This Day In History For 2020-01-17

1 Year Ago (more context)

2019-01-17
20:19		Disable the xfer-optimization if the two tables have the same root page due to schema corruption. (check-in: f31b3bd2a6 user: drh tags: trunk)
20:06		Fix another corruption related crash in fts5. (check-in: 4538d9afe4 user: dan tags: trunk)
19:33		Add the SQLITE_ENABLE_EARLY_CURSOR_CLOSE compile-time option which causes read cursors to be closed after their usefulness ends during a two-pass UPDATE. (check-in: 7def6c8edd user: drh tags: trunk)
19:11		Fix a buffer overwrite that could occur when running an fts5 prefix query against a corrupt database. (check-in: 3910b5639d user: dan tags: trunk)
17:39		Fix problems with joining two or more fts5_vocab tables that access the same underlying fts5 table. (check-in: 49956395e1 user: dan tags: trunk)
15:47		Sync up with the latest enhancements on trunk. (check-in: 475a179a27 user: drh tags: apple-osx)
15:40		Revamp the SrcList allocator routines to be methods of Parse instead of being methods of the "sqlite3" object, so that they can leave better error messages when the SrcList object grows too large. (check-in: df08d472b0 user: drh tags: trunk)
14:34		Limit the size of SrcList objects to 200 entries (compile-time configurable using -DSQLITE_MAX_SRCLIST=n). The maximum number of tables in a join has always been 64, so this is not a real constraint on capability. Limiting the size of a SrcList prevents DOS attacks (discovered by OSSFuzz) using crazy nexted CTE joins. (check-in: 7cac614d5d user: drh tags: trunk)
04:40		Fix a corner-case for the logic that cause an insert of a NULL into an INTEGER PRIMARY KEY column to be converted into a valid integer key, when the NULL results from a CASE expression that lacks an ELSE clause. (check-in: 9a425051e7 user: drh tags: trunk)
03:43		Ensure that the variable-length integer decoder in FTS3 never tries to left-shift a negative number. (check-in: 10ffc1fe00 user: drh tags: trunk)
01:06		Fix a failure in sqlite3ExprCompare() (check-in: 835e2cc55f user: drh tags: trunk)

2 Years Ago (more context)

2018-01-17
21:14		Alternative implementation for the internal sqlite3Pow10() utility for MSVC, which is more accurate on that platform. (check-in: 469b96be53 user: drh tags: trunk)
20:57		Fix a problem in the sessions module with logging sqlite_stat1 rows for which (idx IS NULL) is true. (check-in: 25bf734be1 user: dan tags: sessions-stat1)
17:38		Fix a problem causing the sessions module to occasionally lose track of rows with composite primary keys when there are two rows with the same text value in the leftmost column of the PK. (check-in: 09aed13678 user: dan tags: trunk)
16:11		Fix main.mk so that testfixture can be built either from the amalgamation or from individual source files. No changes to code. (check-in: a8aea925f8 user: dan tags: trunk)
13:15		Fix harmless compiler warnings, mostly unused parameters for UDFs in the CLI. (check-in: bfbeffab77 user: drh tags: trunk)
12:58		Omit the single test from zipfile.test that uses json functionality in non-SQLITE_ENABLE_JSON1 builds. (check-in: 6bb2a10fad user: dan tags: trunk)
12:57		Update test file fts5fault6.test to account for test tokenizers implemented in Tcl returning SQLITE_ERROR instead of SQLITE_NOMEM following an OOM error. (check-in: c232f6424a user: dan tags: trunk)
01:40		Simplifications to winOpenSharedMemory in the Win32 VFS. (check-in: 3e04999dab user: mistachkin tags: trunk)
01:26		Corrections to error code handling in os_win.c, pursuant to walfault.test. (check-in: 568192228c user: mistachkin tags: trunk)
01:15		Fix duplicate test names in 'walfault.test'. (check-in: 7274d05ff4 user: mistachkin tags: trunk)

3 Years Ago (more context)

2017-01-17
10:41		Fix a problem that could cause a spurious SQLITE_NOMEM error when attempting to resume an RBU operation if the previous client failed right after completing the incremental checkpoint. Also a "cannot vacuum wal db" error that could occur when resuming an RBU vacuum if an error (OOM or IO error) occurs during the incremental checkpoint. (check-in: 681d96eb82 user: dan tags: trunk)
00:10		Disable intrinsic functions for Windows using Clang, due to reports of linkage errors. This causes a 0.6% performance reduction. We will want to revisit this change in the future. (check-in: 7fd560c6d2 user: drh tags: trunk)

5 Years Ago (more context)

2015-01-17
20:01		Ensure an up to date copy of the fts5 configuration has been loaded into memory before attempting to modify the same configuration. (check-in: f30afd209a user: dan tags: fts5)
17:48		Improve the performance of the fts5 porter tokenizer implementation. (check-in: 96ea600440 user: dan tags: fts5)
15:26		Fix a harmless compiler warning the VFS for Windows. (check-in: ceac571f53 user: drh tags: trunk)

15 Years Ago (more context)

2005-01-17
22:08		Infrastructure changes to handle name resolution differently. This is needed to fix various long-standing problems with column names in joins. It will also make the implementation of correlated subqueries easier. (CVS 2228) (check-in: 4a7534396a user: drh tags: trunk)
08:57		Fix a bug reported on the mailing list concerning a conflict between "INSERT INTO ... SELECT" statements and the "SELECT max(x) FROM tbl" optimization. (CVS 2227) (check-in: 5a9da62ae3 user: danielk1977 tags: trunk)
07:53		Fix some memory leak problems with corrupt.test and auto-vacuum databases. (CVS 2226) (check-in: 6244252915 user: danielk1977 tags: trunk)
03:42		Add incomplete, preliminary drafts of new documentation. (CVS 2225) (check-in: a01159e8cb user: drh tags: trunk)
03:40		Fix a memory leak that occurs as a result of an IO error. (CVS 2224) (check-in: 1edfdcbf14 user: danielk1977 tags: trunk)
02:12		Change some assert()s that could fail if the database is corrupt to return SQLITE_CORRUPT instead. (CVS 2223) (check-in: 2d58c0afa7 user: danielk1977 tags: trunk)
01:33		Have sqlite3pager_get() return SQLITE_CORRUPT for a page number greater than 2^31. (CVS 2222) (check-in: feb49d10e8 user: danielk1977 tags: trunk)