Patent reform

The Something Awful Forums > Discussion > Debate & Discussion > Patent reform

Paul MaudDib: May 3, 2006; TEAM NVIDIA:
FORUM POLICE

To answer that, and speaking normatively rather than about current law, I'd say the situation should depend on what precisely is novel in the invention. If the patent is mostly about hardware and the software consists of "a computer algorithm to control the system", no, that shouldn't be patentable. If there's a novel, non-obvious combination of algorithm and data structure that solves some significant problem in the system, sure, that should be patentable.

Remember that something doesn't need to be patented to be protected against infringement. Source code and binaries are still copyrighted - if someone actually stole Apple's work, they would still get boned on copyright even if Apple couldn't sue over rounded corners.

# ¿ Dec 9, 2013 00:37

Adbot: ADBOT LOVES YOU

# ¿ Apr 20, 2024 02:30

Paul MaudDib: May 3, 2006; TEAM NVIDIA:
FORUM POLICE

Kalman posted:

Copyright is perfectly good at protecting that... Except for a line of cases that said that aspects of user interfaces can't be copyrighted, stemming from someone (Lotus, I think) copying the WordPerfect user interface.

Basically, copyright is pretty good at protecting the artistic aspects of software and reasonably good for preventing out and out copying, but is a poor mechanism for protecting the technological aspects of software.

Again, I don't see how that's problematic. Word processor interfaces aren't novel, there's no reason you should be able to patent "a system of menus with buttons that accomplish word processor functions". The whole thing is entirely analogous to physical controls on a typewriter anyway. There isn't really anything in a word processor interface that should be patentable, and the actual code itself is copyrightable. The trivial design aspects aren't, but this gets back to the problem where one company shouldn't be able to have a monopoly on word processors for a couple decades.

# ¿ Dec 9, 2013 18:24

Paul MaudDib: May 3, 2006; TEAM NVIDIA:
FORUM POLICE

Kalman posted:

What part of "the interfaces aren't copyrightable" did you miss in that post?

Calm down. The design of the interface is not patentable, the interface itself (the source code and binary) most certainly is copyrightable. Please cite the case where Lotus copied/pasted WordPerfect's source code and got away with it.

I haven't seen a word processor interface yet whose design is sufficiently novel to really warrant a patent. They are all basically some variation on classic typewriter interfaces, with modifications to increase control (more typefaces, etc) and adapt to existing user interface paradigms (drop down boxes instead of physical sliders, menus and panels of related options grouped together, etc). I view the issue of patents for such things to be totally frivolous, and it's actively and seriously detrimental to regular activity in the software field.

Paul MaudDib fucked around with this message at 19:01 on Dec 9, 2013

# ¿ Dec 9, 2013 18:49

Paul MaudDib: May 3, 2006; TEAM NVIDIA:
FORUM POLICE

Space Gopher posted:

I'm pretty sure that Kalman was talking about Lotus v. Borland, where Borland's Quattro Pro spreadsheet copied the Lotus 1-2-3 menu interface and keyboard commands character-for-character. Lotus still lost the case, but it wasn't the "I hereby claim all implementations of x software using current standard UI conventions" that you seem to think it was. There was a legitimate argument there.

So Kalman's implication is that keyboard combinations and menus should be patentable? :psyduck:

I really think that goes to show just how far up its own rear end the entire patent system is.

Space Gopher posted:

Why do you feel that user interfaces are inherently non-novel? I don't think anyone will argue with the fact that the current system is broken, but there are a lot of possible ways to implement a given UI to perform certain tasks. Human/computer interaction is a big field of study in corporate and academic environments for just this reason. Some UI implementations are obviously better than others (you probably don't want the save function - which, incidentally, a typewriter can't do - to be buried four menu layers deep) and some of them are non-obviously better. And, there are some significantly different user interfaces which haven't received a lot of positive attention but are nevertheless novel and patentable - take, for instance, an engine that analyzes a user's writing in real-time and pops up options relevant to what the system predicts the user might want to do (Microsoft did this; you probably know the technology as "that drat paperclip"). Unless you take a hardline "no inventions should be patentable, ever" stance, it's hard to see how user interface elements should never be patentable.

I'll change your example slightly - instead of popping up Clippy, how about predictive typing? Sure, a new engine that produces better predictive typing should be eligible for patent. But it's a novel type predictive engine - we haven't described a novel user interface paradigm with that. It's going to slap the text ahead of the cursor as you type just like every other type prediction system ever. If you can come up with a novel way to display and interact with the suggested type, that would be a UI patent.

I don't think that user interface paradigms should never be patentable, there are certainly some innovations possible there, but they certainly appear to be the vast minority of the UI patents I've seen around. Given that the balance tips heavily towards "frivolous", and every patent causes fairly large losses in terms of economic activity, I'm perfectly OK with either a hardline "no user interface patents ever" or a strong presumption against issuing them.

When issued, they should be for very short terms - something on the order of 3 years non-renewable. Software development cycles are just too fast compared to the 1700s. Whoever said you didn't buy a new cotton gin every year hit the nub of the situation - a patent might have covered 3 development cycles or so at the time, and now it's 20-40. Companies shouldn't be allowed sole use of a basic UI paradigm for the entire lifespan of a device's existence, and that's what current law provides. Copyright will still protect their actual work, but the concepts need to be free for competition within relatively short spans.

Paul MaudDib fucked around with this message at 19:32 on Dec 9, 2013

# ¿ Dec 9, 2013 19:17

Paul MaudDib: May 3, 2006; TEAM NVIDIA:
FORUM POLICE

Kalman posted:

I am not going to gently caress around reading the spec to determine what a calling hierarchy is, but it isn't half as broad as you make it out to be.

A calling hierarchy is a series of functions that call each other. So the operating system calls "main()" and then main calls "printf("hello world");", that forms a hierarchy.

In other words, it's a fancy way to make you think they're saying something other than "in an algorithm running on a computer", because that describes every single computer program in existence except for programs that are hyper-trivial (moreso than even hello world).

Any computer that has an operating system also has a calling hierarchy, plus every computer that has functions/subroutines, so we're talking about something like "every computer since UNIVAC" in that line. The patent clerks got played for a chump.

Paul MaudDib fucked around with this message at 01:22 on Dec 10, 2013

# ¿ Dec 10, 2013 00:59

Paul MaudDib: May 3, 2006; TEAM NVIDIA:
FORUM POLICE

Kalman posted:

Except that that isn't true for this patent. Which is why I said you'd have to read the spec.

Instead, a calling hierarchy is defined in the spec by "A calling hierarchy indicates the routines (e.g. functions, methods) that have been invoked by or on behalf of a principal (e.g. thread, process) but have not been exited." So the definition you've proposed doesn't apply and we use that one instead.

Please stop while you're ahead. You don't know the faintest thing about the thing you're talking about. What you have described is the exact same thing I did. The function tree does describe functions that are active but have not exited yet.

You can find a synopsis here:

Note that they are not a part of an Integrated Development Environment, they are an essential aspect of how all modern computers work. If you want, I will go into detail, but the fact is that the thing that specification describes is a trivial aspect of how every program works on a computer that is more modern than 1950 or so. In fact it would be exceedingly difficult and inefficient to construct a program that does not satisfy that criterion.

e: I like this representation better.

Paul MaudDib fucked around with this message at 01:46 on Dec 10, 2013

# ¿ Dec 10, 2013 01:33

Paul MaudDib: May 3, 2006; TEAM NVIDIA:
FORUM POLICE

Kalman posted:

The patent still doesn't cover a prospective analysis - it only applies to doing permission analysis by tracing back through the permissions already granted to the hierarchical functions above the calling function. Calling a new function which says "what permissions are associated with function X?" wouldn't infringe.

Right, and let me elaborate on that too.

What it actually means by "first association" is almost certainly "first-class function", wherein pointers to functions are treated as an allowable data type. As a simple example, you could have a list of data which takes a pointer to a function. Then you could use pointers to two different functions to cause different actions on different data types.

Or, for example, you could store that function hierarchy as a stack of pointers to the current instruction for that function in memory - in other words, when Program() wants to call HelloWorld, it pushes the last address it was at onto the stack, so that when the printf() function is done, the computer knows where to go back to. In other words the function tree would look like this inside the printf function:

code:

void main()
{
 printf("hello world");
}

code:

Stack:
OS [0xAddrInsideOperatingSystem]
main [0x1]

The significance here is that you could declare certain functions to be "off limits" to unprivileged code. This is commonly used to enforce boundaries between system ("kernel") code and user code, where you want to deny user code the ability to do something, but need the kernel to be able to do it. The "protection domain" in a simple model like this is binary - the system is either operating in kernel mode or user mode - but other domains commonly exist, such as functions marked with a scope of "private" inside a program.

I think pretty much all systems do a "prospective analysis" here - you just don't complete the call to SystemInternalFunction() if the CPU isn't operating in supervisor mode.

quote:

In computer terms, supervisor mode is a hardware-mediated flag which can be changed by code running in system-level software. System-level tasks or threads will have this flag set while they are running, whereas user-space applications will not. This flag determines whether it would be possible to execute machine code operations such as modifying registers for various descriptor tables, or performing operations such as disabling interrupts. The idea of having two different modes to operate in comes from �with more control comes more responsibility� � a program in supervisor mode is trusted never to fail, since a failure may cause the whole computer system to crash.

Supervisor mode is �An execution mode on some processors which enables execution of all instructions, including privileged instructions. It may also give access to a different address space, to memory management hardware and to other peripherals. This is the mode in which the operating system usually runs.� [8]

Again, this is a trivial aspect of computers that first became common in the 1970s when multitasking OS's became a thing. You would have to go out of your way to avoid that too.

Paul MaudDib fucked around with this message at 02:04 on Dec 10, 2013

# ¿ Dec 10, 2013 01:52

Paul MaudDib: May 3, 2006; TEAM NVIDIA:
FORUM POLICE

Kalman posted:

Just to correct you, the reason it uses "first association" instead of just association is either because later on (in dependent claims) it talks about a different association or because the examiner issued an antecedent basis rejection saying that association could be confused with the association between permissions and routines in a calling hierarchy.

It almost certainly has nothing to do with a first order function and would not be interpreted that way in practice.

So what exactly is the restriction, if it's not "these set of functions should be allowed to call RestrictedFunction(), others should not"?

quote:

"in response to detecting the request, determining whether said action is authorized based on permissions associated with a plurality of routines in a calling hierarchy associated with said principal, wherein said permissions are associated with said plurality of routines based on a first association between protection domains and permissions."

That exists within every operating system, pretty much. Certainly anything based on the x86 architecture or a "ring" architecture in general, so pretty much everything since the 70s. Certain functions are inside the domain (have supervisor mode enabled) and certain functions are outside the domain (have supervisor mode disabled) and the domain denotes permission to do something. Or again, the private function access domain (only functions inside a class should call), or users who have authenticated themselves (which is the most trivial interpretation, and I believe the one that actually applied here).

The permissions don't have to be based on the routines, only associated with them.

Paul MaudDib fucked around with this message at 02:23 on Dec 10, 2013

# ¿ Dec 10, 2013 02:03

Paul MaudDib: May 3, 2006; TEAM NVIDIA:
FORUM POLICE

Kalman posted:

My guess, based on the description of the file history up thread, is that the applicant specified in more detail what a protection domain is such that rings don't qualify and that's why it got allowed. The spec at least suggests that rings wouldn't qualify since it says protection domains are associations between the source of code and permission, rather than between the executing thread and permission. Since a currently executing thread might be executing unprivileged but the code ultimately was provided from a privileged source, in theory this would automatically elevate privilege since the code is privileged even though the thread isn't. But like I said, I would want to read the file history to really hazard a guess on protection domains.

Actually it doesn't say the permissions are based on the function hierarchy, it just just says the permissions are associated with a plurality of routines. You don't want the user to have to reauthenticate every time the server makes a function call - you want one user's permissions to percolate through the function calls until you're done handling the whole request. In other words, you need those permissions to associate to a plurality of routines.

quote:

"in response to detecting the request, determining whether said action is authorized based on permissions associated with a plurality of routines in a calling hierarchy associated with said principal, wherein said permissions are associated with said plurality of routines based on a first association between protection domains and permissions."

It's still not really a novel method of determining if someone's authenticated, it's basically "determine if permissions exist (on a computer)". Which is done all the time on every level from basic kernel calls to web servers.

Paul MaudDib fucked around with this message at 02:35 on Dec 10, 2013

# ¿ Dec 10, 2013 02:27

Paul MaudDib: May 3, 2006; TEAM NVIDIA:
FORUM POLICE

Also patented: streaming videos that offer a view similar to a live classroom experience.

quote:

�...instructor at the head of the classroom with live-participants arranged between the instructor and the camera with a direct line of sight between the camera and the instructor allowing for the viewer participant to have unobstructed views while simultaneously allowing for the viewer participant to have live participants in the periphery, as if the viewer was attending a live class."

quote:

As you might remember from a couple months back, YogaGlo applied for two really similar patents. Initially, both were rejected by the U.S. Patent and Trademark Office (USPTO) based upon the existence of �prior art��which is essentially evidence that the thing a patent is filed for already existed before said patent was filed. On October 7th, however, YogaGlo amended one of its previously rejected patent applications, adding the following caveat:That the camera used to record online yoga classes must "provide a participatory view [which means �a view observed by a participant in the rear of the class,� according to YogaGlo] from a height of about three feet."

https://yogainternational.com/article/view/yogaglo-patent-issued

Here's the claim in full:

quote:

I claim:

1. A system for automatically producing a video representation of a yoga class configured so a remote viewer enjoys the experience of being in a real yoga class, the system comprising: a studio having a front area and a rear area; an instructor position located in the front area and facing the rear area; a line of sight corridor disposed between the rear area and the instructor; a plurality of students at student positions, facing the instructor position, distributed across the studio between the instructor position and the image capturing device wherein the student positions do not impinge upon the corridor; an image capturing device for capturing video located in the rear area disposed to provide a participatory view from a height of about three feet by capturing through the line of sight corridor, an unobstructed video of the instructor in the instructor position including images of the students disposed along the sides of the line of sight corridor; and sound capture equipment to capture at least audio of the instructions given by the instructor disposed in the instructor position to the students disposed in the student positions.

2. The system according to claim 1, further comprising remote control systems whereby the image capture device pans and zooms.

3. The system according to claim 1, further comprising means for recording the video.

4. The system according to claim 1, further comprising means for streaming the video via the Internet.

5. The system according to claim 1, further comprising means for adding music to the audio.

6. The system according to claim 1, wherein the instructor position includes an instructor's mat and each student position includes a student mat with each student mat disposed normal to the instructor's mat.

7. The system according to claim 1, wherein the corridor becomes wider near the image capture device.

8. A method for automatically producing a video representation of a yoga class configured so a remote viewer enjoys the experience of being in a real yoga class, comprising the steps of: providing a studio having a front area and a rear area; placing an instructor position at the front area of the studio, the instructor position disposed to face the rear area; locating an image capturing device for capturing video in the rear area at a height of about three feet to allow for a participatory view disposed to provide a participatory view from a height of about three feet by capturing an image of the instructor in the instructor position; laying out a line of sight corridor between the image capturing device and the instructor; distributing a plurality of student positions, facing the instructor position, across the studio between the instructor position and the image capturing device wherein the student positions are distributed so as not to impinge upon the corridor; capturing at least audio of the instructions given by the instructor disposed in the instructor position; and capturing a video of the instructor through the line of sight corridor, which video includes images of the students disposed along the sides of the line of sight corridor.

9. The method according to claim 8, further comprising a step of remotely controlling the image capture device.

10. The method according to claim 8, further comprising the step of recording the video.

11. The method according to claim 8, further comprising the step of streaming the video via the Internet.

12. The method according to claim 8, further comprising the step of adding music to the audio.

13. The method according to claim 8, further comprising a step of placing an instructor's mat in the instructor position and a student mat in each student position with each student mat disposed normal to the instructor's mat.

14. The method according to claim 8, further comprising a step of increasing the width of the corridor near the image capture device.

So it's pretty much just "making a yoga video in a studio, with a camera, with good production values, and then you record it or stream it onto the internet or something I guess, now give me money".

This thing, this image right here? That's a novel innovation that's advancing the arts and sciences and is worthy of a monopoly on the concept for decades.

There just seems to be this huge, gaping disconnect between the theoretical and the patents that actually get issued in real life. I don't have a problem with RSA being patented, I don't have a problem with LZW being patented, I accept most of the arguments in theory (except for design patents), but then in real life the patent office is rubberstamping patents for rounded corners and using a camera to film your yoga workout video. That seems like the real issue to me. You'd think for $1000 an hour (or whatever ridiculous number Kalman threw out there) the patent office could hire some people who have a reasonable understanding of practicing their art.

But this one isn't even a technical thing where someone was trying to confuzzle the patent clerk. Have these clerks never seen a workout video before? :psyduck:

Paul MaudDib fucked around with this message at 23:14 on Dec 12, 2013

# ¿ Dec 12, 2013 22:59

Paul MaudDib: May 3, 2006; TEAM NVIDIA:
FORUM POLICE

Kalman posted:

Imagine that the current American political system controls all research funding (in practice, as private funding is going to be significantly reduced if everything can be legally copied and if they can just rely on public results to base products on.). Are you still comfortable with a direct grant system?

I have no idea what you're talking about, the NSF grant system is generally viewed to be working excellently, with the largest problem being not enough money to give out.

# ¿ Dec 14, 2013 17:18

Paul MaudDib: May 3, 2006; TEAM NVIDIA:
FORUM POLICE

Kalman posted:

It's almost like the only ones that make the news are the ones that upset people or are borderline. Do you think that there might be a selection bias operating?

Not every patent needs to be frivolous to cause serious impacts on innovation. In terms of lawsuits, patent-assertion entities make up 60% of patent lawsuits filed in 2013. Not every patent from a PAE is frivolous, but the number overall is probably fairly high.

quote:

Patent-assertion entities, also known as P.A.E.�s, typically have no operations other than collecting royalties on patents. They accounted for more than 60 percent of the roughly 4,000 patent lawsuits filed last year, up from 29 percent two years earlier.

�There are companies that are engaged in spurious lawsuits, seeking settlements that are less than the cost of litigation. But not us,� said Scott Burt, chief intellectual property officer at Mosaid Technologies, a Canadian company that is nonetheless considered one of the largest patent trolls. �We are a patent-licensing company.�

The types of lawsuits that have been filed or threatened sometimes exceed comprehension. One such suit recently threatened thousands of companies with liability for damages on charges they violated a patent by hooking up a document scanner to a computer network and sending a scanned file by e-mail to an employee.

http://www.nytimes.com/2013/06/20/business/ftc-is-said-to-plan-inquiry-of-frivolous-patent-lawsuits.html

The chance of a patent being revoked once it's opposed is roughly 50%, and in most of the rest of the cases it's significantly narrowed. Obviously frivolous patents are more likely to be opposed than high-quality patents, but it trends towards the same rough numbers. My gut instinct here is that the percentage of frivolous patents is 10-30%, and many of the patents that are not outright frivolous are probably written overbroadly.

It's probably higher than average in software and technology patents because it's easy to use some fancy language to obfuscate "in a program on a computer".

Paul MaudDib fucked around with this message at 03:23 on Dec 17, 2013

# ¿ Dec 17, 2013 03:07

Paul MaudDib: May 3, 2006; TEAM NVIDIA:
FORUM POLICE

Shifty Pony posted:

It is context specific and depends on what a person of ordinary skill would conclude is "about" the setting claimed in light of the specification. In this case this is in the disclosure: "In embodiments of the disclosure, terms such as "about," "approximately," and "substantially" can include traditional rounding according to significant figures of the numerical value."

Someone from the dorkroom would be better able to answer but I think f-numbers and ISO values are pretty standardized. So some hypothetical crazy lens with a setting at an f-stop of 5.6452 would be "about" 5.6, but one set at f-4 or f-8 would not be.

Shifty Pony posted:

I know the values are arrived at in a set way I just didn't know or feel like digging into the standard to find out if vales such as 322 or the like are allowed or it just rounded to the nearest 10.

WhiskeyJuvenile posted:

322 rounded to the nearest significant figure is 320?

F-numbers and ISO are relatively standardized. Both of them are logarithmic-type scales. A "unit" on these scales represents a doubling or halving of the intensity of the light (f-stop) or the sensitivity of the sensor (ISO), and is colloquially referred to as a "stop" ("I opened up the aperture one stop and turned the ISO down one stop").

The usual scale for ISO uses a base number of ISO 100. So then you have 200, 400, 800, 1600, etc, or 50, 25, etc. Because it's a log scale, half or third stop changes have strange units. ISO320 is ISO200 + 2/3 stop or ISO400 - 1/3 stop. I'd say that's probably a "normal" setting for studio work, where you're using lots of artificial illumination, except in one respect. Most cameras offer an ISO setting that is adjusted in half or third stops like that, but it's not the most efficient way to use the camera.

Making a digital image is a three step process. First, the camera reads out the image from the CCD sensor. The CCD signal goes through an analog amplifier, which has a selectable gain. The higher the gain, the more sensitive the sensor is (higher ISO) but the noise in the signal is also amplified, so it gets grainier and you start losing color precision ("bits of color depth"). Second, this signal is converted from electrical values into digital values by an analog-to-digital converter. Most cameras are based on what's called a Bayer filter - the pixels are covered by a mosaic grid of R,G,B filters which record the intensity of light for one color at one pixel as a 24-bit B+W value. Finally, the image is then read into an onboard image processor, which takes the monochrome values from adjacent pixels (representing one R,G,B channel each) and convolves them into 8-bit R,G,B values at each pixel. However it can only work with what it's given, if there's noise in the signal read into the processor, the processor can try to fix it but it can't eliminate it after the fact.

The thing is that the analog amplifier only works in whole stops from the base ISO setting (usually 100). So the way the camera gets things like ISO 320 is to turn the analog image amplifier up to ISO 400 and then subtract 1/3 of the intensity from all three channels. You're increasing the amount of noise in the image and throwing away some of the information in the image. Smart photographers only use the whole ISO stops, so it's interesting that they picked 320 instead of 200 or 400.

However, the difference in practical terms is zero, particularly in a studio where you can easily adjust the intensity of your lighting. Either 200 or 400 would be "about ISO 320" to a photographer, which is why I suspect they picked that - they can cover both bases, which is dead center of where you would put that setting for studio work. Either 200 or 400 is about right. 800 is pretty high, 100 is a bit on the low side for studio work.

The f-stop scale does not have nice even units, but it works in the same way. The scale goes like: f/2, f/2.8, f/4, f/5.6, f/8, f/11, f/16, etc, those are the "whole units" typically used today (not always historically). There's three competing concerns to selecting an f-stop. The first is the amount of light - the lower the f-number, the brighter the image projected onto the sensor. This isn't really a problem under studio lighting. Second, depth of field - the lower the number, the less that can be in focus at one time. This is a big concern with product photography and other macro-like work - when you get up close, the depth of field gets really shallow, so you don't want to use low f-stop numbers. Finally, the diffraction limit - stopping down makes the lens sharper but it increases the amount of interaction between beams of light, causing increased additive/cancellation effects which reduce the resolution past a certain point. That link is the best technical explanation I've found, but the numbers are out of date, modern sensors have more megapixels and run into that limit quicker.

f/5.6 happens to be the exact number where the diffraction limit kicks in with modern sensors. No photographer would use an aperture lower than f/4 or so for product photography - there's just too little in focus up close at f/2.8 and some macro lenses don't even go that low. And at f/11 and higher you're really starting to lose a measurable amount of resolution. That means that there's an ideal range of about f/4-f/8 for this kind of stuff, and one f-stop difference is "about f/5.6" to a photographer. So again they've staked out the exact middle of the reasonable working range with those ambiguous "about" claims.

Paul MaudDib fucked around with this message at 22:01 on May 9, 2014

# ¿ May 9, 2014 21:33

Paul MaudDib: May 3, 2006; TEAM NVIDIA:
FORUM POLICE

Kalman posted:

Given that a camera actually operates differently at ISO 320 versus ISO 200 and 400, decent argument that they don't fall within "about" and certainly don't fall into equivalents. Same for f-stop. Especially since, now that I've read the disclosure, they basically say "about means normal rounding."

The claims also lack written description to support interpretations of "about" to mean anything other than the value plus or minus the normal tolerance you'd expect in a optical electromechanical system, which means that constructions of the claims that included those interpretations are disfavored.

Basically, it's not something they could effectively use against anyone.

That's kind of like saying that a claim that involves driving a car at "about 47 miles per hour" wouldn't cover driving a car at 45 miles an hour or 50 miles per hour because at those settings the high-speed/low-speed jets in the carb are injecting different amounts of fuel and that means the car is operating differently. :iiaca:

That's a tiny nuanced internal detail that most photographers probably don't even know about. You usually have to dig through menus to turn half-stop and third-stop ISO settings off, if it can be done at all. The camera always adjusts the image in the image processor to at least some degree anyway, so this kind of nuance is basically subsumed by the normal process of how the camera converts the RAW file (the digitized signal) to a image (eg JPG).

To go back to the car example above, the precise details of how the carb operates would be significant if we were dealing with a patent for a new carb, but it's insignificant to the act of driving the car, which is what is claimed. This isn't a patent on a novel method of capturing a photo, it's a patent on taking a photo composed like this, and the precise combination of manipulations the image processor applies in the course of taking the photo is irrelevant.

Let's say we take the patent on photo composition and we use all their listed settings as described in the patent, except instead we use a camera with a Foveon sensor (which lack the Bayer filters and thus "operate differently" under your metric). I'm 100% sure the patent would stand in that case. The same as if the car in the example above was fuel injected instead of carburated. In fact, every camera implements the above digitization process slightly differently, so unless they include a laundry list of every camera ever made then the process is guaranteed to be "different" in the sense you're using. Analogously, should patents involving cars have to include a laundry list of specific models, because car A has jets that are 2mm in diameter and car B has 2.4mm in diameter, and car C has differently sized wheels, and so on? That's clearly absurd. Sometimes the process varies down to the firmware version installed on the camera, and the mess would get even deeper if we included third-party tools that can process the RAW off-camera, like Aperture or Lightroom.

The precise technical implementation of how control signals are implemented is just not relevant when the claim is a particular combination of control signals. To the end user, ISO 320 is ISO 320 no matter which digital camera is in their hands.

Paul MaudDib fucked around with this message at 23:09 on May 9, 2014

# ¿ May 9, 2014 22:18

Paul MaudDib: May 3, 2006; TEAM NVIDIA:
FORUM POLICE

Kalman posted:

Yes, ISO 320 is ISO 320, but ISO 200 is not equivalent to "about ISO 320" because it doesn't pass the function-way-result test; arguably a different result, but you've just explained that ISO 320 is achieved in a different way from ISO 200 in that one is directly recorded while the other is achieved via processing.

ISO 200 is also achieved by processing. You're looking at a picture in RGB color, not electrical charge levels in monochrome. If you're going to make that distinction, the distinction has to be that the processing is different, not that no processing is done at ISO 200, because it is. And that really opens up a huge can of worms because no two firmware versions, let alone two different cameras, do that processing quite the same. No two workflows do that processing quite the same either - you're saying that literally every setting from the camera to the gamma and color profile on the user's monitor would have to be the same and that's impossible and absurd.

The better point to dismiss an infringement suit on would be the very specific claims of lighting and camera, but I have a sneaking suspicion they opened up a basic photography book and tried to patent the layout pictures. Sounds like more or less every studio softbox setup.

Those top and side panels are all there to be used to diffuse a light source, and the same arrangement is commonly scaled up for larger stuff. That's pretty much what the claim is regarding the composition.

e: here's another one that is also pretty close

https://www.flickr.com/photos/rapturedmind/5446909504/in/photostream/

That arrangement is really obvious and common and pretty much every photographer has played with a DIY softbox like the one above.

Paul MaudDib fucked around with this message at 23:25 on May 9, 2014

# ¿ May 9, 2014 23:09

Paul MaudDib: May 3, 2006; TEAM NVIDIA:
FORUM POLICE

Kalman posted:

The picture you posted might meet the "camera" claim element, but that's probably about it.

Camera, background, light placement, elevated platform, etc. It's pretty close really.

e: Some of that guy's other stuff is actually even closer. This actually may be a dead-on match for the patent. The lens listed on the photo is just for taking pictures of the setup - you'd use a longer lens to actually photograph the object.

https://www.flickr.com/photos/rapturedmind/5536144990/

quote:

1. A studio arrangement, comprising:
a background comprising a white cyclorama;
a front light source positioned in a longitudinal axis intersecting the background, the longitudinal axis further being substantially perpendicular to a surface of the white cyclorama;
an image capture position located between the background and the front light source in the longitudinal axis, the image capture position comprising at least one image capture device equipped with an eighty-five millimeter lens, the at least one image capture device further configured with an ISO setting of about three hundred twenty and an f-stop value of about 5.6;
an elevated platform positioned between the image capture position and the background in the longitudinal axis, the front light source being directed toward a subject on the elevated platform;
a first rear light source aimed at the background and positioned between the elevated platform and the background in the longitudinal axis, the first rear light source positioned below a top surface of the elevated platform and oriented at an upward angle relative to a floor level;
a second rear light source aimed at the background and positioned between the elevated platform and the background in the longitudinal axis, the second rear light source positioned above the top surface of the elevated platform and oriented at a downward angle relative to the floor level;
a third rear light source aimed at the background and positioned in a lateral axis intersecting the elevated platform and being substantially perpendicular to the longitudinal axis, the third rear light source further positioned adjacent to a side of the elevated platform; and
a fourth rear light source aimed at the background and positioned in the lateral axis adjacent to an opposing side of the elevated platform relative to the third rear light source; wherein a top surface of the elevated platform reflects light emanating from the background such that the elevated platform appears white and a rear edge of the elevated platform is substantially imperceptible to the image capture device; and
the first rear light source, the second rear light source, the third rear light source, and the fourth rear light source comprise a combined intensity greater than the front light source according to about a 10:3 ratio.

We have

White cyclorama
Camera
Elevated platform
Top down light
First rear light aimed at background
Second and third rear lights behind the camera aimed down and toward the background
Front light (camera flash that triggers the optical strobe slaves)

quote:

SB-80dx (Diffusor - 1/8) into 28" Westcott Apollo Softbox from the top as mainlight, SB-80dx (24mm - 1/32) bare and flagged towards camera and background as rim from the right, SB-80dx (24mm - 1/32) bare and flagged towards camera and background as rim from the left, SB-80dx (14mm - 1/8) bare on background with 2 green gels, camera and flashes tiggered via PocketWizard Plus II and optical trigger.

Yep that's the setup all right. I guess he may have put his tripods in slightly different places but that's the same lighting arrangement.

Paul MaudDib fucked around with this message at 23:41 on May 9, 2014

# ¿ May 9, 2014 23:26

Paul MaudDib: May 3, 2006; TEAM NVIDIA:
FORUM POLICE

hobbesmaster posted:

Patent law is not horse shoes.

Yeah, the whole patent system is incredibly navel-gazing and that's really what's wrong with the whole thing. Frankly I really don't think a patent on a lighting arragement would possibly fail based on the fact that two cameras would capture an image slightly differently given the same control settings, but that's the kind of poo poo that patent lawyers spend their days arguing.

I just wish the whole thing weren't so incredibly disruptive to the actual process of innovation and advancing the arts and sciences. I look at the way things like software patents have been weaponized and think that we'd be better off with substantially loosened intellectual property rights. Right now it mostly seems to be big companies filing mountains of patent applications for really obvious and uninteresting things and when a few inevitably slide through, no matter what the prior art (slide to unlock, etc), it becomes a weapon they can use to file SLAPP suits and run their competitors out of business.

Much like the insane British legal system pre-reform, it's really not a good idea to have to be forced to pay a bunch of money to get a legal wizard to spin some crap that Amazon uses a Nikon D40 and I use a Canon 40D and that means their lighting arrangement patent no longer applies, just to use basic object photography arrangements that have been floating around for 50 years. The patent system is that level of broken. Patents like that shouldn't be granted but neither should arguing brands and models of image capture devices and lenses and literally firmware versions.

Paul MaudDib fucked around with this message at 09:28 on May 10, 2014

# ¿ May 9, 2014 23:49

Paul MaudDib: May 3, 2006; TEAM NVIDIA:
FORUM POLICE

karthun posted:

Paul MaudDib, why did you just list claim 1 of the patent instead of the entirety of claims 1-27?

I don't think you realize just how specific this patent is.

Because that's what someone listed on the previous page. Actually Claim 2 is pretty much a dead ringer for that last setup I posted.

quote:

2. A studio arrangement, comprising:
a background comprising a cyclorama;
a front light source positioned in a longitudinal axis intersecting the background, the longitudinal axis further being substantially perpendicular to a surface of the background;
an image capture position located between the background and the front light source in the longitudinal axis;
an elevated platform positioned at a first distance from the elevated platform and between the image capture position and the background along the longitudinal axis, the front light source being directed toward the elevated platform;
a first at least one rear light source positioned between the elevated platform and the background, the at least one rear light source directed towards the background;
a second at least one rear light source positioned in a lateral axis intersecting the elevated platform and being substantially perpendicular to the longitudinal axis;
at least one light shield positioned between the second at least one rear light source and the elevated platform, the at least one light shield configured to shield the elevated platform from light emitted directly from the second at least one rear light source from lighting an upper surface of the elevated platform; and wherein
a top surface of the elevated platform reflects light emanating from the background such that the elevated platform appears a substantially similar color as the background and a rear edge of the elevated platform is substantially imperceptible to an image capture device positioned at the image capture position.

A "light shield" is commonly referred to as a flag. I'm pretty sure most any studio softbox setup can be covered by those claims. It's really fairly obvious how to go about taking a photo of an object against a white background without shadows, there's only so many ways to do it that don't involve having a stand in the photo or shooting a flash straight into your lens or whatnot.

This one, by the way, doesn't even specify what the image capture device is.

Paul MaudDib fucked around with this message at 00:15 on May 10, 2014

# ¿ May 10, 2014 00:01

Adbot: ADBOT LOVES YOU

# ¿ Apr 20, 2024 02:30

Paul MaudDib: May 3, 2006; TEAM NVIDIA:
FORUM POLICE

karthun posted:

Thats just claim two, now what about the rest of the claims? Please show me how "any studio softbox setup can be covered by those claims." I cant wait until you realize that you need to infringe on ALL of the claims in order to infringe on the patent.

That's not how claims work. Claim 2 is independent, the rest of the claims after that are dependent.

quote:

There are two basic types of claims:

the independent claims, which stand on their own, and
the dependent claims, which depend on a single claim or on several claims and generally express particular embodiments as fall-back positions.

Some claims are "stand alone" independent claims, not referring back to any earlier claim. Other are dependent claims that do refer back to an earlier claim, take on all of the limitations of that claim and then add additional restrictions. e.g. "The handle of claim 2 where it is hinged." Each dependent claim is, by law, more narrow than the independent claim upon which it depends. Although this results in narrowed coverage than provided by the independent claim which it depends from, it is additional coverage and there are many advantages to the patent applicant in submitting and obtaining a full suite of dependent claims:

http://en.wikipedia.org/wiki/Claim_%28patent%29

quote:

For infringement to exist, each element (or its equivalent under certain circumstances) of the claim must be present in the accused device or method. Although a patent may have many claims, only one claim need be infringed for the patent to be infringed.

http://books.google.com/books?id=XqDniFJ5CWQC&lpg=PA258&ots=75CU5oCbkG&pg=PA258#v=onepage

Glad I could clear that up for you.

Paul MaudDib fucked around with this message at 00:52 on May 10, 2014

# ¿ May 10, 2014 00:27

The Something Awful Forums > Discussion > Debate & Discussion > Patent reform