Re: [Talk-GB] place=village/town/city
On 15/02/16 14:15, Colin Smale wrote: > On 2016-02-15 13:42, Lester Caine wrote: > >> So Bath is also a >> city despite being below some arbitrary population limit. >> > Bath has around 100k inhabitants, not exactly a hamlet... But it doesn't > have a city council, only Charter Trustees. Bath has not lost it's city status, unlike Rochester, so the designation is correct. >> If we know the >> population then it should be recorded, or a link to some other database >> that can provide a current and possibly historic population record? >> > There is a well-established key population=* > : http://wiki.openstreetmap.org/wiki/Key:population > > Populations change every day of course, so they are never entirely > accurate. But the wiki describes also population:date and > source:population which are important to put the number in the right > context, as is putting the tag on the right geometrical object which > really should be a polygon (so either admin boundaries or landuse or > place) and not a node. There we will have to disagree ... In my book there should be a node for every place in the UK. And it's location should be suitable to the 'centre' of the place. Personally I use the geonames.com as a cross reference and the population figures there are an alternative. It may actually be useful to add the geomnames reference to OSM and then use the name transalations via that ... but for population we still need a more reliable source? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] place=village/town/city
On 15/02/16 11:08, Mark Goodge wrote: > The only way to reconcile this, in the long run, is to have two separate > tags for populated places, one describing the size according to global > OSM guidelines, and one describing the legal status according to local law. Since there is a 'Should normally' in the wiki entry for place=city, then the population is NOT a hard and fast rule. It was left woolly specifically because the legal status should take priority, so St. David's has been a city since 1994 having had that status restored by the request of the Queen. city up until 1888 and as a prominent cathedral location that is the well established rule. So Bath is also a city despite being below some arbitrary population limit. If we know the population then it should be recorded, or a link to some other database that can provide a current and possibly historic population record? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Additional Tagging Suggestions for Schools
On 12/02/16 23:01, John Aldridge wrote: > 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. I suspect that some of the NAMES on edubase are actually wrong as well, certainly for some of the academy schools, although there does seem to be a little more consistency NOW for adding a new entry and closing the old one when a schools status changes, so the logged 'official_name' on OSM realy is just a hook to how the match was confirmed. Now we just need messages from edubase when a schools reference changes :) But I am playing with historic stuff to see if I can create a history for each school from the edubase data. Very much a back burner job, but amazing what you can conjure up using a relational database ... -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] place=village/town/city
On 11/02/16 21:32, Michael Booth wrote: > So my question is, how are we defining villages, towns and cities? Only > by population, or do we also take into account their generally accepted > status (whilst trying to be consistent across the country)? Population is only a rough indicator ... as you say it's more local choice on any final distinction. I think Facebook may have solved the problem ... EVERYTHING is a city! Even if getting the city into the right county/country is something that still has to be fixed :( -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Additional Tagging Suggestions for Schools
On 12/02/16 10:51, Robert Whittaker (OSM lists) wrote: > That said, it's on my To Do list to look at doing a fuzzy name > comparison -- probably by deleting common words / phrases (School, > Academy, CofE, Voluntary Controlled, and, &, etc) from both strings > and then seeing if what's left is equal. This would probably be used > to flag up possible errors in matched pairs (either in the OSM name or > the matching), rather than as a way to get better matching in the > first place. The latter can always be solved by adding the ref:* > values to OSM. I adopted the standard of official_name=xxx where the edubase listing differs from the signage or other local 'format'. A few schools have been 're-branded' as they became academies, but either the local site has yet to updated, or the school IS using a different name to that registered. A quite common one is the CofE and how it's printed on site or on other documents. This is a bit of a mess across the board :( -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Additional Tagging Suggestions for Schools
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, but while cross checking they were a useful reference, and seeing the entry I can see how many are different in edubase. Of cause what would be nice is a tool that imports third party data records given a reference in OSM. But don't assume that the data provided by the third party are more up to date than the OSM version ;) A substantial number of the website links I loaded are out of date on the edubase records :( OSM is now more accurate than edubase ... -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
[Talk-GB] 'Training Centres'
I've got a few loose ends on the HR and WR postcode areas relating to schools that are not ... Worcester Snoezelen is a training centre and activity centre specialising in support for special needs pupils. I put it down as a school since pupils do attend regularly as part of their education, but for them it is more of an activity centre with specialist therapy. It is also a training centre for teachers and the like. A switch to 'activity_centre' seems right, but that is not a current tag. http://www.openstreetmap.org/way/392376528 Vipassana Meditation Centre is essentially a college specializing in teaching meditation but it is not covered by edubase so should it remain as a college or is something else a better option? http://www.openstreetmap.org/way/395464497 I've hacked a few entries where there are two edubase IDs on the same site, but I'm not happy with this setup. Relations have been used for a couple of sites, and sites like 'Hereford Cathedral Collage' need tidying up in this respect, but I'm not happy with the current tags even though they trick the heck tools that things are OK. Abbey Park Schools http://www.openstreetmap.org/way/72534583 Splitting the site in two seems wrong Heart of Worcestershire College http://www.openstreetmap.org/relation/5902498 http://www.openstreetmap.org/way/38737994 And some missing bits ;) The whole spread of sites has just one edubase ID ... -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Schools project - update 3
On 26/01/16 19:38, Rob Nickerson wrote: > 4. HS and ZE are 100% complete! > Wow! The HS (Outer Hebrides) and ZE (Lerwick) postcodes are complete. > Thanks go to OpenStreetMapper seumas. Many other postcodes (with more > schools within them ;-) ) are close to this now too > http://robert.mathmos.net/osm/schools/progress/ Actually the completion list is actually a lot better than that. The current process has problems with special cases which ARE complete as best we can tag them, but the progress tool does not understand the problems. But it would be better to also display what percentage actually have edubase or similar ID tags as well. That coverage needs more work yet. I've hacked the current name check script and can identify those schools that exist but without an ID. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Schools project - update 3
On 26/01/16 21:20, Rob Nickerson wrote: > > BTW: The amenity=college page on the wiki [1] is lacking a lot. Could do > with some extra details (rather than simply linking to wikipedia) and > it's context needs expanding beyond just the UK. Any takers? > > [1] https://wiki.openstreetmap.org/wiki/Tag:amenity%3Dcollege Actually the way that the UK higher education structure has changed it's possibly pointless today bothering with 'college' as a category. Many 'Technical Colleges' now call themselves Universities, and secondary education now extends into the sixth form college sector. When a University includes 'Hairdressing' or 'Car mechanics' then it's degree status looks a little suspect? Even the Wikipedia article is tagged as 'Outdated' for 'Further Education' :) That said, The remaining college sector does seem to be consolidating with a well defined 'non-university' prospectus and the EduBase2 categories for University and College is probably an ideal division to follow, and I was tagging sixth form colleges as College rather than School following that tagging. The question perhaps is should the fancy 'Academy' Secondary school be tagged separately at all? Since these do not have to include that in their titles ... The tricky bit being the 16-19 age range campuses. Of cause once the URN is included one can look up the secondary data anyway. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Schools Progress Tracker Update
On 24/01/16 13:09, Lester Caine wrote: > That just leave 4 more objects to check :) OK most of the 'red dot' items were out of area, but I've tidied them up anyway as they are still Worcestershire. Pershore College has been sort of solved by following the current signage ;) It has 'Part of Warwickshire College' on the entrance signs, so I've added that to the name. The campuses of 'Heart of Worcestershire College' all seem to have new signs with that on but nothing else, so I've added a '- xxx' with the campus name taken from their website. On my todo list still is to add the rest of the missing sites. Think hat this just leaves the Abbey Park Schools problem and the New College Worcester to be sorted. The information is in the database, just the progress display showing a problem. Next month I'll move on to 'DY' as a lot of that is still 'Worcestershire' :) -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Schools Progress Tracker Update
On 24/01/16 10:43, Ed Loach wrote: > But the above and my edits yesterday make me wonder if we can settle on a way > to map two schools that share grounds. I have been tagging the grounds as > amenity=school, and the separate buildings for the two schools with the > different ref:edubase etc tags on. These don't get matched - I end up with a > proximity match for one of the two schools to the school grounds area, and a > blue dot for the other. > Examples: > http://www.openstreetmap.org/way/393006341 > http://www.openstreetmap.org/way/64096389 I have the same problem with Abbey Park Schools http://www.openstreetmap.org/?mlat=52.110348=-2.081149=16#map=18/52.10995/-2.08240 The central block is shared along with the grounds, and both school buildings have their edubase ref, but the site does not so the check fails. I'm three schools down on 100%, the other 'problem' New College Worcester is the RNIB College, and has two edubase references for some reason. I've tagged with the one which matches the website, and added a node for the second, but need some way to make sure it's recognised? The progress list quotes this as 'New College Worcester' but that is not what appears on the edubase list ... New College Worcester (ISP) and New College Worcester (NMSS) have different ids. Not sure how Northwick Manor Primary School slipped through the net ;) My other problems are South Worcestershire College, which is two campuses one in Evesham, which edubase identifies, and the second in Malvern ... which is not listed. However the bigger problem in my todo list is Heart of Worcestershire College, which also has a campus in Malvern along with another dozen sites spread around Worcester, Bromsgrove and Redditch. Pershore College is on the not matching list, and is now part of the Warwickshire College Group. So I think I just need to switch the edubase to that. Worcester Snoezelen is another one of those specialist teaching centres that edubase does not list but is correct, so how do we tag them to be ignored? Same with 'Avoncroft Arts Centre' but that is not in the WR area, having a B postcode I think. That just leave 4 more objects to check :) -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Schools Progress Tracker Update
On 24/01/16 16:14, Dave F wrote: > If we let you know the why certain schools are flagging an error could > you update your list with an extra column to indicate the reason? It > would save users from double checking. Since a number of them are down to me, and ARE correct, that would be useful ;) -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] OSM with Wikidata: 27,232 matches found in England
On 24/01/16 11:46, Edward Betts wrote: > It should be possible to use this a basis for uploading. The results can be > grouped by category and county when uploaded. Looks like you are still using your original data? http://edwardbetts.com/osm-wikidata/england/county/Worcestershire/Schools St. Mary's 'ceased operations at the end of the summer term 2014' (from wikipedia ;) ) Some of the OSM links are to the original building rather than the campus. I've not got up as far as the north of the county yet where the buildings need changing, but I had thought that wikidata was looking to import the entire edubase data to fill in he gaps in the schools list, so personally I would view the small number of current matches for Worcestershire as a crib sheet to add the wikipedia pages and wait to add the 'DfE URN' linkage once that process has completed. The question really is should both the wikipedia link and the wikidata link exist? Malvern St James has the correct wikipedia=en:Malvern St James and that is the one which should gain the wikidata tag as well or should the satellite sites duplicate all the main site tags? This is one of the areas we are still tidying up on the schools project. But I am now at the point where adding the wikipedia tag, but this opens a can of worms since https://en.wikipedia.org/wiki/Category:Schools_in_Worcester only lists some schools and some of them are now closed ... so where does one stop cross project? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Schools Progress Tracker Update
On 24/01/16 17:53, Brian Prangle wrote: > DY almost complete so you need to pick another area. TF needs some > attention rgds Brian I'll head over to HR ... although the corner of 'B' covering Worcestershire still needs some work. Covering areas I've contacts who can check things on the ground. ;) -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
[Talk-GB] OSMSchools - Loose ends ...
OK - I've just finished working down the list for the WR page, and have actually added more schools than were flagged as missing, but I think part of that was because the location data for the post codes did not tie with a number of premises. So this may result in a few miss matched schools. One thing I need to tidy up is where there are multiple sites in some cases across postcodes, so the primary site has an edubase ref linked to the main school in another postcode area. I've tagged every part of a set of school sites with the correct edubase reference, and I think that this should be used as an initial match? This also overrides the situation where the edubase title does not match the name used on the site? The other block of non-matched sites are given by institutions which are not managed via edubase listings. I've flagged the 'Morton Fire College' and now have a couple of schools in Worcester such as the 'Elgar School of Music' and 'Worcester Snoezelen' which I've tagged as ref:edubase='not listed' but I think perhaps some other tag should be used to show that an amenity=school object has be checked and is correct? The one query I think I need to create is one which flags schools that are missing a edubase reference. While all the schools are now listed, I know some do not have a full tag set including street, postcode. edubase and website. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Quarterly Project : Schools - Wales data
On 19/01/16 13:17, Bogus Zaba wrote: > Is there a way of downloading the whole of the edubase database so that > I do not run into the problems of searching on terms that I may not get > quite right? http://www.education.gov.uk/edubase/edubasealldata20160119.csv 36Mb, and a little slow downloading, but has all the data and seems to be updated regularly even if it still has old website url's and other content ;) -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Quarterly Project : Schools - Wales data
On 19/01/16 15:55, Bogus Zaba wrote: >>> Is there a way of downloading the whole of the edubase database so that >>> >> I do not run into the problems of searching on terms that I may not get >>> >> quite right? >> > http://www.education.gov.uk/edubase/edubasealldata20160119.csv >> > 36Mb, and a little slow downloading, but has all the data and seems to >> > be updated regularly even if it still has old website url's and other >> > content ;) >> > > Lester - > > Thanks for this, but I think this link is not fully functional right > now. Firefox, Chromium and a simple > wget command all give me a truncated file about 6.6MB. Needless to say > it's missing the > Welsh Council areas. > > I'll try again tomorrow. I guess it's updated daily with the new date > put into the filename / URL? I just download with Firefox, opened in LibreOffice and filtered the Worcs/WR* entries to give me a working list ;) Nice to not have to jump through hoops to get a list. If you get stuck trying again, shout and I'll drop you a welsh filtered version. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Quarterly Project : Schools - New edit tracker tool
On 18/01/16 13:13, Dave F. wrote: > On 18/01/2016 13:00, Lester Caine wrote: >> On 17/01/16 20:35, Harry Wood wrote: >>> I made a new tracker tool for this school mapping project: >>> >>> http://harrywood.dev.openstreetmap.org/diffreader/schools >> Can I make an appeal for those adding the edubase reference and checking >> the address details with http://www.education.gov.uk/edubase/ can you >> add the website to the school so we have all the contact details for >> later developments. Since it's on the same page as the primary details >> it's just a couple more clicks to copy. > > I've notice a couple in my area with slightly incorrect addresses. (I > find it irritating that businesses appear obsessed with changing them, > especially sub pages). Either use a validator, such as Keep Right to > check after adding them all or if you have the time, do a web search to > ensure the current URL. Some of the schools I've just edited are more accurate than the edubase data ;) Seems since schools 'change ownership' a lot more often than one would think, the official directory can't keep up :) -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Quarterly Project : Schools - New edit tracker tool
On 17/01/16 20:35, Harry Wood wrote: > I made a new tracker tool for this school mapping project: > > http://harrywood.dev.openstreetmap.org/diffreader/schools Can I make an appeal for those adding the edubase reference and checking the address details with http://www.education.gov.uk/edubase/ can you add the website to the school so we have all the contact details for later developments. Since it's on the same page as the primary details it's just a couple more clicks to copy. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Two schools on the same site
On 17/01/16 20:34, Neil McManus wrote: > It seems I am not the only one with this problem. I too have been > contributing to the quarterly project by adding a few more schools in > West Lothian and have come across a problem where two schools share the > same building. They have a few sites where catholic and > non-denominational school share the location. > > How should I tag this in OSM? For now I put both names in the tag like > 'amenity=school' , 'name=school 1 / school 2', but this doesn't seem right. > e.g. http://www.openstreetmap.org/#map=18/55.95730/-3.46847 > After adding it with both names as above, I checked that a search for > school 1 works, but school 2 does not. I've been lucky so far in that the two or three schools can be identified even if the exact area each uses will require an on site survey. If the information is not available, then I would in yur case simply add two nodes with the name, ref:edubase, and amenity=school, inside the area tagged amenity=school ... the verifier will complain, but the facts are cleanly available and the only error reported would be 'no name on boundary amenity' ... but both schools will be listed and can be crossed off. Getting the schools to provide the fine detail would then be the follow up. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Review plan for adding 1, 164 wikidata tags in the West Midlands
On 17/01/16 18:17, Edward Betts wrote: > Wikipedia has an article about a village called Brailes, but OSM has nodes for > two villages, Upper Brailes and Lower Brailes. The matcher solves this problem > by picking the civil parish. Helps if you answer ;) This is actually one of the sites I had fun with when looking at 'missing villages', and since both Upper and Lower Brailes appear in the Local Government directory, then there needs to be two separate nodes. That wikipedia merges them is not our problem, so while both villages have to reference the same wikipedia/wikidata page, it is not the right answer especially since there are separate figures for the two villages which one would EXPECT wikidata to document? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Quarterly Project : Schools - Multiple Schools on one site
On 17/01/16 17:28, Steve Doerr wrote: > One possible approach is simply to draw each school's boundary so as to > include the shared area (i.e. overlapping). If that's not possible, then > it implies that the shared area really 'belongs' to one of the schools > and the other one merely 'borrows' it at certain times. Like Andy, I do feel that the whole area is 'amenity=school' and while the government seems to be planning to sell off any land, on the whole, the council owns the land and their schools use it as required. The shared facilities do not belong to one school or another, and certainly the ones I have personal knowledge off do not have parking areas segregated to one or other building so there is not a 'shaded area' at all? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Abbreviations in OSM and schools
On 17/01/16 20:42, Ed Loach wrote: > So, should we be using the full school name or abbreviating 'church of > england voluntary aided' to CEVA as they do on the school pullovers? > Similar questions for other variations - I've seen CE, CoE, C of E for > example. I've been changing them to match what is listed on edubase for the main name, and retaining an alt_name sometimes. > I've been leaving existing names in the main (occasionally adding VC or > VA) in the hope these are from surveys and are what are on the sign, but > some names are sourced from fhrs or even os opendata. Is the DfE name > better in these cases? The one thing I've been adding is the schools website where available. That often uses a different name to either of the format used on edubase or originally, so just what is right? Since the edubase name equates to the 'registered' name that should be taken as official, and many of the missing schools simply have a large discrepancy in names, but which any human would understand as matching ;) -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Review plan for adding 1, 164 wikidata tags in the West Midlands
On 17/01/16 18:17, Edward Betts wrote: > Wikipedia has an article about a village called Brailes, but OSM has nodes for > two villages, Upper Brailes and Lower Brailes. The matcher solves this problem > by picking the civil parish. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Review plan for adding 1, 164 wikidata tags in the West Midlands
On 17/01/16 17:06, Andy Mabbett wrote: >> Please remove the School list from this. We are currently adding the >> > edubase references to each of these, and this will replace the need for >> > an additional wikidata tag. Better to just have the one primary reference. > There is no restriction limiting us to add only one identifier, and no > harm done by adding both. In particular, the EduBase ID does not > "remove the need" to add a Wikidata ID. I've no objection once the correct objects are tagged for each school, but many of the ones being automatically tagged ALSO need updating to an alternative object. Lets get one tidy finished before then looking for other? Adding all of these just adds to the work tidying the basic objects. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Quarterly Project : Schools - Multiple Schools on one site
On 17/01/16 12:40, Dave F. wrote: > Although I'm uncertain of a perfect solution as both the entrance and > recreation ground appears to be shared in Ed's example, I find there's > usually a defining boundary around schools that are adjacent to each > other. Especially infant schools where they don't want the little ones > wandering off. Looking at the site using a website that shall not be > mentioned, it appears to use a fence & the school building itself as the > barrier. On ground conformation will, of course, be required. Situations where a school has a secure play area which is used by Nursary and first school pupils at different times is not unusual, especially now the 'Nursery' provision for younger children is being added around the country. Ideally for us this would just extend the range of an existing school, but there seems to be financial advantages in creating a separate 'school'? Yes closer inspection may produce different results, but to get the key data in now would be nice, and it can be refined later? > As mapped ATM both the fhrs:id & ref:edubase tags aren't associated with > amenity=school which is not ideal for filtering data. Proper quoting would have included this comment in with mine about whether amenity=school was appropriate on the outer boundary when it is difficult to separate multiple edubase refs inside the area. Just as there are a number of ways off adding 'school' to an item, there may be a case for 'landuse=school' where one is then going to add 'amenity=school' to the internal elements? Be that simple nodes for each occupant of a high rise building, or the primary building of each where several other buildings and play areas are shared during the day. For filtering data I think that 'amenity=school' makes sense when linked with all the primary data for each school, which ever country is looked at, so some means of identifying the landuse for a multiple school area is the logical follow through. I'm very tempted at the moment to simply remove the Evesham boundary 'amenity=school' tag and replace it with on on each primary building which will at least allow the current verification to cross them off the list. What ever way things are progressed, something needs to be changed. ( And in relation to mass adding wikidata tags to the CURRENT school references, this is premature since in many cases the wrong area is tagged ) -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Review plan for adding 1, 164 wikidata tags in the West Midlands
On 17/01/16 13:36, Edward Betts wrote: >> And there should be individual nodes for each village so not sure where the >> > relations come from? > The villages matching relations are the civil parishes. The Wikidata item > represents both the village and the civil parish. This happens when the > matching algorithm can't find the appropriate village node. > > I can change the code so that towns and villages only match nodes. If there is not a node for a village in the OSM data, then it needs adding. While wikipedia may return the same page for the village and the matching parish, I thought that wikidata should distinguish between a village record and a parish one? A search for a city, town, village, hamlet should return a node with the key data located at an appropriate location in the place. A search for the parish should return the boundary, but coverage around the UK for parishes is sporadic? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Quarterly Project : Schools - Multiple Schools on one site
On 17/01/16 00:08, alasd...@dunakin.me.uk wrote: > I'd like some advice please on how to tag 2 Schools that use the one site. Having just had a 3 hour session working my way across Worcestershire, I've hit the same problem in a number of sites. Either the Nursery or Middle School has an independent EduBase reference, or two or three schools share a single set of playing field and parking facilities. What I've done initially is tagged the buildings of each part with the correct name and ref:edubase tag, and not put a tag on the site boundary. This I think will throw errors on the progress page, but in some cases the sites have a public name ... such as Abbey Park Schools http://www.openstreetmap.org/changeset/36626132, but the sites in Evesham I don't think make the distinction, with in one case 5 separate edubase establishments across two sites. Not sure what to do there, but tagging the building has to be correct, and the lack of a site name is not a problem! So should not be flagged as an error. Another little niggle ... Morton Fire Service College :) Not on the edubase list, but definitely a training centre, so again should it just be ignored on the verifier? One thing that IS obvious is that we should be messaging some of the schools to get them to add local details. Something that would encourage new mappers with an interesting local target. I've put in the outline and key detail for http://www.openstreetmap.org/changeset/36625621 but it has a LOT of detail that could be added and if students get the bug, then the surrounding area is a little bare as well. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] 2016 first quarterly project:Schools
On 13/01/16 11:24, Ed Loach wrote: > Is there an easy way to find a wikidata reference (if there is one) if you > have the DfE URN (the ref:edubase value)? > I've not been adding wikidata tags so far as I don't know how to find the > value without a lot of searching. > And if there is an easy way, do we really need both ref:edubase and wikidata > tags on the same object? Where a clean identifier to third party data is available it makes sense to use it. And in this case the fact that we can freely use the edubase reference is nice. There is no need to have to bounce through some other service because of licence restrictions, so only that reference is needed. Using that to look up a mirror record in wikidata seems pointless in this case? That wikidata could return their version via a link from the base reference is their problem? I'm not planning to work on wikidata to add missing records, so making a wikidata tag a 'required' on the project seems wrong! All my local schools have a boundary tagged with amenity=school, but for speed I'm using potlatch2 rather than josm and the existing defaults on the details don't match the project guidelines. I'm a little loathed to re-tag details simply to change from one style to another especially when it then no longer displays in potlatch2, adding to which I have listed building information on some buildings as well ... We do not have any mechanism as yet to use references like this to leaver the third party data which would simplify accessing data such as website, so I am planning to duplicate that entry from the edubase site for website and phone. Just as a little funny. My wife's school had 119 pupils ... 60 boys and 60 girls :) -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] UKOSM - technical aspects
On 10/01/16 21:47, Harry Wood wrote: > I've just set up an osmuk organisation on github as a placeholder for this > stuff: > > https://github.com/osmuk > > I just found you on github and invited you Nick. Anyone else who should be a > organisation developer member, let me have your github user name. Others > (anyone) will be able to make changes via pull requests anyway. OK Guys time for me to bow out ... Seems lots of decisions have already been made and they do not work with my methods or cover my own views of what a 'British' group should support ... -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [OSM-talk] Tile Server manual build 15.10 troubleshooting
On 04/01/16 05:16, Skyler F wrote: > Nevermind, it decided to work finally yay! Skyler ... have been following your efforts and totally understand the frustration ;) I HAD all of this working on an earlier build of SUSE having jumped through all the hoops, and I thought carefully documenting every step. Many of the problems are not so much with OSM but with the way every distribution adds it's own complexity to running essentially the same tools. Currently my own attempts at restoring the service have cost a lot of unpaid time ... service apache2 reload systemctl restart apache2 or what ever esentially running /etc/init.d/apache2 restart or what ever is replacing the init.d service Even running LTS servers in order to keep things working is not always the safest way of doing things as a recent exercise in porting a PHP application has proven. The PHP5.2 builds are running stable in production, but moving several years of data over to the 5.4 version takes time, and once converted, can no long be used on the 5.2 build. But while both are still available on LTS servers, both are end of life, and moving them to 5.6 introduces another incompatible build, and PHP7 messes things further. Python has much the same problem with Python2 and 3 needing incompatible versions of code, and while many projects stick with P2, pressure to upgrade resulted in a number of broken P2 projects over Christmas. Then add in all the other languages used to produce a working OSM setup! MANY of the tools listed on the switch2osm site are no longer actively supported by their originators, so we do need an overhaul of the whole process, but a reliable base framework does not exist and the diversity between linux distributions is making that more and more difficult. Personally Ubuntu is not practical for me as all my production servers are SUSE based, and when I have tried alternatives the subtle command line differences such as restarting apache catch you out at critical times :( And apache is perhaps not the best choice for production web servers anyway ... I HAD a working system which I was hoping to make live to produce a UK style service, but currently only the OSRM layer is working fully. I'm at the pink square stage on the tile server ... and need to pay the bills so don't have time to 'start again' with a new framework. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [Talk-GB] OSM UK group - Sign up to mail lists
On 17/12/15 22:02, Rob Nickerson wrote: > Email sign up form: > Minutes: Do we really have no option but bloody Google ... ANYTHING would be better than that and it will certainly put me off if that is adopted going forward! -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [OSM-talk] Fixing Starbucks Wikipedia Tags (Was Nominatim Weakness)
On 15/12/15 18:05, Eugene Alvin Villar wrote: > On Wed, Dec 16, 2015 at 1:36 AM, Clifford Snow <cliff...@snowandsnow.us > <mailto:cliff...@snowandsnow.us>> wrote: > > My adding the wikipedia tag to the original Starbucks store... > > I disagree with this tagging. You only tag wikipedia=* if the Wikipedia > article and the OSM object refers to the same thing. The Wikipedia > article is about the company/brand and not about the original store even > though the article would certainly mention the store (as part of the > company's history). Just to expand this, a website URL should also only link to a particular branch and most corporate websites have a 'store finder' which provides that link. Certainly the wikipedia link should follow the same convention. The store link if well implemented will provide all of the extra data that some people think should be added directly to the database and the Starbucks finder is a particularly good example! -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Nominatim weakness
On 15/12/15 16:37, Sarah Hoffmann wrote: >> Do end users want to find a coffee shop local to them or one >> thousands of kilometres away just because it has an extra tag >> attached? > > I sincerely hope not. > > Given that we have a simple data issue at hand here and that > the target audience of osm.org are mappers who happen to have > the knowledge and skill to fix data issues, one would hope that > a positive feedback loop unfolds and both the bad data > and the bad search results are gone in no time. I still consider that the initial problem is that there are two different requirements and only one selection option. Results far away are a little of a red herring, and a result of WANTING to find POI's that are not local so you can move quickly to that location. THEN looking for results filtered to that location makes sense. It is not any bad data, but simply overly wide results set. Add a flag for local/global to filter on the current window and the bulk of problems go away without thinking it's the data that is bad. Make the search 'only local' and add a note to search for the area first then one can do 'holland' then 'starbucks', but of cause the 'holland' search needs to be global ;) -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Nominatim weakness
On 14/12/15 09:00, Maarten Deen wrote: > IMHO it is a programming error on the account of importance. No amount > of importance could be so great that local results get flooded and > pushed down so much in importance. > In the Netherlands there is one Starbucks I believe (Schiphol Airport) > and even standing at that location does not return it in the search. Now I can see both sides of this debate. If I am looking for POI's near by it is a different search to asking to find say 'Starbucks, Birmingham' and having the focus change to the selected result. As a new user I'm just as likely to put in my home location and select it as I am to look for something adjacent when the map is not already centred on my location. ASSUMING that the computer IS defaulting to one's physical location is a mistake. *I* use the search to find places that are not 'within range' and can then make an educated guess at wanting 'Birmingham, UK' over one of the American alternatives, and having selected Birmingham, UK, a search for Starbucks, Birmingham, UK fine tunes the results. That is not to say that options for 'near by' would not be useful, but just that 'importance' is difficult to gauge if you don't know what the user expects to see? (An the pigging change of motorway colour really does not help a new user in the UK! One can not readily distinguish the motorways around Birmingham) -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] How does one tag a business that closed and was replaced by something else?
On 14/12/15 17:18, Andrew Wiseman wrote: > In this case, two businesses closed and were replaced by a single > business that took over both buildings (they are adjoining.) How should > I tag this? I didn't see anything about closed: or former: on the wiki. If the units are individually identified simply re-tag each with the new name. While some people will say 'historic data does not belong', the change record for each building will retain the history. If you delete each and create a new single building a lot of history which IS recorded will be lost. At some point the two units may split again. I presume that both already have their own address entries anyway? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [Talk-GB] route relations type=road
On 07/12/15 12:53, Richard Fairhurst wrote: > Historically there is a designated route concept in the UK in the sense that > some countries have them, but it doesn't map to road numbers, has largely > fallen from use, and is neither signposted nor verifiable: > http://www.watsonlv.net/pdf/trunk_roads.pdf I'm not sure of the current situation, but these were the routes that were maintained and managed centrally via what is now the highways agency? Rather than being a local council responsibility? Motorways have replaced many of the original primary routes, but essentially this list is still the base for the green trunk routes? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] route relations type=road
On 07/12/15 19:20, Andrew Hain wrote: > But surely I can see no obvious harm in the presence of the relations. Also > searching the database by reference doesn’t always work, for instance not all > road segments tagged A1 in the UK are part of the road from London to > Edinburgh. >>> >> http://www.openstreetmap.org/relation/103301#map=10/51.2112/-2.5578 The question is more one of what is the point of adding another layer of complexity that has to be maintained. The sample relation here has numerous gaps and parallel elements that make it simply a unordered list of elements. If any element on that list does not currently have an A37 tag then THAT needs to be fixed, and looking up all elements that are tagged A37 should reproduce the same list, but these will be different because currently elements of the A303 for example are incorrectly related as A37. The ONLY way to correctly identify the Route North and South is to trace the traffic path for each rather than randomly linking elements which route to other paths. Every time a component of the road system gets split then the relation also needs updating, but that may not be easy to verify while the new road network needs to be correct. If I want data on 'route' like A37 then the starting point is the ref tag and traffic direction rather than a messed up list of elements. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [OSM-talk] What3words
On 30/11/15 14:28, Ian Dees wrote: > so could we maybe let this thread die? The Abu Dhabi perhaps shows a better understanding of the problems of managing addresses than the w3w project? But I've not looked to see if OSM is picking up on that initiative yet. Certainly since it is in effect a local postcode system it would be more appropriate to be using with OSM data ... especially since it is adding the missing names to many of the new roads in the area as well as giving a smartcode tag which can be converted to a rough geo location. It's not clear from the citymetric article if the smartcode goes down to individual postal addresses, or just the building? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] What3words
On 22/11/15 14:32, Colin Smale wrote: > By the way, just to be absolutely clear, I am not thinking of w3w as a > coordinate system in OSM, but as an addressing attribute similar to postcodes. On one hand, one plugs in the three word location to their app and get a coordinate which takes you approximately to where you want to be. One needs the map to find the location in the first place, so if nothing is mapped one needs a precise coordinate ... so one logs the coordinate as well? I get the idea of 'what3words' but not while it has a page for pricing! It is something that should be a free world standard and there is nothing stopping the likes of HOT providing an alternative? But when one adds proper support for 6500+ languages building something inherently based on English is perhaps not the best starting point? All the uses I am seeing for it ALSO have the coordinates so it seems somewhat contrived trying to make it commercial in the first place? The second you NEED an app to convert from one 'system' to another is there really any need to have some human readable name? Can I go to amuses.sizing.stream without an app, when I can go to uk.worcs.broadway.xxx by following signs? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] What3words
On 30/11/15 11:59, Colin Smale wrote: > I think their big attraction is the 75% (their figure) of the world that > doesn't have a functional address system. The added value in the UK is > indeed zero. In some tribal village in Africa for example where an > address might not get any better than "3rd mud hut on the left after the > group of 3 trees" the idea of giving all the dwellings a simple address > might open the world of e-commerce up to them. They will have an address > to use, and Amazon's drones will be able to find them. Maybe not today, > maybe not even tomorrow, but soon. But 'What3words' can't actually locate them ... you HAVE to convert it to the GPS coordinates. Third hut on the left at least works without needing a mobile phone :) Doing the reverse process you need an accurate GPS system to establish the coordinates before you can convert that TO a w3w title. If the mapping system is only accurate to 10mts your android drone has a selection of targets. I've just looked up my own address in the UK and depending on which map overlay I select I got four different answers, some the next door addresses. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] What3words
On 30/11/15 12:50, Colin Smale wrote: > Correct, but the accuracy issue is a weakness in lat/lon based > coordinates as well. If you use your consumer GPS or phone to find your > lat/lon, you might indeed be a long way adrift and you might get > different values on different occasions. Imagine that you were relying > on that to get your shopping delivered... Which is why you need something on the premises you are trying to access even if that is only 'the pink door' > In my example the party that needs to do the translation from w3w to > lat/lon would be Amazon, and they will probably be paying w3w for a > licence to do that. But if two adjacent doors both have amazon landing mats there is still no guarantee that the w3w tag is correct for one or other of those mats. But I expect at some point some sort of bar code cross check would evolve and the correct mat can be identified. As has already been pointed out on the discussions on these droids, it only works when there is a single story dwelling. Add 100 flats with one front door on a busy street ... but personally I think there will be more losses due to droids being intercepted in flight? :) > On 2015-11-30 13:30, Lester Caine wrote: > >> On 30/11/15 11:59, Colin Smale wrote: >>> I think their big attraction is the 75% (their figure) of the world that >>> doesn't have a functional address system. The added value in the UK is >>> indeed zero. In some tribal village in Africa for example where an >>> address might not get any better than "3rd mud hut on the left after the >>> group of 3 trees" the idea of giving all the dwellings a simple address >>> might open the world of e-commerce up to them. They will have an address >>> to use, and Amazon's drones will be able to find them. Maybe not today, >>> maybe not even tomorrow, but soon. >> >> But 'What3words' can't actually locate them ... you HAVE to convert it >> to the GPS coordinates. Third hut on the left at least works without >> needing a mobile phone :) Doing the reverse process you need an accurate >> GPS system to establish the coordinates before you can convert that TO a >> w3w title. If the mapping system is only accurate to 10mts your android >> drone has a selection of targets. I've just looked up my own address in >> the UK and depending on which map overlay I select I got four different >> answers, some the next door addresses. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] What3words
On 30/11/15 13:43, Marc Gemis wrote: > On Mon, Nov 30, 2015 at 1:50 PM, Colin Smale <colin.sm...@xs4all.nl> wrote: >> > In my example the party that needs to do the translation from w3w to >> > lat/lon >> > would be Amazon, and they will probably be paying w3w for a licence to do >> > that. > Wouldn't it be more likely that Amazon would invent their own system > where customers get a mat with some sensors or build-in GPS tracker > (so I can move and take the mat with me) that can be used by the drone > ? Why would they rely on a (broken) third-party solution ? And of cause why rely on some optical barcode marking when a wireless based tag would be better. May need to change the battery every few years, but it only need to be in receive mode until a drone is in range ... with a button to 'sync' location with base ... -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
[Talk-GB] Sync problem with british-isles.osm.pbf
I've got a copy of british-isles.osm.pbf which was updating fine until
the 1st November when it got stuck at
start import from seq-nr 896702
The osm2pgsql stage is failing with ...
> Osm2pgsql failed due to ERROR: insert_way failed: ERROR: null value in
> column "pending" violates not-null constraint
> DETAIL: Failing row contains (4254992,
> {17947723,17947722,17947721,2891767946,2891767954,2891767909,289...,
> {highway,secondary,ref,302}, null).
> (7)
> Arguments were: 4254992,
> {17947723,17947722,17947721,2891767946,2891767954,2891767909,2891767947,2891767919,2891767905,17947720,2891767938,17947719,2891767929,17947718,17947716,2891767907,2891767922,17947715,2891767944,17947714,2891767959,17947717,2891767953,17947713,2891767939,2891767910,17947712,2891767942,2891767943,17947711,2891767933,2891767937,2891767952,2891767928,17947710,2891767906,2891767936,2891767932,2891767913,2891767958,2891767924,17947709,2891767925,2891767918,2891767920,2892714782,2891767940,17947708,2892714719,2891767948,2891767950,2891767917,2892713977,17947707,2891767955,2891767926,2891767927,2891767908,17947706,2891767916,2891767957,2892712501,2891767956,17947705,2891767935,2892710491,17947704,2891797509,2891767930,2891767923,17947703,2891767911,17947702,2891767934,17947701,2891767931,2891767921,17947700,2891797521,2891767945,2891769852,2891767912,2891767915,2891767941,2891790382,17947699,2891767914,2891767949,2891767951,17947698,17947697,2891721239,17947695,1794
7694,17947693,17947692,17947691,17947689,309822592,17947688,17947687,17947686,17947685,17947684,17947683,17947682,17947681,17947680,17947679,17947678,17947677,17947676,2218967592,17947675,1851431147,17947674,17947673,17947672,17947671,17947670,17947668,17947667,17947666,17947665,17947664,17947663,17947662,1851431102,2890090823,17947661,17947660,17947659,17947658,17947657,17947656,17947655,17947654,1861986735,1861986730,17947653},
{"highway","secondary","ref","302"},
And now I'm stuck as to how to get it past the problem?
Is it a case of wiping the existing copy and downloading a clean copy,
or is there something to fix the update stream?
--
Lester Caine - G8HFL
-
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk
___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb
Re: [OSM-talk] highway=residential_link
On 10/11/15 09:40, Martin Koppenhoefer wrote: > 2015-11-10 9:36 GMT+01:00 Mateusz Konieczny <matkoni...@gmail.com > <mailto:matkoni...@gmail.com>>: > > In theory, it is possible that former motorway/.../tertiary with slip > roads/ramps was converted to residential road, without changing road > infrastructure and traffic is still grade-separated. > > sounds not very likely. At the point where a motorway or even tertiary > road becomes a residential road, it will clearly change a lot (or > otherwise it would remain a tertiary etc. road). A residential link > would be a road between 2 residential roads (or a residential and an > inferior road like a service), because if it were a higher road class to > which the residential connects, the link would be called after that road > class. > > There's not really a problem calling a road that connects 2 residential > roads a residential road as well. > > Examples /(also covering unclassified and even tertiary): > http://www.openstreetmap.org/way/35022603 > http://www.openstreetmap.org/way/24038678 > http://www.openstreetmap.org/way/27854328 The third example is the sort of link I would expect to see approaching a motorway where restricted traffic gets redirected, but I think that as in this case, the lowest level would be 'unclassified' rather than residential or service. Of cause in the UK most roads ARE classified in some way, so the residential or service is simply a way of avoiding having to look at 'landuse=residential' or 'retail/industrial' as these are all still link roads between other higher classification routes? So this is just a matter of agreeing terminology, and while 'tertiary_link' may well pushing the limits still, any other _link not listed on http://wiki.openstreetmap.org/wiki/Highway_link should be re-tagged. The best description of how a link should work is probably when creating routing instructions. What still grates with me is the 'Bare slightly left onto M5' when the instruction should be 'Take slip road onto Axx' and that in my book is where the _link information has an important element but could also be a second tag. However like 'highway=residential', 'highway=trunk_link' is a shorthand that works at many levels ... but should the link be tagged for the road it provides the link to rather than the one it obviously comes from? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Undiscussed (?) edits removing lesser-used highway=* tags
On 10/11/15 10:08, Martin Koppenhoefer wrote: > One of the great strengths of OSM is that you can invent tagging on the > fly and trying to suppress that just so that the data consumers have it > easy, is misguided. In the end the main way our tagging evolves is be > contributors trying to map stuff that doesn't have a popular tagging > scheme associated and not allowing that will reduce new tagging to that > decided by a committee. > > +1, very well put. I have in the past seen this more than once: invented > a new tag to try something out, someone else comes along (typically a > remote mapper with no knowledge whatsoever about the place) and tries to > "normalize" the new tag into something common (but not applicable). Even > fixing "typos" has to be done very carefully and hesitant, limited to > actual typos like "highway=residental" and not extending to assumed > synonyms. Plus and minus on that ... Yes adding tags has to be flexible, but there are still a few areas that we need perhaps a tighter control, and adding highway types that do not then get displayed is probably one? As with the current discussion, key elements need a few ground rules, but that does not prevent additional data being added via extra tags. The rule about not tagging for renderer or router works both ways since one may need to add information that can't easily be accessed or simply does not exist on the base tagging? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [Talk-GB] Secondary, tertiary and unclassified
On 04/11/15 21:49, Mark Goodge wrote: > If a road can be used for through routing then it probably is tertiary > rather than unclassified[1]. That's a reasonable rule of thumb for > making the decision, in the absence of more reliable information. But > secondary is very well defined, and is part of published open data. > There's never going to be a reason to tag a road as secondary that isn't > already known to be. > > [1] This is one of the reasons I don't like the new style, with white > for all smaller roads. The difference between tertiary and > residential/unclassfied is quite significant in UK road topology, and > needs a clearer distinction than just the width of a digital brush > stroke. But that's a different issue. That some 'B' roads now need re-classifying and the alternate currently unclassified roads need upgrading on the council records is the distinction here. Routes through residential areas which are currently classified are being restricted with traffic calming, 20MPH limits or even barriers and alternate routes being improved, but the only documentation on that in the absence of new maps is the minutes of council meetings rather than as yet, the NSG or other council documents. Only primary and main routes get well documented re-designation currently :( In this case we do have to map what is on the ground rather than trying to justify the 'legal status' of those roads. Now if the three equally important secondary idents are treated with the same level of importance, then the addition of other restrictions will make more sense, rather than promoting a 20MPH 'B' road over a 60MPH 'unclassified' ... which the current rendering is doing? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [OSM-talk] New OSM Carto: White Roads & Road Widths
On 04/11/15 12:08, Pieren wrote: > On Tue, Nov 3, 2015 at 5:58 PM, Martin Koppenhoefer > <dieterdre...@gmail.com> wrote: >> > tertiaries are still rendered, they just don't have a different colour >> > than other roads (but they are thicker). In German we say: remain on the >> > carpet ;-) > Well, I would say in English "you play with words". It's not visible > as a specific class. And for the minority who will see it's thicker, > they will think it's based on a physical attribute (width or lanes), > not on its importance. The simple answer is secondary, tertiary and unclassified roads in many areas of the world have the same importance, so rendering them drastically differently is a mistake! -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
[Talk-GB] Secondary, tertiary and unclassified
OK finally spotted what is going thanks to the new style sheet ;) The question is where do we get the 'tertiary' designation from since in many cases there is little to distinguish those roads from 'unclassified'. Both are potentially good quality national speed limit routes and in my case they are regular elements of the routes I take out to the motorway or other primary route and the level of lorry traffic probably confirms they are essential secondary roads. I have been here before and re-tagged roads in the past to remove 'residential' or 'service' from secondary routes. This is not so much 'tagging for rendering or routng', but rather maintaining continuity of traffic routes. Now 'unclassified' has to be included in that grouping so that the secondary routes at least appear even if now in white. For some reason the particular roads I'm looking at are not actually appearing on the French rendering and I STILL have not managed to get tilemill or now kosmtik to render my own style sheet for a comparison. What is bloody annoying here is that tilemill was working nicely a month ago :( Now it's throwing errors when trying to reinstall, and while kosmtik is just throwing deprecated warnings on maxzoom it's not actually producing tiles ... -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [OSM-talk] New OSM Carto: White Roads & Road Widths
On 04/11/15 01:06, Dave F. wrote: > > Hmm... Interesting; I notice nearby that motorway & links are now > rendered as I suggested: > > http://bl.ocks.org/tyrasd/raw/6164696/#18.00/1.29498/103.87800 > > However it doesn't for trunk_link: > http://www.openstreetmap.org/#map=18/51.39961/-2.33758 > > Maybe it should? (I haven't test the other *_links) That is a part of the jigsaw of why the rendering around here is looking 'messy'. The motorways are at least rendering with a central reservation, while the primary routes are loosing that boundary. On smaller devices the problem is worse than on high resolution devices. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [Talk-GB] Secondary, tertiary and unclassified
On 04/11/15 13:31, Andy Townsend wrote: > No - the "rules" for trunk, primary and secondary are as spelt out by > Tom Hughes, and have worked well. There are minor exceptions where the > official classification hasn't caught up (e.g. something that's > "officially" an A road that you can't drive down most of the time) but > let's try not to make things more complicated than they need to be. There are no problems with that ... it is just the whole secondary layer that is in a mess at the moment, with half of the roads difficult to see on both the new rendering and on the Mapquest version. That many 'B' roads are of lower capacity than some of the unclassified rural roads is the problem I'm trying to work around ... but need to work out the licensing on http://www.worcestershire.gov.uk/downloads/download/924/searches_and_adopted_roads_classified_and_unclassified_roads next :) -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Secondary, tertiary and unclassified
On 04/11/15 11:48, Robert Whittaker (OSM lists) wrote: > On 4 November 2015 at 10:25, Mark Goodge <m...@good-stuff.co.uk> wrote: >> > Unfortunately, the NSG is not Open Data, so it isn't available to OSM by >> > default. The nearest we can do is attempt to visually classify by >> > observation. That's one of the weaknesses of a crowd-sourced approach., If >> > that matters to you, you might want to get involved in the campaign to get >> > the NSG released as Open Data. > I fear the NSG will probably never be open data, as it's based on OS > base maps, and replicates a large quantity of key OS data (i,e. the > road network). However, that doesn't necessarily mean that the > classification information is inaccessible to OSM. Under the Highways > Act, each highway authority (usually a County Council or Unitary > Authority) is required to maintain a "list of highways maintainable at > public expense" commonly known as the "List of Streets", and make it > available for public inspection. Any rights in this information are > likely to rest solely with the highway authority, so they should be > able to give permission to use it under e.g. the Open Government > Licence, which would then allow its use in OSM. > > I have some notes about requesting and getting permission to use these > (and some other related) documents, which people may find useful: > http://robert.mathmos.net/osm/prow/council-docs.htmlt Since the source of the material for both NSG and NLPG is a legal requirement on the various local councils, giving that information to a commercial operation is fundamentally wrong, so all should be freely available, and there is no reason that each council could not provide the material on a free licence prior to it's delivery to geoplaces and then we can all benefit from it. I've campaigned for that for all the time that that material has been provided to me for use in our own systems. It is only when graphic elements are added that OS could claim some IP, but the coordinates provided work equally well with OSM data, and we can use OSM graphics to eliminate the OS tie-in. The point I was trying to make was that Secondary, tertiary and unclassified are essentially the same level of importance for road navigation and so treating them differently in rendering ( or routing rules ) adds an incorrect importance to one over the other. In the absence of any other evidence I'm planning to simply re-tag the problem unclassified routes as tertiary for now, but I can make a case for all being secondary so they get rendered with the same separation from the sections of the road system that should not be used for through routing. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [OSRM-talk] Demo Server Down
On 02/11/15 19:34, Patrick Niklaus wrote:
> We are not sure yet what is causing the problems with the current
> dataset. I'm currently digging through the changesets.
http://osrm.rdm2.co.uk/ is up and running again. Looks like the problem
on mine was an SUSE update that messed something up, but not sure what.
However osm2pgsql had a problem with
Osm2pgsql failed due to ERROR: insert_way failed: ERROR: null value in
column "pending" violates not-null constraint
DETAIL: Failing row contains (4254992,
{17947723,17947722,17947721,2891767946,2891767954,2891767909,289...,
{highway,secondary,ref,302}, null).
(7)
However it does not seem to worry osrm.
Now to work out why tilemill has stopped working :(
--
Lester Caine - G8HFL
-
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk
___
OSRM-talk mailing list
OSRM-talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSM-talk] New OSM Carto: White Roads & Road Widths
On 03/11/15 14:39, Christoph Hormann wrote: > Note ultimately design decisions are always subjective. If at the end > of a long process a decision is made without once more listing in depth > all arguments for and against it this does not mean the decision is > arbitrary. A 'long process' carried out in a small area of discussion is proving problematic for the wider user base. This is one area that once exposed *IS* proving to be the wrong subjective decision followed closely by the changes to road width. The world is a large area with a large range of data spread, so trials may well have missed numerous examples of where something that looks good for a few limited examples fails in the wider coverage. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSRM-talk] Demo Server Down
On 02/11/15 15:14, Patrick Niklaus wrote: > What version are you running? Our production servers are definitely > working with the current OSM extracts. My update process had stopped which is why I was asking if your problem was related to something similar or a local hardware problem? I do need to pull in a new version of software, but the current style change is taking all my spare time at the moment. > On Sun, Nov 1, 2015 at 9:53 PM, Lester Caine <les...@lsces.co.uk> wrote: >> > On 01/11/15 20:42, Patrick Niklaus wrote: >>> >> The data on the demo server seems to have gotten corrupted. We are >>> >> currently reverting back to an older dataset, while we investigate >>> >> what went wrong. Should be all up and running in a few hours. >> > >> > My own service seems to have broken so was there anything in particular >> > to explain the problem? A later data update gone wrong perhaps? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [Talk-GB] [OSM-dev] New Map Style feedback
On 02/11/15 12:49, Marc Gemis wrote: > On Mon, Nov 2, 2015 at 12:57 PM, Lester Caine <les...@lsces.co.uk > <mailto:les...@lsces.co.uk>> wrote: > > Now that I've got an easy toggle between the new style and the French > tile server a few more backwards steps become apparent. The other WTF > > Don't you have a problem with all names appearing in French ? The French > tile server shows Londres... That is not a problem with the local embedded elements, but I don't have anything suitable to replace the main osm site as the 'large view' which has been my main promotional tool up until now. Switching to the Mapquest layer is better than nothing, but that has stripped most of the detail that is the whole point of using osm in the first place :( I've got a local display of my own style, but as yet not got that working as a cached tile server on the production servers :( -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [OSM-talk] Improved new Icon set for Open Street map
On 02/11/15 18:08, nebulon42 wrote: > However, there is of course always the possibility to improve things > further. I'm welcoming contributions at https://github.com/gmgeo/osmic, > which is a repository that holds all icons I have done for the Standard > map style (some I have also adapted from Maki and other free sources). > You can also directly contribute to osm-carto at > https://github.com/gravitystorm/openstreetmap-carto. As a dinosaur who actually liked full colour icons on websites, I've my own version of bootstrap using full colour ones, and I'm now looking to extend that into the icon map for my own rendering. While the styling for OSM is quite complicated it can easily be broken down into manageable elements such as road colours, or icon library, so switching elements is not particularly difficult once one has a working local rendering system. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [Talk-GB] [OSM-dev] New Map Style feedback
On 02/11/15 10:56, Andy Robinson wrote: > I agree that the abandonging of the blue for motorways is a bad choice. > It is not only a british color, motorways are signalled in blue also in lots > of other countries in europe. > But that is not really the issue. It is not that a colorscheme should follow > the colorscheme of a particular country per se. The current color scheme just > makes it hard to distinguish roads. Teritary roads, being white, are all but > unrecognizable. Looking at motorways, trunk roads or primary roads, I can not > tell one from the other, except when I see two next to eachother. > Furthermore, on high zooms, roads have gotten too fat. It makes the map look > bulky. > > The colorscheme for roads is defintely a step back from the previous. Now that I've got an easy toggle between the new style and the French tile server a few more backwards steps become apparent. The other WTF I've just hit is the excessive stripping of place names at some zoom levels? Simplifying the map is one thing, but having to toggle in and out of levels to see detail is a problem. Google and Bing are less cluttered because they don't have the same fine detail that OSM has access to and while we can't display everything, but the new stripping seems to me to be a little excessive? My main complaint is one that applied to the old style as well and that is that the general drift over time to more 'pastel' colouring. The new style is a further loss of contrast! That is why I've been looking to a style that restores a more usable map for the visually impaired as demonstrated by the sample I provided earlier. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] New map style
On 02/11/15 10:02, Paul Sladen wrote: > On Sun, 1 Nov 2015, Lester Caine wrote: >> On 01/11/15 07:22, Ed Loach wrote: >>>> It very simple, the colours should match the road sign colours: Blue, >>> Green, Red! >>> Red? >> all the legislation, > > ...stems from one document: > > Geneva Convention on Road Signs and Signals > Part I > Annex 1 > Section G > Sub-section I > Paragraph 3 > > "3. Advance direction signs or direction signs relating to motorways > or roads treated as motorways shall bear white symbols or inscriptions > on a blue or green ground." > > > http://www.unece.org/fileadmin/DAM/trans/conventn/Conv_road_signs_2006v_EN.pdf > (PDF Page 65). As with all 'conventions' ... plenty of wriggle room :) And plenty of abstentions such as the US ... >> Anybody still got signs with red backgrounds in their area? > > "2. Informative signs ... the colour red may be used only > exceptionally and must never predominate." The question was more one of exceptions too the rule, or rather any old signs still not been replaced ;) -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] New map style
On 01/11/15 07:22, Ed Loach wrote: >> It very simple, the colours should match the road sign colours: Blue, > Green, Red! > > Red? Was waiting for someone to pick that one up, and yes it has been some time since red was dropped from the legal framework and therefore the highway code. But the Blue and Green are well documented and just what traffic is restricted from accessing a motorway drummed into people. It would be interesting to find out if our French colleagues have any plans to switch their servers to the new style, but I expect they will be a lot more considerate! The default style they provide is actually a better one for the UK than the 'old' style was (wish I'd found it sooner!), but along with a few useful variations related to France BOTH are available. So I would anticipate that the new style will simply become an option there? Back to the 'Red' question, and the simplification introduced between Primary and non-Primary routes. A quick search on google produces no easy answers, and Wikipedia has references to all the legislation, but many of the links to VIEW the facts no longer work. It's this disregard for maintaining history that annoys me most. The bottom line is that 'non-primary' routes are any road used to link primary routes, and INCLUDES tertiary routes in many rural areas. Apart from the way the the style suddenly appeared rather than a proper roll out, my only complaint about the new style is that tertiary routes are not included in the 'non-primary' grouping. ADD that to the orange routes and it will fix that particular bug. As for the new style ... no I don't find it particularly useful at all. The use of red, orange and yellow to rate the non-primary routes is really only a matter of following the OS conventions. It is one of the areas that I have actually adjusted in my own clone of the style and is a little different on the French version. p.s. - Anybody still got signs with red backgrounds in their area? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Restoring a usable map service!
On 31/10/15 21:51, Frederik Ramm wrote: > As for the remainder of your posting, about a "single base" and how all > this is somehow related to a few colours changing on the osm.org map, I > think you lost me there. Answered the colour problem in the thread on New Map Style. The problem with setting up a tile server is the GENERATION of the tiles and how that can be modified to provide the other element I've been banging on about. Viewing the UK at a point in time rather than simply providing what is the current often incorrect view of the country. (Roads around Coventry are still wrong while the older views were actually better). I switched from Apache to Nginx on the servers for the performance improvements it gives working with the PHP sites I manage, and it SHOULD be simple to get Nginx to front access to both the tile server and the editing tools for playing with the style sheets. It proxies everything else happily enough and serves all the PHP static material directly. I got to http://lsces.co.uk/storage/attachments/71/2071/osm-lsces-z11.png but I need to pull the nice county boundary twiddle from the French style sheet now and work out why some mass 'farm' areas still display incorrectly. That one is a tagging problem in the data! http://lsces.co.uk/wiki/OSM+Development was where I was back in August, but I've not been able to get back to that since, and the current fire fighting will add further to that delay :( -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [OSM-talk] Release openstreetmap-carto v2.36.0
On 01/11/15 12:39, Max wrote: >> not long ago (maybe even today?) pharmacies were not only selling goods but >> > also producing many kinds of ointments and possibly other things as >> > ordered >> > by the prescribing physician. >> > >> > So "shop" would be a too narrow definition. > With that argument you should also change stuff like shop=florist, > because they aren't just selling the flowers they buy in bulk, they > actually make arrangements and bouquets and stuff... The irritating bit is the miss match of things like amenity, shop and now 'healthcare'. In my tag list the local doctors surgery is an building=clinic and has within it an amenity=doctor, nurse, dentist and pharmacy. Just as my local hospital has various additional amenities such as podiatrist (who CAN be a medical practitioner here!), optometrist and the like. But just down the road from the doctors is a shop=chemist which also provides a dispensing pharmacy amenity and has a visiting podiatrist. The doctors pharmacy carries out no over the counter sales ( apart from the extra tax con ;) ) so is an amenity, while the chemist is clearly a shop providing various amenities, some of which may not be 'healthcare' ... do we really need yet another complicated tag in this mix? Adding 'office' as another option of "A place predominantly selling services." also confuses the picture. The building is an office, shopping_mall or retail premise, and once again, the amenity provided is the job function. The estate agent is next to the chemist's shop in the same building, so tag the building 'retail', and each unit with is amenity(s)? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [Talk-GB] Restoring a usable map service!
On 31/10/15 07:39, Ed Loach wrote: >> Bing and Google are almost useless, and the new indistinguishable >> > rendering primary routes and motorways on OSM once everything >> > finally >> > re-renders are going to be useless as well. > As I just posted on talk, I was attached to the old colour scheme and didn't > want it to change for change's sake, but now it has I can see it is visually > much clearer than the old colour scheme and I have no problem distinguishing > between primary routes, trunk routes and motorways. > > I was particularly concerned about tertiary going white, but even those are > significantly different to residential and unclassified that they stand out > enough for me to feel it is an improvement. I'll provide visual examples of the problems that the new colour scheme creates in rural areas. It's the same with Bing and Google, but osm's 'old' style while not perfect did at least allow the tertiary roads which around this neck of the woods provide the key fast routes to be distinguished. I worked through to 5:30am and got a replacement for the embed facility working into the French tile server and I can replace that with my own server once it's updating reliably. The French tiles have provided a couple of bonuses so on my own roadmap are working in the right direction. http://goodflexrubber.com/wiki/Home+Block+10 is 90% there Not quite sure what Bing is up to. get the idea of the spooky stile for halloween but the purple colouring for roads? Anyway ... got to get out to site to fix another bloody sick M$ machine :( -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Restoring a usable map service!
On 31/10/15 15:18, Richard Fairhurst wrote: >> while switch2osm may well produce a working system for >> > some ... I have to also support current paying traffic on the >> > hardware and that prevents running too many different >> > competing web services. > You can run a tileserver for the UK on a £10/month virtual machine. If your > paying traffic can't support £10/month across all your clients, and instead > you have to rely on a third-party server operated by a non-profit > organisation, then there's something wrong with your business model. My current business model is to provide a UK based service which I can share freely with others and provides both current and historic material for the UK. What is preventing this happening is all the disjointed elements that are being used currently between OSM, OHM and other services. Up until now my time has been spent trying to make that come to fruition while still trying to add raw data and updates and trying to ensure current material is not simply dumped when new history is created. The current changes to OSM has changed priorities so what I'm trying to achieve HERE is some help to get the alternative service running ASAP while in the meantime fire fighting once again the support system :( No I don't have a bottomless pit of money to pay for services, so I have to compensate by the best use of what resources are available, and I can provide hardware and use my spare bandwidth to do that. I bought the machine for this job some months back but as yet it is still only partially working ... now if others are working to the same end can we not pool that resource? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Restoring a usable map service!
On 31/10/15 19:51, Frederik Ramm wrote: >> It's the constant moaning whenever something changes that annoys me >> > most. All the effort that volunteers put in to create and improve the >> > style, to run the servers, to write the blog posts and announcements >> > and all we get from Lester is moan moan moan. I've had enough, to be >> > frank. > I'd like to broaden this a bit - it's not only about map styles, and not > only about Lester's complaint. My only complaint is that many decisions are still unilateral rather than considered. Personally I could be contributing a lot more to the software if it was using main stream tools rather than some of the exotic elements that seem to be the in thing because of someone's pet interest. But that is a problem across a whole gamete of open source projects these days. We are now stuck with some elements which are less than ideal and have to now work around them, so perhaps it is time to isolate the data even further from the graphical tools and make it easier to use with all tools? How much bigger would the programmer base have been if a different base had been adopted at the start. It's problematic enough adding python, perl and java to a base of C/C++, PHP, javascript, css and html without adding Ruby into the mix, and the spiralling range of tools and libraries 'improving' javascript, css and the essential responcive browser interfaces is spreading everything thinner and thinner. I'm looking at the OHM rendering problem, and Kartotherian is now being pushed from that side and I can't see why we have to be using different bases to do exactly the same job? Why can't we have a single base? SO I repeat what I said earlier? Is there ANYBODY who has an interest in creating a simple generic system for creating material across the whole spread of data that we have to work with. Initially targeting the vast range of UK current and historic data. A vector based rendering system is probably the right way to go and the OSMAND rendering works well and as far as I can tell ALLOWS a selectable rendering style? But that is yet another tangent and ring fenced solution :( -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Restoring a usable map service!
On 31/10/15 09:27, Matthijs Melissen wrote: >> new indistinguishable rendering primary routes and motorways > For my information, could you take the test on > http://enchroma.com/test/instructions/ and tell me what it says? It's > ok if you prefer to respond off-list. I have Normal 'Color' Vision even though I do now ware glasses. But I work with a number of clients who support visually impaired or people with other disabilities and we do take a great deal of care to get flexible control of both size and contrast of material being displayed. My urgent problem is the one already described ... people with learning difficulties come to rely on stability and when things change to can cause panic attacks. YES I need to remove the problem by building my own servers, but this needs to be a consistent national resource. The simple addition of a 'classic' default which can be restored is often all that is required, but when the whole system keeps changing apparently from day to day it can be frightening to some people, or makes safe usage difficult if as is the case with both windows and android, the sequence of operations a visually impaired user is used to suddenly stops working! They often can't 'see' the fix. That is one of the major brick walls with M$ ramming W10 onto machines without our being able to stop it and retain the original W7 classic desktop. Often *WE* have no control over updates :( Just something simple like the current mess with scroll bars on applications where some have both up and down button at the bottom, or the modern style of hiding the thing off the side, added to the apparently totally random way some page up and down when clicked outside the bar and others just jump to that location. All of this makes for a totally confusing mess. When the renderers finally catch up with all the tiles in the areas I want to use as a demonstration of the problems the new style is creating in several of the rural areas I support then I'll publish a proper document, but while switch2osm may well produce a working system for some ... I have to also support current paying traffic on the hardware and that prevents running too many different competing web services. The MAIN problem which both Bing and Google maaps hve always had and which many routing services also mess up is that around here many of the main roads are 'tertiary' but they are still 60MPH speed limit, and it would probably actually be better to dodge the problem by re-clasifying them as secondary which has actually be done - incorrectly in some places! The whole UK system is essentially SIX levels with motorways, primary trunk roads, main roads, secondary classified roads, secondary unclassified roads and the whole residential and service layer. The B roads need to be clearly separated from the rest of the lower level structure, but so do the tertiary roads and these are now slowly vanishing from the new rendering :( I have now established where some routing problems arise, and this is down to segments of a tertiary route being tagged as 'minor' which certainly on OSMAND used to add a block to the use of that route. A combination of fixing the tagging, and the better defaults in OSMAND does seem to be improving the situation, but IDEALLY what one is using on the sat nav in the car should match display wise what one has reviewed on the desktop prior to planning the route ... so in addition to stabilising the broadband service, the off-line services need to be kept in sync, along with the editors ... and so it goes on :( -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [OSM-talk] Release openstreetmap-carto v2.36.0
On 30/10/15 21:01, Matthijs Melissen wrote: > * Major rewrite of road and railway rendering, as part of Mateusz > Konieczny's Google Summer of Code project. See BUGGER I still haven’t got the backup system working and now all my UK stuff is screwed up :( This really does not work! -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [Talk-GB] Restoring a usable map service!
On 31/10/15 02:21, Steve Doerr wrote: > Lester, > > In the short term, would substituting http://tile.openstreetmap.fr/ > solve your rendering problems? Interesting thought ... that actually solves one of the todo items on my own style sheet tweaks in that the French version actually shows the county boundaries a lot clearer, but switching to a large view which has French attributions and overlays may simply add to the confusion :( Having sold clients on the idea that we can show all the right parking and access around their premises on a UK style map, many have been helping by providing finer detail and corrections in their areas. I KNOW the phone will be ringing Monday if not before. In many of their areas Google and Bing are simply white blanks! OK ... where do I get a copy of the osmfr layer style sheet :) That is rendering the area around here a lot nicer than the 'defa...@osm.org' ... everything is a lot crisper. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
[Talk-GB] Restoring a usable map service!
OK ... what do we need to do to get a working UK map again? Bing and Google are almost useless, and the new indistinguishable rendering primary routes and motorways on OSM once everything finally re-renders are going to be useless as well. I've just spent another couple of hours trying to remember where I got to with the machine I've been trying to build here to provide just the UK and Ireland, but I'm going to have to go back to first principals and ignore the Nginx layer and go back to slow old Apache for the tile cache :( I know why the current tile servers can't handle more than one style, and in my book THAT was the first problem that needed fixing rather than forcing a complete new style sheet with no option to retain the current style in parallel. So what do we need to do to get a UK friendly service working? I would like to throw into the mix that it would make sense to me if a complete new rendering service also handled historic mapping as well which should be relatively easy to do if confined to just the UK, but I think we may need to be open to rendering more than just the UK? It's bad enough at the moment tying to cope with W10 being rammed into my clients systems and breaking perfectly good operational hardware, android apps changing various operating procedures in illogical ways, and the various 'browser improvements' resulting in even more legacy sites no longer working without also having to start reworking every map used across my client base. OSM is the only source that has much of the detail that I need to show, and retaining blue ad green road links is essential in my book ... and one of the reasons I started even developing with OSM. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [OSM-talk] Map key
On 17/10/15 09:03, nebulon42 wrote: > I don't know an ETA of this change on the tile servers yet as it is a > quite major change. Others may be able to comment on that. Is this ACTUALLY a 'done deal'? I though there was a going to be at least some option to ask if there was a more general agreement for the change. But I have no doubt other sources will become live if the change goes ahead with their own style of map key ... -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [Talk-GB] [Talk-gb-westmidlands] New St station platform alignments
On 25/09/15 22:48, Graham Jones wrote: > I'm pretty sure there have been no changes to the tracks or major > building structures at station platform level - I didn't see any > evidence of heavy work like that on my travels anyway. http://maps.nls.uk/view/101584678 The later maps just have black boxes over the station area. I know the turntable area was cleared to put the electrification gear in, but the British Rail line books should clarify if there were other changes. It looks like extra tracks have been added through the platform area? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] [Talk-gb-westmidlands] New St station platform alignments
On 25/09/15 23:22, Lester Caine wrote: > On 25/09/15 22:48, Graham Jones wrote: >> I'm pretty sure there have been no changes to the tracks or major >> building structures at station platform level - I didn't see any >> evidence of heavy work like that on my travels anyway. > > http://maps.nls.uk/view/101584678 > The later maps just have black boxes over the station area. > I know the turntable area was cleared to put the electrification gear > in, but the British Rail line books should clarify if there were other > changes. It looks like extra tracks have been added through the platform > area? Seems the NetworkRail websites are as reliable as the train service ;) http://www.networkrail.co.uk/searchresult.aspx?q=virtualarchive lists https://www.networkrail.co.uk/virtualarchive/new-street/?cd=7 which should have a history of the changes at track level but something is broken :( -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [OSM-talk] recent changes in rendering the map make it worse
On 23/09/15 13:59, Andy Townsend wrote: > To add to that, creating "a personal map style" of a small area using > https://switch2osm.org/serving-tiles/manually-building-a-tile-server-14-04/ > whilst not for everyone, isn't that difficult these days, and it doesn't > require huge server resources. There are also canned options such as > Mapbox of course if you want someone to host it for you. It'd be nice > if there was a way to see the tiles on osm.org without resorting to > cludges such as > http://wiki.openstreetmap.org/wiki/User:SomeoneElse/Your_tiles_from_osm.org > though. I'm STILL struggling to get a private tile server working just covering the UK. I DO have my own tweaked version of the style shhet running via Tilemil, but while the database SHOULD be updating, I'm not seeing changes I made rendered on the local copy, and I still need to get a proper tile cache working. But what I do have working shows in house rendering is not at all difficult. Just need to nail all the cludges in a simple to roll package of tools? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [Talk-GB] Anybody a member of CAMRA?
On 16/09/15 17:36, Dave F. wrote: > I was trying to find out how many GB pubs/bars there were in OSM. Didn't > Taginfo used to have a option to just search the UK? It appears to have > been removed. Is there another way? I'm looking for the same sort of thing myself ... parishes in county and the like, but pub is equally relevant ... -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [OSM-talk] THIS is the kind of enthusiasm some would reject
On 15/09/15 08:42, Martin Koppenhoefer wrote: >> If it wasn't clear already, railway=dismantled, end_date, or any >> > system that mixes past and present in the same namespace is IMHO not >> > acceptable. > > I agree that end_date is not a desirable way to add stuff. > > railway=dismantled on the other hand is not a past feature, it is a > dismantled railway now, in the present. In the past it was a railway=rail etc. The crux of the problem here is 'end_date' and if it is to be supported or not. I'm perfectly happy that features which exist on the ground need to be documented, and even having removed the tracks, a rail bed is still a substantial structure which can be reused or robbed out. The use of the name 'Abandoned Railway' on a cycle track is an alternate compromise, so it is just breaks which we are discussion here. If OHM WAS usable as an alternate source of data in parallel with the main database I would not have a problem, but the discussions there are at a tangent to the main problem of retaining material that has been accurately mapped already and for which 'end_date' is the perfect tag. The difficulty here is distinguishing data that has been deleted or changed because it was simply wrong and changes that result from re-tasking or redeveloping areas. I've already given a good practical example in the case of Tollbar improvements which are still work in progress, but the current data is not as accurate as it could be because elements are mapped that do not yet exist and elements which have been realigned do not actually follow the current state on the ground. The full scheme started life back in 1998 and the various phases were well documented showing what was to be added and deleted at each stage, so careful mapping taking note of both start and end dates COULD have done a better job, and this IS about mapping the present! It shows what is planned so drivers know of the disruption and while dates will change updating them builds the very history *I* am talking about, so ALL we are discussing is what happens to the bits as the forthcoming end_date is reached? And what is shown for sections for which a 'start_date' is still to be achieved. ARCHIVING material is the question here while keeping it available to view in conjunction with the current state on the ground along with the future planned state. History is an inherent part of the current 'namespace'! -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [Talk-GB] Village, Hamlet and populations ...
On 14/09/15 11:11, Tom Hughes wrote: > Broxbourne, Wormley and Turnford don't really have significant centres > and historically would likely have been considered villages. Broxbourne > is now a town in wikipedia with a population of over 13 thousand while > the other two are still listed as villages with a combined population of > around 8 thousand. Much of that is 20th century dormitory for London though. Again on IOP 'Broxbourne and Hoddesdon South' gets a population figure of 9065 and I'm seeing three other Hoddesdon entries, but the whole area is covered by 'Broxbourne' in that data ... rather than Hoddesdon. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Village, Hamlet and populations ...
On 14/09/15 13:02, SK53 wrote: > On the other hand finding a way to access population figures to places > for data consumers is useful. Directly adding population values may work > in Britain where population change is relatively slow, so slowly > outdated data is still useful, but is risky in other parts of the world. > At the very lease also add a link to wikipedia/wikidata as well, which > ultimately should obviate having to maintain population values That is precisely where I started ;) Or rather adding the wikipedia link ... The problem is that even coverage in wikipedia is far from complete so one finds erratic quality of information. Additionally wikidata is still at an early stage and does not have the bulk of the information already in wikipedia, so someone has to go through and change all the information blocks around or add missing ones where needed :( -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Village, Hamlet and populations ...
On 14/09/15 15:18, Richard Symonds wrote: > Perhaps it would be better to, instead of having a hierarchy based on > definitions, instead having a hierarchy based on pure population size. > If this gives odd results, then perhaps you could have a "booster value" > if the town is used as a post town or a seat of local government (for > example). Not going to happen. Wish list! On OSM some places have all their is_in: tags for parish, ward, county. Some rely on having enclosing boundaries to provide that information, and some have nothing where many of the boundaries are still missing. For the UK we have the whole hierarchy from the ONS data so there is no need to create it, we simply need an agreed method to use it. We could create all the is_in: tags from the data so we can search and find all of x in y, or we could pass that off to a third party such as wikidata where we just add a link to the whole gamut of what can be added virtual data wise. Currently all the wikipedia links are being added but I think that to use wikidata efficiently one has to use the designated ID rather than the name? Since Facebook insist on using the names as defined by wikipedia this is where my problem originally arose since they only add county when wikipedia do so often you have no idea which is the right place to use. What is used as a link has other consequences! It does still not get around needing the boundaries IN OSM so one can click anywhere in an area and identify which of multiple zones it is actually in. Adding this data to the places does not fill that hole :( -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Village, Hamlet and populations ...
On 14/09/15 14:23, Richard Symonds wrote: > Is there any reason that a place can't be both? > eg. > "defines self as=town" > "defines self as=village" > "defined by X as village" > > Or the like? The obvious answer is that unless one adds some sort of filter it will get counted twice? Once as a town and once as a village. There should only be one place entry in OSM or so the http://wiki.openstreetmap.org/wiki/Key:place?uselang=en-GB says. I'm not sure but adding a second place tag should not work I think. Place is part of the hierarchy of a location on OSM, so having it appear as both a town and village would be confusing on searches. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Village, Hamlet and populations ...
On 14/09/15 15:47, Colin Smale wrote: > Hi Lester, can you provide a link to the ONS data you are referring to? Main index http://www.ons.gov.uk/ons/guide-method/geography/products/index.html Useful one for now ... http://www.ons.gov.uk/ons/guide-method/geography/products/other/index-of-place-names--ipn-/index.html https://geoportal.statistics.gov.uk/Docs/Products/Index_of_Place_Names_2012_(E+W).zip Prior to 2011 it was IOP, but seems it's now IPN ... It's supplied as an Access database, but I just dumped it as a CVS and am looking to update my http://enquirysolve.co.uk/wiki/NLPG+Data software to use the new coding format. The various boundaries are also available complete with updates since 2012 -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [OSM-talk] Buildings on abandoned railways (via Overpass)
On 14/09/15 02:23, John Eldredge wrote: > In the case of Nashville, Tennessee, an Interstate bypass, I-440, that > loops about one-third of the way around the city core, deliberately made > use of a no-longer-used railway right-of-way to reduce land purchase > costs. If anyone were to go back and map that railroad now, the entire > route would overlap the motorway. One wonders if in the bigger scheme of things such as global warming and pollution if that was really the best way of saving money ;) In the UK rail usages has been growing and lines are being reinstated, or brand new lines built demolishing or 'blighting' housing stock, and in some cases those houses are being demolished where they were build on the original abandoned routes since that is required to restore the links with what does currently remain. http://www.stratford-herald.com/33894-could-rail-line-idea-be-back-on-track.html is a current item on my own pet 'abandoned railway' ... and the preserved railway is fast approaching from he south ... -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [Talk-GB] Village, Hamlet and populations ...
On 14/09/15 08:53, Mark Goodge wrote: > Historically, the distinction between a hamlet, a village and a town was > based on ecclesiastical parishes. A village was a populated area > comprising a parish of its own, with one parish church. A town was a > contiguous populated area comprising multiple ecclesiastical parishes, > while a hamlet was a populated area too small to have its own parish > (and thus being contained within another one, either a village parish or > an outlying area of a town parish). What also muddies this picture is the closure and combination of parish churches while retaining the political distinction. > This official distinction has been lost over the years with multiple > phases of local government reorganisation, but it still provides a good > rule of thumb. > > In England and Wales, a civil parish council can choose to style itself > a town council if it wishes. The majority of those which have done so > are those which, prior to the Local Government Act 1972, would have been > a Municipal Borough (eg, Evesham or Lewes) and which meet the historical > definition of a town, but by no means all of them fall into this category. The exceptions to the rule come to mind. > What that means is that population alone is a no more than a rough guide > to the likely status of a town or village, at least in England and > Wales. There's a significant overlap between the largest villages and > the smallest towns. In the first instance it is what the local preference is. Broadway is likely to remain a 'Cotswold Village' and is tagged as such on OSM, but has an identity crisis on wikipedia. The missing piece of the jigsaw is around 500 people live in the parish, but not in Broadway itself. Our own enclave is probably one unnamed 'hamlet', and I think there are potentially another couple, but I think what is missing is the outline of what ONS classify as the 'built up area' of Broadway ... And the boundaries of the parishes in Gloucestershire are missing as well ... another one for the todo list :( -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Village, Hamlet and populations ...
On 14/09/15 09:51, Tom Hughes wrote: > > To go back to the example of the town where I grew up, namely > Wotton-under-Edge in Gloucestershire. That is a single parish and has a > current population of 5627 according to wikipedia, which also points out > that it had a municipal corporation until 1886 when it was abolished > following the Municipal Corporations Act of 1883. It now has a town > council. On IOP Wotton-under-Edge is a parish with an identical built up area, so the population figures match for both entities. > Hell, try http://www.visionofbritain.org.uk/place/1009 where I now live > which wasn't even a parish in it's own right until 1844 but was almost > certainly considered a town before that, as a coaching stop on the main > road from London to Cambridge. Hoddesdon seems to have several entries on IOP and certainly a substantial population, and would exceed the 1 figure on the wiki guide. The simple fact is that the inclusion of population on the wiki page would seem to be something of a red herring, although for other countries one may need that guide? It would seem that my fingure in air 8000 may be better nearer 3000 although at the end of the day the local comunity my well call themselves a town or a village based on choice rather than size. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Village, Hamlet and populations ...
On 14/09/15 00:41, Tom Hughes wrote: > On 14/09/15 00:16, Lester Caine wrote: > >> The OSM wiki defines 'hamlet' as less than 100-200 people, but village >> supposedly starts at 1000 up to 1 with the proviso that it depends >> on the country. Ideally the two would perhaps meet :) We are perhaps >> looking at a population of around 8000 for a town designation in the UK, >> but anything down to 100 is still classified as a village by the ONS. >> What are actually missing from the OSN data are ANY hamlets despite >> their claiming to include them. > > Please don't try and draw bright lines based on population, and > certainly don't try and mass edit things based on that. It's much more > subjective than that. > > Nobody would ever have described the place where I grew up as anything > other than a town, but we always used to reckon on a population of > around 3000 people (wikipedia says 5627 as of the 2011 census) and > certainly 8000 sounds very high to me. > > Equally there were surrounding villages that aren't that much smaller > but which lack things like a High Street and really wouldn't be > described as towns - basically places which had historically been > smaller but had acquired housing in the second half of the 20th century > without necessarily growing in other ways. I've no intention of changing anything, just add in missing material, but it would be nice to have a consistent set of guide lines for the UK. Rather than the simplistic view Facebook applies ... every hamlet, village and town is a city :) I'm just looking to keep a consistent use of what is well documented data and ensure we same numbers everywhere. But to be honest I see no need to add 'population' to OSM. The wikipedia link gives a lot more than just that ... if the wikipedia page is right. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
[Talk-GB] Village, Hamlet and populations ...
I'm currently working around the area trying to get all the local places cross referenced properly. I've got The 2012 Index of Places from the ONS which has a supposedly complete set of places, but I've been hitting a number of problems which I think I've finally sussed. The OSM wiki defines 'hamlet' as less than 100-200 people, but village supposedly starts at 1000 up to 1 with the proviso that it depends on the country. Ideally the two would perhaps meet :) We are perhaps looking at a population of around 8000 for a town designation in the UK, but anything down to 100 is still classified as a village by the ONS. What are actually missing from the OSN data are ANY hamlets despite their claiming to include them. My first exercise was to add links on the OSM data to the wikipedia entry for each village, and there is a small list of miss matches which I'm trying to sort out. However when cross referencing the population data reported by wikipedia quoting the 2011 census, but what that fails to account for is that only provides totals for the whole parish, which may have more than one hamlet/village. The IOP data contains the 'BUA' population for the villages, but omits the rest of the hamlets that make up the parish/ward. The IOP data is released under the Open Government Licence V2 so I see no problem using it in OSM or Wikipedia? I think what we are still looking for is a consistent list of hamlets to work from to fill in the gaps? The IOP data was supposed to be updated annually, but it seems only annual updates are currently being generated. To add to the fun, the six digit codes I've been using for LLPG data for many years have been replaced by a 9 digit code. While the 6 digit code had a nice three level structure, the 9 digit code has lost the third layer, but these still only go down to the ward/parish level. There is still no ID for the town/village :( And Scotland and Northern Ireland are separate data sets ... -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
[OSM-talk] Mapping what is on the ground ...
This is one for the GB list really, but given the debate here and it's relevance to the discussion I thought it appropriate to widen the exposure. I'm battling Facebook and it's crap set of official 'cites' which one is allowed to select as home places. We won't go on about the location of my own home town 'Broadway' since I simply refuse to use the Facebook offering, but just up the road we have 'Weston Sub Edge' and 'Aston-sub-Edge' as shown on the OS map http://data.ordnancesurvey.co.uk/doc/70020328 and reported by the post office as 'Weston Subedge' and 'Aston Sub Edge'. The road signs and official material are also incosistent, so while the signs into Aston show 'ASTON-sub-EDGE', the ones for Weston vary between 'Weston Sub-Edge and 'Weston-Sub-Edge' but the navigation roadsigns show 'Aston Subedge' and 'Weston Subedge' mixed with 'Weston-Sub-Edge'. Old maps consistently use the 'Subedge' second half and that incudes what started the fun ... the map showing the 'Weston Subedge Halt' on the old railway ;) I'm asking both parish clerks just what the current legal position is, but mapping what you see does not help. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Fw: important... recomend auto deletion of any Fw:important
On 11/09/15 01:27, Warin wrote: > I'll be putting a block on any more messages with the subject containing > Fw:important The buggers are using several subjects, multiple probably hacked relay sites, but all ending at the same crap. Just need that site closing down? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Motion: Dedicated mailing list for abandoned railways
On 09/09/15 08:48, Jukka Rahkonen wrote: > I suggest to make a new mailing list for those who want to talk about > abandoned railways. I can see that such does not exist yet > https://lists.openstreetmap.org/listinfo. It has already suggested to use > the Historic list https://lists.openstreetmap.org/pipermail/historic/ but I > have understood that abandoned railways are special and therefore they > deserve a dedicated list. Sorry but this is missing the point ... There are many specialist areas, and some good lists that already support them, but the whole point here is not a single niche subject, but rather elements that are currently mapped but evolve over time retaining many elements visible on the ground but with gaps created due to new development, or changes in layout as bottlenecks are improved. We are not talking about mapping historic details from maps but simply retaining currently mapped data in a slightly different format so that the information can still be accessed. Railways are just a small but easy to identify element that have become uneconomic over time. As have some waterways, and coastal erosion continues to affect some areas of the UK, and I'm sure the rest of the world. All are slowly evolving ove time and that evolution is as useful as a map that only displays a snapshot of 'today'. HOT is mapping disaster areas and the the currently mapped details get overlaid with the effects of change and the remaining visible elements stick through. Trying to cherry pick what gets kicked to an 'historic' database and what remains in the main one should not be necessary especially if the restoration work slowly returns the original elements. Simply tagging collapsed buildings or blocked roads and railways and then tagging what gets rebuilt and what replaced is exactly the sort of data that should be recorded. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] THIS is the kind of enthusiasm some would reject
On 08/09/15 07:01, Mateusz Konieczny wrote: >> I don't believe anyone's advocating the removal of existing entities. >> > In your viaduct case above, keep the viaduct entity, remove the >> > railway=abandoned tag, use the historical tag to describe the past of >> > the viaduct (which exists) but don't use it to describe the railway >> > (which doesn't). > Note that it is not necessary to use "historical tag" for existing > viaduct. man_made=bridge seems to fit well. Since I live within quarter of a mile of a section of 'abandoned railway' which used to extend several miles south and still extends several miles north I have a perfect example of one which perhaps needs re-tagging. There are still bridges and cuttings and while it is still essentially 'private property' some sections have unofficial footpaths. The section south of Broadway now has a new track (where it was dual track originally) and a growing new station. There is an option to continue north which would allow the preserved line to reconnect with the existing main line. Given the problems the preserved line has had with land slippage and having to replace 100+ year old bridges it may never happen, but the route is protected currently but 'abandoned'. Now if rendering changes do I need to re-tag this 'dismantled' to maintain the current map, or is this another reason for abandoning the main map service and providing an alternative? While my point about the historic aspect has been dropped, this line is also a good example of the managing of current and historic material since the main route of the line is preserved, the goods yard at a local army base is now being built over with a new housing development, and while it currently exists on the ground it will slowly succumb to progress and I have no problem with the CURRENT view of the data only showing the elements that remain, the history is a documented fact which just needs a slightly different style of management. It's NOT actually been removed from the database since it is still fully documented in the change log, so all we are talking about is semantics and allowing a view of the data that can be rendered using the existing tools. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] THIS is the kind of enthusiasm some would reject
On 08/09/15 12:58, Mateusz Konieczny wrote: >> The historical tag can be used to indicate that the viaduct was >> > previously used as a railway. It should be used in conjunction with >> > other tags such as man_made. > Is there anything **currently** making clear (or at least indicating) > that it is constructed as a railway bridge? Is there any difference? > > Historical data should not be added and if present - removed. This is perhaps the sticking point? A structure exists due to the previous construction of say a railway and it gets 're-tasked' to something else. If it's called 'the old railway viaduct' then that is acceptable, but if it's just called 'the viaduct' one is not allowed to add in some way 'formally the xxx railway'? Having to dig back through change log data to establish that was previously mapped while it was a something else when many locals will be looking for 'the old xxx' is wrong. If the object being mapped has an historical aspect there should be no objection to adding that data and no one has a right to remove it. Even 'site of xxx' has a precedent to map it if there is some marker visible on the ground but no other indication it ever existed. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] THIS is the kind of enthusiasm some would reject
On 07/09/15 23:16, Dave F. wrote: > I'm not sure there's been a discussion as you've mostly ignored the > basic comment made - it it's deleted in the real world it gets deleted > in OSM. If there is still a trace of anything related to something being deleted ... it gets it's tags modified. You only remove it completely when/if it is replaced by an alternate structure. A forest may well get felled for timber, become open land until a new crop is finally established. Just as in some cases tracks have been lifted on a viaduct or cutting but the railway use for that land is still documented. One of the problems we had was people removing the way which was actually another structure such as a viaduct and that was removed as well. The request was for people NOT to remove something if they did not understand it's reason for existing. Certainly some of the 'automated' editing of material without any personal intervention is not acceptable. But we still need a proper way to move perfectly valid 'old' data to an alternative if that is what the majority want ... I just happen to think that this is the wrong way of managing material that NEEDS a substantial amount of the existing live data to be able to manage it's complete display so one has to now manage two parallel versions of the same data :( OHM can only work if it is a compete copy of the current visible data, and all of the historic data that has gone before so that as new parts are deleted from one they remain valid in the other. The current OHM is simply a scratch pad to store isolated historic material. It does not have any of the history currently being created daily in the main database. 'It get deleted' is the very history that someone has spent a lot of time previously documenting. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [Talk-GB] Rail maxspeeds being converted
On 05/09/15 09:18, Dave F. wrote: > Are the signs next to the rails still in mph? > > His edits appear to be country wide. > https://www.openstreetmap.org/way/308866549/history > > I've sent him a message asking why he's doing it. The whole system still works to MPH acording to the current working books http://www.networkrail.co.uk/aspx/10563.aspx Anything else is simply incorrect, so these changes should be reverted back to the correct data? -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [OSM-talk] Preserving History ...
On 03/09/15 09:34, Roland Olbricht wrote: >> http://www.openstreetmap.org/#map=16/52.3790/-1.4692 is a perfect >> example, where a small roundabout has evolved into a much larger one, >> but the new through route is still being excavated. The problem with >> this is while it is a good approximation of how work is progressing, >> many of the original roads have been deleted and replaced and all the >> history relating to them lost. > > Please have a look at > http://overpass-turbo.eu/s/bg7 > > You can view that data as well in JOSM: Click "Export > raw data", save > the created file and open it via "File > Open" in JOSM. > > The same works for any other date. The history starts from the license > change on 2012-09-12 and proceeds till today. I can see the history from 2012, but I an convinced that there should be some more history prior to that date since the 2012 data already contains the planed improvements to the junction. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Abandoned Rails
On 02/09/15 11:30, moltonel 3x Combo wrote: > I always understood that land property was out of scope for OSM. Do > you know of any osm data which records property ? So I should remove all the detail on http://www.openstreetmap.org/search?query=wr12%207ep#map=17/52.04851/-1.85665 rather than adding the missing detail to the right? And the 'Abandoned Railway' to the left is a protected route which HAS encroachments onto it, but which is still potentially re-enstatable once the line to Broadway becomes active again. Using the ground for a long period does not always allow to take possession of it and rail routes are one of those documented exceptions. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Abandoned Rails
On 02/09/15 14:25, Paul Johnson wrote: > But in most (all?) of the US, land ownership (and vehicle ownership, for > that matter) records are open and subject to public inspection, and why > land transfers are typically published conspicuously in the regional > news periodical of record. And in the UK you just buy a copy of the Electoral Register ... -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Abandoned Rails
On 02/09/15 12:56, moltonel 3x Combo wrote: >> So I should remove all the detail on >> > http://www.openstreetmap.org/search?query=wr12%207ep#map=17/52.04851/-1.85665 >> > rather than adding the missing detail to the right? > Sorry, I see residential areas, hedges, postcodes, etc, overall a well > maped area with usefull detail, but nothing mapping land property ? I > never suggested that this kind of data was out of scope for OSM, > please go ahead and map more of it. When I mentioned "land property" > I'm talking about legal ownership, parcels, cadastre, etc. The boundaries mapped are property boundaries as are the field boundaries around them. Ideally I would like to add the NLPG reference to each but currently that is blocked by possible licensing problems. It WOULD be nice to complete the 'hidden' data in much the same way the postcode is added here so that searches can be done properly. Just because a post has not been put in the ground to identify a location, the location still exists if properly documented. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Abandoned Rails
On 02/09/15 13:23, p...@trigpoint.me.uk wrote: >> The principle of "what data belongs in OSM" is about the propeties of >> > that data, not what kind of data it is. But as it happens, a given >> > kind of data usually has the same properties, so "this kind of data >> > doesn't belong in OSM" is a usefull simplification. >> > > We can map barriers and visible dividing marks, but land ownership has > massive privacy and data protection issues. Adding actual names against a property is perhaps the question here? And that is a step to far. Even NLPG data does not have personal data included IN it, but simply putting an address into a phone book on the internet will more than likely provide that data. I don't see any problem identifying what fields belong to a particular farm or what area belongs to a particular dwelling. That is essentially 'land use' and part of this came about because of a blanket application of 'farmland' to some areas of the UK but where proper verification of boundaries throws up various problems with THAT data and now tidying that up needs 'non-visible' information in addition to the visible stuff to correct the mistakes. I have tried simply hiding the data which works around here, but another area nearer London is obviously using a different tag to 'farmland' ... just not identified yet what else to ignore. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Abandoned Rails
On 02/09/15 13:43, moltonel 3x Combo wrote: >> Ideally I would like to add the NLPG reference >> > to each but currently that is blocked by possible licensing problems. It >> > WOULD be nice to complete the 'hidden' data > Assuming the license issue gets resolved, how will you import it, > conflate with existing data, tag ? How will you keep it up to date > when a field or a chunk of garden changes hand between neighbours ? > Who do you expect will make use of the data ? My main client base is council based systems, so I HAVE the LLPG data for each client. I just can't use it in OSM. The main problem here is I'm using OSM to display location information so that staff can manage things like 'change of address', and YES recording changes to the LLPG data in order to amend the raw data which is then uploaded to NLPG. Using OSM allows much more flexible options on the user interface than OS does and potentially the whole country can be kept up to date simply because the data is already being digitised and we the UK public are paying for that! Drawing a line around each boundary is still a problem as are areas in general in OSM. I'm not happy with using 'relations' to draw areas, but simply drawing a polygon for a field or property boundary is messy when one needs to add gates, boundary style, and yes 'hidden' elements such as open driveways. So one is essentially limited to creating a relation for each property and then does one include the buildings within the area in the relation? Which is why I've stopped at just drawing boundary types so far. Actually the property boundaries on our road include the area of grass outside the front fence, but not the path which is another detail not yet added ... -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] What's the point of the Wiki 'k=v' icon?
On 01/09/15 12:14, Dave F. wrote: > That the icon is used purely as a place holder & *removed* when a > picture is added proves its irrelevance. > > Why would a newbie care or even need to know the XML structure of the > database? That there is a lot of material that is NOT rendered on the map is a fact. These do not have icons so the k=v is the IDEAL icon for entries in the wiki that ARE only available as XML data. > The wiki has to be a clear & simple as possible. > > "I see that icon as an indication that the detail I'm looking at has a > assortment of icons within the page." > "the only way of viewing data tagged by k=v is to search on the raw data > rather than expecting an icon to appear on the map." > > Sorry, but I don't really understand what you've said there. You seem to think that every page only as a single icon associated with it? A large number of pages have a column for the icon that relates to those sub-entries the page describes. Some sub-elements will again not be available rendered on a map. Perhaps we should simply strip the icon on the page header when there is no single selection, but it is a useful prompt that the page contains key/value data rather than other types of material. Are we now saying that the wiki is only allowed to document key/value elements that are rendered and therefore have to have icons defined? The growing number of maps do have different lists of rendered objects so even the fact that a page has an icon does not mean you will see it on the map you are looking at. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Proposed mechanical edit: surface=woodchip to surface=woodchips
On 31/08/15 13:54, Mateusz Konieczny wrote: > On Mon, 31 Aug 2015 13:32:24 +0100 > les...@lsces.co.uk wrote: > >> > This may just be a language thing, but in my dictionary the mulch >> > material is woodchip, so the incorrect spelling is woodchips ... >> > unless one is using it as a fuel for an 'eco' power plant. So is the >> > mistake simply because people are translating from other languages? > What dictionary is giving this definition? > https://en.wikipedia.org/wiki/Woodchips is in plural form, according to > https://en.wiktionary.org/wiki/woodchip woodchip is for a single > mechanically produced chip of wood. > > In my printed dictionaries (Polish-English and English-Polish, one by > Oxford University Press and second by Wiedza Powszechna) neither term > appears. Back in the office now ... that post should have gone to the list as well :( I'm just watching a item on the telly about biomass and processing woodchip to it's various sizes for the jobs it's destined for. On google woodchip produces results ... woodchips does not ... so in my book surface=woodchips is what is wrong! Playground areas are covered with woodchip or bark as one of the approved safety requirements although it's not uncommon to use the term wood fibre to cover the whole range of material. Will add notes to the soil/dirt thread but equestrian surfaces using wood chip or soil have more stringent requirements in relation to dust control. 'dirt' tends to imply a dusty surface, something one would not expect with a better managed soil surface. -- Lester Caine - G8HFL - Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk Rainbow Digital Media - http://rainbowdigitalmedia.co.uk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
