Re: [Talk-transit] Public transport workshop in Germany

2009-05-29 Thread Jochen Topf
On Fri, May 29, 2009 at 08:54:48AM +0100, Roger Slevin wrote:
 What has not been mentioned specifically in this thread (although I know
 Peter is very much aware of it) is that there is an approved European
 Technical Specification (Identification of Fixed Objects in Public Transport
 - IFOPT) that has built on the experience of NaPTAN and other related work
 to date, and covers the same ground.  A colleague is putting together some
 comments on how the German work relates to IFOPT.  Early indications are
 that the matching of fundamentals is good (as might be expected - given that
 NaPTAN was a key input to IFOPT) ... but I hope something on this will be
 posted here in the next few days.

Lets not forget that this is OSM. We do things step by step here with
lots of experimentation. Its more about evolving to a good model than
creating some complex top-down design. The scheme under discussion is
one such step. Its not as complex as IFOPT and probably can't do all that
IFOPT can do, but it is reasonably similar to the current OSM model yet
more clear and powerful while still beeing understandable.

A complex model created by professionals for professionals is sure to
fail in OSM. In the long run we can work more and more towards this
complex model when the software supporthas improved and we understand
better what we want and what we need. But for now we should try to cram
everything in.

IFOPT seems for instance to allow full recursion on many of its objects
which is really hard to handle properly and has, in my opinion, currently
no place in OSM.

Of course where there are good ideas we should incorporate them. Especially
when naming things it makes sense to follow established practices here.

Jochen
-- 
Jochen Topf  joc...@remote.org  http://www.remote.org/jochen/  +49-721-388298


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


Re: [Talk-transit] Public transport workshop in Germany

2009-05-29 Thread Gerrit Lammert
Hi Jochen.

Jochen Topf wrote:
 The weekend before last we had a workshop in Karlsruhe, Germany on the
 topic of public transport in OSM. The idea was to bring interested
 people together to improve the modelling of public transport
 infrastructure and networks in OSM.

 The results have now been documented. See
 http://blog.geofabrik.de/?p=23 for details.

That is great. I did some work in this area by working out the
unified_stoparea proposal some months ago.
I am really happy there finally seems to be some progress in this area. :)

Here are some notes I have after reading (the German version of) your
wiki page.
I'm writing this before I read responses to this thread, to not get
confused. I hope this won't be obsoleted by what others wrote...

- the usage tag (and similar tags)
Here I would prefer to use hierarchical tagging, so instead of
  railway=rail; usage=branch
I would rather write
  railway=rail:branch

- ref/name tags
If you intend this to be filled with information like
ref=Line 10 or
name=Main station - zoo
I would rather see this information with the appropriate
route-relations. Things like
name=Ringleis
which name the actual rail corpus and not the train traveling upon it,
are fine with me.

- Stadtschnellbahnen (German)
It confuses me, that these only consist of subway and monorail. I think
the term may cause some confusion. I do not think it is necessary to
combine these two and form a group for them.

-separate geometry with tram
I disagree that this is a good idea. Most attributes belong to both (the
road and the tracks within the road). Together they form one entity.
Attributes that do not belong to both should use hierarchical tagging
scheme, e.g.:
  railway=tram:disused
  railway:gauge=100
  railway:operator=railcompany
  highway:operator=citymanagement

-concerning the tagging of transport stops.
Please have a look at my proposal regarding this:
http://wiki.openstreetmap.org/wiki/Proposed_features/unified_stoparea
it is very similar to what you propose and is already used by quite some
mappers throughout Europe.
Please try to build upon it, as you are basically defining something
very similar with new tag names.
I am not sure if it would be necessary or wise to have everything under
the public_transport name space as you propose

All in all, I am very happy with this scheme, foremost that someone
cares about this. :)
I am not sure if the scheme is not to complicated for the casual
not-so-concerned mapper. That is why I was trying to change as little as
possible with my proposal and allow a wide range of detail levels.
The irony is, that I personally rather prefer a complex and complete
model to the change-as-little-as-possible approach I finally proposed
(I'm German after all ;-)
I am eager to see, how your proposal is received by the crowd!

Greetings,

Gerrit


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


Re: [Talk-transit] Public transport workshop in Germany

2009-05-29 Thread Gerrit Lammert
Jochen Topf wrote:
 On Fri, May 29, 2009 at 06:07:55AM +0100, Peter Miller wrote:
 Out of interest where does this leave the following two proposals?

 Unified StopArea
 http://wiki.openstreetmap.org/wiki/Proposed_features/unified_stoparea

 QROTI
 http://wiki.openstreetmap.org/wiki/QROTI

 Should we review these other proposals, integrate any appropriate ideas 
 and then politely suggest that people refer to this new proposal for 
 guidance rather than to the other two pages? Would anyone be offended by 
 that?
 
 For the time beeing I have put links to those pages into the See also
 section. Maybe we should ask those people who have worked on those pages
 to comment.

As I mentioned earlier, I am founder and main contributor to the
unified_stoparea proposal.

Well, I actually _do_ feel a little offended by two things.
- First: I am confused at how little reference to my proposal I find
in the new scheme. It's not that I am vain or anything, but you copied
about anything from that proposal and just gave some items new names.
I wonder if you read it and really copied most of it, in which case I
would have expected some note about it, or if you came up with that on
your own, but than this would show bad research (but still means we both
came independently to the same solution(s)).
- Second: I tried to rally for a coherent PT-tagging scheme for weeks
without end in the beginning of the year (mainly the german list, but
also here) with disappointingly little feedback. I wished you (Jochen
and his group) would have participated in the process back then.
It would have saved me a lot of work and a lot of people that started to
use the unified_stoparea-proposal to eventually re-tag all of their work.

That said: I really _am_ happy that there is some progress in this
field. As you guys have a group (and the geofabrik) behind you, its much
more likely to succeed and become wide spread. And I am very interested
to have a unified tagging scheme, whatever name it bears.

Gerrit



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