[OSM-talk-ie] Tagging townlands in OSM-Ireland

Patrick Casey pcasey at compuserve.com
Wed Sep 10 21:00:11 BST 2008


Thanks for this, Paul. Thanks to your input and that of Dermot I now have
the confidence to fix things that I have direct knowledge of (but don't
worry, I won't go in with the proverbial chainsaw and create havoc. I'm
taking things a few steps at a time).

I wish Google Earth would use the OSM data for its road overlays. Many of
the smaller roads in the Google Earth views of the West of Ireland are at
least 2-300ft from their true positions.

Paddy


-----Original Message-----
From: talk-ie-bounces at openstreetmap.org
[mailto:talk-ie-bounces at openstreetmap.org] On Behalf Of Paul Cunnane
Sent: 10 September 2008 20:57
To: talk-ie at openstreetmap.org
Subject: Re: [OSM-talk-ie] Tagging townlands in OSM-Ireland


 
On Wed, 2008-09-10 at 18:37 +0200, Patrick Casey wrote:

> I now understand why one or two places in 'my' area in the
> Tubber-Ruan-Crusheen triangle aren't where they should be on the map (e.g.
> the village of Ruan). May I take the liberty of moving them a few hundred
> yards to their correct sites  or does a committee have to vet changes. I'm
> new to OSM and not yet sure of the protocol and the last thing I want to
do
> is contravene some existing convention or practice. I'm very conscious of
> the fact that in a project like this the participants need to conform to
> standards to prevent anarchy. Anything else and the OSM becomes useless.

I'll wade in here with my experience. If something's wrong on the map,
fix it. Sometimes mapping starts off approximate and depends on local
knowledge to perfect it, on the basis that approximate data is better
than no data.

In the unlikely event that someone disagrees with your edit, they can
discuss it with you, but I haven't seen that become an issue yet.

> If I interpret your tenor correctly there would be no objection if I were
to
> "re-baptise" e.g. Coolbane as a locality. I would restrict such changes to
> places that I know well and have walked over and can vouch for the fact
that
> they consist of a few scattered houses or farms or cabhals and not tiny
> groups of habitations that one might call a hamlet. Would that be OK ?

Sounds good. I've been leaving the "hamlet" designation alone because I
was unaware of "locality" - I'll be correcting any I come across in
future.

> Some of the lakes in 'my' area seem to have been very roughly drawn (or my
> GPS was faulty). For example, the footpath to Dromore Castle from the car
> park (which I walked with the GPS) appears to cross part of the lake.
There
> is a New Testament example of walking on water but I haven't yet managed
to
> emulate it. Is it possible that those lakes were traced from a relatively
> low-resolution satellite image ? I'm talking about small errors here, e.g.
a
> couple of hundred yards. 

Lakes have generally been traced from satellite imagery, and I've come
across some that are offset. When I determine a more accurate location
with GPS, I move lakes.

> I couldn't understand your statement ".....it's a bad idea to tag for the
> renderer....".  Could you clarify that, please ?

The idea is to have a database of geographical information that
accurately represents the world. The maps, whether Mapnik or Osmarender,
are interpretations of that database. Each renderer has its own ideas
about how to visually interpret the data, and those ideas sometimes vary
over time.

Basically, you should tag nodes and ways so that the information you've
captured about them accurately reflects the real world. If the rendered
maps don't adequately represent that data, it's better to fix the
renderers than to store inaccurate data in order to get a better visual
representation.

-- 
Paul


_______________________________________________
Talk-ie mailing list
Talk-ie at openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-ie





More information about the Talk-ie mailing list