you are not logged in


User login

3. Choosing a path

Which way to choose, or is there a middle ground?

On one hand it is feasible to have as little data as possible for storage convenience, system performance, ease of data delivery and so on. On the other hand it must be observed that preservation or archival of data is just that, preservation and archival! Which means that there should be as little compromise as possible when preserving data.

The complexity of triggers together with lack of trigger support for several database systems in MIXED makes the inclusion of triggers unwanted at the moment. Triggers will be left out in the first version of MIXED.

Triggers: data manipulating triggers

As mentioned in "Data... and the rest." (/node/222), stored procedures are candidate for preservation in Mixed because "there is a high chance that the procedure will contain logic to modify or aggregate data". Dirk comments to leave out triggers because "they are mainly used for data manipulation (on insert, update, delete)". But triggers are also frequently used to manipulate (insert/update/delete) data...... I am not sure if these data-manipulating triggers are as frequently used by scientists in there databases as they are in business software/logic, but it's worthy to keep this in mind.


Another reason to leave out triggers: they are mainly used for datamanipulation (on insert, update, delete).