Preparing a datamodel for internationalisation / localisation

I have a database that should ultimately be included in a reporting engine. This will be a point-and-click setup whereby the user selects columns to include in the report and the users can run and save the report-script themselves. Users will have affinity with the data, but we want to design the system so that it becomes clear in the user interface what the user is selecting. To that end we want to use the column_comment feature in MySQL.

We add column comments as follows:

ALTER TABLE productprices CHANGE price price FLOAT COMMENT 'price of a product at any given moment' 

Given the flat nature of a database, this does not allow for expressions. For example, we would want to prepare the database for translations (localisation), that would in procedural languages be done using gettext for example. What would be the best way to maintain translations of column comments?

On an early version of a utility CIS in oracle 6, we had to have separate schemas - One for Spanish, another for English. There too many side effects otherwise.

We published two data dictionaries one for Spanish, one for English, to a special set of tables we used for prompting. We did not use comments. We used two sets of synonyms that had descriptive table and field names. Base sql code was unaffected by this.

I am sure this is not what you wanted to hear.

Thank you for your response. We thought it was going to go that way. It looks like the work load for both development and changes is not going to be affected by this all that much, even if expressions were allowed.