Elan Microelectronics Turns to U.S. International Trade Commission in Multi-Touch Patent Dispute With Apple

[BobVB]

You seem to be confusing two separate issues. In this post your argument appears to be that the Elan patent is not infringed by Apple. Apple is all digital, or whatever it is you are trying to say.

Ok, fine. I don know what it is Apple actually does, and infringement is an issue that will be determined freshly by this court and on which the previous case has no bearing


But that is a completely different issue than whether or not the patent is VALID.

Because our discussion has evolved and how the patent is presented.

If the patent is presented as being a fundamental 'multitouch' patent that basically says that 'anyone who touches a screen with more than one finger has to pay us money', to that view I say the patent is invalid - that isn't what it says, such a basic idea shouldn't be patentable in the first place.

To what the patent actually does, yes it is valid. They saw a patent for a capacitive touchpad that was patented for a single contact usage. They saw that they could take an x and y profile trace for its existing leads and by making a series of precise 'assumptions' about these they could get it to detect multiple touches. Applause, clapping - they figured out a way to get multitouch usability out of a sensor the inventors didn't design to do that. I agree that patent on that sensor is valid. Anyone who does a multitouch application with that type of sensor should pay them money for the duration of the patent.

And again if the difference was merely cosmetic in that today's sensors filled their data matrix with the same data their patent did I would say they'd have a case to say that the patent should stand no matter what type of sensor is used, but it doesn't. Their patent puts the same X value in every column, the same Y value in every row producing a heck of a lot of noise and false values. The intersection that really has a base line value will be filled with a reading because at a different X Y location there is a 'maxima' reading and that value has been carried through out the entire row/column.

This is the technical limitation that their initial complex flow diagram is compensating for in Figure 6-1, where they are scanning the x trace to find differences, and then in Figure 8-1 where it is clear they are now looking to see if the 'xcompute' OR the 'ycompute' found indications of multiple fingers.

Apple doesn't do that, no one using the new grid transducers does this. They don't have an x trace or y trace to even scan for meaning all that 'matrix math' is useless and pertinent ONLY for a sensor that does produce that kind of 1 dimensional perspective on the axis of a 2 dimensional space.

So I guess as I've learned I am agreeing with you - this patent is very specific about what it does, it works with a data set that only a very specific technology would even produce and current technology has nothing to do with it other than it also is concerned with more than one contact or action at a time.

If the patent is presented as what it does, what it talks abouts, and how it works then yes it is valid, but then how could Elan even think that Apple was infringing on it?

If the patent is presented as some generic 'multitouch' technology that is NOT what it does in gross or fine examination and such a claim must be considered invalid.

And if you look at Apple's patent they are basically doing the same thing Elan did for this type of 'axis profile' sensor with the new grid ones. Both require working on specific data sets produced by different sensor types - and neither is about the data set types of the other.

 

[cmaier]

Because our discussion has evolved and how the patent is presented.

If the patent is presented as being a fundamental 'multitouch' patent that basically says that 'anyone who touches a screen with more than one finger has to pay us money', to that view I say the patent is invalid - that isn't what it says, such a basic idea shouldn't be patentable in the first place.

To what the patent actually does, yes it is valid. They saw a patent for a capacitive touchpad that was patented for a single contact usage. They saw that they could take an x and y profile trace for its existing leads and by making a series of precise 'assumptions' about these they could get it to detect multiple touches. Applause, clapping - they figured out a way to get multitouch usability out of a sensor the inventors didn't design to do that. I agree that patent on that sensor is valid. Anyone who does a multitouch application with that type of sensor should pay them money for the duration of the patent.

And again if the difference was merely cosmetic in that today's sensors filled their data matrix with the same data their patent did I would say they'd have a case to say that the patent should stand no matter what type of sensor is used, but it doesn't. Their patent puts the same X value in every column, the same Y value in every row producing a heck of a lot of noise and false values. The intersection that really has a base line value will be filled with a reading because at a different X Y location there is a 'maxima' reading and that value has been carried through out the entire row/column.

This is the technical limitation that their initial complex flow diagram is compensating for in Figure 6-1, where they are scanning the x trace to find differences, and then in Figure 8-1 where it is clear they are now looking to see if the 'xcompute' OR the 'ycompute' found indications of multiple fingers.

Apple doesn't do that, no one using the new grid transducers does this. They don't have an x trace or y trace to even scan for meaning all that 'matrix math' is useless and pertinent ONLY for a sensor that does produce that kind of 1 dimensional perspective on the axis of a 2 dimensional space.

So I guess as I've learned I am agreeing with you - this patent is very specific about what it does, it works with a data set that only a very specific technology would even produce and current technology has nothing to do with it other than it also is concerned with more than one contact or action at a time.

If the patent is presented as what it does, what it talks abouts, and how it works then yes it is valid, but then how could Elan even think that Apple was infringing on it?

If the patent is presented as some generic 'multitouch' technology that is NOT what it does in gross or fine examination and such a claim must be considered invalid.

If our conversation evolved, someone forgot to tell me. In any event, we seem to be in agreement that the patent is a valid patent. I offer no opinion as to whether Apple infringes it. "How the patent is presented" appears to be what lawyers call the "scope" of the patent claims. The first step in any patent lawsuit is what is called "claim construction." This is decided by a judge, not the jury. The process involves the judge deciding what the terms of the claims mean. Hence the scope of the claims ("how the patent is presented") is determined by the judge. Since this is "a matter of law," the claim construction from the Synaptics case is technically binding in this new case (though it is common for a judge to only construe a few terms, meaning there may be a new fight about the meaning of other terms). Some judges revisit claim construction in any event. I have not read the claim construction from the earlier case, so I don't know what has been determined as far as the meaning of the words in the claims.

 

[BobVB]

If our conversation evolved, someone forgot to tell me.

You can't help yourself, can you?


In any event, we seem to be in agreement that the patent is a valid patent.

Hmmmm well that's iffy - depends on how its presented.

I found the court judgement pdf here. Seems it was Synaptic's 'code 2' that was seen as infringing in that it did an approximate of what the Elan patent 'xcompute' example details. From what I see in the Apple patent they actually go out of their way to NOT do that. Hopefully  can clearly demonstrate this to the court and so it will not be infringing on the Elan patent.

From looking at the judgment it seems to be dependent on specifics we are not privy to - does it seem to you its a acknowledging a generic 'multitouch' patent to Elan or is it specific to the actual Synaptic's approach to multitouch ?

Looking at the judgement and the way Synaptic's 'Code 1' was found not infringing I think  will be found the same. Both don't scan for 'maxima & minima' but rather use thresholds -  in particular doesn't look for a maxima at all but just for the boundaries of touch regions, i.e. it its never looking for a maxima. If the previous ruling is precedence then I don't think  has much to worry about.

 

[cmaier]

If our conversation evolved, someone forgot to tell me.

You can't help yourself, can you?


In any event, we seem to be in agreement that the patent is a valid patent.

Hmmmm well that's iffy - depends on how its presented.

I found the court judgement pdf here. Seems it was Synaptic's 'code 2' that was seen as infringing in that it did an approximate of what the Elan patent 'xcompute' example details. From what I see in the Apple patent they actually go out of their way to NOT do that. Hopefully  can clearly demonstrate this to the court and so it will not be infringing on the Elan patent.

From looking at the judgment it seems to be dependent on specifics we are not privy to - does it seem to you its a acknowledging a generic 'multitouch' patent to Elan or is it specific to the actual Synaptic's approach to multitouch ?

Looking at the judgement and the way Synaptic's 'Code 1' was found not infringing I think  will be found the same. Both don't scan for 'maxima & minima' but rather use thresholds -  in particular doesn't look for a maxima at all but just for the boundaries of touch regions, i.e. it its never looking for a maxima. If the previous ruling is precedence then I don't think  has much to worry about.

This is only one of the judge's orders, relating to Synaptics' motions for summary judgment of non-infringement (and not to Elan's motions for infringement).

As you note from the judge's order, he ruled that "Type 1 Code" cannot infringe, but that "Type 2 Code" might. It is clear that subtle differences in how things are calculated and measured make a difference. Again, this supports my point that the claims are not so broad (i.e.: they do not cover so much) as you started out arguing. Again, since I don't know what Apple actually does (their own patent application may or may not reflect what they do in hardware), I can't guess whether they infringe.

 

[BobVB]

This is only one of the judge's orders, relating to Synaptics' motions for summary judgment of non-infringement (and not to Elan's motions for infringement).

As you note from the judge's order, he ruled that "Type 1 Code" cannot infringe, but that "Type 2 Code" might. It is clear that subtle differences in how things are calculated and measured make a difference.

Fortunately the ruling delineates the difference - code 1 never even looked for a maxima, code 2 effectively found one and the excuse of 'but we never use it for anything' didn't cut it with the judge.

IF  is doing what they patented they aren't even finding a maxima - they find the border of each touch region and never even look inside it for data - this puts it in the 'code 1' type of situation which was found non-infringing and definitely not in the 'code 2' kind that was.

Of course if they aren't then that's 'entirely different' but since doing it the way the Elan patent describes would actually be less useful than the way they patented hopefully they are doing what they said.

 

[cmaier]

Fortunately the ruling delineates the difference - code 1 never even looked for a maxima, code 2 effectively found one and the excuse of 'but we never use it for anything' didn't cut it with the judge.

IF  is doing what they patented they aren't even finding a maxima - they find the border of each touch region and never even look inside it for data - this puts it in the 'code 1' type of situation which was found non-infringing and definitely not in the 'code 2' kind that was.

Of course if they aren't then that's 'entirely different' but since doing it the way the Elan patent describes would actually be less useful than the way they patented hopefully they are doing what they said.

Yes. As I think I said about 6.5 pages ago, if Apple is searching for a gradient and stops at that point, it isn't finding the maxima. Much will depend on implementation details, however. Elan's attorneys must meet their burden under Federal Rules of Civil Procedure Rule 11, so they must have at least some reason to believe infringement is more than a remote possibility.

Elan's lawyers are Alston and Bird - interestingly they are in the same building I am in. They're tiny compared to the firm I work for, but they have enough of a reputation that I'm sure they aren't just fishing.

 

[MagnusVonMagnum]

Anyone else here find this thread beyond ANNOYING with these two going at it non-stop? Here's a hint guys. You two will NEVER *EVER* agree and if you think otherwise you're a darn fool. Both your views suck (for sake of argument at least as in they're both annoying the living heck out of me). So just DROP IT and do us all a big favor. Let a judge decide because neither of you get ANY say in the case what-so-ever. Personally, I'm tired of greedy companies all around so I couldn't care less how the courts decide either way.

 

[cmaier]

Anyone else here find this thread beyond ANNOYING with these two going at it non-stop? Here's a hint guys. You two will NEVER *EVER* agree and if you think otherwise you're a darn fool. Both your views suck (for sake of argument at least as in they're both annoying the living heck out of me). So just DROP IT and do us all a big favor. Let a judge decide because neither of you get ANY say in the case what-so-ever. Personally, I'm tired of greedy companies all around so I couldn't care less how the courts decide either way.

No one is forcing you to read the thread, and judging from the PM's in my inbox, people are interested in what I have to say.

 

[Smoother]

Very interesting that this lawsuit was filed from an Tawainese company just a month after Apple filed one against another Taiwanese company

In my opinion all these lawsuits are stupid, unless a company actually steals something that looks and works exactly the same way they should just let it go.
I mean picture if Nokia/Ericsson were American companies and got patents for GSM technology or whatever etc. Where would we be still stuck using Brick phones?
They need to re-do the whole patent system, cause this is getting out of hand and slowing down innovation.

 

[BobVB]

Here's a hint guys. You two will NEVER *EVER* agree and if you think otherwise you're a darn fool.

Ah the voice of unreason. guess you missed we actually have come to agree on many things which wouldn't have happened without discussion.

 

[rolphi]

Very interesting that this lawsuit was filed from an Tawainese company just a month after Apple filed one against another Taiwanese company

In my opinion all these lawsuits are stupid, unless a company actually steals something that looks and works exactly the same way they should just let it go.
I mean picture if Nokia/Ericsson were American companies and got patents for GSM technology or whatever etc. Where would we be still stuck using Brick phones?
They need to re-do the whole patent system, cause this is getting out of hand and slowing down innovation.

Just for fun, I decided to promote myself to judge here, so I might as well render a ruling.

I think the original question was that the patent was not a valid patent because it failed the obviousness test.

BobVB, I'm sorry to say that in reviewing your arguments against the patent, your basic position was, "they didn't deserve a patent because well, duh..." Unfortunately, at that point I would probably have granted a summary judgement for cmaier if he requested one.

Then you did offer the interesting argument in favor for a "put 5 engineers in a room with the problem and existing technology and see what they come up with in one day" test for obviousness which was actually testable and not circular, so I decided to allow that.

cmaier countered with the argument "everything is obvious in hindsight" as well as the fact that this patent was already tested once in court, which I found compelling. So I would have to say that your test for obviousness is too strict and that if you ever do put 5 engineers in a room and they solve a problem which has no evidence of prior art, then you in fact DO have a patentable invention even if it only took one day.

BobVB, as for the question of whether or not Apple is infringing this patent, I would have to say I am genuinely grateful to you for providing the information about how Apple implement their touch sensors especially compared to what was in the patent. And it does not appear that anyone is offering a contrary opinion at this point, so there you go. We'll all have to see how this one plays out.

Since you do still think the patent is "iffy" then I think we all need to hear some logical argument where the patent office made a mistake in granting the patent. "It's obvious because it's obvious" is not a logical argument. If you are trying to come up with a test for obviousness that the patent office should use instead of its current one, I would love to hear that as well.

Too many forum posters drop into conversations like this one and proffer their "opinion" that the current system is actually inhibiting innovation without offering evidence or logical reasoning why that should be so (see the quote above). Their usual remedies are some unspecific variation of throwing away the whole system and starting over. BobVB, if you have some idea on a way for the patent office to test of obviousness that would have preserved the good things the patent office has done over the years but would have prevented the Amazon one-click patent, I think we would all genuinely love to hear it.

And I don't think it would be productive for me to say to which of you I was referring in my previous post, so I'll just keep quiet on that.

 

[BobVB]

Since you do still think the patent is "iffy" then I think we all need to hear some logical argument where the patent office made a mistake in granting the patent. "It's obvious because it's obvious" is not a logical argument. If you are trying to come up with a test for obviousness that the patent office should use instead of its current one, I would love to hear that as well.

Thank you for your response. As I said, my view had evolved. Like many who hear of such claims I initally thought this was as if some generic 'all your multitouch is ours' patent had been granted. I still feel that would be definitely wrong.

But having read the patents, the dependent patents, and the ruling from the judge on the previous case the 'iffy' part for me is more the court's latitude in deciding what the patent means after the fact.

Reading Elan's actual patent if I were a judge I would see it as a patent on how to extract multitouch capability out of a particular type of sensor and be very limited on any rights it extended beyond that. But in the previous suit the judge seems to have magically waved away many parts of the patent that I would consider integral to its basis and focused on the 'finding maxima and minima' concept - you do it you're infringing, you don't you aren't. Maybe you can tell me now he ended up with the view?

Obviously Elan is going to present their case as the patent being as expansive and all inclusive as possible and the other as narrow as I view it to be or even more so but how does the court decide what are the 'essential' parts of the patent, and what are not? That is the 'iffy' part to me since it seems almost arbitrary, particular if the actual human deciding isn't a tech person to begin with.

Again, look at the ruling from the previous case - the patent does figures and then compares the evaluated data to a threshold, the code that was found non infringing took raw data and compared it to a threshold creating a matrix of binary 'above threshold, not above threshold' 1s and 0s. (which is basically what  is now describing in their patent by only looking for the boundaries of touch regions of raw data that has been compared to a threshold.) That the court focused in on that as the salient difference and not that Elan patented a method that was only necessary on a sensor which gave a single x and y trace is a mystery to me - gives me a feeling of sticking a finger in the patent and saying 'this is the part that was essential'.

Are there courts that are very restrictive? Are there rulings that have been subsequently found too expansive? Are there ones where others have just gone 'WTF?! Was the court high when they made this ruling?'

I've learned a lot - I do see the patents are more specific, now I have to figure out how what is and isn't specific in a particular patent is decided.

 

[cmaier]

Are there courts that are very restrictive? Are there rulings that have been subsequently found too expansive? Are there ones where others have just gone 'WTF?! Was the court high when they made this ruling?'

I've learned a lot - I do see the patents are more specific, now I have to figure out how what is and isn't specific in a particular patent is decided.

In determining how restrictive/expansive the claims are to be interpreted, courts look at the meaning of the words in the claims (claim construction) taking into account the text of the patent, the text of related patents in the same patent family, the file history of the patent (the back-and-forth with the patent office, often taking several years, where the inventor makes arguments as to why his claims are new and novel compared to prior art), etc. Many of those things result in the claims being interpreted more narrowly than might be apparent from a plain reading.

Courts are fairly consistent in applying these tests. All appeals are heard by the same court (the Court of Appeals for the Federal Circuit), to provide even more consistency.

Juries, on the other hand, vary greatly from place-to-place, but that typically has a bigger effect on the infringement result than on the validity result.

I should point out that while people may think, based on my comments here, that I am sort of patent infringement hawk, I almost invariably represent defendants in such suits - most of my experience involves trying to invalidate patents by producing prior art, or rendering them unenforceable by showing inequitable conduct by the patentee, or showing that they are not infringed.

 

[BobVB]

Ok a different tangent...

Obviously you can't actually patent 'multitouch' in all its forms because its a pre-existing idea, Star Trek: The Next Generation had been using it regularly since 1987 as just one example. Same that you couldn't generically patent the idea of nanites, phasers, replicators or transporters (all star trek) or interplanetary travel by way of gravity screen propulsion (Wells). Surely you can't get a patent for a concept that is already in the public domain, all you could do was suggesting a specific way it could be done with some existing technology, right?

In this case you can't using a finger or multiple fingers as interface input device as that's been being done in stories since the 50's at least. All you could patent would be a specific way of doing this well known 'thing' on some existing technology. Elan did this, writing a very specific way of implementing it on a very specific technology, a technology that few use anymore.

So how do you find the limits of 'generalization' for the patent? I find the ones the previous ruling found to be a bit odd and arbitrary and don't understand why these and not others.

Any insights?

 

[cmaier]

Obviously you can't actually patent 'multitouch' in all its forms because its a pre-existing idea, Star Trek: The Next Generation had been using it regularly since 1987 as just one example. Same that you couldn't generically patent the idea of nanites, phasers, replicators or transporters (all star trek) or interplanetary travel by way of gravity screen propulsion (Wells). Surely you can't get a patent for a concept that is already in the public domain, all you could do was suggesting a specific way it could be done with some existing technology, right?

yes.

In this case you can't using a finger or multiple fingers as interface input device as that's been being done in stories since the 50's at least. All you could patent would be a specific way of doing this well known 'thing' on some existing technology. Elan did this, writing a very specific way of implementing it on a very specific technology, a technology that few use anymore.

Ok.

So how do you find the limits of 'generalization' for the patent? I find the ones the previous ruling found to be a bit odd and arbitrary and don't understand why these and not others.

Any insights?

The "limits of 'generalization'" is called the scope. First, the limits are determined by the claims - they are like stakes in the ground that bound off a plot of land.

Second, you can't claim more than you've enabled. In other words, at best you can claim what a person having ordinary skill in the art would be enabled to create based on your entire patent disclosure.

Third, your claims are construed based on what a person having ordinary skill in the art at the time the patent was filed would understand the words to mean, in light of the rest of your patent document (and any back-and-forth and explanation in the patent's file history - the communications between the patent agent/attorney and the USPTO).

Once you've determined what the words in the claim mean with respect to that patent, you've determined the scope. Once you know the scope, you know what constitutes infringement.

There are also some specific rules, like if a claim says "comprising" vs. "consisting of" or contains the word "means."

Finally, you are required to fully disclose the invention so that a person having ordinary skill in the art would know what it is (this is slightly different than the enablement requirement I discussed above), the best mode of practicing the invention, and to to follow various procedural rules, otherwise your patent may be invalid or, if valid, may be unenforceable.

 

[BobVB]

yes.
Once you've determined what the words in the claim mean with respect to that patent, you've determined the scope. Once you know the scope, you know what constitutes infringement.

OK, so a example question:

The Elan patent is defined by their 'current invention', the touchpad someone else patented that could deliver an X axis and a Y axis trace. The flow diagram they provide is showing how if they can't find 2 discrete fingers in one of the profiles they can in the other and act accordingly (with the textual claim they can do more fingers).

So, put your 3 fingers tight together on the pad at a 45° angle to both x and y axis. With the Elan patent the x and y axis projections would look to overlap - their algorithm to determine if there were multiple fingers down would fail and multitouch functionality would not be achieved.

But with the  'threshold then trace touch borders' method explained in their patent there would be no problem in identifying 3 distinct touch areas, and their locations.

Since the 'scope' of the  patent does things the Elan one cannot does that mean they are not infringing on the old technology?

 

[cmaier]

the short answer is that it doesn't mean that. Infringement is determined simply by figuring out if each step in the patent claim is performed in the accused device. Doing additional steps on top of those inthe claim doesn't prevent infringement unless the claim has certain magic words in it.

If apple doesn't infringe, it's because it does not do the require steps.

OK, so a example question:

The Elan patent is defined by their 'current invention', the touchpad someone else patented that could deliver an X axis and a Y axis trace. The flow diagram they provide is showing how if they can't find 2 discrete fingers in one of the profiles they can in the other and act accordingly (with the textual claim they can do more fingers).

So, put your 3 fingers tight together on the pad at a 45° angle to both x and y axis. With the Elan patent the x and y axis projections would look to overlap - their algorithm to determine if there were multiple fingers down would fail and multitouch functionality would not be achieved.

But with the  'threshold then trace touch borders' method explained in their patent there would be no problem in identifying 3 distinct touch areas, and their locations.

Since the 'scope' of the  patent does things the Elan one cannot does that mean they are not infringing on the old technology?

 

[BobVB]

Hmmmmm,

the short answer is that it doesn't mean that. Infringement is determined simply by figuring out if each step in the patent claim is performed in the accused device. Doing additional steps on top of those inthe claim doesn't prevent infringement unless the claim has certain magic words in it.

If apple doesn't infringe, it's because it does not do the require steps.

OK, well since the Elan patent absolutely requires x y trace scans, if  apple is using higher level matrix math to find 'touch regions' as it's patent claims (they say it can use xy or radial coordinate systems), then even at the basest level of the patent its not infringing.

I guess I still don't understand 'scope' if a patent's basic methods fail then that means that situation is outside its 'scope', I would think that if another patented method didn't fail in that situation (and was not just some extra tag on process on top of the first patents to 'catch the misses') it would indicate the two patents had inherently different 'scopes'. even if incidentally they can do many of the same things.

 

[cmaier]

OK, well since the Elan patent absolutely requires x y trace scans, if  apple is using higher level matrix math to find 'touch regions' as it's patent claims (they say it can use xy or radial coordinate systems), then even at the basest level of the patent its not infringing.

I guess I still don't understand 'scope' if a patent's basic methods fail then that means that situation is outside its 'scope', I would think that if another patented method didn't fail in that situation (and was not just some extra tag on process on top of the first patents to 'catch the misses') it would indicate the two patents had inherently different 'scopes'. even if incidentally they can do many of the same things.

Imagine I had a patent on a stool with a first leg and a second leg.

Now I make a stool with two legs. It works fine if the person sitting on it balances with his own legs and leans on it just right, but it's not the best solution. My stool infringes.

Now I make a stool with a third leg. It works much better. It works even if my feet don't touch the ground to keep it from tipping over. It solves a problem not solved by a two-legged stool.

But my three-legged stool still infringes, because it has a first leg and a second leg. I can get my own patent on a three legged stool (maybe), but that doesn't allow me to actually make such a stool, because I would still be infringing the patent on the stool with a first and a second leg.

 

[BobVB]

Imagine I had a patent on a stool with a first leg and a second leg.

Now I make a stool with two legs. It works fine if the person sitting on it balances with his own legs and leans on it just right, but it's not the best solution. My stool infringes.

Now I make a stool with a third leg. It works much better. It works even if my feet don't touch the ground to keep it from tipping over. It solves a problem not solved by a two-legged stool.

But my three-legged stool still infringes, because it has a first leg and a second leg. I can get my own patent on a three legged stool (maybe), but that doesn't allow me to actually make such a stool, because I would still be infringing the patent on the stool with a first and a second leg.

Ok, but what if I use an entire different method of providing support, a semi circle of plastic around the entire bottom perimeter of the seat. At most it could be considered one leg if it was considered a leg at all. (with the analogy being  is using an entirely new way to determine touch areas than the Elan patent).

Would my one support stool patent be considered to infringe on a two leg stool patent?

 

[cmaier]

Ok, but what if I use an entire different method of providing support, a semi circle of plastic around the entire bottom perimeter of the seat. At most it could be considered one leg if it was considered a leg at all. (with the analogy being  is using an entirely new way to determine touch areas than the Elan patent).

Would my one support stool patent be considered to infringe on a two leg stool patent?

No.

Let me modify the hypothetical a bit. Let's say my claim was:

1. A stool comprising:
a first leg;
a second leg substantially parallel to said first leg; and
a seat orthogonally opposed to said first and said second legs.

If my stool had 2 legs that were slightly not parallel, it might still infringe by the "Doctrine of Equivalents." Assuming that the "parallel" requirement was not added to differentiate from prior art, then since legs that are slightly non-parallel perform the same function as parallel legs and peform it in the same manner, it would likely infringe.

But the Doctrine of Equivalents has an "all elements rule." This means EACH element of the claim must be present (literally or by its equivalent). You can't (usually) say that object X is equivalent to multiple claim elements. So in your example pointing at the C-shaped member and saying it is both the first leg and the second leg would certainly not meet the Doctrine of Equivalents.