| [railML3.4 beta4] Changed placement of sub-element elementState [message #4118] |
Tue, 25 August 2026 15:03  |
Dominik Looser
Messages: 66 Registered: March 2020
|
Member |
|
|
Dear all,
While reviewing railML 3.4 beta4 I noticed that the placement of elementState has changed compared to railML 3.3. E.g. for LevelCrossingIS, elementState used to be placed after typeDesignator, now it is placed after name.
I have not checked all the elements, but I see the same change in several other FunctionalInfrastructure subelements as well.
Is there a specific reason for this? For backward-compatibility, I think we should keep the order of subelements as long as there are no compelling reasons to change it. Changing the order of subelements unnecessarily complicates the code of any (exporting) tool that works with several versions.
I am looking forward to your feedback.
Thank you and best regards,
Dominik Looser, trafIT solutions gmbh
|
|
|
|
|
|
| Re: [railML3.4 beta4] Changed placement of sub-element elementState [message #4155 is a reply to message #4119] |
Fri, 04 September 2026 13:53   |
Thomas Nygreen
Messages: 138 Registered: March 2008
|
Senior Member |
|
|
Dear Dominik,
Mathias is correct that this is caused by generalising elementState. It is not inherited from RTM_NetEntity instead of from FunctionalInfrastructureEntity. Any extension of a type in XSD places the new elements after those inherited from the base type. There is nothing we can do about that.
@Mathias, an xsd:sequence is strictly ordered, and the order of sequences likewise. That is not something that is up to the library, it is part of the specification. Using xsd:all has other issues, the most important one being that a type containing xsd:all cannot be the base type for an extension. (It is possible in XSD 1.1, but the support for 1.1 is so lacking that it is unusable for a standard like railML.)
Best regards,
Thomas
Thomas Nygreen – Common Schema Coordinator
railML.org (Registry of Associations: VR 5750)
Altplauen 19h; 01187 Dresden; Germany www.railML.org
|
|
|
|
|
|
|
|
| Re: [railML3.4 beta4] Changed placement of sub-element elementState [message #4171 is a reply to message #4164] |
Mon, 07 September 2026 15:37  |
Thomas Nygreen
Messages: 138 Registered: March 2008
|
Senior Member |
|
|
Oh, how I wish we could use XSD 1.1. Then semantic constraints could instead be modelled in the schema using assertions and alternative typing. And improvements to how you can extend in general, improvements to xs:all (introducing xs:openContent)...
XSD 1.1 is not supported by Microsoft .NET nor by our modelling tool Sparx Enterprise Architect, among others.
Best regards,
Thomas
Thomas Nygreen – Common Schema Coordinator
railML.org (Registry of Associations: VR 5750)
Altplauen 19h; 01187 Dresden; Germany www.railML.org
|
|
|
|