We have a commercial of the shelf application accessed via Citrix and
fat clients.

Citrix Users have reported that when the GBP currency symbol, "£", is
shown on the screen it appears as a lower case 'u' with an 'accent'
above. On Fat Clients the currency symbol is shown correctly.

On the Citrix servers the Regional Options are Standards and Formats -
English (United Kingdom), Location - United States; symbol.tbl contains
II_CHARSETOP WIN1252

On the Fat Clients the Regional Options are Standards and Formats -
English (United Kingdom), Location - United Kingdom; symbol.tbl contains
II_CHARSETOP IBMPC850

Is this difference found in symbol.tbl due to the "Windows Regional
Options" in effect when OpenRoad was installed?


--
SloughBC

Where is the GBP symbol shown coming from?
Is it
- the user who types it in
- coming from the database
- automatically added (for "money" fields") ?


--
Bodo

The currency symbol is entered into a text field, defined as varchar.

when entered via Citrix, it appears as a £; when viewed via Citrix, it
appears as a £

when entered via Citrix and viewed via thick client it appears as a
thick vertical bar

when entered via fat client, it apears as a £; when viewed via fat
cleint, it appears as a £

when entered via fat client and viewed via Citrix it appears as a lower
case 'u' with accent.


--
SloughBC

I assume that the currency symbol is stored to database (or how else can
one client see the text the other client entered?).

Your thick client's configuration setting for II_CHARSET (II_CHARSETOP)
is wrong.
This is a Windows client, but the configuration says IBMPC850, which is
a MS-DOS codepage.

When storing the Windows character "£" (code 0xA3) into the DB from
your fat client,
it gets transferred as ASCII (IBMPC850) character with the the same
character code - which is "ú".
This character is then stored in the database (converted to whatever
charset is specified for the DBMS).

When retrieving with the same (wrong) setting then it's converted back
to 0xA3,
but when retrieving from a correctly set client (II_CHARSET=WIN1252),
then the Windows character "ú" is displayed.

When storing 0xA3 character from your correctly set Citrix client, and
then retrieved back
from your wrongly set fat client the conversion will be done to 0x9C,
which is the IBMPC850 code point for "£". The 0x9C character is then
displayed using the Windows codepage in the entryfield
This is a "œ" character, but not all fonts can display it though, so
the vertical bar is displayed for the "unprintable" character.
E.g. if you change your font from "OpenROAD System" to "OpenROAD
Helvetica" you'll see the "œ" character.


--
Bodo