**Beats head against wall** Don't they realize that this defeats the entire point of the paper trail?! It needs to print as the vote is cast, so that the voter can verify it. By the time they print it out afterwards, it can already be changed!
Of course they do. There's no reason for them to fight so rabidly for abusable voting systems otherwise.
Keyboard styles come in waves. Every vendor out there inexplicably clones their competitors and drops their old models. Perhaps this is because of the "car mentality" -- if you switch styles frequently enough, a three-year-old keyboard just looks "old".
I remember the flat, traditional keyboards.
Then ergo keyboards were the rage, and I saw a ton of split-keyboard, tilted-away-from-the-center keyboards.
Apparently that went out of style, and then most keyboards were split-keyboard, not-tilted-away-from-the-center keyboards.
Now, it seems that wireless is all the rage.
It's a real irritation, because heavy computer users often become comfortable with a keyboard that they like, and then can't ever get ahold of that keyboard again.
I'm really considering getting a Powerbook as my next luggable machine, but the fact that it has a touchpad is the single thing holding me back...
Unfortunately, IBM has a patent on the nipple interface, which is why you don't see a lot of other manufacturers with nipples. It's not as if the addition of the thing would add much cost to the device, but they'd have to pay licensing fees.
The vagina-like design is probably intentional. Industrial designers are a bunch of sneaky bastards that do nothing but figure out how to play off of the subconscious.
You can replace the batteries with longer-life ones.
That being said, I've had the (dis)pleasure of using an MX700, and frankly, these things are massively overblown. It's convinced me that cordless mice suck -- the weight and battery life are a pain in the ass. Heck, Logitech even makes a wireless *trackball*. What are they thinking?
At $25, it seems like a reasonably priced product. I disagree. There are many more interesting mice that can be had for equal or less money. Mice are cheap these days.
And a *hemisphere*? Did this designer learn nothing from the awful Apple puck mouse? Mathematically simple does not mean usable.
Don't get me wrong -- I think that Microsoft's input devices are the best things in their product line, and that their traditional many-button optical mouse line isn't bad. But, frankly, this mouse is just not remotely "news for nerds".
Whether it's fair or not, it's up to each of us to find a way to be valuable to a company.
[shrug] Not necessarily. The point of an econonmy is to provide a framework to supply goods, services, and other resources. While our current model does follow an approximation of what you're talking about, many places are much more socialistic.
I have to agree with the other respondant. I understand how irritating it is for you -- after all, there are no code style constraints on Windows drivers -- but the only software that makes it into the mainstream Linux kernel (or bundled with the Windows distribution) is the very good stuff, and unlike on Windows, on Linux the code, not just the binary, is a factor in the quality.
I *do* realize that this could cause major problems if there are internal reasons for using a particular style or -- God forbid -- if you are trying to get a driver into both Linux and BSD or something similar. I'm not sure that there is a good solution, for trying to get open source drivers included in a kernel. That's an interesting point, though.
Same here, but with a 9200. The card before that was a Matrox G450, and the card before that a G200. In each case, it was because the card was the best-supported-by-open-source-drivers-on-Linux product of its time.
Of course, I am only one person.
It would be to everyone's benefit for someone to maintain a website ranking products based on open-source support for Linux, so that I can just choose the highest-ranked product that fits my needs.
Of course you can say all things about manufacturers 'not being able to open' (nvidia binary drivers cannot be opened, because of rules, etc.), but that is just not true. Nvidia has no advantage of opening their secrets (yes, their competitors will take advantage of this) and therefore they are not doing it (they lose perhaps a handfull of users). A perfect business decision in the light of most linux users using a binary driver. However if ATI IS providing (this is an example) the source, Nvidia has to look different to the situation, especially if, say 20% of the users uses ATI because of this and if ATI is also 'giving up' their secrets.
Video cards are a special case. They are high-innovation, both ATI and Nvidia retain a large number of very highly-paid and well-known computer scientists on staff, and they hide a lot of their magic in the driver. It is a hyper-competitive arena where a 10% performance difference can make or break a product cycle. There are almost no other devices with drivers that operate like this -- certainly not webcams.
Yeah -- he might be able to do something legally, but companies can do a lot of nasty things legally, like using bogus patents.
The point is that he's trying not to burn Phillips for letting him have the data. It would be *infinitely* better if Phillips would just let the code be GPLed, but they might not want that to happen, if they think that it's a major competitive advantage (and that their competitors are willing to go GPL or are just interested in taking ideas, not code, from their GPL code).
No. It's fine to write code and put it in both GPLed and non-GPL-compatible software if you hold all copyrights. I have plenty of utility code that I've written that has been in both.
The problem is that you can't write code and then run it in a manner where it is simultaneously linked with both GPLed and GPL-incompatible software, which is what the webcam module guy is doing.
Libusb is cool, and lets you do lots of stuff that would otherwise be in kernelspace, but was changing not long ago -- I was going to write an X10 libusb controlling program a while back, but the libusb software shipped by Red Hat at the time was missing necessary functionality.
But if he decides not to license his work under the GPL anymore, what right do you or I have to use his code in a new project? We have no license from him, so we'd be violating his copyright.
It doesn't work like that. He can choose not to license his *future* work on the codebase under the GPL *if* he is the sole copyright holder (or all copyright holders agree), but even if you are the sole copyright holder, you can never "remove the rights you granted under the GPL to existing code". It might be that v2.3.1 or whatever is the final GPL version, but that version will stay GPL.
He can request that the code be removed from the mainline kernel -- a kernel guy doing so would do so as a courtesy to him, not because of any legal requirement.
Anyone could pick up the last GPLed version and start maintaining it.
The whole thing kind of sucks. The kernel people are understandably cranky because they're trying to work in a world where NDAs aren't that feasible, the GPL is important, and binary modules play hell with their ability to troubleshoot, platform compatibility and the like. The guy is understandably cranky because the kernel people are basically throwing out work that he went to the trouble of coding up for free under NDA (which seems to work pretty well). End users are understandably cranky because they paid $100 for their webcam that suddenly (at best) drops in resolution or stops working. Philips *could* become justifiably cranky because of the way they've given out some technical data and gotten kinda screwed, with Linux support for their products being dropped with no warning.
I have a bit of code I'm thinking of open-sourcing -- I've kept it closed simply because it's essentially a web-spider, and I incorporated pauses in it, to spare other people's servers, that I don't want selfish users to remove.
As a completely unrelated aside, I'm not sure that this is a good idea. Web servers can always limit rates to heavy users. However, a user can always get an "abusable" tool from somewhere. And honestly, wget in recursive mode doesn't cause problems (if wget allowed parallel operation, *then* there might be fireworks). I've never modified open-source software to remove delays, but I *have* binary-hacked a closed source piece of software to remove an imposed delay on web site downloading.
Can anyone recommend new feasible choices for low-cost video cameras for Linux use? I've been wanting to pick one up for a bit, and the camera I was looking at was an affected model.:-(
I'd really like to get a camera that has an all-GPL driver, if possible.
Post to LKML. This is probably the "official way", and even if not, they'll definitely tell you the right thing to do.
Also helpful is #kernelnewbies on Freenode, where a bunch of kernel hackers hang out, but it sounds like this doesn't require low-latency interaction.
I would like to thank you for stepping up. All of us Linux users benefit (I, have been batting around the idea of playing around with some image recognition and processing ideas on Linux, and this would allow me to continue to do so).
I'd like to point out that there is userspace software (GPLed) to do superresolution processing that may be able to provide PWXC-like functionality, but still keep the kernel hackers happy.
I would like to see the PWC/PWXC guy reconsider abandoning ownership (and all the nice work he did) -- he probably did what he did in a fit of frusteration with the other kernel hackers, but if he becomes willing to allow PWC to just pass its data through some userspace software, this might not hurt end users (and if the userspace software is better, might even be a win).
I work for a Fortune 100 corporation that outsourced its internal help desk to India. The main effect seems to have been that everyone now pesters me for technical support instead of calling the supposed IT department.
Been there, done that (well, Fortune 500, not 100, but still big and stuffy enough to be into outsourcing IT). The fact that all the local people wind up doing their own tech support is a cost that will probably never wend its way into corporate overseas cost analysis.
As another great anecdote: I remember finally getting in touch with the (clearly Indian) people at the network operations center because I needed to know whether they could add a CNAME pointing at a name in a domain outside of company control, or whether that would violate IT policy. I don't think I've ever run into a DNS admin before that draws a blank on the term "CNAME" but responds to "alias" and has no idea what an "IP block" is.
Experts are not likely going to be on the helpdesk. I know helpdesk workers often consider themselves experts, but come on.
The funny thing is that people expect to have an serious domain expert on the other end of the line at their beck and call (and no $2/minute charges or anything like that) when they call up for tech support for a service that they pay a lousy $20/month on. Or maybe on a $500 computer with essentially no profit margin. You can't do it. It's just non-economical. The only sane way to deal with this is to try to make products that are as idiot-proof as possible so that people never call tech support.
Frankly, I'd like to see a "zero tech support" sticker, so that I could buy only ZTS products, so that I don't have to subsidize people that need tech support. It's just more of my money down the tube. Have a good web site with the same data that you'd feed your tech support people, and I'll be happy.
Support on software is SolidWorks, but then I pay nearly $2000/yr for upgrades and phone support...out of Utah, in my case. So I cancelled my ELink service & asked to release the Elink hold on my Ph# and that resulted in "That will take 7-10 business days." which saved me $20/month.
And there you have it.
Everybody wants tech support, but nobody wants to pay for it.
Let's say that, instead of a minimum-wage script-reading monkey, you get a really competent network admin to answer calls. Also keep in mind that phone support is not a very fun job to do, and that said network admin may demand more money to do support, and that phone support, being a customer-facing job, has requirements on accent and behavior that being a regular network admin may not.
Let's suppose that this network admin makes, say, $50K/year, which, for a competent troubleshooter, is certainly not very generous. There are ~250 working days in the year, ignoring holidays and vacation and sick leave and the like. That means that our network admin is making $200/day. Given an eight-hour day, that's $25/hour.
Now, let's assume that this guy works *constantly* every minute of his day, doesn't have a lunch break or anything, never fails to close a ticket in a single call, and has to escalate anything or the like. Let's assume that he maintains a call average of only 15 minutes.
That means that this guy, not counting management overhead, the guy's benefits, the infrastructure (A/C, the phone line, the costs of the building he's sitting in, etc) run $6 per call. I don't work at an ISP (I suspect some Slashdotters can provide better numbers than I can), but I'm guessing that on $20/mo service, that call has pretty much blown all profits for the month, even with all our generous assumptions.
Now, sure, Earthlink could provide really great tech support. But they'd have to charge $30/mo instead of $20/mo to pay for the tech support. And they'll lose out to their competitors.
I have a friend who just had his mother call him twice -- first that she couldn't get online, second that the wireless hub was malfunctioning, and third that she had called someone else and he had figured out that she had unplugged the wireless hub. She's been calling just about every other day with technical problems.
So people want this good, free tech support, but they simply do not want to pay for it. Expert knowledge simply is not cheap -- most people that really understand a system can get a more comfortable job than tech support.
Really, the only tech support I've ever thought even remotely decent is the MSDN Universal support -- which costs even more than your Solidworks package does.
Retail figured this out a long time ago. The guy at the appliance store doesn't know how to fix any problems you have -- he can take any returns you have, but that's about it. Problem? Just return the thing.
I'd like to see more services have "no support" options, or at least "charge per call" support, so that I don't have to subsidize others. I'd strongly prefer web pages with common problems than to pay to have some guy waiting for me to maybe call.
Full of factual errors (higher mercury levels, over-extended military, intelligence reform, etc) and it gets modded insightful.
I'm unaware of the mercury issues. I can't understand how you can say that our military is not over-extended -- remember the controversy about length of tours of duty and moving National Guard to Iraq and the like? And, frankly, I'm not very happy with the existing intelligence situation.
Slashdot Mirror
**Beats head against wall** Don't they realize that this defeats the entire point of the paper trail?! It needs to print as the vote is cast, so that the voter can verify it. By the time they print it out afterwards, it can already be changed!
Of course they do. There's no reason for them to fight so rabidly for abusable voting systems otherwise.
Keyboard styles come in waves. Every vendor out there inexplicably clones their competitors and drops their old models. Perhaps this is because of the "car mentality" -- if you switch styles frequently enough, a three-year-old keyboard just looks "old".
I remember the flat, traditional keyboards.
Then ergo keyboards were the rage, and I saw a ton of split-keyboard, tilted-away-from-the-center keyboards.
Apparently that went out of style, and then most keyboards were split-keyboard, not-tilted-away-from-the-center keyboards.
Now, it seems that wireless is all the rage.
It's a real irritation, because heavy computer users often become comfortable with a keyboard that they like, and then can't ever get ahold of that keyboard again.
I'm really considering getting a Powerbook as my next luggable machine, but the fact that it has a touchpad is the single thing holding me back...
Unfortunately, IBM has a patent on the nipple interface, which is why you don't see a lot of other manufacturers with nipples. It's not as if the addition of the thing would add much cost to the device, but they'd have to pay licensing fees.
The vagina-like design is probably intentional. Industrial designers are a bunch of sneaky bastards that do nothing but figure out how to play off of the subconscious.
You can replace the batteries with longer-life ones.
That being said, I've had the (dis)pleasure of using an MX700, and frankly, these things are massively overblown. It's convinced me that cordless mice suck -- the weight and battery life are a pain in the ass. Heck, Logitech even makes a wireless *trackball*. What are they thinking?
At $25, it seems like a reasonably priced product.
I disagree. There are many more interesting mice that can be had for equal or less money. Mice are cheap these days.
And a *hemisphere*? Did this designer learn nothing from the awful Apple puck mouse? Mathematically simple does not mean usable.
Don't get me wrong -- I think that Microsoft's input devices are the best things in their product line, and that their traditional many-button optical mouse line isn't bad. But, frankly, this mouse is just not remotely "news for nerds".
Whether it's fair or not, it's up to each of us to find a way to be valuable to a company.
[shrug] Not necessarily. The point of an econonmy is to provide a framework to supply goods, services, and other resources. While our current model does follow an approximation of what you're talking about, many places are much more socialistic.
* Indians, Chinese, Japanese, and anyone else who may take a US job are the spawn of Satan
.com, not during Outsourcing.
I've yet to see Slashdotters hammering on Chinese (and, to an even greater extent, Japanese).
Americans have an inalienable right to overpaid jobs
No. People thought that during
Anything under 100% employment represents a failure of the US economy
The removal of long-term unemployed from the unemployment statistics *does* skew the numbers. I don't think anyone is demanding complete employment.
Bush is responsible for every job that leaves the US
No. He's just helping promote outsourcing.
John Kerry will fly in and save the world from the evil "Benedict Arnold CEOs"
No. He's just more likely to engage in protectionism than Bush.
I have to agree with the other respondant. I understand how irritating it is for you -- after all, there are no code style constraints on Windows drivers -- but the only software that makes it into the mainstream Linux kernel (or bundled with the Windows distribution) is the very good stuff, and unlike on Windows, on Linux the code, not just the binary, is a factor in the quality.
I *do* realize that this could cause major problems if there are internal reasons for using a particular style or -- God forbid -- if you are trying to get a driver into both Linux and BSD or something similar. I'm not sure that there is a good solution, for trying to get open source drivers included in a kernel. That's an interesting point, though.
Same here, but with a 9200. The card before that was a Matrox G450, and the card before that a G200. In each case, it was because the card was the best-supported-by-open-source-drivers-on-Linux product of its time.
Of course, I am only one person.
It would be to everyone's benefit for someone to maintain a website ranking products based on open-source support for Linux, so that I can just choose the highest-ranked product that fits my needs.
Of course you can say all things about manufacturers 'not being able to open' (nvidia binary drivers cannot be opened, because of rules, etc.), but that is just not true. Nvidia has no advantage of opening their secrets (yes, their competitors will take advantage of this) and therefore they are not doing it (they lose perhaps a handfull of users). A perfect business decision in the light of most linux users using a binary driver. However if ATI IS providing (this is an example) the source, Nvidia has to look different to the situation, especially if, say 20% of the users uses ATI because of this and if ATI is also 'giving up' their secrets.
Video cards are a special case. They are high-innovation, both ATI and Nvidia retain a large number of very highly-paid and well-known computer scientists on staff, and they hide a lot of their magic in the driver. It is a hyper-competitive arena where a 10% performance difference can make or break a product cycle. There are almost no other devices with drivers that operate like this -- certainly not webcams.
Yeah -- he might be able to do something legally, but companies can do a lot of nasty things legally, like using bogus patents.
The point is that he's trying not to burn Phillips for letting him have the data. It would be *infinitely* better if Phillips would just let the code be GPLed, but they might not want that to happen, if they think that it's a major competitive advantage (and that their competitors are willing to go GPL or are just interested in taking ideas, not code, from their GPL code).
No. It's fine to write code and put it in both GPLed and non-GPL-compatible software if you hold all copyrights. I have plenty of utility code that I've written that has been in both.
The problem is that you can't write code and then run it in a manner where it is simultaneously linked with both GPLed and GPL-incompatible software, which is what the webcam module guy is doing.
Libusb is cool, and lets you do lots of stuff that would otherwise be in kernelspace, but was changing not long ago -- I was going to write an X10 libusb controlling program a while back, but the libusb software shipped by Red Hat at the time was missing necessary functionality.
But if he decides not to license his work under the GPL anymore, what right do you or I have to use his code in a new project? We have no license from him, so we'd be violating his copyright.
It doesn't work like that. He can choose not to license his *future* work on the codebase under the GPL *if* he is the sole copyright holder (or all copyright holders agree), but even if you are the sole copyright holder, you can never "remove the rights you granted under the GPL to existing code". It might be that v2.3.1 or whatever is the final GPL version, but that version will stay GPL.
He can request that the code be removed from the mainline kernel -- a kernel guy doing so would do so as a courtesy to him, not because of any legal requirement.
Anyone could pick up the last GPLed version and start maintaining it.
The whole thing kind of sucks. The kernel people are understandably cranky because they're trying to work in a world where NDAs aren't that feasible, the GPL is important, and binary modules play hell with their ability to troubleshoot, platform compatibility and the like. The guy is understandably cranky because the kernel people are basically throwing out work that he went to the trouble of coding up for free under NDA (which seems to work pretty well). End users are understandably cranky because they paid $100 for their webcam that suddenly (at best) drops in resolution or stops working. Philips *could* become justifiably cranky because of the way they've given out some technical data and gotten kinda screwed, with Linux support for their products being dropped with no warning.
I have a bit of code I'm thinking of open-sourcing -- I've kept it closed simply because it's essentially a web-spider, and I incorporated pauses in it, to spare other people's servers, that I don't want selfish users to remove.
As a completely unrelated aside, I'm not sure that this is a good idea. Web servers can always limit rates to heavy users. However, a user can always get an "abusable" tool from somewhere. And honestly, wget in recursive mode doesn't cause problems (if wget allowed parallel operation, *then* there might be fireworks). I've never modified open-source software to remove delays, but I *have* binary-hacked a closed source piece of software to remove an imposed delay on web site downloading.
Can anyone recommend new feasible choices for low-cost video cameras for Linux use? I've been wanting to pick one up for a bit, and the camera I was looking at was an affected model. :-(
I'd really like to get a camera that has an all-GPL driver, if possible.
It could also be due to (unlikely as it sounds) superresolution processing on-camera or on-host that requires multiple frames to be captured.
Post to LKML. This is probably the "official way", and even if not, they'll definitely tell you the right thing to do.
Also helpful is #kernelnewbies on Freenode, where a bunch of kernel hackers hang out, but it sounds like this doesn't require low-latency interaction.
I would like to thank you for stepping up. All of us Linux users benefit (I, have been batting around the idea of playing around with some image recognition and processing ideas on Linux, and this would allow me to continue to do so).
I'd like to point out that there is userspace software (GPLed) to do superresolution processing that may be able to provide PWXC-like functionality, but still keep the kernel hackers happy.
I would like to see the PWC/PWXC guy reconsider abandoning ownership (and all the nice work he did) -- he probably did what he did in a fit of frusteration with the other kernel hackers, but if he becomes willing to allow PWC to just pass its data through some userspace software, this might not hurt end users (and if the userspace software is better, might even be a win).
And you're capable of doing better?
And willing?
No?
Then I don't see how you have any grounds to complain about him being "lazy". He's an unpaid volunteer giving his work to you.
I work for a Fortune 100 corporation that outsourced its internal help desk to India. The main effect seems to have been that everyone now pesters me for technical support instead of calling the supposed IT department.
Been there, done that (well, Fortune 500, not 100, but still big and stuffy enough to be into outsourcing IT). The fact that all the local people wind up doing their own tech support is a cost that will probably never wend its way into corporate overseas cost analysis.
As another great anecdote: I remember finally getting in touch with the (clearly Indian) people at the network operations center because I needed to know whether they could add a CNAME pointing at a name in a domain outside of company control, or whether that would violate IT policy. I don't think I've ever run into a DNS admin before that draws a blank on the term "CNAME" but responds to "alias" and has no idea what an "IP block" is.
Experts are not likely going to be on the helpdesk. I know helpdesk workers often consider themselves experts, but come on.
The funny thing is that people expect to have an serious domain expert on the other end of the line at their beck and call (and no $2/minute charges or anything like that) when they call up for tech support for a service that they pay a lousy $20/month on. Or maybe on a $500 computer with essentially no profit margin. You can't do it. It's just non-economical. The only sane way to deal with this is to try to make products that are as idiot-proof as possible so that people never call tech support.
Frankly, I'd like to see a "zero tech support" sticker, so that I could buy only ZTS products, so that I don't have to subsidize people that need tech support. It's just more of my money down the tube. Have a good web site with the same data that you'd feed your tech support people, and I'll be happy.
Support on software is SolidWorks, but then I pay nearly $2000/yr for upgrades and phone support...out of Utah, in my case. So I cancelled my ELink service & asked to release the Elink hold on my Ph# and that resulted in "That will take 7-10 business days." which saved me $20/month.
And there you have it.
Everybody wants tech support, but nobody wants to pay for it.
Let's say that, instead of a minimum-wage script-reading monkey, you get a really competent network admin to answer calls. Also keep in mind that phone support is not a very fun job to do, and that said network admin may demand more money to do support, and that phone support, being a customer-facing job, has requirements on accent and behavior that being a regular network admin may not.
Let's suppose that this network admin makes, say, $50K/year, which, for a competent troubleshooter, is certainly not very generous. There are ~250 working days in the year, ignoring holidays and vacation and sick leave and the like. That means that our network admin is making $200/day. Given an eight-hour day, that's $25/hour.
Now, let's assume that this guy works *constantly* every minute of his day, doesn't have a lunch break or anything, never fails to close a ticket in a single call, and has to escalate anything or the like. Let's assume that he maintains a call average of only 15 minutes.
That means that this guy, not counting management overhead, the guy's benefits, the infrastructure (A/C, the phone line, the costs of the building he's sitting in, etc) run $6 per call. I don't work at an ISP (I suspect some Slashdotters can provide better numbers than I can), but I'm guessing that on $20/mo service, that call has pretty much blown all profits for the month, even with all our generous assumptions.
Now, sure, Earthlink could provide really great tech support. But they'd have to charge $30/mo instead of $20/mo to pay for the tech support. And they'll lose out to their competitors.
I have a friend who just had his mother call him twice -- first that she couldn't get online, second that the wireless hub was malfunctioning, and third that she had called someone else and he had figured out that she had unplugged the wireless hub. She's been calling just about every other day with technical problems.
So people want this good, free tech support, but they simply do not want to pay for it. Expert knowledge simply is not cheap -- most people that really understand a system can get a more comfortable job than tech support.
Really, the only tech support I've ever thought even remotely decent is the MSDN Universal support -- which costs even more than your Solidworks package does.
Retail figured this out a long time ago. The guy at the appliance store doesn't know how to fix any problems you have -- he can take any returns you have, but that's about it. Problem? Just return the thing.
I'd like to see more services have "no support" options, or at least "charge per call" support, so that I don't have to subsidize others. I'd strongly prefer web pages with common problems than to pay to have some guy waiting for me to maybe call.
Just like the hemp advocates are a front for the drug legalization crowd, fetal stem cell advocates are a front for pro-abortionists.
Isn't this just a slippery slope fallacy on your part?
Full of factual errors (higher mercury levels, over-extended military, intelligence reform, etc) and it gets modded insightful.
I'm unaware of the mercury issues. I can't understand how you can say that our military is not over-extended -- remember the controversy about length of tours of duty and moving National Guard to Iraq and the like? And, frankly, I'm not very happy with the existing intelligence situation.
It's not "indirect theft". It's copyright infringement.