Nine trademarks, one target: how solomining.de claimed the names of open-source projects

What happened

On 19 April 2026 Google notified us that a counterfeit complaint had been filed against bitaxe.de. A second notification about the same matter followed on 25 April. In the end 71 URLs of our shop were affected. A few days before the first notice, the same party had already filed a copyright complaint with our hosting provider. Google acted on the complaint. The pages dropped out of the index.

The complainant was Bolt Engineering GmbH of Fellbach, Germany — the company that operates the shop solomining.de. Its legal notice identifies it as “Bolt Engineering GmbH, represented by its managing director Jannis Langer, Höhenstr. 10, 70736 Fellbach”, registered at Mannheim district court under HRB 743598. They never contacted us. We learned of both actions only through the platforms they were aimed at.

What came afterwards — a letter from their lawyers, the public discussion in the Bitaxe project’s repository, and above all a look at the trademark registers — no longer resembles an ordinary dispute between two shops. It shows a portfolio of rights over other people’s project names, built up over two years out of public view, and a first exercise of those rights aimed at exactly one competitor: us. And for now it ends with the discussion being closed and locked to everyone outside the project.

Nine trademarks over four projects that are not theirs

The names at stake here come from an open developer community. Bitaxe was created by Skot / bitaxeorg and released under an open licence. Nerdaxe and Nerdminer come out of BitMaker’s work, NerdQaxe++ from shufps. All four are community projects, not commercial brands.

A search of the official registers — the German Patent and Trade Mark Office and the European Union Intellectual Property Office, retrieved through TMview — returns the following portfolio for Bolt Engineering GmbH, all in class 9:

Three things stand out.

First, the date of the earliest filing. The German word mark “Bitaxe” was filed on 28 September 2024. bitaxe.de had been selling under that name for months by then — the Internet Archive records our shop from May 2024 onwards. A first complaint against us had already been filed half a year earlier, on behalf of “the trademark owner”, at a time when no such trademark existed. The name was neither free nor unused when it was claimed.

Second, the pattern. Bitaxe and Nerdaxe were filed one week apart. Nerdminer and NerdQaxe++ followed exactly a year later, two days apart. On 20 March 2026 three EU applications went out on a single day. These are not isolated decisions. This is a programme.

Third, the reach. Four projects, four different authors, one applicant. None of the authors is Bolt Engineering GmbH.

The Nerdminer test

A trademark is meant to protect a business you actually run. It stops someone else from trading under the name you trade under. So the most revealing question you can ask about any filing is simply: does the applicant sell the product at all?

We went through the full catalogue of solomining.de — 26 items as of 24 August 2026, titles and descriptions. The result:

The German word mark Nerdminer was filed on 5 October 2025 and registered on 17 April 2026; the EU application followed on 20 March 2026. The word “Nerdminer” does not appear anywhere in that shop — not as a product, not as an accessory, not in a single description. For Nerdaxe there is no device either, only accessories for hardware that other people sell.

The justification the register does not support

There is a stock defence for trademarking open-source project names: someone has to protect the name from cheap Chinese copies. It sounds reasonable. It is also testable — because if it were true, the register would show action taken against Chinese filings.

It shows the opposite. The same names have long since been filed by Chinese companies — in Europe, in Germany, in the UK, New Zealand, Canada and the US. A selection:

The first row is already telling: “BITAXE” was registered in China on 7 September 2024 — three weeks before Bolt Engineering GmbH filed its first German application. Anyone hoping to protect the name from China was already too late at that point.

The second row is the one that matters. The EU mark “BITAXE MINER”, owned by a company in Shenzhen, was open to opposition until 28 June 2025. Bolt Engineering GmbH had held the German word mark “Bitaxe” since January 2025 — valid grounds for an opposition. They were demonstrably occupied with trademark matters in exactly that window: on 23 June 2025, five days before the deadline expired, they filed their own EU mark. No opposition against the Chinese application is on record. It was registered on 6 July 2025.

It goes further. Two Chinese companies now hold German word marks for “NerdMiner” and “NerdQaxe++”. Zhanjiang Zhuocai’s Nerdminer mark was registered on 16 March 2026 — a month earlier than the identical mark held by Bolt Engineering GmbH. Again: no visible opposition, no complaint, no public reaction.

And the argument fails twice over. An EU trademark bites on import and distribution inside the single market. It does not trouble a direct seller in Shenzhen shipping single parcels to European consumers. It works precisely where a competitor has a legal notice, a VAT number and an address you can serve papers at — which is to say, against the sellers who play by the rules.

The complaint to Google

The complaint that removed our pages is publicly documented. The Lumen database records it under number 84146531. Sender: Bolt Engineering GmbH. Recipient: Google LLC. Type: counterfeit. Target: bitaxe.de, 71 URLs. The text, verbatim:

“The website https://bitaxe.de/[…] is selling products under my registered trademark ‘Bitaxe’ without authorization. The domain name and product listings falsely suggest an official source. The site impersonates my brand, uses my trademark without permission, and misleadingly claims ‘Made in Germany’. Additionally, the legal imprint contains false information and a non-existent address. This constitutes trademark infringement, counterfeit marketing, and consumer deception.”

The wording is the point: my registered trademark, my brand. An open hardware project is presented to Google as the complainant’s own brand, and the sale of genuine open-source devices by another shop as counterfeiting. Google acted on it and removed every URL listed, which makes this not an attempt but a measurable loss: 71 entry points of a working shop, gone from search.

The two side allegations are quickly dealt with. The legal notice complained of — a phone number that ended up in our imprint through a copy-paste error — had been corrected long before the complaint was filed. We reviewed our manufacturing claims and tightened the wording where it could be misread. None of that touches the core of the complaint, which rests on a trademark over a name its owner did not create.

What the other side’s lawyers wrote

We sent a cease-and-desist letter. Their lawyers replied. This is the passage we published in May 2026, translated from the German original:

“Our client had the Bitaxe trademarks you refer to in your letter registered in coordination with the holder of the title rights to the open-source project “Bitaxe”. The enforcement of the trademark rights likewise took place with the knowledge of the holder of the title rights and of the so-called “OSMU” (Open Source Miners United), in which our client is heavily involved and which it has been actively shaping and advancing for years.

Furthermore, our client has in the past granted trademark licences to two substantially larger competitors than your client.”

That paragraph makes three claims, all of them by the other side, all of them binding on them:

• The registration happened in coordination with the holder of the title rights — that is, with the author of the Bitaxe project.
• The enforcement happened with his knowledge and with the knowledge of OSMU. They are therefore conceding that the mark was used against a competitor, and that OSMU knew.
• Trademark licences have already been granted to two larger competitors. A licensing programme exists over the name of an open-source project.

The third point is the most remarkable. Whoever grants licences decides who may use the name and who may not. That kind of control over a project’s name is precisely what open hardware is supposed to rule out.

What the same people said in public

In May 2026 the matter was discussed openly in the Bitaxe project’s repository — in the list of trusted vendors, the legitlist. On 14 May someone requested a review of the solomining.de listing, citing the Lumen record. Four days later the case was closed. Two statements from that thread cannot be reconciled with the lawyers’ letter.

The developer WantClue wrote on 18 May 2026 at 14:15:

“There is no evidence of any manufacturer in Europe need to ask for permission to sell Bitaxe. Even tho solomining.de owns the trademark it seems this company has never used it for any purpose of restricting competitors. […] the legal problem between solomining and another company exceeds the purpose of this repository.”

Fifty-five minutes later Skot, the founder of the Bitaxe project, closed the case with “I agree, this can be closed as resolved” — and stated in the same comment:

“For the record bitaxeorg owns the Bitaxe trademark and does not recognize or endorse local jurisdictional trademark registrations by others.”

Neither statement fits the lawyers’ letter. The declaration that the project expressly does not recognise such registrations stands against the claim that the registration was made in coordination with the holder of the title rights. And the assertion that the mark had “never” been used to restrict competitors was made 23 days after a complaint aimed at removing 71 URLs of a competitor — and four days before the other side stated that it had licensed two competitors.

We are not saying anyone knowingly said something untrue. We are saying the accounts contradict each other and cannot both be correct. How that contradiction was built up becomes clear when you look at the earliest event in the whole story.

A complaint for a trademark that did not exist yet

In March 2024 our infrastructure provider received a complaint against bitaxe.de. It was filed by the OSMU developer WantClue, expressly, as it says there, “on behalf of the trademark owner”.

Except that in March 2024 there was no trademark owner. The first Bitaxe mark was filed half a year later, on 28 September 2024, and registered only in January 2025. Acting “on behalf of the trademark owner” against a competitor in March 2024 means acting for a right that nobody held at the time.

Who gave permission?

The other side’s lawyers gave a clear answer: registration and enforcement took place “in coordination with the holder of the title rights” and “with the knowledge” of OSMU. There is no room to leave open who that means. Title rights in a work belong to whoever created it. The Bitaxe project was created by Skot, who released it under an open licence and who states himself that bitaxeorg holds the Bitaxe trademark. The lawyers cannot mean anyone else.

So the other side’s statement stands against that same man’s own words. On 18 May 2026 Skot wrote publicly that the project does not recognise third-party national registrations. Four days later the trademark owner’s lawyers wrote that this very registration had been made in coordination with him, and enforced with his knowledge. Both cannot be true: either the registration was Bolt Engineering’s own initiative, or it was agreed with the project’s author.

Coordination was denied from another direction too. The same developer who had acted “on behalf of the trademark owner” in 2024 wrote in May 2026:

“There is nothing we can do as this company did file the trademark on their own. Even if we would like to revoke that trademark there is 0 possibility as they did everything according to their jurisdiction. OSMU is not an Organization nor a company.”

A solo effort nobody can do anything about — that is one account. A registration coordinated with the author and enforced with his knowledge — that is the other, from the trademark owner through her lawyers. One of the two is not true. We cannot say which. We can say the question has been open ever since, that a single sentence would settle it, and that the place where it could be asked in public has since been closed and locked.

What happened in May 2026

How the review of all this ended can be read off the timestamps in the project’s public repository, to the minute:

1. 14 May 2026: a review of the solomining.de listing is requested, citing the Lumen record. The project founder replies within 23 minutes that this is a matter of German law and outside the repository’s scope.
2. 18 May, 14:15: WantClue states that the company has “never used it for any purpose of restricting competitors” — 23 days after a complaint aimed at deleting 71 URLs of a competitor.
3. 18 May, 15:10: fifty-five minutes later the case is closed. “I agree, this can be closed as resolved.”
4. 18 May, 16:51: in the same thread, aimed at us: “Nothing about Bitaxe.de is looking legit.”
5. 22 May 2026: four days after the closure, the trademark owner’s lawyers state that licences were granted to two larger competitors — precisely the opposite of what the exoneration had asserted.
6. 18/19 August 2026: the same developer moves to have bitaxe.de removed from the list. Two other participants agree.

You do not have to impute a motive to anyone to see that two different standards were applied. The review of the trademark holder ended after four days on the strength of an assertion with no evidence at all. The review of us has been running since May, demanding documents, screenshots and deadlines — and ended in a motion to delist us.

The donations: confirmed, not documented

On its “About us” page, solomining.de advertises ongoing support for the developers:

“We support the efforts of the open-source developers by giving a small contribution to the open-source developers for every miner we sell.” (translated from the German original)

Who receives those payments, how much they amount to and on what basis is not stated. That same evidence had previously been demanded of us — and WantClue himself had set the standard: any vendor publicly claiming to support OSMU financially should be able to substantiate it.

On 24 August 2026 that question was put in the Bitaxe project’s repository, in exactly three parts: who received the money, how much was it, what verifiable evidence exists? At 13:05:58 UTC the project founder answered:

“I can also confirm donations from solomining.”

After that, things moved quickly:

Forty-one seconds from the answer to a closed door. Let us be precise about what that sentence does and does not do. It confirms, for the first time from the project side, that payments were made — until then that was only the vendor’s own marketing. It answers none of the three questions asked: not who received the money, not how much, not what evidence exists. And with the lock in place, the question can no longer be asked there.

The comparison of standards belongs here. When our own wording about donations was at issue, solid evidence was demanded — and we corrected the wording accordingly. When the same question was put to the vendor holding the trademarks, one sentence without evidence sufficed and the discussion was ended. Same process, same list, same people — different standard.

Can you trademark an open-source project at all?

Legally the answer is that the offices do not check who invented a name. Whoever files first and pays the fee gets the registration. Which makes the more interesting question not whether you may, but what you do with it.

Open hardware runs on a simple trade. The developer gives everything away — schematics, board layout, firmware, bill of materials. Anyone may build it, change it, sell it, commercially, better than he does. What he keeps is one single thing: that the object is named after his project. The name is the last remaining link between the work and the person who gave it away. Registering that name puts into private hands the one thing the developer never handed over.

Then there is the purpose of trademark law itself. A mark is meant to indicate origin: it tells a buyer which house a product comes from. A mark over a community name does not do that — it asserts an origin that does not exist. A Bitaxe does not come from Fellbach. It comes from an open project many people worked on, and it is built by dozens of workshops worldwide. The Google complaint states that false origin quite openly: my registered trademark, my brand.

There is an honest way, and it looks different

You could object that somebody has to file the name, or a third party will. That is true — the table above shows exactly that happening. But the objection comes with a condition, and everything turns on it: who holds the mark, and what happens next?

The template has existed for decades. The Linux Foundation holds the “Linux” trademark and licenses it free of charge to anyone distributing Linux, for the sole purpose of stopping anyone else from locking it up. There are even dedicated legal instruments for this: the collective mark, held by an association, and the certification mark, open to anyone meeting published criteria. A defensive filing in the project’s name, administered for everybody, would be entirely welcome. We would support it and abide by it.

The difference between defence and appropriation comes down to three questions:

• Who is on the register? The project or a trustee — or a single vendor competing in the same market?
• What happens to the rights? Are they kept open to everyone — or granted to some and withheld from others?
• Who is pursued? Free riders abusing the name — or the next-largest competitor, while the free riders are left alone?

By that measure the case answers itself. The registrant is a vendor, not a project. Licences were granted, by their own account, to two larger competitors — and not to us. And exactly one market participant was pursued, while in the same period Chinese applicants registered the same names unopposed.

There is one more test, the simplest of all: would the developers have agreed, if anyone had asked them? The author of the Bitaxe project answered that in public — bitaxeorg does not recognise third-party national registrations. It is hard to be much clearer than that.

Why this goes beyond us

You could file all this under “two shops fighting”. We think it is more than that, for one reason: the names are not a single vendor’s to dispose of.

The Bitaxe became what it is because anyone may download the files, build the device, improve it and sell it. That freedom turned a hobby board into a family of machines built all over the world, from the Gamma to the NerdOctaxe. If the name of that family belongs to one seller, the freedom applies only to the engineering, not to the word the engineering happens under. And without the word, nobody finds it.

In practice, a trademark monopoly over a project name means:

• Whoever builds the device may no longer be allowed to call it that — or needs a licence.
• Whoever sells it can be removed from search engines and platforms by complaint, without a court ever seeing the matter.
• Whoever contributes as a developer finds their project’s name registered to a third party.
• And the buyer ends up finding only the vendors permitted to use the name.

That is what happened to us in April 2026. Not as a threat, but as an accomplished fact: 71 URLs, removed from the index.

What we are asking for

• Clarity about who agreed. The other side, through its lawyers, relies on the consent of the holder of the title rights — Skot — and on the knowledge of OSMU. In public, the opposite has been stated. That question needs answering, by the people named, not by us.
• Disclosure of the licensing programme. If two larger competitors were granted licences: which ones, on what basis, on what terms?
• Verifiability of the confirmed payments — recipient, amount, evidence. That they exist is now confirmed; everything else is open. It is the same standard that was applied to us, and that we accepted.
• An open discussion. Asking for evidence is not an attack. An issue closed and locked 41 seconds after the answer does not remove the doubt — only the ability to voice it.
• And the heart of it: that the names of open projects belong to nobody in particular. We are pursuing cancellation of the marks in so far as they cover project names their owner did not create.

In closing

We would gladly have done without this article. A shop that sells miners ought to be writing about hashrate and heatsinks, not file numbers. But silence stops being an option when 71 pages vanish from search, when the answer to a request for evidence is a locked door, and when the register lists the names of four projects that other people designed in their spare time and gave away.

We know how this looks: a competitor writing about a competitor. So there is nothing in this article you have to take on trust. The trademark data are in the DPMA and EUIPO registers, most easily searched together through TMview. The complaint is in the Lumen database under number 84146531. The quotations are in the Bitaxe project’s public repository, the catalogue figures in the solomining.de shop, the company details in their legal notice. The extract from the lawyers’ letter comes from our own proceedings; we published it in May 2026. All retrieved on 24 August 2026. Go and check — that is the point of listing them.

If anything here is wrong, we will correct it, promptly and in this place. Bolt Engineering GmbH and everyone named may demand a right of reply; we will publish it in full inside this article, with no commentary before or after. The address is in our legal notice. That undertaking has no expiry date.

Beyond that, what held from the beginning still holds: the Bitaxe belongs to nobody. That was rather the point.