|
|
NIPerlin Patch and Kosada/smokris I was looking where I probably shouldn't be poking and found something kind of interesting. I noticed that there is a Perlin Noise kernel in the NI Image Unit thingy recently, and was looking at the description (see attachment). I noted that the copyright was given to Perlin, but hey, Perlin didn't write it in CI. I also noted the lack of NI copyright, which is usually pretty prominent. Something didn't quite rub me right about it since I, like an elephant, remember insanely stupid details, and remembered smokris porting Perlin's kernel. So, I looked in the actual plugin and saw this in the actual kernel... // Improved Noise - Copyright 2002 Ken Perlin. // =========================================== // // Conversion to CIKernel by smokris@kineme.net, 2009.02.26 // // Copyright (c) 2009 Kosada Incorporated. // // Permission is hereby granted, free of charge, to any person // obtaining a copy of this software and associated documentation // files (the "Software"), to deal in the Software without // restriction, including without limitation the rights to use, // copy, modify, merge, publish, distribute, sublicense, and/or sell // copies of the Software, and to permit persons to whom the // Software is furnished to do so, subject to the following // conditions: // // The above copyright notice and this permission notice shall be // included in all copies or substantial portions of the Software. // // THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, // EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES // OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND // NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT // HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, // WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING // FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR // OTHER DEALINGS IN THE SOFTWARE. Again, I do note that Perlin's copyright isn't missing from the patch description, and NI doesn't take any kind of claim, so it seems like some thought went into the m.i.a copyright in the description, but who knows. To err is human.
|
I think its safe to say this is covered by NI's copyright clause because you installed this via FXFactory, which has its own licensing clause when you install it, and is a software component thereof. If you are curious about Noise Industries products and licensing, (or code since you are clearly poking around the CIKernel sources), your best bet would be to contact Noise Industries directly.
I'm not knowledgable on our policies to be frank, so I can't really respond definitively.
Oh, I think I've miscommunicated something.
I have this code totally apart from NI. I can use it however I want without involving NI, as that is part of the Kosada clause on it. I have to note in a place where it can actually be seen that it is a Kosada copyright though, which NI has arguably done, if it can be assumed that all users go into the plugin file to look at resources. This is my point. In my eyes, it seems like it should have a Kosada copyright in the patch description to accurately convey the origin of the patch.
So, I'll definitely use it however I see fit, and try to be respectful of the copyright holder by placing their copyright in a notable place, as is asked by the agreement in the source code. NI doesn't really come into the picture here if I wish to use this.
The only reason I'm poking around the code is because when I saw it, it was obvious where it came from, and that credit was not properly given, so I was thinking "hey, maybe the smokris's copyright notice is still in here somewhere".
This is where he posted it:
http://kineme.net/Applications/Compositions/PerlinNoiseCIKernel
It may have been posted first a bit earlier (?)
Smokris is really loose about i.p. so for all I know, he knows about this, and totally doesn't care. It is a port, and a pretty small thing, so it's more of a pat on the back than anything. It's just amusing to me that the only condition is to print the copyright prominently, and that this hasn't been fulfilled.
I do know that if I was using this plugin and had no clue about kineme/kosada/smokris, and saw the copyright that should seemingly be there, I would google kineme, and then see smokris's work.
I'm jaded in that I note that society has come to a point where you can give something for free under the condition of a "shout out" (as I see this scenario, more or less), and even that very small request isn't followed. When someone distributes free code, I see it as a business card for them, and try to follow whatever they ask.
I kind of already touched on this, but again, there is this vibe that I should be asking about licensing b/c I was looking at the kernel source. The only reason I looked at the kernel source is because it appeared that the appropriate credit wasn't placed in the patch description. The only way NI can really say that the criteria has been met is if it's ok to look at the kernel source, because the copyright isn't anywhere else. Also, there is little reason to take/use any of the NI Kernels because they are either textbook stuff, or already existing in the QC patch list (thought they are exceedingly handy for speed of workflow).
Not always -- there are IP battles worth fighting, they're just the rare exception (contrary to what modern legal departments and patent offices would have you believe). Smokris knows about this, and in fact, if I recall my history, NI actually contacted Kineme to get most of that blabbing in there to ensure that we wouldn't/couldn't pull a fast one on them legal-wise. Most of that text is therefore at NI's request.
Depends on what you mean by "shout out" -- the terms of the license are such that the notice stays intact (it did). The terms as stated have been met. Looks like you're looking for a fight when none exists.
If not that, then where? On the box? In the instruction manual? On the app UI? What's the limit? This quickly grows beyond scalability. The copyright only applies to that kernel. Not the patch. Not the application. Not the product, the website that sells it, the internet it's sold on, or anywhere else.
I'm as rabidly anti-IP as anyone (and with just enough law training to not know anything, but sound convincing ;), but this isn't a fight, and if it were, it probably wouldn't be worth fighting. Human life is worth so much more than fighting over who invented things. Truth be told, all modern inventors are irreversibly and eternally indebted to those that came before. Jack E. Bresenham and Xiaolin Wu are personally responsible for everything you've done that involves a screen. Do you know what they did? A couple co-workers wrote all of CG -- where are their props? Who invented CoreImage, which made this thread possible? (Probably jharper, but I haven't asked him) What about the intel engineers who came up with clever tricks to make things fast (branch prediction, hardware-assisted prefetchers, store buffers, etc)? That's all done in code (microcode), so shouldn't apps that use it (e.g. by executing instructions) have to tip their hat to it?
This is a slippery slope, my friend. Pull out now and make something awesome. It's more personally satisfying.
(For the record, Noise Industries asked permission before including my port of the Perlin CIKernel in their product, and helped fund its development. I don't personally have issue with their fulfillment of the legal requirements, since, as you noted, they did retain the MIT license and copyright in the kernel source.)
Oh no, not looking for any fight. My actual tone isn't coming across. My curiosity was simply piqued.
Where? On the patch description, where copyright info usually goes, and where it is for Perlin.
There is a large difference between work done for hire and work that's pinched, as far as your second to last paragraph goes. When work is done for hire, you go into it understanding what the deal is. It's a trade of labor. However, if one has made something of value independently, then this has a value that is set by the marketplace for the item that has been made.
If I work as a carpenter, I expect my weekly wage, and make a chair that is defined within a box drawn by my boss. If I am an artisan, and make a particularly ornate chair independently, I'm my own boss, and can sell that for what I see fit, or what the market bears.
This drifts OT and into philosophy, but I do believe that intellectual property is as valid as the concept of physical property. If not for both, we may as well have a system of anarchy/"might=right".
On the last paragraph, agreed.
Again, my main "huh?" was from the fact that I thought NI's agreement was that we weren't supposed to be looking at the source or backwards engineering, so to have to look at the source to see copyright seemed weird.
but is the kernel the patch? Or is the patch something else that uses the kernel? (Hint: I am confident smokris's code pales in comparison to the patch code that coordinates its operation). What if there's a limitation where the patch cannot have a custom copyright string? (I think that's the case, but I'm not certain -- it all depends on how it's tied together).
Where's the line between art and non-art? If I design a concrete slab, can I charge royalties for every sidewalk tile in the US (or the world)? someone designed that. what about a chair -- do I get a cut from every chair sold to every elementary school? Fluorescent bulbs? Toasters? C# chords?
movinstructions? Each of those is simultaneously trivially simple, and fascinatingly complex. Who decides if it's art?Well, the patch does say copyright Perlin, and others say NI, in the same place where you would write up whatever else you want to about the patch - in the description, as shows up in the patch list.
These descriptions are unique among various kernels in the Image Unit when they show in the patch list, and this is a supported function of image units when loaded by QC; it's not a limitation in this case.
I see what you mean about one kernel vs. the whole Image Unit, but programming an Image Unit is truly one of the easiest entry points to expanding QC, and barely requires anything above the programming of the kernel.
Just to make sure that this is taken in the right context, I have really appreciated NI's offerings. If I didn't, I wouldn't have seen the NIPerlin patch, and noted that it looked identical to the kernel I already had loaded - the one I got off of Kineme.net. I thought it was most likely that smokris knew and didn't particularly care if the copyright was in the description or not (as he has said is the case). (the following stuff doesn't relate to this scenario at all, but is spurred by Chris's points about I.P) On that last thing, I wasn't making a distinction between art and utilitarian design, I was making a distinction between for-hire work, and when you see something that is already made, that then you have to negotiate to obtain. I'm also making a distinction between the fact that when something is done for-hire, in conjunction with a company, said company informs large portions of the criteria, so a lesser level of compensation or even total anonymity, makes sense.
If you designed road reflectors though, yeah, you could have made royalties on them, and the concept of reflection existed years prior ;) Not so sure about concrete slabs, or permutations of the wheel at this point in time. It may be public domain.
This discussion reminds me a bit of the movie "Flash of Genius" where the automotive industry is arguing that the windshield wiper design was obvious because all of the parts were in prior existence. The protagonist pulls out "A Tale of Two Cities", begins reading the first lines, and asks if those words were in existence prior to the book, and if this makes the book unoriginal.
"I see what you mean about one kernel vs. the whole Image Unit, but programming an Image Unit is truly one of the easiest entry points to expanding QC, and barely requires anything above the programming of the kernel."
You clearly have not done much Core Image development.
I've made many Image Units, and it's the one thing that I do habitually do to expand my own install of QC.
You are incorrect.
Maybe I was oversimplifying to say it's the easiest thing, but for me, this has proven to be the case. It's not necessarily trivial to setup the .m file, but it's not horrible either.
Image Units/Core Image is easier for me to understand than, for instance, something that seems way more complex to me, like the Freeframe plugin project, where I can understand what each function is doing basically, but aspects go over my head.
Do you feel that using GLSL to make a QC plugin that processes image is harder/easier/equivalent to using a CIKernel as the basis for an Image Unit? I'm not asking this to prove any points or make any argument, it's a legitimate question. Making Image Units has made sense to me, but my eyes sort of glaze over when trying to implement GLSL in a QC plugin. Did you use one of the standard 10.5 QC Plugin projects to inform the way that you do your GLSL based plugins?
My point is that complex Image Units are not just simple wrappers to complex CIKernels. Doing anything that hits limitations of CI such as non variable loop iteration at compile time (you cannot pass loops of variable iteration amounts) require either brute force or clever work arounds in Obj-C, for example.
So while yes, some/most CI Units are simple wrappers that pass variables to a CIKernel, you occasionally have to do some hard work to make sure you handle the ROI/DOD transforms as well as overcome issues with limitations in CI, let alone doing funky things like creating images procedurally in Obj-C with proper bit depth/gamma to pass into units as secondary image inputs to a kernel, where you may only want to publish 1, just as a random example.
Edit, regarding GLSL, frankly I found GLSL much ore intuitive than Core Image when I first started working with it, due to the lack of assumptions about sample space and transforms. As far as my plugins are concerned, no I did not, but 95% of it is just wrapper code to get it to get the GLSL loaded and set up OpenGL state (and restore it).
GLSL is significantly faster than CI, I will tell you that. Which is why for my own non color sensitive work I use it. As for complexity, once you get one plugin working via GLSL, its trivial to get other shaders working similarly.
Ok, I get where you're coming from on that one.
I've also found GLSL (in QC) way more intuitive, mainly because of the fact it's way better documented, and I don't have to mess around with essentially converting GLSL to CI through trial and error.
Thanks for sharing your thoughts on that. I guess I should really go ahead and try it again.
I guess when you do this, the image that results has to be QCImageBuffer?
Yes. When you output an image from a QC plugin (using the standard public API) you can do it 4 ways. 2 ways use predefined 'helper' functions that output a GPU based texture, or a CPU based pixel buffer. The other 2 ways its up to you to create a class that uses the appropriate protocol and create methods that output the proper values for each one, ie, bounds, texture ID or buffer base address, etc.
Any of the 4 output to a 'opaque' object called a QCImageBuffer, that acts as a wrapper for the various image formats that QC supports via plugins, as well as their reported colorspace, bit depth and pixel format. The QCImageBuffer wrapper around your image allows QC to transparently 'move' your image from the CPU to the GPU and vice versa (for example, in situations where outputting a texture from say a v002 Blur, it outputs a QCImageBuffer representation, which then lets another plugin (assuming I've not done anything buggy and stupid), download it to the CPU by calling the standard methods. The v002 Blur does not need to know about anything downstream and how it handles images, the QCImageBuffer class handles that for you.
The private patches tend to not use QCImageBuffers internally because the image class types QC handles natively all have internal code that handles the cases for, say, CVPixelBuffer, CVOpenGLTextureRef, NSImage, CGImage, NSBitmapImageRep, CIImage, etc etc. It "knows" what to do with those in each case with a pre supplied patch because they are written to handle it. QCImageBuffers are, as far as I am aware, only output from 3rd party plugins? (I may be wrong on that last point in some cases, but the point is they are there to abstract differences in image formats for patches that just want to access an image ONE way, regardless of the source format).
Is that helpful?
It seems like factors of ten speed difference, why so?
You're so right on, with that one (I don't think that's a GLSL thing, but an OpenGL aspect, as far as sprites/billboard, and built in blending (?)). It can actually be faster to blend w/ two billboards, even THEN putting them in an RII, and running that output through CI, than using composite blend CI stuff.
However... CI seems blazingly faster, in some instances, than it did in Leopard for me.
The differences I get in fps w/ stuff like ZoomBlur vs v002 ZoomBlur, is really trivial, and every so often I run into odd scenarios where CI seems to gain some fps, but I'm not sure of the rhyme or reason behind this yet.
It's usually when filtering an image, which I am then rendering with GLSL. In that case, I sometimes see modest fps improvement from using something like CI to pre-treat the image, vs. v002. This seems pretty darn computer specific, as v002 basically always outperforms CI w/ my white macbook w/ intel, but this can drift into opposite territory w/ my macbook pro that has the nvidia combo.
It definitely is. Thanks for filling in some gaps in my knowledge, you made some things much clearer for me.
I'd be curious to see patches illustrating CI being faster. I've never seen it if personally and have off list confirmations from folks at Apple saying its faster. Not that I don't believe you, its just It may be on hardware I've not looked at or seen. The only thing I can think of is that Core Image invokes the tiler sometimes aggressively not just when hitting texture size limitations but when it can get an added performance boost since multiple, smaller texture access may be faster than one larger one? Speculation, but multi-pass is meant to be fast on GPUs, so I could see that maybe under some fragmented VRam situations HD frames may be processed faster in CI.... Hrm. Things to think of.
Oh, and with regards to compositing being more performant via two billboard with over blend mode vs two images, core image over and then a billboard, consider this.
image a -> linearizing + color correction -> core image over image b -> linearizing + color correction -> core image over
then new image -> render on billboard
vs
image a -> render on billboard image b -> render on billboard
;)
I'm pretty darn sure it's when I've been inputing image into a shader or openCL type patches, and never any other time, but I certainly wouldn't swear to it. Next time I'm doing something where this pops up, I'll send something your way. Like you say, I suspect that some of it could be gpu specific.
Ah, maybe there is some optimal format for textures that Core Image -> OpenCL handles automatically that perhaps QC and thus v002 Blurs etc do not? So many possibilities ;)
In those cases, the gain w/ CI has usually been modest - like 3~5 fps on average. It probably pretty much always been when using blur to "smooth" extrusion, whether with GLSL or OpenCL.
I definitely know that on Leopard, using my white Macbook, the thought that CI would ever be faster than v002 would be ludicrous. The difference in performance is night and day, with v002 being tens of frames per second faster most times.
Yeah, good point.
...and if you put those two billboards in an RII, and do further processing, it's usually still faster.
It seems like there used to be a kind of buzz about "don't use RII, it slows stuff down", but if I didn't use RII the way I have in some setups, I'm 100% confident that some of the coolest stuff I've done would have never even run fast enough to work. Especially considering that if you are heavily processing a scene with filters, that you can usually lower the res and get away with it easily, and speed everything up.
I just wish that RII didn't kill the MSA that is in SL. This really puts a thorn in the side of media server-esque activities, where one would WISH that the image you're processing didn't have aliasing. I know it sounds weird that I'm advocating downsampling using RII on one hand, and wanting MSA on the other, but I find myself in scenarios where both techniques would work hand in hand for optimal performance.
My understanding is that in 10.6, the rendering backend changed significantly so Render in Image now uses operations that are far faster in GL, and those operations are available across the platform that 10.6 supports, whereas in 10.5, the GL drivers did not support those operations across the board, so slower alternatives were often used. I think thats the gist of it.
(frame buffer read/draw and blit seemed to be used now and are the state changes I had to track down in 10.6 to get things working again (or I got garbage)).
As soon as I read a recent comment about linearizing (was it on this thread or the Apple list?) I thought, well that's gotta be a part of it. So there's colour conversion while not on Renderer too. I was wondering if the QC renderers are more optimised to GPU hardware level instructions or are just closer to the business end of the pipeline and have an edge there too?
Good point -- it's probably doable (but again, not required strictly by the license).
why should I get paid multiple times for exerting effort designing something once? This happens in software because the cost is astronomically high, and by selling several copies at Astronomical_Cost/N, the hope is to in aggregate recoup the cost by selling >= N copies. (whether or not you think that is legitimate, that's how it's phrased). But if I can spend just 8 hours designing a road reflector, why should I get paid for more than 8 hours of work? Because it's such a good idea that showering me with money will encourage my inner genius to make more useful things?
who decides obviousness? While in highschool, smokris and I "infringed" numerous obvious patents by simply thinking about solving problems (IBM's thick lines patent, IBM's XOR cursor patent, possible premultiplied alpha patents, possibly others?). If a couple random nerds in highschool in suburban ohio can come up with those ideas in a vacuum, they're almost certainly not "novel" or "original".
Dickens wasn't inventing a tool with "A Tale of Two Cities", he was entertaining (or writing an exposé or something -- authors write for lots of reasons). Claiming it's not original because its constituent parts already existed means everything is unoriginal because it's composed of already-extant pieces. However, that counterexample does not imply that all compositions of existing things is unique or original.
I guess my point is that there's a gigantic gray area between the two ends, and nearly all IP-related debates are forged by 1) pulling tons of examples out of this gray area to build a nice straw man (what I've done, and a bit of what you've done -- I started it though ;) and then 2) making a mockery of how silly their straw man is, and thus the whole spectrum should be black or white.
Until there's an actual rational way to determine if something is one case or the other, I choose to apply occam's razor -- I don't necessarily see it as a good solution, but until people stop talking about "registering" their Totally Awesome And Amazing Ideas, I honestly don't see much credibility in an alternative.
I'm not making the argument that this should be, merely describing that this is.
An alternate view is that this is functional for society, because it fosters industry, and provides for multiple people, not just whoever invented said product. To use the road reflector example, a company has to form to produce this, people install them, etc., and all of those people make a livelihood off of said invention.
It seems that you're making the argument (for the point of illustration) that the person that invented the reflector, and spent 8 hours, should likely make less then a person who works on the road crew and installs them for a week. To be fair - and I'm not equivocating - I can actually buy that argument, and do see a certain kind of social justice in it, but it's not how society works at this point in time.
Is the worth of that item in the 8 hours spent making it, or in the countless hours of time spent using it?
"Yesterday", by Paul McCartney started out with the mock lyric, "scrambled eggs, oh my baby how I love your legs". While that is entertaining, he refined this into the classic song that it came to be. His choices were informed by his unique life experience, his intellect, and all of the things that influence all of us when we work.
Now, in the years since that song was written (over the course of a few weeks it sounds like), it has been listened to countless times, and while many may have wanted to flip the station, I'm sure many were entertained. How many minutes of time have people been entertained by the song Yesterday, if you were to add them up? Does that amount of time have any influence on the amount that the creator of this should be getting paid?
Also, every time Paul McCartney goes on tour, or makes a record, how many people in his organization profit? How many people get to eat? Should that factor into compensation? I don't know.
I touched on this above, but how can one quantify the work that was put into gaining the experience to do something quickly? How much time was spent leading up to the "8 hours" spent inventing and designing said road reflector that informed the ability to make this in a brief 8 hours weren't paid for?
There are numerous examples of patents that seem totally inane, and I totally sympathize with that take and have the same concerns. I also realize the inherent foibles involved with having random people judging validity of patents and such, who may not even really be competent enough to understand said issues.
We have law to try to ensure some type of justice, but the fact is that the legal system is a bureaucracy, and bureaucratic systems don't handle unique scenarios well. Our society hasn't developed a very functional escape valve for scenarios like you describe above, and thus, common sense sometimes does not prevail.
The converse is true; sometimes there are actually novel ideas that haven't quite been done. I don't think there is social justice in someone profiting off of the idea or concept of another, and the person that had said concept not profiting simply because they took the idea and did an implementation.
If I write sheet music for a song down, my song is still in the realm of "idea", but I've transcribed it to a physical form that's a blueprint, or instruction for making the music, but it's not the actual "record" - not a full implementation. When one conveys an idea, it can sometimes be a similar scenario in my estimation, though not always.
Agreed, there is a gigantic gray area.
Overall, I weigh in on the side of defending the concept of intellectual property, if we are going to have any kind of system of order and law at all. I see more benefit than harm overall but I'm not dogmatic about it, because in the bigger picture, the structure of our society is very flawed.
Hmm. I don't know about occam's razor, because I don't actually believe in the existence of simplicity/complexity as being "real", only subjective human perception. Things are also both infinitely similar and different at the same time as well, despite subjective appearance... whether we're talking about two of the same model macbook, or a macbook and a squirrel. It just depends on your vantage point.
You have managed to crack me up with the line, " 'registering' their Totally Awesome And Amazing Ideas".
[first, a disclaimer: I'm aware of current laws, and am not advocating disregarding them, even if I don't necessarily agree with them. Disregard law at your own peril.]
I guess I don't buy the whole "fosters industry" argument -- in fact, I'd be willing to bet that a surprising portion of innovation is from passion, not from economic incentive. I agree that an invention can benefit (economically) many, but that isn't dependent upon an invention. Once upon a time I worked for a grocery store. There wasn't an R&D department, and they likely held no patents. Yet they were able to build/renovate buildings, order trucks full of stuff, and weather freak power outages that destroyed their entire refrigerated/frozen stock.
ahh, now we're getting somewhere: do you pay the "cost" (actual creation costs, as decreed by the seller), or the object's "worth" (which is subjective)?
If cost == worth, no one would buy anything and capitalism would fail -- it works by 2 parties exchanging things, both to their mutual advantage. That's in part why monopolies like Comcast suck: they can set arbitrarily high prices, never compete, and they're under no demands to benefit the other. As such, the trade is "one way". If kineme charged "worth", there'd be fewer paying customers, but those that did pay would pay a whole bunch. But since kineme charges "cost", there are people who merely tinker (at a small price), and then a few that actually earn money in part because of the product. Is it reasonable for kineme to expect payment in those cases? (In practice, that's been fairly subjective, and there cannot be a hard-and-fast rule).
With current industry practices, "yes". ripping DVDs wasn't possible till DVD jon, ripping bluray discs is still hairy, recording CDs is illegal. The attempted equivocation is that if I use 1000kWh in a month, I should pay more than someone who uses 100kWh -- that's correct. But that's because the difference (900kWh) had to come from something (coal, uranium, wind, sun, whatever -- likely coal here in the US, because it's been successfully marketed as "clean" and nuclear has been successfully marketed as "dangerous" by greentards). Copying a disc incurs no additional cost to the manufacturer, unlike more energy. Multiple copies downloaded add bandwidth, but it's generally so trivial that is has no cost.
Why should it factor in? PM sets a ticket price (based on venue cost, his and his people's time cost, etc), and if people feel it's beneficial to them to trade their dollars for a ticket, then PM gets what he gets, and he divides it up however he wants. In this case, you're not just paying for "the music", you're paying for the environment (PM's presence, other like-minded people attending, merchandise, whatever). If you were to attend 2 such concerts, I feel it reasonable to pay twice because you're paying for more than just "listening to the disc twice". Ditto for going to a movie twice.
who cares? I mean that seriously. By reckoning previous time, prodigies would get paid less because they spent less time, even for otherwise identical results. Whatever time/resources were spent acquiring that skill, whether it's natural talent (which is extraordinarily rare), time (the most common factor in those that excel), cybernetic implants, whatever, those are all sunk costs, and shouldn't factor in. If I solidly outperform a PhD at work, should I be less compensated because I'm just "that good"? way to motivate! :)
It's more than that though -- there's an oft-ignored aspect of humanity at work too: each successive generation is getting smarter. What wasn't obvious 50 years ago may be obvious today. What wasn't obvious last year may be obvious today (I think some of our personal discussions have touched more concretely on that subject). Using "obviousness" is fundamentally flawed because 1) it cannot be measured 2) it isn't constant and 3) those that determine it are the least qualified. Any one of those is a total deal breaker.
Rather, (not to go all tinfoil) we have legal departments, where dollars are traded for "justice". That's not always the case, but it in cases where it matters it seems to be. A mockery of protection (above) combined with a mockery of justice is my main beef with the current scenario. (Mind you, I've profited off of it more than I've lost, but that doesn't mean I see it as something responsible/sustainable).
I don't think justice is required 99% of the time for those cases. Jet pack. flying car. Faster Computer. There, I've got 3 new ideas -- cash in when they hit the scene! One profits because they take (hopefully calculated) risk -- actually spending resources to bring an invention into reality. Not because they spend resources (whether it's time or money).
Let's take it farther: I fire up Illustrator, and after making a fool of myself for a while because I cannot draw, I sketch the jet pack as concept art. What now? (and do I factor Illustrator lessons into the cost?)
Farther still: I actually study mechanical engineering, learn autoCAD, and design a blueprint. What now?
Farthest: I actually build said device. What now?
And here we see an interesting distinction: the first two (idea, sketch with no specs) are laughably don't-care territory. (At best, the second is copyright fodder, because it's a work of art). The third is where it gets interesting: they're actual instructions/plans to implement said device. With that piece of information, I can now clone said device with no R&D costs (I don't need to solve any problems, other than how to get materials).
It's identical with software: I conceptualize a program. I write some comments describing how I think it should work. Then I write the code. At that point, the thing exists -- the code is "instructions" that can be fed to a machine to produce a new thing (in this case, fed to a compiler, to produce an application or framework). Whether or not I ever compile said code, it exists in reproducible form.
So the only question that remains is: what happens if the plans are wrong (they simply cannot work, or they involve non-extant elements)? What happens if the (never-before-compiled) source code doesn't work? Do they lose their "exists" status? I don't know. This is where things are interesting (and testable!).
Using the pattern above, it now "exists", even if it's never been played. Additionally, music cannot "fail" (it always "works", even if it sounds like crap or something ;), so it's not an "invention", it's copyrighted. Performances of said work also cannot "fail", they can suck -- thus, also not invented, but copyrighted (as a derivative work, perhaps?).
I think "defending the concept of IP" is oversimplified: I support copyright as a limited means of protection for "unfailable" works (music, art, prose, etc) -- I feel the term is too long presently.
I support patent on devices that have a definite "pass/fail" implementation. If I make plans for an airplane that doesn't fly, it's not patentable.
I do not believe software can reasonably be patented, because it is simply a mathematical description of steps required to solve a problem. (in real life, it can currently be patented, which is disappointing). So while software can "fail", it's equivalent to a GPS route (which cannot be patented) or a recipe (which hopefully still cannot be patented).
That's why I favor testability -- tests are not subjective (ideally -- crappy tests can be subjective). Using the "all truth is relative!" argument is kinda sophomoric, and self defeating ("If you say that there's no truth, and who cares, how come you say it like you're right?").
Testing for "failability". if it cannot fail, it cannot be patented. If it can fail, it might be patentable -- physical things are likely a requisite to patentability (to exclude recipes, software, and "plots"), but I honestly haven't thought about more domains than these. I'm sure others far brighter than I have spent more time than I thinking about these things. (I wonder if a patent lawyer has tried to patent a legal argument?)
You owe me 5¢ for quoting me ;) (I kid, I kid)
My thought is that it cuts both ways. I am sure there are times when the concept of intellectual property hurts industry, and the kind of scenario where you describe goliath corporations patenting mind-numbingly obvious concepts is a great example of that.
I see the tradition of folk song, for example, as a case the speaks to the benefit of a free, anarchist, exchange of ideas, where very well honed art results from people being able to freely edit that which has come before.
I also truly agree with the remark that passion fosters innovation.
The grocery store may have not had an R&D department, and may have held no patents, but it falls into the case of a company that is benefitting off of the concept of i.p, still. I'm sure a high percentage of products sold were branded, had trademarked logos, many manufactured products had certain trade secrets, and that some even had patented processes involved in manufacture.
Now, is that better than a co-op, or a farmers market? Hmm. Not really. So, you've actually proven a point to me here, but maybe not in the way you intended! The concept of i.p makes the economy less localized, can sometimes hurt local business; this is not my preference.
I agree with you, with emphasis on the "there cannot be a hard-and fast rule".
The problem I see with the comparison to tangible resource is that none of us inherently have right to those tangible resources, but we still have to pay for it, and people can decree ownership, and have ownership based on wherever they are in the social structure, not through any kind of inherent merit. So, it seems reasonable that intellectual property can exist as well.
I understand your comment about copying a disc incurring no additional cost to the manufacturer, and the argument that if you didn't copy it, you would not have bought an additional copy anyway, thus there is no monetary damage. In application to the entertainment industry, I think it ignores a larger issue.
The actual art and media which is created is becoming a loss-leader for the tangible substance that the media is transported on, at the cost of the people producing the element that gives the physical media value.
By that, I mean, if Sony/Phillips owns a large music company, and also owns a large company that makes blank discs, if someone doesn't buy a record, but copies it, they still make money, because they are producing the physical media that it is being written to, and own patents involved, if other companies wanted to produce similar media. If you are a company that makes portable music players in the 90's, like the Walkman, it makes great sense to approach this from both sides of the table as they did.
When the music division, that fuels the reason to sell the blank media, loses money, who really loses money? Sony/Phillips, or the people who the actual losses get foisted upon- the artists involved.
Obviously, we are a few steps past this point in history, and into a scenario where things are being stored on HD's, sent around via the net, etc., and there is no putting pandora back in the box. People's perception of i.p has also been forged by this set of events, and traditionally, people involved in computer technologies are somewhat cavalier about it, because of the perception that putting limitations on technologies like torrents, and stuff like copy protection stymies innovation and the rights of the user, instead of thinking about them as a tool being used by big industry to facilitate the total bankrupting and marginalization of music, cinema, and other humanities as instruments of cultural change or ways for "the small guy" to make money.
I see what you mean, and agree. I was coming from the stand point of "should the boss/leader make more money"?
I thought this was the tact that you were taking actually.
For instance, one of my best friends wrote a top ten single, has had some other moderate successes, and has been able to make a decent living out of that. The guy wrote the song on the toilet in the time it took to think of it b/c he is a prodigy. I guess my argument is that "prodigy" is the result of mulling stuff over in the back of your mind, and weighing things, establishing sensibilities that forge one's ability to do something quickly when it's go time. People that are not prodigious don't really engage in that kind of "off the clock" thought process.
So, it's a little complex. I think that it is totally just that concept of copyright has helped my friend have a career. My take is that the concept of i.p benefits and rewards prodigious people that may not have to spend much time on something of substantial benefit.
I think that areas of intellectual concentration evolve. I have to consider your theory that each successive generation is getting smarter a bit more. I totally agree with your comments about obviousness.
100% agree.
Totally agree. It's a continuum.
As far as a system that depends on a non-extant element, this could possibly be reasonable. If one has really refined something to the point where there are tightly described missing elements that cannot be produced at this point in time, I wonder if it is just for them to be able to profit from their contribution. Someone who has the ability to make the "missing piece" may not ever think of making the missing piece without knowledge of the system that's has the lacking element.
This is total philosophy at this point. Note that I say " I wonder ". I'm not sure how I feel about it.
Agreed. What I mean is that, I think that it is possible that in some cases, a description of a system, that literally COULD fit on the proverbial cocktail napkin, is more than enough to let the cat out of the bag. Maybe a handful of words could suffice to describe a totally novel concept, but still inform many parties about what they need to do to implement it. "Faster computer" is weak, but "rubber tire", maybe not.
I guess I should state that I have been talking about intellectual property in broad terms, but there are certainly distinctions between copyright, patent, trademark, etc., and I'm only talking about i.p broadly for the sake of simplicity.
I agree with you in general. I'm not 100% sure that I think something has to work to be patentable, but I see the validity in the argument, and tend to agree.
I tend to agree with you about the concept of software not being able to be reasonably patented, but I do agree with the concept of being able to copyright aspects of software, and also think that someone's particular way of solving a problem may have some element of art in it.
I wasn't trying to use an all truth is relative argument, but I do think that conception of simplicity vs. complexity is very relative. I don't personally know how to judge simplicity vs. complexity, so I have never been able to relate to the concept of occam's razor. The concept is very appealing, and I understand why it "makes sense", but I can see many times when something that is the "simplest explanation" isn't the case. I see it is a rejection of reality in some cases.
I'm pretty sure "I kid, I kid" may be owned by NBC, in association with Triumph the Insult Comic Dog. Not 100% on that one. (har har har)