I'm sure this must be covered elsewhere on the forums, but 10 minutes of searching didn't yield a result, so here goes...

Most long-term Omniscope users will be familiar with the similarities and differences of Category and Text data types. Most, I'm sure, will have experienced the odd headache when defining a Category field, and then 'over-filling' the Category with "too many" entries so that Omniscope decides it should be held as Text. Some will have experimented with various technical/advanced settings to try to control this.

At present we have 3 separate (but related) situations where this *appears* to be the case (i.e. a Category field is forced to become a Text field), but we're struggling to identify the exact cause. Ideally, we would like *absolute control* (in a Field Manager block, in DataManager, preferably) over a field's data type - and we're happy to take the consequences (e.g. a flaky IOK file!). In the absence of such control, it would be *very* helpful to understand the 'rules' that Omniscope uses to make this decision - this way we may be able to protect against it.

So, the cases we have identified are:

1. Too many separate values in the Category
We're all familiar with the warning ("you have more than 500 entries...") which Omniscope displays when a user converts a Text field into a Category. We can override this (by accepting the risk), but we have experienced cases where refreshing a model with new data causes Omniscope to reassess this risk, and convert from Category to Text.

2. Too long a text value / too many long text values
We're not exactly sure of the 'rule' here, but we have documented cases where a Category field with only 10-20 entries is converted to a Text field, and this seems to be where individual field values are long i.e. LEN([FieldName])>100 (or 200?). It may even be some form of aggregate, where more than 5 entries of more than 100 characters force the switch to Text?

3. Dynamic content from a Formula
We have cases where the content of a Category field is switchable according to a Variable setting i.e. a formula like:
IF([Variable]="X",[CatFieldOne],[CatFieldTwo])
We can define the formula field, and the two source fields, as Category data type... and both the 'source' fields (CatFieldOne and CatFieldTwo) remain as such, but changing the Variable value - and hence altering the content of the formula field - causes Omniscope to 'reasses' and flip to a text data type.

I would be interested in:
* Anyone else's experiences of Category to Text conversion - are there other cases I haven't identified?
* Visokio's summary of the 'rules'
* Any workarounds that people have discovered
...
* and a nice, simple solution from Visokio :)

Thanks!