Linked by Thom Holwerda on Fri 4th Jul 2014 12:57 UTC
PDAs, Cellphones, Wireless

A Sailfish developer (third party, so not affiliated with Jolla) has developed a swipe keyboard for Jolla. It's essentially done and ready to go, but he was too afraid to release it. The reason?

I'd like to release this as an open source project, but at the moment I'm not comfortable with the patent issue (I'm interested in any advice on this topic). I live in a country outside the US (and without software patents), so should I just find a code hosting service with no relation with the US?

Fellow Sailfish developers and users chimed in, arguing he should be fine with releasing it as open source and hosting it outside of the US, with a warning that it should not be used in the US. He has accepted this advice, and is currently working on releasing it. While this is great news for Sailfish users, this does highlight the destructive nature of software patents.

Since he's going to release the code as open source, we can be 100% sure that none of the code in there is stolen from Swype and that none of it violates the open source license governing possible other swipe-like functionality (e.g. Google's Android keyboard). Ergo, he has developed this on his own, and has produced his own code, or used code that is freely available. It's a fruit of his labour, possibly infused with code that was meant to be used in a sharing manner.

And yet, despite the above, it's very likely that yes, he is violating a bunch of patents by producing this keyboard, and is, potentially, running a risk. I'm not so sure the legal advice given in the thread holds up - I'm not a lawyer, and neither are (I'm assuming) the people in the thread - but I'm at least happy he is willing to run the risk for us.

Now, I ask you: is this fair? Is this the future that we want for developers and programmers? Is this the message that the United States government, its technology companies, and said companies' public advocates want to send to aspiring hobby developers the world over? Should Europe, India, China, and the rest of the world just accept this?

I'm sure the proponents of software patents will wave this away to solve their state of cognitive dissonance, but I'm honestly and seriously worried about the developers who have not released, are not releasing, or will not release their code because of the bribes changing hands from Apple, Microsoft, IBM, Google, and the rest to Washington legislators.

Patents are supposed to spur innovation, not hinder it.

Thread beginning with comment 591983
To view parent comment, click here.
To read all comments associated with this story, please click here.
RE[6]: Help me
by oskeladden on Sun 6th Jul 2014 20:59 UTC in reply to "RE[5]: Help me"
oskeladden
Member since:
2009-08-05

Software developers are ok with this. Most of the work goes into writing the code, which is copy protected.


Alfman,

Software developers are indeed OK with this, but copyright still doesn't serve anyone's interests because it's the wrong tool. If you don't mind, I'd like to continue a conversation which we started in another thread but which closed to comments before I could respond.

Copyright represents a bargain: it protects the expression of an idea in exchange for the idea itself being put out there for anyone else to take and run with. Let's say I invent a new literary genre - Japanese Buddhist steampunk detective fiction to pick a slightly absurd example. I publish the novel, and get copyright (and, thus, commercial exclusivity) for it. In exchange, a whole lot of other stuff - the idea, the genre, the technique, etc. - are put out there for others to use.

This bargain simply doesn't work with software copyright: because my source code is never published, I don't actually make my technique available for others to use and learn from. The contribution to public culture is a fundamental part of the justification for copyright, but you don't really get it in relation to computer software. The result is that the rightsholder gets copyright, and because source is closed and software licenses are extremely restrictive, the rightsholder also gets a type of protection not too dissimilar in its effects from trade secrets. In just about every other industry, people have to choose - they can't get both trade secrets and some other form of IP. Except in software, where they get a close equivalent of both.

This would be bad enough in any industry, but it is particularly disastrous in computer software because advance in programming is quite deeply cumulative. You said elsewhere that programmers tend to converge on the same solutions to problems, but this is a relatively new phenomenon. Contrast it with early developments in software - for instance text editors. It is striking just how different the solutions to the same problem were back then - vi, emacs, pico and sam were all responding to pretty much the same problem, but they did so in very different ways. The contrast with modern office suites could not be more stark. I would argue that this is because the fundamental engine of progress back then was being able to see how other programmers did things - an engine that modern IP law basically shuts out.

Patents are even worse. Back in 1994, Columbia Law School held a fascinating symposium on computer software and IP law, which should've gotten more traction than it actually got. Some of the participants there predicted that if we stuck with software copyright, the result would be that companies would find some way of dragging software into the patent dragnet with very unhappy results. They were of course right. The essential problem is that people on the management side see software as a product - as, in essence, a way of solving a problem and meeting a need; and the thing they want protected is that solution. They can afford the best lawyers - far better ones that the patent office and ordinary joes can - so if copyright doesn't give them what they want, they will find a way of stretching patent law to fit. As they indeed proceeded to do.

Back then, my thinking was that we needed to have a clear prohibition against patenting any form of microprocessor instruction, and instead create a sui generis software 'method' right, which granted a protection for a very limited period to the particular method of solving a particular problem a piece of software embodied. The quid pro quo would be a filing which disclosed the full technical details of the method (as you have to do with patents). The functionality would not be protected - only the method used to achieve that functionality (much as was done with semiconductors). But that boat has sailed. We're stuck in the worst of all possible worlds (again, as some participants at the Columbia symposium predicted twenty years ago), where computer software - uniquely - gets copyright, patents AND something akin to trade secret protection.

TL;DR: Software copyright is a bad idea; software patents are one of the few things that manage to be worse; we've managed to end up with both.

Reply Parent Score: 3

RE[7]: Help me
by Alfman on Sun 6th Jul 2014 23:06 in reply to "RE[6]: Help me"
Alfman Member since:
2011-01-28

oskeladden,

This bargain simply doesn't work with software copyright: because my source code is never published, I don't actually make my technique available for others to use and learn from. The contribution to public culture is a fundamental part of the justification for copyright, but you don't really get it in relation to computer software. The result is that the rightsholder gets copyright, and because source is closed and software licenses are extremely restrictive, the rightsholder also gets a type of protection not too dissimilar in its effects from trade secrets. In just about every other industry, people have to choose - they can't get both trade secrets and some other form of IP. Except in software, where they get a close equivalent of both.


Well this is very insightful, and you are right it is a sort of difference. While technically an author can apply copyright to BOTH the source code and the binaries, many developers choose not to use copyright for source code at all and simply do not release the source.

Interestingly the topic of source code access came up under the threads of another recent article too.
http://www.osnews.com/thread?591781



Back then, my thinking was that we needed to have a clear prohibition against patenting any form of microprocessor instruction, and instead create a sui generis software 'method' right, which granted a protection for a very limited period to the particular method of solving a particular problem a piece of software embodied. The quid pro quo would be a filing which disclosed the full technical details of the method (as you have to do with patents). The functionality would not be protected - only the method used to achieve that functionality (much as was done with semiconductors).



In your opinion, would a software developer who creates all his own software have to worry about accidentally infringing?

The sui generis protection of chipsets that you described seemed reasonable because it's very unlikely for two engineers to develop the exact same board by coincidence. But I ask myself to what extent this property carries over to software especially given the number of software developers is many magnitudes greater than board designers.

I'm concerned with overhead & scalability. The work needed to uphold the patent system grows exponentially with the number of developers contributing work to be cross checked. The patent system is still relatively tiny compared to the number of developers we have, so it would be even worse if participation was compulsory and everyone needed to register. If we could eliminate the laborious and error prone uniqueness constraints, then it might be scalable, but is that useful?

An idea would be to try and "normalize" submissions in such a way that each unique algorithm automatically gets a unique identity. This would help with the uniqueness problem, but then I'd worry that we'd get someone who deliberately iterates through the unclaimed algorithms in rapid succession to claim ownership.

But that boat has sailed. We're stuck in the worst of all possible worlds (again, as some participants at the Columbia symposium predicted twenty years ago), where computer software - uniquely - gets copyright, patents AND something akin to trade secret protection.


Yea, this seems to be the case. I'm glad we're discussing alternatives, but like you say, it's hard to change course at this point.

Reply Parent Score: 3