Maybe. They shouldn't but I wouldn't put it past them.
The answer is "yes", as it turns out. The appropriate sections of the license agreement are already copied a buncha times on this thread, so I won't repeat them, but yes, Adobe can, under their license, use the images, sell the use of those images to other people, and pretty much do with them anything they want short of preventing you from doing the same.
Does Adobe use retain share or gain any legal use of my uploaded images?
Yes. As I understand it from discussing the matter with a lawyer earlier today, the license agreement gives them the right to take images you load to public areas of their web site, and gives them free reign to do with them as they will. They *don't* have your copyright, you can do with them as you will as well, but they can, for example, put those images in a TV ad for their product, they can sell rights to use your image for money, they can sell the rights to use your image to someone else to sell for money, they can sell the rights to use your image to causes you are vehemently opposed to.
I very much doubt that you are correct in the extreme examples of what the license allows, e.g., putting Aunt Martha's picture with snot coming out of her nose on a network television broadcast.
That having been said, I doubt that particular commercial is in Adobe's business plan, either.
I'm pretty annoyed by the way that licensing is hidden under a banner of "free". I charge for the rights to use my images, if I have to give those rights away to use a web site then, from my perspective, that site isn't "free".
Yes, indeed. This is the fourth time I've addressed that mistake underneath the post you've replied to.
Yes, it was a mistake. Yes, that isn't an exceptional error rate. Mea maxima culpa.
Indeed, I did screw that up.
(I caught that and replied to myself with a correction and apology just after posting, but it's harder to see with the partial-comments display.)
I retract my comment about the 35/year, obviously I misremembered what I'd read, the true SSA number is much higher than my comment would indicate. Mea maxima culpa.
almost every time I come across a 'bug' in our ERP system, it's because a clerk did something wrong.
*That's* a bug in your ERP process. I've run projects that required large-scale, high-quality data entry. E.g., 600,000 French verb conjugations. Of the following factors:
- the extent to which the UI helps the clerk enter the data quickly and easily
- the extent to which intelligence can be and has been applied to detect errors in entered data via checks against other data sources and/or sanity checks, or to detect possible errors in entered data
- whether or not data was entered redundantly by multiple clerks and cross-checked
- how "wrong" the clerk was, that is, the overall error rate of the individual clerk
the latter was by far the least significant in every case.
That people type the wrong things sometimes is, for the most part, unavoidable. It's how you cope with that reality that makes the most difference.
In the case of the SSA, I'm surprised the false death rate is only 35 a year, I actually think that's an error rate to be proud of (out of 300,000,000 people in the US)
I am surprised nobody mentioned that those other companies (RIM, Nokia and Synaptics) also hold spurious patents that could block iPhones?
Perhaps they didn't mention it because they don't know it, or don't believe it? What do you mean by spurious? Which patents? Are all three of those companies prosecuting those patents so far?
seems Apple is just joining the fray
The article directly states that Apple has been filing for gesture patents for years, so... I'm not sure I could agree that they're "just joining" the fray, unless you mean something different than filing patents at all.
I guess my final comment on all of this is that reforming abuses of the patent system isn't going to happen, or should it, by individual companies opting out of the process. Opting out of the patent system unilaterally in many cases will simply mean "you lose". Fixes need to come at a more systemic level, IMHO.
The price isn't that bad, it's instructive, I think, to realize that using memory cards as one-time disposable resources is arguably half the cost of storing photographs on slide film used to be.
For that price, one could reasonably send a memory card home and keep one with you, which seems a completely sensible backup policy.
And, of course, you *can* reuse the cards.
I don't see why 5000K LED light differs from 5000K incandescent light. A mixture of different-colored LEDs formulated to produce 5000K light wouldn't have any 'coldness', as it would use more of the spectrum than the first generation of blue-white LED lamps that you could get at the Sharper Image.
Indeed, the problem with trying to recreate black-body spectra with (a small number of, say 3 or 4) single-frequency band LEDs is more subtle than a problem of coldness or warmness. The problem is that in a spiky-spectrum light, looking at a variety of different colored objects, for light to look right we expect the colors of known objects to maintain stable perceptual relationships with each other--and in spiky-spectrum light they don't. The general term for the perceptual shift of colors despite color adaption within the brain is called "metameric failure".
A contrived example will give you a sense of the problem. Imagine that we use a red, green and blue LED in your system, and that the green LED emits at a wavelength of 555nm. That spike is pretty sharp. If a green object's reflectance spectra isn't very smooth, for example, if the object reflects a huge pile lot of light at 550 nm but no light at 555nm, we're not going to "see" any green under the 3-LED "white" light, despite the large green component.
That's a contrived example, but in real-life examples the perceived color relationships between real world objects are yanked around enough by spiky-spectrum light to be perceived widely as "something ain't quite right."
Now, nothing is saying you couldn't do a much better job with 100 different wavelength LEDs, or maybe even 20, or 8.
My point here is that the color problems of flouros are more complicated than a mere issue of color temperature, and even three LEDs don't really solve the problem here.
The spiky spectrum of fluorescents does cause things to look weird. While changes in smooth spectrum light are things that the eye and brain can adapt to (which is why we can still see color in both sunlight and indoor light), those mechanisms don't and can't adapt well to spiky emissions spectra, having only three types of color sensors in the eye (well, for a few folks, four), and the fact that objects in the real world also don't have smooth reflectance spectra, all combine to lead to Metameric failure.
I'm pretty sensitive to color (professional danger, I suppose), and I notice these effects often.
One other concern I have about CFLs are for my customers. I'm a photographer, and many of my prints over the years have been made using with a digital enlarger that still uses traditional color photographic chemistry. Archival color prints are still very sensitive to ultraviolet light, which CFLs produce a metric buttload of. I've warned customers about exposing photographs to direct sunlight or fluorescents for years, I use specially coated glass in my framed prints to reduce UV transmission from about 30% to a few percent, I've suggested some of the UV-blocking sheathes for fluorescent tubes for some applications. But, in the end, some of my customers are going to miss some of my suggestions and some are going to end up with their prints fading in years instead of scores of years, and that isn't good for me or for them.
(Going forward, I have other options, of course, but that's a whole 'nother post...)
HAKMEM item 93 is solved. Back in 1972 when HAKMEM was written, the AI Lab folks estimated a year or so of computer time then, I'm guessing, given how long has passed, that this was a bit optimistic.
My guess, and it's not much more than a guess, is that the neural net will dominate the compression/decompression time, and you'll have to run that both during compress/decompress time the same way, as well as turning it's outputs into probabilities for arithmetic coding. If I'm right, then yeah, compress and decompress would have fairly similar performance.
Actually, it depends on the precise wavelength, and how transparent/opaque you mean. A typical sheet of framing glass will eat about half your "UV load". UV-coated glass will pick up significantly more than that, but still less than 100%. And again, the numbers depend precisely on how you weight across the spectrum.
I did that Google, and my mileage still varies. I didn't find a good glass absorption spectrum, but what I did find was this, in a comparison of camera filters (whose glass tends to be a bit thinner than framing glass, but not much). Note the visible difference between different brands of filters in the absorption under 400nm based on coatings.
UV absorbing glass isn't really that much more expensive than similarly sized clear glass, what's expensive is glass with anti-reflective "museum" coatings. (Not the matte finishes, but actual interference-based optical coatings.) I purchase that regularly, and it is pricey, but plain old UV coatings are relatively cheap.
As I recall, the Chomira and Noritsus use LEDs, the Cymbolic Sciences LightJet uses lasers (with friggin sharks...). I've seen the (room-sized) LightJet at Calypso Imaging opened up, the thing prints on a 50" roll of photographic paper, it's just incredible.
You don't think your traditional chemistry prints are in for that long a life, do you? If you've got standard color chemistry prints (e.g., Fuji Crystal Archive), you may very well be watching those prints slowly fade from exposure to UV light. This problem will accelerate over time as governments (like those in Australia and California) begin to require UV-emitting compact fluorescents in home lighting, very few people invest in UV-protecting glass.
And as a production manager of a newspaper (http://thecatholicrecord.org) I have never heard of a photoshop printing digital using 'traditional methods.'
I do it all the time, using labs that have the Cymbolic Sciences LightJet, or Chomira-type printers. These printers can really be thought of as digital enlargers, putting digitally controlled light onto traditional materials like Fuji Crystal Archive. Traditional chemistry to develop the result gives you an essentially "traditional print" from digital bits, and if you start from a digital original there's no need to rephotograph anything. (If you start from a slide, which is what I originally did, then a drum scan of the slide is typically your highest-quality option, and then use the above process to produce a result.)
While it may seem paradoxical to use a digital printing mechanism when you want to start from a slide and end up with a traditional chemistry print, the benefits of doing so in terms of color management and repeatability far outweigh the extra work, at least for fine art prints. This workflow allows you to not suffer with, say, the sorts of necessary contrast increases one typically suffers when printing from slides. If you have a digital original, it's even easier, of course. The only real catch is that these printers are large and expensive, but a large number of top-tier fine art nature photographers use this workflow for their prints today.
Slashdot Mirror
The answer is "yes", as it turns out. The appropriate sections of the license agreement are already copied a buncha times on this thread, so I won't repeat them, but yes, Adobe can, under their license, use the images, sell the use of those images to other people, and pretty much do with them anything they want short of preventing you from doing the same.
Yes. As I understand it from discussing the matter with a lawyer earlier today, the license agreement gives them the right to take images you load to public areas of their web site, and gives them free reign to do with them as they will. They *don't* have your copyright, you can do with them as you will as well, but they can, for example, put those images in a TV ad for their product, they can sell rights to use your image for money, they can sell the rights to use your image to someone else to sell for money, they can sell the rights to use your image to causes you are vehemently opposed to.
I very much doubt that you are correct in the extreme examples of what the license allows, e.g., putting Aunt Martha's picture with snot coming out of her nose on a network television broadcast.
That having been said, I doubt that particular commercial is in Adobe's business plan, either.
I'm pretty annoyed by the way that licensing is hidden under a banner of "free". I charge for the rights to use my images, if I have to give those rights away to use a web site then, from my perspective, that site isn't "free".
Thanks, and sorry if I came off a bit grumpy, my bad. Take care.
Yes, indeed. This is the fourth time I've addressed that mistake underneath the post you've replied to. Yes, it was a mistake. Yes, that isn't an exceptional error rate. Mea maxima culpa.
Indeed, I agree that I had misread.... I replied to myself catching my error and retracting this a minute after I posted.
Indeed, I did screw that up. (I caught that and replied to myself with a correction and apology just after posting, but it's harder to see with the partial-comments display.)
I retract my comment about the 35/year, obviously I misremembered what I'd read, the true SSA number is much higher than my comment would indicate. Mea maxima culpa.
almost every time I come across a 'bug' in our ERP system, it's because a clerk did something wrong.
*That's* a bug in your ERP process. I've run projects that required large-scale, high-quality data entry. E.g., 600,000 French verb conjugations. Of the following factors:
- the extent to which the UI helps the clerk enter the data quickly and easily
- the extent to which intelligence can be and has been applied to detect errors in entered data via checks against other data sources and/or sanity checks, or to detect possible errors in entered data
- whether or not data was entered redundantly by multiple clerks and cross-checked
- how "wrong" the clerk was, that is, the overall error rate of the individual clerk
the latter was by far the least significant in every case.
That people type the wrong things sometimes is, for the most part, unavoidable. It's how you cope with that reality that makes the most difference.
In the case of the SSA, I'm surprised the false death rate is only 35 a year, I actually think that's an error rate to be proud of (out of 300,000,000 people in the US)
Perhaps they didn't mention it because they don't know it, or don't believe it? What do you mean by spurious? Which patents? Are all three of those companies prosecuting those patents so far?
seems Apple is just joining the fray
The article directly states that Apple has been filing for gesture patents for years, so ... I'm not sure I could agree that they're "just joining" the fray, unless you mean something different than filing patents at all.
I guess my final comment on all of this is that reforming abuses of the patent system isn't going to happen, or should it, by individual companies opting out of the process. Opting out of the patent system unilaterally in many cases will simply mean "you lose". Fixes need to come at a more systemic level, IMHO.
Indeed, it's that last point that gives me the most concern about using DVDs on long trips.
The price isn't that bad, it's instructive, I think, to realize that using memory cards as one-time disposable resources is arguably half the cost of storing photographs on slide film used to be. For that price, one could reasonably send a memory card home and keep one with you, which seems a completely sensible backup policy. And, of course, you *can* reuse the cards.
Well, they obviously don't believe the share price represents the value of the company.
... or honesty ... or something ... disturbing.
So, they could buy the stock for $12, thought it was worth more than $31, and weren't buying more?
I find their lack of faith
Indeed, the problem with trying to recreate black-body spectra with (a small number of, say 3 or 4) single-frequency band LEDs is more subtle than a problem of coldness or warmness. The problem is that in a spiky-spectrum light, looking at a variety of different colored objects, for light to look right we expect the colors of known objects to maintain stable perceptual relationships with each other--and in spiky-spectrum light they don't. The general term for the perceptual shift of colors despite color adaption within the brain is called "metameric failure".
A contrived example will give you a sense of the problem. Imagine that we use a red, green and blue LED in your system, and that the green LED emits at a wavelength of 555nm. That spike is pretty sharp. If a green object's reflectance spectra isn't very smooth, for example, if the object reflects a huge pile lot of light at 550 nm but no light at 555nm, we're not going to "see" any green under the 3-LED "white" light, despite the large green component.
That's a contrived example, but in real-life examples the perceived color relationships between real world objects are yanked around enough by spiky-spectrum light to be perceived widely as "something ain't quite right."
Now, nothing is saying you couldn't do a much better job with 100 different wavelength LEDs, or maybe even 20, or 8. My point here is that the color problems of flouros are more complicated than a mere issue of color temperature, and even three LEDs don't really solve the problem here.
The spiky spectrum of fluorescents does cause things to look weird. While changes in smooth spectrum light are things that the eye and brain can adapt to (which is why we can still see color in both sunlight and indoor light), those mechanisms don't and can't adapt well to spiky emissions spectra, having only three types of color sensors in the eye (well, for a few folks, four), and the fact that objects in the real world also don't have smooth reflectance spectra, all combine to lead to Metameric failure.
I'm pretty sensitive to color (professional danger, I suppose), and I notice these effects often.
One other concern I have about CFLs are for my customers. I'm a photographer, and many of my prints over the years have been made using with a digital enlarger that still uses traditional color photographic chemistry. Archival color prints are still very sensitive to ultraviolet light, which CFLs produce a metric buttload of. I've warned customers about exposing photographs to direct sunlight or fluorescents for years, I use specially coated glass in my framed prints to reduce UV transmission from about 30% to a few percent, I've suggested some of the UV-blocking sheathes for fluorescent tubes for some applications. But, in the end, some of my customers are going to miss some of my suggestions and some are going to end up with their prints fading in years instead of scores of years, and that isn't good for me or for them.
(Going forward, I have other options, of course, but that's a whole 'nother post...)
*rolls eyes*
HAKMEM item 93 is solved. Back in 1972 when HAKMEM was written, the AI Lab folks estimated a year or so of computer time then, I'm guessing, given how long has passed, that this was a bit optimistic.
My guess, and it's not much more than a guess, is that the neural net will dominate the compression/decompression time, and you'll have to run that both during compress/decompress time the same way, as well as turning it's outputs into probabilities for arithmetic coding. If I'm right, then yeah, compress and decompress would have fairly similar performance.
I have a few emails, somewhere, that date from 1981. Scary stuff. :)
UV absorbing glass isn't really that much more expensive than similarly sized clear glass, what's expensive is glass with anti-reflective "museum" coatings. (Not the matte finishes, but actual interference-based optical coatings.) I purchase that regularly, and it is pricey, but plain old UV coatings are relatively cheap.
Dangit, if you'd only said "27 years" instead of 30 (for the digital data, that is.)
As I recall, the Chomira and Noritsus use LEDs, the Cymbolic Sciences LightJet uses lasers (with friggin sharks...). I've seen the (room-sized) LightJet at Calypso Imaging opened up, the thing prints on a 50" roll of photographic paper, it's just incredible.
You don't think your traditional chemistry prints are in for that long a life, do you? If you've got standard color chemistry prints (e.g., Fuji Crystal Archive), you may very well be watching those prints slowly fade from exposure to UV light. This problem will accelerate over time as governments (like those in Australia and California) begin to require UV-emitting compact fluorescents in home lighting, very few people invest in UV-protecting glass.
I do it all the time, using labs that have the Cymbolic Sciences LightJet, or Chomira-type printers. These printers can really be thought of as digital enlargers, putting digitally controlled light onto traditional materials like Fuji Crystal Archive. Traditional chemistry to develop the result gives you an essentially "traditional print" from digital bits, and if you start from a digital original there's no need to rephotograph anything. (If you start from a slide, which is what I originally did, then a drum scan of the slide is typically your highest-quality option, and then use the above process to produce a result.)
While it may seem paradoxical to use a digital printing mechanism when you want to start from a slide and end up with a traditional chemistry print, the benefits of doing so in terms of color management and repeatability far outweigh the extra work, at least for fine art prints. This workflow allows you to not suffer with, say, the sorts of necessary contrast increases one typically suffers when printing from slides. If you have a digital original, it's even easier, of course. The only real catch is that these printers are large and expensive, but a large number of top-tier fine art nature photographers use this workflow for their prints today.