Refactor: scheduler: functionize checking for anonymous instance probe

Refactor: scheduler: best practices for promotion_order()

Refactor: scheduler: functionize checking whether node is allowed to run…

Refactor: scheduler: best practices for filter_anonymous_instance()

Refactor: scheduler: best practices for lookup_promotion_score()

Refactor: scheduler: best practices for promotion_score()

Low: scheduler: calculate group promotion scores correctly

Test: scheduler: update regression test for promotion score correction

API: libcrmcommon: add pcmk_readable_score()

API: libcrmcommon: deprecate score2char()

Refactor: libpacemaker,libpe_status: avoid score2char()

Refactor: libpacemaker,libpe_status: avoid score2char_stack()

API: libcrmcommon: deprecate score2char_stack()

Test: libcrmcommon: rename score2char_stack_test.c

Refactor: scheduler: avoid unnecessary test

Test: libcrmcommon: update unit test for pcmk_readable_score()

Refactor: tools: functionize individual resource checks in crm_resource

Refactor: tools: use a flag to indicate locked resources in crm_resource

Fix: tools: crm_resource target-role check should use meta-attribute table

Fix: tools: crm_resource is-managed check should use meta-attribute table

Refactor: tools: pass node to cli_resource_check()

Refactor: tools: drop unused argument from cli_resource_check()

Refactor: tools: drop unused argument from resource-reasons-list message

Feature: tools: crm_resource --why now checks node health status

Low: schemas: copy crm_resource API schema in preparation for changes

Fix: tools: correct crm_resource --why schema to match actual output

Test: cts-cli: use validated XML output for crm_resource --why test

Low: schemas: simplify crm_resource --why schema

Low: schemas: update crm_resource --why schema for new health check

Test: cts-cli: add tests for checking resource status on unhealthy node

Log: libcrmcommon: no need for "Cleaning up memory" message

Build: RPMs: provide a virtual package needed by booth

Refactor: scheduler: make "ordered" clone variant member a flag

Refactor: scheduler: make more clone variant members into flags

Refactor: scheduler: drop unnecessary arguments to child_promoting_constraints()

Refactor: scheduler: drop dead code in child_promoting_constraints()

Refactor: scheduler: best practices when ordering instance promotion

Refactor: scheduler: drop dead code in child_demoting_constraints()

Refactor: scheduler: drop unnecessary arguments to child_demoting_constraints()

Refactor: scheduler: best practices when ordering instance demotion

Refactor: scheduler: best practices when checking for role changes

Refactor: scheduler: best practices for apply_promoted_location()

Refactor: scheduler: best practices for node_to_be_promoted_on()

Refactor: scheduler: pcmk__node_available() optionally checks guest

Refactor: scheduler: pcmk__node_available() optionally checks score

Refactor: scheduler: best practices for sort_promotable_instance()

Low: controller: validate controller requests better

Low: controller: validate execution requests better

Low: controller: validate transition requests better

Log: controller: improve logs when routing messages

Low: libcrmcommon: validate attribute manager IPC messages better

Low: pacemakerd: validate IPC messages better

Low: pacemaker-schedulerd: validate IPC messages better

Log: libcrmcommon: validate controller IPC messages better

Low: libcrmcommon: validate pacemaker-schedulerd messages better

Low: libcrmcommon: validate pacemakerd messages better

Low: libcib: validate requests better

Refactor: cib: remove dead code

Low: scheduler: improve validation of resource XML

Low: fencer: improve topology level validation

Refactor: fencer: simplify default request handler

Refactor: controller: make process_te_message() easier to follow

Log: various: improve messages that use pcmk__s()

Refactor: controller: check pointer for NULL *before* using it

Log: libcrmcommon: don't try to log NULL

Refactor: scheduler: break line at 80 characters

Log: libcib: quote ID in trace messages

Log: libcrmcommon: improve pcmk_controld_api_reprobe() debug message

API: libcib: deprecate cib_api_operations_t:delete_absolute()

Log: libcrmservice: fix log message arguments

Doc: Pacemaker Explained: document ACL groups are now supported

Doc: books: fix upload directory for build information

Doc: Pacemaker Development: rearrange API sections

Doc: Pacemaker Development: reorganize C chapter sections

Doc: Pacemaker Development: add a section about C macros

Doc: libcrmcommon: correct error in pcmk__xml_update() doxygen description

Doc: Pacemaker Development: mention closing task from commit message

Doc: libcrmservice: fix typo in comment

Low: tools: improve error message wording in crm_resource

Low: libcib: make static analysis happy

Log: attrd: improve dampening logs

Test: cts-lab: ignore fencer connection loss if pacemaker-based is shot

Refactor: scheduler: best practices for parent_node_instance()

Refactor: scheduler: rename sort_clone_instance()

Refactor: scheduler: functionize creating a hash table with one node

Refactor: scheduler: functionize applying parent's colocation scores

Refactor: scheduler: drop unused argument from pcmk__cmp_instance()

Refactor: scheduler: best practices when sorting instances by colocation scores

Refactor: scheduler: best practices when sorting instances

Refactor: scheduler: best practices when checking whether resource failed

Refactor: scheduler: functionize checking whether node is allowed

Refactor: scheduler: expose instance number comparator in libpacemaker

Test: cts-scheduler: update scores for instance number change

Low: scheduler: compare instance numbers not IDs for assignment order

Refactor: scheduler: remove dead code

Test: cts-scheduler: update expected scores for instance sorting change

Low: scheduler: don't prefer greater anti-colocation when ordering instances…

Refactor: scheduler: drop pcmk__is_unfence_device()

Refactor: scheduler: drop redundant resource/node arguments

Fix: scheduler: restore line dropped by 34209e2e63