VeloViewer Explorer Score and Max Square

[Peat]

Getting there with Oxfordshire, bit near Reading to go, and then up in the North/Banbury & beyond...

oxon0820 by a oxon, on Flickr

(click to show/hide)

Oxfordshire is somewhere in the middle, by area...
https://www.adducation.info/earth-general-knowledge/england-counties/ 

Andy, could you tell me how you generated that map? That would be really handy when planning a square bagging ride.

[citoyen]

Having never played the game before, I am finding it such a great way to explore my new local area

One of the things I enjoyed about yesterday's ride is that it didn't take me far from some of my usual routes, yet it uncovered some really lovely lanes I've never ridden before. That's been the biggest benefit to Veloviewer for me (aside from the satisfaction of filling in new tiles for its own sake).

It will take a lot of walking (annoyingly, I should have recorded many of our hikes already) as the high moors are largely impenetrable by bike.

Same here. A lot of the coastal paths are unsuitable for cycling - but again, it's been a great way of discovering new places to walk the dog.

[citoyen]

Andy, could you tell me how you generated that map? That would be really handy when planning a square bagging ride.

IIRC, you can export the outlines of blank tiles surrounding your main cluster as a KML from Veloviewer. Then convert it to GPX and import it to something like GPXeditor to view the map.

(The process is explained earlier in the thread if you can be bothered to trawl through to find it.)

[Auntie Helen]

if you cross into Germany at Siebengewald and ride down the German side you get
A. Quieter roads with fantastic road surface (the roads in NL at this point are rather buzzy asphalt)
B. Three cracking cake emporia. One near Weeze airport, one just outside Twisteden, one west of Straelen

Might take you slightly overdistance by the time you get back to Venlo station though.

Something like this:

https://www.strava.com/routes/2724715401175676904

I'm not so sure I'd do it as a 200 yet, I'm still rebuilding my fitness, Maybe I'd do it as a 2 day trip 100k each day... or just get a train to Nijmegen and cycle from there...

J

Train to Nijmegen is a good option. Then if I can persuade you it's a lovely ride south within Germany if you leave your track at Gennep. Your current track goes within about 1km of one of the great cafes so if you are wedded to sticking in NL you only have a minor divert for decent cake (about 500 metres over the border into Germany).

[andyoxon]

...
Andy, could you tell me how you generated that map? That would be really handy when planning a square bagging ride.

I used the BRouter website:  https://brouter.de/brouter-web/#map=5/50.990/9.860/standard
Downloaded the kml file of unbagged tiles surrounding my cluster, from the veloviewer summary page.
Manually traced (from OSM) the boundary of Oxfordshire in RWGPS, saved as a gpx.  (Can post link if wanted)
Uploaded both these gpx tracks files into BRouter.

I use the vv kml file & BRouter* route planning, since the free version of Strava routeplanner 'disappeared'. 

*also super easy to share routes with.   Just copy URL of route, which contains the route info, and email...

[Peat]

Thanks. Perfect.

[Lightning Phil]

...
Andy, could you tell me how you generated that map? That would be really handy when planning a square bagging ride.

I used the BRouter website:  https://brouter.de/brouter-web/#map=5/50.990/9.860/standard
Downloaded the kml file of unbagged tiles surrounding my cluster, from the veloviewer summary page.
Manually traced (from OSM) the boundary of Oxfordshire in RWGPS, saved as a gpx.  (Can post link if wanted)
Uploaded both these gpx tracks into BRouter.

I use the vv kml file & BRouter* route planning, since the free version of Strava routeplanner 'disappeared'. 

*also super easy to share routes with.   Just copy URL of route, which contains the route info, and email...

You can use overpass turbo to pull out OSM county borders as a single GPX.

http://overpass-turbo.eu/

[quixoticgeek]

Ooph. After a year or so of no movement on square or cluster, in a week I've got it up by 69. Now at 712. Not as satisfying as 666, but it now goes all the way to Utrecht. I need to fill in the gaps between Utrecht and Hilversum. And From Utrecht to Alphen. But it's positive movement. A few more rides, and I should be able to get my square to move close to Utrecht, and stop being limited by the Noord Holland Peninsular...


https://pbs.twimg.com/media/EerfvwJWsAIX1A0?format=jpg&name=large

J

[quixoticgeek]

Is there an easy way to get the lat/long of the 4 corners of a specific tile?

J

[Greenbank]

Plot a route that touches the 4 corners and then examine the resulting GPX file.

There obviously is a formula/algorithm for finding the corners but it'd be a huge amount of work to derive it.

[quixoticgeek]

Plot a route that touches the 4 corners and then examine the resulting GPX file.

There obviously is a formula/algorithm for finding the corners but it'd be a huge amount of work to derive it.

That is sub optimal, the four corners are all in the middle of a river...

Isn't there some KML interface for this?

J

[Kim]

Today's headwindfest filled in the last of what I've been thinking of as the Worcester Gap - my cluster now runs uninterrupted between Worcester and Stratford.  (It's done bugger all for my max square until I can perform some sort of slingshot manoeuvre around the dark side of Coventry, and fill in an assortment of annoying gaps to the south-east.)

This involved some trespassing, nettles, some muddy cyclocross, wedging a recumbent vertically in a kissing gate, carrying the bike up and then down some slippery steps on a narrow footpath, discovering that the path was comprehensively blocked off by building works, and having to re-trace the afore-mentioned slippery steps to a marina.  Which I only managed to escape by Indiana Jonesing an electronic gate as it closed after a car passed through (it was either that or the mud and the kissing gate again).

Sort of thing that really makes you appreciate the Dudley Bypass...

[Greenbank]

Isn't there some KML interface for this?

You can download a KML file of your unexplored tiles near your cluster. It'll be in there.

[quixoticgeek]

You can download a KML file of your unexplored tiles near your cluster. It'll be in there.

Bingo, got it.

Thanks. Can add that as waypoints to my gps.

J

[PeteBart]

You can download a KML file of your unexplored tiles near your cluster. It'll be in there.

Bingo, got it.

Thanks. Can add that as waypoints to my gps.

J

Not sure how the JavaScript is going to render but here is the maths that RideEveryTile and VeloViewer use. (You can use size=1 for an ordinary tile)

——

function tile_to_lat(y) {
      var z = 14;
       var n = Math.PI - 2 * Math.PI * y / Math.pow(2, z);
      return (180 / Math.PI * Math.atan(0.5 * (Math.exp(n) - Math.exp(-n))));
   }
   
   function tile_to_long(x)
   {
      var z = 14;
       return (x / Math.pow(2, z) * 360 - 180);
   }

    function tile_to_latlong_rectangle(x, y, size) {
      var left = tile_to_long(x)
      var right = tile_to_long(x+size)
      var top = tile_to_lat(y)
      var bottom = tile_to_lat(y+size)
      return [[top, left], [bottom, right]];
   }

———


Sent from my iPhone using Tapatalk

[Davef]

Is there an easy way to get the lat/long of the 4 corners of a specific tile?

J

Vv tiles are Osm tiles at zoom level 14
To find all 4 corners you would call the below functions for x and x+1 and y and y+1

#define VVZOOM 14

// lon from tilex and zoom level
double osm_tilex2lon(int x, int z=VVZOOM)
{
return x / (double)(1 << z) * 360.0 - 180;
}

// lat from tiley and zoom level
double osm_tiley2lat(int y, int z=VVZOOM)
{
double n = M_PI - 2.0 * M_PI * y / (double)(1 << z);
return 180.0 / M_PI * atan(0.5 * (exp(n) - exp(-n)));
}

[Davef]

Snap. JavaScript and c++. Other languages are available.

[R_nger]

If you don't want to do the maths, there are other options...the updated route planner on RWGPS allows you to right click to popup a context menu which includes a "show lat/long" option (I think you can also left click on google maps)

[quixoticgeek]

Using the ride every tile heat map to investigate target tiles, I came across this ride, which was tagged as having got 3 tiles that are hard to get...

https://www.strava.com/activities/1291592803

Is it me or is that blatantly cheating?

J

[JonBuoy]

I guess that you are talking about the straight line section at the northeast of the trace.  The rider appears to have lost nearly 30 minutes of the recording and Strava will just join the two points.  Veloviewer has some algorithm to try to ensure that sections with dropped signals don't score tiles.  If the VV algorithm hasn't done this and the rider has deliberately taken advantage of the situation to grab those tiles then you might indeed call it blatant cheating.  Maybe the rider didn't realise what had happened - in which case it isn't.

Now you're going to tell me that you were talking about something entirely different 

[quixoticgeek]

I guess that you are talking about the straight line section at the northeast of the trace.  The rider appears to have lost nearly 30 minutes of the recording and Strava will just join the two points.  Veloviewer has some algorithm to try to ensure that sections with dropped signals don't score tiles.  If the VV algorithm hasn't done this and the rider has deliberately taken advantage of the situation to grab those tiles then you might indeed call it blatant cheating.  Maybe the rider didn't realise what had happened - in which case it isn't.

Now you're going to tell me that you were talking about something entirely different 

Nope, that's what I was referring to!

J

[JonBuoy]

If you view the ride in question on VeloViewer and look at the tiles that it is credited with I think that you owe Mr van Kamp an apology.



I also owe an apology to Mr VeloViewer for doubting his algorithm.

It would appear that the discrepancy* is in the heatmap.



*I really can't bring myself to call it an error.

[PeteBart]

If you view the ride in question on VeloViewer and look at the tiles that it is credited with I think that you owe Mr van Kamp an apology.



I also owe an apology to Mr VeloViewer for doubting his algorithm.

It would appear that the discrepancy* is in the heatmap.



*I really can't bring myself to call it an error.

Interesting ... the heatmap comes directly from a list of tiles exported from VV, not from points/lines. So I don’t have an explanation until I get in front of a big computer and can look it at properly. Possibly the VV algo has improved over time.

[salar55]

That might be due to a Gamin hiccup with the satellite reception. Have one issue where its a straight line missing out part of the track along the shore on east side of Loch Lomond side. Lost a tile where the track just caught a corner. Picked up a few new tiles doing the circuit of Loch Lomond, started Drymen east Loch Lomond (WHW) Inverarnan, Balloch and back to Drymen.

[L CC]

I have one of those. No hardship to walk Simonside again, though.

https://veloviewer.com/athletes/188837/activities/2305282663