Here is the catch with EarthCape: you cannot modify the underlining database structure. Well, you can but that would an undocumented hack which I will document later.

Given that, there is still a way to store data that does not fit into the predefined schema in EarthCape database: Descriptive module comes into play allowing adding Tags (with TagValues) to any object in the database:

Note that apart from "standard" TagValue types there is one that handles FileAttachments. It is still part of the experiment of making annotation system reasonably (but, potentially, overly) diverse and flexible. I will talk more about file attachments and links in later post.

What the approach above allows is to add more "data fields" to your base data (e.g. taxa, specimens) without restructuring the database. Another benefit is the potential to explicitly treat Tag assignments with audit data (who recorded, when) and to record several values for the same Tag (e.g. repeated measurements). This format is also better suited for EC Analysis module.

One problem UI related problem came up right away during field testing - it is a bit cumbersome to enter tag values, especially during rapid and high volume measurement lab experiment. User essentially has to choose from the drop down list which type of value he/she wants to create, choose the tag, fill the form, close it and repeat the process.

Currently I have introduced a shortcut that pops up a form that has list of tags that are assigned to current project. This form can be filled for several values of different types and when it is saved, all values get attached to all selected objects (e.g. specimen records).

Next step would be to bring TagValues  editing close to real table editing experience (as if those Tags are actually columns in the base data table) using pivot setup in Analysis module. At this point it is just an idea and experimenting will show if this solution is usable.

If YOU have any ideas on how to improve user experience when working with for instance character data and would like to share it, please comment on this blog or drop me a line. It would be most interesting to know what kind of setup different labs are using and what are the major bottlenecks in the area of descriptive data management.