Many hyperlinks are disabled.
Use anonymous login to enable hyperlinks.
Changes In Branch snapshots-lock-wal Excluding Merge-Ins
This is equivalent to a diff from 2c03d8b8f0 to d71eeaab9e
|13:39||Lock the wal file for all snapshot transactions, even if they would not otherwise require this, preventing checkpointers and writers from wrapping the wal file. This means that if one connection has an open snapshot transaction it is guaranteed that a second connection can open a transaction on the same snapshot. (check-in: b81a31495b user: dan tags: trunk)|
|09:37||Add an extra snapshot related test case. (check-in: 24a95e1437 user: dan tags: trunk)|
|07:46||Experimental change so that snapshot transactions always lock the wal file - preventing writers or truncate-checkpointers from wrapping it. (Closed-Leaf check-in: d71eeaab9e user: dan tags: snapshots-lock-wal)|
|20:18||Merge in all the trunk enhancements of the previous 7 months. The LIKE optimization has stopped working when there is an ESCAPE - that problem will be addressed in a subsequent check-in. (Leaf check-in: 8d2a1cca61 user: drh tags: est_count_pragma)|
|16:23||Use the updated Win32 VFS semantics for winOpen from check-in [5d03c738e9] for WinRT, et al, as well. (check-in: 2c03d8b8f0 user: mistachkin tags: trunk)|
|12:52||Partial backout of check-in [e0af9a904076]. It turns out we do need some extra space at the end of the record blob as an overrun area to use when decoding a maliciously malformed record. (check-in: 403b88a894 user: drh tags: trunk)|
Changes to src/wal.c.