Skip to content

0.15.0
Compare
Choose a tag to compare
@github-actions github-actions released this 05 Dec 22:11
v0.15.0
4155b93

Merged PRs

doltgresql

  • 1035: Tests and schema fixes for dolt.status, dolt.merge_status and dolt_constraint_violations_*
  • 1033: switch DoltgresType to be used as pointer
  • 1029: Fix schemas for dolt system tables with boolean columns
    Dolt PR: dolthub/dolt#8634
  • 1014: Reworked DoltgresType function access
    This reworks how functions are accessed within DoltgresType, in addition to adding a faster path for functions that fall under specific criteria.
  • 1004: Add tests for dolt_statistics
    GMS: dolthub/go-mysql-server#2762
    Dolt: dolthub/dolt#8589
  • 1003: Add diff tests for dolt.docs
  • 998: implement to_char for timestamp (mostly)
    This PR adds support for most of the date format specifiers mentioned here: https://www.postgresql.org/docs/current/functions-formatting.html
  • 996: Support dolt.rebase, add tests
  • 995: Tests for dolt_ignore per schema
    Dolt PR: dolthub/dolt#8581
  • 994: Tests and fixes for dolt_rebase and dolt_ignore
    Dolt PR: https://github.com/dolthub/dolt/pull/8578/files
  • 988: Added a mini sysbench test that runs on PRs
    This adds a new workflow that runs on every PR, similar to the regression tests. This runs a small version of the sysbench tests, primarily to detect large performance changes that a PR may unexpectedly introduce. For now, the variance is set to ±10%. That is, if it's under 10% in either direction, it'll just display as a white color. If it's greater, then it'll display as green or red respectively to indicate a possible performance regression. There's variability in the tests, that we'll want to account for, but perhaps the 10% window is too large. Or maybe we should run the tests for longer.
  • 986: Revert "Merge pull request #904 from dolthub/jennifer/type"
    This reverts commit 40dba4f, reversing changes made to 165c30e.
    Reverting this change because of performance regression.
  • 983: Major improvements to auth
    This adds a myriad of tests, in addition to:
    • Schema-Wide Table Privileges
    • Schema Privileges
    • Database Privileges
    • Role Memberships
      There are also many smaller changes, such as disallowing non-superusers from modifying superusers, and many more.
  • 981: support insert DEFAULT VALUES
    This is postgres special syntax that allows you to just insert all default values.
    This is equivalent to insert into <tbl> values () in MySQL.
  • 979: support EXCEPT set operation
  • 978: support EXISTS subquery
    This PR adds support for EXISTS <subquery> clause.
    Additionally, we have to add a bunch of skips namely in the setup queries in the testing/go/regression/results directory; these queries encounter performance issues for joins over pg_catalog tables discussed here: #940
  • 977: update Regression test to display 50 tests each at most with total number of tests
  • 970: catch more unsupported partition of queries
    These were panicking before, but now we throw unsupported
  • 965: Fix DEFAULT insert for NULL or unspecified column default values
  • 964: Add dolt_schema and dolt_procedure tests
  • 963: TestUpdate and TestDelete
  • 949: Release v0.14.1
    Created by the Release workflow to update DoltgreSQL's version
  • 945: support with
    somewhat related: #928
  • 904: create DoltgresType struct used for all types
    This PR replaces current DoltgresType interface with DoltgresType struct. All built-in types will have pre-defined values for this struct. User-defined types will populate the struct fields with appropriate values.
    Depends on dolthub/go-mysql-server#2736

Closed Issues

  • 985: pull requests need #benchmark so latency regressions can be caught before they're merged.
  • 943: Unable to set statement_timeout
  • 937: doltgres 0.14.0 regression test failure
  • 843: Error on restore into doltgres (pg_restore)
Assets 8
Loading