I'd bet on managers pushing through the update before it was ready.
More likely it was an underfunded test configuration. If test systems even exist. When systems are upgraded to meet production needs, rarely are test systems upgraded. Test systems tend to be 'trickle down' systems that are no longer useful in the production environment. They never have enough processor, disk space, or the ability to put a production equivalent load on them.
But; for any moderatly sized OSS project there is a quantity of people who are are all to happy to help (provide support) to you for free through mailing lists, bulletin boards, IRC etc...
That support isn't good enough for most businesses. If you ran a corporate website on open source software and had a problem, you probably couldn't post it on a mailing list and hope for a response. Businesses buy support contracts for priority. That's why they'll pay IBM to support a server, with a 4 hour response time, that is under warranty. Service contracts are a simple expense, a cost of doing business at a set cost. Just like software licensing. A system outage is an unknown, and the 'free' part is an insignificant part of the total loss.
Managers feel more comfortable having a phone number they can call when there is a problem, even when, as is often the case, the response is that the support people don't have a clue what the problem is, much less how to fix it. That's just the way people are; we want to know someone is listening when we have a problem.
And it's not that corporations don't understand free. They understand that you get what you pay for. Is it practical for me to learn how to build and configure a firewall, or pay a consultant to do it (and maintain it) for me?
But you are correct that most non-business users wouldn't buy support contracts. The time they spend working on a problem isn't lost revenue. For example, I've been tinkering with MythTV for months. I even bought new hardware. But my time is less valuable than getting something that works exactly like I want it to. As a business decision, I would have just bought a Tivo.
Hmm, I originally read your post as $40 for the device, plus development costs. Doesn't really matter though. Just like the software industry, financial services need to be continously improved. They're already spending the money; for instance RFID credit cards so you don't have to swipe your card. The only thing difficult about swiping your card is they can't make up their minds whether the reader is going to be on the left or the right.
Credit card companies are looking for ways to get people to spend more money on the internet. Make plastic more convenient than cash, because they get a bite out of every transaction. They prefer plastic to checks because they are easier to process. Banks are on an efficiency kick right now. Not only are they pushing their online banking, I was recently offered free bill-pay just for signing up for electronic statements. They save money on both ends of that deal if I write fewer checks. Kind of like a few years ago my bank used to charge 15 cents to use an ATM card. Then they found out if they encouraged ATM usage, they needed fewer tellers.
So what it comes down to is spending money to make money. Security isn't a selling point for customers, but convenience is. Am I going to carry (and type) one use passwords? Probably, but my mom would not. It's too complicated. Give her a USB dongle that fits on a keychain, it's almost as convenient as RFID, but it can be used with the home PC without special hardware.
The only people I've known who've been ripped off, had their credit cards abused, etc, had it happen by not taking the carbons when they use Visa to pay for gas.
What the hell is a 'carbon'..... Seriously, from a customer standpoint, pay-at-the-pump is good because no one ever touches your card but you. Something you should watch for are merchants that still put the CC number on receipts. They are only supposed to be putting the last 4 digits now.
BTW, I had one of my credit card #s stolen a couple years ago while I was on vacation. Since I have a special card for that, I know that it was either taken at a motel in Colorado or a mom and pop gas station in Utah. The morons actually used it to buy Bonzi Buddy. (and a $1500 camera on eBay)
Equipping your 1 million customers with some kind of secure random password generation smart card probably costs $40 each, both for the card and programming as well as associated infrastructure and overhead costs.
I've seen USB devices that were considerably less expensive than that. At least the hardware was... (I was looking for some computer access technology)
You could use devices with a static identification, combine it with an account number and pin/password, and provide access that would be as secure at your home PC as it would be at an ATM. It would probably be cost effective to add single use password/credit card numbers capabilities too. One of the big drawbacks with American Express' Blue card was the requirement of a special chip reader. (and the requirement for merchants to have affiliate accounts rather than creating single use CC #s) For a million customers, you could probably get the devices in the $1-5 range pretty easily.
You don't actually believe your bank checks the signatures on your checks, do you? Not likely, unless it's a really small bank or there is a dispute.
Some banks are processing checks like EFTs (Electronic Funds Transfers) already. For instance, if you have a Citibank credit card, you write them a check to pay your bill, they process it like a debit card and then destroy the check. Hopefully they image it first, but that wasn't mentioned as part of the process when they explained it to my mom.
If you have Direct Deposit on your paycheck, the same routing information that is used to deposit your paycheck can be used to remove it.
As far as I can tell, the only protection customers of US banks have is the goodwill of banking institutions. Rather than authenticate, they seem to use a 'trusted peer' methodology.
So, compared to their EFT systems, it looks like their web sites are pretty secure.
BTW, I have one account that requires a 6+ digit password, with the first character being alpha, one character has to be numeric, and one punctuation. Sure, it's safe from a dictionary check, but I can never remember the damn thing.
And my college assigns passwords with random characters for e-mail and enrollment. I'm sure it's so much more secure for me to write it down and carry it in my wallet than it would be to let me choose it myself.
It cost me $850 for the control panel alone but it was WELL worth the price.
That's a nice looking panel, but $850? That's outrageous. I'm working on two cocktail cabinets right now. I have spent $40 each for my IPACs and $30 each for the controls. (2 joysticks, 12 player buttons, Player 1 and player 2 buttons, and a couple extra. Excluding the computer equipment, I won't be spending $850 for both cabinets.
Of course the interpreters for those languages are usually written in C or C++, so we haven't quite escaped the problem yet.
Well, since Delphi doesn't have an interpreter, I would say as least some of us have escaped the problem. Haven't had a buffer overflow in an application in 10 years. Pointers and range checking hasn't been a problem either.
Another Gateway favorite was "it's a problem with your modem (or video) drivers, download news ones from our website and install them. If you still have a problem, then call us back." They would use that for any problem, not just modem/video issues. The problem they were trained to fix was how to end a call in less than a minute.
(I've never been a Gateway tech, but several of my friends were. As a result, I've never purchased a Gateway computer)
There's always a risk that any application that's handling data, especially unclean internet data, can be the victim of a buffer overflow.
Insightful? Not even close. Buffer overflows aren't a given. They aren't a fact of life. Quite simply, all you have to do is simple bounds checking. If you allocate a 4k buffer, don't try to copy 6k to it. Buffer overflows are a 'feature' of C/C++. There are plenty of other languages that don't have that problem.
Unchecked buffers are the result of poor program design. No programmer, or company, should say that they are security conscious if they haven't done a code review and fixed them.
Copyright infringement is a civil matter, not a crime.
At least it wasn't until the creation of the DMCA. Such a nicely orchestrated move to shift enforcement burdens from the copyright holder to the public taxpayer.
But when a pirate does it, gets sued, and settles, somehow it's evil that the RIAA sued in the first place and the pirate is the good guy martyr.
First of all, I could care less whether people get their free music. I think it's a bad business decision to not make it available in some form, because so many of us won't buy a new CD until we have heard all of it since we are tired of getting burned buying CDs with a solitary good song.
But downloading music is a bad comparison. People aren't downloading MP3s so that they can incorporate it on their own CD and sell it. If they are, they are distributing and should be sued by the copyright holder. SCO was incorporating someone else's IP in their product without prior authorization.
One graph has ext2 first, the next has xfs. ext3 is light blue in one, purple in the next. I quickly found that after having to recheck the key for each graph, I really didn't care anymore.
> How about geothermal? Iceland has had a lot of success with that.
Well, maybe because it is a small island directly placed above a contintental rift. I don't know for how many regions that would work.
Technically, ground sourced heat pumps are geothermal too. Just at a lesser degree. That is available anywhere.
It's not like we need a silver bullet.
My original point was that wind and tidal power aren't silver bullets. The original poster was pointing to wind and tidal power as the solution to those evil dams. But they have drawbacks too. It's going to have to be a balance of several technologies.
I doubt there is anywhere on earth that is entirely unlivable. Maybe by us, but we aren't the only organisms here. Keep in mind that drastic changes in other habitats tend to cause drastic changes in ours.
Or, you can look at it another way. It's not possible for man to destroy the ecology of the earth. It will still be here (in some other form) long after it can no longer support us.
At wholesale energy prices. So you pay them full retail price for power you use, and they pay you a lot less for the power you sell back to them.
Don't oversize your system. This isn't a money making adventure. Unless you have time and date metering, they are just checking your meter once a month. If it moves forward, you pay. Backwards, they do. The wholesale/retail problem would only come into effect when you're trying to carry over from month to month.
Most off-grid people use very little eletricity, even to the point of unplugging wall warts when not in use, because they draw a few watts even when you aren't using what they are connected to.
Also, unplug all of your vampire appliances. These are appliances that use power even when turned off. A lot of times, off does not mean off. Every manufacturer wants to install clocks and LEDs. For instance, my Apex DVD has a power switch and you can turn it off with the remote. The remote just suspends parts of the system, the LED stays lit, the circuitry is live to listen for commands from the remote. A lot of stereos are like that too. My AudioTron has that annoying problem too. The power switch on the front causes it to hibernate, the true power switch is inaccessibly on the back.
Hydro power is now on the way out as a major power source. Many dams have been removed in Western countries because they lead to salinization of cropland, destruction of hatcheries, and they just cost so bloody much.
Salinization isn't caused by dams, it's caused by irrigation. Cost isn't the issue, it costs more to tear them down. While they are there, their operation profits cover expenses. The problem is the destruction of natural habitats. I'll admit that a thorough environmental analysis needs to be done before building a dam, but I have to wonder how many habitats are restored by a dam's destruction. You don't just drain the reservoir, blow up the damn and everything magically returns.
Besides, wind turbines will cool the atmosphere by some tiny amount to offset global warming.
Oh, I'm sorry, that wasn't meant to be a joke??? Wind turbines kill birds. A significant number, hard to say. I haven't seen any serious studies. Tidal power requires changing the shoreline. What effect is that going to have? Ever wonder why Florida is building artificial reefs? It's because development in one area is causing beach scouring in another.
How about geothermal? Iceland has had a lot of success with that.
The sell high/buy low only works if your meter keeps track of the time of day you are producing and consuming. And you have to be on a special rate program, something power companies don't like to use on residential service. But the grid makes a nice, maintenance free battery. Even when you buy and sell at the same price.
If you want to be completely solar, you have to attack the problem on several fronts. You need to find any way that you can to reduce power consumption. Then there are storage and backup power problems to deal with.
Since you didn't post capacity, it's hard to say what the $17k covers. I don't know if the programs are currently active, but in the past California had a tax rebate program that could offset about 1/2 of the cost of installation. I would be surprised if they missed that in the quote, but you might want to check. It would drastically cut payback time. Also remember that part of your payback comes in non-monetary benefits.
The type of system you were looking at is a good one, but probably needs a few adjustments. Being connected to the grid has a lot of advantages. The grid serves as your batteries. If your usage spikes (air conditioning?), the grid will make up the difference. And the grid supplies your power at night and when you can't produce.
When the grid goes down, you don't necessarily have to shut down too. When the grid goes down, you DO have to disconnect your PV units from the grid, regardless of whether they are producing or not. Neither you nor PG&E wants you powering their lines and electrocuting their linemen.
And if you look at it that way, compare the cost of lost revenue from over production to the cost of batteries. If losing the money bothers you that much, slightly undersize your system so you don't produce an excess amount. PG&E will happily cover the difference.
Actually, I'd say "stick to unstable" and immediately replace whatever is in Knoppix's apt sources with the Debian unstable sources, then update and dist-upgrade your system
Unstable isn't very good for new users. They'll get too many broken packages and no idea how to fix them. And god forbid they have a windows mentality when they learn the power of apt-get. A dist-upgrade will almost surely kill their system with the default 'testing' settings.
In regard to my problems with KDE, it came from not paying attention to what I was doing with apt. I had MythTV stable with apt-get testing. Both should have been set to unstable. When it had dependancy problems, I was stupid enough to try -forcing a package that backleveled KDE. Even installing the unstable KDE didn't clean up the mess. But if you install Knoppix, set apt to unstable, and then apt-get install KDE it takes you to 3.2.2 just fine.
Now that 3.4 is out, I need to burn it and see if it can now detect the sound card in my Dell. And they've added some support for wireless too.... Fortunately I have plenty of test boxes to work with.
Well, I can see that telling people who "want cell phones to be just cell phones" to buy secondhand is a little extreme.
Extreme? I was thinking condescending.
It's like telling everyone who wants a car to buy a limo. And if you don't like all the features of the latest Towncar, you should just buy a second hand one.
I had to replace my phone in December, and I had a rough time finding a decent phone that didn't include a camera. (I have a real digital camera, thank you very much) I ended up with a Samsung E105, and even it has a ton of features I won't use. The only non-phone feature I need is inbound text messaging, which is a much more logical extension than games or PIMs.
Give me a well made, no frills, unlocked GSM phone and that's all I need or want. If they want to sell me additional services, give me a reasonably priced PC Card w/ data service.
Overall, I'd say the risk of a patch breaking something on your specific machine (as opposed to a few random thousand of the 100s of millions out there) is much lower than the risk of a virus hitting you while you're "testing" the patches.
That hasn't been our experience here. Less than a year ago we specifically put together a plan for staged rollouts of patches. It started with a get tough plan to make sure all servers were up to date, followed by several applications on all of our middletiers working erratically. It took a week for the programmers of the effected apps to get the problem fixed and working reliably. Things were starting to get a little ugly and users were not happy. Result, we have three stages of rollouts; test systems, first half production, last half production. None of which install automatically.
I wasn't effected on that case, but I have had MS 'fixes' break critical systems. A while back a 'fix' of the generic text printer driver caused it to eat the first character of each line. Barcode printers stopped working. And no barcodes, no shipping. Spent a day finding it, added a sacrificial space to each line, system is back online. A year later, MS fixes the 'fix' and the driver is working correctly again, but now the printers are choking on the extra space. Pull our fix for their 'fix', and our systems are back in a couple hours. But only because I remembered the previous problem and work around.
As to timeframe; it takes time to test complicated systems. Add to that the effects of the ecomony and companies are expecting more from fewer developers. So we have to balance our time between business requirements and testing MS patches. Being late installing a patch doesn't show up on my annual review, missing development deadlines does.
As far as getting hit; we don't get hit very often, today is the first case of an infected server that I can remember since code red hit our website. We have up-to-date scanning on our systems, SUS for desktop patches, email scanning, and properly configured firewalls.
Today we are fighting with a variant of a worm that isn't being detected by our scanners. But also doesn't appear to be using a vuln fixed by any patch. But that's a problem for Operations; developers are coding today, not chasing MS bugs.
Slashdot Mirror
I'd bet on managers pushing through the update before it was ready.
More likely it was an underfunded test configuration. If test systems even exist. When systems are upgraded to meet production needs, rarely are test systems upgraded. Test systems tend to be 'trickle down' systems that are no longer useful in the production environment. They never have enough processor, disk space, or the ability to put a production equivalent load on them.
But; for any moderatly sized OSS project there is a quantity of people who are are all to happy to help (provide support) to you for free through mailing lists, bulletin boards, IRC etc...
That support isn't good enough for most businesses. If you ran a corporate website on open source software and had a problem, you probably couldn't post it on a mailing list and hope for a response. Businesses buy support contracts for priority. That's why they'll pay IBM to support a server, with a 4 hour response time, that is under warranty. Service contracts are a simple expense, a cost of doing business at a set cost. Just like software licensing. A system outage is an unknown, and the 'free' part is an insignificant part of the total loss.
Managers feel more comfortable having a phone number they can call when there is a problem, even when, as is often the case, the response is that the support people don't have a clue what the problem is, much less how to fix it. That's just the way people are; we want to know someone is listening when we have a problem.
And it's not that corporations don't understand free. They understand that you get what you pay for. Is it practical for me to learn how to build and configure a firewall, or pay a consultant to do it (and maintain it) for me?
But you are correct that most non-business users wouldn't buy support contracts. The time they spend working on a problem isn't lost revenue. For example, I've been tinkering with MythTV for months. I even bought new hardware. But my time is less valuable than getting something that works exactly like I want it to. As a business decision, I would have just bought a Tivo.
Hmm, I originally read your post as $40 for the device, plus development costs. Doesn't really matter though. Just like the software industry, financial services need to be continously improved. They're already spending the money; for instance RFID credit cards so you don't have to swipe your card. The only thing difficult about swiping your card is they can't make up their minds whether the reader is going to be on the left or the right.
Credit card companies are looking for ways to get people to spend more money on the internet. Make plastic more convenient than cash, because they get a bite out of every transaction. They prefer plastic to checks because they are easier to process. Banks are on an efficiency kick right now. Not only are they pushing their online banking, I was recently offered free bill-pay just for signing up for electronic statements. They save money on both ends of that deal if I write fewer checks. Kind of like a few years ago my bank used to charge 15 cents to use an ATM card. Then they found out if they encouraged ATM usage, they needed fewer tellers.
So what it comes down to is spending money to make money. Security isn't a selling point for customers, but convenience is. Am I going to carry (and type) one use passwords? Probably, but my mom would not. It's too complicated. Give her a USB dongle that fits on a keychain, it's almost as convenient as RFID, but it can be used with the home PC without special hardware.
The only people I've known who've been ripped off, had their credit cards abused, etc, had it happen by not taking the carbons when they use Visa to pay for gas.
What the hell is a 'carbon'..... Seriously, from a customer standpoint, pay-at-the-pump is good because no one ever touches your card but you. Something you should watch for are merchants that still put the CC number on receipts. They are only supposed to be putting the last 4 digits now.
BTW, I had one of my credit card #s stolen a couple years ago while I was on vacation. Since I have a special card for that, I know that it was either taken at a motel in Colorado or a mom and pop gas station in Utah. The morons actually used it to buy Bonzi Buddy. (and a $1500 camera on eBay)
Equipping your 1 million customers with some kind of secure random password generation smart card probably costs $40 each, both for the card and programming as well as associated infrastructure and overhead costs.
I've seen USB devices that were considerably less expensive than that. At least the hardware was... (I was looking for some computer access technology)
You could use devices with a static identification, combine it with an account number and pin/password, and provide access that would be as secure at your home PC as it would be at an ATM. It would probably be cost effective to add single use password/credit card numbers capabilities too. One of the big drawbacks with American Express' Blue card was the requirement of a special chip reader. (and the requirement for merchants to have affiliate accounts rather than creating single use CC #s) For a million customers, you could probably get the devices in the $1-5 range pretty easily.
You don't actually believe your bank checks the signatures on your checks, do you? Not likely, unless it's a really small bank or there is a dispute.
Some banks are processing checks like EFTs (Electronic Funds Transfers) already. For instance, if you have a Citibank credit card, you write them a check to pay your bill, they process it like a debit card and then destroy the check. Hopefully they image it first, but that wasn't mentioned as part of the process when they explained it to my mom.
If you have Direct Deposit on your paycheck, the same routing information that is used to deposit your paycheck can be used to remove it.
As far as I can tell, the only protection customers of US banks have is the goodwill of banking institutions. Rather than authenticate, they seem to use a 'trusted peer' methodology.
So, compared to their EFT systems, it looks like their web sites are pretty secure.
BTW, I have one account that requires a 6+ digit password, with the first character being alpha, one character has to be numeric, and one punctuation. Sure, it's safe from a dictionary check, but I can never remember the damn thing.
And my college assigns passwords with random characters for e-mail and enrollment. I'm sure it's so much more secure for me to write it down and carry it in my wallet than it would be to let me choose it myself.
It cost me $850 for the control panel alone but it was WELL worth the price.
That's a nice looking panel, but $850? That's outrageous. I'm working on two cocktail cabinets right now. I have spent $40 each for my IPACs and $30 each for the controls. (2 joysticks, 12 player buttons, Player 1 and player 2 buttons, and a couple extra. Excluding the computer equipment, I won't be spending $850 for both cabinets.
Controls
IPAC
Of course the interpreters for those languages are usually written in C or C++, so we haven't quite escaped the problem yet.
Well, since Delphi doesn't have an interpreter, I would say as least some of us have escaped the problem. Haven't had a buffer overflow in an application in 10 years. Pointers and range checking hasn't been a problem either.
Another Gateway favorite was "it's a problem with your modem (or video) drivers, download news ones from our website and install them. If you still have a problem, then call us back." They would use that for any problem, not just modem/video issues. The problem they were trained to fix was how to end a call in less than a minute.
(I've never been a Gateway tech, but several of my friends were. As a result, I've never purchased a Gateway computer)
There's always a risk that any application that's handling data, especially unclean internet data, can be the victim of a buffer overflow.
Insightful? Not even close. Buffer overflows aren't a given. They aren't a fact of life. Quite simply, all you have to do is simple bounds checking. If you allocate a 4k buffer, don't try to copy 6k to it. Buffer overflows are a 'feature' of C/C++. There are plenty of other languages that don't have that problem.
Unchecked buffers are the result of poor program design. No programmer, or company, should say that they are security conscious if they haven't done a code review and fixed them.
Copyright infringement is a civil matter, not a crime.
At least it wasn't until the creation of the DMCA. Such a nicely orchestrated move to shift enforcement burdens from the copyright holder to the public taxpayer.
But when a pirate does it, gets sued, and settles, somehow it's evil that the RIAA sued in the first place and the pirate is the good guy martyr.
First of all, I could care less whether people get their free music. I think it's a bad business decision to not make it available in some form, because so many of us won't buy a new CD until we have heard all of it since we are tired of getting burned buying CDs with a solitary good song.
But downloading music is a bad comparison. People aren't downloading MP3s so that they can incorporate it on their own CD and sell it. If they are, they are distributing and should be sued by the copyright holder. SCO was incorporating someone else's IP in their product without prior authorization.
Nicholas Petreley uses the tired term "paradigm shift" in his article!
It could be worse, he could have used "wake-up call".
One graph has ext2 first, the next has xfs. ext3 is light blue in one, purple in the next. I quickly found that after having to recheck the key for each graph, I really didn't care anymore.
> How about geothermal? Iceland has had a lot of success with that.
Well, maybe because it is a small island directly placed above a contintental rift. I don't know for how many regions that would work.
Technically, ground sourced heat pumps are geothermal too. Just at a lesser degree. That is available anywhere.
It's not like we need a silver bullet.
My original point was that wind and tidal power aren't silver bullets. The original poster was pointing to wind and tidal power as the solution to those evil dams. But they have drawbacks too. It's going to have to be a balance of several technologies.
I doubt there is anywhere on earth that is entirely unlivable. Maybe by us, but we aren't the only organisms here. Keep in mind that drastic changes in other habitats tend to cause drastic changes in ours.
Or, you can look at it another way. It's not possible for man to destroy the ecology of the earth. It will still be here (in some other form) long after it can no longer support us.
local energy untility has to BUY BACK power
At wholesale energy prices. So you pay them full retail price for power you use, and they pay you a lot less for the power you sell back to them.
Don't oversize your system. This isn't a money making adventure. Unless you have time and date metering, they are just checking your meter once a month. If it moves forward, you pay. Backwards, they do. The wholesale/retail problem would only come into effect when you're trying to carry over from month to month.
Most off-grid people use very little eletricity, even to the point of unplugging wall warts when not in use, because they draw a few watts even when you aren't using what they are connected to.
Also, unplug all of your vampire appliances. These are appliances that use power even when turned off. A lot of times, off does not mean off. Every manufacturer wants to install clocks and LEDs. For instance, my Apex DVD has a power switch and you can turn it off with the remote. The remote just suspends parts of the system, the LED stays lit, the circuitry is live to listen for commands from the remote. A lot of stereos are like that too. My AudioTron has that annoying problem too. The power switch on the front causes it to hibernate, the true power switch is inaccessibly on the back.
Hydro power is now on the way out as a major power source. Many dams have been removed in Western countries because they lead to salinization of cropland, destruction of hatcheries, and they just cost so bloody much.
Salinization isn't caused by dams, it's caused by irrigation. Cost isn't the issue, it costs more to tear them down. While they are there, their operation profits cover expenses. The problem is the destruction of natural habitats. I'll admit that a thorough environmental analysis needs to be done before building a dam, but I have to wonder how many habitats are restored by a dam's destruction. You don't just drain the reservoir, blow up the damn and everything magically returns.
Besides, wind turbines will cool the atmosphere by some tiny amount to offset global warming.
Oh, I'm sorry, that wasn't meant to be a joke??? Wind turbines kill birds. A significant number, hard to say. I haven't seen any serious studies. Tidal power requires changing the shoreline. What effect is that going to have? Ever wonder why Florida is building artificial reefs? It's because development in one area is causing beach scouring in another.
How about geothermal? Iceland has had a lot of success with that.
The sell high/buy low only works if your meter keeps track of the time of day you are producing and consuming. And you have to be on a special rate program, something power companies don't like to use on residential service. But the grid makes a nice, maintenance free battery. Even when you buy and sell at the same price.
If you want to be completely solar, you have to attack the problem on several fronts. You need to find any way that you can to reduce power consumption. Then there are storage and backup power problems to deal with.
Since you didn't post capacity, it's hard to say what the $17k covers. I don't know if the programs are currently active, but in the past California had a tax rebate program that could offset about 1/2 of the cost of installation. I would be surprised if they missed that in the quote, but you might want to check. It would drastically cut payback time. Also remember that part of your payback comes in non-monetary benefits.
The type of system you were looking at is a good one, but probably needs a few adjustments. Being connected to the grid has a lot of advantages. The grid serves as your batteries. If your usage spikes (air conditioning?), the grid will make up the difference. And the grid supplies your power at night and when you can't produce.
When the grid goes down, you don't necessarily have to shut down too. When the grid goes down, you DO have to disconnect your PV units from the grid, regardless of whether they are producing or not. Neither you nor PG&E wants you powering their lines and electrocuting their linemen.
And if you look at it that way, compare the cost of lost revenue from over production to the cost of batteries. If losing the money bothers you that much, slightly undersize your system so you don't produce an excess amount. PG&E will happily cover the difference.
The real question is, "How much could the Russians do it for"?
No, no, no. You outsource it to China and India and then buy your parts from Walmart. No wonder it takes $10m to build this thing.
Actually, I'd say "stick to unstable" and immediately replace whatever is in Knoppix's apt sources with the Debian unstable sources, then update and dist-upgrade your system
Unstable isn't very good for new users. They'll get too many broken packages and no idea how to fix them. And god forbid they have a windows mentality when they learn the power of apt-get. A dist-upgrade will almost surely kill their system with the default 'testing' settings.
In regard to my problems with KDE, it came from not paying attention to what I was doing with apt. I had MythTV stable with apt-get testing. Both should have been set to unstable. When it had dependancy problems, I was stupid enough to try -forcing a package that backleveled KDE. Even installing the unstable KDE didn't clean up the mess. But if you install Knoppix, set apt to unstable, and then apt-get install KDE it takes you to 3.2.2 just fine.
Now that 3.4 is out, I need to burn it and see if it can now detect the sound card in my Dell. And they've added some support for wireless too.... Fortunately I have plenty of test boxes to work with.
Well, I can see that telling people who "want cell phones to be just cell phones" to buy secondhand is a little extreme.
Extreme? I was thinking condescending.
It's like telling everyone who wants a car to buy a limo. And if you don't like all the features of the latest Towncar, you should just buy a second hand one.
I had to replace my phone in December, and I had a rough time finding a decent phone that didn't include a camera. (I have a real digital camera, thank you very much) I ended up with a Samsung E105, and even it has a ton of features I won't use. The only non-phone feature I need is inbound text messaging, which is a much more logical extension than games or PIMs.
Give me a well made, no frills, unlocked GSM phone and that's all I need or want. If they want to sell me additional services, give me a reasonably priced PC Card w/ data service.
Overall, I'd say the risk of a patch breaking something on your specific machine (as opposed to a few random thousand of the 100s of millions out there) is much lower than the risk of a virus hitting you while you're "testing" the patches.
That hasn't been our experience here. Less than a year ago we specifically put together a plan for staged rollouts of patches. It started with a get tough plan to make sure all servers were up to date, followed by several applications on all of our middletiers working erratically. It took a week for the programmers of the effected apps to get the problem fixed and working reliably. Things were starting to get a little ugly and users were not happy. Result, we have three stages of rollouts; test systems, first half production, last half production. None of which install automatically.
I wasn't effected on that case, but I have had MS 'fixes' break critical systems. A while back a 'fix' of the generic text printer driver caused it to eat the first character of each line. Barcode printers stopped working. And no barcodes, no shipping. Spent a day finding it, added a sacrificial space to each line, system is back online. A year later, MS fixes the 'fix' and the driver is working correctly again, but now the printers are choking on the extra space. Pull our fix for their 'fix', and our systems are back in a couple hours. But only because I remembered the previous problem and work around.
As to timeframe; it takes time to test complicated systems. Add to that the effects of the ecomony and companies are expecting more from fewer developers. So we have to balance our time between business requirements and testing MS patches. Being late installing a patch doesn't show up on my annual review, missing development deadlines does.
As far as getting hit; we don't get hit very often, today is the first case of an infected server that I can remember since code red hit our website. We have up-to-date scanning on our systems, SUS for desktop patches, email scanning, and properly configured firewalls.
Today we are fighting with a variant of a worm that isn't being detected by our scanners. But also doesn't appear to be using a vuln fixed by any patch. But that's a problem for Operations; developers are coding today, not chasing MS bugs.
Is there a Knoppix->Gnome faq out there?
You might try Gnoppix.