At some point, you may find that the CSV DB doesn't fit your needs in terms of features or performance. If that happens, I would recommend SQLite. And also check out SQlite Manager for Firefox. SQLite does a lot of what MySQL can do but without the admin hassle. In a multi-process environment, SQLite requires that an exclusive lock be acquired to update the dB for a write. So performance is not the same for writes, but for reads, this thing works fantastic! For complex operations, I found that you can dynamically vary the amount of memory that SQLite uses. That kind of fine tuning can have a big performance impact.
Update: The reason that SQLite is so cool is that it uses a simple file for its work (rough comparison to DBD CSV) So it does not require a server to be running (along with the admin hassles with that).
In reply to Re: dbi::CSV works with text, but not numbers
by Marshall
in thread dbi::CSV works with text, but not numbers
by Sandy_Bio_Perl
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |