The standard is open, but it costs to play, that's for certain (LonWorks), which is, IMHO, wrong.
It *has* opened up the HVAC markets quite a bit, although it's really only the big boys that can afford to use it, but it's no longer a case of where only say, Honeywell, can provide the parts for a building HVAC system, because you can use siemens, trane, or a number of other thermostats and sensors in the system. But a low-volume startup simply cannot afford the development kits to get started.
LonMark hasn't helped much, either.
*****
The problem with the consumer electronics guys is that they really do *not* want to use an open standard. Sony only wants you to buy Sony gear, why would they want their Sony DVD player to talk to someone's Toshiba plasma screen, when they can use the fact that someone already owns Sony gear to help push them towards a Sony display?
We know it works both ways, but talking companies into seeing that just does not happen. And if a company is really good in 80% of the home theater components, but not so great in that last 20%, then they can leverage the proprietary control protocol to get people who have bought from the 80% to also buy from the 20% not so good, as it fits better into the system.
*****
I think the low data rate control protocols will do some amazing things for home automation, but companies need to have affordable ways to do it, and open standards to work with, otherwise the hobbyists that will drive it simply aren't going to be able to do anything. And X10 sucks, so that's not really a good option.
I really think it's in both interests to be in a position to work together like that. Gmail is highly interactive html, and XUL is way cool, but is limited to the moz platform.
But gmail's work with dynamic html coupled with working on the browser, they get both perspectives, the server-side of the application provider, and the client-side browser, and with both views, they can lay out the right way to break up the work.
Developing both interfaces at the same time is much better than guessing what the other will need in the future.
Because frequency response is measured with a purely resistive dummy load. Speakers tend to have wildly varying impedances depending on the frequency, and if the impedance goes low enough, then the driver which has a flat frequency response at high impedances now can no longer push enough current to keep the frequency response flat.
Numerous high-end headphones will try to pull more current than a lot of consumer equipment can push, which is why there are headphone amps (well, that and to add some cross-mixing to help the ears out a bit).
The resolution of a TV is based on the phosphor grid layout, and they purposefully design it to overlap with itself, so that you can't get perfect square pixels. The result is that the display anti-aliases itself in analog. Which means that all resolutions are fuzzy.
CRT monitors do the same, but using much, much smaller phosphors in the grid than does a TV. Display a white image on a tv, and grab a magnifying glass, and then do the same with a CRT computer monitor and an LCD computer monitor.
Good digital displays (LCD, DLP and Plasma) should up-covert to their native resolution internally. So if you send a 480p signal to a display that's 1280x720, it will upconvert it to 720 internally to give the right resolution.
Feed them via DVI, and even better, as the DVD player can send a signal for exactly the right resolution (DVI allows some bidirectional data for asking the display what aspect ratio/resolution it is).
The new plasmas do last 10 years, with little to no degradation in brightness.
A friend of mine owns a video rental store, and has two WEGAs and a high-end plasma for showing movies in the store. The plasma is running 12 hours a day, 7 days a week. And it's still an amazing picture. Granted, it was the most expensive unit in that size when it was new, but it's been used as a display model for over 2 years now. That's like 60K hours.
But even high end CRTS like Princeton units suffer from the problems that plague all CRTS. I helped a friend with calibrating his display, and with a proper blackpoint set, full contrast white and black still bleeds over. CRTs just do that.
What's especially noticeable is the fadeout from full white to full black, especially on a moving image. If it's moving fast enough, you can see each frame ghosting out.
A comparable quality plasma doesn't have this problem at all, as the response rate is much higher.
Due to the encrypted DVI datastream from cable-boxes and DVD players, you need to get an LCD "TV" vs. an LCD comptuer monitor.
!@#!@#!@$$$ MPAA....
The only difference is really that the LCD TV has the decryption keys in it for the DVI streams. They may also have a tuner, but the really important distinction (to me), is the ability to decrypt the digital stream from DVDs.
However, you can still buy Black. You can't buy copies of the originals...
Re:Quick Question
on
Defining Google
·
· Score: 2, Interesting
Supervisory Control and Data Aquisition. Also called HMI (Human-Machine Interface), previously MMI (Man-Machine Interface).
Virtual guages and logging of data, sometimes with alarming. Basically a GUI for a plant-floor, or a chemical plant, or a water treatment facility, etc.
Largish commercial market, but there's not a lot to know, although I don't understand why so few people actually grok it who are involved in it.
I used to work for Rockwell Software, as an apps engineer on their RSView32 HMI development and runtime engine. You could do a lot of nice stuff with it, but it was obvious most people hae no idea how to make a useful UI.
In order to have any kind of automated backup solution, a human attacker will be able to get to it.
I see a couple easy blocks to these, though:
1) write a shell script for mounting the backup drive, both onto the SCSI chain and into the filesystem, performing the backup, and then unmount it.
2) round-robin the drives on a regular basis, so an IT monkey can physically swap out sets when needed to provide off-site storage (basically use hot-swap bays like very large, fast tape jukeboxes).
3) encrypt the pertinent scripts, and use yet another script with a bening name to perform the decryption of the shell script, the chmod to executable of it, and then exec'ing it.
****
Yes, it's still hackable, but it ups the bar considerably, and if you're swaping the drives out nightly/weekly, you've got good backups that are offline, and not too old.
Hasn't this been known in security circles for decades? (I first read this in Out of the Inner Circle).
If you broadcast who you are, what you're running, and especially the version (and patch-level) you're running, you are actively saying, "Hello, you can use exploits X, Y, and Z to p0wn me!!"
Tried that for a few minutes on my IBM T30... didn't go over so well. The sub-pixel rendering kept using non-grey colors, and as a result small fonts would appear in multiple colors at once, and when scrolling text, it would turn lots of different colors. Although from a little bit further away (say using a projector in a meeting room), it could look really good.
The major problem is that they have a value, but it's unknown. Which means that they essentially don't have a value.
Now, if the options are granted below the end-of-day market value for the stock on the date of the grant, then they have a value (the difference).
But if I get options for 1000 shares at $12.00, but they don't vest for 12 months to 48 months (standard), then what's their value at the time of grant? Nothing.
They are utterly worthless. For 12 months, they have no value. If I leave within twelve months, they have NO value, they didn't exist.
After 12 months, they have a *possible* value, which changes minute by minute during the hours that the market is open. Could be positive, could be negative. Nobody knows.
When I exercise them, they suddenly have a value. They are worth the market price. If I sell them that day, I get the difference between the market price and the option price. The company receives the option price as an investment by me into it.
If I exercise them but not sell them, usually a collossally bad idea, then I get no money. I have invested into the company, and received shares, but at option price instead of market price, and I have to worry about AMT on the difference.
*****
Stock options are great for a private company (because the FASB doesn't apply in the same way, since there's no market value).
For public companies, I don't think it makes sense. Because now part of your compensation is tied to the market, which is risky, because most of the market doesn't have a clue about what high tech companies do. It's amazing how many investors have NO IDEA what a company that they are investing in does.
The real danger of options is options to the executive staff. They are in a position to receive a lot of stock, and to easily manipulate the stock to their short-term advantage, and then run with the money (Enron). But that's something that can be worked with.
Calling the system flawed because it has a loop-hole that can be closed by itself is inane.
*****
Options fueled the Valley. People would work for peanuts (literraly), at a startup, because there was the promise that if the company got sold, they'd get money or shares in the new company, or if the company IPOd, the original founders would be rich as hell. If you're young, no family, the risk/reward ratio is awesome. Not for someone that has to worry about thier kids.
But even away from that extreme, I'd gladly take a hit in the paycheck for a shot at building a company up to be worth a LOT of money, if it was going to pay that back out to me in the form of investment in the company.
****
Flip-side. I think the market is broken. Seriously broken. It's ruled by short-term gains, which cause very bad business decisions to be made. Sure, the company could change things to be more profitable, but could the company sustain it's customer base at those levels of quality?
Rebuild natural massive carbon sinks like plankton, sea algae, and rainforests.
The total amount of carbon on the planet (including CO2 in the atmosphere) should be constant (aside from whatever lands via meteorites).
As we burn it, it ends up in the atmosphere as CO2. We can continue to burn it, if we also provide for a mechanism to re-sink it back below the surface.
Burning coal is half the problem. Burning coal while at the same time cutting down millions of acres of rainforest, or poisoning off large algae growths in the ocean compound the problem.
The silicone valley area is an excellent example. Over the last 40 years, as more and more people have moved in, the orchards have been removed, and it's now mostly concrete. Very little in the way of greenery (more than most midwest cities, though...) In addition to this, there're now also a ton more people using electricity and driving around, dumping CO2 from tailpipe emissions, but also using meagwatts of electricity (which become heat, in the end).
The trees that used to be here are no longer here to provide a sink for the C02, and they no longer shade the area. As a result, summer temps are higher, winter temps colder, the storms more violent, and the local ambient CO2 levels keep going up.
I agree that we should use whichever method is best, but I think that we can determine what that is.
Invalid votes (hanging chads, two presidential choices, etc.) I think can be viewed as errors, and a system devised that works to minimize those kinds of problems.
We cannot know if the voters intent made it to the ballot, but if the ballot is clear in what the ballot is trying to say, then it's just a matter of the voter confirming that the ballot says what they want it to say.
And since the only mechanism we *can* look at is the number of votes that couldn't be read, or had errors on them, then that's our error of margin. Some percentage of votes cast had to be ignored, because they couldn't be understood.
What're you calling problematic ballots, and what are errors?
If it cannot be determined, absolutely, what the intent of the ballot was, then that's an error in the voting process (even user error for something like this points to voting process).
However, when votes can be improperly cast, it starts to look a lot more like statistics, and less like exact counts of the opinion of all registered voters.
Except, we're not getting 100% voter turnout, and we're not getting 100% properly cast votes. And not 100% of the eligible voters are registered.
So, the voting process we have is a statistical analysis of the big picture. And apparently, not a great one, if it's this questionable.
They're not perfect, but anything rejected by the computer can be counted by hand. The idea (to me) behind optical scan is that it can be counted both by hand and by machine. And the machine is only there to assist in the counting, not to be "the counter"
Slashdot Mirror
The standard is open, but it costs to play, that's for certain (LonWorks), which is, IMHO, wrong.
It *has* opened up the HVAC markets quite a bit, although it's really only the big boys that can afford to use it, but it's no longer a case of where only say, Honeywell, can provide the parts for a building HVAC system, because you can use siemens, trane, or a number of other thermostats and sensors in the system. But a low-volume startup simply cannot afford the development kits to get started.
LonMark hasn't helped much, either.
*****
The problem with the consumer electronics guys is that they really do *not* want to use an open standard. Sony only wants you to buy Sony gear, why would they want their Sony DVD player to talk to someone's Toshiba plasma screen, when they can use the fact that someone already owns Sony gear to help push them towards a Sony display?
We know it works both ways, but talking companies into seeing that just does not happen. And if a company is really good in 80% of the home theater components, but not so great in that last 20%, then they can leverage the proprietary control protocol to get people who have bought from the 80% to also buy from the 20% not so good, as it fits better into the system.
*****
I think the low data rate control protocols will do some amazing things for home automation, but companies need to have affordable ways to do it, and open standards to work with, otherwise the hobbyists that will drive it simply aren't going to be able to do anything. And X10 sucks, so that's not really a good option.
Curious what your experience with LonWorks is?
Not a bad idea. Use XUL as the engine for handling a rich display (either instead of, or in addition to, dyn html).
Giving gmail some more "pim"ish functionallity would definitely be good for it (and me too, as I use gmail).
That was my thought as well.
I really think it's in both interests to be in a position to work together like that. Gmail is highly interactive html, and XUL is way cool, but is limited to the moz platform.
But gmail's work with dynamic html coupled with working on the browser, they get both perspectives, the server-side of the application provider, and the client-side browser, and with both views, they can lay out the right way to break up the work.
Developing both interfaces at the same time is much better than guessing what the other will need in the future.
Yes, and no.
Because frequency response is measured with a purely resistive dummy load. Speakers tend to have wildly varying impedances depending on the frequency, and if the impedance goes low enough, then the driver which has a flat frequency response at high impedances now can no longer push enough current to keep the frequency response flat.
Numerous high-end headphones will try to pull more current than a lot of consumer equipment can push, which is why there are headphone amps (well, that and to add some cross-mixing to help the ears out a bit).
Try using 1024x720, and what are you using for the interconnect? DVI? Component? Composite?
If analog, tell the TV you're giving it a 16x9 signal, and send it that resolution, and you should be good.
From a computer, 1024x720 will be 720p.
The resolution of a TV is based on the phosphor grid layout, and they purposefully design it to overlap with itself, so that you can't get perfect square pixels. The result is that the display anti-aliases itself in analog. Which means that all resolutions are fuzzy.
CRT monitors do the same, but using much, much smaller phosphors in the grid than does a TV. Display a white image on a tv, and grab a magnifying glass, and then do the same with a CRT computer monitor and an LCD computer monitor.
Good digital displays (LCD, DLP and Plasma) should up-covert to their native resolution internally. So if you send a 480p signal to a display that's 1280x720, it will upconvert it to 720 internally to give the right resolution.
Feed them via DVI, and even better, as the DVD player can send a signal for exactly the right resolution (DVI allows some bidirectional data for asking the display what aspect ratio/resolution it is).
The new plasmas do last 10 years, with little to no degradation in brightness.
A friend of mine owns a video rental store, and has two WEGAs and a high-end plasma for showing movies in the store. The plasma is running 12 hours a day, 7 days a week. And it's still an amazing picture. Granted, it was the most expensive unit in that size when it was new, but it's been used as a display model for over 2 years now. That's like 60K hours.
But even high end CRTS like Princeton units suffer from the problems that plague all CRTS. I helped a friend with calibrating his display, and with a proper blackpoint set, full contrast white and black still bleeds over. CRTs just do that.
What's especially noticeable is the fadeout from full white to full black, especially on a moving image. If it's moving fast enough, you can see each frame ghosting out.
A comparable quality plasma doesn't have this problem at all, as the response rate is much higher.
Due to the encrypted DVI datastream from cable-boxes and DVD players, you need to get an LCD "TV" vs. an LCD comptuer monitor.
!@#!@#!@$$$ MPAA....
The only difference is really that the LCD TV has the decryption keys in it for the DVI streams. They may also have a tuner, but the really important distinction (to me), is the ability to decrypt the digital stream from DVDs.
What resolution are you trying to output to it?
However, you can still buy Black. You can't buy copies of the originals...
Supervisory Control and Data Aquisition. Also called HMI (Human-Machine Interface), previously MMI (Man-Machine Interface).
Virtual guages and logging of data, sometimes with alarming. Basically a GUI for a plant-floor, or a chemical plant, or a water treatment facility, etc.
Largish commercial market, but there's not a lot to know, although I don't understand why so few people actually grok it who are involved in it.
I used to work for Rockwell Software, as an apps engineer on their RSView32 HMI development and runtime engine. You could do a lot of nice stuff with it, but it was obvious most people hae no idea how to make a useful UI.
In order to have any kind of automated backup solution, a human attacker will be able to get to it.
I see a couple easy blocks to these, though:
1) write a shell script for mounting the backup drive, both onto the SCSI chain and into the filesystem, performing the backup, and then unmount it.
2) round-robin the drives on a regular basis, so an IT monkey can physically swap out sets when needed to provide off-site storage (basically use hot-swap bays like very large, fast tape jukeboxes).
3) encrypt the pertinent scripts, and use yet another script with a bening name to perform the decryption of the shell script, the chmod to executable of it, and then exec'ing it.
****
Yes, it's still hackable, but it ups the bar considerably, and if you're swaping the drives out nightly/weekly, you've got good backups that are offline, and not too old.
is not a good idea.
Hasn't this been known in security circles for decades? (I first read this in Out of the Inner Circle).
If you broadcast who you are, what you're running, and especially the version (and patch-level) you're running, you are actively saying, "Hello, you can use exploits X, Y, and Z to p0wn me!!"
Tried that for a few minutes on my IBM T30... didn't go over so well. The sub-pixel rendering kept using non-grey colors, and as a result small fonts would appear in multiple colors at once, and when scrolling text, it would turn lots of different colors. Although from a little bit further away (say using a projector in a meeting room), it could look really good.
Lock == Login Request ??
The major problem is that they have a value, but it's unknown. Which means that they essentially don't have a value.
Now, if the options are granted below the end-of-day market value for the stock on the date of the grant, then they have a value (the difference).
But if I get options for 1000 shares at $12.00, but they don't vest for 12 months to 48 months (standard), then what's their value at the time of grant? Nothing.
They are utterly worthless. For 12 months, they have no value. If I leave within twelve months, they have NO value, they didn't exist.
After 12 months, they have a *possible* value, which changes minute by minute during the hours that the market is open. Could be positive, could be negative. Nobody knows.
When I exercise them, they suddenly have a value. They are worth the market price. If I sell them that day, I get the difference between the market price and the option price. The company receives the option price as an investment by me into it.
If I exercise them but not sell them, usually a collossally bad idea, then I get no money. I have invested into the company, and received shares, but at option price instead of market price, and I have to worry about AMT on the difference.
*****
Stock options are great for a private company (because the FASB doesn't apply in the same way, since there's no market value).
For public companies, I don't think it makes sense. Because now part of your compensation is tied to the market, which is risky, because most of the market doesn't have a clue about what high tech companies do. It's amazing how many investors have NO IDEA what a company that they are investing in does.
The real danger of options is options to the executive staff. They are in a position to receive a lot of stock, and to easily manipulate the stock to their short-term advantage, and then run with the money (Enron). But that's something that can be worked with.
Calling the system flawed because it has a loop-hole that can be closed by itself is inane.
*****
Options fueled the Valley. People would work for peanuts (literraly), at a startup, because there was the promise that if the company got sold, they'd get money or shares in the new company, or if the company IPOd, the original founders would be rich as hell. If you're young, no family, the risk/reward ratio is awesome. Not for someone that has to worry about thier kids.
But even away from that extreme, I'd gladly take a hit in the paycheck for a shot at building a company up to be worth a LOT of money, if it was going to pay that back out to me in the form of investment in the company.
****
Flip-side. I think the market is broken. Seriously broken. It's ruled by short-term gains, which cause very bad business decisions to be made. Sure, the company could change things to be more profitable, but could the company sustain it's customer base at those levels of quality?
Rebuild natural massive carbon sinks like plankton, sea algae, and rainforests.
The total amount of carbon on the planet (including CO2 in the atmosphere) should be constant (aside from whatever lands via meteorites).
As we burn it, it ends up in the atmosphere as CO2. We can continue to burn it, if we also provide for a mechanism to re-sink it back below the surface.
Burning coal is half the problem. Burning coal while at the same time cutting down millions of acres of rainforest, or poisoning off large algae growths in the ocean compound the problem.
The silicone valley area is an excellent example. Over the last 40 years, as more and more people have moved in, the orchards have been removed, and it's now mostly concrete. Very little in the way of greenery (more than most midwest cities, though...) In addition to this, there're now also a ton more people using electricity and driving around, dumping CO2 from tailpipe emissions, but also using meagwatts of electricity (which become heat, in the end).
The trees that used to be here are no longer here to provide a sink for the C02, and they no longer shade the area. As a result, summer temps are higher, winter temps colder, the storms more violent, and the local ambient CO2 levels keep going up.
Incandescants don't flicker at 60Hz, they take too long to heat up. CF and cheap flourescents, however, will flicker (drives me up the wall).
Ok, that makes more sense to me now.
I agree that we should use whichever method is best, but I think that we can determine what that is.
Invalid votes (hanging chads, two presidential choices, etc.) I think can be viewed as errors, and a system devised that works to minimize those kinds of problems.
We cannot know if the voters intent made it to the ballot, but if the ballot is clear in what the ballot is trying to say, then it's just a matter of the voter confirming that the ballot says what they want it to say.
And since the only mechanism we *can* look at is the number of votes that couldn't be read, or had errors on them, then that's our error of margin. Some percentage of votes cast had to be ignored, because they couldn't be understood.
Was that more or less this time around?
That's not exactly clear.
What're you calling problematic ballots, and what are errors?
If it cannot be determined, absolutely, what the intent of the ballot was, then that's an error in the voting process (even user error for something like this points to voting process).
Only on /. would people actually bitch about the spelling mistake that people make while posting online.
It's not worth my time to spell-check my posts.
However, when votes can be improperly cast, it starts to look a lot more like statistics, and less like exact counts of the opinion of all registered voters.
Except, we're not getting 100% voter turnout, and we're not getting 100% properly cast votes. And not 100% of the eligible voters are registered.
So, the voting process we have is a statistical analysis of the big picture. And apparently, not a great one, if it's this questionable.
They're not perfect, but anything rejected by the computer can be counted by hand. The idea (to me) behind optical scan is that it can be counted both by hand and by machine. And the machine is only there to assist in the counting, not to be "the counter"