[comp.databases] Informix inconsistency in 8-bit characters

jw@pan.UUCP (Jamie Watson) (09/17/88)

We have are loading information from IBM mainframe systems into Informix
databases.  There are accented characters in the data we recieve, which
are represeted as 8-bit characters according to the IBM Graphics tables;
for example, hex 81 is u-umlaut, 84 is a-umlaut and so on.  We weren't
sure if Informix would process 8-bit characters correctly, so we just
decided to give it a try.

We found that we were able to load the data with no problem, and it is
stored and processed properly within the database.  It comes out right
in reports, and on the screen when queried.  However, when we tried to
actually type in such 8-bit characters on a terminal, we found that the
Informix SQL Perform program accepts everything very happily, but the
4gl forms refuse to accept 8-bit input.  Anytime an 8-bit character is
typed in a 4gl program, the terminal beeps and the input is refused.

Why the inconsistency?  Which is correct?  Will 4gl be upgraded in the
future to accept 8-bit input, as SQL currently does?  Since it appears
that everything else in the database system, and even in 4gl, processes
8-bit characters correctly, there are no obvious reasons for 4gl to
reject this input.

jw