in reply to Single file RDBMS w/o system install

I wouldn't recommend DBD::CSV but DBD::RAM is supposed to be pretty wild from what I've seen. Haven't done any real world with it but it seems to try to do good things with the catalog mechanism and it's fast if you have RAM to feed it.

Incidentally it's not like an Access MDB file plus Jet engine plus ODBC plus Windows is anything as simple or efficient as slurping a text file. Why not use the Berkeley system? It probably gives you plenty of power as an embedded object store without necessarily using SQL.

The answer to an information retrieval strategy has a lot to do with what you are trying to do, how much it will be accessed, whether it is going to cost you money if it goes down or loses data, etc. That said maybe you'd like to rent yourself a relatively cheap account somewhere that will host your (small) mysql or maybe postgresql database for you.

Move SIG!

  • Comment on Re: Single file RDBMS w/o system install