the Federal Government. I mean, if you want to successfully lead a conspiracy, you have to be competent and you have to cover your tracks.
This latest example of bumbling incompetence shows us that you cannot trust the Feds to do either.
That's why we should fear the Feds when they want to help us: considering their track records at taking care of their own problems, only a suicidal madman would trust them to manage other people's lives.
So lemme get this straight. The employee objects to the Googling that reveals his past problems with another Federal agency.
That Google search probabky took 5 minutes and actually saved the taxpayers money. Had the Air Force archive not been available on the Web, Mullins's employer would have issued the proper redtape-littered procedure to unearth Mullins's file from Federal archives. Such a thing involves weeks, if not months, of inter-agency sleuthing, and ends up costing a pretty penny. For the taxpayer, of course.
But such a search would have been so much more official, so much less oh-noes-high-tech-got-me, so Mullins would not have objected, right?
The Pentagon is a public employer. Its agents are susceptible to public scrutiny when it comes to how their salary is justified. As such, it's normal that its employment archives are conveniently available. Don't like it and still want to be a Federal employee? Get a Masters in maths and get hired by the NSA. They don't plaster the web with their employment records, for some reason.
Otherwise, get into the private sector and deal with the longer hours and the sucky retirement and heath plans, like the rest of us.
Kudos to the Weather Forecast Office in Indiana for getting rid of cheating, lying, dishonest employees.
Are you trying to extract flow information from non-instrumented valves? If so, you need to either replace the valve with something modern that outputs its data cleanly, or just insert flowmeters in series with the values to be monitored. Either way, get something clean and professional, don't do a DIY job that's hard to maintain. Also, once you have an instrumented device, I'd not recommend putting a homebrew micro-controller system on a factory flow, especially a glass factory, which abounds in abrasive dust. Any consumer cooling fan, for example, would quickly see its bearings ground and clogged by silicate dust. Get a real industrial device with sealed bearing, etc. Devices that can widthstand factory floor abuse, high temperature and high dust levels are expensive, but there is a good reason.
I must agree. That replacement exist. It's call SVG and Java.
I said: The way normal people would do it is use an X11 graphic application
You replied: Ah, it appears that a Slashdot-crazed unix fanatic could be sillier!
Silly? Why? Please explain.
I didn't say X11 was THE way to do it. But I've seen large deployment of X11-based graphical applications running on Unix servers (AIX to be precise) that were accessed first from a variety of thin clients (X terminals among others) and later from Windows machines running X11 display servers. The whole thing was reasonably fast. They worked well enough that management was in no hurry to recode the apps to make them browser-based.
These days, when I am spec'ing new distributed GUI-based applications, I don't recommend writing against the X11 stack, but your mileage may vary.
From the ComputerWorld article: Deal and an IT employee, who spoke to Computerworld on the condition of anonymity, said part of the problem with the HealthConnect system is that the Citrix Application Delivery infrastructure implemented by Kasier just can't handle the load of the Epic Systems.
"We're the largest Citrix deployment in the world," Deal said. "We're using it in a way that's quite different from the way most organizations are using it. A lot of users use it to allow remote users to connect to the network. But we actually use it from inside the network. For every user who connects to HealthConnect, they connect via Citrix, and we're running into monumental problems in scaling the Citrix servers."
So instead of deploying the app on N thousands Windows desktop, they deciced to use Citrix to remotely connect to a pool of servers. The Citrix server and the Windows machine at the other end could not stand the load. Big surprise.
The way normal people would do it is use an X11 graphic application (X11 is available for Win32), or use a Java webstart client, or even do everything within a browser, or... But there are many, many way to architect a distributed app these days.
The ONE thing you shouldn't do is deploy lots of Windows servers, use the half-baked ICA protocol, and expect everything to be peachy.
Remember, CIO boys and girls: Uncle Bill's broken OS just cost lil' Cliff Dodd his job. Don't be the next one. Keep Win32 where it belongs, outside the server room.
Not sure about the 128 SIMDs. The PS3's cell processor has 7 so-called SPUs (Synergistic Processor Units), basically coprocessors that work on a 256k separate internal storage. If your code is embarassingly parallel, you can greatly benefit from this architecture.
r y/pa-linuxps3-1/index.html?ca=drs-
The hard-coded floating point unit is single-precision indeed, and doesn't implement the full IEEE floating point spec, which raises eyebrows in the scientific double-precision junky crowd. The individual registers are 128-bit, although they are organized as 4 32-bit words.
Here is a nice little intro to the beast: http://www-128.ibm.com/developerworks/linux/libra
Do data centers really use as much power cooling the server farms as running them?
A bit more actually. For every watt of energy burned by a server in a DC, you need another 1 to 2 watts to handle the air, run the AC units, run the pumps (for chilled water systems), etc. A very efficient DC has a Power Utilization Ratio of 2.5 - That is, it consumes a total power of "only" 2.5 W for each watt of computer power inside. In other words, for each watt of computer power, it uses another 1.5 W for ventilation and cooling.Power now costs about 60% as much as server expenditures in a DC.
There are already several offers on the market for spraying CPU with chilled inert liquids. Oil is just another way to do it.