• Issue number: #49

Task Description

We want all license metadata to be present and to be correct. We also need full accounting for it.

Tasks

• [ ] Improve licensing info on deployed website.
  • [ ] Create a license overview page, that links to all license relevant documents and also lists the license of all available pages.
    • [x] Start with empty page. -> LicensePageExtension provides this at http://localhost:8443/net/splitcells/website/license-page.html
    • [x] Create and use a website server config, that contains a list of all license pages. The new option is located at net.splitcells.website.server.Config#licensePages
    • [x] Add name/description to every license page.
    • [o] Create a static Cell methods, that lists all relevant license resources. Use these methods in the static website server construction functions. -> For now it is ok, to use website server config.
    • [x] Note that user content is not affected by this.
    • [x] Link relevant existing pages.
      • [x] Fix all links.
      • [x] net.splitcells.network
      • [x] net.splitcells.network.system
      • [x] net.splitcells.martins.avots.distro -> Does not have relevant license files.
      • [x] net.splitcells.martins.avots.website -> Does not have relevant license files.
    • [x] Add license overview page to menu.
  • [ ] List license of every page centrally.
    • [x] Create empty page for listing all files. -> ResourceLicensingExtension
    • [x] net.splitcells.maven.plugin.resource.list has also to create and bundle a directory containing the metadata into the jar. The directory contains a file tree, where any file corresponds to a resource managed by the resource list. The file format is a UTF-8 encoded property file.
      • [x] Create meta data files.
      • [x] Read license metadata in the file headers.
      • [x] Read copyright text metadata in the file headers.
      • [x] Read REUSE.toml for list files without license metadata.
        • [x] Double check results in jars by hand.
        • [x] Correctly, filter paths of parent REUSE.toml, that are not part of the current project.
        • [x] Parse actual license data for each path REUSE.toml.
        • [x] Write all MetaData after parsing of any source to a file to MetaData map.
        • [x] Correct the MetaData#filePath. Currently, it is absolute.
          • [x] Fix file content parser.
          • [x] Fix REUSE.toml parser.
        • [x] REUSE parser does not correctly determine copyright and license data of a file.
        • [x] After creating MetaData map write these to target/classes.
        • [x] Support ** in the REUSE paths.
          • [x] Detect pattern.
          • [x] Check all project files for pattern.
          • [x] Create one meta file per file match.
      • [x] Merge results of REUSE.toml parsing and individual file parsing.
      • [x] Process binary files with MetaData class. -> This is only supported via ** matching in REUSE.toml. Currently, there is no supported metadata inside binary files and REUSE.toml is required by the REUSE tool.
    • [ ] Provide license metadata for page via ProjectsRenderer#metaData(String). Use this to list the licensing of every HTML file.
      • [ ] Extended the filesystem API, where an additional method returns a metadata for a given file. A write method is not needed for this, as there is currently no use case for that.
        • [x] The default implementation should return empty metadata.
        • [ ] Implement this for FileSystemViaClassResourcesImpl.
        • [ ] Implement this for the DevDistroCell.
      • [ ] Only provide access to metadata, to which the user has access to.
        • [ ] Use RenderRequest for ProjectRendererExtension#metaData.
          • [x] Create new method with empty default.
          • [ ] Implement new method for one class. -> XmlProjectRendererExtension is used for this.
          • [ ] Use new method in ProjectsRendererI. The old metaData method is used via the new default version. If the new metaData method is implemented, there should not be an old variant at the same class or the new one should use the old one as well.
          • [ ] Migrate all old implementations to new methods.
          • [ ] Remove old method.
      • [ ] Render the metadata in HTML layout.
      • [ ] Go over all extensions and implement this.
  • [ ] Show license of every page on every page.
  • [x] Link to NOTICE.md and LICENSE.md.
  • [x] net/splitcells/network/legal/Developer_Certificate_of_Origin.v1.1.txt is not correctly rendered.
• [ ] Images are licensed under the CC BY-SA 4.0.
  • [ ] Update licensing guidelines.
  • [ ] Consider add licensing metadata in images itself. See German court cases (to be refound), why no license metadata in CC BY-SA 4.0 can be a problem for other distributors.
  • [ ] Consider https://creativecommons.org/licenses/by-sa/4.0/deed.en
• [x] Ensure, that every repo has proper licensing data. -> This is done via the daily CI workflow at Codeberg.
• [ ] Simplify Java's license header, in order to simplify Java files and make them shorter. Document that the Linux kernel uses the same format and that such a header is REUSE compliant as well.
• [ ] Add denial notice of EU's copyright exception regarding machine learning.
• [ ] Add REUSE files to license pages.
• [ ] Remove big license header from Java code as this only wastes resources.
• [ ] Check and/or generate license NOTICE file, in order to ensure its correctness.
• [ ] Ensure via Maven plugin, that all files have license info.

Done Tasks

• [x] test.everything should verify license data as well. -> This is done via repos.test.sh instead.
• [x] https://www.it-cisq.org/software-bill-of-materials/index.htm
• [x] Use https://reuse.software/ in order to check licensing in source code.
  • [x] Use tmp branch for this experiment.
  • [x] https://reuse.software/tutorial/
    • [x] reuse init
    • [x] ~/.local/bin/reuse lint
  • [x] Check this in daily CI for all repos. Just create the workflow for now. The actual license info will be fixed via the weekly test deployments task.
    • [x] Create command bin/repos.license.check in order to check this.
    • [x] Make license processes completely independent of GPL3 code, by providing a default license check implementation, that does nothing and inject REUSE dynamically into the process instead.
    • [x] Install OSI in CI.
      • [x] Test this in worker.execute first.
        • [x] cd projects/net.splitcells.os.state.interface && ./bin/install
        • [x] Create a general shell environment setup command. -> This is done inside bin/worker.bootstrap. A real general command is not needed for now.
          • [x] Create a project command that list all default command repositories managed by the hub repo and install these with this command via the shell project. -> The project command is bin/net.splitcells.shell.projects.peers.for.commands.
          • [x] Create a project command that list all command repositories, that are present in the current repository. -> net.splitcells.shell.repo.subs
          • [x] Use new project commands, in order to install all command repos.
          • [x] Install OSI first. Then use command.repository.register, in order to easily use command.repository.register. After that, reinstall OSI, in order to have all repos installed.
      • [x] Create general license test command. -> This command is named repos.license.check.
      • [x] Add this test to Codeberg's CI.
    • [x] Maybe this is part of the worker bootstrapping? -> Yes, this is integrated into bin/worker.bootstrap.
  • [x] The reuse-tool is licensed under the GPL-3. Is that a problem?
  • [x] Register main repo at: https://api.reuse.software/register
• [x] Create licensing guidelines.
  • [x] Create documentation stating all licence checks, that should be done.
  • [x] Add REUSE to licensing guidelines.
• [x] Document/link licensing overview in README. -> This is not relevant for now.
• [x] Add Reuse to project file system standard and document, that its LICENSES folder is exceptional.
• [x] When rendering CommonMark files, the licence info at the start of the file should be moved to the end of the file.