A coworker of mine had a word document he was trying to open... word started to read it, then the computer crashed. He reboot only to find his bios zorched. He replaced it with a spare (identical) motherboard, tried it again, same result. He got a new motherboard, a fresh install of windows (95 or NT, it was a while ago), and a fresh install of office, tried to open the same file... EXACT SAME RESULT! He sent the word document to ms, they confirmed it would zorch the particular motherboard he had been using, and that it sucked to be him.
My guess is that it was a very lucky buffer overflow in the corrupted document that did it, but it's not what most people expect from a brand-spanking-new microsoft install.
A lot of times, the manufacturer of the USB chipset you're using will sublet their product ID's for free or nominal cost. (example: Answer 3, part 2). They're happy to sell parts, and it costs them only 1.1 cents a year per PID (they've got a block of 64k; that's a lot of numbers)
Also, correction to Qeygh's original question: it's not just $1500, it's $1500 every two years, for just the numbers Or you can join the USB org for $2500/year and get the numbers for free!
It's not the megapixels, it's the lens that differentiates professional cameras. If it doesn't have depth-of-field control (either aperature and/orr a seperate DC setting) and manual focus, you've got little creative control and it's just a snapshot. Cameras with small lens and high f-stops might as well be focus free, so they don't count. Also, a good range of lenses is useful, as is macro.
My point is that you're not going to get all that in a phone camera, irregardless of the resolution -- it's a bulkiness/user friendlyness problem.
We developed a special high-speed 10+MB/sec parallel port for a customer. We tested the software on our system, and then took it to the customer's site to try out on their spiffy 4-processor Sun.
The test program would read data off one computer and transfer it to the hard drive of another. We hooked it up, and it worked great... then we checked out the setup a little closer... they were using a network drive, not a local drive. The two computers were talking to the same server, connected with fibre. The server was probably a RAID because they did a lot of serious image manipulation, but it was just as fast or faster than the IDE's on our test system.
I've used them for years with no hassle. But they have a don't-screw-the-customer policy I like. The first thing in their contract is: The Client owns the Domain Name registered. That's a big difference from other places, like NetSol, who don't view names as property and feel like they can keep it (and auction it off to the highest bidder) should you ever quit hosting the name with them.
Also, when I joined, Gandi was a not-for-profit service... I can't find that on their webpage now, but that's the spirit they operate under.
I used to do embedded programming where it was really costly to fix in the field.
Here's a similar project's repair estimates (pdf). Mind you, this product cost 1000x our product, but since they were at similar customer sites, the repair cost wouldn't be significantly different.
Service trip #1 = $413 million Service trip #2 = $497 million Service trip #3 = $547 million Service trip #4 = $400 million
(note: these prices don't include airfare)
In fact, it would be far cheaper to just toss out our old hardware and start from scratch ($13 million total costs) than it would be to try to fix it in the field.
A machine shop owner could just find spamming the forums of their favorite magazine:-)
Excellent arguement and well put, but there is the exception that the dongle-protection-creator is exceptionally dumb and implements the same software interface in all products. Then, someone could write a rouitine that searches the executables for a function called Check_Dongle_Attached and patch it so that it always returns 1.
EDA tools us flexlm extensively - they've virtually got a lock on the market because phb's won't allow more than one kind of licensing scheme to keep up with (and run servers for). Flexlm is also priced attactively enough for software producers so that it's just under the cost for any one company to roll-their-own (~$30k). It offers good features, but security-wise, it still sucks and is regularily broken for all products that use a particular version number of flexlm.
Thanks. I was using only conductance because that's the proported use of this material - as an insulator. (Also, I'm a bit rusty on the formulas of convection and radiation!).
Also, for the radiation measurement, I'd think you'd be facing the earth on reentry, and thus use its temperature instead of 2.7K. In low earth orbit, there are big variations in the temperature seen (deep space/earth/sun), but since about half of your view is always earth, a proper thermal system will keep the average temperature in a satellite close to room temperature. Far less stringent than automotive temperatures. But, they are still designed to -55/+125C just in case. (that's just satellites, where there isn't the massive heat of reentry)
You're on the right track. The $60 million doesn't get you just shuttle heat tiles, it also gets you a warehouse full of paper documenting every single test and the certificates-of-compliance of every chemical/assembly used. Even bolts have lot numbers and are totally tracable.
I suspect they do other testing, including:
Water resistance. Not only so it doesn't wash off on rainy days, but doesn't absorb water so that freezing causes it to crack.
Free oxygen erosion. Low earth orbit exposes the leading edge of spacecraft to free oxygen (O, not the stable O2), which tends to 'rust' things quickly.
Thermal coefficient of expansion matching to the aluminium body, so it doesn't flake off. If it isn't matched, then you need a good adhesive system.
Impact resistance. Does it chip or flake? You don't want a catastrophic failure mode (a super high-speed micrometeroite should make a hole instead of shatter the whole thing)
Weight. They stopped painting the booster tank and saved a lot of weight. Current shuttle tiles are foam-like in weight.
Repairability. Do you need to resurface the whole shuttle for the slightest chip, or is it fixable?
Lastly, NASA wants a proven scientific theory of operation... something better than "It dissipates heat at an exponential rate, it's beyond belief, and I have no idea why it does, all I know is that it does." All things dissipate heat at an exponential rate - heat flow is usually related to a difference in temperatures, so as an object reaches the temperature of its surroundings, the heat flow slows down to aproach zero. That's pretty basic to understanding heat flow, and not novel.
That was a cool game, but bullfrog got hit with a crazy frivolous class-action lawsuit. It seems that about 3% of the gamers couldn't handle it and were maimed for life. In the settlement, Bullfrog agreed to build a special colony just southeast of Detroit where these gamers could live out the rest of their days... since being permanently cross-eyed made them lose their depth perception, all of the buildings in the colony are painted in a 1 foot x 1 foot grid so that accurate depths can be estimated. Outsiders visiting must wear special outfits marked with a horizontal stripe every foot. I've been there and it's a horrible way for any 3d gamer to end up. Let's only pray that something similar doesn't happen with this new technology.
And if not, they have a 6 megawatt coal-fired power plant within spitting distance of the engineering buildings. That's 5 kilowatts per computer; plenty for either a house or a computer.
Incidently, after years of putting up with coal dust, I thought we would finally see the benefits of living next to a huge coal pile. We had a severe ice storm and most of blacksburg lacked power. but so did campus!.
Ahh, that explains it - thanks. I haven't heard that term since TFT took over. I never thought of it until now, but maybe the lack of ambiguity is what TFT became preferred.
p.s. wow, my original post was poorly written. sorry.
Since AMD laptops seem few and far between, I looked for some more info. It seems that QLI doesn't offer a 15" AMD option among the laptops they offer. The only AMD laptop they offer is 14.1", and if this is the one they meant, I have no idea why it would win. Other than the linux preinstall, it's totally undistinguished. The screen is 1024x768, has shared graphics memory, and the main memory runs at only 133MHz -- hardly a "workstation".
Is this a mistake, or did they used to offer an awesome computer?
Ugh- as an electrical engineer, that smart start product looks dangerous.
Two problems: 1. It ramps up voltages. This usually isn't a good thing for digital logic - it means its going to spend some time at a marginal voltage level. Hopefully this ramp is fast.
2. It sequences the seperate power supplies without sequencing the logic. When you've got something connected but not powered, the signal lines power the chip through protection diodes - the effect is that the IDE controller is trying to power the drive logic. Maybe the drive is made to disconnect itself in this situation, but it would take extra circuitry and is not standard practice. Powering up the motor 4.5 seconds after the logic probably isn't a good idea either - the drive is probably wondering why it can't spin up. (good thing it probably has an interlock to prevent dragging the heads across the platter when the drive isn't spinning, otherwise you'd have a head crash)
The best solution is probably the free, built-in one. Usually drives have a jumper setting for delayed spin-up. Read the drive manual and use that. It's a drive firmware solution, without messing with the drive's power supply.
It's not that I want something for free just because, and it's not about shelling out the muulah.
When I bought the machine, I knew there was a risk that I wouldn't get an upgrade to Panther, and if they had a consistant policy, I'd be happy to live by it.
But they are favoring one particular product - the G5 - over all others, even though my machine is newer a newer model, costs more, was bought later than most of the G5s. It's even aimed at the same professional market. Just an explanation - Isn't this supposed to be the year of the laptop?
Sounds like this may be a inrush current problem... maybe something is taking extra juice to start up, or the supply can't deliver the initial kick. A multichanel scope would help a lot to see that all voltages are getting to where they need to be in a reasonable time. Otherwise, try removing everything that you can and see if it comes on (drives, ALL ram, usb, pci cards, modem, etc.) - maybe eliminate it that way. Remember, a bad card (one that's drawing too much current) may not cause the system to fail until another good card is inserted and pushes the power supply over the threshold.
Ha ha - Me too, once you pointed it out. But were' laughing with you PiusII, not at you. Actually, I think Xcode will be free to anyone - I remember a blurb saying dev tools would be free, but couldn't find it again on their website.
It's not so much the actual features that rubs me wrong, its just getting left out of the update. I know I'm whining. I mean, anyone who bought a laptop yesterday is getting screwed, too; I just want some logic: "30 days" or "latest round of hardware" are both good, "16 days or the only a certain line of computers that aren't the newest" just doesn't make sense.
Also, I misspoke - the current preview doesn't have postscript support, only pdf -- I generate my own ps and only the newer version will work for me. Of course there are other programs, or I could convert ps2pdf (which isn't in 10.2.7).
When I bought my last laptop, Office 2000 was on the horizon, and I was upgraded three months later when it was released. Maybe my case - 5 weeks - is pushing it, but did they really have to screw everyone who bought a laptop yesterday?
Three years ago I switched completely to Linux. I wanted a unix machine that was well integrated because I was tired to trying to get Red Hat to do simple stuff (example: native MP3 support was discontinued because of license issues, DVD playback). RH is great - I've bought 2 copies retail and my desktop runs RH8 for work, but I just didn't want to do admin chores while on my own time.
The pdf support is supposed to be "the fastest of any machine ever"... probably marketing hype, but I use a lot of large postscript files and render speed is important.
That sucks. I shelled out $3k for their latest revision of the 17" powerbook on the second day it was released (Sept 17), only to find that the features they have been touting for so long (X windows, native & fast PDF support, X code) are going to cost. Why only the top of the line desktops - why not my more expensive top of the line laptop? My machine is newer than many G5s (although maybe they were trying to satisfy customers with long preorder times).
Slashdot Mirror
A coworker of mine had a word document he was trying to open... word started to read it, then the computer crashed. He reboot only to find his bios zorched. He replaced it with a spare (identical) motherboard, tried it again, same result. He got a new motherboard, a fresh install of windows (95 or NT, it was a while ago), and a fresh install of office, tried to open the same file... EXACT SAME RESULT! He sent the word document to ms, they confirmed it would zorch the particular motherboard he had been using, and that it sucked to be him.
My guess is that it was a very lucky buffer overflow in the corrupted document that did it, but it's not what most people expect from a brand-spanking-new microsoft install.
A lot of times, the manufacturer of the USB chipset you're using will sublet their product ID's for free or nominal cost. (example: Answer 3, part 2). They're happy to sell parts, and it costs them only 1.1 cents a year per PID (they've got a block of 64k; that's a lot of numbers)
Also, correction to Qeygh's original question: it's not just $1500, it's $1500 every two years, for just the numbers Or you can join the USB org for $2500/year and get the numbers for free!
So, at that rate, they should be #2 by Haloween!
It's not the megapixels, it's the lens that differentiates professional cameras. If it doesn't have depth-of-field control (either aperature and/orr a seperate DC setting) and manual focus, you've got little creative control and it's just a snapshot. Cameras with small lens and high f-stops might as well be focus free, so they don't count. Also, a good range of lenses is useful, as is macro.
My point is that you're not going to get all that in a phone camera, irregardless of the resolution -- it's a bulkiness/user friendlyness problem.
We developed a special high-speed 10+MB/sec parallel port for a customer. We tested the software on our system, and then took it to the customer's site to try out on their spiffy 4-processor Sun.
The test program would read data off one computer and transfer it to the hard drive of another. We hooked it up, and it worked great... then we checked out the setup a little closer... they were using a network drive, not a local drive. The two computers were talking to the same server, connected with fibre. The server was probably a RAID because they did a lot of serious image manipulation, but it was just as fast or faster than the IDE's on our test system.
I can do this right now. I just fire up the powerbook and download a song from iTunes... no IBM PC required!
Oh god, 4 weeks with the new machine and I'm turning into a mac zealot!
I've used them for years with no hassle. But they have a don't-screw-the-customer policy I like. The first thing in their contract is: The Client owns the Domain Name registered. That's a big difference from other places, like NetSol, who don't view names as property and feel like they can keep it (and auction it off to the highest bidder) should you ever quit hosting the name with them.
Also, when I joined, Gandi was a not-for-profit service... I can't find that on their webpage now, but that's the spirit they operate under.
I used to do embedded programming where it was really costly to fix in the field.
Here's a similar project's repair estimates (pdf). Mind you, this product cost 1000x our product, but since they were at similar customer sites, the repair cost wouldn't be significantly different.
Service trip #1 = $413 million
Service trip #2 = $497 million
Service trip #3 = $547 million
Service trip #4 = $400 million
(note: these prices don't include airfare)
In fact, it would be far cheaper to just toss out our old hardware and start from scratch ($13 million total costs) than it would be to try to fix it in the field.
A machine shop owner could just find spamming the forums of their favorite magazine :-)
Excellent arguement and well put, but there is the exception that the dongle-protection-creator is exceptionally dumb and implements the same software interface in all products. Then, someone could write a rouitine that searches the executables for a function called Check_Dongle_Attached and patch it so that it always returns 1.
EDA tools us flexlm extensively - they've virtually got a lock on the market because phb's won't allow more than one kind of licensing scheme to keep up with (and run servers for). Flexlm is also priced attactively enough for software producers so that it's just under the cost for any one company to roll-their-own (~$30k). It offers good features, but security-wise, it still sucks and is regularily broken for all products that use a particular version number of flexlm.
Thanks. I was using only conductance because that's the proported use of this material - as an insulator. (Also, I'm a bit rusty on the formulas of convection and radiation!).
Also, for the radiation measurement, I'd think you'd be facing the earth on reentry, and thus use its temperature instead of 2.7K. In low earth orbit, there are big variations in the temperature seen (deep space/earth/sun), but since about half of your view is always earth, a proper thermal system will keep the average temperature in a satellite close to room temperature. Far less stringent than automotive temperatures. But, they are still designed to -55/+125C just in case. (that's just satellites, where there isn't the massive heat of reentry)
Apparently the Chewbacca defense works!
I suspect they do other testing, including:
Water resistance. Not only so it doesn't wash off on rainy days, but doesn't absorb water so that freezing causes it to crack.
Free oxygen erosion. Low earth orbit exposes the leading edge of spacecraft to free oxygen (O, not the stable O2), which tends to 'rust' things quickly.
Thermal coefficient of expansion matching to the aluminium body, so it doesn't flake off. If it isn't matched, then you need a good adhesive system.
Impact resistance. Does it chip or flake? You don't want a catastrophic failure mode (a super high-speed micrometeroite should make a hole instead of shatter the whole thing)
Weight. They stopped painting the booster tank and saved a lot of weight. Current shuttle tiles are foam-like in weight.
Repairability. Do you need to resurface the whole shuttle for the slightest chip, or is it fixable?
Lastly, NASA wants a proven scientific theory of operation... something better than "It dissipates heat at an exponential rate, it's beyond belief, and I have no idea why it does, all I know is that it does." All things dissipate heat at an exponential rate - heat flow is usually related to a difference in temperatures, so as an object reaches the temperature of its surroundings, the heat flow slows down to aproach zero. That's pretty basic to understanding heat flow, and not novel.
That was a cool game, but bullfrog got hit with a crazy frivolous class-action lawsuit. It seems that about 3% of the gamers couldn't handle it and were maimed for life. In the settlement, Bullfrog agreed to build a special colony just southeast of Detroit where these gamers could live out the rest of their days
Yep, the numbers must be wrong.
And if not, they have a 6 megawatt coal-fired power plant within spitting distance of the engineering buildings. That's 5 kilowatts per computer; plenty for either a house or a computer.
Incidently, after years of putting up with coal dust, I thought we would finally see the benefits of living next to a huge coal pile. We had a severe ice storm and most of blacksburg lacked power. but so did campus!.
The raincoat idea ... we already have completely waterproof raincoats. What people really want is a surface that is more breathable, too. Does this have the potential to be better than Goretex?
"make xconfig" or "make menuconfig" ?
Ahh, that explains it - thanks. I haven't heard that term since TFT took over. I never thought of it until now, but maybe the lack of ambiguity is what TFT became preferred.
p.s. wow, my original post was poorly written. sorry.
I noticed this:
Favorite Portable Workstation: QLI 15" AMD NOTEBOOKS
Since AMD laptops seem few and far between, I looked for some more info. It seems that QLI doesn't offer a 15" AMD option among the laptops they offer. The only AMD laptop they offer is 14.1", and if this is the one they meant, I have no idea why it would win. Other than the linux preinstall, it's totally undistinguished. The screen is 1024x768, has shared graphics memory, and the main memory runs at only 133MHz -- hardly a "workstation".
Is this a mistake, or did they used to offer an awesome computer?
Ugh- as an electrical engineer, that smart start product looks dangerous.
Two problems:
1. It ramps up voltages. This usually isn't a good thing for digital logic - it means its going to spend some time at a marginal voltage level. Hopefully this ramp is fast.
2. It sequences the seperate power supplies without sequencing the logic. When you've got something connected but not powered, the signal lines power the chip through protection diodes - the effect is that the IDE controller is trying to power the drive logic. Maybe the drive is made to disconnect itself in this situation, but it would take extra circuitry and is not standard practice. Powering up the motor 4.5 seconds after the logic probably isn't a good idea either - the drive is probably wondering why it can't spin up. (good thing it probably has an interlock to prevent dragging the heads across the platter when the drive isn't spinning, otherwise you'd have a head crash)
The best solution is probably the free, built-in one. Usually drives have a jumper setting for delayed spin-up. Read the drive manual and use that. It's a drive firmware solution, without messing with the drive's power supply.
It's not that I want something for free just because, and it's not about shelling out the muulah.
When I bought the machine, I knew there was a risk that I wouldn't get an upgrade to Panther, and if they had a consistant policy, I'd be happy to live by it.
But they are favoring one particular product - the G5 - over all others, even though my machine is newer a newer model, costs more, was bought later than most of the G5s. It's even aimed at the same professional market. Just an explanation - Isn't this supposed to be the year of the laptop?
Sounds like this may be a inrush current problem... maybe something is taking extra juice to start up, or the supply can't deliver the initial kick. A multichanel scope would help a lot to see that all voltages are getting to where they need to be in a reasonable time. Otherwise, try removing everything that you can and see if it comes on (drives, ALL ram, usb, pci cards, modem, etc.) - maybe eliminate it that way. Remember, a bad card (one that's drawing too much current) may not cause the system to fail until another good card is inserted and pushes the power supply over the threshold.
Thanks. I just did. And on to the apple store on my way home tonight...
Ha ha - Me too, once you pointed it out. But were' laughing with you PiusII, not at you. Actually, I think Xcode will be free to anyone - I remember a blurb saying dev tools would be free, but couldn't find it again on their website.
It's not so much the actual features that rubs me wrong, its just getting left out of the update. I know I'm whining. I mean, anyone who bought a laptop yesterday is getting screwed, too; I just want some logic: "30 days" or "latest round of hardware" are both good, "16 days or the only a certain line of computers that aren't the newest" just doesn't make sense.
Also, I misspoke - the current preview doesn't have postscript support, only pdf -- I generate my own ps and only the newer version will work for me. Of course there are other programs, or I could convert ps2pdf (which isn't in 10.2.7).
When I bought my last laptop, Office 2000 was on the horizon, and I was upgraded three months later when it was released. Maybe my case - 5 weeks - is pushing it, but did they really have to screw everyone who bought a laptop yesterday?
Three years ago I switched completely to Linux. I wanted a unix machine that was well integrated because I was tired to trying to get Red Hat to do simple stuff (example: native MP3 support was discontinued because of license issues, DVD playback). RH is great - I've bought 2 copies retail and my desktop runs RH8 for work, but I just didn't want to do admin chores while on my own time.
The pdf support is supposed to be "the fastest of any machine ever"... probably marketing hype, but I use a lot of large postscript files and render speed is important.
That sucks. I shelled out $3k for their latest revision of the 17" powerbook on the second day it was released (Sept 17), only to find that the features they have been touting for so long (X windows, native & fast PDF support, X code) are going to cost. Why only the top of the line desktops - why not my more expensive top of the line laptop? My machine is newer than many G5s (although maybe they were trying to satisfy customers with long preorder times).