Just a note

[Ionel Simionescu]

|
| I will store HUGE amounts of data, so I think Gadfly would be "too small"
| as DBMS, since it stores all it`s data in memory ( as I recall ) .

Just a note.

You can still store huge amounts of data with Gadfly, if you delegate the
storage itself to disk files and keep in Gafly just the filenames.

This scheme, however, would impose on the data access. Namely, you will not
be able to easily apply operators on the full database, but only on the
items you can keep in memory.
And you will need to manage what's kept in memory all by yourself.

However, if you do not need access over the full data domain at the same
time, then you can load data in Gadfly as required and still have a smooth
workflow.

Because this is the usage pattern most of the times, you may still give
Gadfly a thought. Its advantage is better integration with your code
(because Gadfly will do many type conversion for you autmatically) and the
opportunity to hack in Gadfly your specialized operators.

I hope this helps,
ionel