[ALL] let's think about Query Language for 2.0

[Keith DeLong]

> Not exactly.
> 
> You here speak about mixture of 2 things
> 
> 1) VIEW in SQL. When we can create virtual table that is based on some SQL
> query in fact. You myTable do exactly this.
> 
> 2) API with functions Select(), Where(), OrderBy(), exists in e.g. ADO of
> Microsoft. Common for VisualBASIC developers and MFC developers.

Nice to know my ideas have merit even if they are not original  ;-)
 
> This functions in fact in background build SQL string, and later send it to
> DBMS server. Because Servers do not have other way as SQL strings.
> So this API just play role of wrapper.

Consider The API wrapper idea a feature request...
 
> 3) YES, we also can have this.
>  But I have talk about totally new feature for V4RB,
>   -- ability do search and sort on table without SQL at all!!!

I generally understood this. I did get carried away moving to an example as
a wrapper for what would have to execute as a SQL select.

> IF somebody want make size of his app smaller,
>   and faster and he have simple enough queries.
>   Or may be he like work with navigational model,
>   or his task better fit navigational model,
>   then such developer can use in V4RB only BaseObjects and this new search
>   features. NO CURSORS at all!   No SQL strings at all!

Okay, I understand more clearly -- Though not executing as SQL I would still
propose that SQL naming be used for method names where possible for
clarity's sake.
     
> As result of search he get for example RB ARRAY with found RecIDs.
> So now he can do BaseObject.RecID = arr[i] to iterate found records.

I agree these are common searches in applications. A faster and simpler
method would be a benefit.
... 
 
> Well, may be REALbasic developers do not need all this to avoid confusing.

I don't know if your develop efforts and the API complexity are worth the
simplicity/speed gained on the back side. That¹s a valid question.

Keith