Licensing Status¶
Unknown / unresolved — no licence is currently granted
Fed-MaxFuse previously carried an MIT licence. That declaration has been withdrawn pending review. No permission to use, copy, modify, or distribute this code is granted at present.
Why the licence was withdrawn¶
Fed-MaxFuse builds on, and benchmarks against, MaxFuse, which is not distributed under an open-source licence. It is covered by the Stanford Academic Software License Agreement, whose terms include:
- use permitted solely for internal academic non-commercial purposes;
- no distribution or transfer to another location or person without prior written permission from Stanford and Penn;
- no reverse engineering, together with a clause stating that "any programs created based on the Software will be considered a derivative of Software and owned by STANFORD AND PENN";
- supplied as-is, with no warranty or support.
The reverse-engineering and derivative-ownership clauses are unusually broad. Read expansively, they could be argued to reach work that builds on the method, including an independent re-implementation of the published algorithm.
Whether they actually do is an open question that has not been settled here. Rather than continue to assert an MIT licence that may not be ours to grant, the licence declaration has been removed until the position is clarified.
What this means in practice¶
| Question | Answer |
|---|---|
| Can I use this code? | Not currently. Absent a licence, default copyright applies and all rights are reserved |
| Can I copy parts of it into my own project? | No — please ask first |
| Can I use it commercially? | No |
| Can I cite the work? | Yes. Citation is not a licensing matter — see Publications |
| Can I read and learn from it? | Yes. The repository is public; reading is not restricted |
If you want to use any part of this work, please get in touch; see Contributing for contact details. The intent is to reach a clear position, not to restrict access.
Two different objects, two different terms¶
Worth separating, because they are often conflated:
- The algorithm is published in the peer-reviewed literature (Chen et al., Nature Biotechnology 2024), open access under CC BY 4.0. Published methods are generally free to implement.
- The reference implementation is separately licensed under restrictive academic terms.
An independent implementation written from the paper occupies different ground than a fork of the package. But "different ground" is not the same as "clearly safe", which is precisely why this page says unknown rather than offering a conclusion.
Note also that the original run type in this project invokes upstream MaxFuse directly for
baseline comparison. That is a straightforward dependency on the licensed package, and its terms
apply to that mode regardless of how the rest of the codebase is characterised.
Data¶
The benchmark datasets are distributed by the MaxFuse authors and derive from published studies (Hao et al. 2021; Kennedy-Darling et al. 2021; King et al. 2021). Their terms are set by those publications and the hosting repository, and are unaffected by anything on this page. Cite the original studies when using them; see Publications.
Documentation and figures¶
This documentation shares the repository's unresolved status. Figures reproduced from Chen et al. (2024) are used under CC BY 4.0, the licence of that open-access article, which is clear and unaffected.
This is not legal advice
Nothing on this page is legal advice, and it should not be relied on as a legal conclusion in either direction. Licence interpretation depends on jurisdiction, on the precise relationship between the works, and on facts not assessed here. Anyone with a concrete use in mind should consult their institution's technology transfer office or legal counsel. For this project, the appropriate route is TU Delft's legal and IP support.
Attribution¶
- Fed-MaxFuse — Krzysztof Baran, with Swier Garst, Marcel Reinders, and Jérémie Decouchant (TU Delft, 2025)
- MaxFuse — Chen, Zhu, Huang, Hickey, Lin, Snyder, Greenleaf, Nolan, Zhang & Ma (Stanford / University of Pennsylvania)