Delphi 2009 and Firebird 2.1 = Full Unicode?

Question

Has anyone started creating Unicode applications or converting existing applications to Unicode?

How do you tune Firebird for the least problems, especially the CHARSET attribute?

Are there any problems? Anything else I should know about?

I am just getting ready to get fewer surprises before jumping on a Unicode train.

Answer 1

Use UTF8 as the character set for your Firebird database. This is the only option that supports all possible characters since UNICODE_FSS is a thing of the past.

This, however, is orthogonal to your question about creating Unicode programs with Delphi. Even for the new Ansi Delphi program using UTF8 encoding, it will be your best option, since you can translate Ansi programs (for example, using the GNU gettext for Delphi http://dybdahl.dk/dxgettext/ ) into other languages, and only with the base UTF8 data you can store all the text that can be sent to the database, depending on the Windows codepage. An alternative would be to use different sets of database characters for different deployments, and then you can no longer simply exchange database files or transfer data between installations without first overwriting them.

For your development, make sure your database administration program works flawlessly with UTF8 databases. Since most of them are written in Delphi, you may have problems unless it uses Unicode compatible controls.

Answer 2

these are the "edges" of the application where you will see problems (reading / writing files, DLLs, storage, etc.).

ini files: I switched from TIniFile (only ansi) to TMemIniFile (do not forget to call UpdateFile).

as for the Firebird part of your question, I have no experience with this.

Answer 3

As far as I know, the only part of Delphi 2009 that does not support unicode is console functions like readln and writeln. The rest should be fully unicode.