Re: [Talk-GB] Tagging of shared use paths

[John Aldridge]

On 12/10/2020 12:41 PM, Ken Kilfedder wrote:
As a break from 'tagging for the renderer', I'd like to see rendering 
for the tags.


A long standing grump of mine!

We see lots of excellent effort put in to designing tagging schemes 
which could support a wide variety of applications, but rather little 
into creating those applications themselves, most especially 
consumer-directed maps.


As it is, the conversation typically goes

   Please don't tag for the renderer!

   Oh, sorry, how do I get The Map to look right then?

   You don't, The Map is intended for map creators, not end users

   Eh?

   But it's no problem, you can create your own map on your own server
   which renders however you want it to. All you need is a degree in
   computer science.

In fact, various people have bravely done this, but (a) you can't find 
those maps because they're not indexed from the OSM or OSMUK pages, (b) 
they sometimes don't cover the whole country, (c) they run on 
underpowered hardware, and (d) oh, that map worked last year, but it's 
gone missing now.


There'd be a whole lot less temptation to tag for the renderer, if the 
renderers rendered for the tags a bit better!


I appear to be in a grumpy mood today. Sorry! I love OSM really :)

John

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

Re: [Talk-GB] National Cycle Network removal/reclassification

[John Aldridge]

On 8/13/2020 3:41 PM, Simon Still wrote:
 
Width of cycleyway is definitely useful if separated from traffic but 
some way of reflecting the comfort of the riding experience on marked 
routes would be a big step forward. Traffic Volumes,. Lane widths, 
traffic speed all contribute (as does surface - gravel bad, cobbles bad, 
smooth tarmac good)


Definitely. I have no solution to offer, but it's perhaps worth noting 
that there are several classes of user here - what's fine for a mountain 
bike may be downright uncomfortable for a racing bike, and what's fine 
for a leisurely Sunday excursion might be wholly unsuitable for a daily 
commute.


Another aspect which has a big effect on the cycling experience is the 
frequency of encountering walkers (especially dog walkers). If you 
frequently have to slow to a walking pace to get past safely, it rapidly 
spoils the experience.


Being on an NCN is definitely no guarantee: one of the least comfortable 
couple of mile's cycling I recall was on the NCN 3 at Penzance, for example.


John

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

Re: [Talk-GB] Ficticious embankments? Vandalism.

[John Aldridge]

On 17-Mar-20 02:08, Warin wrote:

A single GPS trace is fine if that is all there is, better to average many GPS 
traces, in some locations I have 50+.


Though, AIUI, once you've reached this level of precision, remaining 
errors are likely to be systematic (e.g. satellites in a particular 
direction being generally received via a -- delaying -- reflection 
rather than directly). No amount of averaging will help with that.


John

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

Re: [Talk-GB] What is farmland?

[John Aldridge]

On 14-Dec-19 16:52, SK53 wrote:
Like Dave I have come to the view that mapping individual fields as 
farmland is a good way to do it.


I too concur. Here's the diary entry I wrote when I was doing the fields 
round here...


https://www.openstreetmap.org/user/jpsa/diary/17738

Cheers,
John

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

Re: [Talk-GB] Common Land has stopped rendering

[John Aldridge]

On 16-Mar-19 13:23, SK53 wrote:
No, Carto is just one way of viewing OSM data. It has a small number of 
maintainers and although highly visible does not represent any 
'official' view.


Yes and no.

For many people, the *point* of OSM is that it's a better version of 
Google maps. If data is not visible on the 'front page' rendering, it 
might as well not be there.


I agree, of course, with the principle of not tagging for a particular 
renderer. But along with espousing that principle should come an 
obligation to provide a renderer which does reflect (and hence 
encourage) what we regard as good tagging practice. Lacking that, people 
will keep tagging for Carto, and I can't entirely blame them.


There should, IMO, be a general purpose UK styled rendering on the front 
page of https://osmuk.org/


--
Cheers,
John

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

Re: [Talk-GB] We're missing changes to M1 Junction 36 which have apparently been in place for a year.

[John Aldridge]

On 11-Feb-19 09:37, Paul Berry wrote:
However I'm not overly keen on attempting to map the confluence 
of four major roads on foot.


What have others done in similar circumstances?
Driving round the roads with a GPSr in the car should get good enough 
data to be going on with.


--
Cheers,
John

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

Re: [Talk-GB] How to map houses

[John Aldridge]

On 27-Nov-18 11:50, Lester Caine wrote:

On 27/11/2018 11:40, John Aldridge wrote:
It would be useful if there was a means of splitting buildings in the 
editor(s).


I'm probably in a minority here, but since the mapper usually can't 
tell how the building is divided internally, it's more honest to leave 
the building undivided and put the housenumber etc. tags on nodes on 
the building boundary which represent the front doors.


I also think this is more useful to someone using the map, as it shows 
where to find the doorbell!


My source material has all the house divisions and we could even include 
the internal floorplans...


Perfect! Though doing that job properly would sometimes involve 3D 
mapping - there's one smallish building around here with four front 
doors which I suspect is split into two upstairs and two downstairs flats.


But that's very much the exception, I suggest. Mostly people seem to 
have arbitrarily divided a semi or terrace so as to give separate 
polygons to tag. Better all round to just tag front door nodes under 
those circumstances, I suggest. And where a single exterior door is 
shared between several dwellings (e.g. in a block of flats) something like


addr:housenumber=8-24
addr:interpolation=even

on the node deals neatly with the case.

--
Cheers,
John

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

Re: [Talk-GB] How to map houses

[John Aldridge]

It would be useful if there was a means of splitting buildings in the 
editor(s).


I'm probably in a minority here, but since the mapper usually can't tell 
how the building is divided internally, it's more honest to leave the 
building undivided and put the housenumber etc. tags on nodes on the 
building boundary which represent the front doors.


I also think this is more useful to someone using the map, as it shows 
where to find the doorbell!


--
Cheers,
John

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

Re: [Talk-GB] 'historic' county boundaries added to the database

[John Aldridge]

I'd like to register a +1 in favour of accepting these historic counties.

I *generally* agree with your principle of 'only mapping what is on the 
ground', but if we followed that strictly we wouldn't map current 
administrative boundaries either. These historic counties do, rightly or 
wrongly, form part of some people's sense of identity *today*, and I 
think that crosses the bar for inclusion.


--
Cheers,
John

On 10-Aug-18 09:38, Stuart Reynolds wrote:

Hi

I’ve watched this from afar, but thought that I would add my two 
pennyworth, as a more casual mapper.


Historic county boundaries have some merit (in a very general sense), 
but where do you draw the line? As it happens, I was discussing where, 
exactly, Middlesex was with my son only yesterday, and I looked it up on 
Wikipedia. Turns out that Middlesex has changed quite significantly over 
time. First of all, it existed. Then, some of it got plonked into London 
- and it had already lost the City of London and Westminster by then. 
Bits of it got hived off to Hertfordshire. Then the rest of it got 
incorporated into Greater London. So what would you map, historically? 
Do you map every single variation of it, and try and date them all? If 
you were going to map historic counties properly, then you must.


But think what this does to the data. Think what this does for the new 
mapper (who we are trying to encourage). There is now a mass of 
overlapping, conflicting entities to edit. You need to go through every 
one, laboriously, working out which ones you need to edit, and which 
ones you need to leave alone. It’s a data management nightmare, and the 
chances of the wrong thing being edited, or being edited incorrectly, 
rises exponentially.


Personally, I have never particularly liked the variety of ways that OSM 
attempts to map disused / demolished entities (e.g. bus station 
rebuilds, etc) even now. I am firmly of the opinion that we should be 
mapping existing, current, objects, and that things that don’t exist on 
the ground should be ripped out. If OSM as an organisation wants to take 
annual snapshots for posterity, or to set up a separate “historic OSM” 
then I am all for it - I won’t be mapping in it, myself, although I 
would have an interest in using it. As in my Middlesex example, though, 
you would still have data management issues unless you compartmentalise 
it by year - but that is a whole new interface or workflow.


So I am very strongly in favour of NOT mapping historic counties, and 
only mapping what is on the ground (or verifiably shortly to be there, 
as in new builds)


Stuart


On 10 Aug 2018, at 09:24, Sean Blanchflower > wrote:



I guess you at least acknowledge that not everyone agrees with your 
views below though.


A quick factual error though: the traditional/historic counties were 
not administrative in the sense that current areas are. The changes of 
the Local Government Act 1888 were to create administrative areas for 
the first time, and it was the fact that they were called 'counties' 
that has caused all the trouble since then. The government 
acknowledged that the new areas were distinct from the existing 
counties and were not replacing them, and in fact the Ordnance Survey 
continued to print them on maps after then.


How do we reach some compromise here? We seem to be at an impasse.

> I'm sorry, but this is complete and utter bullshit. The "historic" 
> county boundaries are no more "real" than the current ones. They were, 
> at the time, the administrative boundaries. They are no longer the 
> administrative boundaries.

>
> I do appreciate that there are matters where the historic boundaries are 
> relevant (primarily genealogical research). But that's not really a 
> mapping issue., And the emotional attachment to the pre-1974 boundaries 
> is just that - emotion, not based on any objective assessment. And the 
> fact that, in retrospect, the 1970s changes were over-reaching and did a 
> lot of harm does not change that.

>
> Describing the historic boundaries as "real" is like insisting that we 
> map, say, the old Euston station the way it was before it was rebuilt, 
> because it was a lot nicer then. It may well be the case that it was. 
> But we map what exists now, not what existed in the past and in 
> rose-tinted memory. The same with county (and other administrative) 
> boundaries. We map what is, not what was.



On Wed, Aug 8, 2018 at 3:49 PM Sean Blanchflower > wrote:


Hi all,
I'm smb1001 and have been adding the traditional county boundaries
recently. DaveF kindly let me know of the discussion thread here
so I've joined Talk-GB to add my side of things.

I'm not alone in thinking the traditional county boundaries have a
place on current maps. It's unfortunate here that these counties
are known as 'historic counties' as this implies that they are no
longer extant. The debate as to their current 

Re: [Talk-GB] Implicit speed limits: What to tag in built-up areas?

[John Aldridge]

On 02-May-18 11:55, Philip Barnes wrote:



I believe it's DoT policy not to allow 30mph repeaters (at least,
someone told me that).


True for roads with street lighting, but quite common, and required, on 30 mph 
roads with no street lights.


Ah, yes, thank you (both) for the clarification. And, for completeness, 
here's the reference



http://www.legislation.gov.uk/uksi/2016/362/schedule/10/made



http://www.legislation.gov.uk/uksi/2016/362/schedule/10/part/4/paragraph/2/made


--
Cheers,
John

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

Re: [Talk-GB] Implicit speed limits: What to tag in built-up areas?

[John Aldridge]

On 01-May-18 16:29, Philip Barnes wrote:

And yes, you may have to go back several roads before you see the
speed limit sign. No all local authorities put up the repeater signs
but that doesn't mean that the speed limit stops applying.


And 30mph limits don't need repeaters, for example it is possible to pass a 
30mph limit outside Glenfield and certainly get several miles across the city 
without seeing another sign.


I believe it's DoT policy not to allow 30mph repeaters (at least, 
someone told me that).


--
Cheers,
John

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

Re: [Talk-GB] Public Rights of Way data for Cambridgeshire

[John Aldridge]

One bit of feedback, from a first try at doing this for real: footpaths 
often cross parish boundaries, and at least in this area change their 
reference when they do so. But your slippy map only displays geometry 
for a single parish at a time, meaning that tracking the prow_ref value 
for the full length of a single path can take a lot of navigation within 
your tool.


Would it be hard to display geometry for all ROWs overlapping the 
current slippy map extent, whichever parish they are from?


--
Cheers,
John

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

Re: [Talk-GB] Public Rights of Way data for Cambridgeshire

[John Aldridge]

On 11-May-17 00:20, Robert Whittaker (OSM lists) wrote:

The presence of prow_ref=*
tags to allow matching with official data is almost non-existent.)


So what's the best way to fix this?

If I click on your map, it shows me something like

  Teversham FP 3
  (MS: 0 | ΔL/L: —)

which is a bit cryptic! Is the first line the prow_ref? Should I just 
copy it from there?


--
Cheers,
John

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

Re: [Talk-GB] Propose automated edit to update NAPTAN data in the west mids

[John Aldridge]

On 03-Feb-17 17:52, Brian Prangle wrote:

In line with the automated edits policy there's a wikipage

with full details


I realise that this proposal is only for West Midlands, but if it were 
extended nationwide, I'd have a concern about globally assuming NAPTAN's 
positional accuracy is better than OSM's existing data. I rather doubt 
that's the case for the ones I've re-surveyed round here, for example.


Even where it is true, there's a follow on task needed which will be to 
make sure nearby non-bus-stop geometry is also improved to match. I know 
there are many places where OSM data has been added from poorly 
georeferenced aerial photography, resulting in a map which, although 
perhaps up to 5 or 10 metres out, is nevertheless useful and usable. 
Simply fixing the bus stop coordinates in isolation will produce some 
very odd results like moving them into the middle of people's houses!


All the stuff about fixing route information sounds excellent, though! 
That's certainly in rather poor repair locally.


--
Cheers,
John

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

Re: [Talk-GB] Quarterly task

[John Aldridge]

Perhaps also worth trying to get the flow direction right.

http://tools.geofabrik.de/osmi/?view=water

can be helpful here.

--
Cheers,
John

On 04-Jan-17 15:22, Andy Townsend wrote:

On 03/01/2017 22:34, Rob Nickerson wrote:

Water sounds good.


... and to me too - in fact it's what I've been doing on and off for the
last couple of months, mainly in South Wales bordering areas that I've
been walking where I've noticed that streams are in the wrong places.

OS OpenData StreetView is quite predictable here - it's usually correct
except in the following (checkable) situations:

1) a mountain stream has obviously moved (which you can see from Bing
imagery, even if the latter is offset somewhat)

2) there's been some recent man-made intervention (e.g. a fishing lake)

3) it's a ditch beside a road - the OS tends to draw these in the
"wrong" places for cartographical purposes, but again you can align via
Bing, even if the latter is offset, because you can align the road.

4) the OS often adds a "catographic flourish" at the upper ends of
mountain streams which in my experience (Black Mountain / Beacons /
south of Hay) doesn't exist.

5) Sometimes the OS has the top ends of streams and ditches extending
further than you might think (e.g. through a boggy area). That's the
only one of these 5 that really needs surveying to check.


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

Re: [Talk-GB] access:psv

[John Aldridge]

On 13-Oct-16 18:51, Chris Hill wrote:

I have written about this process more than once in the past, for
example
http://chris-osm.blogspot.co.uk/2012/01/homogenised-data-no-thanks.html


I agree with this... any formal tagging schema is going to end up 
obstructing useful mapping of circumstances which the schema authors 
didn't consider.


I don't think that general principle should rule out the correction of 
simple errors, though. In this case I think I might send a message to 
the original mapper to ask them whether these unusual tags were created 
inadvertently and whether they'd mind if they were changed to correspond 
to more common practice.


I'd then happily take "please don't" for an answer, though.

--
Cheers,
John

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

Re: [Talk-GB] Autumn Quarterly Project

[John Aldridge]

On 10-Oct-16 20:26, Dan S wrote:

Really sorry for the stupid question, but: what is a participant in
this quarterly project actually supposed to *do*?


I asked a similar question a week or so ago. From the responses I got, I 
don't think there is a single shared vision of the goals of this project.


What /I'm/ doing is checking the Greg's map, and trying to make all the 
blobs round here go green :)


In most cases that involves filling in addr:postcode and fhrs:id on 
existing OSM features. I'm not, however, trusting that the postcode 
recorded on FHRS is accurate, and I'm not setting addr:postcode unless I 
can find corroborating information (e.g. the establishment's web site).


I'm definitely /not/ copying any other FHRS data into OSM tags. If 
people want to see the hygiene rating, they need to follow fhrs:id to 
the FHRS page, rather than looking at a stale OSM copy of that data 
which unlikely ever to be maintained.


Which reminds me... Greg's tool seems to have recently added two local 
churches (which don't in fact seem to have FHRS records at all), so I 
must wander over there to see if I can find a postcode on the 
noticeboards, since there doesn't seem to be one on either of their web 
sites!


--
Cheers,
John

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

Re: [Talk-GB] Autumn Quarterly Project

[John Aldridge]

On 02-Oct-16 17:30, SK53 wrote:

My personal rules on this have always been two independent sources of
information OR a survey...



FHRS data should contain full address details most of the time, so there
should be no need to add anything from the website other than the url...


By url do you mean the fhrs:id tag?


I'm confused, then, by the assertions on the web page for this project


https://wiki.openstreetmap.org/wiki/UK_2016_Q4_Project:_Food_Hygiene_Ratings


that

(a) this process might be completely automated [how does that square 
with requiring two sources of information], and


(b) that one of the goals is to accelerate our completion of UK postcode 
data [I'd assumed that implied we needed at least to add addr:postcode 
too, with or without further checks]


--
Cheers,
John

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

Re: [Talk-GB] Autumn Quarterly Project

[John Aldridge]

On 02-Oct-16 10:32, Jez Nicholson wrote:

I have added the page
to https://wiki.openstreetmap.org/wiki/UK_Quarterly_Project


I'm not quite clear what the concrete goals are here: is it just to add 
fhrs:id and addr:postcode tags to relevant establishments (and perhaps 
to map those establishments if they're not already there in OSM?)


What degree of verification is expected on the part of the mapper? Is 
the existence of a (probably fuzzy) name match between an OSM feature 
and a nearby FHRS record adequate justification for adding the relevant 
tags, or is some local knowledge or survey expected?


(On a related topic, suppose an establishment has a web site which 
includes useful information like address & postcode details. Would one 
be breaking rules to copy that information to OSM? After all, the text 
on the web site will be, AIUI, copyright.)


--
Cheers,
John

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

Re: [Talk-GB] Summer quarterly project

[John Aldridge]

On 11-Sep-16 18:45, Brian Prangle wrote:

Only a couple of weeks left - time to start thinking about our next
quarterly project...


Speed limits? There've been a whole lot of changes round here -- mostly 
new 20 mph limits -- recently!


--
Cheers,
John

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

Re: [Talk-GB] Does "Great Britain" need a relation with "place=island" on it?

[John Aldridge]

On 18-Aug-16 14:47, Craig Wallace wrote:

The Garmin maps is another question. They seem to label the whole area
with the island name, which is often unhelpful and confusing, even for
smaller islands. Probably better just to have a label in the middle of
the island, or on the coast.


If anyone else needs a workround for this, I noticed this issue when I 
rebuilt my Garmin maps the other day, and resolved it at least for now 
by commenting out the lines.


# place=island & name=* [0x53 resolution 19]
# place=islet & name=* [0x53 resolution 20]

from the 'polygons' style file. It was particularly irritating, because 
as well as the name scattered everywhere, it put a hatch pattern over 
the whole island (at least when viewed in BaseCamp)!


--
Cheers,
John

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

Re: [Talk-GB] Quarterly Project April-June 2016

[John Aldridge]

On 01-Apr-16 13:06, Steve Doerr wrote:

On 01/04/2016 11:51, Ed Loach wrote:

With noon rapidly approaching I'll withdraw my earlier suggestion
regarding deleting sadly-lacking-in-information buildings as part of a
future project.


LOL. Brilliant! Took me in.


Doh! Me too!

--
Cheers,
John

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

Re: [Talk-GB] Quarterly Project April-June 2016

[John Aldridge]

On 01-Apr-16 11:51, Ed Loach wrote:

With noon rapidly approaching I'll withdraw my earlier suggestion regarding 
deleting sadly-lacking-in-information buildings as part of a future project.


In my earlier reply I should have said that the general principle of a 
quarterly project to add information to un-attributed buildings seems 
admirable. It's only the deletion part I didn't like.


--
Cheers,
John

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

Re: [Talk-GB] Quarterly Project April-June 2016

[John Aldridge]

On 01-Apr-16 08:30, Ed Loach wrote:

I was thinking of a different possible project which perhaps we can
postpone until a different quarter rather than starting today.

There are a lot of buildings which have been traced from various imagery
layers but with no other useful information on them. Is it a shop, a
house, what is its address, etc?

I was going to suggest that we spend three months trying to gather the
address information for such buildings, and at the end of the quarter
remove all the others to give us a better idea of how complete the map
is. If we just want approximate building shapes we’ve got OS OpenData
for that. At the moment they give a false impression of completeness,
and also make it harder without zooming in to work out which areas still
need a proper survey.


I vote against that. Outlines traced (with appropriate care) from 
imagery will often be more accurate than OS OpenData outlines.



If the project proves successful we could extend it to farmland areas
where hedges and fences haven’t been added to break up the area.


I know several places round here where there is no hedge, fence or any 
other substantial boundary between fields: the farmer just leaves an 
uncultivated strip.


--
Cheers,
John

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

Re: [Talk-GB] Open data (Was: Parliamentary debate mentions OSM)

[John Aldridge]

On 29-Mar-16 10:19, Rob Nickerson wrote:

Should we attempt to include everything that is in the open data
datasets plus our on the ground additions (manually or, unless we
suddenly get many more mappers, by some form of controlled merge) or
should we leave the end users with the task of mixing OSM with the open
data?


I'm generally not in favour of importing 'definitive' data from other 
sources, and it would be better to have some dynamic overlay procedure 
(which must of course available to normal end-users of the map on 
www.openstreetmap.org, not just to sophisticated OSM data processors).


An example is parish boundaries which, I understand, have been imported 
from Ordnance Survey data. The problem with these are that they often 
get inadvertently corrupted in OSM: they tend to lie along other 
features, which means that it's rather easy to get them inadvertently to 
share nodes, which in turn ends up with them being dragged around by 
mistake.


I appreciate that we don't have such a dynamic overlay procedure, and 
nor do I have a solution to offer, but the problem is real: if we import 
data which is definitively specified elsewhere, we are pretty much 
guaranteeing that OSM's version of that data will be inferior to other 
mapping, even if only because it'll get out of date.


I suppose a satisfactory alternative to dynamic overlay would be if any 
such import were required to have an automated procedure for adopting 
regular updates from the definitive source.


--
Cheers,
John

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

Re: [Talk-GB] Open data (Was: Parliamentary debate mentions OSM)

[John Aldridge]

On 26-Mar-16 06:30, Rob Nickerson wrote:

 >And there seems to be some more
 >open data on the way from the OS.

Interesting. A good opportunity but maybe also a threat - I wonder what
quality of map can now be produced from OGL and other open data and how
that compares to OSM. We already have some users that mix and match
between OSM and other sources.

How do we ensure the mix continues to contain a lot of OSM data?


By ensuring that OSM data is of higher quality, or contains useful 
information still absent from those other sources. If we can't or don't 
do that, OSM (in the UK) will cease to have a purpose, and can be left 
to wither un-mourned.


--
Cheers,
John

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

Re: [Talk-GB] Pubs as areas: should be map the property or the building?

[John Aldridge]

On 14-Mar-16 10:43, Stuart Reynolds wrote:

The one pub that I plotted, I added when I was doing a couple of nearby
schools and noticed that it was missing. I used exactly the same
principle as for the schools - an outer “amenity=pub”  polygon and an
inner “building=pub” for the actual building.


Sounds good to me... similarly for restaurants which have their own 
car-park and perhaps outdoor seating area?


--
Cheers,
John

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

Re: [Talk-GB] place=village/town/city

[John Aldridge]

On 15-Feb-16 13:48, Colin Smale wrote:

And would this
mean that St Davids is place=town, place:designation=city or the other
way round?


I have no axe to grind here (the city I live near has a population 
>100,000 anyway), but if the former, I suspect the residents of St 
David's would not be happy with this, and they'd still keep editing it 
back to place=city!


I don't really see the point, either: if you want to signal the 
population, then use the population=* tag. I'm content with letting 
place=* be the name by which the place is widely known: if the residents 
call it a city, then that's what it should be.


--
Cheers,
John

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

Re: [Talk-GB] Additional Tagging Suggestions for Schools

[John Aldridge]

On 12-Feb-16 20:22, Lester Caine wrote:

On 12/02/16 19:34, John Aldridge wrote:

On 12-Feb-16 17:31, Lester Caine wrote:

I adopted the standard of official_name=xxx where the edubase listing
differs from the signage or other local 'format'...


I'm not objecting, but why did you feel you wanted to clone information
from edubase into OSM at all?

Given the presence of the ref:edubase tag, I can look the official
information up if I want to know (and without the risk of it becoming
out of date).


I have no problem on them being removed,


I have no intention of removing anything -- I'm merely curious!


but while cross checking they
were a useful reference, and seeing the entry I can see how many are
different in edubase.


You mean they serve the same function as not:name does for street names 
in reducing noise from automated consistency checking? OK, I can see 
that's useful.



But don't assume that the data
provided by the third party are more up to date than the OSM version ;)


But in this case of official_name I thought you were saying the 
appropriate value was the value from edubase, so by definition it 
couldn't be out of date there.



A substantial number of the website links I loaded are out of date on
the edubase records :( OSM is now more accurate than edubase ...


Now that doesn't surprise me at all! I absolutely agree that putting 
(e.g.) website information into OSM is appropriate -- especially if 
edubase is wrong.


--
Cheers,
John

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

Re: [Talk-GB] UK style rendering port ...

[John Aldridge]

I hope this can be done, too, if it makes it easier to be much more 
responsive about making the renderer respect best UK tagging practice.


One of the reasons, IMO, that its difficult to prevent people tagging 
for the renderer, is that there currently seems to be little attempt, 
when a tagging scheme is devised, to consider whether and how the 
results of such tagging should be made visible on the map, and then to 
get any necessary renderer changes implemented.


--
Cheers,
John

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

Re: [Talk-GB] Postbox refs

[John Aldridge]

On 03/04/2015 18:49, Rob Nickerson wrote:

Hi,

I've noticed recently that postbox references have started appearing
with a D on the end (same ref as before just with a D added to the
end). This seems to be in areas where RM have updated the collection times.

Does anyone know what the D means?


I'm not certain, but the ones I've noticed round here seem to be those 
with latest collection times very early in the day (Perhaps done along 
with the delivery round, rather than a separate collection round? That's 
conjecture, though).


--
Cheers,
John

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

Re: [Talk-GB] Quarterly Projects Update

[John Aldridge]

On 03/04/2015 15:55, Colin Smale wrote:

So now instead of having two versions of the name (1. according to the
council and 2. according to the sign) we are to have a third version,
according to OSM?


That's not what I meant, unless I've misunderstood you.

There's the definitive name according to the council, which I believe 
we can't use because it's copyright information (perhaps even 
inaccessible to the general public, I'm not sure). If it were to become 
available in the future I'd be quite happy for that to go in an 
official_name=* tag.


There's what appears on road signs, which is what I think OSM should be 
putting in name=*. Granted this principle can cause problems in cases 
like the one I described earlier where those signs are inconsistent with 
each other, but those are fairly rare and in the end, at least in the 
cases I've seen, rather unimportant.


There's also what appears on the Ordnance Survey OpenData maps, which 
might or might not match either of the first two. As I understand it we 
could legally copy this information into a third os_name (or 
ordnancesurvey_name) tag, but I don't see what that buys us.



I am also thinking of our conventions with regard to
punctuation and abbreviations (not to mention capitalisation), which
lead to mappers not copying the sign verbatim into OSM but entering some
kind of normalised version. If the sign is gospel we need to stop that
as well.


Pedantically I agree with you, but it's not what's done, and I don't 
think it's worth changing.



According to the National Street Gazetteer, the official source of
street names is the local authority in their Local Street Gazetteer
which they have to feed into the NSG. Is this the other database to
which you refer?


Probably, I couldn't remember the details.

--
Cheers,
John

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

Re: [Talk-GB] Quarterly Projects Update

[John Aldridge]

On 03/04/2015 14:59, Colin Smale wrote:

Why not tag both spelling variants? They are both correct in their own
frame of reference.

If it differs to what is on the ground, we can use official_name=* for
the name given by the local authority, warts an' all.


I wouldn't have a problem with this at all, provided the official data 
is licensed in such a way that we can use it.


As I recall, however, although the OS map is suitably licensed, it is 
not itself definitive, but is derived (perhaps with errors) from another 
database which we are not entitled to copy. I don't think there's much 
value in adding an os_name=* which may differ from both the ground-truth 
and the definitive data.



Even council employees and contractors make mistakes occasionally.
Should we be legitimising and propagating manifest errors by putting the
errors into OSM?


Because IMO they're *not* errors in OSM, whose job is to map physical 
reality, not to be a repository for various geographic information 
databases. I'm aware that not everyone shares this view.


--
Cheers,
John

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

Re: [Talk-GB] Quarterly Projects Update

[John Aldridge]

On 03/04/2015 13:43, Robert Whittaker (OSM lists) wrote:

While I'm here, one thing it might be good to discuss in relation to
this project is how best to tag some of the different types of
delivery/collection infrastructure...


I had a go a while ago at working out what the consensus was on this 
topic, and offered to write it up. After some discussion it became clear 
that there was no general consensus (although some people had very 
strong opinions), and that what was perhaps the majority opinion 
conflicted with the documentation on the Wiki. There also seemed little 
interest then in reaching a documented conclusion to the discussion, so 
I gave up.


I do think it would be useful to document a tagging scheme on the UK 
guidelines page before starting this activity.


--
Cheers,
John

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

Re: [Talk-GB] Advice on footpaths - when should they be separate, when not?

[John Aldridge]

On 01/12/2014 11:39, Stuart Reynolds wrote:

Looking for some advice in Bletchley, specifically, but to answer a more 
general point about footpaths.

 :

So what is the guidance here? Ought the road have a distinct footpath both 
sides? Or not footpath, and use the tags on the road, or just connecting spurs 
from the footpath to the road at key points (e.g. opposite Selwyn Grove), or 
what…?


I think mapping the path explicitly is perfectly reasonable if it is 
(for at least some of its length) separated from the road by a 
non-trivial distance (say more than a couple of feet of grass).


I agree that managing the transitions between such explicit paths and 
implicit sidewalk tagged sections is clunky!


--
Cheers,
John

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

Re: [Talk-GB] Voting mechanical edit: UK shop names

[John Aldridge]

On 31/10/2014 22:51, Matthijs Melissen wrote:

Dear all,

Voting is now open for the proposal to unify the names of chain shops
within the UK by renaming them.


I'm genuinely not sure about this proposal.

Philosophically, I think OSM is a *mapping* project (duh!), and so the 
correct name is exactly the text which appears on the shop's sign. If 
a chain of shops were to be inconsistent in how they are signed (I don't 
know how common this is), then so be it. This leads me to the view that 
correcting shop names (although desirable) can only safely be done by 
a ground survey (to check what the sign actually says).


On the other hand, I have no doubt that the overwhelming majority of the 
variations this proposal will eliminate are simply mapping errors, and 
do not reflect actual signing differences. The average quality of the 
map will be significantly improved by the automated edit, even if a few 
genuine differences are accidentally lost.


--
Cheers,
John

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

Re: [Talk-GB] Post NaPTAN edits (Cambridge area)

[John Aldridge]

Thanks for your help, Richard  Brian.

--
Cheers,
John

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

[Talk-GB] Post NaPTAN edits (Cambridge area)

[John Aldridge]

I've been reading the NaPTAN import thread, and saw a reference to the 
Novam viewer, which thinks  that I should fix a few stops round here. 
Before I do that, I'd like to check that I'm going to do it right!



Presumably I should simply update the positions of the stops, if 
necessary, based on ground survey (actually I've already done that -- I 
hope that wasn't wrong).


I assume I should check the naptan:* tags are accurate and fix them if 
necessary.


I also assume I should add a shelter=(yes|no) tag.

But what about route_ref? Should I be setting this to a list such as 
route_ref=16;17? Some of the colour schemes in Novam complain about a 
missing route_ref, others don't, and I see the documentation 
http://wiki.openstreetmap.org/wiki/Bus_stop says Use of the tags 
'route_ref'... in no longer recommended.


If it makes any difference, route relations seem to have been set up 
round here (though I haven't checked whether they are complete). Though 
the stops are only linked to those relations by proximity, I


Having done all that, should I simply delete the naptan:verified tag? 
Or should I leave it there and change it from no to yes?



Anything else?

--
Cheers,
John

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

Re: [Talk-GB] Life Ring - British English

[John Aldridge]

I've always heard them called lifebelts.

--
Cheers,
John

On 16/06/2014 10:30, Richard Mann wrote:

en-gb is probably lifebuoy

I've never heard it called a life ring - that's too vague a name. Most
people would probably refer to it by starting to describe it - one of those
red ring things that you can use to help someone who is drowning.


On Mon, Jun 16, 2014 at 10:05 AM, Andreas Goss andi...@t-online.de wrote:


I'm trying to clean up the emergency tags in the Wiki and found
emergency=life_ring as well as some less used other tag combinations with
amenity and buoy.

Is life ring how it is commonly referred to in British English. Just
wanted to make sure it's not literal translation from German and isn't used
in the UK at all. Wikipedia lists a lot of different names. I guess
lifebuoy is more American? And is it written life ring or lifering? Both
correct?

http://wiki.openstreetmap.org/wiki/Tag:emergency%3Dlife_ring
https://en.wikipedia.org/wiki/Lifebuoy


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

Re: [Talk-GB] Somerset Levels Flooding

[John Aldridge]

On 07/02/2014 12:37, John Baker wrote:

Always to play the devils advocate.
We have all heard about mapping for the renderer but are you mapping
for the third party data providers that is slow at updating the planet
data. They should use the data correctly. It is annoying the stale
content that some devs/providers provide, maybe it would encourage them
to put some effort into updating the content for frequently.


Something along the lines of

  http://wiki.openstreetmap.org/wiki/Proposed_features/temporary

would help, though I think it would benefit from some way of specifying 
that the start  end dates are estimates only.


That proposal seems to be languishing, though.

--
Cheers,
John

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

[Talk-GB] Fwd: Re: Royal Mail Parcelforce delivery offices

[John Aldridge]

I received this reply from SK53, inadvertently sent to me alone, which 
I'm resending to the list with his permission...
---BeginMessage---
First of all don't go by definitions on the wiki. The consensus tag (at
least in UK) is amenity=post_depot.

Second, we have the perfectly reasonably tag post_depot to distinguish
these things.

Third, DONT TAG FOR THE RENDERER. Just because you cant see them on the OSM
front page does not mean that you cant search for them. These are truly
different things. Locating delivery offices is not always easy, they are
often in some obscure parts of an industrial estate. I actually mapped a
local one because my father had tried to find it and failed. They need to
be distinguished, if they are just tagged post office they disappear in the
noise of real post offices.

As for DHL, CityLink etc, this is a more involved problem, particularly as
the Royal Mail is now not necessarily a universal service provider.
However, I think some other tag should be used for these distinct from
post_office and post_depot. parcel_depot or parcel_office come immediately
to mind.

Typically a delivery office will have a post code ending 1AA, as in the
Maidenhead Delivery office. Unmapped ones can be found by looking for large
numbers of post codes in the same place. I wrote a bit about adding them to
OSM 
herehttp://sk53-osm.blogspot.co.uk/2013/12/british-postcodes-on-openstreetmap.html
.

Jerry


On 15 January 2014 17:59, John Aldridge j...@jjdash.demon.co.uk wrote:

 On 15/01/2014 17:49, Robert Whittaker (OSM lists) wrote:

 On 15 January 2014 16:03, John Aldridge j...@jjdash.demon.co.uk wrote:

 If I believed there was a snowflake in hell's chance of getting
 amenity=post_depot objects displayed on the default map, I'd be more
 sympathetic, but as it is I think I'd prefer something like...


 If everyone followed that reasoning, there would never be any new tags
 created! Besides, deliberately tagging known-not-to-be-a-post-office
 objects as post offices is incorrectly tagging for the renderer.


 I do understand your position, but I don't think that's a fair criticism
 of mine. These things *are* post offices by the definition in the Wiki, so
 I don't think it's unreasonable to tag them that way.

 --
 Cheers,
 John


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

---End Message---
___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb

[Talk-GB] Fwd: Re: Royal Mail Parcelforce delivery offices

[John Aldridge]

To which I replied...
---BeginMessage---
[[ Did you mean to email me privately, rather than the list? Should I 
copy this email to the list? ]]



On 17/01/2014 09:48, SK53 wrote:

First of all don't go by definitions on the wiki. The consensus tag (at
least in UK) is amenity=post_depot.

Second, we have the perfectly reasonably tag post_depot to distinguish
these things.


OK, I'll take your word for it, but is there any chance of getting this 
consensus documented somewhere, then, perhaps on


  http://wiki.openstreetmap.org/wiki/United_Kingdom_Tagging_Guidelines

which would have saved me starting the thread :)



Third, DONT TAG FOR THE RENDERER. Just because you cant see them on the
OSM front page does not mean that you cant search for them. These are
truly different things. Locating delivery offices is not always easy,
they are often in some obscure parts of an industrial estate. I actually
mapped a local one because my father had tried to find it and failed.
They need to be distinguished, if they are just tagged post office they
disappear in the noise of real post offices.


No need to shout :)

I'm trying to distinguish between tagging for the current renderer 
(which is obviously wrong), and tagging in a way which is never going to 
be rendered (which, at least for concrete things like this, is pointless!)


I'm delighted (if a little surprised, given the Wiki) that Robert 
Whittaker thinks that a patch to render post_depot objects would be 
accepted by the renderer team. If so, I'm disappointed that, given 
there's a consensus around using that tag, that hasn't been done already.


The fact is that the renderer on openstreetmap.org *is* OpenStreetMap 
for most people, and it's crucial that it be kept up to date with best 
mapping practices.




As for DHL, CityLink etc, this is a more involved problem, particularly
as the Royal Mail is now not necessarily a universal service provider.
However, I think some other tag should be used for these distinct from
post_office and post_depot. parcel_depot or parcel_office come
immediately to mind.


I'm not necessarily disagreeing, but could you expand on why? Is it just 
that there's still a broad cultural belief that Post Office Ltd post 
offices are special, even though there's little actual difference left?


What about Parcelforce depots (Parcelforce is part of the Postal 
Services Holding Company plc, along with Post Office Ltd and Royal Mail 
itself): should their depots be amenity=post_office, or would you 
suggest parcel_depot for them too?


--
Cheers,
John
---End Message---
___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb

[Talk-GB] Fwd: Re: Royal Mail Parcelforce delivery offices

[John Aldridge]

And received the following clarification...
---BeginMessage---
Yes of course if should have gone to lists (its a while since I used the
gmail client, which I dislike).

Tagging for the renderer is very important, and pushing things in so they
get rendered has a serious deleterious affect on the data. I use OSM data
as data: when someone degrades the meaning by overloading a tag, this can
mean that all of the data become useless. landuse=grass for farmland is a
classic example: it is now impossible to get useful landuse figures or maps
out of OSM.

As for Parcel Force yes I'd use parcel_depot, as I would for British Rail
Red Star parcels if they still existed!

As for adding depot to the renderer, I think I agree with Robert its
unlikely to clash with other icons and completes existing information. Mind
you showing doctors is rather more important. A more significant thing
would be to add this to Nominatim which would enable people to find such
things without them being rendered.

By all means copy back to list.


On 17 January 2014 10:41, John Aldridge j...@jjdash.demon.co.uk wrote:

 [[ Did you mean to email me privately, rather than the list? Should I copy
 this email to the list? ]]



 On 17/01/2014 09:48, SK53 wrote:

 First of all don't go by definitions on the wiki. The consensus tag (at
 least in UK) is amenity=post_depot.

 Second, we have the perfectly reasonably tag post_depot to distinguish
 these things.


 OK, I'll take your word for it, but is there any chance of getting this
 consensus documented somewhere, then, perhaps on

   http://wiki.openstreetmap.org/wiki/United_Kingdom_Tagging_Guidelines

 which would have saved me starting the thread :)



  Third, DONT TAG FOR THE RENDERER. Just because you cant see them on the
 OSM front page does not mean that you cant search for them. These are
 truly different things. Locating delivery offices is not always easy,
 they are often in some obscure parts of an industrial estate. I actually
 mapped a local one because my father had tried to find it and failed.
 They need to be distinguished, if they are just tagged post office they
 disappear in the noise of real post offices.


 No need to shout :)

 I'm trying to distinguish between tagging for the current renderer (which
 is obviously wrong), and tagging in a way which is never going to be
 rendered (which, at least for concrete things like this, is pointless!)

 I'm delighted (if a little surprised, given the Wiki) that Robert
 Whittaker thinks that a patch to render post_depot objects would be
 accepted by the renderer team. If so, I'm disappointed that, given there's
 a consensus around using that tag, that hasn't been done already.

 The fact is that the renderer on openstreetmap.org *is* OpenStreetMap for
 most people, and it's crucial that it be kept up to date with best mapping
 practices.



  As for DHL, CityLink etc, this is a more involved problem, particularly
 as the Royal Mail is now not necessarily a universal service provider.
 However, I think some other tag should be used for these distinct from
 post_office and post_depot. parcel_depot or parcel_office come
 immediately to mind.


 I'm not necessarily disagreeing, but could you expand on why? Is it just
 that there's still a broad cultural belief that Post Office Ltd post
 offices are special, even though there's little actual difference left?

 What about Parcelforce depots (Parcelforce is part of the Postal Services
 Holding Company plc, along with Post Office Ltd and Royal Mail itself):
 should their depots be amenity=post_office, or would you suggest
 parcel_depot for them too?

 --
 Cheers,
 John

---End Message---
___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb

Re: [Talk-GB] Royal Mail Parcelforce delivery offices

[John Aldridge]

On 17/01/2014 15:54, SK53 wrote:


Tagging for the renderer is very important, and pushing things in so
they get rendered has a serious deleterious affect on the data. I use
OSM data as data: when someone degrades the meaning by overloading a
tag, this can mean that all of the data become useless. landuse=grass
for farmland is a classic example: it is now impossible to get useful
landuse figures or maps out of OSM.


Agreed, of course. Note that I did not believe I was proposing to 
overload the post_office tag: I believed (based on the Wiki) that I 
was proposing to use it accurately, but to add subsidiary tag which 
allowed the various kinds of post_office to be distinguished.



However, there does seem to be a consensus (at least amongst those who 
feel strongly) that this is the wrong approach, and that having separate 
amenity=* tag values is a better way to go. So to make this concrete, 
here's a second draft proposal...


**

'Regular' post office branches (as currently run by Post Office Ltd) are 
the only things which should be tagged 'amentity=post_office'. The 
'operator=*' tag is not required for these (and perhaps better omitted, 
since the name of the company keeps changing!)


Royal mail delivery offices are the only things which which, at present, 
should be tagged 'amenity=post_depot'. The tag 'operator=Royal Mail' 
should also be specified.


The depots of parcel services (including Parcelforce as well as DHL, 
CityLink etc.) should be tagged 'amenity=parcel_depot'. The additional 
'operator=*' tag is essential.


That leaves us, in principle, with the question about what to do if a 
private company starts offering letter post services. Perhaps 
'amenity=post_depot' would be reasonable in that eventuality, too.


**

If these proposals (or some refinement thereof) seem to meet with 
approval, I'm happy to draft an update to


http://wiki.openstreetmap.org/wiki/United_Kingdom_Tagging_Guidelines

to document it (is this something I can just do, or do I need to be an 
'approved person' to edit the Wiki?)



In addition, it would be really nice if someone could start the process 
of getting the default renderer updated to display the two new tags


 amenity=post_depot, and
 amenity=parcel_depot

does anyone know how to go about getting this done?


I guess there's a meta-question too: is my help appreciated trying to 
get this agreed and documented, or would this be better left to others?


--
Cheers,
John

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

Re: [Talk-GB] Royal Mail Parcelforce delivery offices

[John Aldridge]

On 16/01/2014 09:34, Pierre Riteau wrote:

On Wed, 15 Jan 2014, at 05:59 PM, John Aldridge wrote:

On 15/01/2014 17:49, Robert Whittaker (OSM lists) wrote:

On 15 January 2014 16:03, John Aldridge j...@jjdash.demon.co.uk wrote:

If I believed there was a snowflake in hell's chance of getting
amenity=post_depot objects displayed on the default map, I'd be more
sympathetic, but as it is I think I'd prefer something like...


If everyone followed that reasoning, there would never be any new tags
created! Besides, deliberately tagging known-not-to-be-a-post-office
objects as post offices is incorrectly tagging for the renderer.


I do understand your position, but I don't think that's a fair criticism
of mine. These things *are* post offices by the definition in the Wiki,
so I don't think it's unreasonable to tag them that way.


Then maybe the definition in the Wiki should be modified to correspond
to what most people will intuitively see as a post office?

I would say that a post office is not only a place where you can send
*or* collect letters and parcels, but rather a place where you can send
letters and parcels (including buying stamps and weighting your
parcels), *and possibly* collect letters and parcels.


Heh... that definition excludes 'regular' (Post Office Ltd) post 
offices, though! If you try to send a letter at one of those, they tell 
you to put it in the (Royal Mail) post box outside.


--
Cheers,
John

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

Re: [Talk-GB] Royal Mail Parcelforce delivery offices

[John Aldridge]

On 15/01/2014 17:49, Robert Whittaker (OSM lists) wrote:

On 15 January 2014 16:03, John Aldridge j...@jjdash.demon.co.uk wrote:

If I believed there was a snowflake in hell's chance of getting
amenity=post_depot objects displayed on the default map, I'd be more
sympathetic, but as it is I think I'd prefer something like...


If everyone followed that reasoning, there would never be any new tags
created! Besides, deliberately tagging known-not-to-be-a-post-office
objects as post offices is incorrectly tagging for the renderer.


I do understand your position, but I don't think that's a fair criticism 
of mine. These things *are* post offices by the definition in the Wiki, 
so I don't think it's unreasonable to tag them that way.


--
Cheers,
John

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

[Talk-GB] Royal Mail Parcelforce delivery offices

[John Aldridge]

Is there a consensus on how to tag Royal Mail  Parcel Force delivery 
offices?


Are these amenity=post_office, or something else?

Presumably Parcelforce offices are distinguished by being tagged 
operator=Parcelforce?


Apologies if I've missed something in the documentation or help -- I did 
look!


--
Cheers,
John

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

Re: [Talk-GB] Question regarding OS Opendata

[John Aldridge]

On 27/12/2013 15:07, tony wroblewski wrote:

I have a question regarding the OS Opendata. How accurately aligned is
it? I've noticed that in some places the bing orphophotos match up
almost perfectly to the buildings below, and it other areas there can be
some difference (sometimes up to a meter of).


I don't know how accurate OS Opendata is, but I do know that Bing is 
sometimes several metres off (and sometimes differently off depending on 
how far you are zoomed in).


I've had carefully surveyed data buggered up a few times by people 
'correcting' it to match the Bing photography!


--
Cheers,
John

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

Re: [Talk-GB] walls versus landuse=field

[John Aldridge]

On 30/04/2013 09:38, Henry Gomersall wrote:

I noticed that lots of fields, for example in
http://www.openstreetmap.org/?lat=53.92332lon=-1.7091zoom=15layers=M
are shown as closed loops of landuse=field. Clearly walls/fences and
enclosed fields are somewhat equivalent, but subtly different in terms
of what they describe (certainly, walls are not always around fields)


Sometimes fields do not have any physical barrier, such as a fence, 
wall, hedge or ditch. Rather, the farmer just leaves an uncultivated 
margin separating adjacent fields. The position of this margin is 
(often, at least) stable from year to year, so seems worth representing 
in OSM. This seems to argue for mapping each field as a landuse=field 
loop, at least in such cases.


--
Cheers,
John

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

[Talk-GB] Invisible/impassable rights-of-way

[John Aldridge]

All this discussion of rights of way reminds me: is there a consensus 
about how (and whether) to map rights-of-way which are either impassable 
or invisible?


I've encountered examples of both round here, and have so far chosen not 
to map them at all, on the grounds that we're trying to map the actual 
state of the ground, not some legal fiction.


Do people concur?

--
Cheers,
John

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

Re: [Talk-GB] Invisible/impassable rights-of-way

[John Aldridge]

On 25/01/2013 18:52, Andy Street wrote:

On Fri, 2013-01-25 at 13:58 +, Dave F. wrote:

Blockages of ways are often just temporary.
I disagree with Andy Street's comment:
If you can't traverse a right of way then it shouldn't have a highway
tag.


Okay perhaps I could have been clearer but I wasn't suggesting omitting
the highway tag on paths that have the occasional fallen tree or
something that is likely to be rectified quickly, what I had in mind was
when someone builds a house over a public right of way or where you'd
need power-tools because the path is completely non-existent. It's the
same principal as roadworks where we don't change how we tag unless they
are going to close the road for a significant length of time.


In one case, the location of the path is clear, because it runs between 
two walls and the green Public Footpath signs are present, but a 
section of it has become completely and densely overgrown with brambles. 
It also has a large pile of earth obstructing it, which makes me wonder 
whether the obstruction is deliberate. It's been like this for at least 
a couple of years.



If there is a problem with the path notify the local authority
regardless of how you tag it in OSM.


I've done that already, though I would sympathise with the council if 
they chose to do nothing about it, because the section is only a couple 
of hundred yards long, and there is an equally convenient alternative 
walking route.



In the other case, the right of way runs diagonally across a field, but 
on the first few times I first visited it (over a period of a couple of 
years, so it wasn't just a temporary issue) the field was full of crops, 
and there was no sign of the path on the ground. There is an alternative 
path running round two sides of the field, but it is not a right of way.


This case is now of only theoretical interest, though, because the last 
time I went there the farmer had reinstated the diagonal path, and I was 
able to map it properly.


--
Cheers,
John

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

Re: [Talk-GB] [OSM-talk] Rendering of Farmland not 'Light' enough?

[John Aldridge]

On 06/01/2013 10:13, Robert Whittaker (OSM lists) wrote:

On 6 January 2013 01:54, Eugene Alvin Villar sea...@gmail.com wrote:
In the first proposal, I find it very difficult to see the difference
between the farmland and the non-tagged areas. It's a bit easier in
the second proposal. It could maybe be made a bit lighter, but not by
that much.


Agreed -- a little lighter would be nice, but not as much as the first 
proposal.



In this final image, I have adjusted the Hue and Saturation to provide
more of a 'green':

https://docs.google.com/open?id=0B6J5ZA1hu93bZXhzdVJMVU44X2M


I think it would be a pity to lose the hue relationship between fields  
farmyards (which are currently different lightnesses of an otherwise 
similar brown).


--
Cheers,
John

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

Re: [Talk-GB] Marking landuse and field boundaries

[John Aldridge]

On 31/12/2012 21:59, Graham Jones wrote:

I would like to see field boundaries and land uses in OSM, for the same
reason as you.   I think the main reason that there are not many in
there, is that they are very difficult to survey.


I second that! See my diary entry
  http://www.openstreetmap.org/user/jpsa/diary/17738
for the issues I had mapping fields round here.

--
Cheers,
John

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