>Dear Cetin Basoz
>
>Suppose I have a table with two fields, PName and EName. The contents of the field Pname is customer name with the Persian (or Arabic) characters and the contents of the field Ename is same names in English.
>The following setting should be made at the beginning of the program to properly index Persian names.
>SET COLLATE TO "ARABIC"
>
>I indexed the table based on Pname and Ename.
>I have put this file in a shared folder on windows XP.
>I run the following commands:
>
>SET OREDR TO TAG PNAME
>BROWSE
>
>SET ORDER TO TAG ENAME
>BROWSE
>
>So everything is right here.
>
>But if I connect to the shared folder with a Windows 7 client, the following result is achieved.
>
>SET OREDR TO TAG PNAME
>BROWSE ( wrong result )
>
>SET ORDER TO TAG ENAME
>BROWSE ( Correct result )
>
>Now if I put my table in a shared folder on windows 7 and reindex it, all results is correct. On windows 7 and from windows XP client.
>
>Why is this happening?

Do you have the same language packs installed on each of the systems?

What language is set for the non-Unicode programs on each of the computers?
Win7: Control Panel -> Clock, Language, and Region -> Region and Language -> Administrative
XP: Control Panel -> Date, Time, Language, and Regional Options -> Regional and Language Options -> Advanced

I've run into situations where workstations configured differently for non-Unicode programs would result in consistency problems. It definitely becomes problematic when the workstations aren't all configured the same due to certain requirements of software (e.g. typically most Chinese/Japanese/Korean software don't work properly on systems that are not configured on their respective languages, and in some cases English-language software may do weird things if run on systems configured for Chinese/Japanese/Korean. Even more confusing is that Chinese, Japanese and Korean configurations *each* have at least two or three different modes as there are often different standards for double-byte encodings)