Pragmatic generic coding-rules. Such structures are known from a long time as ``frames'' a classical
[
ref] data-structure.There are always several ways to represent an information
as a logical-structure. Here are a few key ideas to make such choices:
- Maximizing atomicity. I.e. structure the data with a maximal decomposition (e.g. atomic values must
only contain ``words'' else there is still a ``structure'' and is thus to be decomposed itself in terms of elements).
- Maximizing factorization. I.e. prohibit data redundancy, but use references to index a data fragment from another
part of the data. This saves place and time, but also avoid data inconsistency.
- Maximizing flat representation. I.e. avoid complex tree structures, when the data can be represented as uniform list of
data, i.e. table with simple records, such as a field-set.
- Maximizing generic description. I.e. abstract representation, without any reference to file format or
operating-system syntax: independent of how the data is going to be used.
- Maximizing parametrization of functionality. I.e. specify, as much as possible, the properties (i.e. characteristics / parameters / options) of a
software module or a function as a static set of data (instead of ``putting-it-in-the-code'').
