search

DATEX-II-EU / DatexII

Main repository for issues and bugs for the DATEXII standard
0 stars 0 forks source link

Support OpenLR binary (Bugzilla Bug 455) #455

Closed datexii closed 7 months ago

datexii commented 1 year ago

This issue was created automatically with bugzilla2github.py

Bugzilla Bug 455

Date: 2023-08-10T15:12:39+02:00 From: @iancornwellmottmac To: Jean-Philippe Mechine <Jean-Philippe.Mechin@cerema.fr>

Last updated: 2024-02-07T14:14:20+01:00

datexii commented 1 year ago

Comment 1849

Date: 2023-08-10 15:12:39 +0200 From: @iancornwellmottmac

It has been requested that we directly support OpenLR binary. It has been suggested that all known uses of OpenLR with DATEX II use binary, via an extension, rather than the existing fully decomposed class model in the LocationReferencing namespace. An extension has already been created and registered by the Data Task Force. Two TB members have suggested that this should be adopted as an approved extension in the next revision (3.5). (We may even remove the fully decomposed class model in v4 but that would be another issue.)

datexii commented 1 year ago

Comment 1850

Date: 2023-08-10 15:13:51 +0200 From: @iancornwellmottmac

Ian as overall UML model editor to review the extension to check he can support its approval. It would be useful for France as Part 2 editor to also review it.

datexii commented 1 year ago

Comment 1851

Date: 2023-08-14 17:57:09 +0200 From: @iancornwellmottmac

When I see the model, that reminds me I have already reviewed it. Need to find that...

datexii commented 1 year ago

Comment 1852

Date: 2023-08-14 17:59:22 +0200 From: @iancornwellmottmac

See the review thread in Teams 4.1.4 from 27/3/23 to 23/6/23.

datexii commented 1 year ago

Comment 1864

Date: 2023-09-26 12:18:21 +0200 From: @iancornwellmottmac

This was discussed at the Sep 2023 TB meeting. The discussion on Teams was endorsed. Although discontent with the idea of an opaque binary-only location in DATEX II, the prevailing view was that it is a requirement we must support.

There was some doubt over whether OpenLR specifies an encoding that is different to base 64, but since OpenLR documentation shows base 64, we assume that is OK, so it would be consistent for the DATEX II model to choose our existing base 64 binary type because that is already supported.

A risk on licensing was observed, but then said to be probably no problem. Checking, we see that the OpenLR specification is provided under an Apache licence. If DATEX II was to include an OpenLR binary attribute, merely linking to the OpenLR specification by name, not distributing any of the OpenLR specification, then IC doesn't see any problem.

datexii commented 7 months ago

Comment 1898

Date: 2024-02-07 14:14:20 +0100 From: @iancornwellmottmac

A model has been committed with an extension added, for inclusion in the forthcoming release 3.5.