Low: cts, xml: New upgrade-3.10-2.xsl XSLT stylesheet
Test: cts-schemas: Add test for dropping duplicate nvpairs within nvset
Test: cts-schemas: Add test for multiple top-level rules in rsc_location
Refactor: xml: Drop duplicate nvpairs within a given nvset
Feature: schemas: Disallow multiple top-level location rules
Feature: libpe_rules: Drop support for name with id-ref attribute
Fix: libcrmcommon: Detect newly created alerts section
Feature: schemas: Disallow nvpair name with id-ref attribute
Test: cts-schemas: New test for setting unset validate-with attribute
Low: cts, xml: New upgrade-3.10-1.xsl XSLT stylesheet
Low: xml: Bump validate-with cib attribute in 3.10 XSL transformation
Refactor: xml: Sort nvset elements by score within a given parent
Test: cts-schemas: Add test for sorting nvsets
Test: cts-schemas: Add test for dropping value-less nvpairs
Feature: schemas: Disallow nvpairs without value attributes
Refactor: libcrmcommon: Fix overlong lines from 67d63ee7
Test: libcrmcommon: Use proper XML functions in unit test
Test: libcrmcommon: Add pcmk__cib_element_in_patchset() unit tests
Fix: libcrmcommon: Detect newly created alerts section
Closed by commit c6a8725f. Ping node support has not been dropped yet, but it will be soon.
Refactor: libpe_status: Clean up nvpair processing in populate_hash()
Fix: libpe_status: Ignore nvpair with invalid id-ref
Fix: libpe_status: Don't look for value in nvpair with id-ref
Feature: libpe_status: Drop support for multiple top-level nvset rules
Feature: libpe_status: Drop support for nodes with invalid type
Refactor: xml: Add prefix for XSLT-generated IDs
Refactor: xml: Resolve id-refs at beginning of 3-to-4 upgrade pipeline
Test: cts-schemas: Add test for resolving id-refs
Feature: xml: Allow previously disallowed primitive meta-attributes
Feature: xml: Allow previously disallowed cluster properties
Refactor: xml: New upgrade-3.10-0.xsl XSLT stylesheet
Test: cts-cli: Update upgrade and validity outputs for 4.0
Feature: xml: Allow previously disallowed "required" op meta-attribute
Test: cts-schemas: Fix cleanup of .up and .up.err files
Low: xml: Copy schemas in preparation for changes
Refactor: cts-schemas: Use local variables
Test: cts-cli: Update cibadmin digest for pacemaker-4.0
Doc: Pacemaker Explained: Render footnote correctly
Doc: Pacemaker Explained: Render footnote correctly
Log: libpacemaker: Log correct score in pcmk__apply_coloc_to_priority()
Test: cts-scheduler: Add tests for cloned group promoted colocations
Refactor: libpacemaker: apply_coloc_score() returns priority delta
Refactor: libpacemaker: Clean up pcmk__apply_coloc_to_priority()
Fix: libpacemaker: Apply promotion priority to cloned group instance
Test: cts-scheduler: Test for doubled colocation scores for promotion
Fix: libpacemaker: Don't apply colocation twice for promotion priority
Test: cts-scheduler: Update outputs for double colocation fix
Test: cts-scheduler: Test for doubled colocation scores for promotion
Test: cts-scheduler: Update outputs for double colocation fix
Fix: libpacemaker: Don't apply colocation twice for promotion priority
Log: libpacemaker: Log correct score in pcmk__apply_coloc_to_priority()
Test: cts-scheduler: Add tests for cloned group promoted colocations
Refactor: libpacemaker: Clean up pcmk__apply_coloc_to_priority()
Fix: libpacemaker: Apply promotion priority to cloned group instance
Refactor: libpacemaker: apply_coloc_score() returns priority delta
Fix: libpe_status: Use first action meta-attr within nvset if duplicates
Fix: libcib: Don't overwrite property values for CIB config hash
Refactor: libcrmcommon: Staticize pcmk__xml_update()
Refactor: libpe_status: Handle overwrite directly in add_node_attrs()
Refactor: libpe_status: Drop pe__unpack_dataset_nvpairs() overwrite arg
Doc: Pacemaker Explained: Document option precedence
Fix: tools: crm_node -i must initialize nodeid before passing pointer
Fix: various: Don't set cluster-layer node ID as XML ID
Fix: tools: crm_node -i must initialize nodeid before passing pointer
Refactor: various: Don't set cluster-layer node ID as XML ID
nrwahl2 renamed
T847: crm_node -i passes pointer to uninitialized variable to pcmk__query_node_info() from
crm_node -i passes uninitialized pointer to pcmk__query_node_info() to
crm_node -i passes pointer to uninitialized variable to pcmk__query_node_info().
nrwahl2 renamed
T847: crm_node -i passes pointer to uninitialized variable to pcmk__query_node_info() from
crm_node -i passes invalid pointer to pcmk__query_node_info() to
crm_node -i passes uninitialized pointer to pcmk__query_node_info().
Test: cts-schemas: Drop cts_scheduler test suite
Test: cts-schemas: Implement only one test per upgrade pipeline
Test: cts-schemas: Split test input/output data into separate dirs
Test: cts-schemas: Remove output files on success
Test: cts-schemas: Drop -B option
Feature: test: Move xml/regression.sh to cts/cts-schemas
Test: cts-schemas: One test output per transformation
This is actually only allowed by the schema, so there's nothing to transform in the 3.10 to 4.0 upgrade. How do we want to go about this? I'm thinking just drop the Pacemaker code for it and leave the schema alone until we drop support altogether.
Re: content, be aware that Pacemaker currently strips text nodes from the CIB. That can be changed but it's possible that certain code (for example, related to XML private data being non-NULL) is relying on the assumption of no text nodes in the CIB. Also be careful that whitespace-only text nodes don't cause problems.
API: tools: Drop crm_mon --as-html option
API: tools: Drop crm_mon --disable-ncurses option