Not that anyone actually needed it, but it was fun to filter and summarize. (caffeine sink)
On Jul 11, 11:41 pm, Warwick Prince <warwi...@mushroomsys.com> wrote: > Thanks for the 'heads-up' Eric :-) > > > > > > > > > ! Nothing to see here, move right along ! > > > Except... Couple of interesting additions coming up in PostgreSQL 9.1 > > (still in beta) for anyone who's interested. Release notes: > >http://developer.postgresql.org/pgdocs/postgres/release-9-1.html A > > couple of selected items I found of interest: > > > * New support for CREATE UNLOGGED TABLE -- a new type of table that is > > sort of in between temporary tables and ordinary tables. They are not > > crash-safe as they are not written to the write-ahead log and are not > > replicated if you have replication set up, but the tradeoff is they > > can be written to a lot faster. Could use these to speed up testing, > > or in other non-production scenarios where crashproofness is not a > > concern. > > > * New support for Synchronous replication -- primary master waits for > > a standby to write the transaction information to disk before > > acknowledging the commit. This behavior can be enabled or disabled on > > a per-transaction basis. Also a number of new settings related to > > keeping a 'hot standby'. > > > * They added a true serializable transaction isolation level. > > "Previously, asking for serializable isolation guaranteed only that a > > single MVCC snapshot would be used for the entire transaction, which > > allowed certain documented anomalies." The old snapshot isolation > > behavior will now be accessible by using the "repeatable read" > > isolation level. > > > --This one might be particularly interesting for SQLAlchemy-- > > * INSERT, UPDATE, and DELETE will now be allowed in WITH clauses; > > these commands can use RETURNING to pass data up to the containing > > query. While not strictly necesary, this can improve the clarity of > > SQL emitted by eliminating some nested sub-SELECTs. There is other > > cool stuff you can accomplish with this such as deleting rows from one > > table according to a WHERE clause inside of a WITH...RETURNING, and > > inserting the same rows into another table in the same statement. The > > recursive abilities of WITH statements can also be used now to perform > > useful maneuvers like recursive DELETEs in tree structures (as long as > > the data-modifying part is outside of the WITH clause). > > > * New support for per-column collation settings (yawn... but someone > > somewhere needs this while migrating or something) > > > * New support for foreign tables -- allowing data stored outside the > > database to be used like native postgreSQL-stored data (read-only). > > > * Enum types can now be added to programmatically (i don't know if > > they can be removed from) via ALTER TYPE > > > * Added CREATE TABLE IF NOT EXISTS syntax -- seems like SA's DDL > > machinery might want to use that in the checkfirst=True case to > > eliminate the separate check operation? A minor matter, but nice. > > > * Added transaction-level "advisory locks" (non-enforced, application- > > defined) similar to existing session-level advisory locks. > > > * Lots more (a dump + restore will be required between 9.0 and 9.1) > > > -- > > You received this message because you are subscribed to the Google Groups > > "sqlalchemy" group. > > To post to this group, send email to sqlalchemy@googlegroups.com. > > To unsubscribe from this group, send email to > > sqlalchemy+unsubscr...@googlegroups.com. > > For more options, visit this group > > athttp://groups.google.com/group/sqlalchemy?hl=en. -- You received this message because you are subscribed to the Google Groups "sqlalchemy" group. To post to this group, send email to sqlalchemy@googlegroups.com. To unsubscribe from this group, send email to sqlalchemy+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/sqlalchemy?hl=en.
