Cycle Travel question!

Cycle-touring, Expeditions, Adventures, Major cycle routes NOT LeJoG (see other special board)
LittleGreyCat
Posts: 495
Joined: 7 Aug 2013, 8:31pm

Re: Cycle Travel question!

Postby LittleGreyCat » 15 Jun 2019, 6:11pm

Richard Fairhurst wrote:
gloomyandy wrote:Quick heads up. For some reason Google has decided that the .tcx file download from cycle.travel is not a safe file and will not allow the download to a Chromebook (well to my Chromebook at least). I had to disable the website protection feature to allow me to do it. It seems to be only the .tcx files that are being blocked, .gpx works fine. I've no idea why it is happening but you may find that it impacts other users as well. Possibly something to do with this https://www.ghacks.net/2019/04/11/chrom ... oads-soon/ Do you use https or http for tcx downloads?


Just to say that I raised this as an issue with Google's Search Console, and they've now come back to me and said "we've reviewed your site and determined that it's ok"! So hopefully these warnings will stop appearing.


I've noted that Windows Mail seems to lock any email with the cycle.travel URL in it.

Nothing to do with attached files because these are replies to emails I sent out with attachments and a link to a saved map.

Has the site crept onto a blacklist?

User avatar
RickH
Posts: 4434
Joined: 5 Mar 2012, 6:39pm
Location: Horwich, Lancs.

Re: Cycle Travel question!

Postby RickH » 19 Jun 2019, 9:30pm

Psamathe wrote:Many thanks but limited success. I'd already tried the cycle.travel profile speed but I'd not noticed the pace in the Garmin Connect web site import. I set the av speed for the course there and when the course loaded onto my GPS the summary had the correct speed I'd entered in the "pace". But during the ride the GPS still seemed to be using something around 15 mph. I'd entered 10 mph as an average to GC but with 14 miles to go the GPS was saying less than an hour for time to go and in the ETA.

It's not a crucial feature as 10 mph is fine and is easy mental arithmetic and my ETA is not crucial anyway.


I tried creating 2 identical routes, 1 in cycle.travel & the other in Ride with GPS, and then importing them to Garmin Connect (not adding an average speed, as per my previous post, in either case). When I sent them to my Garmin (Edge 1000 in my case) they both ended up showing a 15.54mph average speed. That suggests that it is something that the Garmin units make up (or Garmin Connect does it on exporting it) if there is no other info included. :?

I've not yet tried riding a course with a (sensible) pace set in GC to see what I get.

Woodtourer
Posts: 203
Joined: 23 Jan 2018, 1:51pm

Re: Cycle Travel question!

Postby Woodtourer » 21 Jun 2019, 6:10pm

Current challenge/problem.
We are currently cycling in Poland. When we use the Find lodging option is seems like after looking at one option it freezes up and I have to refresh the page and start over.

User avatar
mjr
Posts: 13265
Joined: 20 Jun 2011, 7:06pm
Location: Norfolk or Somerset, mostly
Contact:

Re: Cycle Travel question!

Postby mjr » 24 Jun 2019, 1:02pm

When deciding which side of the road to cycle on (when the contraflow side has a two-way cycleway), does cycle.travel take into account the number of road crossings? The number of toucan lights? The traffic volume of the road being crossed? I ask because I feel the default routing choice in https://cycle.travel/map/journey/110871 is incorrect to prefer the contraflow when you've got to cross back and join the flow anyway.
MJR, mostly pedalling 3-speed roadsters. KL+West Norfolk BUG incl social easy rides http://www.klwnbug.co.uk
All the above is CC-By-SA and no other implied copyright license to Cycle magazine.

Richard Fairhurst
Posts: 1363
Joined: 2 Mar 2008, 4:57pm
Location: Charlbury, Oxfordshire

Re: Cycle Travel question!

Postby Richard Fairhurst » 24 Jun 2019, 1:29pm

Woodtourer wrote:Current challenge/problem.
We are currently cycling in Poland. When we use the Find lodging option is seems like after looking at one option it freezes up and I have to refresh the page and start over.


That's strange - I can't get that to happen but other people have reported something similar (but maybe it might be specific to something I can't test on?). Can you give me an exact step-by-step - i.e. plan this route, click this, click that - so I can try and track it down?

mjr wrote:When deciding which side of the road to cycle on (when the contraflow side has a two-way cycleway), does cycle.travel take into account the number of road crossings? The number of toucan lights? The traffic volume of the road being crossed? I ask because I feel the default routing choice in https://cycle.travel/map/journey/110871 is incorrect to prefer the contraflow when you've got to cross back and join the flow anyway.


It does all of that, yes (though traffic volume is approximated when it comes to calculating turn costs). I think the issue in this case is probably that c.t simply likes off-carriageway facilities, doesn't like paint, and doesn't like dual carriageways. In OSM, the junction is mapped with a dual carriageway with a painted lane on the north side. Google Street View suggests that's not actually the case, and that the north-side provision is a fairly sketchy shared-use path, but I don't know the location to know if that's changed.

Unfortunately routing through horrible UK urban junction design can be a losing battle at times!
cycle.travel - maps, journey-planner, route guides and city guides

User avatar
mjr
Posts: 13265
Joined: 20 Jun 2011, 7:06pm
Location: Norfolk or Somerset, mostly
Contact:

Re: Cycle Travel question!

Postby mjr » 24 Jun 2019, 3:39pm

Richard Fairhurst wrote:
mjr wrote:When deciding which side of the road to cycle on (when the contraflow side has a two-way cycleway), does cycle.travel take into account the number of road crossings? The number of toucan lights? The traffic volume of the road being crossed? I ask because I feel the default routing choice in https://cycle.travel/map/journey/110871 is incorrect to prefer the contraflow when you've got to cross back and join the flow anyway.


It does all of that, yes (though traffic volume is approximated when it comes to calculating turn costs). I think the issue in this case is probably that c.t simply likes off-carriageway facilities, doesn't like paint, and doesn't like dual carriageways.

I wondered about that, but in the nearby "mini-Holland", it refuses to use some of the kerbed cycleways such as via point 1 in https://cycle.travel/map/journey/110904 - what's the decision-making there?

Richard Fairhurst wrote:In OSM, the junction is mapped with a dual carriageway with a painted lane on the north side. Google Street View suggests that's not actually the case, and that the north-side provision is a fairly sketchy shared-use path, but I don't know the location to know if that's changed.

I think Street View is more current, but I have not confirmed that recently - but I'd prefer joining a painted lane on the north side 100m earlier to probably waiting for an additional two push-button crossings, so even as mapped, I'm not sure that's the correct decision there.

Richard Fairhurst wrote:Unfortunately routing through horrible UK urban junction design can be a losing battle at times!

Riding it isn't always great, either...
MJR, mostly pedalling 3-speed roadsters. KL+West Norfolk BUG incl social easy rides http://www.klwnbug.co.uk
All the above is CC-By-SA and no other implied copyright license to Cycle magazine.

Richard Fairhurst
Posts: 1363
Joined: 2 Mar 2008, 4:57pm
Location: Charlbury, Oxfordshire

Re: Cycle Travel question!

Postby Richard Fairhurst » 24 Jun 2019, 4:51pm

mjr wrote:I wondered about that, but in the nearby "mini-Holland", it refuses to use some of the kerbed cycleways such as via point 1 in https://cycle.travel/map/journey/110904 - what's the decision-making there?


Ah, that's an interesting one. :)

That's a challenge that basically springs from OSM's data model. Because roads are mapped as centrelines, the OSM data shows that as two sharp turns (near 90° in each case). c.t has a slight penalty for each turn, so in this case the benefit of taking a few metres of cycleway is outweighed by the cost of the two turns.

Ideally OSM needs some way to represent this as "this isn't actually a turn, because cyclists would already be on that side of the road". In reality I'm not too worried about the routing choice here - if you were cycling along that road, you'd pop onto the kerbed cycleway anyway - and the biggest challenge is getting the turn instructions right: something saying "turn left onto cycleway" would clearly be wrong. I've done some work on improving this but there's more still to do. (Looking at CycleStreets, it has similar issues too.)
cycle.travel - maps, journey-planner, route guides and city guides

Psamathe
Posts: 9860
Joined: 10 Jan 2014, 8:56pm

Re: Cycle Travel question!

Postby Psamathe » 14 Aug 2019, 11:58am

RickH wrote:
Psamathe wrote:Many thanks but limited success. I'd already tried the cycle.travel profile speed but I'd not noticed the pace in the Garmin Connect web site import. I set the av speed for the course there and when the course loaded onto my GPS the summary had the correct speed I'd entered in the "pace". But during the ride the GPS still seemed to be using something around 15 mph. I'd entered 10 mph as an average to GC but with 14 miles to go the GPS was saying less than an hour for time to go and in the ETA.

It's not a crucial feature as 10 mph is fine and is easy mental arithmetic and my ETA is not crucial anyway.


I tried creating 2 identical routes, 1 in cycle.travel & the other in Ride with GPS, and then importing them to Garmin Connect (not adding an average speed, as per my previous post, in either case). When I sent them to my Garmin (Edge 1000 in my case) they both ended up showing a 15.54mph average speed. That suggests that it is something that the Garmin units make up (or Garmin Connect does it on exporting it) if there is no other info included. :?

I've not yet tried riding a course with a (sensible) pace set in GC to see what I get.

I've reported the issue to Garmin who are investigating (they have replied, asked for more details so they can reproduce the issue). I'll report further if/when I get more responses from Garmin.

Ian

doodah
Posts: 89
Joined: 9 Aug 2008, 2:26pm

Re: Cycle Travel question!

Postby doodah » 16 Aug 2019, 8:51am

Morning All,

First off, I have used cycle travel quite a lot and am very content with it.

I know how to use it but would like to understand how it arrives at some of its planning.

eg. If a route is asked for from Nantes to Constanta, CT gives a route at variance to EV6 - which is for cyclists. Not a problem for me but why?

If I plan a route from A to B and it passes through a mountain range with 3 passes, will CT give me the shortest route with, say the highest pass, or will it give me the lowest pass with a 50km 'detour'.

I only ask because I would like to put my complete faith in CT for providing me with an optimum route for a laden touring cyclist. I appreciate that I can carve the route up and get elevation profiles that show % elevations but this can be tedious and then generate doubt into what CT is really offering on other parts of the route.

Again, I have no beef with CT at all and I also have a limited knowledge and understanding of mapping technicalities and jargon.

Apologies if this has been raised before and will readily accept a redirect link.

Thanks

Doodah

Psamathe
Posts: 9860
Joined: 10 Jan 2014, 8:56pm

Re: Cycle Travel question!

Postby Psamathe » 16 Aug 2019, 10:11am

Psamathe wrote:
RickH wrote:
Psamathe wrote:Many thanks but limited success. I'd already tried the cycle.travel profile speed but I'd not noticed the pace in the Garmin Connect web site import. I set the av speed for the course there and when the course loaded onto my GPS the summary had the correct speed I'd entered in the "pace". But during the ride the GPS still seemed to be using something around 15 mph. I'd entered 10 mph as an average to GC but with 14 miles to go the GPS was saying less than an hour for time to go and in the ETA.

It's not a crucial feature as 10 mph is fine and is easy mental arithmetic and my ETA is not crucial anyway.


I tried creating 2 identical routes, 1 in cycle.travel & the other in Ride with GPS, and then importing them to Garmin Connect (not adding an average speed, as per my previous post, in either case). When I sent them to my Garmin (Edge 1000 in my case) they both ended up showing a 15.54mph average speed. That suggests that it is something that the Garmin units make up (or Garmin Connect does it on exporting it) if there is no other info included. :?

I've not yet tried riding a course with a (sensible) pace set in GC to see what I get.

I've reported the issue to Garmin who are investigating (they have replied, asked for more details so they can reproduce the issue). I'll report further if/when I get more responses from Garmin.

Ian

Well that was embarrassing. Maybe not. I can't test at the moment as I can't go out for a ride (dental operation ...) but on creating a test course, loading and "riding" it the initial "Time to Dest" is now using the "Pace" entered in https://connect.garmin.com.

I wonder if this happens to have just been fixed as since returning from tour where I noticed this issue the 1030 has had a firmware upgrade. It's not any changes to cycle.travel as I also re-checked a courses I was using when I 1st raised this in this thread and the one I re-checked is now using the "Pace" (went back in Garmin Connect and found the courses I'd set the "pace" to 10 mph in following having the setting pointed out to me here).

But all I've done on the 1030 is to start the course (i.e. load to 1030 and select "Ride" and check the initial Time to Dest) - not actually ridden the course and I've no idea if the Time to Dest might then revert to 15.?? mph. But I'm not going to assume it does not work in the absence of evidence and when the test suggest it does now work.

Ian

Richard Fairhurst
Posts: 1363
Joined: 2 Mar 2008, 4:57pm
Location: Charlbury, Oxfordshire

Re: Cycle Travel question!

Postby Richard Fairhurst » 16 Aug 2019, 10:46am

doodah wrote:Morning All,

First off, I have used cycle travel quite a lot and am very content with it.

I know how to use it but would like to understand how it arrives at some of its planning.

eg. If a route is asked for from Nantes to Constanta, CT gives a route at variance to EV6 - which is for cyclists. Not a problem for me but why?

If I plan a route from A to B and it passes through a mountain range with 3 passes, will CT give me the shortest route with, say the highest pass, or will it give me the lowest pass with a 50km 'detour'.

I only ask because I would like to put my complete faith in CT for providing me with an optimum route for a laden touring cyclist. I appreciate that I can carve the route up and get elevation profiles that show % elevations but this can be tedious and then generate doubt into what CT is really offering on other parts of the route.


It weighs up a lot of different factors: surface type, climb (both total and gradient), signposted routes, motor traffic, minimising turns, and so on. So there's no hard and fast answer - it won't necessarily choose the lowest pass or the shortest route, but will look for a "best of both worlds" route. There's a few thousand lines of code that work out the weightings.

That's why, although it does like EuroVelo routes (and National Cycle Network routes, and any equivalent), it doesn't slavishly follow them: it will often cut off long doglegs or divert past rough surfaces. For example, for this big loop in the Danube, it will cut off the loop to save 30 miles although it does entail more climbing. If you do want to do the loop, you can drag the route back to the river. (I'm in the early stages of looking at options to make it easier to follow signposted routes faithfully if that's what you want to do.)

Generally for a long, laden route I'd suggest planning an outline route with via points in the places you want to go. Then split it up into stages for each few days - you can do this quite easily by putting a via point at your overnight stop, right-clicking the via point and selecting "Delete after", then saving under a different name. You can then fine-tune the legs by dragging the route around, using the "Paved only" option, and so on.


Ian - glad it's working now!
cycle.travel - maps, journey-planner, route guides and city guides

doodah
Posts: 89
Joined: 9 Aug 2008, 2:26pm

Re: Cycle Travel question!

Postby doodah » 16 Aug 2019, 11:26am

Thanks Richard,

[quote]It weighs up a lot of different factors: surface type, climb (both total and gradient), signposted routes, motor traffic, minimising turns, and so on. So there's no hard and fast answer - it won't necessarily choose the lowest pass or the shortest route, but will look for a "best of both worlds" route. There's a few thousand lines of code that work out the weightings.
/quote]

That is well explained and confirmed what I was tentatively thinking.

[quote]That's why, although it does like EuroVelo routes (and National Cycle Network routes, and any equivalent), it doesn't slavishly follow them: it will often cut off long doglegs or divert past rough surfaces. For example, for this big loop in the Danube, it will cut off the loop to save 30 miles although it does entail more climbing. If you do want to do the loop, you can drag the route back to the river. (I'm in the early stages of looking at options to make it easier to follow signposted routes faithfully if that's what you want to do.)/quote]

I don't believe in slavishly following the route as these are sometimes manufactured to take the touring cyclist into a town of no consequence in an effort to make said cyclist unnecessarily part with his money. The big loop that was omitted is precisely what I had spotted. Cutting the loop off means missing the view of the Iron Gates at Drobeta Turna - so you can't have it both ways.

[quote]Generally for a long, laden route I'd suggest planning an outline route with via points in the places you want to go. Then split it up into stages for each few days - you can do this quite easily by putting a via point at your overnight stop, right-clicking the via point and selecting "Delete after", then saving under a different name. You can then fine-tune the legs by dragging the route around, using the "Paved only" option, and so on.
/quote]

This is what I am doing and I thought I was being pedantic in not putting my trust in CT. You have have now confirmed that it gives a good outline for a long distance that might need to be tweaked - depending on your circumstances, time, money, load, physical condition, reason for travelling, etc. A shorter route is easy to identify instantly if one needs to modify it, whereas the long route has to be broken up into stages and then looked at. Good advice.

Thanks once again for your help. I realise - as others on this forum also do - that you must be a busy man, not needing a bombardment of continual 'simple' questions. Could it not be possible to expand the help section to give further explanations on 'How to...' etc? I would be readily available to help you in this and a guaranteed Horlicks can be assured. In fact, it could be subtitled 'How not to....' Would make for a lively forum with the ensued chaos:)

Appreciated

Doodah

Richard Fairhurst
Posts: 1363
Joined: 2 Mar 2008, 4:57pm
Location: Charlbury, Oxfordshire

Re: Cycle Travel question!

Postby Richard Fairhurst » 16 Aug 2019, 3:16pm

doodah wrote:Thanks once again for your help. I realise - as others on this forum also do - that you must be a busy man, not needing a bombardment of continual 'simple' questions. Could it not be possible to expand the help section to give further explanations on 'How to...' etc? I would be readily available to help you in this and a guaranteed Horlicks can be assured. In fact, it could be subtitled 'How not to....' Would make for a lively forum with the ensued chaos:)


100% a good idea. It's probably one of those scenarios where I haven't put aside the time to write the help pages, but in the long run it'll save time, so I really should...

Incidentally, two gotchas worth noting for the Danube. Don't entirely trust c.t's elevation read-out: where a road is very close to a steep valley side, the elevation data is sometimes not precise enough and will give you an averaged midpoint. (Higher-resolution elevation data does exist but it's quite memory-hungry and I need to experiment with compression options.) Second, the source OSM data is not always 100% reliable in Eastern Europe for surfaces: on occasion unpaved roads aren't recorded as such. So if you're planning an excursion on minor roads away from the river, it might be worth double-checking on satellite imagery that the road looks tarmacked.
cycle.travel - maps, journey-planner, route guides and city guides

User avatar
andrew_s
Posts: 4835
Joined: 7 Jan 2007, 9:29pm
Location: Gloucestershire

Re: Cycle Travel question!

Postby andrew_s » 16 Aug 2019, 9:05pm

doodah wrote:Thanks Richard,

[quote]It weighs up a lot of different factors: surface type, climb (both total and gradient), signposted routes, motor traffic, minimising turns, and so on. So there's no hard and fast answer - it won't necessarily choose the lowest pass or the shortest route, but will look for a "best of both worlds" route. There's a few thousand lines of code that work out the weightings.
/quote]

Appreciated

Doodah

When dividing up quoted text to interpose separate replies to different parts, there's a Quote button on the "post a reply" toolbar that will avoid the quote errors in your reply.
Just highlight a block of text, and click the Quote button, and it will add the open and close quote tags for the highlighted block.

doodah
Posts: 89
Joined: 9 Aug 2008, 2:26pm

Re: Cycle Travel question!

Postby doodah » 16 Aug 2019, 9:36pm



Thanks for the info.

Highlighted your text and pressed the quote button and got what you see on the first line.



Did it again - just to be sure I was doing it wrongly.

As I said, my help would make for a lively debate.