TEC released new data on 8 november of 2014. Why would we still want to
work with stale data from July?
There is no need to include metadata like a release date on all 3
nodes. When there is a change, my scripts will pick up on it. source is not
needed either, it is included in the changeset.
What is a whellchair=yes and where does this data come from? It was not in
the datasets I converted. I also 'integrated' a stop in the middle of
nowhere and I would be very surprised there was special accomodation like a
higher platform present there to ease boarding with a wheelchair.
All village names from Osmose are ALL CAPS, which looks ugly. I converted
them all and reincluded accents where needed.
I communicated with Frédéric to work together. I would convert and
preprocess the data. He preferred to do it himself, but something like
calculating route_ref is quite hard if you can't load all of the tables
into a database and perform a JOIN query.
And then there is our Belgian situation. Not only Brussels poses problems,
there are also the faciliteitengemeentes around it and Commines - Voeren.
And then there is our decision to drop the city name for the larger cities,
or rather to include them for the smaller ones and the villages.
My scripts cater for all that 'business logic'. If Osmose uses a different
way to convert the data and that is what Frédéric was planning to do, then
there will inevitably always be differences flagged as 'errors'. That is
why I had asked him to not provide the possibility to use Osmose for
integrating TEC stops.
All this is moot now. Initial integration of TEC stops is over. I did it in
a way that was way more efficient than what could have been achieved with
Osmose. I'm in the process of merging the ones I had included using more
than 1 node. This will probably take another week or two. Part of the
conversion is automatic. The actual merging is done manually by (virtually)
visiting many of them all over again one by one.
Jo
2014-11-16 1:27 GMT+01:00 Jo winfi...@gmail.com:
I asked Frédéric Rodrigo explicitly in July to exclude TEC stops from his
Osmose integration tool. He wanted to work directly from the source data
and not with the data as I would have provided it.
All stops are already in OSM now, so initial integration/import is over.
I sent a mail to the list a few days ago telling everybody I'm converting
all Belgian stops to a way of tagging which will allow 1 node to be used
per stop. (Well 2, 1 for the 'platform'/pole and often another for the
stop_position). I had started to use 1 node per operator, where TEC behaves
like 6 operators, each assigning refs to the stops.
Leave the route_ref alone, it's useful to me. It became route_ref:De_Lijn
and route_ref:TECx where x can be BCHLNX. It's something that can be easily
added/checked while surveying in the field. For the stops of MIVB/STIB I'm
still using route_ref.
If Frédéric wants to include Osmose tests for bus and tram stops in
Belgium, ask him to read the following:
http://wiki.openstreetmap.org/wiki/WikiProject_Belgium/Conventions/Bus_stops
Jo
2014-11-16 0:58 GMT+01:00 André Pirard a.pirard.pa...@gmail.com:
Hi,
@Julien: not an interesting info.
Apparently all Wallonian TEC bus stops are hit by an Osmose report like
the following:
Please note that:
- there may apparently be a July update to TEC data
- source=...07-2014 to introduce (was obstinately refused for the
present data)
- should be 2014-07 (- is used for ISO format and / otherwise)
- route_ref= to be removed
- spelling mistakes
- *whellchair*=: extra w?
*Wallonia stop not integrated*
Wallonia stop of THEUX Place du Centre, RUE CHARLES RITTWEGER, THEUX
+ *name* = THEUX Place du Centre
+ *bus* = yes
+ *source* = tec-wl.be - 07-2014
+ *public_transport* = platform
+ *whellchair* = yes
+ *operator* = TEC
+ *ref* = LTHcent
+ *highway* = bus_stop
Error reported on: 2014-11-15
so-called help:
public transport
[image: Osmose-marker-l-8040.png]
http://wiki.openstreetmap.org/wiki/File:Osmose-marker-l-8040.png
item=8040 map http://osmose.openstreetmap.fr/map/?item=8040 list
http://osmose.openstreetmap.fr/errors?item=8040
*Detail*
- Class 1 RATP station not integrated : subway, Paris, France.
- Class 11 RATP station not integrated : RER, Paris, France.
- Class 21 RATP station not integrated : tramway, Paris, France.
- Class 41 TransGironde stop not integrated : bus of Gironde,
France.
- Class 51 TBC stop not integrated : public transport of Great
Bordeaux, France.
- Class 61 CG71 stop not integrated :
*Fix*
Add new points or merge existing. See also item 8041.
- Class 1 RATP station not integrated :
- Class 11 RATP station not integrated :
- Class 21 RATP station not integrated :
- Class 41 TransGironde stop not integrated :
- Class 51 TBC stop not integrated :
- Class 61 CG71 stop not integrated :
