Skip to content

RFC: a plan for false positive license detection #2878

Open

Description

@pombredanne

Context

We are reporting too many false positive licenses. We need to fix this!

Problem

There are several false cases, yet they boil down to these types:

  1. False detection of very short and weak license detection rules detected exactly such as:

  2. Detection of a license text or notice fragment which is too weak to represent a bona fide license detection alone.

  3. Detection of longer unknown license references such as

    • a "license introduction" (as in "This is licensed under....") that may be noisy when followed by a bona fide license notice or text.
    • a license reference to the license in a file (as in "See file COPYING for license") where we can follow the reference
  4. Lack of proper detection of a structured license tag found in a package manifest which is returned as an unknown license

  5. When fragments of the same license are detected with only copyrights added in between as in license detection: Add the nunit license #2859

  6. When sequence of SPDX licenses id are found in license detection tools

  7. Please add yours!

Solution elements

We could treat and report separately mere clues such as this one: they could be an interesting insight in some cases, but alone they are too weak to be considered a license detection

The upcoming two-step process where license matches are grouped in a license detection is another way to consider. We could detect patterns of license matches that could be resolved in a detection. For instance a license intro followed by a license notice.

The scancode-analyzer heuristics and ML-based detection of false positive is another way

Activity

  1. porsche-rishisaxena commented on Feb 28, 2022

    @porsche-rishisaxena

    Hi Philippe,

    In reference to our collective ORT community meeting, we touch base on the false positive license detection 2 weeks ago on version v30.1.0 where Porsche AG OSO also consolidated a report of false-positive cases. Please find attached the report for your kind reference and review.

    report_false_positives.xlsx

    CC: @sschuberth

  2. PatteSI commented on Feb 28, 2022

    @PatteSI

    Thank you for taking action here.
    I will now have a deeper look into our false positive findings as well.
    EDIT: I forgot to mention that everything I mention below was found using scanCode 30.1.0
    At first glance it seems that many LicenseRef-scancode-free-unknown and LicenseRef-scancode-unknown-license-reference findings in our Java projects are actually found in META-INF/LICENSE files created by Maven inside the JARs Example: https://repo1.maven.org/maven2/com/fasterxml/jackson/dataformat/jackson-dataformat-yaml/2.13.1/jackson-dataformat-yaml-2.13.1.jar
    In this case line 3: "Jackson is a high-performance, Free/Open Source JSON processing library." and line 11-13: "Jackson core and extension components may be licensed under different licenses.
    To find the details that apply to this artifact see the accompanying LICENSE file.
    For more information, including possible other licensing options, contact"
    The latter one would probably fall into point 3 mentioned above, a reference to another license file. No idea about the first one though. Not sure why they are only found in the JAR binary and not also in the actual source repo with the same text: https://github.com/FasterXML/jackson-dataformats-text/blob/2.14/properties/src/main/resources/META-INF/NOTICE

    Another interesting example is okhttp3 because the false positive that is found was actually introduced by yourself @pombredanne ;-) : lysine-dev/okhttp#4569 , The current file in my example is this one: https://github.com/square/okhttp/blob/parent-5.0.0-alpha.3/okhttp/src/main/resources/okhttp3/internal/publicsuffix/NOTICE . The license LicenseRef-scancode-unknown-license-reference is found in line 4: "It is subject to the terms of the Mozilla Public License, v. 2.0:" I don't understand why in this case MPL v.2.0 is not recognized correctly.

  3. PatteSI commented on Mar 1, 2022

    @PatteSI

    I created a python parser that can parse the evaluated-model.json file create by the ORT Reporter.
    It is currently scanning a list of problematic licenseRefscan codes which are mostly (always?) causing false positives: https://gist.github.com/PatteSI/5904f4bdfb149dc1ce8c73da53e2f6ae
    I parse a couple of our component and this is the result. Of course it still contains a lot of duplicates (it's a json file but github won't let me upload it as .json):
    falsePosFindingFinal.txt

  4. pombredanne commented on Mar 3, 2022

    @pombredanne
    MemberAuthor

    @porsche-rishisaxena Thank you ++ for the list of false positive in #2878 (comment) ... this is great and actionable!

  5. pombredanne commented on Mar 3, 2022

    @pombredanne
    MemberAuthor

    @PatteSI re: #2878 (comment)

    In this case line 3: "Jackson is a high-performance, Free/Open Source JSON processing library." and line 11-13: "Jackson core and extension components may be licensed under different licenses.

    These tow look like basic license-related clues, but are not real license statement alright.

    Here is the detection I get:

    headers:
        -   tool_name: scancode-toolkit
            tool_version: 31.0.0
            options:
                input:
                    - jackson-dataformat-yaml-2.13.1.jar-extract/META-INF/NOTICE
                --license: yes
                --license-text: yes
                --license-text-diagnostics: yes
                --yaml: '-'
            notice: |
                Generated with ScanCode and provided on an "AS IS" BASIS, WITHOUT WARRANTIES
                OR CONDITIONS OF ANY KIND, either express or implied. No content created from
                ScanCode should be considered or used as legal advice. Consult an Attorney
                for any legal advice.
                ScanCode is a free software code scanning tool from nexB Inc. and others.
                Visit https://github.com/nexB/scancode-toolkit/ for support and download.
            start_timestamp: '2022-03-03T065555.808073'
            end_timestamp: '2022-03-03T065557.886174'
            output_format_version: 2.0.0
            duration: '2.078113555908203'
            message:
            errors: []
            extra_data:
                spdx_license_list_version: '3.16'
                files_count: 1
    files:
        -   path: NOTICE
            type: file
            licenses:
                -   key: free-unknown
                    score: '100.0'
                    name: Free unknown license detected but not recognized
                    short_name: Free unknown
                    category: Unstated License
                    is_exception: no
                    is_unknown: yes
                    owner: Unspecified
                    homepage_url:
                    text_url:
                    reference_url: https://scancode-licensedb.aboutcode.org/free-unknown
                    scancode_text_url: https://github.com/nexB/scancode-toolkit/tree/develop/src/licensedcode/data/licenses/free-unknown.LICENSE
                    scancode_data_url: https://github.com/nexB/scancode-toolkit/tree/develop/src/licensedcode/data/licenses/free-unknown.yml
                    spdx_license_key: LicenseRef-scancode-free-unknown
                    spdx_url: https://github.com/nexB/scancode-toolkit/tree/develop/src/licensedcode/data/licenses/free-unknown.LICENSE
                    start_line: 3
                    end_line: 3
                    matched_rule:
                        identifier: free-unknown_85.RULE
                        license_expression: free-unknown
                        licenses:
                            - free-unknown
                        referenced_filenames: []
                        is_license_text: no
                        is_license_notice: no
                        is_license_reference: yes
                        is_license_tag: no
                        is_license_intro: no
                        has_unknown: yes
                        matcher: 2-aho
                        rule_length: 3
                        matched_length: 3
                        match_coverage: '100.0'
                        rule_relevance: 100
                    matched_text: Free/Open Source
                -   key: unknown-license-reference
                    score: '92.86'
                    name: Unknown License file reference
                    short_name: Unknown License reference
                    category: Unstated License
                    is_exception: no
                    is_unknown: yes
                    owner: Unspecified
                    homepage_url:
                    text_url:
                    reference_url: https://scancode-licensedb.aboutcode.org/unknown-license-reference
                    scancode_text_url: https://github.com/nexB/scancode-toolkit/tree/develop/src/licensedcode/data/licenses/unknown-license-reference.LICENSE
                    scancode_data_url: https://github.com/nexB/scancode-toolkit/tree/develop/src/licensedcode/data/licenses/unknown-license-reference.yml
                    spdx_license_key: LicenseRef-scancode-unknown-license-reference
                    spdx_url: https://github.com/nexB/scancode-toolkit/tree/develop/src/licensedcode/data/licenses/unknown-license-reference.LICENSE
                    start_line: 11
                    end_line: 13
                    matched_rule:
                        identifier: unknown-license-reference_224.RULE
                        license_expression: unknown-license-reference
                        licenses:
                            - unknown-license-reference
                        referenced_filenames:
                            - LICENSE
                        is_license_text: no
                        is_license_notice: no
                        is_license_reference: yes
                        is_license_tag: no
                        is_license_intro: no
                        has_unknown: yes
                        matcher: 3-seq
                        rule_length: 28
                        matched_length: 26
                        match_coverage: '92.86'
                        rule_relevance: 100
                    matched_text: |
                        licensed under different licenses.
                        To find the details that apply to this artifact see the accompanying LICENSE file.
                        For more information, including possible other licensing options,
            license_expressions:
                - free-unknown
                - unknown-license-reference
            percentage_of_license_text: '24.37'
            scan_errors: []
    

    Not sure why they are only found in the JAR binary and not also in the actual source repo with the same text:

    This is weird and I got them the same way in both case. Could it be ORT handling things differently in these cases?

    Another interesting example is okhttp3 because the false positive that is found was actually introduced by yourself

    Oh well.... as the saying goes, "no good deed goes unpunished!"

    https://raw.githubusercontent.com/square/okhttp/parent-5.0.0-alpha.3/okhttp/src/main/resources/okhttp3/internal/publicsuffix/NOTICE scans this way:

    headers:
        -   tool_name: scancode-toolkit
            tool_version: 31.0.0
            options:
                input:
                    - NOTICE.1
                --license: yes
                --license-text: yes
                --license-text-diagnostics: yes
                --yaml: '-'
            notice: |
                Generated with ScanCode and provided on an "AS IS" BASIS, WITHOUT WARRANTIES
                OR CONDITIONS OF ANY KIND, either express or implied. No content created from
                ScanCode should be considered or used as legal advice. Consult an Attorney
                for any legal advice.
                ScanCode is a free software code scanning tool from nexB Inc. and others.
                Visit https://github.com/nexB/scancode-toolkit/ for support and download.
            start_timestamp: '2022-03-03T065940.353825'
            end_timestamp: '2022-03-03T065942.146784'
            output_format_version: 2.0.0
            duration: '1.792968988418579'
            message:
            errors: []
            extra_data:
                spdx_license_list_version: '3.16'
                files_count: 1
    files:
        -   path: NOTICE.1
            type: file
            licenses:
                -   key: unknown-license-reference
                    score: '60.0'
                    name: Unknown License file reference
                    short_name: Unknown License reference
                    category: Unstated License
                    is_exception: no
                    is_unknown: yes
                    owner: Unspecified
                    homepage_url:
                    text_url:
                    reference_url: https://scancode-licensedb.aboutcode.org/unknown-license-reference
                    scancode_text_url: https://github.com/nexB/scancode-toolkit/tree/develop/src/licensedcode/data/licenses/unknown-license-reference.LICENSE
                    scancode_data_url: https://github.com/nexB/scancode-toolkit/tree/develop/src/licensedcode/data/licenses/unknown-license-reference.yml
                    spdx_license_key: LicenseRef-scancode-unknown-license-reference
                    spdx_url: https://github.com/nexB/scancode-toolkit/tree/develop/src/licensedcode/data/licenses/unknown-license-reference.LICENSE
                    start_line: 4
                    end_line: 4
                    matched_rule:
                        identifier: license-intro_3.RULE
                        license_expression: unknown-license-reference
                        licenses:
                            - unknown-license-reference
                        referenced_filenames: []
                        is_license_text: no
                        is_license_notice: no
                        is_license_reference: no
                        is_license_tag: no
                        is_license_intro: yes
                        has_unknown: yes
                        matcher: 2-aho
                        rule_length: 4
                        matched_length: 4
                        match_coverage: '100.0'
                        rule_relevance: 60
                    matched_text: subject to the terms
                -   key: mpl-2.0
                    score: '100.0'
                    name: Mozilla Public License 2.0
                    short_name: MPL 2.0
                    category: Copyleft Limited
                    is_exception: no
                    is_unknown: no
                    owner: Mozilla
                    homepage_url: http://mpl.mozilla.org/2012/01/03/announcing-mpl-2-0/
                    text_url: http://www.mozilla.com/MPL/2.0/
                    reference_url: https://scancode-licensedb.aboutcode.org/mpl-2.0
                    scancode_text_url: https://github.com/nexB/scancode-toolkit/tree/develop/src/licensedcode/data/licenses/mpl-2.0.LICENSE
                    scancode_data_url: https://github.com/nexB/scancode-toolkit/tree/develop/src/licensedcode/data/licenses/mpl-2.0.yml
                    spdx_license_key: MPL-2.0
                    spdx_url: https://spdx.org/licenses/MPL-2.0
                    start_line: 4
                    end_line: 4
                    matched_rule:
                        identifier: mpl-2.0_90.RULE
                        license_expression: mpl-2.0
                        licenses:
                            - mpl-2.0
                        referenced_filenames: []
                        is_license_text: no
                        is_license_notice: no
                        is_license_reference: yes
                        is_license_tag: no
                        is_license_intro: no
                        has_unknown: no
                        matcher: 2-aho
                        rule_length: 6
                        matched_length: 6
                        match_coverage: '100.0'
                        rule_relevance: 100
                    matched_text: 'Mozilla Public License, v. 2.0:'
                -   key: mpl-2.0
                    score: '50.0'
                    name: Mozilla Public License 2.0
                    short_name: MPL 2.0
                    category: Copyleft Limited
                    is_exception: no
                    is_unknown: no
                    owner: Mozilla
                    homepage_url: http://mpl.mozilla.org/2012/01/03/announcing-mpl-2-0/
                    text_url: http://www.mozilla.com/MPL/2.0/
                    reference_url: https://scancode-licensedb.aboutcode.org/mpl-2.0
                    scancode_text_url: https://github.com/nexB/scancode-toolkit/tree/develop/src/licensedcode/data/licenses/mpl-2.0.LICENSE
                    scancode_data_url: https://github.com/nexB/scancode-toolkit/tree/develop/src/licensedcode/data/licenses/mpl-2.0.yml
                    spdx_license_key: MPL-2.0
                    spdx_url: https://spdx.org/licenses/MPL-2.0
                    start_line: 5
                    end_line: 5
                    matched_rule:
                        identifier: spdx_license_id_mpl-2.0_for_mpl-2.0.RULE
                        license_expression: mpl-2.0
                        licenses:
                            - mpl-2.0
                        referenced_filenames: []
                        is_license_text: no
                        is_license_notice: no
                        is_license_reference: yes
                        is_license_tag: no
                        is_license_intro: no
                        has_unknown: no
                        matcher: 2-aho
                        rule_length: 3
                        matched_length: 3
                        match_coverage: '100.0'
                        rule_relevance: 50
                    matched_text: MPL/2.0/
            license_expressions:
                - unknown-license-reference
                - mpl-2.0
                - mpl-2.0
            percentage_of_license_text: '33.33'
            scan_errors: []
    

    You will note that I am using these command line options:
    --license --license-text --license-text-diagnostics --yaml- which means: license, with actual license text, but limited to the exact portion of text that was matched and reported as yaml directly on screen (with the dash) rather than to a file.

    This overall looks like a case of where we could merge the license intro "subject to the terms" with a following notice.
    There are also some missing rules separately that will help catch more of the MPL URL and more of the MPL details in general.

  6. pombredanne commented on Mar 3, 2022

    @pombredanne
    MemberAuthor

    @PatteSI re: #2878 (comment)

    I created a python parser that can parse the evaluated-model.json file create by the ORT Reporter.

    This is great! Ideally what I would need is a script that would fetch the code. With that I could run extractcode to extract any archive and run a scan to get the actual details. I think this can be derived from you JSON.

  7. pombredanne commented on Mar 3, 2022

    @pombredanne
    MemberAuthor

    @porsche-rishisaxena re: #2878 (comment)

    The CSV is super useful and I can derive a script to automate re scanning from this too.

    In your case and @PatteSI case, creating these data required a lot of (useful) work.
    I am wondering what could be the tools that would make it easier to help you report these false positive.

  8. sschuberth commented on Mar 3, 2022

    @sschuberth
    Collaborator

    Could it be ORT handling things differently in these cases?

    ORT is not handling findings in binaries or sources differently per se, and is taking ScanCode findings mostly as-is (except some post-processing to remedy #2873). But it might be that some project-specific path excludes were applied in that particular case.

  9. sschuberth commented on Mar 3, 2022

    @sschuberth
    Collaborator

    I am wondering what could be the tools that would make it easier to help you report these false positive.

    For ORT, if false-positives were addressed via package configurations, we could quite easily extract the detected_license vs. the concluded_license.

    @fviernau, is that something that could be done from HERE's (probably massive) amount of package configurations?

  10. pombredanne commented on Mar 3, 2022

    @pombredanne
    MemberAuthor

    @fviernau

    is that something that could be done from HERE's (probably massive) amount of package configurations?

    If there is something that can be shared, that could be used to fix massively some of these false positive! :)

  11. pombredanne commented on Mar 5, 2022

    @pombredanne
    MemberAuthor

    Here are some related issues:

    These ones can likely be fixed with the new key phrases feature:

    These could help with the diagnostic false positive:

    This is an example of an weak detection for a new license:

    This may help with some false positives:

    @rspier ping too

    I think we should have a live call to discuss the options to fix these. What do you think?

  12. bennati commented on Mar 7, 2022

    @bennati
    Contributor

    @pombredanne I attach the false positives from a bunch of HERE curations, as produced by @PatteSI 's script.
    Hope this helps,
    falsepositives.txt

  13. sschuberth commented on Mar 8, 2022

    @sschuberth
    Collaborator

    I think we should have a live call to discuss the options to fix these. What do you think?

    To be frank, I believe having a live call with all reporters of false-positive mentioned here would be overkill. Also, I guess most people don't care too much how their issue is fixed as long as it is fixed.

    From my side, however, I'd strongly vote against hard-coding just the reported cases as false-positives. Instead, we should

    • ensure that rules always contain enough words / context to confidently identify licenses in general.
    • never allow a score of 100% for unknown licenses.
    • think about tweaking the score to be based on user feedback instead of being calculated: If a rule reportedly causes many false-positives, its score could be manually lowered.
  14. 10 remaining items

  15. PatteSI commented on Mar 23, 2022

    @PatteSI

    So the core question really is: what do you think "unknown license" means? Is it "there is a license but scancode doesn't know which one because it is not in its knowledgebase" (whether or not it is open or closed) or "scancode couldn't detect which license it is and threw its hands up"? This is conceptually a big difference.

    I think this discussion is a bit deviation from the original problem here. It doesn't matter what anyone thinks "unkown license" means. We are discussing how the heuristics could be improved and ways to give the end users more options to evaluate findings. I am talking here as an end-user of ORT, which is using ScanCode as a scanner. Now we started to migrate away from NexusIQ to ORT and we see sometime hundreds of those "unknown-license" findings in big projects. There are many examples of trivial finding where the heuristics/rules used in ScanCode are just to broad and get triggered for simple comments using the word "license". We are not only discussing the general problem here of how to improve the rule based findings. There are many example given in the first post. It's not only about "unknown" licenses.

  16. pombredanne commented on Mar 26, 2022

    @pombredanne
    MemberAuthor

    I have attached a presentation to better grap a summary of the issue:

    ScanCode-licenses-false-positive-2022-03.pptx.pdf

  17. pombredanne commented on Mar 26, 2022

    @pombredanne
    MemberAuthor

    @richardfontana I would be interested to get some feedback too
    @opensourcepilot I was re-reading your (thought-provoking) article in https://opensource.com/article/21/7/open-source-scanning-error and we are trying to fix false positives license detections for ScanCode with this issue. Your insights would be much appreciated!

  18. richardfontana commented on Mar 26, 2022

    @richardfontana

    @sutula may find this of interest

  19. alext34ms commented on Mar 28, 2022

    @alext34ms

    On the topic of making the current rule set a bit more stringent:
    The feature of making certain tokens required by putting them in {{ }} is an awesome tool to sharpen SCTK even more.
    The challenge is that we have 31.000+ rules. But that should not stop us. 😉

    Proposal: doing an "automated" retro-fit of all rules to include SPDX identifiers in {{ }}.
    I made a very KISS "one liner" that does just that. The result seems quite OK when looking at some sample rules.

    Some thoughts for this update:

    • How does it affect AND/OR rules
    • How does it affect (false-positive*) rules and what are those even?
    • Does it add any accuracy?
    • Does it pass integration tests?
    • Any others that I have missed?
    cd src/scancode-toolkit/src/licensedcode/data/rules
    for identifier in `tac ~/tmp/spdx_identifier.list`;
    do
      echo $identifier;
      for rule in `egrep -l '([^A-Z]|^)('$identifier')([^A-Z]|$)' *.RULE`;
      do
        sed -i -E s/\(\[\^A-Z\{\{\\]\|\^\)\($identifier\)\(\[\^A-Z\}\}\]\|\$\)/\\1\{\{$identifier\}\}\\3/g $rule;
      done;
    done

    spdx_identifier.list

    Above one liner makes changes to about 6400/31000 rules.

  20. pombredanne commented on Mar 29, 2022

    @pombredanne
    MemberAuthor

    @alext34ms re:

    doing an "automated" retro-fit of all rules to include SPDX identifiers in {{ }}.
    I made a very KISS "one liner" that does just that. The result seems quite OK when looking at some sample rules.

    Sleek! very smart. I like it

    • How does it affect AND/OR rules

    I do not think there it would have any impact.

    • How does it affect (false-positive*) rules and what are those even?

    These should be left alone. These are rules that can be matched only exactly and that are about licenses but are NOT license notices or texts. They should be used sparingly as a last resort. For instance, this text:

    copyright info have been adapted to avoid the violation of the GPL license

    is NOT a GPL-related notice, but some commentary about the GPL license and would be a typical case for a "false positive" rule.

    • Does it add any accuracy?

    It should, but it may also degrade and miss some matches in a few corner cases. These could be caught separately by the --unknow-license option though.

    • Does it pass integration tests?

    It will likely make some fail.

    • Any others that I have missed?

    I think the approach could be refined using a Python script as we have code that handle the RULEs and has all SPDX licenses alright and we could also expand this to a few more things:

    • SPDX id alright as you suggest
    • the license short and long name
    • potentially the SPDX names, and URL

    Some scripts examples to use as a base are in https://github.com/nexB/scancode-toolkit/blob/develop/etc/scripts/licenses/

  21. pombredanne commented on May 12, 2022

    @pombredanne
    MemberAuthor

    @vargenau Another case to track here #2905

  22. borisbaldassari commented on May 17, 2022

    @borisbaldassari

    Hi Philippe, all,

    First things first: thanks for the good work people! You're great!
    I'd like to contribute some feedback regarding false positives (/wrong license) we get at the Eclipse Foundation. One of our main issues is with the canonical headers used with the EPL-2.

    /*********************************************************************
     * Copyright (c) 2019 Red Hat, Inc.
     *
     * This program and the accompanying materials are made
     * available under the terms of the Eclipse Public License 2.0
     * which is available at https://www.eclipse.org/legal/epl-2.0/
     *
     * SPDX-License-Identifier: EPL-2.0
     **********************************************************************/
    

    Some lines (typically {4,5}) are recognised as LicenseRef-scancode-unknown-license-reference even with the SPDX tag sitting right behind. It seems that another license text (license-intro_29.RULE) is matched before the EPL-2.0 text, so even adding various variations of the headers (line ends, etc.) doesn't help. Setting the license-score to 100 helps a bit (i.e. less wrong violations), but still not enough to make this case go away.

    Would it be helpful to provide a list of false positives / wrong identifications? I'll be happy to provide one if so.

  23. AyanSinhaMahapatra commented on May 17, 2022

    @AyanSinhaMahapatra
    Member

    @borisbaldassari Thanks for your feedback and report.

    We are working on this and this specific issue of a license_intro being present before detections is going to be fixed, this is WIP and hasn't landed yet.

    Would it be helpful to provide a list of false positives / wrong identifications? I'll be happy to provide one if so.

    This would be extremely helpful, we will use this for testing this new feature extensively, as we are using the other lists contributed here. Thanks a lot!

  24. borisbaldassari commented on May 18, 2022

    @borisbaldassari

    Hi @AyanSinhaMahapatra Thanks for the head-up! I'll wait for the landing and give it a try. :-)

    Please find below a list of unknown-license false positives found in a few Eclipse projects (Che, JGit, CDT, Tycho). If needed I can analyse more projects -- but since we're using ORT we don't have direct access to the scancode output, so I need to run it separately (and manually).

    scancode_fp_eclipse.tar.gz

  25. borisbaldassari commented on May 18, 2022

    @borisbaldassari

    Please also find attached the Python script used to generate the csv's, if it's useful.

    extract_false_positives.py.tar.gz

  26. pombredanne commented on May 20, 2022

    @pombredanne
    MemberAuthor

    @borisbaldassari Thank you ++

  27. pombredanne commented on May 23, 2022

    @pombredanne
    MemberAuthor

    Another short SSPL false positive #2975

  28. PatteSI commented on Aug 2, 2023

    @PatteSI

    As @pombredanne also asked in my initial issue for a list of false positives I just wanted to mention that the ORT community also started sharing curations for false positives. I guess ScanCode ist still one of the most widely used scanning component in ORT so they might all be relevant for you. Check out their curantions and package configurations: https://github.com/oss-review-toolkit/ort-config

  29. added
    RFCRequest for comment, from the maintainers, users and the larger aboutcode community.
    on Jun 4, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    RFCRequest for comment, from the maintainers, users and the larger aboutcode community.bugimprove-license-detection

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions