ESP Wiki is looking for moderators and active contributors!

Talk:Free software

Revision as of 21:53, 6 December 2010 by Ciaran (talk | contribs) (2009 Thesis on patents and free software: ::There's an article by Glyn Moody about the Thesis at http://www.computerworlduk.com/community/blogs/index.cfm?entryid=3019&blogid=14 ::~~~~)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

2009 Thesis on patents and free software

Title: The Effects Of Software Patent Policy On The Motivation And Innovation Of Free And Open Source Software Developers

I just looked at this briefly; I couldn't find what the conclusions are:

If anyone has the time to take a look, it probably contains a few interesting stats and conclusions. Ciaran 17:05, 6 June 2010 (UTC)

p.27: "Software patent presence. Although the general approach of applying a broad concept of software patent presence and using different metrics is still useful, most of the metrics used have shown little predictive power. One clear result, however, is that software patent incidents are a very rare event: Only 2.7% of the respondents reported such an incident.
"The newly constructed patent pressure index (used in chapter III and IV) proved useful, as it clearly shows the differences in perceived patent pressure between different software domains. The fourteen categories of SourceForge that were also used in the survey (see the list of abbreviations for all explanations) indicate three groupings: few domains such as multimedia (MIM) and formats and protocols (FMP) show a high patent pressure, and few domains such as education (EDU) and text editors (EDT) show nearly no pressure, which leaves a rather large group in the middle with only slight variation in the patent pressure level.
...
"Innovation behavior. Software patent presence per se has no observable empirical effect on innovation behavior, be it positive or negative, with the exception of reverse-engineering, which seems to be more frequent when software patents are present. One possible explanation is that FOSS developers who are motivated by the philosophical idea of ‘free software’ choose fields in which software patents are strongly present – exactly to provide free alternatives that are also ‘freed’ from patent claims."
p.28: "Concerning the effects of software patent presence on innovation behavior, the empirical results are less conclusive."
pp.138-139: "Concerning the effects of SWP presence on innovation behavior, the empirical results are less conclusive. Neither opponents nor proponents of SWP will find support for their positions that the presence of SWP decrease or increase respectively the odds for innovative, algorithm-based contributions by FOSS developers. None of the three metrics used to capture SWP presence lends sufficient support to either side – be it positive or negative. Support, however, is found for a hypothesis related to reverse-engineering: stronger SWP presence attracts reverse-engineering-based contributions by FOSS developers."
In other words, the only measurable effect that patents have on FOSS developers is to make them reverse-engineer their workarounds to be sure of avoiding infringement. However if patents became more embedded and widespread in the system and related incidents were thus encountered more often than 2.7% of the time, then I guess that would have to significantly bite on our ability to innovate. steelpillow 18:50, 6 June 2010 (UTC)
Hmmm, I suppose he has something interesting there. Could be useful info for refining our descriptions for how patents affect software - i.e. not all types of functionality are affected, but a few crucial functions (displaying video) are severely affected. The table at the end of page 89, 90 goes in this direction.
The lists of references and studies is also a good source for documents this wiki should have articles about. Ciaran 20:32, 6 June 2010 (UTC)
There's an article by Glyn Moody about the Thesis at http://www.computerworlduk.com/community/blogs/index.cfm?entryid=3019&blogid=14
Ciaran 20:53, 6 December 2010 (EST)

Free Software has important differences as concerns arguments

Because open source software provides something very significant to the public, entire blueprints, it does fulfill a special role when we consider arguments based on promoting the progress or free speech.

It is easier to justify giving open source a "free pass" because it enables synergies not possible with closed source. Second, it competes very favorably and directly with patents in terms of disclosure given to the public. Third, it enables a whole lot more people to write much more sophisticated software because of all the sophisticated software that is provided at $0 and with generous terms (some proprietary competes with this through "community editions" but not as a rule or to nearly the same extent). It also allows new inventions at almost $0 cost because blueprints (which can trivially be forked into a new product) are provided also for $0 and generous terms. By increasing the pool of those able to participate, the liability of a patent monopoly is much higher. Also, it is specifically because of FOSS that we frequently state that writing software requires near $0 (at least marginally). Finally, in terms of free speech, FOSS is clearly intended as an important and very valued exercise in free speech on a level beyond how proprietary software might be considered to be free speech. And this is part of what makes it a trivial step to go from free speech to a running product that some want to try and patent.

I think Free Software clearly stands out. In fact, those wanting closed source to remain patent free owe a debt to FOSS because the overall arguments are better because of the proof of concept and extensive library of products and education built through FOSS (including very large parts of the "Internet" of today). There is a further motivation to support no software patents if you like closed source: if you don't make good arguments and build up significant support, FOSS might get the "free pass" and leave you at a greater competitive disadvantage. Jose X 19:53, 16 October 2010 (EDT)

Special exception for free software

I've reverted two edits, but the text in those edits might be useful elsewhere, so here's a direct link:

I reverted because I can't see what use they would be for a campaigner who consults this page

The freedom of speech issues are important, but we have to keep our information clear and avoid mixing topics. When we talk about free software, we should include mentions of freedom of speech, and links to that article and its sections. If the reader wants to dig into that topic, they can go to the article for that topic.

Similarly for discussion of why non-FOSS companies don't support a FOSS exception (it's not their core business) - that's true, but campaigners who come to this page have probably already analysed the situation to the point of knowing that already. If it's worth a mention at all, it should be kept to a short sentence, to avoid hurting the readability.

That said, if someone wants to take a look at the link I posted here, there might be material for developing other articles. Ciaran 08:33, 21 October 2010 (EDT)

Free software defined

The article claims that a free software exception "requires a legal definition of 'free software'". As I see it, that'd be the easy part and desirable on its own merits. I took the liberty to translate FSF's free software definition and a couple key terms from GPLv3 into the writing style of Titles 17 and 35, United States Code:

To "convey" a work means to make and distribute copies or phonorecords of the work, to transmit it publicly, or to do anything else that allows other parties to receive copies or phonorecords of the work.

The "source code" of a work of authorship is the preferred form of the work for making modifications to it.

A "free software license" is a non-exclusive copyright license that allows the owner of a copy of a work made available under the license to reproduce the work, prepare derivative works based on the work, and convey the work or any derivative works based thereon to the public, without any royalty.

"Free software" is a work that

(a) is available as source code, and which may be available in another form; and
(b) is made available by its authors under a free software license or not copyrighted.

A "copyleft license" is a free software license that requires a licensee who conveys derivative works based on the work to convey them as free software. A "copylefted work" is a work distributed under a copyleft license.

An "information processing claim" is a claim that covers a new method of information processing with a known computing machine.

It shall not be an infringement of patent to make, use, convey, or offer to convey free software implementing the method of an information processing claim.

Any comments from the lawyers in the audience? --Damian Yerrick 11:20, 7 November 2010 (EST)

Hi. You might have a good point. I'll have to think it through a few times. In the mean time, one problem I remember is the worry that by putting a definition of free software in the legislation, we lose control of it. Our licences might contain errors or their meaning might get twisted by future changes in the legislation or case law, but if this happens we can react and try to fix the licences and try to propagate the fixed versions. If it's in the legislation, we can't do anything when the context changes. (I'm not saying this is a big problem, just mentioning what has been discussed.)
You could make a section "==What exclusion text might work==" (or whatever) in the article if you want to develop this idea.
One suggestion is to remember to keep the GPL as just an example. Be careful about using it's words and concepts - the exception should be broad enough to protect free software under all licences. (For example, it might be better to remove all mentions of copyleft.) Ciaran 14:51, 7 November 2010 (EST)