Re: [Talk-transit] Defining service on railway=tram

2019-02-17 Thread Jarek Piórkowski
On Sun, 17 Feb 2019 at 13:43, Stephen Sprunk  wrote:
>
> The current four service values are based on physical characteristics of the 
> track that are easily observed on the ground and unlikely to change.
>
> This proposal seems to overload that with an indication of how the track is 
> used, and we already have a tag for that: usage. Granted, none of its 
> existing values seem like a great fit, but if we're going to add new values, 
> wouldn't that be the right place?
>
> I can't recall having seen a tram siding, but I have seen light rail sidings. 
> Given the fuzzy line between the two, it seems unwise for any of their (many) 
> common tags to have different meanings.
>
> Also, does this problem even need solving? With route relations, consumers 
> can easily deduce that a given track is not normally used, so why have a 
> redundant method of indicating the same thing? They're certainly more work to 
> create and maintain, but they also provide more benefits, so that seems fair.

Hi Stephen,

The problem I was initially trying to solve initially was lack of
definition or standardization. Similar types of tram track
("non-revenue", "auxillary", "irregular" - as you wish to call them)
are being tagged inconsistently as service=spur, service=siding, or
service=yard, even within the same city, because a standard was never
suggested.

I wanted to tag some non-revenue track and there wasn't a
specification of how it should be tagged. As I wrote in the initial
message to tagging list, on-ground difference might be that standing
street-side, on regular track one might see a tram go by every 5
minutes, whereas on non-revenue trackage at least hours and possibly
days might go by between trams. Relations can indicate this, but
service tag was already used and rendered - just not used
consistently.

It is true that usage is a more correct word for this, but in looking
at several dozen cities I saw hardly any tram track currently tagged
with usage. If making a new tag/value, using usage might be a good
idea.

thanks,
--Jarek

___
Talk-transit mailing list
Talk-transit@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-transit


Re: [Talk-transit] Defining service on railway=tram

2019-02-17 Thread Stephen Sprunk
The current four service values are based on physical characteristics of the track that are easily observed on the ground and unlikely to change.This proposal seems to overload that with an indication of how the track is used, and we already have a tag for that: usage. Granted, none of its existing values seem like a great fit, but if we're going to add new values, wouldn't that be the right place?I can't recall having seen a tram siding, but I have seen light rail sidings. Given the fuzzy line between the two, it seems unwise for any of their (many) common tags to have different meanings.Also, does this problem even need solving? With route relations, consumers can easily deduce that a given track is not normally used, so why have a redundant method of indicating the same thing? They're certainly more work to create and maintain, but they also provide more benefits, so that seems fair.S___
Talk-transit mailing list
Talk-transit@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-transit


Re: [Talk-transit] Defining service on railway=tram

2019-02-17 Thread Jarek Piórkowski
On Sun, 17 Feb 2019 at 10:55, Markus  wrote:

> detour seems a bit unsuitable for turn tracks or connections because afaik it 
> implies a longer route, but the tracks i have in mind are rather shortcuts. 
> Maybe deviation doesn't have this meaning?

Right, I withdraw the detour suggestion. Irregular is still best so far I think.

> But don't you also want to include tracks that are only used for drives to 
> the depot? If so, auxiliary/irregular/secondary seems like a better fit 
> anyway.

Tracks that are only used for drives to depot would have service=yard
in my mind.

>> I chose "siding" because I didn't want to invent new tag value, to
>> avoid too big and slow of a change. But maybe we should do it, what do
>> you think?
>
> Imo if a tag or key doesn't fit it's better to invent a new. It would be nice 
> though to hear opinions from other mappers.
>
> Besides, are you sure that siding tracks for trams similar to those for 
> trains don't exist somewhere? If they exist and we use service=siding for 
> auxiliary tracks, there won't be a distinction anymore (or a new tag would 
> have to be invented for real tram siding tracks).

Honestly - I've looked through many of the systems and there isn't
much that functions like a real siding in railway sense. Wiki
describes railway siding as "These tracks are used by slower trains to
be overtaken or to let passengers enter/leave the train if the main
tracks do not have platforms." which doesn't really apply on tram
systems I've seen. The wiki even notes "These tracks might be hard to
differ from the main tracks in some cases." ... if unsure, we could
just leave the tracks with no service.

The closest thing that comes to mind for tram sidings are tracks for
parking or bypassing trams that aren't being used right now, usually
near route termini (e.g. https://osm.org/way/46140380,
https://osm.org/way/69049487). But I've tried to come up with a
description of these that wouldn't also include other tracks not used
in regular service, and was unable to do so. I am leaning towards
deciding it's not really worth drawing the distinction where it's
difficult to actually articulate one. I was able to explain
non-revenue/irregular, yard, and crossover; I don't know if I can
define more categories well.

Regarding input from other mappers, I have also gotten a response from
User:Tigerfell on the wiki at
https://wiki.openstreetmap.org/wiki/Talk:Tag:railway%3Dtram#service.3D.2A_for_tram
- maybe that gives you more ideas.

While we think some more, I might try to look into what the current
tagging is on light rail systems (in the Porto Metro sense, not the
Berlin S-Bahn sense...) to see if it makes sense to include them in a
proposal together with trams, what with trams and light rail forming a
kind of a continuum ("Service classes on street-crossing passenger
rail transport networks"?).

Thanks,
--Jarek

___
Talk-transit mailing list
Talk-transit@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-transit


Re: [Talk-transit] Defining service on railway=tram

2019-02-17 Thread Markus
On Sat, 16 Feb 2019, 20:20 Jarek Piórkowski  I agree with you that "siding" is not a good description, but it
> seemed the least-wrong one of the 4 railway values in use. "Spur" to
> me sounds like something that branches off and ends (as illustrated
> for railways). Meanwhile with service=siding I am looking for a
> description for tracks that connect two stretches of regular tracks.
>

You're right, spur doesn't seem to fit either.

I do like your suggestions - "irregular" should be clear enough, and I
> like "auxillary" as well except I don't know if its meaning is
> commonly-enough understood. "Minor" seems like it could be
> misunderstood: if you have two lines, one of which runs more
> frequently than the other, are the tracks of the less-frequent line
> "minor"? Or how about service=detour?
>

I agree that minor could be misunderstood.

detour seems a bit unsuitable for turn tracks or connections because afaik
it implies a longer route, but the tracks i have in mind are rather
shortcuts. Maybe deviation doesn't have this meaning?

But don't you also want to include tracks that are only used for drives to
the depot? If so, auxiliary/irregular/secondary seems like a better fit
anyway.

I chose "siding" because I didn't want to invent new tag value, to
> avoid too big and slow of a change. But maybe we should do it, what do
> you think?
>

Imo if a tag or key doesn't fit it's better to invent a new. It would be
nice though to hear opinions from other mappers.

Besides, are you sure that siding tracks for trams similar to those for
trains don't exist somewhere? If they exist and we use service=siding for
auxiliary tracks, there won't be a distinction anymore (or a new tag would
have to be invented for real tram siding tracks).

Or how about if we were recommend that of the current options, siding
> should be used (to attempt to standardize what we have); and in
> parallel launch a formal proposal process for adding more proper tram
> service=irregular?
>

I wouldn't recommend this. It seems too confusing and i don't see a benefit.

>
___
Talk-transit mailing list
Talk-transit@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-transit