Google's influence over the free software community: a culture of threats, sanctions and bullying

People have been asking victims about this on a daily basis, leaving little choice but to write down the details about what really happened. The increasing public scrutiny of these issues is one of the sad consequences of leaders in various organizations failing to talk to people and failing to resolve sensitive issues fairly and privately.

In January 2018, the last FSFE Fellowship representative elected by the community before elections were abolished, who is also a Debian Developer, heard a story about a Fellow who had asked questions about Google's donations to FSFE. The Fellow distributed a leaflet at the FSFE table at the 34C3 event questioning the relationship between FSFE and Google. An FSFE staff member, who is also in the FSFE "CARE" team, ordered him to remove the leaflet from the booth. He complied and started distributing the leaflet from another location in the event. Even though he was a volunteer, not staff, this was seen as insubordination. FSFE staff, their salaries are partly paid with Google money, forcibly excluded him from his local FSFE Fellowship group. An email came out of Berlin with the rather sinister title Konsequenzen and was discussed vigorously in the FSFE members private list (the "GA" in FSFE-speak).

That email announcing the "CARE" team told us that reports can be sent encrypted and would be handled confidentially, yet the email thread below shows how all the members of FSFE were invited to comment on the Fellow in question. How would you feel in this Fellow's position?

Google's funding to FSFE is a non-trivial sum. It is more than a staff member's annual salary. The implication is clear: if Google withdrew funding, one of the staff would have to leave or be sacked. FSFE staff know which side their bread is buttered on. This is the taste of corporate influence.

It may be acceptable for a free software organization to accept money from Google, FSFE's mistake may simply be that they have kept hiring people every time revenues increase. Maintaining a lower ratio of salary-to-revenue would preserve the organization's resilience to influence from the larger sponsors. A debate about budgeting could never take place because the budgets are hidden and discussions like this are shut down.

This incident and the severe punishment demonstrates the real nature of the "safety" teams and "codes of conduct" being established in many free software organizations: protecting the incumbent management and protecting the heavyweight corporate sponsors from being held to account. Silencing dissent. Creating an illusion of happiness rather than talking openly about the truth.

The elected representative of the community had a clear mandate to question what was going on and stand up for the rights of the Fellow concerned. FSFE management immediately turned against the representative and started plotting the fastest way to get rid of him.

So started a bumper year for Google influence in free software organizations.

Control and obedience

The trailer for Das Experiment makes it really clear what is going on in some free software communities today when a volunteer puts on a uniform or badge and starts talking about enforcing the so-called code of conduct:

Enforcement is rarely the right way to resolve a dispute. Even when it is, the arbitrary volunteers who step up for these roles don't have the training or experience to do any of what they claim to be doing. In the end, they simply end up hurting people, like the randomly selected guards in the real-life Stanford prison experiment.

Yet the people volunteering for these vigilante roles in free software organizations are not always random. Some are quite political and militant with their own agendas. They don't wear a uniform to maintain order, they wear a uniform because they see it as a way to get what they want.

The safety mission is an important one: by using their powers for political agendas, these teams lose people's trust and undermine the real mission.

Google and Outreachy

Shortly after FSFE's Konsequenzen scandal, a discussion started on the Outreachy mentors private mailing list. One of the administrators was concerned that important announcements were not reaching many interns because of mail filters, for example, the infamous Google "Promotions Tab". To work around this Gmail flaw, they wanted to implement a completely new web-based Outreachy discussion forum. It would become mandatory for all Outreachy interns and mentors to log in regularly. Some mentors were concerned that this would undermine efforts to integrate interns with the communications platforms used by the host communities. One mentor raised another possibility: why not simply insist that interns provide a non-gmail address when registering, rejecting gmail addresses?

The discussion went off the rails. A member of the Outreachy team complained that interns couldn't be expected to run their own mail servers, although nobody had actually asked for them to do so. Extrapolating from non-gmail addresses to running your own mail server is a hideous misrepresentation of the original idea. Twisting somebody's words like that to discredit their opinion is just another form of bullying, most commonly used when the bully's own position lacks credibility and can't be defended by any logical argument.

This was incredibly disappointing for those who believed that we were all on the same team and that diversity was the goal of Outreachy. Mentors said little about this storm-in-a-teacup but sadly other people raised it again many months later, blowing it out of all proportion.

Many interns successfully use alternative email accounts with providers like Protonmail and Riseup. What, then, is the real reason for the outrage that erupted? Well, Google is also an Outreachy donor and the two work closely together. Some Google funds trickle down into the paycheques of Outreachy and Conservancy staff. Even if blocking gmail addresses would be the easiest technical solution to solve Outreachy's problem with the Promotions tab, it was unthinkable. Further discussion was shut down and heavy-handed intimidating emails were sent.

Let me make that clear: the person speaking up about Google was bullied and belittled by a member of the Outreachy organizing team. This is not an incident that involved an intern, it was bullying of a volunteer.

At the same time, and completely by coincidence, one of Debian's internwrote an amazing blog about how she feels about Google. Meeting people like Renata has been one of the highlights of Debian participation in Outreachy.

Outreachy does a lot to raise awareness about diversity issues in technology. But how will women ever be empowered if seventy-six percent of applicants to this illustrious program are fully dependent on a company run by men, Google? How will Outreachy fulfill its mission if they argue that women can't survive without Gmail?

The volunteer concerned completed mentoring his last intern right up to the end of her internship and then he didn't volunteer for Outreachy mentoring again, amongst other things, because Google Summer of Code (GSoC) was about to start. The Outreachy mentors page starts with a nice photo of two former interns from Outreachy and GSoC: Urvika and Pranav were both mentored by the same mentor referred to in this post.

Debian and GSoC

Debian's GSoC delegation is long out of date. Three out of four admins had retired after putting in a huge effort over many years. The only remaining member of the delegation is Molly de Blanc.

We would have preferred to publish this report without including people's names. The fact that Debian's admin team experienced a number of controversies in GSoC would not alone be enough for us to put the name of any admin here. Even the fact that de Blanc was the only admin team member to ever be subject to a complaint from an intern would not have motivated us to name her publicly. What brings her name into the spotlight is that after all this, she has side-stepped into the "anti-harassment" team where she appears to now have immunity. That team has been hurting other members of the community with their vigilante tactics.

The ongoing harm that "anti-harassment" causes to other volunteers through their tactics and political agendas means they deserve an extra level of scrutiny, just as the police are subject to scrutiny from internal affairs and external watchdogs.

I have never personally received a complaint or reprimand from Debian's "anti-harassment" team. My concern about their misconduct is based entirely on the reports I've heard from other victims of such teams in Debian and other communities. A number of these people confided in me personally after I spoke up about bullying in Debian.

In 2017, Debian wasn't accepted in GSoC.

In 2018, de Blanc tried to organize it at the last minute.

I mmissed this on the application before! We need 2-5 administrators for the application. Who else wants to be one?

As the code of conduct tells us, we should assume good faith. So better late than never. But maybe we just weren't ready.

Another experienced mentor wanted to be helpful. But he couldn't fully commit to the role. So he replied to de Blanc privately, telling her:

You can use my name temporarily while looking for other people to help you in this role... However, I can't officially commit to help with the duties of an administrator right now.

What none of us knew at the time, but only found out later, is that de Blanc wasn't going to be fully committed to the team either. This is disturbing given that other mentors were also telling de Blanc that they couldn't fully commit. At least others had been up front about their status.

On 14 February, the same Debian Developer had written to Chris Lamb, the Debian Project Leader (DPL), asking if he could fund anybody else to go to the Tirana BSP. For various reasons, he didn't feel he could go but he also knew that all our Albanian friends put a huge effort into organizing events and he felt it would be really bad if their BSP went ahead without a developer. Lamb's response was negative and appears highly unsympathetic.

On 16 February, the same Debian Developer sent another email reminding the other members of the GSoC admin team and Lamb that he was not fully available for the admin role.

On 19 February, another member of the admin team wrote that they were only participating as an admin so that their own project would go ahead.

24 February, he advised Lamb and the FOSSASIA team that he would be unable to attend due to serious reasons outside the project. He didn't give any reasons but he wanted to make sure the previously authorized travel funds would be available for another developer if they wanted to go.

March 1, the Developer was in Tirana, Albania for the very successful Bug Squashing Party (BSP). Several people went to him personally about challenges they were facing. He wrote to Lamb about some things and Lamb sent back a sharp reply accusing him of being a gatekeeper. Lamb appeared to be insisting that every tiny little query has to be submitted to him directly by email, even if another volunteer is out in the field dealing with people face to face. The reality is, many people are far more likely to contact the volunteer who is present rather than some stranger like Lamb in a remote country. That isn't gatekeepering, it is just human nature.

March 4, the day after the BSP and participants from outside Tirana were leaving their accommodation. Lamb had sent them emails authorizing their travel and accommodation but given them no instructions about payment. On a Sunday morning, many of them were nervously contacting the local organizers and the sole Debian Developer in the city asking how to pay the hostel. The Developer volunteered to go over there, gathered up all the documents for FFIS and settled the bill with cash. In all his history of doing things with Debian, this is the first time that a reimbursement had not been paid to him. It wasn't even his own expenditure, it was money that Lamb authorized for other people and the volunteer paid it to help them out and ensure a smooth experience for Debian's partners. It is in RT ticket #7180 and there has been radio silence from Martin "Joey" Schulze at FFIS for almost a year now.

March 5, he sent a thank-you email to one of the local team, CC to Lamb. Lamb appeared to resent the relationships the developer was establishing with people in the Balkans and moaned about this email:

I'm writing to thank you personally for the tremendous effort you have put in to making the Bug Squashing Party a success in Tirana over the weekend. You have only recently joined Open Labs and if I understand correctly, this was the first event you were involved in organizing. People are already talking about it around the world. The high ratio of female participation was an outstanding feature of the event and your ongoing contact with the women from Kosovo was a major factor in that. Without your decision to initiate this and your persistence in contacting people locally and in the wider Debian community it may never have happened. I'm sorry I wasn't able to give you an earlier confirmation of my own participation, you did everything as well as you possibly could. Do you know if anybody from the team will write a blog about the event? ----- already put the photos online, I would write something myself but I would prefer to let somebody in Albania or Kosovo be first.

Since when is the Debian Project Leader the only person with a right to send thank-you emails? Since when does he have the power to prevent other developers sending thank-you emails?

The developer wrote an email suggesting to Lamb that maybe they should meet in person to talk through whatever issues were at play, breaking out of the cycle of unhelpful emails:

It seems we are both sometimes disappointed with the communications between ourselves. We both believe in the same things and we both believe in the integrity and reputation of the Debian project. Maybe the mode of communications isn't ideal. Could it be better for us to find an opportunity to discuss things in person perhaps? I am usually in the UK once per month, usually around Herts, currently I'm here until Thursday

Lamb refused.

During March and April, Debian Developers who volunteered for GSoC mentoring received hundreds of emails, some privately, some on public lists, about GSoC participation. The most well known mentors have been a victim of their own success, each year, their projects have become more popular and the number of applicants has snowballed. This is a situation that is difficult for any mentor to handle, especially with everything else going on in life.

One former intern had also seen de Blanc's call for new admins and replied to de Blanc privately. de Blanc had accepted them into the admin team too, the former intern's mentor hadn't personally been aware of the process de Blanc followed onboarding the admins for GSoC.

Given the way the team had come together, even though one of the admins had expressed strong reservations about participating, he decided to try and continue, hoping that the larger group would be able to share the workload. He also felt that as some of his former students were keen to participate again in various ways it would be helpful for him to be around and support them. He did this entirely in good faith.

In April, one of the admins expressed concern about Debian being an "umbrella" organization, but Google explicitly permits that and some other members of the team were in favour. What's more, many potential mentors and students had already put in significant effort believing Debian, being a Linux distribution, would perform the umbrella organization role as Debian had in previous years. It wasn't something we could radically overhaul at the last minute. Nobody in the admin team had put up any constructive proposal for selecting or prioritizing projects, in fact, none of them had even remembered to send an official email calling for mentors because everybody thought Molly de Blanc was in charge.

On 10 April, admins received the email from Google advising that Debian had been granted funding for 29 students in GSoC 2018. This reflects the enormous amount of work some volunteers had put into building up a mentoring team and Google's confidence in them.

Despite hundreds of emails, an enormous flow of information, one particular fact stood out. When admins had a team IRC meeting on 16 April to confirm selections, one admin made a point of explicitly reminding fellow admins about it. As it had come up several times, he believed they were already well aware anyway. (One of those Debian conflicts of interest, which became public after this report on bullying). Molly de Blanc, who was the only remaining member of that delegation, immediately acknowledged it with the comment:

<mollydb> nice responsibile decision making
<mollydb> thanks for being so consciencious

On the admin's part, nothing was ever hidden from the rest of the team. If they got things wrong, it was a team mistake, no one member should be used and abused as a scapegoat.

For the most experienced mentors, thousands of GSoC-related emails come through their inboxes each year. One that stood out was a comment from a student who wasn't selected: he told us this was his last chance to participate in GSoC as he is about to graduate. This got him thinking: "when I was a student, there was no GSoC. In fact, when I started my undergraduate studies, Google didn't even exist. We set up a community network and successfully applied for a $100,000 grant to run Debian." Is there a danger that programs like GSoC are preventing students from aiming higher? Another example of corporate benevolence having side effects.

Lamb's gatekeepering anxieties continued to grow. In May, once again, the same volunteer was the one out in the field with some of Debian's new GSoC interns and he submitted a single request for a group of expenses. He wasn't being a gatekeeper, he was simply being expedient given that Developers only have a finite amount of time to contribute to the project and would prefer to spend it on development rather than bureaucracy. Lamb snapped back:

Can I ask why they do not contact me myself?

After this incident in May, the volunteer decided not to make any further requests to Lamb. It just didn't feel like Lamb was being respectful at all.

At the same time, while he was the one out in the field, the FSFE High Command in Berlin was having an extraordinary general meeting behind his back to cancel elections and try to pass an obfuscated motion expelling the current representative, the same Debian Developer, without any due process:

The chair asks the members to vote on how to deal with existing Fellowship representatives and puts the following options to a vote:
  1. The current Fellowship representatives' membership ends as soon as the constitutional change is successfully registered, or 2 years after their election, whichever comes later.
  2. The current Fellowship representatives' membership ends immediately after the next ordinary General Assembly.
  3. The current Fellowship representatives' membership ends as soon as the constitutional change is successfully registered.
Ulrike Sliwinski (staff) requests a secret vote. The Chair asks who else is in favor of a secret vote. 3 members voted for a secret vote. This does not meet the threshold of 1/3 for having a secret vote.

This isn't the first time that a dictator in Berlin decided he was above democracy and due process. It isn't the first time that a group of Germans in Berlin felt they were entitled to speak for all of Europe. Despite the "E" in the name, FSFE is predominantly a German organization. Despite the "FSF" in "FSFE", they also broke their deal with FSF.

By removing elections from the FSFE constitution, Kirschner demoted 1500 volunteers, including the Debian Developer who was elected as a representative, from being Fellows to unpaid interns.

Any trust the Fellows had in FSFE and Kirschner was vaporized by that meeting. For the Fellowship representative, resigning as fellowship representative became a question of when, not if.

In June, with GSoC now getting into full swing, mentors suddenly find themselves in a position where they are trying to work out how to apply Google's rules. Numerous queries are sent privately to admin teams like Debian's and also on the mentors private mailing list, where over 1,000 volunteer mentors are subscribed. Each year mentors witness some examples of bullying and discrimination. In one of the more outstanding examples, an intern had confided in their mentor about a mental health problem. The mentor circulated the details to the full GSoC mentors mailing list:

We have a student who was doing good work, passed the mid-term review, then disappeared shortly after. Following up, one of the org admins had a discussion with the student and discovered he was being treated for (suppressed). (Personally identifying details withheld)

Mentors from various other organizations started speculating about whether the illness was real or how to deal with it:

For example, giving health advice:

send the poor kid a copy of ...

and attacks on the intern's integrity:

The rule of thumb is: students always tends to have serious accidents and hardware malfunction just on the day of their exams. It might be harmful for some cases, but statistics don't lie.

Not all mentors are like this. Several others were quick to point out the discussion was inappropriate. Some took their time to write about more constructive attitudes to mental health challenges.

Regardless of what support the student received, would Google allow their own employees' medical histories to be debated by 1,000 random strangers like this?

How can interns trust their mentors and program administrators when this type of thing is going on?

Google management eventually replied that the mentor didn't have to pay the intern for work already done, or in GSoC-speak, they could "fail" the intern.

Then there was this...

In several cases each year, we've either seen mentors threaten interns or seen mentors recount stories about how they threatened their interns. One mentor shared this strategy for keeping students motivated:

I'd recommend hangout with the student, get exactly what their commitments are and when and then manage tightly with the threat of midterm fail.

and another mentor appeared to be implementing that with an email like this:

Subject: Final Warning Mail I didn't want to write this and tried my best to avoid writing such kind of mail from long. ... Treat this as a final warning, that if you're not able to show considerable progress in the coming weeks, and complete the project according to scope decided earlier, within the GSoC duration, we'll be *FAILING YOU*.

The email did nothing to help the interns understand where they were going astray or how to get back on track. It was sent on the weekend when many people would probably prefer to be resting. Is Google encouraging a 24x7 culture that is harmful to both mentors and students?

Having seen this so many times, it isn't fair to blame any individual mentor for this type of communication. It is a cultural problem in the program and in some free software organizations.

Once again, this shines a light on the effects of corporate influence. Mentors are clearly concerned with appeasing Google and each year some go too far. The culture of the organization contaminates the community.

When some of us were students ourselves, developing our own solutions with Linux and free software in the nineties or earlier, there was no Google and we never saw threats like this.

The Debian Developers who volunteer as admins become mentors to the mentors and try to help them find more effective ways to motivate their interns. This can be both tiring and rewarding and as it is usually done through private communication channels, there is rarely any recognition or thanks for this effort.

In July, the same admin/mentor/Debian Developer who withdrew from FOSSASIA also informed Stephanie Taylor, head of GSoC at Google, Chris Lamb and the rest of the GSoC admin team that there had been extraordinary personal circumstances that had an impact on his role as a mentor. Given the utter lack of privacy and respect in the community, he didn't give any more details than that. Taylor replied, only to Lamb and the mentor, suggesting he take a rest from mentoring. That was hardly an unusual response in the circumstances. This private exchange never should have gone any further, let alone used opportunistically for political purposes, bullying and harassment.

Nobody bothered to ask what was wrong or how Debian could deal with that as a team. People only expressed frustration and blame, much like the first-time mentors expressing frustration with their students.

At the same time, Molly de Blanc admitted what had become increasingly obvious to other team members:

I generally check my email once a week.

In other words, given that GSoC and Outreachy generate such a huge volume of email, most of it deserving prompt attention, de Blanc was avoiding it. de Blanc had signed up other volunteers to participate and then quietly stepped back in the hope that other people would do the work.

While admins had sensed de Blanc was a bit flippant about her responsibilities, none had ever said anything publicly. Nonetheless, one of the Outreachy interns wrote a damning email to de Blanc on 5 August, before the payments had even been made. Incredibly brave.

The main thing I want to note is that you do your work not so good. You haven't responded to me and also during the last round of Outreachy you have provide almost to no response to applicants of Outreachy FSF project. I asked some of them and they said that you haven't helped them: . Though, you were a mentor of this project. At the same time, I can't see where you were useful for me. You haven't answered to me, you haven't answered to applicants. So, it would be great if you can improve your usefulness.

Incredible stuff. de Blanc replied to the intern:

My work with the FSF and my work with Debian are two different things, and ideally need to be kept separate from one another. They use different parts of my time, come with different responsibilities, and are managed by different people.

Other members of the admin team didn't find de Blanc's reply helpful.

A number of mentors and students reported feeling unwelcome in the Debian community or at Debian events. Some stalwarts seemed to be trying to belittle those who were not "real" Debian Developers working on their "real" Debian tools. It is remarkably similar to the "us v them" phenomenon described in Amnesty's bullying crisis.

Nobody saw fit to make these issues public at the time. The admin who had earlier reduced his role for personal reasons simply wrote an email thanking the team and advising he wouldn't volunteer again in 2019. There is an expression that comes to mind: let sleeping dogs lie.

A number of things have changed in GSoC and Outreachy over the years and it is not the same thing any more. For example, both programs have become more and more like jobs, where interns have fixed working hours and response times for answering emails, yet they are not being paid like a job and they don't have basic benefits that everybody else has like sick pay and accident insurance. Some mentors are no longer comfortable encouraging students to engage in that.

On 30 August, de Blanc went even further, telling admins:

Filling out the GSoC paperwork and writing a call for participation was a response to community desire and no one else stepping up to get that started.

As de Blanc is named in the delegation, most people thought she was responsible for a lot more than simply filling out the GSoC paperwork at the beginning.

In September, Matthias Kirschner at FSFE started sending fresh threats and accusations to try and extinguish the last trace of an elected representative, the same Debian Developer who had just withdrawn from his GSoC responsibilities for personal reasons. Kirschner's menacing emails set a deadline on 20 September for him to respond to an accusation that he had "broken the bond of trust" in FSFE. That accusation is easily refuted: Kirschner had broken the bond of trust himself in May when the representative was out in Kosovo doing real free software activities and Kirschner tried the backstabbing vote, holding that extraordinary general meeting in Berlin. Therefore, how can Kirschner accuse the volunteer of breaking something that was already broken by Kirschner's own actions five months earlier?

The FSFE constitution requires a reason to be given for terminating a member and it also gives the member a right of reply. No reason had been given and due process had not been followed when Kirschner had people vote on the representative's membership in May. The original vote had failed and Kirschner had made up this false accusation so he could have a second vote to try and remove the last representative. Forcing a member to go through two votes on their status is harassment by a sore loser who didn't get what he wanted the first time.

Note to Kirschner: harassing the developers doesn't help free software.

On September 18, SPI sent an email announcing that Google would shortly send $17,200 to Debian.

There was never any team de-briefing or review of problems Debian's admin team faced.

On September 20, just hours before the deadline set by Kirschner at FSFE and just as the Google money was hitting SPI's bank account (which is not under the control of the entire Debian membership, because only a small number of volunteers are also SPI members), another member of the Debian community started sending abusive emails to the same volunteer. It was an extraordinary act of bullying and intimidation. A particular point to note about that email is that rather than making any reference to FSFE, they callously whinged about the inconvenience when he reduced his involvement in GSoC in July, complaining how others "had to step in to provide them support". Nobody made any attempt to inquire about his welfare or the circumstances that had also forced him to withdraw from FOSSASIA and reduce his involvement in so many other things throughout 2018. Had Lamb hidden these details when he encouraged other developers to harass the volunteer?

When the volunteer replied to that abusive email asking if they knew about those circumstances, they dismissed it as "minutiae". A truly hideous response.

The emails had the distinctive feeling of a veiled threat: that the Debian conspirators behind them wanted to exercise influence over his elected role in FSFE. That he either does as they expected or they would drag his name through the mud, as Debian has done to various other volunteers in recent history.

It was absolutely clear that Chris Lamb was involved in constructing those abusive emails and running roughshod over the volunteer's privacy and the privacy of other people in various ways.

What victims didn't know at the time was that Lamb was simultaneously sending nasty emails to other people outside Debian to try and hurt this volunteer. Can you imagine the leader of a highly respected organization like Debian trying to force out a volunteer and then immediately sending a "kill confirmed" email to overlords at Google? Did Google order a hit on a Debian Developer, was it a condition of that $17,200 payment, the DPL making a deal with the devil? Or did Lamb simply feel he needed to serve up a scapegoat to save face with the paymasters? Or was Google's name simply being used out of convenience, to add to the pressure on this volunteer at the time when FSFE was fighting tooth-and-nail to extinguish democratic representation?

Every significant discussion and decision during GSoC 2018 was escalated to the entire admin team. There were a few times when people contacted admins personally about sensitive issues and in each case the volunteer concerned told them they needed to send their email to the whole team. If there was any problem in the GSoC land, the whole team was equally responsible for it, especially the absent-minded Molly de Blanc.

If Lamb really felt that Debian needed to appease Google, he should have simply offered his own resignation. In February, the volunteer had personally told Lamb and the rest of the team that there were limits on his availability in 2018 so the buck stops with Lamb. Given the disclosure the volunteer had made in July, it was unthinkable for Lamb to try and play dirty tricks like this, at least if he was serious about being a leader.

The volunteer decided to resign from his role at FSFE. The bullying and veiled threats from Lamb had an impact on that decision. It is important to remember that the volunteer is not really the victim of that resignation: the victims are the 1,500 fellows who no longer have a representative or vote. Now FSFE can do whatever they want without scrutiny.

Even after resigning from FSFE, Kirschner continued to send him abusive and threatening emails.

The volunteer wrote to Lamb and simply requested a meeting to talk about whatever was bothering him. Lamb refused and sent a series of inflammatory emails. Lamb was the first one to start sending emails to volunteers telling them they were demoted: in doing so, he simply demoted himself from being a leader to a being a cyberbully. He also demonstrated that he was making things up as he goes along, the Debian constitution does not give him authority to demote volunteers.

Around the same time, the volunteer was still following up on his responsibilities from GSoC, adding some of the students to Planet Debian. Looking at the commit log, he noticed another long standing developer's blog being censored by Chris Lamb over a grammatical error. It occurred to him that the DPL may be burned out. This is a sign that the problems in Debian right now are not down to any one individual, the failures of individual relationships may be a symptom of an organization structure that isn't working.

In October, de Blanc, an FSF employee, attended a five-star all-expenses-paid trip to Google's GSoC mentor summit in California.

de Blanc was also promoted to become a Debian Developer and joined Debian's anti-harassment team.

During the last years, many of Debian's leading mentors and admins have personally had to assist in several stressful incidents that required a cool-headed approach to mediate and de-escalate. This type of thing is always a strain on people but leaders accept that is part of being in a leadership role in any community organization. There were some occasions where the volunteer in question witnessed things that bothered him and privately asked other members of the community for advice. In Debian, none was forthcoming.

Given recent events, several people confided in us about their interactions with the "anti-harassment" team. When we look at the reports that have been shared, we find that various patterns emerge:

The overall impression is that the "anti-harassment" team is basically taking sides and harassing people. The very name of the team creates an adversarial posture, although they have already acknowledged that. One Debian Developer previously suggested some new names for the team and made comments on the risks.

Yet what is obvious now is that the anti-harassment team is not about stopping real harassment, it is stifling any attempts to hold power to account and building up dossiers on people for future ambushes. When people talk about creating a "safe space", they appear to be creating a safe space for corporations like Google to exert influence.

What is more astounding is that given de Blanc's own shortcomings as an administrator in GSoC and Outreachy, she has somehow maneuvered herself into the anti-harassment team where she can enjoy immunity while the people who did do some work are abused and used as convenient scapegoats.

Coincidentally with de Blanc's appointment to the team, people have noticed a range of gloating emails using the title "bits from the Anti-harassment team" as well as a variety of blogs and comments from the DPL trolling victims of this team. It is extraordinary that de Blanc has been absent from her basic responsibilities in the Debian Outreach team but she attends free software events to preach about imposing a regime of enforcement upon developers.

By its nature, this type of work is normally done in complete confidence. By gloating about their actions and rubbing salt in the wounds of their victims, they demonstrate that they are more concerned with bastardization than respect and harmony. Many of these references to the anti-harassment team have a triggering effect for victims of bastardization and bullying. This is harmful for anti-harassment's victims and the community as a whole.

At Christmas, things become even more difficult for the community. The Debian Account Managers attacked another developer, trying to snuff him out under the radar just days before Christmas. At the same time, the volunteer who had been attacked after GSoC went down to the Balkans to see some of his friends there and started getting hints about nasty emails Lamb had been sending behind his back. Lamb smugly sent a public email denying he had compromised anybody's privacy:

I have been nothing but scrupulous and gentlemanly with regards to your personal privacy

yet when challenged with evidence, Lamb has gone into hiding.

This lie started an avalanche, with other developers speaking up about harassment from Debian's leadership. On Christmas day, Martin Krafft wrote:

I know that there's been at least another case, in which DAM and AH have acted outside their mandate, threatening with project expulsion, and choosing very selectively with whom they communicate. I know, because I was being targeted.

As another developer wrote, this snowballed further into "monster threads" on expulsions on the debian-private mailing list. If that isn't cyberbullying, what is?

While it has been an extraordinarily difficult experience for many people in the free software community, it has also been helpful in demonstrating the flawed nature of the "anti-harassment" team and just about every other equivalent in other free software organizations. Trying to invent our own solutions is part of Debian mythology. Yet for something as serious as welfare and harassment, none of us have the competence to deal with certain situations. Debian has basically built a vigilante group that runs around looking for witches to burn. It simply isn't fit for purpose and it is hurting people. It is time to disband it and get outside advice from people with professional experience.

Observing Google's influence in three different organizations, FSFE, Outreachy and Debian, with similar hostility arising in all of them, it is clear that these unpleasant outcomes are more than a simple coincidence.