Hi,

Our company will be upgrading a software analysis tool and the thinking is, if we store the data we gather in a database format, we can export that database so end users can query against it however they want to. Firebird is highly regarded online as a deployable database.

It seems hubristic for us to install a complete client/ server architecture and have a db server service running all the time when it's only needed when gathering and processing the data ourselves. So my initial reaction is to (once again) derive a flatfile record structure and generate my own queries, then write a DBF (or even CSV) export filter. There could be ~200k records with ~10 fields/ record.

But it may be that the Windows embedded Firebird engine will do what we need without service/ resource issues - message #42825 seems to indicate so.

My questions are -
1. Does Windows embedded simply start/stop whenever a query is presented to its database (e.g., using Firebird's iSQL command line queries)?

2. If so, are there significant delay times as the engine starts/ stops? One can envision integrity checks, garbage collection, and other such background activities.

3. What is the linkage between the application query and the starting of the embedded engine? It does not appear as if there's the equivalent of a net start, for instance.

4. Since we support Linux and most flavors of Unix as well: what are the plans for an equivalent engine outside of Windows? I'm thinking here of avoiding issues with installing a daemon on those platforms (user may not have root privilege) and running all the time even though the service is just required during the time we're gathering/ processing our data...

Thanks very much!