Battling the Hydra: FSFE’s work on Open Standards
On Friday we published an FSFE analysis on patents and standards, and shared it with the European Commission. This was in reaction to a letter [pdf] which the Business Software alliance (BSA) had written to the Commission in a last-ditch attempt to eradicate the last traces of Open Standards from the European Interoperability Framework. We replied to the BSA’s arguments, picking them apart one by one.
Why Open Standards matter for Free Software
Here’s a bit of perspective to this episode. Open Standards are important because they can be implemented in any software model, whether Free or proprietary. Where Open Standards are used, programs compete on their merits rather than on the number of users they manage to lock in.
For a standard to be open means, among other things, that it can be implemented without restrictions such as an obligation to pay a licensing fee to the patent holder for every copy that’s made of the software. Such “running royalties” are a typical feature of patents licensed under so-called “(fair,) reasonable and non-discriminatory” ((F)RAND) conditions. Glyn Moody has a great introduction to the topic, and here’s an in-depth discussion of the relation between patents and standards. To be sure, zero royalties vs running royalties is just one of several issues when it comes to standards that involve software, but it’s certainly an important one.
That’s why FSFE puts a lot of effort into promoting Open Standards in different fora, such as WIPO, the Internet Governance Forum and the European Commission. We also organise Document Freedom Day each year at the end of March. We read the lenghthy documents from the European Commission, WIPO and any other relevant policy fora. We analyse. We publish. We form alliances where it serves to strengthen the Free Software message, without ever compromising on substance. Sometimes we coordinate with industry groups such as Open Forum Europe. If we have a common agenda, we work together. If we disagree, each party goes their own way. Such is the business of policy work.
A global campaign against Open Standards
Recently, the Business Software Alliance and its friends have been waging an all-out campaign against Open Standards. Last week’s letter was only the latest instance of what looks like a co-ordinated international effort by the BSA and its friends to get their message out to policy makers. Like the legendary many-headed Hydra, people pushing out the party line are seemingly everywhere. Over the past month or so, I kept hearing the same routine wherever I went, always in similar words.
A few quick words if you’re unfamiliar with the BSA: Prominent BSA members include Microsoft, IBM, SAP, Adobe, Siemens, and many others. The BSA is perhaps best known for everybody’s favourite stats on illicit copying of software, which routinely claim that everyone who uses unlicensed proprietary software would have bought a licensed copy if the unlicensed one hadn’t been available – an error known as the “lost sale myth”.
This time, the BSA argues that most Free Software licenses are compatible with RAND licensing of patents as we know it. That is clearly bunk. More than 85% of Free Software projects are distributed under licenses that are incompatible with RAND licensing. The exception are non-protecting MIT and BSD-type licenses.
Microsoft’s Wilfried Grommen argued at the Internet Governance Forum 2010 that RAND licensing is compatible with Open Standards. A week later, at the Westminster eForum in London, Francisco Mingorance, the BSA’s Senior Director Government Affairs, made the same point (have a look at Glyn Moody’s article on the event here).
Last week at WIPO, during the 15th session of the Standing Committee on Patents, a side event took place to inform WIPO delegates on the relation between patents and standards. Two figures on the panel were Benoit Mueller, formerly the BSA´s Director of European Software Policy; and Jonathan Zuck of the “Association for Competitive Technology“, a group that pushes for software patents. Both of them argued that RAND licensing of patents in standards was compatible with Free Software, and that governments shouldn’t express even the mildest of recommendations for Open Standards in software.
Somehow we weren’t surprised to see exactly those arguments repeated in the BSA’s letter to the European Commission, signed by Francisco Mingorance.
False flags?
Then there’s Florian Mueller, who is a bit of a special case. He claims credit for opposing software patents in Europe back in 2005, which used to buy him some credibility in the Free Software community. Recently, however, he has been running around the web as an “independent software developer” (here’s a typical quote), echoing the same message as Microsoft and the BSA in his blog, and in various comment threads. When challenged, he admits to being paid, but won’t say by whom:
What I have done is some work in connection with competition cases, which is explicitly exempt under the European Commission’s transparency rules. You’re not in a position to ask for more of a disclosure than the Commission does.
While I have no problem in principle with people lobbying for whatever they care about – hey, free speech –, doing it without disclosing who’s backing you strikes me as morally challenging. (You can find FSFE’s list of supporters here.)
In it for the long run
The European Interoperability Framework is just one battle among many. Besides the topic of interoperability in the public sector, there’s the task of reforming standardisation systems so that they produce Open Standards, and educating policy makers about the importance of the issue. I’m sure we’ll meet the BSA and its friends many times during this long-term effort. We’ll have to keep dealing with the Hydra’s heads for some time yet.
Comments
[...] This post was mentioned on Twitter by Georg Greve and Carlo Piana, Karsten Gerloff. Karsten Gerloff said: Battling they Hydra – FSFE's work on Open Standards http://ur1.ca/24plt #fsfe #gnu #fsf #eifv2 #swpat [...]
Since you linked above to that recent ComputerworldUK discussion, why didn’t you just go there and reply to the question I raised, about you apparently not being worried about the approach to interoperability and “open standards” taken by your lobbying allies such as IBM and Oracle? A free software developer from Austria also asked you to take a position on my civilized, legitimate questions.I appreciate your link to my blog. That link is apparently meant to substantiate the claim that I would be “echoing” Microsoft/BSA positions. If people follow that link (here it is again), they can see for themselves that it’s a compromise roadmap — a proposal that’s about as much in the middle between your position and that of BSA as it gets.Since Eben Moglen blessed Red Hat’s royalty payment that was part of the FireStar settlement, it would be helpful if you, too, could clarify once and for all that royalties and the GPL (in Eben’s opinion, even GPLv3) aren’t inherently incompatible with each other. Your “analysis” of the BSA letter has 17 occurrences of the word “royalty” and only one of “freedom” (“freedom from lock-in”). How about the four freedoms and free as in speech vs. free as in beer? I can see why a corporate lobby group would focus on royalties (saving costs, increasing margins) instead of freedom, but the FSFE?You falsely claim that I’m “lobbying”. Whom am I lobbying now and whom have I ever lobbied over the EIF? Please name just one official I would have lobbied over this. There isn’t. My last lobbying assignment was in 2007, just for an update.On your final sentence, I hope you can still sleep at night despite all those “Hydra’s heads” purportedly surrounding you.
Enable comments on your blog and you never can tell what kind of people will respond. The question still appears to remain – should open standards allow RAND or FRAND licensing of patents, a requirement not compatible with the vast majority of open source?
Requiring licensing of any patent as a precondition to an ‘open standard’ either provides preferential treatment to proprietary software interests allowing embrace and extend or blocks competition by preventing a wide range of open source adoption. Allowing the requirement for explicit licensing to practice an open standard provide the patent holder with information on the competitive landscape. All these provide advantage to the patent holder and proprietary software interest while disadvantaging the majority of open source.
Should open standards provide competitive advantage or a level playing field?
@Dio Gratia: You just repeated an important talking point of the royalty-free camp: “a requirement not compatible with the vast majority of open source” — that’s not true. Some FRAND is FOSS-compatible, some is incompatible. The constructive approach is to identify specific problems and to suggest solutions. In terms of a level playing field, I agree on the objective but royalty-free isn’t the only way to get there. You mention advantages of patent holders — there’s nothing that prevents FOSS developers, contributors and distributors from holding their own patents, and many of them do (Red Hat, for instance).
@Florian: “Some FRAND is FOSS-compatible, some is incompatible. he constructive approach is to identify specific problems and to suggest solutions.”…
Please explain what you view as constructive.
a) Insuring that the principles of Free Software (or the OSD if you prefer) are upheld.
b) Using loop holes in some Free Software / OSI certified licenses to circumvent these principles for FRAND licensing policies
(Side Note: I get the impression that what you wrote infers that Red Hat is licensing patents under FRAND terms… I hope that is not what you intended. We know that many Free Software players hold many patents as that is the world they need to deal with. But I would I would be extremely surprised if there was any evidence that Red Hat would endorse anything but RF terms (save counter claims))
@Florian: Paying royalties for patent license is Danegeld, ceding ‘open standards’ to mean competitive advantage to or ownership by the patent holder.
Some of the most popular open source licenses require warranting that a recipient of a distributed work be free to use software for whatever purpose. Unfortunately, the Non Discriminatory part of (F)RAND doesn’t refer to use rather inferring eligibility for a license. We have seen at least one vendor limit their patent pledge to implementation of a particular ‘open’ standard in a manner precluding adoption by the aforementioned open source software licenses without further clarification, the particular vendor also quite reluctant to provide clarification. You could note that any control of the delay in clarification can also provide competitive advantage.
Would either such case represent an open standard or is the meaning being suborned or limited to impracticality for purposes of competitive advantage? It certainly doesn’t appear to be open if open should infer a level playing field.
@florian, Mr. Mueller, please don’t play with words, you are such an embarrassment!
Lobby — v.intr.: “To try to influence the thinking of legislators or other public officials for or against a specific cause: lobbying for stronger environmental safeguards; lobbied against the proliferation of nuclear arms.” — The American Heritage® Dictionary of the English Language, Fourth Edition
“Lobbying (also Lobby) is a form of advocacy with the intention of influencing decisions made by legislators and officials in the government by individuals, other legislators, constituents, or advocacy groups. A lobbyist is a person who tries to influence legislation on behalf of a special interest or a member of a lobby. Governments often define and regulate organized group lobbying that has become influential” — Wikipedia
You are clearly lobbying for the same for which BSA lobbies: to remove a reference to open standards, however vague, from the EIF, and against Open Standards. So you are denying to be lobbying in the same paragraph where you do lobby, proposing “an alternative roadmap” (to what?), I believe this speaks for itself.
[...] Karsten on Free Software » Battling the Hydra: FSFE’s work on Open Standards [...]
@David Ayers: It would be great if the FSFE just wanted to uphold the principles of free software. However, you should look at the FSFE’s analysis of the BSA letter to the European Commission and ask yourself what that kind of letter has to do with free software. It mentions royalties 17 times and “freedom” only once (“freedom from lock-in”). Concerning Red Hat’s patents, I meant inbound — not outbound — patent licensing.
@Dio Gratia: you just bring up an example of clause that some may consider FRAND but which would in the opinion of some not be compatible with free software. I don’t deny that some FRAND may not work for free software. What I reject is the notion that all FRAND is incompatible with free software. In terms of a possible limitation to (or exclusion of) a particular license, how about the GPL? Even GPLv2 and GPLv3 aren’t compatible. So this is a problem that apparently affects free software all the time.
@Guglielmo If you’re so sure of being right with what you claim, why do you have to start your comment with an insult? I have no problem with a reference to open standards. The problem is that there will only be an EIFv2 if a middle ground between the opposing camps can be found somewhere. An extreme definition of “open” (although “open” and “royalty-free” are two different aspects of the same thing) is one of the key obstacles to such an agreement, and in my view it’s unhelpful to keep insisting on a vision that ignores the reality of the patent regime we live under.
@David Ayers: an addition, Red Hat’s patent policy doesn’t specify how Red Hat will use those patents in connection with proprietary software (which is basically like the dual licensing approach MySQL pursued) and even FOSS isn’t protected against certain scenarios, one of which would be that of Red Hat selling its patents to a strategic aggressor or troll.
@Florian:
Fair, reasonable and nondiscriminatory terms can affect the ability to distribute open source software. See Clause 7 of GPL v2. Imagine a patent license grant qualifying as FRAND that doesn’t inherit (doesn’t cover the work once distributed). The first distributor would have the unequivocal right to distribute, a recipient would not.
The fourth paragraph of Clause 11 of the GPL v3 is intended precisely to prevent such an occurrence insuring the rights of the recipient are identical to the rights of the distributor.
Wishing to exploit the differences between the GPL v2 and GPL v3 is only useful for stopping this cure to the loss of recipient rights and limiting distribution of open source.
A cure for open source licenses not having the foresight of Clause 11 of the GPL v3 would be a requirement that FRAND terms inherit with distribution, avoiding the necessity of a public statement of the patent holder required for estoppel, as an affirmative defense.
There is undoubtedly proprietary advantage in blocking the rights of down stream recipients, effectively tying open source projects as satellites of a patent holder offering less than open FRAND license rights.
@Florian:
Another issue with FRAND is restrictions on use such as found in MPEG LA no fee license grants, while the intent of open source distribution licenses is generally not to restrict
The effect is to require FRAND be not limited by use to comply with the most popular open source licenses.
@Dio Gratia: focusing on just the last point you made, all I proposed on my blog was for a FOSS-compatible solution within the FRAND framework. If that sounds reasonable to you, then we’re in agreement. It is, however, not what FSFE asks for. FSFE promotes “royalty-free” (without caring about all other aspects of software freedom, as if “free software” was just free as in beer, not as in speech). I can give you a royalty-free FOSS-incompatible license. I can also give you a royalty-based FOSS-compatible license.
@Florian:
Open source and open standard as used by the Open Source Initiative are compatible with majority share open source licenses. Note clause 3 of the Open Standard Requirement for Software http://opensource.org/osr .
“royalty-free terms for unrestricted use, or be covered by a promise of non-assertion” is not enclosed by fair, reasonable and non-discriminatory (FRAND). Your position appears untenable by OSI definitions.
@Dio Gratia: open standards existed in all sorts of industries before the OSI was founded. The OSI doesn’t own that term. It has its views — others have different views. My point was always about what the license allow, not about what someone may propose for a definition. If you want to read about the general line in the EU, here’s a summary. In connection with competition and standardization, a clear distinction is made between openness and royalties. In that tradition, the logical middle ground would be for FOSS to meet FRAND and vice versa.
@Florian:
Clause 2, sentence one of GPL v2:
“You may modify your copy or copies of the Program or any portion of it, thus forming a work based on the Program, and copy and distribute such modifications or work under the terms of Section 1 above, provided that you also meet all of these conditions:
…” (no use restrictions placed in conditions).
Clause 7, first paragraph last two sentences of GPL v2:
“If you cannot distribute so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not distribute the Program at all. For example, if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you, then the only way you could satisfy both it and this License would be to refrain entirely from distribution of the Program.”
The intent of which is mentioned in the GPL v2 preamble, second to last paragraph, last sentence:
“To prevent this, we have made it clear that any patent must be licensed for everyone’s free use or not licensed at all.”
From the EUPL V1.1 Guidelines for Users and Developers, document Page 7 (PDF page 10, English version), left column, fifth paragraph:
From the EUPL point of view, there are compatible copyleft licences listed as follows: (1st February 2009):
• the GNU General Public Licence (GPLv.2)
• the Open Software Licence (OSL) v. 2.1 and/or v. 3.0
• the Common Public Licence v. 1.0 • the Eclipse Public Licence v 1.0 • The CeCILL v. 2.0
The EUPL v1.1 is compatible with GPL v.2, which tells us explicitly that Royalty-Free licensing is required for recipients for any patents.
Fair, Reasonable And Non-Discriminatory (FRAND) licensing cannot be reconciled with Royalty-Free for an open ended distribution as noted in the Glyn Moody blog post linked in this article. It doesn’t appear possible to support FRAND in the EUPL without breaking GPL v2 compatability mentioned in Article 5 of the EUPL v.1.1 as supported in the Appendix without the original works author having obtained a paid in full no royalty license from the patent holder(s) without use restrictions and allowing distribution within the OSI OSR guidelines (or GPL). For any recipient that would by Royalty Free. You could note that the GPL v2 supports regional (geographic) restrictions on distribution to reflect patent license (Clause 8). This has been the intent of the GPL v2 all along, that the works author could have licensing rights that would be passed to any recipient for any use without royalty payment.
We’re still quibbling over whether open in open standards allowing FRAND means the same as open in open source, wherein downstream distribution must be royalty free. You’re butting heads with people who contend that if the original works author can’t implement a standard without prepaying for all possible recipients who use the resulting code for any purpose, then it isn’t an open standard, that license fees have no place in open standards.
Trying to change the argument to one of that it is possible for someone to ‘bell the cat’ under open source does not affect whether or not a standard is open. Such a definition doesn’t provide a level playing field.
An non-level playing field insures proprietary interests dominate, while an open source implementation might require someone to pay up after prolonged negotiation, the negotiation enabling competitive advantage. As one might expect such a definition of ‘open standard’ might be held in the same regard as ‘office open’, where under mandate to provide non-competitive access Microsoft only provided sufficient promise allowing open source support after more than fourteen months, where open source adoption might languish in the mean time in the face of willful infringement.
Open standards should not give competitive advantage in exchange for the ‘open’ label. That the subject even comes it would appear someone seeks competitive advantage for gain, or would it be simply a side effect, either undesirable, unrecognized or discounted?
Open in open source doesn’t convey competitive advantage, why should it in ‘open standard’?
Why Open Standards matter for Free Software ?
My opinion on your question. An open source when it was first developed will be shared for everyone, so it was free and then people with the same passion for open source development and it can also be part of trade but most software is still free open source so it is usually free.
[...] just graduated and my insurance runs out in about a month. I have part time work now at a medical board but I need a full time job and that is hard to come by in this economy. I am [...]
[...] the rest here: Karsten on Free Software » Battling the Hydra: FSFE's work on Open … Tags: bsa, time, under-licenses Category: Uncategorized | Comment (RSS) [...]
[...] Karsten on Free Software » Battling the Hydra: FSFE's work on Open … [...]
[...] are the top 11 Exercises for to Get Faster and Stronger for Football. Add them to your rotation and work ‘em [...]
[...] Karsten on Free Software » Battling the Hydra: FSFE's work on Open … This time, the BSA argues that most Free Software licenses are compatible with RAND licensing of patents as we know it. That is clearly bunk. More than 85% of Free Software projects are distributed under licenses that are incompatible . [...]
[...] have to have a parking sticker on my car for work and when I receive a new, never gets old. The last time I tried to remove the parking sticker left [...]
[...] President tribute "to Filipino workers (OFW). BABYSITTER – Fisher- Price Precious Planet Open Top Cradle [...]
[...] China Leather Association, said: "self-tanning industry, environmental action groups formed and work is a passive commitment to energy conservation by the leather industry emission reduction [...]