I'm starting to leverage on the tons of things the QV Document Analyzer by Master Rob Wunderlich offers.
The Count, Size & Bytes columns in the Memory tab in particular is starting to make me think on the approaches I took with my existing projects.
Currently in one of my projects, I'm loading the following in the script: a numeric Materials field, a separate string Description field, and another Materials - Description field which is a concatenation of the 2.
Materials - Description
12345 - Sample 1
12346 - Sample 2
12347 - Sample 3
From an ease of use standpoint, loading Materials - Descriptionmeans that I wouldn't need to Concatenate in the application layer and would just need to call the field directly. But after seeing that the memory consumption was almost doubled, would it be more efficient if I just loaded the Materials and Descriptions field and concatenate instead in the application layer for a Materials - Description at the expense of probably CPU time?
It will depend on various things (available CPU + RAM, exists already any performance issue, kind of data-structure, fields comes from one or more tables, are key-fields, in which way use within the gui and some more ...) which way might be more suitable in your case.