Add versioning approach documentation #2653

Signed-off-by: Philippe Ombredanne <[email protected]>

Author: pombredanne

docs/source/misc/index.rst

@@ -7,3 +7,5 @@
    faq
    support
    perf_report
+   versioning
+   

docs/source/misc/versioning.rst

@@ -0,0 +1,57 @@
+.. versioning:
+
+
+Versioning approach
+==========================
+
+ScanCode is composed of code and data (mostly license data used for license
+detection). In the past, we have tried using calver for code versioning to also
+convey that the data contained in ScanCode was updated but it proved to be not
+as clear and as effective as planned so we are switching back to semver which is
+simpler and overall more useful for users. We also want to provide hints about
+JSON output data format changes.
+
+Therefore, this is our versioning approach starting with version 30.0.0:
+
+- ScanCode releases are versioned using semver as documented at
+  https://semver.org using major.minor.patch versioning.
+
+- Significant changes to the data (license or copyright detection) is considered
+  a major version change even if there are no code changes. The rationale is
+  that in our case the data has the same impact as the code. Using outdated data
+  is like using old code and means that several licenses may not be detected
+  correctly. Any data change triggers at least a minor version change.
+
+- We will signal separately to users with warnings messages when ScanCode needs
+  to be upgraded because its data and/or code are out of date.
+
+
+In addition to the main code version, we also maintain a secondary output data
+format version using also semver with two segments. The versioning approach is
+adapted for data this way:
+
+- The first segment --the major version-- is incremented when data attributes
+  that are removed, renamed, changed or moved (but not reordered) in the JSON
+  output. Reordering the attributes of a JSON object is not considered as a
+  change and does not trigger a version change.
+
+- The second segment --the minor version-- of the output format is incremented
+  for an addition of attributes to the JSON output.
+
+- We store the output format version string in the JSON output object as the
+  first attribute and display that also in the help.
+
+- This output format versioning applies only to the JSON, pretty-printed JSON,
+  YAML and JSON lines formats. It does not apply to CSV and any other formats.
+  For these other formats there is no versioning and guaranteed format stability
+  (or there may be some other rationale and convention for versioning like for
+  SPDX).
+
+- The output format version is incremented by when a new ScanCode tagged release
+  is published
+
+- We document in the CHANGELOG the output format changes in any new format version.
+
+- For any format version changes, we will provide a documentation on the format
+  and its updates using JSON examples and a comprehensive and updated data
+  dictionary. See https://github.com/nexB/scancode-toolkit/issues/2008 for details.