Developer Forum

Hello,

I have recently written too much procedural code to do
inner & outer joins with dotnet dataset tables.

Somebody must have written a library, that allows this.
I tried google & sourceforge, but dozens of promising
leads always end up being something else.

Any hints?

FYI: my data sources are external and diverse, I don't want
to dump the data to SQL server just to do this.

Thank you, Kari

That may be the best choice. There have been some attempts in this area, but they have
generally been failures because of efficiency and resource problems. After all, it's
what an RDBMS is intended for.

--
Lee Fe *** an, FFE Software, Inc. ( http://www.yqcomputer.com/ )
==============================================================
* The Ultimate DBMS is here!
* FirstSQL/J Object/Relational DBMS ( http://www.yqcomputer.com/ )

Exactly. There's a saying in the database world that you take the
processing to where the data is, not the other way around. A RDBMS is
designed to handle all kinds of processing of data. With all due respects
Kari, a problem exists today among database-weary software developers that
refuse to embrase the power of RDBMS since they believe databases are just
dumping grounds like they were in the old days where data was stored in a
haphazard format and then tons of code was written to retrieve and process
that data. Software developers need to flush out their heads and begin to
embrase database technology for the powerful programming environment they
provide. Database developers write plenty of code too. It's not just
software developers the write code. The difference is that database
developers only write code that is necessary for the data they are working
on, unlike software developers who feel they can get the data locally than
write complicated code instead of implementing SQL programs to simplify it.





but they have
all, it's

I completely agree, that's why I'm looking for a tool that allows me to do
declarative programming instead of those stupid loops.

What I'd hope is that database vendors would understand, that relational
model
has nothing to do with persistence, some users like me just have temporary
data
and want simply to process it: in memory, no locking, no transactions.
Microsoft
had IMDB in w2k beta, but it got dropped for some reason or another.

<snip>



After

Performance will not be a problem, inner join of two max 10k records
in-memory
is a snap even with unoptimized algorithms. On the other hand, dumping the
tables to
some sql system will be significantly slower.

Thanks for suggestions anyway, Kari

I'll take the chance to do a little advertising. Our ORDBMS (see my sig) provides an
in-memory mode, however disk storage is still involved for persistence. We also provide
something we call 'flash' mode which is in-memory mode without persistence, no data on
disk at all. All data is completely transient. It has transactions though in memory, so
it is very fast. It's implemented in Java, but we have an ODBC driver.

--
Lee Fe *** an, FFE Software, Inc. ( http://www.yqcomputer.com/ )
==============================================================
* The Ultimate DBMS is here!
* FirstSQL/J Object/Relational DBMS ( http://www.yqcomputer.com/ )