| Re: UIC train transport id [message #778 is a reply to message #771] |
Thu, 26 April 2012 20:47   |
Joachim Rubröder railML
Messages: 0
Registered: November 2019
|
|
|
|
I discussed this question with Andreas at the meeting on 24.04.2012 and he
agreed that such a field would be fine but it's a bit too eary to decide
how to define the field in the proper way. I therefore created an issue
#147 for version 2.x
Dirk Bräuer wrote:
>
> I'm sorry, there has been again a crossing at the double-track Internet...
> (It seams that an old "Reichsbahner" like me is not familiar with too much
> double-track.)
>
> I've seen that Andreas has already answered so of the open questions.
>
> I hat a look into the source which Andreas sent. It seams to be an
> 'approach' to the problem of train numbers but only from the view of one
> Infrastructure Company and also more for freight traffic than for
> passenger traffic.
>
> Some of the problems are well-known. So for instance they write:
>
> "Uniqueness of the code"
> "It is the responsibility of the company that creates the code to ensure
> its uniqueness. Because a code
> is prefixed with the type and company codes a company only needs to ensure
> the âcore elementâ is
> unique in each timetable period."
>
> It deals a little bit with the problem of more than one Operator being
> involved in one train by defining a 'lead operator' (lead RU). Even this
> is - from my opinion - a little bit blauäugig because in most cases one
> Train Operating Company (TOC) does not 'subordinate' under another. At
> least it is my experience.
>
> But what it does not deal with it the problem of one train crossing over
> the infrastructure of more than one Infrastructure Company - and that is
> by far not an academic problem! There is no "leading infrastructure
> company" and it would also be unrealistic. So that means, there may be a
> 'train' (in general) München - Zürich which has a different "UIC train
> transport id" at DB Netz tracks than it has at SBB tracks. The same
> applies to a train Hagenow Land - Neustrelitz which crosses three
> Infrastructure Companies, one double. So this train then has three "UIC
> train transport ids"? Or four because DB Netz cannot use one double?
>
> Anyway, we should be very careful here.
>
> - It is now-days hard to imagine that a 'company' like DB Netz follows
> the rules of this document. At least it will take a 'few' years until the
> "big" Infrastructure Companies of our world recognise this "UIC train
> transport ids" and again a 'few' years until they implement it in their
> software...
>
> - If we must implement it now in RailML, it seams to be better to put it
> to the train part because I guess (I'm afraid) that each Infrastructure
> Company uses its own "UIC train transport id".
>
> - It is the philosophy of RailML to put everything which can change into
> a train part and let the train do only the grouping of train parts. We
> should not break this principle.
>
> I would prefer to wait with this "UIC train transport id" and to handle it
> in a wider topic like "compatibility to Pathfinder" where we could assign
> a Diplomarbeit or so.
> But this is only my opinion.
>
> Best regards,
> Dirk.
>
>
--
----== posted via PHP Headliner ==----
|
|
|
|
Members
Search
Help
Login
Home
