Commit graph

13 commits

Author SHA1 Message Date
Gregory Burd
1da1168747 Update aclocal to latest version, fix s_* scripts in dist. 2009-08-31 18:52:36 -04:00
Gregory Burd
faa6e61f60 GDBv6 requires a bit more config to deal with dynamic libs and breakpoints. 2008-09-10 12:53:46 -04:00
Gregory Burd
b9a7f69e0b updated copyrights to include 2008 2008-05-15 19:08:18 -04:00
Gregory Burd
598977fa88 Update license to GPL version 3. 2007-10-21 02:57:28 +01:00
Gregory Burd
67da5e7057 Change license to GPL version 3 from version 2.1 2007-10-21 01:42:57 +01:00
Gregory Burd
9ab52d89be Clean up and type fixes.
After discussions with Keith Bostic I've learned that always specifying the
full type (u_int32_t rather than int) as a general rule is not always a good
idea.  Now, as I review code I'm reverting some of those changes.  The new
rule of thumb is to use explict typing only in places that require it.
2007-10-21 01:33:16 +01:00
Gregory Burd
2529f39db1 Update call to srand48 to match prototype. 2007-10-21 01:31:56 +01:00
Gregory Burd
d55ee94646 Update flags for in-memory databases to DB 4.6 API 2007-10-21 01:29:45 +01:00
Gregory Burd
ed5cf2a8d3 Allocate SQL functions on heap, not end of struct.
Allocating the functions hash_t struct as an extension to the dbsql_db_t struct
was foolish.  Allocate it on the heap, remember to free it when we close the
database.
2007-10-21 01:24:20 +01:00
Gregory Burd
89646d2c84 added a bit more to the smoke test 2007-03-24 23:15:20 +01:00
Gregory Burd
f1037cdd9a fixed portability to mac os/x 10.4.9
There was no need to maintain random number generation state within the DBSQL
structure, that is removed and now static within the two functions requiring
random number generation.  The placement of the function hash table in the
memory allocated for DBSQL was wrong, that is now fixed.
2007-03-24 22:59:14 +01:00
Gregory Burd
4a8fcf6cd1 remove generated files 2007-03-24 05:28:22 +01:00
Gregory Burd
a04d39bc52 Initial import. 2007-03-10 20:04:07 +01:00