-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Tim Mooney wrote: > I can't find any test target whatsoever with the amalgamation source.
Note that fulltest also includes more code to do things like fake IO errors and running out of memory. > I can download that would add at least some > rudimentary tests for the version compiled from the amalgamation? Since the SQLite team already extensively test their code, there are only two things that could hit your compilation of the amalgamation. The first is the compiler on the platform mis-optimizing code sufficiently to break it (extremely rare) and the other is changes in performance for your performance sensitive queries. I'd recommend writing your own performance test based on your queries. A single file that creates the database and runs various queries can be fed to the shell - you can time execution to catch differences in performance and run a diff against previous runs to catch potential optimizer problems. Roger -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (GNU/Linux) iEYEARECAAYFAkm1mjsACgkQmOOfHg372QTINwCgiDGyRLJmSxfzi1gBpc/2BbOq 0u0AoKfQjtkZI+aEyddAhaOf3mqvpfhH =8RjQ -----END PGP SIGNATURE----- _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users