The Intersection of Trademarks and Open Source

By Gail Hashimoto and Chiara Portner

When open source developers call us asking to confirm that they can use the trademark or name of an open source project for their newly forked project, they do not get the black and white answer “Yes” that they desire but rather the grey area lawyer response – “It depends on what you propose to use, how you propose to use it, the license, and whether there is a naming or trademark policy.”

When a project is forked in the open source world, a new fork is essentially a breakoff project or new software lineage apart from the main repository of code or original project. You can think of it as the original project being a strawberry plant, and the forked project like a runner that appears growing out of the side establishing its own separate plant. A forked project won’t get the benefits of contributions of bug fixes and improvements to the original project and is separately maintained from the original project. Many in the open source community assume that if they fork a project, they have the right to use the title or name, or the trademark of the original project in their forked project because “it” is open source. This assumption, albeit seemingly logical, is misguided. The “it” is the software which is what the open source license applies to, not the name or trademark of the project.

If you are in this situation where you have forked code, you may be inclined to pick a very similar name because you want to be associated with the original project in order to “gain traction” for your new project. While your interests make business sense, they may not make sense from a trademark or a community sense.

In the open source community, licensors tend to concentrate their focus on the licensing of copyrights in the underlying source code for the software and, to a lesser extent, patents in their software. Trademarks, while certainly an important intellectual property right, are often unmentioned in many open source licenses and are the source of misperceptions in the community. If you take a look at the GPL V2.0, for example, there is no mention of trademarks. The GPL V3.0’s language grants only an express copyright license. This is for a good reason:  there is nothing “open” about trademark use, which requires monitoring and quality control by the trademark owner.

In those open source licenses, such as Apache, that do address trademarks specifically, the reference to trademarks is not to grant trademark rights but rather is in the negative to expressly clarify that no rights to use the names or trademarks are granted: Trademarks. This License does not grant permission to use the trade names, trademarks, service marks, or product names of the Licensor, except as required for reasonable and customary use in describing the origin of the Work and reproducing the content of the NOTICE file.

Note that the Apache license lists several distinct types of identifiers: trade names, trademarks (including service marks), and product names. These types of identifiers each serve a different purpose, and are outside the scope of copyright. The name of the company and product identify a thing – the company or product, respectively. In contrast, trademarks and service marks identify the source of a product. Although sometimes a name and mark are identical words, they serve different purposes and a mark need not be the name of the source in order to identify a product as originating from that source.

Setting aside the related complications of risks of dilution or loss of trademark rights, consumer confusion, infringement, and product liability, if the original project uses a license that does not address naming and trademark rights this raises the question as to how to name the forked project. Can one use the name of the original project? Usually not, unless you have permission and a license from the owner of the original project. Can you use the original’s name or trademark to describe the forked project’s origin? Perhaps, so long as the description is not misleading as an indicator of still being sponsored by or associated with the original project, which depends on how you use the trademark.

As you can see, when setting up a project initially, it is a good idea to have a trademark policy for the project name and marks. Such a policy would prohibit use of the project name or mark in domains and in the names or marks of other forked projects. If the project is so unique and its uniqueness may cultivate more adoption of the project and more contributions, the community will benefit, but one must consider protecting the name and trademark to avoid confusion.  For example, the Linux Foundation posts its trademark use policy at In this policy, the foundation also recognizes the legal concepts of fair use or nominative use, which allows use of a trademark in a fair factual manner but not as to indicate origin or sponsorship. Referring to the name of an original project may be considered fair or nominative use if the reference is to indicate the forked project is dependent on the original project. But one cannot conflate factual descriptions or a dependency with an indication that the forked project and the original project are the product of the source licensor, or that the original project owner sponsors or approves of the forked project. New forked projects must create their own unique name and marks so as not to confuse users; such confusion is the hallmark of trademark infringement.

We recently ran across the following projects which demonstrate some of the tension and potential confusion around open source and trademark rights. See and and see below websites of the original PrestoDB and the fork named PrestoSQL The names and logos appear almost identical.

presto-1.png presto-2.png

 From what we can tell, the PrestoDB project was first developed by Facebook and is hosted by the ubiquitous Linux Foundation which means that there are overarching open source community guidelines and principles. PrestoDB was forked and the fork was named PrestoSQL. There are now also two foundations: PrestoDB (the original product name and mark) which is hosted by Linux Foundation’s Presto Foundation, and the Presto Software Foundation which hosts PrestoSQL (product name and mark for the fork). Are you confused yet? The problem is based on the common use of “Presto” in the entity names, product names, and product trademarks by the second foundation. Unless there is a license between the parties, the problem is not mitigated by use of other non-distinctive, descriptive words or phrases such as “DB” and “SQL.” The similar graphics and use of an identical logo on both websites give the strong impression that the products are linked and the entities are related rather than being independent from one another.

A counterexample to this would be looking at the history of the popular open source project MySQL. Following the acquisition of MySQL AB by Oracle via Sun, the project was forked. Instead of naming the project something similar like “MySQL++” or “MySQLpro”, which would have been confusing to the community, the developers decided to name the new project MariaDB. Both projects live on in their own right. 

In the end, one must keep in mind that simply because a piece of software is licensed under an open source license it does not automatically come with a grant of trademark rights to use the project name or related marks. While open source may be free as to the software, it does not amount to a free license to use names or trademarks.

Stay up to date on the latest news, alerts, events and legal insights: