After adjusting to life on a laptop, I had to do something about the flaking desktop. As frequent followers of this blog remember, the Compaq Presario SR1750NX started doing sudden death incidents. It was sudden, no blue screen of death, no error messages, it just stopped, monitor blank. Fearing total loss of email, bookmarks, check book, letters, plans, software, and you know it all, I backed everything up to CDs and moved operations to a laptop.
But, the Presario was still cluttering up the desk, with monitor and wires and computery stuff hogging too much space. It was either fix it or scrap it.
A new motherboard ought to do the trick, since everything is on the motherboard. So, I took the ailing machine to the workshop and did a motherboarderectomy, looking for part numbers, model numbers, dimensions, hole patterns and anything that might help me find a new board cheap. The Presario had a slather of look alike cables going to lots of wierdo connectors all over the board. But with some handwritten cable labels made from masking tape, and a sketch, I figured it would all go back together.
While I had it apart, I took the shop vac and sucked humungous dustbunnies out of everywhere. I copied down every number I could see. The maker (Asus) didn't bother to put his name on the board (poor marketing) but did put his model number A8AE-LE on it. Two thoughts crossed my mind. The works (all the hi-tech) of this American desktop were designed and manufactured in Taiwan. Americans only do sheet metal work. Second thought was only another A8AE board was gonna fit the casework without a lot a sheet metal butchery. The motherboard carries connectors for VGA, USB, keyboard, mouse, mike, headset, LAN, and Firewire, and the connectors all fit neatly into matching holes in the case. Chances of finding another make of motherboard with the same connector arrangement are slim. From web surfing I knew the A8AE board was special to Compaq and not sold to anyone but Compaq. Getting a new one thru Compaq might be pricey. Very pricey.
So, I put it back together, while I still remembered how everything went. For grins I used a Pink Pearl pencil eraser to polish the gold fingers on the RAM sticks. With it all back together, I plugged it in, just to see. Sun of a gun. She powered right up, just like new.
Damn. Is this for real, or is the mystery crapout just biding its time? I left it running for two days, and it keep right on ticking.
This morning, I backed up last months work off the laptop and loaded it back onto the Presario. Everything looks good. It may be that a simple remove and replace has fixed it for good. Stay tuned for further developments.
Sunday, May 3, 2009
Friday, May 1, 2009
Follow on to Horror Story
Apparently the electronic medical records business is in as bad shape as I feared. Yesterday's WSJ had a story about the electronic medical record system used by the Veteran's Administration. The system is public domain open source and very-low-cost to free. VA has been using it for twenty years. One hitch, the VA merely offers the source code. The hospital has to employ some computer geeks to compile it and install it and do maintainance (bug fixes). The VA software is good except for one thing. No billing functions (VA doesn't bill veterans). The hospital computer geeks will have to find or write a billing module, but this has been done.
The article went on to say that competing commercial medical records programs cost heavily and are not compatible with each other. This means that electronic patient records created by system X cannot be read by system Y. So change hospitals or doctors or health plans and your medical record is in jeopardy.
You might have known.
Looks like a business opportunity for some computer geeks to set themselves up in business offering the VA program to hospitals.
The article went on to say that competing commercial medical records programs cost heavily and are not compatible with each other. This means that electronic patient records created by system X cannot be read by system Y. So change hospitals or doctors or health plans and your medical record is in jeopardy.
You might have known.
Looks like a business opportunity for some computer geeks to set themselves up in business offering the VA program to hospitals.
NYT presents a 1920's idea as new technology
Good old NYT, life would be boring without them. They ran a big spread on ocean thermal power, presenting it as the answer to the energy problem and a promising new technology. Of course the highly educated Times men didn't know that ocean thermal power was pioneered in the 1920's by Georges Claude, a wealthy French scientist. Claude built a test plant on the shore of Cuba in 1929 which worked but produced little power. He built a sea borne ship mounted unit in 1934 which didn't work much better. In the end Claude gave up and scuttled his ship in a place where the ocean was especially deep.
The concept of ocean thermal power is fairly simple. You build a steam engine that runs on a fluid that boils from the heat of the warm surface water, and condenses in the coolth of the deep bottom water. This requires a long pipe reaching down to the cold bottom water layers. This pipe has been the bane of ocean thermal power experiments. Either it breaks in a storm, or the feed water pump uses up most of the output of the plant. Heat engines work off a thermal difference. The greater the temperature difference the more power the engine will deliver. Old fashioned steam locomotives worked off a temperature difference of 300 degrees F.
Warm tropical ocean water might be 80 degrees F. The cold bottom water is never colder than 32 degrees F, yielding a temperature difference of only 50 degrees F at the most favorable locations, like the Carribbean. Off the New England coast, the ocean water never goes above 50 degrees F, reducing the temperature difference to a mere 20 degrees F.
Since Claude's experiments 70 years ago, other's have tried, but no one has been able to squeeze enough power out of a 50 degree temperature difference to make the effort worth while. We are talking steam engines here. Steam engines have been well understood for better than a hundred years. It is unlikely that some intrepid inventor will make a great breakthru in steam engine technology.
The concept of ocean thermal power is fairly simple. You build a steam engine that runs on a fluid that boils from the heat of the warm surface water, and condenses in the coolth of the deep bottom water. This requires a long pipe reaching down to the cold bottom water layers. This pipe has been the bane of ocean thermal power experiments. Either it breaks in a storm, or the feed water pump uses up most of the output of the plant. Heat engines work off a thermal difference. The greater the temperature difference the more power the engine will deliver. Old fashioned steam locomotives worked off a temperature difference of 300 degrees F.
Warm tropical ocean water might be 80 degrees F. The cold bottom water is never colder than 32 degrees F, yielding a temperature difference of only 50 degrees F at the most favorable locations, like the Carribbean. Off the New England coast, the ocean water never goes above 50 degrees F, reducing the temperature difference to a mere 20 degrees F.
Since Claude's experiments 70 years ago, other's have tried, but no one has been able to squeeze enough power out of a 50 degree temperature difference to make the effort worth while. We are talking steam engines here. Steam engines have been well understood for better than a hundred years. It is unlikely that some intrepid inventor will make a great breakthru in steam engine technology.
Thursday, April 30, 2009
Bye Bye Pontiac
It's sad. No more Pontiac GTO's. We had a Pontiac once. Big black '59 wagon with the 389 V8. It was big enough for a family of five with skis, gear, luggage and the family cat. The cat hated it. Once it jumped out the drivers window into a toll basket trying to escape from the car. The driver (father) was quick enough to grab the cat out of the basket, heave it into the back seat, roll up the window and drag race out of the toll booth like nothing ever happened. And before the cat made a second try for freedom.
That wagon had the worst transmission known to man. Three on the tree. First gear was too high, you had to slip the clutch to get the car rolling on the flat. Starting up hill was bad. Lots of burning clutch smell, and without a master's touch on the pedal, bucking, surging and engine stalling. Plus an unreliable shift linkage. It would occasionally get stuck in reverse, and you had to pop the hood and fiddle with the linkage to get unstuck. It acted up on the old man on Boston's central artery, during rush hour, in a driving rainstorm. He was plenty wet (and mad) before he got the car to run forward again.
Lot of talk on some car enthusiast blogs about "the channel" and the need (or lack of need) for two or three or four GM "channels". Those guys were probably car dealers worrying about loosing their franchise. GM doesn't need "channels". It does need product, cars that people will buy. Pontiac doesn't make GTO's any more, and the last interesting Pontiac was the Firebird. Except Firebird wasn't a real car model, it was a Chevy Camaro with a Pontiac nameplate. Everyone knew that, the styling was distinctive, and anyone with two brain cells firing instantly recognized the simularity of Firebird/Camaro. GM would have saved money and raised sales by marketing the Camaro under just one name. Consumers are saturated with advertising, in fact most of us consumers automatically ignore commercials. It takes a LOT of advertising to cut thru the mental filters TV watchers have evolved. Better to spend the money on one car brand than split it between two brand names attached to the same car. Guys (pony cars are a guy thing) who would buy a Firebird are equally likely to buy a Camaro. Why dilute the advertising by selling the same car under two different names?
That wagon had the worst transmission known to man. Three on the tree. First gear was too high, you had to slip the clutch to get the car rolling on the flat. Starting up hill was bad. Lots of burning clutch smell, and without a master's touch on the pedal, bucking, surging and engine stalling. Plus an unreliable shift linkage. It would occasionally get stuck in reverse, and you had to pop the hood and fiddle with the linkage to get unstuck. It acted up on the old man on Boston's central artery, during rush hour, in a driving rainstorm. He was plenty wet (and mad) before he got the car to run forward again.
Lot of talk on some car enthusiast blogs about "the channel" and the need (or lack of need) for two or three or four GM "channels". Those guys were probably car dealers worrying about loosing their franchise. GM doesn't need "channels". It does need product, cars that people will buy. Pontiac doesn't make GTO's any more, and the last interesting Pontiac was the Firebird. Except Firebird wasn't a real car model, it was a Chevy Camaro with a Pontiac nameplate. Everyone knew that, the styling was distinctive, and anyone with two brain cells firing instantly recognized the simularity of Firebird/Camaro. GM would have saved money and raised sales by marketing the Camaro under just one name. Consumers are saturated with advertising, in fact most of us consumers automatically ignore commercials. It takes a LOT of advertising to cut thru the mental filters TV watchers have evolved. Better to spend the money on one car brand than split it between two brand names attached to the same car. Guys (pony cars are a guy thing) who would buy a Firebird are equally likely to buy a Camaro. Why dilute the advertising by selling the same car under two different names?
Dawn over Marblehead. Microsoft wises up slightly.
Microsoft is going to partially disable autorun in Windows 7. Of course they couldn't bring them selves to disable it all the way, CD-ROM drives will still autorun, but USB flash drives will not.
Me, I have disabled autorun with a hand patch to my registry. On earlier versions of Windows you could turn autorun off in Device Manager. But Microsoft loved autorun so much that XP makes it harder to kill. You now have to run regedit and do a hand patch to the registry. User friendly that is.
Autorun is that feature that makes music CD's start to play for just inserting the CD into the drive. It also makes software CD's start to install hands off. It is a gaping security hole in Windows. Autorun will load and execute any kind of code off the CD or flash drive for just inserting the media into the drive or USB slot. Put a virus on a CD and infect every machine the CD ever sees. Sony used autorun to infect user's machines with an anti copy root kit to prevent copying CD's. Malware can spread by copying itself to USB flash drives.
If Microsoft actually cared about Windows security, they would remove autorun completely. Apparently they still love it too much to kill it all the way.
Computers work just fine without autorun. Users just have to click on the CD or flash drive to make it play or run. Much safer that way. I'll click on a million CD's to play them rather than autorun just one ugly piece of malware that eats my hard drive.
Me, I have disabled autorun with a hand patch to my registry. On earlier versions of Windows you could turn autorun off in Device Manager. But Microsoft loved autorun so much that XP makes it harder to kill. You now have to run regedit and do a hand patch to the registry. User friendly that is.
Autorun is that feature that makes music CD's start to play for just inserting the CD into the drive. It also makes software CD's start to install hands off. It is a gaping security hole in Windows. Autorun will load and execute any kind of code off the CD or flash drive for just inserting the media into the drive or USB slot. Put a virus on a CD and infect every machine the CD ever sees. Sony used autorun to infect user's machines with an anti copy root kit to prevent copying CD's. Malware can spread by copying itself to USB flash drives.
If Microsoft actually cared about Windows security, they would remove autorun completely. Apparently they still love it too much to kill it all the way.
Computers work just fine without autorun. Users just have to click on the CD or flash drive to make it play or run. Much safer that way. I'll click on a million CD's to play them rather than autorun just one ugly piece of malware that eats my hard drive.
Tuesday, April 28, 2009
Electronic Medical Records Horror Story
The writer, an asthma patient, describes a near death experience in hospital. All the doctors and nurses were too busy working with a clunky computer system to treat him. The asthma nearly killed him.
The writer works in IT and he talks about data models and message models and other technical stuff. In simpler words, this is an example of the don't know squared problem. The medical people don't know, that they don't know what they need. It's a typical situation. The customer wants "it" computerized but that's as far as his thinking goes. The vendors do their best and present the customer with a vast written specification, 100's of pages, which only makes sense to the vendor who wrote it. "My eyes glaze over" (MEGO) . Eventually the spec is signed, so the job can proceed, even though the customer still doesn't understand it. If the resulting system isn't too too bad, it goes into service.
I had thought that the medical business would have come up with a workable system by now. Manufacturing has SAP, Engineering has Orcad, administrators have Office. Each system has been in use for 10 years or more and is basically satisfactory. Sounds like the medics are not there yet.
And this story doesn't get into security issues. An electronic medical record system based on Windows computers hooked to the public internet will be totally vulnerable to high school hackers. Your medical records will be "accessed" by anyone who cares; employers, insurers, private investigators, the ex-wife, bill collectors, cops, lawyers, parents, children and god knows who else.
Neither did the author talk about portability issue. For the system to be worth much, records entered by your doctor, at his office, should readable at the hospital should you later get hospitalized. Or at the next hospital after you move. Without the fiercest sort of federal supervision, competing vendor's systems will be unable to read records created or edited on other systems.
The writer works in IT and he talks about data models and message models and other technical stuff. In simpler words, this is an example of the don't know squared problem. The medical people don't know, that they don't know what they need. It's a typical situation. The customer wants "it" computerized but that's as far as his thinking goes. The vendors do their best and present the customer with a vast written specification, 100's of pages, which only makes sense to the vendor who wrote it. "My eyes glaze over" (MEGO) . Eventually the spec is signed, so the job can proceed, even though the customer still doesn't understand it. If the resulting system isn't too too bad, it goes into service.
I had thought that the medical business would have come up with a workable system by now. Manufacturing has SAP, Engineering has Orcad, administrators have Office. Each system has been in use for 10 years or more and is basically satisfactory. Sounds like the medics are not there yet.
And this story doesn't get into security issues. An electronic medical record system based on Windows computers hooked to the public internet will be totally vulnerable to high school hackers. Your medical records will be "accessed" by anyone who cares; employers, insurers, private investigators, the ex-wife, bill collectors, cops, lawyers, parents, children and god knows who else.
Neither did the author talk about portability issue. For the system to be worth much, records entered by your doctor, at his office, should readable at the hospital should you later get hospitalized. Or at the next hospital after you move. Without the fiercest sort of federal supervision, competing vendor's systems will be unable to read records created or edited on other systems.
Why does swine flu get more press than plain flu?
Dunno. The swine type sure does drive the newsies into a headline orgy. How does it differ from plain old influenza? Is it more contagious, more virulent, or what? Or does it just give a DNA match with cases from Mexico, or swine, or something. Do I care if complex lab procedures show a match with something else?
Unless I hear something that matters, I think it's just another flu varient, the likes of which we hav been seeing for centuries. Anyone remember the Asian flu of 1957? It went thru boarding school like a house afire, every kid in my school came down with it. They all recovered too.
Unless I hear something that matters, I think it's just another flu varient, the likes of which we hav been seeing for centuries. Anyone remember the Asian flu of 1957? It went thru boarding school like a house afire, every kid in my school came down with it. They all recovered too.
Subscribe to:
Posts (Atom)