Dev:Certification

From wiki.railML.org
Jump to: navigation, search

Certification of railML® interfaces

As to provide for a smooth data exchange, you have to get your interfaces certified, if they employ railML® in productive use (afield).

A certification manual was developed and published in 2016. Please contact the railML® coordinator for any question or inducing a certification procedure.

General rules

  • Every application of railML® needs a certified interface before it is used productively for the first time.
  • Application have to be developed by railML® partners only. So please register at railML.org before starting implenentations.
  • As railML® schemes of 1.x versions are not supported anymore, they are excluded from certification.
  • As railML® schemes of 3.x alpha and beta versions are not yet fixed, interfaces cannot be certified for it until now.
  • If you let us certify an interface you have to bear the costs of the procedure.

Validity

  • The certification is valid only for defined versions of railML®. If you switch to a newer version of railML® or you'll extend or change your interface, your interface has to be certified again.
  • The certification is valid only for the tested version of an interface. If you extend or change your interface, your interface has to be certified again.
  • There are different requirements for import and export interfaces (but both can be certified parallely).
  • The certification is restricted to defined subschemas.
  • The certification is only valid for use in the context of the use case applied in the certification procedure, and for the defined data fields.

Method

  • For import interfaces you have to proof that you can successfully process certified sample files which we provide to you.
  • For export interfaces you have to proof that you can export syntactic and semantic railML compliant data successfully, either in the scope of one of railML's use cases, or in the complete specification of your system.