Comparison of SCE and R/3 Variant Configuration 
The master data that the SCE uses to configure products is maintained in R/3, so both the SCE and R/3 use:
However, R/3 is written in ABAP and the SCE is written in Java, and there are a number of structural and functional differences.
General Differences Between the SCE and R/3
SCE |
R/3 |
Master data in the SCE knowledge base (or CRM consolidated database) is read only. |
Master data is maintained in the R/3 database. |
The SCE uses only the master data tables and fields that are relevant to sales (for example, the SCE does not use table AUSP). |
R/3 master data tables contain data that is relevant to all application areas. |
The SCE only configures configurable materials, in the form of material BOMs. |
R/3 configures a range of objects, including different BOM categories and task list types. |
The configuration profile of each material you want to configure with parts and subparts must have the setting for nested KMAT or SET. These settings are different for different releases of R/3, so for more information, see the R/3 Library Variant Configuration (LO-VC). |
All available settings in the configuration profile are supported, including order BOMs and single-level "sales order". |
The configuration profile of configurable materials must have the setting Multi-level or None for the BOM explosion. |
The configuration profile of configurable materials can have any of the settings Single-level, Multi-level or None for the BOM explosion. |
The SCE is not integrated with non-sales applications, so there is no availability check (ATP), for example. |
R/3 Variant Configuration is fully integrated with all R/3 applications, so availability check and other related functions are supported. |
Type matching is not supported for configured assemblies, because material variants are not part of the knowledge base. |
You can search the database for material variants that match configured (or partly configured) assemblies. |
All master data for configuration is stored in the SCE knowledge base (or CRM consolidated database). This speeds up configuration, because there are fewer database accesses and source code for dependencies is stored as a binary large object (BLOB). |
Master data for configuration is stored in separate tables, according to object type (for example, MARA for material master records). |
The knowledge base profiles for a product define which materials in a structure you can use to start configuration. |
You can start the configuration process from any material in the structure, provided that its configuration profile allows you to. |
Only sales-relevant BOM items can be configured. |
All configurable materials can be configured. |
Bills of Material
SCE |
R/3 |
Each material has at most one BOM. |
A material can have several alternative BOMs for different lot sizes. |
The following objects can be BOM items:
|
The following objects can be BOM items:
|
A BOM item contains only:
Other BOM item data is not downloaded. |
A BOM item contains data that is relevant to all application areas. |
Each combination of BOM item number and component number (material or class) can occur only once in a BOM. |
You can use the same combination of BOM item number and component number more than once in a BOM, because R/3 uses an internal number to identify items internally. |
In SCE* mode only, you can create several instances of one BOM item. |
You can create new BOM items and delete items by processing an order BOM. However, you cannot define an item as optional in a super BOM (items without dependencies are always selected, and items with dependencies are selected according to the dependencies), |
Only BOMs of configurable materials are downloaded. |
All BOMs are in the system. |
If a class node has multiple classified materials matching the assigned values, the SCE does NOT automatically specialize the node to any of them – it leaves the class node as it is. |
If a class node has multiple classified materials matching the assigned values, R/3 picks the first matching classified material. |
SCE does not support unit conversion. All input for component quantity and characteristic values characteristic values must be entered in the base unit. |
You can define conversion factors for units of measure that do not normally match. |
You cannot use the classification of materials to select components. |
You can use the classification of materials to select components. |
Classification System Objects
SCE |
R/3 |
All characteristics must be assigned to classes. Any characteristics that are not assigned to a class in the knowledge base when you download the knowledge base are automatically assigned to all classes. |
Reference characteristics need not be assigned to a class. |
Class networks must allow a precedence list to be calculated. The precedence list is used to define the inheritance logic for default values of characteristics. In exceptional cases, it may not be possible to calculate the class precedence list, so the system issues an error message when you generate the runtime version of the knowledge base. Example of an inconsistent class network: |
Class networks are totally flexible. The system does not recognize the example as inconsistent. |
You cannot use characteristics with a user-defined data type (data type UDEF). |
You can use characteristics with a user-defined data type (data type UDEF). |
You cannot use value hierarchies in characteristics. |
You can use value hierarchies in characteristics. |
You cannot maintain intervals as values for numeric characteristics. |
You can maintain intervals as values for numeric characteristics. |
Characteristic facets (attributes and allowed values) can be changed dynamically using the extended syntax available in SCE* mode. |
Characteristic facets can be changed dynamically by using an additional characteristic. |
You can only use reference characteristics for STPO-MENGE and for fields in structures SDCOM, VBAK, and VBAP. The content of fields in VBAK (sales order header data) and VBAP (sales order item data) must be downloaded by the sales application, because they are not included in the SCE knowledge base. For more information, see Reference Characteristics. |
You can use reference characteristics for a range of fields in STPO and SDCOM. |
You can define no more than one reference characteristic for each table field, and each reference characteristic can only refer to one table field. |
You can define more than one reference characteristic for each table field, and a reference characteristic can refer to more than one table field. |
Graphics for characteristics and/or values can be saved in the file structure on your PC or laptop, and links to the characteristics and/or values are maintained in a properties file. However, as of R/3 Release 4.6C, your can use the document management system, as for R/3, for the SCE standalone and Mobile Sales. |
You can use the R/3 document management system to link the graphics that you want to display to characteristics and/or values in your knowledge base. |
The attributes Not ready for input and No display are always those maintained in the characteristic itself. |
You can overwrite the characteristic attributes Not ready for input and No display for use in a specific class. |
The characteristic value descriptions are always those maintained in the characteristic itself. |
You can overwrite characteristic value descriptions for use in a specific class. |
Organizational areas defined for the user interface are ignored. |
Organizational areas can be used to influence which characteristics appear on the user interface. |
If using the Web UI, you cannot specify a class node manually. This reduces conflicts between values, because all options are selected according to the characteristic values assigned and the object dependencies that they trigger. (If using the Swing UI, you can still specify a class node manually.) |
You can always specify a class node manually. |
Direct assignments of variant conditions to characteristic values only work if you assign a reference characteristic for table field SDCOM-VKOND to the variant class. |
You can assign a variant condition directly to a characteristic value, instead of using object dependencies. |
Object Dependencies
SCE |
R/3 |
All procedures (and actions) must be assigned to a configuration profile or a BOM item. |
Procedures (and actions) can be assigned to characteristics and characteristic values, as well as configuration profiles and BOM items. |
You cannot use actions unless you select the option to automatically convert them to procedures. If you select this option, all actions are converted to procedures with position number 0 and processed as procedures in the SCE (once each time you confirm during configuration). |
Actions are declarative (not procedural) and are processed every time you confirm, as many times as it takes, until no more values can be inferred from the information available. |
All constraint nets are loaded at the beginning of a configuration session. |
Constraint nets that are assigned to a configuration profile are only loaded if the corresponding material is selected in the configuration. |
You cannot use $SUM_PARTS or $COUNT_PARTS in procedures. |
You can use $SUM_PARTS and $COUNT_PARTS in procedures. |
Variant functions and user exits must be written in Java, or written in C and called from a Java application. (You can use the Java Native Interface (JNI) to implement the second option.) For more information, see User Functions for Variants. |
Variant functions and user exits are usually written in ABAP (but can also be written in C). |
The names of variant tables and their columns must consist entirely of alphanumeric characters and underscores, and must not start with SCE. |
The names of variant tables and their columns can also contain hyphens, for example, and can start with any character string. |
Each line of a variant table must contain one value for each characteristic. |
A line of a variant table can contain multiple values or interval values for a characteristic, unless the variant table is connected to a database table. |
Engineering Change Management
If you use R/3 Engineering Change Management to control changes to the objects in your configuration model, the following differences apply:
SCE |
R/3 |
Use change numbers with a valid-from date. |
You can use change numbers with a valid-from date or change numbers with effectivity. |
The date you enter for the runtime version (together with the plant and BOM application you enter) determines which version of each object is valid. |
The date you enter when you configure your product determines which version of each object is valid. |
A runtime version of a knowledge base contains one version of each object. |
The database contains several versions of any object that has been changed with engineering change management. |