Linux Format forums

[guy]

It's not often that Pamela Jones, celebrated driving force behind Groklaw, begins a news item with "OMG."

The U.S. Federal Circuit, a bench of ten judges, has just shot software patents ragged. Being a legal ruling, just how ragged is of course shrouded in lawyer's fees. This is bound to reverberate across the world, even as far as our own little Blighty.

For more, see: http://www.groklaw.net/article.php?story=20130510155818152
_________________
Cheers,
Guy
The eternal noob

[Dutch_Master]

OMG indeed! This'll end up before the US Supreme Court and as such, will drag on for a while. But if upheld, software patents may be a thing of the past

[johnhudson]

It's another step in a battle that has been raging since the 1990s - see http://codingfreedom.com/.

But New Zealand has recently decided not to uphold any software patents. So the tide may be turning.

[Rhakios]

Another angle on the importance of patents, software or otherwise. Indeed, it made me think that perhaps inventions are more likely to be seen as significant if they are not patented.
_________________
Bye, Rhakios

[johnhudson]

There is also the anti-patent view held, for example, by Michael Faraday who published everything he discovered as soon as he had discovered it in order to prevent anyone from patenting it.

Though not his intention, it has been argued that, because he published in English, this gave Victorian industry a headstart. It had been argued earlier than Boulton and Watts' steam engine patents had held up innovation in the 18th century.

In the 1960s the designers of a BBC studio speaker published the design to prevent anyone patenting it. It has been suggested that this prompted the boom in British speaker production.

[normanHLinux]

.Software patents do not make any sense. A patent was designed to protect the creator of a tangible object, to enable others to make similar items if they paid a 'reasonable' fee to the innovator.

Software patents are not relatd to any tangible object. They are, in effect, giving people the right to claim ownership of an idea. Worse, because the US patent office does not have the resources or expertise properly to investigate these applicatons they are licensing a)nebulous, wide-ranging claims and b)claims that should be ineliggible due to prior art.

Leaving aside the fact that software should nott be patentable (and coyright covers the expression of an idea - not the idea itself) the US system, instead of requiring the applicant to show that the idea was originallyy theirs, license spurious claims and then leave it to those affected to chllenge and demonstrate prior art. The costs have now moved from those intent on making money from an idea to those who wish to use that idea, knowing it did not originte with the patent holder.[/i]
_________________
If at first you don't succeed, give someone else a go

[Rhakios]

[Dutch_Master]

It may introduce abuse, if executed that way. A large company can dissuade sponsors to invest in a patent held by a new/small company, to see it thrown out and then use the exact patent to manufacture their own stuff. That's not what patents were intended to do, not by a long shot. Even then, patents may be so far ahead of their time that technology needs some time to catch up.

[wyliecoyoteuk]

A patent is not meant to protect ideas, but true inventions.
If a patent is applied for on something that cannot exist given current technology, it is an idea, not an invention.
If someone needs more than 5 years to develop it, they should be able to prove that it will be available within a realistic timeframe.
Many of the software patents are extremely vague, and design patents are even worse.
A patent on a rectangle with rounded corners for example.
The best defence is prior art.
Barnes and noble took on microsoft's android patents with a book of prior art examples. MS settled, because most if not all of their Android patent portfolio would have dissolved in court.
_________________
The sig between the asterisks is so cool that only REALLY COOL people can even see it!

*************** ************

[guy]

The question as to whether software can in principle be patented remains open. While agencies in various countries happily grant endless patents of dubious originality, it is a general principle that you can patent physical inventions but not ideas. (You can copyright the expression of an idea, but that is not the same as a patent on the idea itself).

There is a popular argument that software is logic, logic is maths, maths is ideas, therefore software cannot be patented and any such patents granted are actually invalid. This is very, very slowly being tested through the American legal system, while Europe (last I looked) remains in unhappy internal conflict.

Being schooled (at a very elementary level) as a philosopher and logician, I don't buy the "software is maths" argument.

Trivially, Maths is actually just one branch of logic, along with things like reasoned argument or plugging a load of hydraulic actuators into each other. It is mainly mathematicians who like to think that their discipline underlies all forms of reason, and there are more of them than there are of us, so they make a louder noise on the interweb. Doesn't mean they are right, though. Either way, software is certainly logic and logic may be abstracted as pure ideas.

However, I think there is a more important flaw in the "software is maths" argument. Suppose I invent a novel mechanical linkage that performs some arbitrary function such as drawing a straight line. I can certainly patent such a novel linkage, even if its function is performed equally well by some other well known linkage. If I invent a novel digital device dedicated to drawing straight lines, I can patent that too.

And that is the key point. I cannot patent the idea of a straight line, but I can patent my linkage - my mechanical algorithm - for drawing it. (I can copyright it too, but someone need only change the design details to evade that - rounded corners, anybody?)

Can anybody explain to me how software embodying my algorithm is supposed to differ from a mechanical or electronic device embodying that same algorithm?

If not, then it seems that software is indeed patentable. The fact that logic can be abstracted as a pure idea does not necessarily mean that it has to be.

Software does not exist in isolation, it must always be stored. So one might argue that a storage device, containing a program which embodies my algorithm, is patentable while the software itself is not. All one needs to do to avoid the patent is to use a different kind of storage device. But isn't that like saying that my linkage is patentable while the arrangement of connections between the links is not? Change the materials of which the mechanism is constructed, and you do not escape my patent, so why think you can change the storage medium? You are not even avoiding my software copyright, a much weaker thing.

Mathematicians might also like to contemplate the distinction between pure and applied mathematics. One might suggest that pure math cannot be patented while applied math at least sometimes can.

Just to clarify: I certainly believe that software patents are in general a Bad Thing. It's just that I don't buy the "software is maths" argument. They are bad for other reasons.
_________________
Cheers,
Guy
The eternal noob

[Rhakios]

[normanHLinux]

guy wrote