Re: [OSM-talk] Is it technically and legally possible to add the Open Location Code to the OSM search?

[Barry Hunter]

On Fri, Aug 10, 2018 at 8:18 PM Oleksiy Muzalyev <
oleksiy.muzal...@bluewin.ch> wrote:

> On 10.08.18 21:07, Mark Wagner wrote:
> > On Fri, 10 Aug 2018 09:32:50 -0700
> > Vao Matua  wrote:
> >
> >> Plus code can be calculated on the fly, but if they are
> >> to be used we will need to have hardcopy maps with the addresses that
> >> can be used to direct aid workers to a specific location.
> > Plus codes form a hierarchical grid, so supporting them on hardcopy
> > maps can easily be done when the maps are prepared for printing.
> >
> > I don't know if you're familiar with the UGSG topo maps, but if you
> > aren't, I recommend looking at one of the 1:24000-scale maps from the
> > late 1970s/early 1980s.  It's got three location grids on it: UTM
> > coordinates and latitude/longitude markings on the outside, and PLSS
> > township/range/section markings on the map itself.  Adding a plus-code
> > grid to the map would be no problem, and wouldn't require importing
> > billions of tags into OSM.
> >
> It is absolutely clear. A plus-code is generated by a mathematical
> formula from coordinates almost instantaneously, and vice versa.
>
> The same as say the binary code is generated from the C++ programming
> language, or words are created from letters, etc. It is just another
> layer of abstraction, which makes it easier to perform a task.
>
> In principle it is possible to write a computer program in assembler,
> the low-level programming language. But it is a bit easier to do it in
> C++, Java, PHP, etc. The same is here.



> It is easier to memorize a
> plus-code, to transmit over the telephone, to put it on the address
> plaque, etc.


So do that *now*. It works.

The 'receiver' just needs a program (unless they know how to decode in
head!) to decode it and use it.

OSMAnd is one example quoted that can ALREADY decode it.

https://osmand.net/blog?id=osmand-2-6-released

Support for Open Location Code (OLC)

OsmAnd now also supports the Open Location Code (OLC) way of representing
coordinates. OLC coordinates are a combination of letters and numbers, and
is considered to be handled easier than the traditional latitude and
longitude coordinates. Please read more about OLC here
. You can now also search locations via this
code in the Search menu - Address - Coordinates Search, there select 'OLC'
under Coordinate format. Also, the context menus of any location selected
now displays OLC in addition to Latitude and longitude.


Nominatim, or any other 'search box' or geocoder could just as easily
implement the decoding of the Open Location Code.

It's a tiny (in the grand scheme of things) but of code to add to the
application.







> Yes, it is possible to do the same thing with coordinates'
> digits, but nobody does it.
>

How many people do enter a coordinate in the OSM serach box? Apart from map
geeks not many.

With little effort the box could easily understand a plus code.



It doesnt need millions of tags imported into OSM database to enable it.
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Is it technically and legally possible to add the Open Location Code to the OSM search?

[Barry Hunter]

On Fri, Aug 10, 2018 at 6:09 PM Martin Koppenhoefer 
wrote:

>
>
> sent from a phone
>
> > On 10. Aug 2018, at 19:02, Barry Hunter  wrote:
> >
> > another issue with it being added as tags, if the node is moved to
> correct its location, the editor would have to remember to update the
> plus-code tags as well (not just the lat/long)
>
>
> this only if you see it as location information, not if it is used as an
> address (the location where to go to, see the example of the long driveway
> above)
>

Haven't seen that example.

But in the case of a long driveway wouldnt the address be attached to the
entryway (so that directions etc, can route to the right location)?

The plus-code can be computed from that node.
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Is it technically and legally possible to add the Open Location Code to the OSM search?

[Barry Hunter]

> It is interesting that this effort for addressing is being trashed because
> it is savvy technology.
>

Dont see anyone has 'trashed' the idea of using Plus Codes as such.

Just the bulk import of them as *data* to the core OSM database. Its
redundant data.





> Plus code can be calculated on the fly,
>

Exactly, the address search interface/gazetteer could show the address
*with* a plus code.



> but if they are to be used we will need to have hard copy maps with the
> addresses that can be used to direct aid workers to a specific location.
>

So have a fancy renderer than can render the Plus Codes as needed.

Never tried, but it should be possible to render special plus code aligned
grids on the map.


I don't understand what problems would be created by adding valuable
> information to an address point.
>

another issue with it being added as tags, if the node is moved to correct
its location, the editor would have to remember to update the plus-code
tags as well (not just the lat/long)

If want to find where a plus code is, a search interface, can just decode
the plus code, get a lat/long and run a standard geospatial query. That
works *now* worldwide, without any sort of bulk import of tags.

... as tags it would need a text index of the tags, and search that.

Apparently there are currently 4665583767 nodes in OSM, the tags mentioned
seem to be about 83 bytes long. That's 390 *Gigabytes* of data just to add
plus codes to them all.


A basic mobile phone could easily compute the location for a plus-code with
the algorithm. If had to look it up in this database would be way more than
390gb of data on the device! (or need an active internet connection to the
online database)




>   So far I see no practical solution for giving an address to the billions
> of people that do not have one,
>

Every single point on earth already has a plus code already. Its already
been assigned by the algorithm.

Just like every single point as a lat/long coordinate.
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Paper/Article about stagnation in OSM

[Barry Hunter]

On 1 August 2018 at 08:37, oleksiy.muzalyev wrote:

>
>
> At the same time it is possible to create 208 billion of 8-letter unique
> quasi-words with 26 letters of English alphabet (26 in the power of 8 =
> 208827064576). Even more if numbers are included. It's enough for all
> dwellings on Earth. It is easy to transmit a 8 letter word via telephone
> with ICAO Phonetic Alphabet [1].
>
> Then when we call in browser something like: osm.org/?address=hj3u878s or
> type the unique quasi-word into a search of of the OSM map: the distinctive
> geo-marker appears at the respective location with the additional
> information, such as entrance door code, apartment level, etc.
>

This sounds remarkably like geohash. https://en.wikipedia.org/wiki/Geohash

Algorithm is public domain, and not commercial .


Also
https://github.com/google/open-location-code/wiki/Evaluation-of-Location-Encoding-Systems
and https://xkcd.com/927/
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Can we utilise the popularity of Pokémon Go for OSM?

[Barry Hunter]

Maybe a mobile app, could be running in the background. It checks the
current status of the local area in OSM, and pops up a notification if the
there are problems*.

Things like
* There doesnt seem to be many features mapped in the local area, does*
this map* match what you see around you?
* This area was last updated over 2 years ago, please compare *this map*,
and see if it matches what you see around you.
* Someone made a *note *nearby, are you available to check it out?


Where 'this map' is a link to a map of the local area. Would possibly need
a bit of an AI, to avoid over 'nagging' the user. Tries a few trial
notifications, if the user responces, show more. If not show less. Adapts
to the user activity.






On 14 July 2016 at 12:46, Svavar Kjarrval  wrote:

> Hi.
>
> Now that Pokémon Go[1] is gaining popularity, people in general will
> spend more time outside walking around with their phone in hand (and
> some with mobile USB chargers). I'm wondering if there might be
> opportunities for OpenStreetMap to utilise that activity to encourage
> the gamers to collect (non-infringing) data for OSM while they're
> playing the game anyway.
>
> I'm not claiming I have any specific ideas at this point. Just wanted to
> bring this point forward.
>
> [1] https://en.wikipedia.org/wiki/Pok%C3%A9mon_Go
>
> With regards,
> Svavar Kjarrval
>
>
> ___
> talk mailing list
> talk@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk
>
>


-- 
Barry

- www.nearby.org.uk - www.geograph.org.uk -
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] What3words

[Barry Hunter]

On 22 November 2015 at 11:07, Colin Smale  wrote:

>  I guess there would be no objections to someone adding
> addr:w3w:en=nice.place.here ? Or addr:w3w=en:nice.place.here ?
>

Ok, so turning it around, what would be the benefit of this? Why bother?
What purpose does it serve?




The only purpose I see, it then serves as an independent database of w3w's.
Persuming the database is comprehensive enough could perform a lookup via
the OSM database to find the location of a given set of words.

But as then it effectively makes the proprietary 'app' redundant,
 what3words themselves might then take exception to it anyway. And/or might
infringe database right.



As I see it what3words is not really 'andress' system. Its a coordinate
system. Its a way of encoding a location in a series of words. OSM
generally uses decimal Lat/Long as its encoding.

OSM doesnt store every known coordinate in the system. Like
addr:mgrs=4QFJ.1234.6789
or British National Grid. Or Indian Grid. Or UTM Coordinates. etc

There are lots of ways of encoding coordinates. Each with (generally) a
well known conversion from WGS84 lat/long (complicated by the fact there
may be a datum transformation needed)

If end user wants to use a given system, then can use that conversion. They
dont need to exist in the database.
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] How best to create a single point of interest online map with OSM data?

[Barry Hunter]

http://cartodb.com has OSM Basemaps (via MapBox)

It can do server side rendering of datasets like fusion tables does.


On 18 Sep 2014 00:38, Bryce Nesbitt bry...@obviously.com wrote:

 What's the best way to create a global single point of interest map,
 with OSM?


 I'm thinking something like this local pay phone map:

 https://a.tiles.mapbox.com/v4/brycenesbitt.j82lj086/page.html?access_token=pk.eyJ1IjoiYnJ5Y2VuZXNiaXR0IiwiYSI6ImNFME9IckkifQ.Nd85HRRFP3Jy3gx8nQ3ATA#14/37.8699/-122.2603
 But global, and with all the tags for each node shown when the node is
 clicked on.


 Or this global drinking water map:

 https://www.google.com/fusiontables/embedviz?q=select+col3+from+1usHO73s_NDGKOx-2jbj0xtSHuHjxvWVo_2MvX_oviz=MAPh=falselat=41.571877511144756lng=-83.65702047624372t=1z=4l=col3y=2tmplt=2hml=GEOCODABLE

 But it feels wrong to use Google maps as the backdrop for OSM data,
 despite the advantages (the map above has 35,000 nodes many with photos,
 and yet it is snappy fast on any browser).

 What's a better way to do this?



 
 Note: The mapbox map started with an overpass API query.

 The fusion table example was an extract from the planet file, merged into
 a fusion table.  Google's servers create and cache bitmaps with the POI's.
 User clicks look up the matching data.  Thus it renders as fast as a slippy
 map, but has all the POI's readily available.

 ___
 talk mailing list
 talk@openstreetmap.org
 https://lists.openstreetmap.org/listinfo/talk



-- 
Barry

- www.nearby.org.uk - www.geograph.org.uk -
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Welcome box on the new map page

[Barry Hunter]

On 1 December 2013 12:37, Lester Caine wrote:


 The new setup seems to be saying 'if you are not here to map go away!' If
 this is really the case then perhaps we need a NEW site for those of us who
 contribute data with a view to actually using it live?

http://open.mapquest.co.uk/

? That is a site for end users of OSM created maps.


In many ways OSM is a data provider, it doesnt have the resources to
create website for end users (as such).

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Good use of time ...

[Barry Hunter]

On 2 December 2013 13:17, Lester Caine wrote:
 what is 'export' intended to do?

It gives option to select an area, then download a osm XML file.
Exports the map as data. If the area is big, gives direct links to
various bulk options.

Works in Chrome and Firefox for me, guess its a browser related issue,
with what you using?

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Good use of time ...

[Barry Hunter]

On 2 December 2013 13:51, Martin Koppenhoefer  wrote:

 offering a list of
 services where you can actually get more and all data (i.e. planet.osm.org,

already listed there

 geofabrik,

already there

 metroextracts,

already there

 coastlines, overpass, torrent, etc. (if those are
 fine with being mentioned prominently there)).

The last link is to
http://wiki.openstreetmap.org/wiki/Download

so they can all be listed there.



 Also it was said that our main API was for editing purposes only, how does
 that relate to an export function like this?

Its intended for Editing. But doesnt meant it can't be used (in a
limited capacity) for downloads...

That wording I would think is mainly to discourage people using the
main API as a general downloading API. There are alternatives that
would almost certainly work better.

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Welcome box on the new map page

[Barry Hunter]

On 2 December 2013 13:54, Lester Caine les...@lsces.co.uk wrote:
 Barry Hunter wrote:

 The new setup seems to be saying 'if you are not here to map go away!'
  If
 this is really the case then perhaps we need a NEW site for those of us
  who
 contribute data with a view to actually using it live?

 http://open.mapquest.co.uk/

 ? That is a site for end users of OSM created maps.

 In many ways OSM is a data provider, it doesnt have the resources to
 create website for end users (as such).


 Now I've now found the 'html' option on the OSM map, and along with the link
 and short link these target the same map. Simply replacing the URL with the
 mapquest one does not provide a working link,

Why should it, they are different sites. MapQuest has their own embed option.

 and in any case the rendering
 of roads in this area is simply wrong ...

Good excuse to badger MapQuest to update their rendering!

If you direct more users at MapQuest, they will have more incentive to
update it :)


 If a service is not going to be provided then it should not be prominent on
 the front page?

The feature is to embed the map as seen on the OSM page. Which it does.

You seem to be wanting to use it for a different purpose.


 In any case, the embeded map is provided via OSM and so
 maintaining a compatible rendering in the follow on link is important,

You seem to be giving mixed signals what you expect this 'follow on
link' to be 'for'.

Is it for users to find out more about OSM? For users wanting to get
directions to the location shown on the map?

They are very different use cases.


 switching to one with a different style on another site seems wrong? If we
 were providing these pages ourselves then we would still be using the OSM
 tile server?

As long as adear to the usage Policy, then you could
http://wiki.openstreetmap.org/wiki/Tile_usage_policy


  I got around 50% of the way through rendering my own set -
 after 7 days - but the tiles were already well out of date :(

Nobody said it would be easy.

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Good use of time ...

[Barry Hunter]

 Yes, they are all in the wiki, of course, but where did you find that link
 to the Download page on the main page?

When click Export :) Under the box for selecting a BBOX for the export.

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Irish placenames anomaly

[Barry Hunter]

This suggests that the data has come from GNS/NGA

I noticed the same anomoly in Geonames,
http://forum.geonames.org/gforum/posts/list/268.page
which used the same source.

Apprently its to do with how the data was digitized from paper maps.

(the dense blob, is likly to be a local mapper!)



On Sun, Jul 14, 2013 at 8:52 PM, Steve Chilton s.l.chil...@mdx.ac.ukwrote:

 The distribution of Irish placenames from OSM presents a couple of
 geographical anomalies. (See
 http://steve8.dev.openstreetmap.org/OSM_Placenames_Ireland.jpg). First,
 the west-east central belt has suspiciously far fewer names than
 immediately to the north and south. And secondly, bang in the middle of
 this belt is a roughly circular area with a dense mass of placenames.
 Any explanations spring to mind...?

 Cheers
 Steve


 ---


 Please note that Middlesex University's preferred way of receiving all
 correspondence is via email in line with our Environmental Policy. All
 incoming post to Middlesex University is opened and scanned by our digital
 document handler, CDS, and then emailed to the recipient.

 If you do not want your correspondence to Middlesex University processed
 in this way please email the recipient directly. Parcels, couriered items
 and recorded delivery items will not be opened or scanned by CDS.  There
 are items which are exceptions which will be opened by CDS but will not
 be scanned a full list of these can be obtained by contacting the
 University.


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




-- 
Barry

- www.nearby.org.uk - www.geograph.org.uk -
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk