leupold wrote to give you a bit mire insight: UDT stores its data in tables at the sql database, but in a specific (serialized) structure, that allows to store different table layout per module in the same database tables. Therefore UDT provides functions to generate the table, add, edit, delete and retrieve the data in it. Please take it as a "separate microcosmos" inside the DNN world, that has only limited options of data level interoperability with other modules (like reports) due to its design.

Interesting information there Leupold! What are the pros and cons in short to this in-house solution. Is this in any way a more efficient than providing a standardized SQL-based solution that anybody could build on and develop? Just curious! Regards, OJ, Stavanger.

ojnord wrote  leupold wrote   to give you a bit mire insight: UDT stores its data in tables at the sql database, but in a specific (serialized) structure, that allows to store different table layout per module in the same database tables. Therefore UDT provides functions to generate the table, add, edit, delete and retrieve the data in it. Please take it as a "separate microcosmos" inside the DNN world, that has only limited options of data level interoperability with other modules (like reports) due to its design.   Interesting information there Leupold! What are the pros and cons in short to this in-house solution. Is this in any way a more efficient than providing a standardized SQL-based solution that anybody could build on and develop? Just curious! Regards, OJ, Stavanger.

OJ,

you obviously missed my UDT session at OpenForce '07 Europe conference last week...

major advantage of the serialized storage is the fact, that there is no need to create a custom table whenever a module is created - which may be limited by the sysadmin to module installation only. Serialized storage is less efficient for densely populated tables, because there are two extra indexes FK fields per value, but for records with an number of columns of which only few values are filled, serialized storage is more efficient. However there is always a (small) overhead during loading and saving data, as well as the difficulties to access the data directly using standard approaches - as you already noticed.

HTH