wspec and the configuration process

One of the most important features of ACEDB is that its configuration files have been seperated from its source files, meaning that reconfiguration for a new organism can be altered without writing code and recompiling. Moreover the configuration files are readable and editable by individuals who have had little or no training in computer science. This allows biologists, who are the producers and consumers of the data, to take control of database design.

The configuration files are found in a subdirectory called wspec, located in the main database directory. The contents of wspec completely determine whether ACEDB "wakes up" as a worm, arabidopsis, or pine tree database. The models are found in models.wrm, one of the wspec files, and this is where the curator will spend most of his or her time.

The other files control different aspects of database behavior. Below is a partial list of the files in wspec and what they are used for. The intact files often contain additional information that you should read.

Most often the curator will modify the models file. Changes will not take effect until one of two things happens. First, if the database is reinitialized, the new database will reread the contents of wspec and use the revised models. Second, the curator can choose to read in the revised models while the database is running, even if it already contains data. The option is available under the main window's pop-up menu. Selecting it brings up a serious warning (which may be preceded by an offer to save the current session if write access has already been taken)

Changing the models in a running database is often convenient, but be aware that modifying structures already occupied by data may have strange effects and is not recommended. On the other hand, modifying structures not yet used, or adding entirely new structures, is generally safe. You did back up, didn't you? And you know all about session control in case you didn't.

Contents of wspec