Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UqXYdJDrM65gxsCn2PrbfK
21 KiB
Third-party software notices
Jungfraujoch is licensed under GPL-3.0 (see LICENSE); the FPGA design is
licensed under CERN-OHL-S-2.0 (see fpga/LICENSE), and the PCIe kernel driver
(fpga/pcie_driver/) under GPL-2.0-or-later (see fpga/pcie_driver/LICENSE). It builds on a number of
third-party components, acknowledged below as required by their licenses.
This file is the human-readable manifest. The verbatim license texts live in the
licenses/ directory (regenerate with bash licenses/COLLECT.sh). The frontend's
bundled JavaScript dependencies are listed separately in
frontend/dist/THIRD_PARTY_LICENSES.txt, generated at build time (npm run licenses).
All licenses below are GPL-3.0-compatible, with one exception: the NVIDIA CUDA Toolkit is used under its own EULA (see the notes at the end of this file).
Fetched at build time and statically linked into the C++ binaries
These are downloaded by CMake (FetchContent / ExternalProject) during the first configure and
linked into the Jungfraujoch executables.
| Component | Version | Copyright | License (SPDX) | License text |
|---|---|---|---|---|
| spdlog | 1.17.0 | Gabi Melman | MIT | spdlog.txt |
| Zstandard | 1.5.7 | Meta Platforms, Inc. | BSD-3-Clause OR GPL-2.0-only (used under BSD-3-Clause) | zstd.txt |
| HDF5 | 2.2.0 | The HDF Group; UIUC | BSD-3-Clause-style + Apache-2.0 | hdf5.txt |
| slsDetectorPackage | 8.0.2 / 9.2.0 | PSI | LGPL-3.0-or-later | LGPL, GPL |
| cpp-httplib | 0.56.0 | Yuji Hirose | MIT | cpp-httplib.txt |
| libzmq (ZeroMQ) | 4.3.5 | iMatix and contributors | MPL-2.0 | libzmq.txt |
| libtiff | 4.7.2 | Sam Leffler; SGI | libtiff (BSD-like) | libtiff.txt |
| Ceres Solver | commit 0c70ed3a1a2d | Google Inc. and contributors | BSD-3-Clause | ceres-solver.txt |
| Abseil | 20250127 (required by Ceres) | Google Inc. | Apache-2.0 | abseil.txt |
| Mbed TLS (with TF-PSA-Crypto) | 4.2.0 | The Mbed TLS Contributors | Apache-2.0 OR GPL-2.0-or-later (used under Apache-2.0) | mbedtls.txt, tf-psa-crypto.txt |
| Catch2 | 3.16.0 | Catch2 Authors | BSL-1.0 | catch2.txt |
| zlib-ng | 2.3.3 | Jean-loup Gailly and Mark Adler; the zlib-ng contributors | Zlib | zlib.txt |
| Eigen | 3.4.1 | Benoit Jacob, Gael Guennebaud and contributors | MPL-2.0 (+ BSD parts) | eigen.txt, README |
| MD5 implementation (from OpenSSL), bundled in slsDetectorPackage | slsSupportLib/src/md5.c of slsDetectorPackage |
The OpenSSL Project Authors; modifications PSI | Apache-2.0 | slsDetectorPackage-md5-Apache-2.0.txt |
| RapidJSON, bundled in slsDetectorPackage | libs/rapidjson/ of slsDetectorPackage |
THL A29 Limited, a Tencent company, and Milo Yip | MIT | rapidjson.txt |
Mbed TLS is fetched and statically linked only for viewer builds (JFJOCH_VIEWER_BUILD /
JFJOCH_VIEWER_ONLY), where it is the TLS backend of jfjoch_viewer's cpp-httplib client; the
broker and writer never link it.
Catch2 is used only to build the test binary (jfjoch_test) and is not part of any shipped
artifact; it is listed here for completeness.
zlib is supplied by zlib-ng built in its zlib-compatible mode (same zlib.h, same API and symbol
names), so it is the zlib that HDF5, libtiff, cpp-httplib and GEMMI all link. It is
built during the configure rather than added as a subproject; the licence is the zlib licence
either way. Eigen is header-only: only its headers reach the binaries, and no Eigen CMake runs.
Vendored directly in the repository
These live in the source tree (see the path) rather than being fetched; traccc, the GEMMI-, Ceres- and fast-feedback-indexer-derived files and the Xilinx-derived HLS stream are the exceptions - code adapted into first-party files rather than a vendored directory, see the notes at the end of this file.
| Component | Path | Copyright | License (SPDX) | License text |
|---|---|---|---|---|
| nlohmann/json (3.12.0) | include/nlohmann/ |
Niels Lohmann | MIT | nlohmann-json.txt |
| Macaron Base64 | include/base64/ |
tomykaira | MIT | base64-macaron.txt |
| TinyCBOR (0.6.0) | frame_serialize/tinycbor/ |
Intel Corporation | MIT | tinycbor.txt |
| Bitshuffle | compression/bitshuffle/ |
Kiyoshi Masui | MIT | bitshuffle.txt |
| Bitshuffle (h-perf) | compression/bitshuffle_hperf/ |
Kal Conley | Apache-2.0 | bitshuffle-hperf.txt |
| LZ4 (1.9.4) | compression/lz4/ |
Yann Collet | BSD-2-Clause | lz4.txt |
| stb_image_write (v1.16, commit 31c1ad3) | preview/stb/ |
Sean Barrett | MIT OR Unlicense | stb_image_write.txt |
| HLS arbitrary-precision types | fpga/include/ |
Xilinx, Inc. | Apache-2.0 | xilinx-hls-headers.txt |
Vitis HLS hls_stream.h (2023.1, adapted) |
fpga/hls/parallel_stream.h |
Xilinx, Inc. | Apache-2.0 | xilinx-hls-headers.txt |
Vivado HLS hls_burst_maxi.h (simulation model of burst transfers on M-AXI, modified; to be removed when the FPGA part is frozen) |
fpga/include/hls_burst_maxi.h |
Xilinx, Inc. | not stated in the file; upstream terms not established | — |
| ICMP responder and IP header checksum (HPCN Group, UAM) | fpga/hls/icmp.cpp, fpga/hls/ip_header_checksum.h |
HPCN Group, UAM Spain; modifications PSI | BSD-3-Clause | hpcn-uam.txt |
| fast-feedback-indexer (commit a94fd4f) | image_analysis/indexing/fast-feedback-indexer/ |
PSI | BSD-3-Clause | fast-feedback-indexer.txt |
| GEMMI (0.7.5, trimmed) | gemmi_gph/ |
Global Phasing Ltd. | MPL-2.0 | gemmi.txt |
| stb_sprintf (v1.10, bundled with GEMMI) | gemmi_gph/stb/stb_sprintf.h |
Jeff Roberts / RAD Game Tools; Sean Barrett | MIT OR Unlicense | stb_sprintf.txt |
| 3x3 symmetric eigensolver (Connelly Barnes, after JAMA; bundled with GEMMI) | gemmi_gph/eig3.cpp |
— | public domain (no SPDX id) | — |
| CCP4 Monomer Library (modified subset: restraint dictionary data) | image_analysis/structure_refinement/monomers/ |
CCP4 / the Monomer Library authors (Vagin et al.) | LGPL-3.0-only | ccp4-monomer-library.txt |
| Larch (its Cromer-Liberman routines, as GEMMI ports them) | gemmi_gph/fprime.cpp |
Matthew Newville, The University of Chicago, et al. | MIT | larch.txt |
| PEGTL | gemmi_gph/gemmi/third_party/tao/ |
Dr. Colin Hirsch and Daniel Frey | MIT | pegtl.txt |
| sajson | gemmi_gph/gemmi/third_party/sajson.h |
Chad Austin | MIT | sajson.txt |
| fast_float | gemmi_gph/gemmi/third_party/fast_float.h |
The fast_float authors (Daniel Lemire et al.) | Apache-2.0 OR MIT OR BSL-1.0 | fast-float.txt |
| half | gemmi_gph/gemmi/third_party/half.hpp |
Christian Rau | MIT | half.txt |
| pocketfft | gemmi_gph/gemmi/third_party/pocketfft_hdronly.h |
Max-Planck-Society; Peter Bell; MIT (FFTW-derived parts) | BSD-3-Clause | pocketfft.txt |
| tinydir | gemmi_gph/gemmi/third_party/tinydir.h |
Cong Xu, Lautis Sun, Baudouin Feildel, Andargor | BSD-2-Clause | tinydir.txt |
| traccc (ACTS) | image_analysis/spot_finding/StrongPixelSet.cpp, SpotExtractorGPU.cu |
CERN, for the benefit of the ACTS project | MPL-2.0 | traccc.txt |
| fast-feedback-indexer (adapted) | image_analysis/indexing/FFBIDXIndexerCPU.cpp |
PSI | BSD-3-Clause | fast-feedback-indexer.txt |
| Ceres Solver (adapted) | image_analysis/geom_refinement/LMSolver.h, LMSolver.cpp |
Google Inc. | BSD-3-Clause | ceres-solver.txt |
| GEMMI (adapted) | image_analysis/structure_refinement/ModelScaling.cpp, ModelMaskGPU.cu, ModelStructureFactorsGPU.cpp, ModelDensityMetal.metal, AtomGradientMetal.metal |
Global Phasing Ltd. | MPL-2.0 | gemmi.txt |
| Durin (adapted) | xds-plugin/plugin.h, plugin.cpp |
Diamond Light Source Ltd.; Global Phasing Ltd. | BSD-3-Clause | durin.txt |
| xbflash.qspi | tools/xbflash.qspi/ |
Xilinx / AMD | Apache-2.0 | xbflash-qspi.txt |
| wingetopt | tools/wingetopt/ |
Todd C. Miller; The NetBSD Foundation | ISC AND BSD-2-Clause | wingetopt.txt |
Data files in the repository
Not code and not linked into any binary: input files of the battery's timing benchmark
(tools/battery/timing.json), kept verbatim as downloaded.
| Component | Path | Copyright | License (SPDX) | License text |
|---|---|---|---|---|
| AlphaFold Protein Structure Database predictions AF-P68082-F1, AF-P62894-F1, AF-P17090-F1, AF-P32322-F1 and AF-Q65MG6-F1 (v6), with their predicted aligned error | tools/battery/models/AF-* |
DeepMind Technologies Limited; EMBL-EBI | CC-BY-4.0 | alphafold-db.txt |
| wwPDB entries 4DC8, 4NWV, 9GJX, 9S02 and 9ZLO, deposited coordinates (mmCIF as the RCSB PDB serves them) | tools/battery/models/pdb/ |
the depositors; the wwPDB | CC0-1.0 | wwPDB usage policy |
The deposited entries are the reference models the benchmark's placements are checked against, and 4NWV's is also moved by a fixed transform at run time to make one start model.
FPGA bitstreams
The prebuilt FPGA images in fpga/images/ (jfjoch_fpga_pcie_100g.mcs.gz,
jfjoch_fpga_pcie_8x10g.mcs.gz) are generated from the CERN-OHL-S-2.0 design together with AMD/Xilinx
IP cores, which are licensed by Xilinx under its own terms and not under CERN-OHL-S-2.0:
| Component | Image | Copyright | Terms |
|---|---|---|---|
| UltraScale+ Integrated 100G Ethernet (CMAC) | jfjoch_fpga_pcie_100g.mcs.gz |
AMD / Xilinx, Inc. | Xilinx IP licence ("Non-cost license ... has to be requested from AMD/Xilinx website", docs/FPGA.md) |
| 10G/25G Ethernet Subsystem for UltraScale+ | jfjoch_fpga_pcie_8x10g.mcs.gz |
AMD / Xilinx, Inc. | Xilinx IP licence, received by PSI from the Xilinx University Program; per docs/FPGA.md: "usage of bitstreams generated by PSI continuous integration pipeline for 8x10g is only allowed for non-commercial use" |
Both images also contain the IP cores included with Vivado, under Xilinx's terms.
Runtime libraries and SDKs (shipped in binaries, not in the source tree)
| Component | Used by | License | Notice |
|---|---|---|---|
| Qt 6 | jfjoch_viewer |
LGPL-3.0 | notice, LGPL-3.0 |
| NVIDIA CUDA Toolkit (cudart, cuFFT) | CUDA builds | NVIDIA CUDA EULA | notice, EULA |
Frontend (npm) dependencies
The React/TypeScript frontend (frontend/) bundles a large transitive tree of npm packages,
overwhelmingly MIT/ISC/BSD/Apache-2.0 licensed. Their full notices are generated automatically:
cd frontend && npm run licenses # writes dist/THIRD_PARTY_LICENSES.txt
The generated file is produced as part of the frontend build target and installed alongside the
served frontend, so the shipped web UI carries its own attribution.
Notes on weak-copyleft and attribution-sensitive components
- MPL-2.0 (Eigen, GEMMI, libzmq, traccc): file-level copyleft. GEMMI is vendored in
gemmi_gph/in trimmed form; libzmq and Eigen are fetched at build time (Eigen is header-only). The corresponding source is available from each project upstream. - GEMMI's own bundled third-party headers — PEGTL, sajson, fast_float, half, pocketfft and
tinydir, all under
gemmi_gph/gemmi/third_party/— are listed separately above rather than being absorbed into GEMMI's row: they are other authors' code under other licences (MIT, BSD and Apache/MIT/BSL), and GEMMI's MPL-2.0 does not speak for them. Their terms are carried in the headers themselves rather than inLICENSEfiles, so thelicenses/*.txtcopies are kept by hand; only PEGTL ships aLICENSE, whichCOLLECT.shcopies. - traccc is the one entry that is not a vendored directory. Its sparse connected-component
labelling enters two otherwise first-party files:
StrongPixelSet.cppadapts the SparseCCL source, andSpotExtractorGPU.cufollows the design of its GPU counterpart. MPL-2.0 is file-level, so both files name the origin at the top and are covered bylicenses/traccc.txt. See ACKNOWLEDGEMENT.md for the citation. - Ceres Solver is also fetched and linked (table above); separately,
LMSolver.h/.cppre-implement its trust-region Levenberg-Marquardt minimiser, projected line search, polynomial step choice and sphere manifold for the crystal refinement, following its source. Both files name the origin at the top and are covered bylicenses/ceres-solver.txt. - fast-feedback-indexer is also vendored and linked in CUDA builds (table above); separately,
FFBIDXIndexerCPU.cppports its CUDA kernels to the CPU. The file names the origin at the top and is covered bylicenses/fast-feedback-indexer.txt. - GEMMI is also vendored (table above); separately,
ModelScaling.cpp,ModelMaskGPU.cu,ModelStructureFactorsGPU.cpp,ModelDensityMetal.metalandAtomGradientMetal.metalinimage_analysis/structure_refinement/follow GEMMI's scaling, Levenberg-Marquardt, solvent-mask, reciprocal-grid and form-factor code. Each names the GEMMI source it follows and is covered bylicenses/gemmi.txt. - Durin (
xds-plugin/plugin.h/.cpp): the XDS plugin's entry points are based on the Durin plugin of Diamond Light Source and Global Phasing; both files name the origin at the top and are covered bylicenses/durin.txt. - slsDetectorPackage's bundled code: the slsDetectorPackage sources compiled into the binaries
include an MD5 implementation taken from OpenSSL (Apache-2.0) and the RapidJSON headers (MIT); their
msinttypes/headers (BSD-3-Clause) are for old MSVC compilers only. - Apache-2.0 components: where upstream ships a
NOTICEfile, it is reproduced in the correspondinglicenses/text. - CCP4 Monomer Library (LGPL-3.0) is data, not code: the restraint dictionaries of the compounds
rugnux's refinement restrains, in
image_analysis/structure_refinement/monomers/, with the library's ownCOPYING. The subset is a modified version of the library: the per-compound files are copied unmodified, whilelinks_and_mods.cifandener_lib.cifare cut down to the entries those compounds use andmetals.jsonis converted tometals.cifholding only the metal-site distances of its ions. Each of these three files says so in a comment at its top, with the date it was generated.image_analysis/structure_refinement/monomers/update_monomers.pyregenerates the subset from the upstream library. The build embeds the files in the binary as text. As the LGPL-3.0 requires, the modified files remain under LGPL-3.0 (with the GPL-3.0 text inLICENSE), the source tree carries them in their preferred (text) form together with the script that makes them, and a user can replace the embedded copy with a version of their own by replacing the files in that directory (or rerunning the script on another library version) and rebuilding. - Qt (LGPL-3.0) and NVIDIA CUDA (EULA) carry redistribution conditions beyond a copyright
notice; see their dedicated notice files. The verbatim LGPL-3.0 and CUDA EULA texts are bundled
(
licenses/Qt6-LGPL-3.0.txt,licenses/NVIDIA-CUDA-EULA.txt); the CUDA EULA is the one shipped with CUDA Toolkit 12.8 — replace it if you build against a different toolkit version.