OS vs OSM for cycling

General cycling advice ( NOT technical ! )
Richard Fairhurst
Posts: 1597
Joined: 2 Mar 2008, 4:57pm
Location: Charlbury, Oxfordshire

OS vs OSM for cycling

Postby Richard Fairhurst » 30 Sep 2020, 10:19am

(split off into a new topic to avoid derailing the KAW thread)

Zulu Eleven wrote:I really don’t understand why people are so wedded to OSM as a mapping source.

The rights of way data on there is very often utter sphericals.

For £20 quid a year you get access to the whole country on OSmaps - an organisation that spends hundreds of millions of pounds every year developing the best mapping in the world, that you can actually navigate with, and if you know how to use one can even work out where you are with when your GPS gets jammed by the Russians

You even also get a downloadable OS Map with every paper map you buy.


Ok, I'll bite.

I'm a big fan of Ordnance Survey and always have been - from here I can reach out to a shelf full of modern OS maps, a complete collection of the New Popular Edition, and CD-ROMs of 1:50k from about 15 years ago. Their 1:50k cartography is probably the best mass-market mapping in the world (with due deference to Herr Imhof). OS maps are great.

If what you want is to look at a complete map of public rights of way in England & Wales, then OS is unbeatable. OS (by law) gets updates directly from PRoW officers, while OSM relies on volunteer edits. OS, as you say, has a budget of hundreds of millions to spend on the UK alone; OSM's annual budget is about £200k for the whole world. If a display map of PRoWs is what you want, then OS will always win hands down.

But there are many, many other considerations for cycle mapping.

OS PRoW data isn't routable (other than in National Parks, weirdly). The footpath and bridleway lines drawn on a Landranger or Explorer map that you see are just that; lines drawn on a map. OS doesn't have a database saying "this bridleway connects to this byway here". That's why there are no bike routing websites that route using OS data.

OS has a very partial selection of points of interest. For example, there are two public water points on the Ridgeway between Streatley and Ogbourne. Neither of these are marked on OS maps. (They're not mentioned or marked in the KAW guidebook, either, which slightly surprised me.) OS has some pubs but no cafes, let alone bike shops.

OS maps only come at the scales they choose. 1:25k is ideal for walking and 1:50k is ok for some types of cycling. For longer distance cycling then a scale of 1:80k—1:125k-ish is often best (Richard Oliver, the unofficial historian of the OS, has written extensively on this) - but OS cartography at 1:100k is an ugly mess, and their data products don't permit you to make your own countryside mapping at that scale, even if you could afford them.

OS is not great at showing waymarked cycle routes. (It has the NCN, sometimes, a bit out of date.) OS doesn't even attempt to map urban cycling infrastructure. OS doesn't record path surfaces, or obstacles like stiles and gates. And so on.

OSM fixes all of these. Not consistently, because it's made by volunteers. But the fact that almost all cycle mapping tech is now OSM-based - Garmin, Wahoo, Strava, countless mapping sites including RWGPS, CycleStreets, Komoot and (ahem) cycle.travel - illustrates that it's good enough for most purposes. It's clearly even good enough for Cycling UK, given that the King Alfred's Way page uses OSM (though given that CUK is using OSM's volunteer server resources for free, it would be kind to have the mandatory credit in a font size you can actually read). There is much more to OSM data than is surfaced on openstreetmap.org itself, and a smart app or site can really bring out the richness of it.

And I'm a bit bemused by your point about the Russians - if you have offline mapping on your phone (or Garmin or whatever), you can still view it without a GPS signal whether the data is from OS or OSM.
cycle.travel - maps, journey-planner, route guides and city guides

sjs
Posts: 865
Joined: 24 Jan 2010, 10:08pm
Location: Hitchin

Re: OS vs OSM for cycling

Postby sjs » 30 Sep 2020, 1:06pm

And the pub symbol on OS explorer maps is stupidly large and vague, about 80m in size and often a similar distance from the real location. This is the most important shortcoming of OS maps.

User avatar
Sweep
Posts: 6609
Joined: 20 Oct 2011, 4:57pm
Location: London

Re: OS vs OSM for cycling

Postby Sweep » 30 Sep 2020, 1:56pm

Great points Richard.
Sweep

mattsccm
Posts: 3463
Joined: 28 Nov 2009, 9:44pm

Re: OS vs OSM for cycling

Postby mattsccm » 30 Sep 2020, 3:00pm

Surely they are complementary rather than conflicting? They have their good and bad points.

OS has the detail but it is too much at times whereas OSM can be adequate or lacking.
OS can be safely relied on to be virtually definitive . (or as good as its updates allow) .OSM can have inaccuracies.
Rather daft to make sweeping statements either way . Bit like comparing apples and pears.
If you don't like one, use the other.

User avatar
TrevA
Posts: 2520
Joined: 1 Jun 2007, 9:12pm
Location: Nottingham

Re: OS vs OSM for cycling

Postby TrevA » 30 Sep 2020, 3:04pm

I paid for the OS app on my phone last year, which gives access to 1:25000 and 1:50000 maps for the whole country. It tells you where you are on the map and you can zoom in or out to determine where you want to go. I’ve found it invaluable on walks and rides over the past year. I plan routes on Garmin Connect or MapMyRide but I always check the route against OS, as the others will often auto-route you down an overgrown bridleway.
Last edited by TrevA on 30 Sep 2020, 3:09pm, edited 1 time in total.
A cart horse trapped in the body of a man.
http://www.jogler2009.blogspot.com

User avatar
mjr
Posts: 16225
Joined: 20 Jun 2011, 7:06pm
Location: Norfolk or Somerset, mostly
Contact:

Re: OS vs OSM for cycling

Postby mjr » 30 Sep 2020, 3:05pm

mattsccm wrote:Surely they are complementary rather than conflicting? They have their good and bad points.

Yes, indeed.

OS has the detail but it is too much at times whereas OSM can be adequate or lacking. OS can be safely relied on to be virtually definitive . (or as good as its updates allow) .OSM can have inaccuracies.

OK then: look at a bridleway on OS and tell me if it's tarmac? If it's obstructed by barriers? What's its reference number to report a problem with it? I can often find all those on OSM.

As for accuracy, OS seem to believe whatever updates councils send them, but many councils seem to have no naffing clue what they've built for cycling and that gets reflected in the OS map for at least a while, showing stuff that doesn't exist and not showing stuff that's been built. OSM editors may not catch everything, but most mistakes seem to get fixed fairly quickly.

As I posted elsewhere, at least we can submit updates to OSM and we don't have to pay for full access.

Rather daft to make sweeping statements either way

O! The irony! ;)

If you don't like one, use the other.

Amen.
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.

Queens Park Pete
Posts: 42
Joined: 7 Aug 2020, 8:08pm
Location: Brighton

Re: OS vs OSM for cycling

Postby Queens Park Pete » 30 Sep 2020, 9:58pm

I spent half of my lockdown "improving" my 1:50 000 paper maps by adding additional highlighting to their bridleways, white roads and some footpaths (if there were no stiles and they were actually rideable tarmac) once I had checked them on cycle.travel and streetview/geograph and then ridden them. Well there wasn't a whole lot happening and I now know whether they are summer routes or all year escapes for my Panorama! I then added these to my Master Maps- the half finished blueTravel Tour OS series. This series could have been a decent replacement to the Bartholomew 1:100 000 maps but are so crudely drawn they have only become useful to me after I have added the rideable bridleways to them by hand. And of course half this series has been discontinued and some of others are on a random scale of up to 1:175 000!

My pet hate on my beloved OS Landrangers are the white and pecked lines which are sometimes roads and lanes closed to the public but other times useful, concealed rights of way. Occasionally Cycle.travel spots these- but also regularly directs me away from useful connections- presumably because OSM hasnot registered a continuation of a route even when it has drawn it !

Of course the best cycle maps are the ancient Tourist Maps with colour contours, one inch to a mile, old but it will do the speed I ride. Works of art. Shame they never did a proper one of Snowdonia!

mjr wrote:
mattsccm wrote:Surely they are complementary rather than conflicting? They have their good and bad points.

Yes, indeed.

OS has the detail but it is too much at times whereas OSM can be adequate or lacking. OS can be safely relied on to be virtually definitive . (or as good as its updates allow) .OSM can have inaccuracies.

OK then: look at a bridleway on OS and tell me if it's tarmac? If it's obstructed by barriers? What's its reference number to report a problem with it? I can often find all those on OSM.

As for accuracy, OS seem to believe whatever updates councils send them, but many councils seem to have no naffing clue what they've built for cycling and that gets reflected in the OS map for at least a while, showing stuff that doesn't exist and not showing stuff that's been built. OSM editors may not catch everything, but most mistakes seem to get fixed fairly quickly.

As I posted elsewhere, at least we can submit updates to OSM and we don't have to pay for full access.

Rather daft to make sweeping statements either way

O! The irony! ;)

If you don't like one, use the other.

Amen.

JakobW
Posts: 412
Joined: 9 Jun 2014, 1:26pm
Location: The glorious West Midlands

Re: OS vs OSM for cycling

Postby JakobW » 1 Oct 2020, 2:30pm

Richard Fairhurst wrote:OS maps only come at the scales they choose. 1:25k is ideal for walking and 1:50k is ok for some types of cycling. For longer distance cycling then a scale of 1:80k—1:125k-ish is often best (Richard Oliver, the unofficial historian of the OS, has written extensively on this) - but OS cartography at 1:100k is an ugly mess, and their data products don't permit you to make your own countryside mapping at that scale, even if you could afford them.


I'm always bemused at the awfulness of the OS Tour maps when compared to the excellence of their 1:25k and 1:50k mapping; is it just that nobody at the agency cared about them? Do you have a link for any of the Richard Oliver pieces on cycling cartography (are they online, or in print?) - I wouldn't mind having a look, as they sound interesting.

Richard Fairhurst
Posts: 1597
Joined: 2 Mar 2008, 4:57pm
Location: Charlbury, Oxfordshire

Re: OS vs OSM for cycling

Postby Richard Fairhurst » 1 Oct 2020, 3:22pm

There are (at least!) two relevant Richard Oliver pieces in back issues of Sheetlines, the journal of the Charles Close Society:

https://s3-eu-west-1.amazonaws.com/char ... e/Sh51.pdf

https://s3-eu-west-1.amazonaws.com/shee ... /Sh104.pdf

He also did a piece in the Cartographic Journal for June 2001 on "Mapping for Cycle Touring in Britain; Past, Present, and a Possible Future", which is a great read but not available unless you pay £££ to the journal publishers.
cycle.travel - maps, journey-planner, route guides and city guides

Queens Park Pete
Posts: 42
Joined: 7 Aug 2020, 8:08pm
Location: Brighton

Re: OS vs OSM for cycling

Postby Queens Park Pete » 1 Oct 2020, 5:01pm

Thanks for the link. I had forgotten about the Charles Close Society. And my wife thinks I am a map obsessive!

Concur with the writer's take on both articles- especially the delightful croydecycle maps (The walking guides are fab too). I would be interested in his take on the Tour maps-as far as I can see they have been designed for people who can't read maps and don't want to wear their glasses- I actually think the Green OS Road series are better for long distance cycling than the Tour! (but you need your glasses on).

It is also interesting to note that IGN in France seem to be dropping their 1:50 000 series and pushing 1:25 000- superb but expensive and no good for cycling any distance and the 1: 100 000 series- nice overall but lacking in enough topographic info for cycling really and rather vague about minor roads and rideable tracks (at least they number them unlike our Tour series). I have got a subscription via Viewranger which seems to get the best of both worlds and the Saturday Walking Club https://www.walkingclub.org.uk/maps/ign-france/ has links to free IGN for planning .

Traction_man
Posts: 36
Joined: 10 Jan 2020, 5:30pm
Location: Bangor NI

Re: OS vs OSM for cycling

Postby Traction_man » 1 Oct 2020, 5:12pm

Richard Fairhurst wrote:
1:80k—1:125k-ish is often best


for those interested, the 'historic' OS half-inch to one-mile series is featured on the CCS website:

https://www.charlesclosesociety.org/halfinch

here in Ireland we had half-inch OS maps covering NI and RoI until relatively recently, and very nice they are too (actually 1:126,720) and a handy scale for cycling as well as having relief shading. I wish I had more!

my Dad used Bartholemew's GB half-inch maps when he cycled as a youngster in the 1940s/1950s, I still have his old collection of these lovely maps, though they can now be accessed online via National Library of Scotland (https://maps.nls.uk/geo/explore/#zoom=5 ... rs=192&b=1).

all the best,

Keith

Zulu Eleven
Posts: 64
Joined: 26 Oct 2018, 9:25pm

Re: OS vs OSM for cycling

Postby Zulu Eleven » 1 Oct 2020, 8:26pm

Richard Fairhurst wrote:
Ok, I'll bite.

I'm a big fan of Ordnance Survey and always have been - from here I can reach out to a shelf full of modern OS maps, a complete collection of the New Popular Edition, and CD-ROMs of 1:50k from about 15 years ago. Their 1:50k cartography is probably the best mass-market mapping in the world (with due deference to Herr Imhof). OS maps are great.

If what you want is to look at a complete map of public rights of way in England & Wales, then OS is unbeatable. OS (by law) gets updates directly from PRoW officers, while OSM relies on volunteer edits. OS, as you say, has a budget of hundreds of millions to spend on the UK alone; OSM's annual budget is about £200k for the whole world. If a display map of PRoWs is what you want, then OS will always win hands down.


Given that this discussion originally came up on an off-road thread, then yes, precisely that, we were discussing the suitability of a mapping format that allows you to navigate off-road along rights of way and other routes in rural areas, rather than navigation for road cycling.

But there are many, many other considerations for cycle mapping.

OS PRoW data isn't routable (other than in National Parks, weirdly). The footpath and bridleway lines drawn on a Landranger or Explorer map that you see are just that; lines drawn on a map. OS doesn't have a database saying "this bridleway connects to this byway here". That's why there are no bike routing websites that route using OS data.


Complex discussion starts here about whether Local Authorities are fulfilling their duties under New Roads and Street Works Act and electronic street works register regulations, some counties include PROW in their SWR, some don't, SWR handbook suggests they should. OSM is just as poor here as it often falls down on Byways and ORPA.

OS has a very partial selection of points of interest. For example, there are two public water points on the Ridgeway between Streatley and Ogbourne. Neither of these are marked on OS maps. (They're not mentioned or marked in the KAW guidebook, either, which slightly surprised me.) OS has some pubs but no cafes, let alone bike shops.


But as we've all seen the problem is the reliability of the databases - pub closures are a huge problem (and i expect will lead to the removal of that data set from the maps), but nothing as compared to bike shop closures. similarly, theres no reliable dataset for water points, they are the properly of the landowner and can be withdrawn at any time. As a general rule in life, if I have the choice of no information or potentially duff information - I believe I am safer in the knowledge that that only one of those two is likely to get me killed.

OS maps only come at the scales they choose. 1:25k is ideal for walking and 1:50k is ok for some types of cycling. For longer distance cycling then a scale of 1:80k—1:125k-ish is often best (Richard Oliver, the unofficial historian of the OS, has written extensively on this) - but OS cartography at 1:100k is an ugly mess, and their data products don't permit you to make your own countryside mapping at that scale, even if you could afford them.


Again, may be relevant for road cycling, but 1-80k plus simply cannot be used safely for off-road navigation, and this conversation was predicated upon that.

OS is not great at showing waymarked cycle routes. (It has the NCN, sometimes, a bit out of date.) OS doesn't even attempt to map urban cycling infrastructure. OS doesn't record path surfaces, or obstacles like stiles and gates. And so on.


I'm going to argue that the fault there lies with Sustrans, who for years refused to supply their NCN mapping data to other sources for financial reasons, and only finally entered into agreement with OS last year.

OSM fixes all of these. Not consistently, because it's made by volunteers.


exactly, not consistently - garbage in, garbage out.

But the fact that almost all cycle mapping tech is now OSM-based - Garmin, Wahoo, Strava, countless mapping sites including RWGPS, CycleStreets, Komoot and (ahem) cycle.travel - illustrates that it's good enough for most purposes.


apart from when it isn't. eg. locally Komoot will route me across a red flagged firing range. Try plumbing KAW into Komoot (even pressing the button telling it to not move you off the recorded route) and outside reading it will route you into the A33 dual carriageway rather than the Byway alongside
(all very reminiscent of what happens with drivers following satnavs off cliffs, into rivers or along railway lines - I reiterate the point I made earlier, Where I have the choice of no information or potentially duff information, only one of those two is likely to get me killed)

It's clearly even good enough for Cycling UK, given that the King Alfred's Way page uses OSM (though given that CUK is using OSM's volunteer server resources for free, it would be kind to have the mandatory credit in a font size you can actually read).

If it were anything like my choice then I'd be happy to put memberships up a few extra pence to pay for OS data, but we've all seen what a bunfight that would turn into :D

There is much more to OSM data than is surfaced on openstreetmap.org itself, and a smart app or site can really bring out the richness of it.
an argument that applies even more so to OS data, which goes far beyond what is available via consumer sites.

And I'm a bit bemused by your point about the Russians - if you have offline mapping on your phone (or Garmin or whatever), you can still view it without a GPS signal whether the data is from OS or OSM.
You can view the map, but being able to actually fix your location and navigate from the information contained within it is different, let alone a system that works seamlessly across platforms and hard media

LittleGreyCat
Posts: 810
Joined: 7 Aug 2013, 8:31pm

Re: OS vs OSM for cycling

Postby LittleGreyCat » 1 Oct 2020, 8:48pm

Very useful information about what isn't in OS maps.

I was riding the other day using my Garmin Edge Explore and looking at the map and thinking that it would be nice with a little more detail, like contour lines and local features.

I usually use it with a pre planned map for navigation, but it can also navigate you to a chosen point and the lack of enough planning detail would be enough to scupper that.

slowster
Posts: 1686
Joined: 7 Jul 2017, 10:37am

Re: OS vs OSM for cycling

Postby slowster » 1 Oct 2020, 9:15pm

The mass trespass on Kinder Scout helped to make rights of way official which had previously had not been recognised by law, and as the official mapping body the OS put those rights of way on its maps. If you don't think the current rights of way network is adequate, then you might not regard OS maps PROW information so positively, because it reinforces the idea that the PROW (outside Scotland) are now set in stone with little or no prospect of significant change or improvement.

I think OSM may act as a driver to improve the PROW network, possibly even to pass legislation giving similar countryside access as Scotland already has. It might ultimately have a similar impact to the Kinder Scout trespass.

Zulu Eleven
Posts: 64
Joined: 26 Oct 2018, 9:25pm

Re: OS vs OSM for cycling

Postby Zulu Eleven » 1 Oct 2020, 10:00pm

LittleGreyCat wrote:Very useful information about what isn't in OS maps.

I was riding the other day using my Garmin Edge Explore and looking at the map and thinking that it would be nice with a little more detail, like contour lines and local features.

I usually use it with a pre planned map for navigation, but it can also navigate you to a chosen point and the lack of enough planning detail would be enough to scupper that.


Indeed. In fact contour lines have to be one of the most important considerations for both navigation and cycle route planning.

Perhaps I’m just lucky enough to have been properly trained in how to use a map- ie plonked somewhere on a hill and taught how to work out my location from what I could see and the map in front of me, and then navigate to another point. The OS map gives you a true idea of landform in a way other sources don’t.