The following limits apply.
There must not be any ASCII TAB characters in the data. This is the primary limit as the ASCII TAB character is the delimiter in rdbtables. There must not be any pipe characters `|` used or entered as data when using the operator 'nsq-ed'. All of the Compare operators and Logical constructors are reserved and should not be used as column names (they are all lower case and four characters or less). They are: 'gt', 'ge', 'lt', 'le', 'eq', 'ne', 'or', 'and', 'null', 'mat', and 'nmat'.
One way of overcoming (elegantly :-) the few limitations on the special characters that may be contained in an rdbtable is to use their ASCII representations, the same way it is done in sgml, html and other text processing systems. So the <TAB> (horizontal tab) character can be represented by the ASCII sequence '', the <N> (newline) character will be ' ', and so on.
The number of columns in an rdbtable may be limited to 32.768 by some AWK implementations (mawk is one of those). It should not be a problem though, as it is a very high number anyway.