News and Updates

SQLAlchemy 1.0.8 Released

permalink

SQLAlchemy release 1.0.8 is now available.

Release 1.0.8 comes almost immediately after 1.0.7, as a new issue involving the connection pool has been identified and fixed which impacts any application that relies upon consistent behavior of the .info dictionary on a connection that is undergoing reconnect attempts. Applications and libraries which make use of connection pool event handlers may benefit from this release, as it repairs the behavior of the .info dictionary and reduces the likelihood of stale connections being passed to the "checkout" handler.

Changelog for 1.0.8 is at:

Changelog

SQLAlchemy 1.0.8 is available on the Download Page.

SQLAlchemy 0.9.10 Released

permalink

SQLAlchemy release 0.9.10 is now available.

This is a maintenance release of the 0.9 series, delivering an array of bug fixes and small feature enhancements which typically have been backported from the 1.0 series. As the 1.0 series has been in production use for some months now, the 0.9 series is expected to receive only critical bugfixes subsequent to this version.

Changelog for 0.9.10 is at:

Changelog

SQLAlchemy 0.9.10 is available on the Download Page.

SQLAlchemy 1.0.7 Released

permalink

SQLAlchemy release 1.0.7 is now available.

This release contains a series of regression fixes, some additional bugfixes, and a few new Core level features, including a new .cast() method available on column constructs as well as support for many new keywords used in creating sequences.

Next on the roadmap for SQLAlchemy are the beginnings of the 1.1 series; the 1.1 milestone has already accumulated several dozen proposed bugfixes and features. Look for development and documentation of the 1.1 series to begin within the coming weeks.

Changelog for 1.0.7 is at:

Changelog

SQLAlchemy 1.0.7 is available on the Download Page.

SQLAlchemy 1.0.6 Released

permalink

SQLAlchemy release 1.0.6 is now available.

This release includes a handful of new bugfixes and some small feature adds, pushed along by a high level of community involvement.

The most critical issue fixed involves the versioning feature of the ORM, where the version counter of a row could be incremented when there was otherwise no net change on the row; this was a regression from 0.9. A few other minor issues were also regressions or "semi-regressions" (e.g. a bug that existed in 0.9, but is more likely to be encountered in 1.0 due to greater use of a feature). Issues involving the SQL Server VARBINARY type, ORM queries for "single inheritance" mappings, and support for the very latest (2.7.1) release of psycopg2cffi are resolved.

The new features are all Postgresql centric, and include support for setting the storage parameters of a Postgresql index, as well as reflection of those parameters and the "using" algorithm of an index; fixes to the ExcludeConstraint construct to allow more open-ended SQL expressions, and support for controlling the size of the row buffer used when "server side cursor" mode is used.

Changelog for 1.0.6 is at:

Changelog

SQLAlchemy 1.0.6 is available on the Download Page.

SQLAlchemy 1.0.5 Released

permalink

SQLAlchemy release 1.0.5 is now available.

Release 1.0.5 is mostly a bugfix release, in particular repairing some issues with MutableType as well as improving support for Postgresql under Pypy.

It also has an improvement to the SQL Server dialect which removes an old legacy system of attempting to rewrite SQL statements that include a schema qualifier; the system remains in place by default for the 1.0 series, however can be disabled using a new create_engine() flag. As the flag will most likely be reversed in the 1.1 series, a warning is emitted when schema-qualified tables are used without a true/false setting for this flag. SQL Server applications that use remote schemas will want to begin experimenting with disabling this behavior in preparation for it being disabled by default.

Changelog for 1.0.5 is at:

Changelog

SQLAlchemy 1.0.5 is available on the Download Page.