Re: Accessing data type 'G' in RPG ???

On 11/27/2009 11:27 AM, James H. H. Lampert wrote:

I did *NOT* say that DBCS *IS* 16-bit Unicode. I said that this
particular *instance* of DBCS fields is *probably* 16-bit Unicode, which
*IS* one of the possibilities (and *at least two* of the valid code
pages) for type-G DBCS fields, and probably the MOST COMMON one for new
work.

As far as I know ... 16bit Unicode is *NOT* a DBCS character set.

Yes, it does take two bytes to represent a single character in 16bit Unicode ... but DBCS has specific meaning on the i.

And as to the so-called controversy, BOTH usages of "DBCS" are correct,
as BOTH usages of the term refer to characters being encoded at 16 bits
each. Limiting it to one usage or the other is NEVER correct.

Context is key here ... "DBCS" on the i is (again, as far as I know) generally referring to a way of encoding Asian characters in an EBCDIC environment.

Although I haven't done an exhaustive investigation, I think you will find that the IBM i manuals reference the term DBCS in this way.

FWIW: On the i, DBCS is supported in the 5250 data stream but Unicode is not. At least as of V5R4 (http://publib.boulder.ibm.com/infocenter/iseries/v5r4/topic/nls/rbagsucsdspandpnlgrpcon.htm).

david