Given that this is Slashdot, I guess I shouldn't be terribly surprised to discover that nobody has pointed out that Office 2k3 has an XML document format: http://www.microsoft.com/downloads/details.aspx?Fa milyId=FE118952-3547-420A-A412-00A2662442D9&displa ylang=en
The product activation gauntlet is a few button clicks. Unless you rebuild your computer with completely different parts 4 times a year, it is never a problem.
Having tried Bibble, I'm not very impressed with the product.
It is probably one of the better 3rd party raw converts I've seen, but I wasn't particularly happy with the output... nothing I can really put my finger on though. Just something about the color seems off a hair.
The demonstration shown isn't very useful as they don't tell you what the camera settings were when the image was shot or what settings were applied in NC4. It would appear that NC4 was configured to output a high contrast image (instead of a low contrast one).
The "auto white balance" is also not something I'd brag about, which turned the whites from red to blue (meaning they trimmed too much red data and amplified the blue channel to much, resulting in higher noise levels from the blue channel). Personally, I'd rather shoot a grey card in the various light conditions than depend on an auto processing setting.
Auto anything is generally not something you'd want to depend on during your post processing (another reason I'm not a particular fan of Bibble, as it's big features woudl appear to be auto gizmos), as it doesn't give consistant results across images and never gets things perfect.
Back to the point, it doesn't matter if one app in your workflow supports color management -- all of them have to; not only that, but you still have the problem of no apps (other than bibble) which understand 16bpc images.
All Bibble is good for is a raw converter. I've tried it before; and while it does deal with raw images well I wasn't impressed with anything else it could do (or its stability for that matter). Doesn't change the fact that anything else in the image processing chain will likely trash that data and won't work with 16bpc images.
An exposure compensation "control" means that you don't have to do the math to properly manipulate the curve, which would be very painful to do/manipulate by hand.
I'm guessing you've never used dcraw. It discards most of the information present in the raw file (so forget about white balance, exposure information, using the proper colorspace, etc). And for 16bpc NEF images the output is messed up.
You forgot a few of the biggiest: 16bit per channel image manipulation and decent raw support from modern DSLRs (dcraw decodes NEFs for 8bpc images ok, but totally honks up 16bpc images) would be the other big items I would highlight.
Linux seems to have image processing tools for the "simple" types of image processing you'd want to do. It doesn't have software for more advanced operations, such enhancing shadows while preserving the highlights (it isn't a simple curve manipulation) or noise reduction for example. Hell, I haven't seen any tool which attempts to alter image exposure in EV steps.
However, the lack of any decent color management would be enough to keep any professional away from using linux IMO.
- 16 bit per channel support (8bit precision is woefully inadequate if you do any amount of image processing, especially when you start mixing colorspaces around) - Passable noise reduction software - Raw image support for modern cameras - Decent color profiling capabilities
Those are the big ones. I'm sure if I spent more than 2 minutes looking into it, I'd find more.
Granted, one might fall in the category of "infringer" by downloading the source and compiling the infringing software for themselves, but that argument is tenuous at best.
Tenous perhaps, but I wouldn't be surprised if this was where the line was drawn. And as that line has never been tested in court, nobody can really say if that arguement would succeed or fail.
Actually, it isn't. An application button is defined as a button on a palm sized PC which is used to launch an application. No mouse has an application button, as 1) it is not a palm sized pc and 2) the purpose of a mouse button is not to launch applications (the purpose of the mouse button is to provide input to the UI, which in turn may launch an application in response to that interaction).
The mouse fails claim one, as it does not have application buttons. It fails claim 2, because the mouse does not have application buttons. It fails claim 3, because claim 3 is based on claim 2. It fails claim 4, because the mouse does not have an application button. It fails claim 5, because claim 5 is based on claim 4. And so on.
The patent you cited is completely unrelated to the use of a mouse.
It covers a modification to hardware 'quicklaunch' buttons commonly seen on a portable device, where the specific enhancement in this case is using the button to perform different yet related application functions based on how the user manipulates that button and what the state of the currently open application is.
EA games has 3,300 programmers. Hire another 1,650 at $60,000 a pop, and the wages cost you $100M a year. Adjust to ~$150M a year for benefits, and you're still taking up less than one third of EA's operating profits from last year.
The true cost of employment is far greater than $60,000 per person. You have to factor in benefits (which you attempted to do), taxes the employer pays on your behalf, equipment, office space, bills (heat, electricity, water...), support staff, overhead of beauracracy, etc; this generally averages about 2.4 times the employee's base salary. The cost to employ a single person @ $60,000 a year is close to $150,000. That's about $250m, 66% more than you estimated.
This is still something they could very well accomplish, but it isn't as simple as cutting the marketing budget by 30%.
Please cite the patent number that you are using as an example. It sounds like you're making the (common) mistake of reading the abstract/summary and making assumptions about what exactly the patent covers.
Most likely, the patent you are referring to covers a small, specific enhancement to the mouse. Kind of like how auto manufacturers patent small specific enhancements they make to engines.
Last time *I* checked, 283 was a definite number and "over 228" was an INdefinite range... spanning all the way up to infinity.
While 283 is "over 228", "over 228" doesn't mean "283".
Yet, 283 is in the set of "over 228", so the statement is a true one. And who's to say that 283 is THE definitive number? New patents are granted every day -- I can't imagine that the number of potential infringments would remain static over time.
Not necessarily. It depends on who holds the patents in the first place, and who is liable. If the holder is a large monolithic software company, and the end users of OSS software are liable, there is no MAD. The end users likely don't have any patents to throw back at the large monolithic software company.
No, "potentially violates" means that a case could be made for infringement, but that it has never been tested in court, so it could very well NOT be an infringement as well. In addition, an untested patent has a fairly good chance of being overturned or limited. Patents that have been upheld in a court are much more dangerous.
Like I said, the distinction between "potentially voilates" and "violates" is one having actually brought a case to court. You couldn't even talk about the difference without mentioning something occuring in a courtroom.
Where Open Source has an advantage over Closed Source is that a patent holder who hasn't sued in a timely manner has much less of an excuse
They don't have to sue in a timely manner. Otherwise "submarining" wouldn't be an issue. This doesn't help OSS at all.
"But Ravicher said Ballmer misinterpreted his study's findings. 'He misconstrues the point of the OSRM study, which found that Linux potentially, not definitely, infringes 283 untested patents, while not infringing a single court-validated patent.'"
If the difference between potentially violates and violates means "nobody has sued anyone yet about this stuff", then I don't think the semantics make much of a difference.
There is a bigger reason why the patent attorney went ape.
If, at a later point in time, you are sued for some sort of patent infringement, and the person bringing the suit can show that you looked at their patent, the infringment becomes "willful". Even if you didn't think their patent didn't apply, it doesn't matter. Willful infringmeent means that the damages awarded are trippled.
Unfortunately, the way you "access" the name doesn't effect trademark ownership. If they knew about it, and failed to go after it, and then a someone else demonstrates such, they lose their 'Microsoft' trademark. No ifs, ands, or buts about it.
Slashdot Mirror
Given that this is Slashdot, I guess I shouldn't be terribly surprised to discover that nobody has pointed out that Office 2k3 has an XML document format: http://www.microsoft.com/downloads/details.aspx?Fa milyId=FE118952-3547-420A-A412-00A2662442D9&displa ylang=en
The product activation gauntlet is a few button clicks. Unless you rebuild your computer with completely different parts 4 times a year, it is never a problem.
Having tried Bibble, I'm not very impressed with the product.
... nothing I can really put my finger on though. Just something about the color seems off a hair.
It is probably one of the better 3rd party raw converts I've seen, but I wasn't particularly happy with the output
The demonstration shown isn't very useful as they don't tell you what the camera settings were when the image was shot or what settings were applied in NC4. It would appear that NC4 was configured to output a high contrast image (instead of a low contrast one).
The "auto white balance" is also not something I'd brag about, which turned the whites from red to blue (meaning they trimmed too much red data and amplified the blue channel to much, resulting in higher noise levels from the blue channel). Personally, I'd rather shoot a grey card in the various light conditions than depend on an auto processing setting.
Auto anything is generally not something you'd want to depend on during your post processing (another reason I'm not a particular fan of Bibble, as it's big features woudl appear to be auto gizmos), as it doesn't give consistant results across images and never gets things perfect.
Back to the point, it doesn't matter if one app in your workflow supports color management -- all of them have to; not only that, but you still have the problem of no apps (other than bibble) which understand 16bpc images.
All Bibble is good for is a raw converter. I've tried it before; and while it does deal with raw images well I wasn't impressed with anything else it could do (or its stability for that matter). Doesn't change the fact that anything else in the image processing chain will likely trash that data and won't work with 16bpc images.
An exposure compensation "control" means that you don't have to do the math to properly manipulate the curve, which would be very painful to do/manipulate by hand.
I'm guessing you've never used dcraw. It discards most of the information present in the raw file (so forget about white balance, exposure information, using the proper colorspace, etc). And for 16bpc NEF images the output is messed up.
You forgot a few of the biggiest: 16bit per channel image manipulation and decent raw support from modern DSLRs (dcraw decodes NEFs for 8bpc images ok, but totally honks up 16bpc images) would be the other big items I would highlight.
Linux seems to have image processing tools for the "simple" types of image processing you'd want to do. It doesn't have software for more advanced operations, such enhancing shadows while preserving the highlights (it isn't a simple curve manipulation) or noise reduction for example. Hell, I haven't seen any tool which attempts to alter image exposure in EV steps.
However, the lack of any decent color management would be enough to keep any professional away from using linux IMO.
- 16 bit per channel support (8bit precision is woefully inadequate if you do any amount of image processing, especially when you start mixing colorspaces around)
- Passable noise reduction software
- Raw image support for modern cameras
- Decent color profiling capabilities
Those are the big ones. I'm sure if I spent more than 2 minutes looking into it, I'd find more.
If the software you're using on Windows/Mac doesn't have a batch mode, you're not using software any self respecting photographer would use.
There is no software for linux that has anything resembling a passable workflow for digital photography.
Granted, one might fall in the category of "infringer" by downloading the source and compiling the infringing software for themselves, but that argument is tenuous at best.
Tenous perhaps, but I wouldn't be surprised if this was where the line was drawn. And as that line has never been tested in court, nobody can really say if that arguement would succeed or fail.
Actually, it isn't. An application button is defined as a button on a palm sized PC which is used to launch an application. No mouse has an application button, as 1) it is not a palm sized pc and 2) the purpose of a mouse button is not to launch applications (the purpose of the mouse button is to provide input to the UI, which in turn may launch an application in response to that interaction).
The mouse fails claim one, as it does not have application buttons. It fails claim 2, because the mouse does not have application buttons. It fails claim 3, because claim 3 is based on claim 2. It fails claim 4, because the mouse does not have an application button. It fails claim 5, because claim 5 is based on claim 4. And so on.
The patent you cited is completely unrelated to the use of a mouse.
It covers a modification to hardware 'quicklaunch' buttons commonly seen on a portable device, where the specific enhancement in this case is using the button to perform different yet related application functions based on how the user manipulates that button and what the state of the currently open application is.
EA games has 3,300 programmers. Hire another 1,650 at $60,000 a pop, and the wages cost you $100M a year. Adjust to ~$150M a year for benefits, and you're still taking up less than one third of EA's operating profits from last year.
The true cost of employment is far greater than $60,000 per person. You have to factor in benefits (which you attempted to do), taxes the employer pays on your behalf, equipment, office space, bills (heat, electricity, water...), support staff, overhead of beauracracy, etc; this generally averages about 2.4 times the employee's base salary. The cost to employ a single person @ $60,000 a year is close to $150,000. That's about $250m, 66% more than you estimated.
This is still something they could very well accomplish, but it isn't as simple as cutting the marketing budget by 30%.
Please cite the patent number that you are using as an example. It sounds like you're making the (common) mistake of reading the abstract/summary and making assumptions about what exactly the patent covers.
Most likely, the patent you are referring to covers a small, specific enhancement to the mouse. Kind of like how auto manufacturers patent small specific enhancements they make to engines.
Last time *I* checked, 283 was a definite number and "over 228" was an INdefinite range ... spanning all the way up to infinity.
While 283 is "over 228", "over 228" doesn't mean "283".
Yet, 283 is in the set of "over 228", so the statement is a true one. And who's to say that 283 is THE definitive number? New patents are granted every day -- I can't imagine that the number of potential infringments would remain static over time.
Balmer wasn't quoting the study though, so why does it matter?
Not necessarily. It depends on who holds the patents in the first place, and who is liable. If the holder is a large monolithic software company, and the end users of OSS software are liable, there is no MAD. The end users likely don't have any patents to throw back at the large monolithic software company.
No, "potentially violates" means that a case could be made for infringement, but that it has never been tested in court, so it could very well NOT be an infringement as well. In addition, an untested patent has a fairly good chance of being overturned or limited. Patents that have been upheld in a court are much more dangerous.
Like I said, the distinction between "potentially voilates" and "violates" is one having actually brought a case to court. You couldn't even talk about the difference without mentioning something occuring in a courtroom.
Where Open Source has an advantage over Closed Source is that a patent holder who hasn't sued in a timely manner has much less of an excuse
They don't have to sue in a timely manner. Otherwise "submarining" wouldn't be an issue. This doesn't help OSS at all.
"But Ravicher said Ballmer misinterpreted his study's findings. 'He misconstrues the point of the OSRM study, which found that Linux potentially, not definitely, infringes 283 untested patents, while not infringing a single court-validated patent.'"
If the difference between potentially violates and violates means "nobody has sued anyone yet about this stuff", then I don't think the semantics make much of a difference.
"In fact, the study said Linux potentially violates 283 software patents, not
...
'over 228' as Ballmer said in his speech."
Last time I checked 283 was over 228
There is a bigger reason why the patent attorney went ape.
If, at a later point in time, you are sued for some sort of patent infringement, and the person bringing the suit can show that you looked at their patent, the infringment becomes "willful". Even if you didn't think their patent didn't apply, it doesn't matter. Willful infringmeent means that the damages awarded are trippled.
I think their inspiration had to be grash-ooms.
...because we all know there are far more Solaris machines than Windows boxes out there running SFU...
Unfortunately, the way you "access" the name doesn't effect trademark ownership. If they knew about it, and failed to go after it, and then a someone else demonstrates such, they lose their 'Microsoft' trademark. No ifs, ands, or buts about it.
Your problem is with Trademark law, not MS.
I think you're confusing Microsoft with Apple in this department...