> Actually it would add at most 1 - 2% to our document-based memory > requirements. Most of the memory requirements come from the the > piecetable frags, fp_run, all the const char strings used to define > text properties and fl_BlockLayout classes used to define our > document. Our postings must have crossed :-). > I won't accept UTF-8 for internal storage. I was not going to put it that strongly, but that is exactly how I feel about the issue.
I have just tried to change UT_UCSChar to 32-bit, and there will be some work associated with the transition, as in places our code uses UT_uint16 in place of UT_UCSChar!. Also, on Windows, we will need to provide some 32->16 bit translation, but overall these changes should be straight forward -- I think we should add this on the list of immediate post 1.0 changes. Tomas
