Home » railML newsgroups » railml.infrastructure » <track>@mainDir
<track>@mainDir [message #1710] Thu, 22 February 2018 13:23 Go to next message
christian.rahmig is currently offline  christian.rahmig
Messages: 103
Registered: January 2016
Senior Member
Dear all,

following discussions with some railML partners, I want to bring the
topic to the forum and ask for a general feedback in order to adapt
implementation in railML 3.x:

In railML 2.x we have an attribute <track>@mainDir that describes the
preferred driving direction on that track. Possible values are "up",
"down", "unknown" and "none" and they always refer to the track
definition direction (trackBegin to trackEnd).

The question to be answered:
Is the information relevant for any use case? Who uses @mainDir and for
what?

If there is no specific interest in keeping @mainDir, we may also think
of declaring the attribute DEPRECATED with upcoming version 2.4.

Any feedback is highly appreciated...

Best regards
Christian

--
Christian Rahmig - Infrastructure scheme coordinator
railML.org (Registry of Associations: VR 5750)
Phone Coordinator: +49 173 2714509; railML.org: +49 351 47582911
Altplauen 19h; 01187 Dresden; Germany www.railml.org
Re: <track>@mainDir [message #1711 is a reply to message #1710] Fri, 23 February 2018 15:33 Go to previous messageGo to next message
Dirk Bräuer is currently offline  Dirk Bräuer
Messages: 228
Registered: August 2008
Senior Member
Dear Christian,

we do not use <track>.@mainDir. We would not miss it.

We use <sectionTT>.@trackInfo with:
=1 for "regular track" (dt: Regelgleis)
=2 for "track of the opposite direction" (dt: Gegengleis).

This usage of <sectionTT>.@trackInfo is redundant to <sectionTT>.<trackRef>.@ref: If a train runs on the Gegengleis (<sectionTT>.@trackInfo='2'), it has also <sectionTT>.<trackRef>.@ref set to the proper track of the opposite direction.

This redundancy is a "short-cut" for cases if there is no full <infrastructure> at the railML file or if a reading software does not like the effort of parsing the <infrastructure>.

In spite of there is a (small) usage of "regular direction" and "wrong direction", it is rather a short-cut around the infrastructure details but not necessarily a demand for infrastructure. It could explain why somebody could want to define a regular direction of tracks. But it is not so obvious or demanding that we (iRFP) need it. So again: no opposition from us against deprecating of <track>.@mainDir.

Dirk.
Re: <track>@mainDir [message #1715 is a reply to message #1710] Mon, 12 March 2018 12:51 Go to previous messageGo to next message
Tobias Bregulla is currently offline  Tobias Bregulla
Messages: 7
Registered: June 2017
Junior Member
Hello Christian,

thank you for that speech. We do not use this modeling in our programmes
and data exports (any longer). Such a predominant direction of track
usage is no longer evident in many modern mainline railways. To the best
of our knowledge, Austrian and Swiss railways have already abolished
this definition in their regulatory frameworks and thus also the
associated signals.

On our part, we have no objections to a statement as DEPRECATED as of
railML 2.4.

Best regards,

Tobias Bregulla
Bahnkonzept Dresden/Germany
***************************
Hallo Christian,

danke für diese Wortmeldung. In unseren Programmen und Datenexporten
verwenden wir diese Modellierung nicht (mehr). Die überwiegende
Benutzungsrichtung ist ohnehin bei vielen modernen Vollbahnen nicht mehr
festzustellen. So haben unseres Wissens die Österreichischen und
Schweizer Bahnen diese Festlegung in ihren Regelwerken bereits
abgeschafft und damit auch die zusammenhängenden Signale.

Unsererseits bestehen keine Einwände gegen einer Erklärung als VERALTET
ab railML 2.4.

Beste Grüße,

Tobias Bregulla
Bahnkonzept Dresden/Germany


Am 22.02.2018 um 13:23 schrieb Christian Rahmig:
> In railML 2.x we have an attribute <track>@mainDir that describes the
> preferred driving direction on that track. Possible values are "up",
> "down", "unknown" and "none" and they always refer to the track
> definition direction (trackBegin to trackEnd).
> ...
> If there is no specific interest in keeping @mainDir, we may also think
> of declaring the attribute DEPRECATED with upcoming version 2.4.
Re: <track>@mainDir [message #1721 is a reply to message #1710] Mon, 12 March 2018 17:19 Go to previous messageGo to next message
christian.rahmig is currently offline  christian.rahmig
Messages: 103
Registered: January 2016
Senior Member
Dear all,

I created a Trac ticket for this issue ideally to be solved with railML
2.4 [1].

Am 22.02.2018 um 13:23 schrieb Christian Rahmig:
> In railML 2.x we have an attribute <track>@mainDir that describes the
> preferred driving direction on that track. Possible values are "up",
> "down", "unknown" and "none" and they always refer to the track
> definition direction (trackBegin to trackEnd).
>
> The question to be answered:
> Is the information relevant for any use case? Who uses @mainDir and for
> what?

[1] https://trac.railml.org/ticket/324

Best regards
Christian

--
Christian Rahmig - Infrastructure scheme coordinator
railML.org (Registry of Associations: VR 5750)
Phone Coordinator: +49 173 2714509; railML.org: +49 351 47582911
Altplauen 19h; 01187 Dresden; Germany www.railml.org
Re: <track>@mainDir [message #1728 is a reply to message #1721] Fri, 16 March 2018 08:50 Go to previous messageGo to next message
christian.rahmig is currently offline  christian.rahmig
Messages: 103
Registered: January 2016
Senior Member
Dear all,

Am 12.03.2018 um 17:19 schrieb Christian Rahmig:
> Dear all,
>
> I created a Trac ticket for this issue ideally to be solved with railML
> 2.4 [1].
>
> Am 22.02.2018 um 13:23 schrieb Christian Rahmig:
>> In railML 2.x we have an attribute <track>@mainDir that describes the
>> preferred driving direction on that track. Possible values are "up",
>> "down", "unknown" and "none" and they always refer to the track
>> definition direction (trackBegin to trackEnd).
>>
>> The question to be answered:
>> Is the information relevant for any use case? Who uses @mainDir and for
>> what?
>
> [1] https://trac.railml.org/ticket/324

the attribute @mainDir has been marked DEPRECATED with railML version
2.4. However, for railML 3.x implementation the topic shall be reviewed
once again w.r.t. infrastructure related use cases.

Best regards
Christian

--
Christian Rahmig - Infrastructure scheme coordinator
railML.org (Registry of Associations: VR 5750)
Phone Coordinator: +49 173 2714509; railML.org: +49 351 47582911
Altplauen 19h; 01187 Dresden; Germany www.railml.org
Re: <track>@mainDir [message #1763 is a reply to message #1728] Sat, 14 April 2018 15:23 Go to previous messageGo to next message
Torben Brand is currently offline  Torben Brand
Messages: 40
Registered: March 2016
Member
We use track@mainDir in Norway in railML2.3nor for our use case SCTP and capacity planning.

<track @mainDir>
We use mainDir for three things:
1. to indicate on a macroscopic level if the line is single (mainDir="none") or double track (mainDir="up" or "down")
2. to indicate the usual driving direction of the double track
3. to indicate the position of the track in a mesocopic level (mainDir="up" track on the right side in increasing mileage)

We ask to reverse the deprecation or give us an alternative suggestion towards our mapping needs.
Re: <track>@mainDir [message #1765 is a reply to message #1763] Mon, 16 April 2018 14:57 Go to previous messageGo to next message
Thomas Nygreen is currently offline  Thomas Nygreen
Messages: 7
Registered: February 2017
Junior Member
I would like to use the opportunity to question the "unknown" value for @mainDir. As the attribute is optional, it seems more natural not to specify it if the value is unknown, and to interpret an unspecified @mainDir as "unknown". Consequently, the "unknown" option could be removed.
Re: <track>@mainDir [message #1766 is a reply to message #1763] Mon, 16 April 2018 16:51 Go to previous messageGo to next message
christian.rahmig is currently offline  christian.rahmig
Messages: 103
Registered: January 2016
Senior Member
Dear Torben,

thank you for sharing your application view that I would like to comment
on from railML.org coordinators view. However, the view of the whole
community is important to us.

Am 14.04.2018 um 15:23 schrieb Torben Brand:
> We use track@mainDir in Norway in railML2.3nor for our use
> case SCTP and capacity planning.
>
> <track @mainDir>
> We use mainDir for three things:
> 1. to indicate on a macroscopic level if the line is single
> (mainDir="none") or double track (mainDir="up" or "down")

Actually, reasoning is only possible in one direction: if there is a
single track line, it cannot have a main driving direction and the usage
of @mainDir="none" is recommended (see [1]). It is not possible to
conclude from the main driving direction whether a line is single track
or double track. The remaining question in particular: how to
distinguish between single track line and double track line without
preferred main direction?

> 2. to indicate the usual driving direction of the double
> track

Agreed. This is the original and described intention of <track>@mainDir
as modelled in railML 2.x. Since we thought to recognize a lack of usage
in the community and additonally noticed a wider usage of free float in
modern railways with doubled signalling in electronic signalling
systems, we tried to adapt and clarify the current model to serve the
community.

> 3. to indicate the position of the track in a mesocopic
> level (mainDir="up" track on the right side in increasing
> mileage)

Sorry, but I do not understand the meaning of this usage scenario. Could
you please provide an example for a track with mainDir="up" and
mainDir="down" following your proposal? From my current understanding,
this approach does not work since the rule "driving on right side" is
not unique among European railway lines and even can differ along a
single railway line. However, maybe your examples may convince me ;-)

> We ask to reverse the deprecation or give us an alternative
> suggestion towards our mapping needs.

I respect your request and leave it up to the discussion here in the
forum to come to a conclusion for railML 2.4.

[1] https://wiki.railml.org/index.php?title=IS:track

Best regards
Christian

--
Christian Rahmig - Infrastructure scheme coordinator
railML.org (Registry of Associations: VR 5750)
Phone Coordinator: +49 173 2714509; railML.org: +49 351 47582911
Altplauen 19h; 01187 Dresden; Germany www.railml.org
Re: <track>@mainDir [message #1767 is a reply to message #1765] Mon, 16 April 2018 16:55 Go to previous messageGo to next message
christian.rahmig is currently offline  christian.rahmig
Messages: 103
Registered: January 2016
Senior Member
Dear Thomas,

Am 16.04.2018 um 14:57 schrieb Thomas Nygreen:
> I would like to use the opportunity to question the
> "unknown" value for @mainDir. As the attribute is optional,
> it seems more natural not to specify it if the value is
> unknown, and to interpret an unspecified @mainDir as
> "unknown". Consequently, the "unknown" option could be
> removed.

your approach sounds reasonable to me. We may think about implementing
this "unknown issue" as a fundamental railML 3 pattern. This means:
whenever an optional attribute (or optional element) is not provided,
this means that the related information is unknown.

Any comments from the community are highly appreciated.

Best regards
Christian

--
Christian Rahmig - Infrastructure scheme coordinator
railML.org (Registry of Associations: VR 5750)
Phone Coordinator: +49 173 2714509; railML.org: +49 351 47582911
Altplauen 19h; 01187 Dresden; Germany www.railml.org
Re: <track>@mainDir [message #1768 is a reply to message #1767] Wed, 18 April 2018 19:05 Go to previous messageGo to next message
Thomas Nygreen is currently offline  Thomas Nygreen
Messages: 7
Registered: February 2017
Junior Member
christian.rahmig wrote on Mon, 16 April 2018 16:55

your approach sounds reasonable to me. We may think about implementing
this "unknown issue" as a fundamental railML 3 pattern.


That is a good idea. It should probably be discussed broader than just in this thread.

christian.rahmig wrote on Mon, 16 April 2018 16:55

whenever an optional attribute (or optional element) is not provided,
this means that the related information is unknown.


Unless a default value is specified. We should strive to always define a default value for optional attributes when possible. A lot of attributes will still not have a meaningful default, in which case a rule interpreting them as unknown will be useful.
Re: <track>@mainDir [message #1769 is a reply to message #1766] Wed, 18 April 2018 19:47 Go to previous messageGo to next message
Thomas Nygreen is currently offline  Thomas Nygreen
Messages: 7
Registered: February 2017
Junior Member
christian.rahmig wrote on Mon, 16 April 2018 16:51

Am 14.04.2018 um 15:23 schrieb Torben Brand:
> We use mainDir for three things:
> 1. to indicate on a macroscopic level if the line is single
> (mainDir="none") or double track (mainDir="up" or "down")


Actually, reasoning is only possible in one direction: if there is a
single track line, it cannot have a main driving direction and the usage
of @mainDir="none" is recommended (see [1]Wink. It is not possible to
conclude from the main driving direction whether a line is single track
or double track. The remaining question in particular: how to
distinguish between single track line and double track line without
preferred main direction?


I am not sure that Torben is correct in this case. It is rather the number of connections between two OCPs that determine if the line is single or double track.

However, there will always be operational rules or practices in addition to those represented in railML. In Norway we have no double track lines without preferred main direction. Neither do we have double tracks with mandatory track directions, but the performance will in many cases be degraded by running on the left track. The operational rules and practices will differ from country to country, and so will the amount of information you can interpret from the values of mainDir. The value itself does however have the same meaning in all countries.

christian.rahmig wrote on Mon, 16 April 2018 16:51

> 3. to indicate the position of the track in a mesocopic
> level (mainDir="up" track on the right side in increasing
> mileage)


Sorry, but I do not understand the meaning of this usage scenario. Could
you please provide an example for a track with mainDir="up" and
mainDir="down" following your proposal? From my current understanding,
this approach does not work since the rule "driving on right side" is
not unique among European railway lines and even can differ along a
single railway line. However, maybe your examples may convince me ;-)


I do not have a relevant railML example at hand, but again, this is a case of differing practices. A renderer for drawing schematic line plans must always have a set of national drawing rules. And in this nation we always drive on the right side under normal operations. For a renderer, or a macroscopic simulator or infra parser, it is much simpler to look at the track's mainDir than to lay out the full topology to determine which track is to the right (or left) and thereby used in increasing (or decreasing) mileage.

Best regards
Thomas
Re: <track>@mainDir [message #1774 is a reply to message #1769] Mon, 23 April 2018 16:55 Go to previous message
Benno Kuehnl is currently offline  Benno Kuehnl
Messages: 1
Registered: March 2017
Junior Member
Dear all,

I haven't needed the <track>@mainDir attribute for anything up to now, but reading this discussion, a use case came into my mind:

On a station with two tracks and one platform for each track (thus, no island platform), it is often very clearly stated on the signs, which platform is meant for trains going into which direction („<-- Züge Richtung München | Züge Richtung Weitdraußen -->"). When driving in the "wrong" direction at such a station, the passengers need significantly more time to switch to the correct platform, thus it is necessary to inform them more prominently. If it happens on a regular basis, you might need to send some workers to change the signs. That needs could be determined by the mainDir attribute.

On the other hand, I think that any "it would be possible in both ways, but usually it is done this way" can impose severe safety issues when someone falsely relies on things happening the usual way all the time.

Best regards,
Benno
Previous Topic: [railML 3.1] Simplification of area engineering over more than one netElement element
Goto Forum:
  


Current Time: Tue Apr 24 04:49:56 CEST 2018