When reports and screen panels are designed in multilingual applications, "width" details for the column headings, labels, descriptions, etc are established from the widest available multilingual definition.
In bi-directional languages the width is calculated from the right, in all other languages the width is calculated from the left (including shift-out/shift-in characters for DBCS languages).
The need to make space for the "widest" language on screen panels and reports can have a bearing on how multilingual details such as column headings are recorded into the dictionary.
Consider an alphanumeric field of length 3 that has a multilingual column heading that is, for example, 20 characters wide in German.
Ignoring the fact that using a 20 character wide column heading for a 3 character wide field is fairly unlikely and violates the guidelines in the LANSA Application Design Guide, consider if it was placed onto a report, the space used would be like this (where GG..GG is the German column heading and XXX the printable field value):
GGGGGGGGGGGGGGGGGGGG
XXX
Now if the system also allowed English and Hebrew, and both of these column headings only used 5 characters, the run time change to either of these languages would produce a result like this:
English .... EEEEE
XXX
Hebrew ..... HHHHH
XXX
Both of these results would look really odd on the report.
The solution is to remember that column headings should generally be centered within the space used by the "widest" language entry.
The use of a scaling or test card language can aid in doing this.
If these column headings were all centered in the 20 characters, the results would be like this (which are far more acceptable):
German ..... GGGGGGGGGGGGGGGGGGGG
XXX
English .... EEEEE
XXX
Hebrew ..... HHHHH
XXX
