Micron Lands Broad "Slide To Unlock" Patent

Zordak writes "Micron has recently landed U.S. Patent 8,352,745, which claims priority back to a February 2000 application---well before Apple's 2004 slide-to-unlock application. While claim construction is a highly technical art, the claims here are (for once) almost as broad as they sound, and may cover the bulk of touch screen smart phones on the market today. Dennis Crouch's Patently-O has a discussion."

Re:The USPTO is holding roundtables

[h4rr4r]

Here is a simple suggestion.

MATH IS NOT FUCKING PATENTABLE YOU IDIOTS! STOP ALLOWING SOFTWARE PATENTS SINCE THEY ARE JUST MATH.

This should be pretty simple, but they would never accept that.

Slashdot admins; The caps are supposed to be yelling, that is why I used them. Sometimes yelling is needed.

Re:The USPTO is holding roundtables

[Theaetetus]

Here is a simple suggestion.

MATH IS NOT FUCKING PATENTABLE YOU IDIOTS! STOP ALLOWING SOFTWARE PATENTS SINCE THEY ARE JUST MATH.

Software patents that are just math are not allowed. You'll see that this one involves hardware, which is not just math*.

*unless you're saying that all hardware can be abstracted as mathematical relationships, and are therefore arguing that all machines are unpatentable too?

Re:The USPTO is holding roundtables

[Dr_Barnowl]

The very first test should be - ask a bunch of software guys - "If you had to do X in software, how would you do it?"

If ANY of them gets even close the patent should be thrown out.

That said, there shouldn't be any software patents. Asking how to improve the process of patenting software is like asking how to improve the process of circumcision. Just because you do it now, and lots of tribal elders say it's a great idea, doesn't mean it's true.

Re:The USPTO is holding roundtables

[AwesomeMcgee]

I think you hit the nail on the head but have the wrong idea. Yes, hardware is patentable; the way in which your hardware did something may be patented by all means. If it's something stupidly simple there will be a million ways to do it so good luck having a patent that matters. The thing you're doing being just math should be unpatentable, but the way you do it in the physical world should be.

Think about it like this, if nvidia comes up with some phenomenal new shader technique in the hardware that blows everything else away, they should be able to patent that hardware technique, chances are if it was easy to come up with ATI would have done it so this patent is worthwhile. ATI can surely do the same thing less efficiently or in software, but the patent gives advantage how it should while not stifling completition because ATI's software that does the same shader technique less efficiently wouldn't land them in court, even a less efficient hardware design (or a more efficient hardware design, which ATI should then patent!). As soon as you take the die apart and find the identical circuit implementation in both chips you have a patent violation.

Re:The USPTO is holding roundtables

[ColdWetDog]

Asking how to improve the process of patenting software is like asking how to improve the process of circumcision. Just because you do it now, and lots of tribal elders say it's a great idea, doesn't mean it's true.

I was going to nominate you for some sort of award, but I am puzzled as to exactly how to proceed. Non sequitur of the month? Most inappropriate Off Topic rant of the week?

Were you trying for anything in specific or was this a toss-it-against a Library of Congress to see if it was saluted?

Re:They need to go

[jd659]

The problem with the software patents (or the current interpretation) is that it's not the code that gets patented, but rather the problem. In the old days, when the inventors had to submit the prototypes, there were hundreds of mouse trap patents issued. In the current "patent the problem" world we would see the wording for a mouse trap to be something close to "a device that stops a creature from free movement by using physical, mechanical, electromagnetic force in a restrain device".

Zero value

[Qzukk]

The patent has no informative properties at all. It does not explain any process or algorithms used to ie detect that a finger is moving on the screen, which direction its moving in, what shape it draws, whether the shape is "close enough" to the shape in memory etc (though to be fair claim #1 doesn't even check the shape, only that a shape was drawn, claim #2 is to check the shape and deny access if it's wrong). If all of the above is obvious to someone of regular skill in the art, then the claims should be invalid as obvious. If it is not obvious to someone of regular skill in the art, then the patent fails to live up to the Constitutional mandate to advance the sciences and arts by not disclosing how these claims are to be achieved, and if current patent law does not make the patent invalid on this basis, then patent law should be changed to comply with the Consitution.

Re:The USPTO is holding roundtables

[gstoddart]

What gets patented is math.

Really? What gets patented is ideas, not math (because you can't actually patent that).

Nobody patented the mathematical formula for swipe to unlock. Nobody patented the mathematics of the Zev-Limpel algorithm, they patented the idea of using it for compression of data.

Other than the fact that it's describable in math, if I patented something like file-sharing, I'm not patenting a single mathematical concept. I'm patenting an implementation, or the idea for an implementation of a NON MATHEMATICAL concept, but done on a computer.

Almost every thing we do in computers involves trying to come up with a digital analog for a real world thing. Nobody says "Hey, I have this mathematical equation, let's patent it" ... they say "hey, let's patent doing this with a computer".

Sorry, but I've been writing code for almost 25 years, and despite the fact that my output can be expressed in math, usually in my day to day life, the fact that there is a connection to mathematics is either irrelevant or secondary.

Comp. Sci is no more a strict subset of mathematics than engineering is, because there's other things that come into play that the mathematicians don't factor in.

Debugging software isn't like solving an equation -- nor is writing it, and nobody writes out a mathematical equation for a piece of enterprise software and says the rest is just an exercise for the reader (unless you're Donald Knuth).

Only mathematicians believe that even most aspects of software is math.