H.Merijn Brand wrote:
In case of my use for foreign_key_info (), the set will only be big in case
all arguments are undef,
FYI, ODBC doesn't allow this:
HY009 Invalid use of null pointer.
The arguments PKTableName and FKTableName were both null pointers.
<http://msdn.microsoft.com/library/en-us/odbc/htm/odbcsqlforeignkeys.asp?frame=no>
The DBI spec doesn't state explicitly such arbitrary restrictions:
<http://search.cpan.org/~timb/DBI/DBI.pm#foreign_key_info>
Thus, my implementations also allow the selection of all foreign keys.
BTW Nice link about that Dutch database: http://monetdb.cwi.nl/
Looks quite promising. How does it's performance and functionality compare to
Progresql?
For functionality, see e.g.:
<http://monetdb.cwi.nl/Home/Features/>
<http://monetdb.cwi.nl/Development/Roadmap/>
Usually, these guys don't speak much about missing features, they implement it
;-)
What's more interesting, IMO, is the architecture of MonetDB:
<http://monetdb.cwi.nl/TechDocs/Core/monet/>
I'm always cautious with performance statements. Here is a published benchmark:
<http://monetdb.cwi.nl/Research/Benchmarks/>
But why do I tell you? Take the next tram and go figure it out ;-)
Steffen
