This chapter will explain the usage of refdba for the administration of refdb. Please refer to the reference chapter about refdba to learn about command-line switches and interactive commands of this application.
Create new databases with the refdba command createdb foo. The new database will be empty initially. If you start adding datasets, the first dataset will have the ID 1.
In order to access the new database, users need to have their access rights set accordingly.
There are some restrictions for the name of a database. MySQL stores databases as folders on the filesystem where the database server runs. A database name can therefore be any valid folder name on the particular filesystem that mysqld runs on (this may be a different computer than the one you sit in front of). This means that e.g. on a Linux system using ext2fs the database name is case-sensitive, whereas on a Windows computer using NTFS the database name is case-insensitive (the 32-bit Windows filesystems are case-preserving, not case-sensitive).
A second restriction is imposed by the way bibliographies are created for SGML documents. If you use the feature to pull references from different databases within the same document, the database name has to be encoded in a SGML attribute. These attributes are case-insensitive in SGML. This essentially means that even if the database server runs on a case-sensitive filesystem, refdb has to treat database names as case-insensitive in that particular case.
In conclusion, you should not treat database names as case-sensitive. Database names should always be completely lowercase unless you have very good reasons not to do so.
Tip: Your SQL database server may manage more databases than your refdb databases. In this case it may be prudent to use a common prefix for all refdb reference databases that you create. This way you can list only the refdb reference databases by providing a simple regular expression to the listdb command.