Below, there is a list of good practices, that lis suggesions, thanks to which the work with processes can turn out easier and processes more effective.

Readibility

• Using flowchart as a main avctivity
• It is neccessary to group activities with the use of the activity Sequence/Flowchart activities with the same scope (e.g. adding a SI item or setting a customer/vendor) and name and localize them in a proper way.

Multiculturalism

• Names of variables and input/output arguments should have English names or be abbreviations of English names
• Default names shoud be changed and <<translations>> should be added to them.
• Descriptions of parameters should contain information about the use of a given parameter.
• A process should have name and added desription.
• The names of exported .wdf files should be in English.

Handling of errors

• It is necessary to complete fields left empty, e.g. e-mail address, to prevent eocurring of exception NullReferenceException
• It is necessary to use TryCatch activity, if it’s possible, in order to catch busness exceptions. Information about errors can be sent to the task inbox

Configuration possibilities

• It is necessary to perform parametrization of process operation, e.g. by adding the possibility of selecting a center with permissions of which the process is to be performed.
• Oermaters of SQL queries should be associated with variables or arguments to facilitate modifciation of process working.

Performance

• When preparing processes, it is necessary to remember that they must work on production databases. A test database used for verification of processes should correspond to the production database in termos of the quantity of data, so that already in the production phase it is possible to identify performance problems.
• A user should pay attention to the method of determining scopes for variables. In case a given variable is used e.g. only inside a given sequence, its scope should not be set.
• To improve memory performance, it is possible to set unused variables to null in the further part of a process. When a process is waiting for decisions, the data regarding it is saved in the datatabse. The less data is in the database, the less cache memory and memory in the database is occupied.