SAP Basis Tech Supports

Symptom

Problems without Unicode
Other terms

Unicode, MDMP, Single code page
Reason and Prerequisites

Non-Unicode code pages support only a small portion of the world-wide existing character sets. Any character conversion based on these code pages therefore bears the risk of silent data loss.
Solution

New releases of SAP NetWeaver and SAP applications based on SAP
NetWeaver, released in 2007 or later,will no longer support new installationsof non-Unicode systems. See also the attached customer letter Customer_Letter_64_U.pdf from April 11th 2006 (also available at www.service.sap.com/platforms).
Existing non-Unicode systems are still supported, but due to the intrinsic limitations of non-Unicode code pages SAP recommends a conversion to Unicode, especially when upgrading system to Netweaver 7.0 (ERP 2005) and higher releases.
Problems can occur after any upgrade, even whithout applying new applications, e.g. in case.existing applications are redesigned based on new technologies.
The following problems can arise from implied limitations of non-Unicode code pages:
1. Only languages that can be covered by one single non-Unicode code page can be supported within each non-Unicode system.2. Some characters that are used in your language are not supported by the installed non-Unicode code page:a) e.g. Western European (ISO-Latin-1) Installations:
EURO SIGN and TRADE MARK SIGN are not supportedb) e.g Japanase Installations:
The characters listed in attachment Missing_Japanese.rar are not supported.
JIS X 2004 standard is not supportedc) e.g. Traditional Chinese Installations:
The characters listed in attachment Missing_trad_Chinese.rar are not supported3. Integration with Unicode based technologies
Most technologies used within an SAP application context are Unicode based. When using these technologies together with a non-Unicode SAP system uncontrolled data loss can occur. This affects:a) Components written in Java
See also note 975768 (Deprecation of Java features with non-Unicode backend)
Note: ERP 2005 contains Java components which should not be used in a Non-Unicode ERP 2005.b) MS-Office Integration
In MS-Office user can enter any Unicode character. These might get lost silently when data are passed to non-Unicode backend.c) HTTP/HTML
In Web based frontend any Unicode character can be entered. These might get lost silently when data are passed to non-Unicode backend.d) SOA
Intention of SOA is to provide standardized and reliable business processes between independent components (which may be provided by different vendors). A conversion from Unicode to non-Unicode during communication would bear the risk of silent data loss and contradict to the intendend reliability. Therefore SOA in connection with Non-Unicode backend systems is deprecated and should not be implemented in new projects. See also note 1358929.e) Technology independent
During conversion the length and the binary sort order of the data may change. This can cause unexpected side effects (especially for Asian languages).

It is a strategic target of SAP to move our customer base to Unicode. End of support for non Unicode based systems for future releases is being discussed internally.