After taking a look at the source code, I think I have an idea as to what is going on. When Sesame creates an application it tags pertinent parts of the design with a unique ID. The Q&A translator usually is creating a new application and so tags the elements, forms, reports, etc, ... with new ID numbers.
The Reconciler uses the ID number to determine if an element, form, etc, ... is derived from current design. So when you translated from Q&A originally, Sesame assigned the ID numbers. SDesigner retained these ID numbers to use when reconciling any application that was (or would be) built using that translated application as a starting point. But when you retranslated the Q&A databases, you essentially started a new (and as far as Sesame can tell) unrelated application, with new and different ID numbers.
So, yes, it is by design. But the intent of the design is not so much to prevent what you are doing (which we really want to encourage) but to provide some way for Sesame to tell if elements in an application design are related.
In terms of recommendation, many customers translating from Q&A will translate, merge databases, and then remove all but a few test records from the resulting Sesame application. They will then redesign as needed. Then, when it is time to switch over to Sesame, do an ASCII export out of Q&A and an ASCII import into Sesame for each of the databases in Sesame.
One of advantages of this approach is that while designing, you are working on much smaller files and your ongoing Q&A users are not disrupted.
Remember to save a copy of the .dsr and .ddt files without data records, and use those as your masters for further changes and updates.
|