railML data in one vs. splitted files (de: railML Daten in einer oder getrennten Dateien)
Description
Background
In railML's IL developer telco on March 14th, 2018 the question of cross references for railML data in splitted railML files was raised.
It was agreed, that some investigations and description in railML's Wiki have to be done, covering the following aspects:
- Advantages / Disadvantages
- When to use, when not?
- Rules for splitted railML files (e.g. same validity of data)
- How to build the cross reference between files when using splitted files?
- Best practice and example
Even this question is independent from railML 2.x or railML 3.x it was assigned to railML 2.x for immediate action.
Links
- Forum discussion:
- Thomas Nygreen, 10.01.2019: [https://www.railml.org/forum/index.php?t=msg&th=637&goto=2083&#msg_2083]
- Trac tickets
- Wiki documentation
Proposed solution railML 3.2
Splitting of infrastructure data
- starts at topology level: one NetElement can only belong to exactly one file
- NetElements don't necessarily have to know connected NetRelations
- NetRelations require referencing of both connected NetElements
- UUIDs are needed for external referencing of NetElements by NetRelations
Splitting of data from TT, IL, RS or CO still open for discussion.