| Commit message (Collapse) | Author | Age |
... | |
| |
| |
| |
| | |
"sqlcache" in the all-one-word style. Still has issues to do with concurrency, retrying transactions, and foreign function calls that either rely on state or have side effects.
|
| |\
| |/
|/| |
|
| | |
|
| |\
| |/
|/| |
|
| |
| |
| |
| | |
table-match-based invalidation. Still has problems parsing non-Postgres SQL dialects properly.
|
| | |
|
| |
| |
| |
| | |
(based on patch by Evan Danaher)
|
| | |
|
| | |
|
| |
| |
| |
| | |
serialization failures
|
| |
| |
| |
| | |
We assume that FFI code may create new contextes with id left unassigned
|
|/ |
|
| |
|
| |
|
| |
|
|\ |
|
| | |
|
| | |
|
|\| |
|
| | |
|
| |
| |
| |
| | |
system call
|
| |
| |
| |
| | |
rollback a failed transaction
|
| |
| |
| |
| | |
of client ID in RPCs
|
| | |
|
| | |
|
| | |
|
| |
| |
| |
| | |
program analysis
|
| | |
|
| | |
|
| |
| |
| |
| | |
optimizations
|
| |
| |
| |
| | |
initialization of database connections, to avoid the overhead in handlers that don't use SQL
|
| | |
|
|/ |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|