MeatballWiki | RecentChanges | Random Page | Indices | Categories

I think this is probably not worth keeping as it is a troll magnet. LimitTemptation. I'll kill the backlinks if this is the general consensus (unless I am reverted). -- SunirShah

I see no trolling going on here. Delete the trolling when it happens, but if you delete anything remotely controversial because it MIGHT attract trolls, they've already won. --ChuckAdams

I think Sunir means that it is a potential troll magnet. I think it should be deleted because it's silly. ;) -- StephenGilbert

1. Summary
2. Background
3. Legal threats and risk
4. Public relations
5. Power
6. Alternatives
7. Discussion

1. Summary

RichardStallman is a passionate ideologue who founded and leads the FreeSoftwareFoundation and the GNU project, the pre-eminent organizer of the OpenSource arena, and the principle author of the GeneralPublicLicense (GPL). Despite the explosion of GPL-related projects, Stallman still personally attempts to police developments. He sent a tactless letter to David Fiddes of the GNU Coldfire project informing him that he was violating the GPL, to which Fiddes promptly responded by quitting OpenSource (exercising his RightToLeave). Both positions were extreme. On one hand, OpenSource will survive without Fiddes, and therefore it can ignore such reactions. On the other hand, this was another example of Stallman's poor communication skills, and it further indicates he should better restrain himself when dealing with the public. Not all of us can exercise our RightToLeave as Fiddes has since for many practical purposes GNU/GPL/FSF and therefore Stallman is the only game in town.

2. Background


RichardStallman discovered that David Fiddes, once the most active contributor to the GNU Coldfire cross-compiler project, was violating the exact terms of the GPL. Instead of distributing the full source of his changes, he only distributed the binaries plus the necessary patches to the GNU C Compiler (the full source exceeds 100 MB and is widely available, and Fiddes' ISP has space and bandwidth limits that made hosting the entire source cost prohibitive). "Because you are distributing binaries, to comply fully with the GPL, the GNU Coldfire cross-compiler source distribution must include complete source code, not just patches." Stallman finished off his letter with, "It is clear that you don't mean any harm, and we are not going to take a vindictive approach, but you need to correct this. I hope we can resolve the matter without resorting to lawyers."

Interpreting the above in part as a hammy LegalThreat, and perhaps since David is a type that will AvoidLegalRisk perhaps more than necessary, David said GoodBye:

"Due in part to this [less than helpful] email from Richard Stallman I have come to realise that I perhaps do not have the technical, political and legal skills as well as time and funds required to make what I feel is an effective contribution to Free Software. I intend to cease all development of GCC and Binutils shortly once I have attempted to submit any useful features in my patches to the official maintainers of these packages. This is as much due to lack of time to do things properly as anything else." [1]

Fiddes also took down most of an informative web site with information about a variety of Coldfire topics. A mirror still exists at http://sca.uwaterloo.ca/www.calm.hw.ac.uk/davidf/coldfire/download.htm . He also discontinued his postings to compiler-related mailing lists. He had written over 200 posts over the preceeding three years.

Further, Fiddes found the amount of work necessary to contribute to a GNU project (or any CopyLeft project) excessive. As he writes:

Apparently providing just the source patches for the compiler against standard FSF releases is not sufficient under the GNU GPL and I should have released the entire modified source used to build the compiler as well.

In order to comply fully with the licence:

I apologise unreservedly for any inconvenience this has caused anyone.

3. Legal threats and risk

First, the FreeSoftwareFoundation has yet to take anyone, including corporations, to court over GPL violations; all cases have been settled out of court. Nonetheless, the FSF has a responsibility to enforce the GPL as it is the very thing that allows free software to exist. As AlexSchroeder comments:

I also find it quite astonishing that people seem to like free software and the licenses that make free software possible, except when they are told to follow the terms. . . . Since we're talking about violations of the GPL, and the GPL is what makes free software free, then the situations we are talking about are not 'trivial misuse' -- they are the very raison d'ĂȘtre of the FSF.

Maybe this is a problem of perception: We use the GPL to defend against misappropriation, therefore we want it to be strong and undisputable. If we weren't adamant about copyleft, we'd be using a BSD license, or the PublicDomain.

As far as we can tell, however, the letter was not intended as a LegalThreat. As Stallman himself pointed out in his letter, he felt that Fiddes was acting in good faith. Given Stallman's self-professed utter lack of social skills, he simply failed miserably to communicate the opposite of a legal threat. Not only did he mention the license, but he mentioned his lawyers and carbon copied the e-mail to them. Whether the threat was intentional or not, it was a ham-fisted way to approach the situation.

That Stallman himself had to contrast what he intended to a LegalThreat indicates he relies on that blunt tool too much. To provide an analogy that those in the free software might appreciate, constantly framing disputes in a legal sense with a contingent of lawyers standing behind you is the same strategy as when the RIAA sends CeaseAndDesist? letters to bully small time music distributors around. Even if the intent isn't to bully, it isn't an effective communication strategy. When SunirShah attempted to explain the ethical ramifications of copying Meatball content by speaking through the vocabulary CopyrightLaw, a language he thought CopyLeft people might understand better than ethics, it spawned the CommunityWiki fork. For those who AvoidLegalRisk, legalese will perceived as an attempt to ExpandScope in the conflict, even if it is merely an attempt to find a CommonContext or vocabulary for a discussion. It was unfair to even begin with a LegalThreat during the initial communication. Even if we might most charitably surmise that Stallman did not intend to make a LegalThreat, it may be the case that after his many license disputes, his diplomatic vocabulary is now limited to license disputes.

Rather, given Fiddes' contributions, the FSF could have either ignored the situation, or tried to broker some sort of solution--like hosting Fiddes' patches and binaries on an FSF server that is already distributing full source. Ignoring the situation may have been wisest of all. Since relatively few people use the Coldfire compiler extensions, it is hardly a high-profile case suitable for setting an example to the public.

Conversely, the problem with this case is that we have only Stallman's admittedly poor initial letter and then Fiddes final, public response. What was Fiddes initial response? Did he immediately close up shop? That may be unlikely, but if so, the blame falls on both participants. Fiddes might have sent back an email explaining why he was unable to distribute the entire source, and ask if the FSF could help him with that. If this was done, and Stallman still insisted that Fiddes do something to correct the problem without offering any help, then it would be justified to throw in the towel.

Still Fiddes' reaction was overly dramatic and rapid, it is a reflection of what a certain contingent of people who AvoidLegalRisk will do in response to anything that even looks like a LegalThreat. In a way, the RightToLeave is meant to keep overly dramatic people out of your hair. On the other hand, the RightToLeave also balances out problematic leaders. In the case of the GPL and the FSF that controls it and Stallman who controls both, though, there is no meaningful way to leave since they are the only worthwhile games in town.

4. Public relations

When dealing with Stallman, it's important to remember that he has very very little time, often due to his own poor time management skills. He reads about 400 mails a day. He gives speeches all around the world. He managed to continue discussing Emacs maintenance issues on emacs-devel evethough he had a broken arm in 2003. The problem is that there is not much money going around, and not many people have the same firmness of conviction you would need to satisfy Stallman's requirements. As AlexSchroeder writes, "It's a typical problem for leaders, in my humble opinion: Without this fanatical drive, you would be unable to dedicate such a big part of your life to the mission. The same quality makes you hard to cooperate with. Many people (me included) have just adapted to the situation."

EmacsWiki has even collected helpful hints to communicate with Stallman on EmacsWiki:RichardStallman. But that page raises some questions. For SunirShah, it reads as "accepting a man with power over you who is obsessive, and then coping with him. Why does he post to the EMACS list at all any longer when he has more important things to do; and why bother with a broken arm? . . . I would never tell a friend he or she should deal with a person like that. I would rather drop him out of my life." (DissuadeInteraction)

In any case, Stallman has a history of being difficult to work with, and he's even admitted this. No one should have to make excuses for his lack of social skills. Stallman seems like the only very busy person who requires special handling instructions, at least on a volunteer project where no one should have enough power to demand such aides. As StephenGilbert puts it, "as if the person was a program and I needed to learn the interface." To borrow from FidonetPolicyFour, while people shouldn't be excessively annoyed, Stallman is excessively annoying: he can't communicate without being curt, impolite and using expressions like "The decision is final" in order to "save time".

And this isn't to say Stallman is overly nasty (at least in writing), although his poor style tends to be received poorly, and he garners a lot of flames for it. As AlexSchroeder says, "at least I've never seen him insult anybody via email, whereas I have seen lots of people insult him. He has never been abusive to me. I don't know anybody who has been abused by RMS. He is curt, he is sometimes impolite. But that's it." Of course, an annoying person can normally be ignored. If you had just met Stallman in the street and he treated you like that, you would just move on with your life. For many, when snubbed by those with more power, they will respond in the only way they can: verbally. PowerIsCriticism.

Nonetheless, thus given the information presented, and taking Stallman's past behaviour and methods of communicating into account, his letter is not an intentional attempt to snub, bully, nor intimidate, and the final clause seems to be a sincere attempt to assure Fiddes that Stallman does AssumeGoodFaith and is not making a LegalThreat. The fact that it manages to convey the exact opposite meaning is because of Stallman's "abysmal social skills," to one commentator. That is, telling volunteers how to behave is demeaning and disrespectful.

Not everyone knows his past behaviour and methods. That's just not an excuse to avoid the PrincipleOfConstantRespect. Perhaps it's difficult for people to understand what it's like to not know him at all, but the first impression he gives off to other people is incredibly negative. As the frontman for the CopyLeft OpenSource movement, this really puts people off OpenSource and CopyLeft. He's particularly agitating when he yells at people, as he has done. For those wanting to support OpenSource in more mainstream environments, his behaviour makes supporting OpenSource or OpenContent is embarassing, and frankly not wise. It's not a good idea to advise people you like to put themselves in legal risk.

If that's the case, then we may conclude:

Therefore, he should let someone else do most of the speaking for the FSF.

However, there's a problem. Many people struggle trying to separate Stallman's personal positions from those of the FSF. In the area of software, there is no discernible distinction. Thus, he either must do the speaking, or find people who won't disagree with him to speak. the former is easier than the latter.

But that doesn't seem like the full story either. What social skill in particular was lacking? Reading more carefully, the substance of the letter doesn't support the idea that Stallman meant well -- there was no offer to work with or to help or any other hint at the possibility of assistance or compromise -- just "you need to correct this." These are words that are used by someone exercising power.

Sidebar: The wider issue. One theory put forward by SunirShah suggests that excusing Stallman's behaviour as "a lack of social skills" seems to me euphemistic at best. He's simply a child of the network. We all have our reasons to be online, usually bad; according to him, his was ["crushing loneliness."] SunirShah writes, "I wonder if he doesn't have so many followers because many of us on the net suffer the same afflictions, and the camraderie of OpenSource gives people something to do." [ed: This discussion continues on StarlaPureheart]

5. Power

The questions naturally arise from this discussion, How has Stallman become president of the FreeSoftwareFoundation? And for that matter, as president of FSF, and suffering the absence of social skills as described above, why would he not delegate the responsibility for such communications to others?

To answer the first question, he became president of the FSF because he basically is the FSF. He put together the philosophy, determined the goals, and the reasons why they were right and good. He created the GeneralPublicLicense and started the GNU project. He contributed vast amounts of code to the GNU project. He's a one man whirlwind. But that much energy comes with a flipside. Various former GNU volunteers and others who have worked with him claim the need to micromanage is one of Stallman's characteristics. To repeat what AlexSchroeder says again, "It's a typical problem for leaders, in my humble opinion: Without this fanatical drive, you would be unable to dedicate such a big part of your life to the mission. The same quality makes you hard to cooperate with. Many people (me included) have just adapted to the situation."

If the question was about Stallman's power within the FSF, he necessarily has to share that for domains outside his core expertise (programming). As far as the GPL and other GNU licenses go, Stallman is not the sole "master". He relies on laywers, particularly EbenMoglen?, to help the FSF develop them. Moglen is an excellent balance to Stallman. While his views many been seen as equally extreme, Moglen is as social as Stallman is not, and is better able to relate to people.

The real question is his power over other volunteers "out there" in the world. Because so much content already exists under the GPL, particularly the GNU project--a program-by-program replacement of the Unix system--it's natural for people to want to draw on this existing rich base of content. This was what led Wiki:LinusTorvalds to license his Linux kernel under the GPL ever so long ago, and which has in turn led to an explosion in material under the GPL. While the GNU project was initially a project under Stallman's almost complete control and auspice, it has since outgrown him to encompass a global movement with major political implications (including the overthrow of some governments).

The caveat is that once you become a member of the CopyLeftNation?, you cannot leave. Projects under the GPL stay under the GPL forever. Admittedly, this isn't a unique problem to the GPL, as other CopyLeft licenses suffer from CopyleftIncompatibility?; and although people are working to fix this problem, for now that means Stallman has a large degree of power over them since he controls how the license--like their OneText, their Constitution--changes over time. Given the nature of the GPL that allows one to license "under the GPL version 2, or any later version at your option," all GPL software is affected by changes to the license that Stallman makes.

Reinforcing this, the FreeSoftwareFoundation also accepts and demands contributors rescind their copyrights to them as a prerequisite to contributing to the GNU project so that they might make credible LegalThreats. So if you contribute to Emacs, they won't accept your patch unless you assign your copyright to the FSF. The GNU project is only one part of the GPL corpus, however, and you are free to license your software using the GPL without handing over your copyrights, which most people do in order to avoid the paperwork. Still, the point is, they can and will sue you, which is bad news for those who AvoidLegalRisk.

Without this much power and say, no one would have had to make a list of instructions on how to "cope" with him, such as EmacsWiki:RichardStallman. Each of us will know when our power has exceeded our capacity for effectiveness when people start distributing suggestions on how best to deal with us. It has happened to many of us, and the inevitable subsequent chain of events are never particularly pleasant. Therefore, DevolvePower sooner.

But that brings us to the second question. Why doesn't Stallman DevolvePower, DelegateResponsibility (particularly regarding public relations), and otherwise give up control? Certainly many feel frustrated that Stallman's intolerance and inflexibility threatens the credibility of the OpenSource movement to the commercial and public sectors. Further, the current culture of CopyLeft teaches and perpetuates offensive or anti-social means of ConflictResolution (e.g. LegalThreats, forking) that a more devolved power structure would, in some way, check and balance. Or at least it would dilute Stallman's influence. As leader and RoleModel for the OpenSource movement, there is a large contingent of people who duplicate his foibles and style. When people from other sectors of society interact repeatedly with this attitude, they can't help but get a bad taste for the whole affair.

Stallman, however, objects to the request to make the FreeSoftware movement more tolerant. He's expressed his [concern] that the "OpenSource" movement might obscure his views and make his task harder. As stated on that page, "For the Open Source movement, non-free software is a suboptimal solution. For the Free Software movement, non-free software is a social problem and free software is the solution." The OpenSource movement--led by EricRaymond--is an attempt to appear more credible than free software by removing the politics--which is a reason for actively trying to not DevolvePower. The interesting question here is has the competing OpenSource movement effectively functioned to devolve power enough--given that is largely the rationale? The answer is probably not, since it's not devolved by the owner, and most of the power still resides with Stallman, the originator and controller of the GPL.

Rather than trying to build a vibrant OpenSource movement, Stallman is actively seeking to bring about deep social and sturctural change. The GPL is massively political (some have compared it--quite seriously--to a digital-age Communism). People who do try to deeply change society, whether knowingly or otherwise, seek to take power, wield it for their own causes, and then once change is wrought sometimes (hopefully) attempt to set down that mantle. One cannot see RichardStallman's apparent goal (from the GnuManifesto?) being acheived overnight so it's not surprising that he does not DevolvePower at this time--he hasn't finished using his power.

6. Alternatives

Many causes have been and are being completely run into the ground because they have a malfeasant leader at the helm of their respective charities. Not to say the FSF is in this position, but rather, that in general accountability for charities and their management is lacking. Probably because people tend to keep giving money to charities if their "mission" seems on target, and if they are one of the main players pursuing that goal, even if their management is malfeasant. And unlike joint-stock corporations, being a "member" of most charities doesn't give you voting rights to kick out management. Perhaps the proper thing to do in such a case is to organize a competing charity whose selling point is "we do the same thing as this other, well-known organization, but we aren't crazy". You can see how difficult that might be.

However, it's very difficult to create a competing charity within the GPL CopyLeftNation? since the FreeSoftwareFoundation controls the GPL and owns a significant amount of IntellectualProperty of the GNU Project itself. The real alternative within the realm of CopyLeft is to use alternative CopyLeft licenses, but that's just more of the same problem due to CopyleftIncompatibility? (oligopoly vs. monopoly). On the other hand, you might use BSD or Apache style licenses since they are decentralized, uncontrolled, and therefore personal (i.e. no ideologies). FreeBSD is already a credible operating system, and it is the basis for the wildly popular MacOsx.

Another solution is to pressure the FreeSoftwareFoundation to open themselves more to FairProcess from those underneath them. If they sued too many of their own members, the board would lose the election and be out of power. Minimally, they might be pressured to readjust their public relations strategy to reduce the role of RichardStallman. Or at least they could be pressured to work out better strategies that follow the PrincipleOfConstantRespect. Of course, without a means to EnforceResponsibility, this won't happen, so the first thing to do is to stop giving them donations.

The legal alternative can be seen in the SCO case: SCO may be seeking to have [GPL code declared public domain]: "The implicit claim is that the GPL is not simply void but also that the act of releasing software under the General Public License is tantamount to having donated it to the public domain".

CategoryCase CategoryCopyright CategoryConflict

7. Discussion

Ok, I summarized the discussion. Now knock it down again. My frustration is simply stated: Stallman's antics make my job harder, and my equally disappointing antics already make my own job harder. (Well, I seek to learn from others' mistakes.) -- SunirShah

The thing that struck me about the discussion was (as I gave away when I inadvertantly wrote "excuses, excuses") how people kept explaining the underlying causes of RMS as if that made it ok. Sure, StallmanVsFiddes was at root caused by an excess of passion, and a paucity of social skills. Equally, MrBungle was caused by an excess of testosterone and a paucity of empathy. Still, since I believe in free will, I think that Stallman could choose to DefendAgainstPassion, or he could choose to develop his social skills, or both. That he does not is both his crime, and his punishment. --MartinHarper

I agree with Martin. The idea that it is OK for VestedContributor with strong technical skills and a long history of contributions to cultivate an abrasive personality is passé. Most such people don't realize how tiresome they are, nor how much damage they do. And there are lots of talented, capable people who are blessed with social skills as well as technical ones out there. It's Lent, so a perfect time for all of us to reflect upon our own social traits and choose whether or not to change them. --Steve

Ok, I'll take a knock at it: I think this page is silly. Stallman sends an email to Fiddes, saying you're not complying with the GPL, here's why, you need to fix it, don't worry we won't sue. Fiddes says (publically), *sigh*, here, it's fixed, and I quit. Although many of the points raised on this page are worthy of discussion, the example case used is not, and does not support those points. Notice that Fiddes quit "[d]ue in part to this [less than helpful] email from Richard Stallman" (emphasis mine); it sounds like this was just the trigger for something that was already building.

If you want to examine Stallman's rocky relations with FreeSoftware contributors, there are far better examples, such as:

Furthermore, there is much weirdness in this page:

My knuckles are sore from too much knocking. ;) To sum up: let's take the various points of discussion on this page and apply them to a better example case. Stallman "vs" Fiddes is a non-starter. -- StephenGilbert

I only put the SCO claims here (the "unconstitutional" claim is the extreme end: the claims of unenforcability or selective enforcement are very slightly less insane) because I thought it was interesting in a kind of "what if?" manner. --MartinHarper

I think there are some unique things about the Fiddes example. First of all, Fiddes wasn't a player in open source/free software politics. He didn't choose the fight, and wasn't trying to spar. That contrasts with the other examples you cite, because they involve people who were engaging in politics. So, they knew there could be consequences. Fiddes, on the other hand, was very much minding his own business and trying to make a contribution.

Also of interest, Fiddes was adding value to GCC by adding support for a new target CPU. There can hardly be anything less controversial to do to improve a compiler. People can sit and pontificate for hours and days on the value and tradoffs of one or another optimization or error handling strategy, or the value of some particular piece of symbolic information for the debugger, or the merits of one or another object code format. But for a new target, there's no room to criticize the endeavor. Stallman, original author of the GCC, was virtually eating his own children in choosing to harass Fiddes, of all people.

I think this is an example of exactly the sort of community problems that the free/open software movement will have to solve to achieve its goals. And Stallman isn't the only one. The lack of respect for new contributors is widespread.


"He didn't choose the fight, and wasn't trying to spar."

There was no fight. Fiddes was unknowingly in violation of the GPL. Stallman pointed this out, as the representitive of the copyright holder (the FSF), and emphasised that he knew it was not an intentional violation. How can this be interpreted as Stallman picking a fight? It's not as if Stallman was sending notices out to non-GNU GPL projects. The FSF has consistantly refused to get involved in any GPL violatons in which it is not a copyright holder.

"Stallman, original author of the GCC, was virtually eating his own children in choosing to harass Fiddes, of all people."

He sent a single email pointing out a problem with a GNU project. Was it less helpful than it could have been? Sure. But "eating his own children"? Come, now. You or I could have easily made the same mistake in an initial communication, and all it would have taken to fix it would be a reply from Fiddes pointing out his situation. We would offer him an account on [Savanna], and the problem would be solved.

We both agree that there are some serious community problems in the Free/Open Source worlds. But this just isn't a good example, especially with this page placing most of the blame on Stallman's shoulders. -- StephenGilbert

Well, we don't know the rest of the story. Fiddes may have replied to Stallman, or not; he hasn't shared any additional communication and does not reply to emails on the subject. I think the example is still valid for the reasons I have stated, though I will admit to a little hyperbole in making my points.

One aspect of this is that Fiddes is in the UK. I wonder if perhaps, in the UK, speakers and writers measure their words more carefully. It seems that the incidence of unintentionally offensive writing (and speaking) is lower there, and in like fashion there is less tolerance when it does occur.

And to your point that "[y]ou or I could have eaisly made the same mistake" -- I don't think that's entirely true. Stallman is, by his own choice, a leader. His words carry weight, and in being a leader, he has a responsibility to watch what he says (PowerIsCriticism). If he can't, he should delegate such actions to someone who is able to do the job properly. Or he should be removed. That is what would happen in any other organization. Diplomacy is a skill and can be learned. I sense that Stallman doesn't think it's valuable and chooses to speak his mind without regard to the consequences.


My opinion concurs mostly with StephenGilbert's. In addition there is a load of hyperbole that, even if one doesn't subscribe to the NeutralPointOfView, goes over the line that i tend to draw, especially when discussing subjects that in themselves already contain a surfeit of passionate overreaction. Aside from that, i am feeling deeply uncomfortable with the amateur psychoanalysis of persons who're not even participating here, and can't bitch-slap the offender. --AlixPiranha

We don't do the NeutralPointOfView here, but we should apply AssumeGoodFaith more readily to Stallman. This case has a lot of problems, not least of which is the lake of detail, and the implicit assumption that David Fiddes was in the right. The discussion was worthwhile though, but it's not a Meatball-quality discussion (mostly my fault). We might criticize rms more in the future, though. Although, while I'm convinced the copyleft movement is flush with assholes, owing mostly to rms and figureheads like him, I'm starting to think LimitTemptation should be applied before we attract more assholes to Meatball. -- SunirShah

It's worth saying explictly that people who know RMS personally (I do; have since the late 1970s) pretty much universally consider him a first class technical genius, and one who has done a lot of good work for the world in obvious ways -- but also pretty universally agree that socially he's a serious asshole, there's no two ways about it, and any soft peddling of this point (calling it "lack of social skills" is a serious understatement even by the low standards of programmers) will lead to misunderstandings.

From the beginning this has caused any number of problems for FSF/Gnu, and it will continue to, and it's just inherent. Talking to him won't change anything in any regard, he is the way that he is, and he's a big mixed blessing. BTW this is not just a matter of his zealous philosophy; he literally is not fit to mix in polite company. I have been to e.g. dinner parties with him. Big mistake, seriously; do not invite him to social occasions attended by anyone except GNU-sympathetic extreme technophiles, and even then, expect him to behave inappropriately and revoltingly with any females present. He did the world a huge favor by spearheading FSF/GNU and continuing to act as a focal point for such things, but this is clearly despite his personality (aside from obvious qualities like persistence, zeal, stubborness). You can check all this with anyone who's had him to parties or as a houseguest.

The point is that analyzing defects of Stallman's interactions with people is always pointless. It's the time that he plays well with others that are noteworthy. -- DougMerritt


MeatballWiki | RecentChanges | Random Page | Indices | Categories
Edit text of this page | View other revisions