@DaveF, your response is the more complete than mine.
Plus also "*=designated".
In my defence, I was trying for an uncluttered response.
regards, Alan

@DaveF, you have made many statements. Some of which are about a standard of mapping that you obviously hod dear. I can only wish my training was a detailed as yours.

You ask for evidence that the Kenett and Avon trail extends into Bristol.

There are several direct and a number indirect.

But so I can reply in context I first ask these questions of you.
Do you have any evidence:
1) ocean going boats went up the Avon from Bristol?
2) canal boats never (or hardly ever) went from Bath to Bristol and return?
3) goods were lightered from Bristol to a quay side at Bath to be transferred there to canal boats?
4) any other possibility?

If you do have such evidence, please provide so I can reply with mine couched appropriately.

kind regards

@trigpointg, thank you.

My question was "what interest do you have in the various matters you have raised?"

By way of example, are you only interested in all things near your home?
Or do you have training in mapping from, say, NCR? Or something completely different?

My purpose is to see what we have in common and to work forward from there.

kind regards

If you say so.

But you didn't ask a question about why.

They evidence the tag "highway=cycle and foot way".

My understanding is that tag has been removed and the advice is to replace with "=path"

You have now reverted to a redacted tag.

By your own actions you declare yourself to be lacksadaiscal and proves you are a mapper possibly caught in a time warp of long ago.

I suggest you consider refraining from taking a high position and add value to the entire process.

With kind regards

The quotes are not mine but from the wiki.

I agree, the gold standard is to have a name a regular intervals along a route.

When a route has been noted, explicitly or implicitly, by others then naming it in OSM is useful to those who come later. And that is what OSM is all about, users adding their bit of information.

So, nothing is a shot in the dark. The canal and the current trail are well indicated in online material.

kind regards

@DaveF, a tag of cycleway does imply cycling is the dominant use.
What is there is a path.
That a Euro, NCR or local cycle route is signposted only goes to confirm it is a shared path. And imagery indicates as many (if not more) on foot as on other means of getting about.
So "=Path" is the proper tagging with appropriate access tagged as "=yes".

First map what is "on the ground", then tag the access. To tag a shared path as "=cycleway" indicates a conscious (or, possibly, an unconscious) bias.

kind regards

@DaveF, thank you for your comments.
OSM is an international publication. This partly explains, from the Wiki, why "highway=sidewalk" is the preferred tag in suburban areas at least for what I, and I suspect you also, would be happier to tag as "=footpath".
You are correct for all "highway=*" tags "cycle=yes" must be present for a CycleRoute relation to be effective. In the same way as "foot=yes" is also required for HikingRoute relations, and so on.
It is my practice to note what access and route relations were marked before and ensure they are continued after any changes. If I have missed some, that is regretted but is not not intentional.

regards

@ndm.
A few days ago I have looked at the talk-GB mailing list for the first several days of July 2019.

A phrase I saw, and took to heart, is to map what is physically there.

I will continue to look from time to time.

Kind regards
I will

@SomeoneElse, thank you.
But what does this mean.
How does your comment relate to the Wiki, where licences are not mentioned.

King regards

@trigpoint, greetings
You have asked many questions recently. Too many at one sitting to answer.
I have asked you one.
I think it not unreasonable for you to answer that one.

kind regards

I should add, reversion of a complete change set for one apparent transgression (which can be fixed by itself, if the wiki description is wrong) seems like a blunderbus strategy.

Please look again at the image. You can see a sidewalk to the left and a road to the right. They are mapped separately.

regards

@ndm, thank you
I looked at available imagery.
I found an urban sidewalk with signage (blue roundel showing a bicycle and child holding the hand of an older person.
The location of the sidewalk indicates use is a foot way as well as a cycle way.
Neither use is dominant.
A route of any type is not recorded.

From the wiki, the tags are sidewalk, footpath or path with both foot and bicycle showing yes.

The wiki suggests sidewalk is most relevant in this situation. See osm.wiki/Sidewalks

regards

@ndm, thanks, a good suggestion

Does "duck test" in this context mean "if it quacks like a duck ..."?
Or does it mean something else?

regards

Of course, marking as a path does not in anyway affect any cycling route relationships. These will still be displayed in the "Cycle Map" layer and in "cycling.WayMarkedTrails.org": these are 'fed' from the OpenStreetMap database and can be as up to date as posts made within the last few minutes before an enquiry.

@ndm, I looked at your issue a day or two later.
In looking at the detail I think I can see where I "went wrong". To the east of Welsh Back, in what was quite a complex set of paths crossing one another and I had not noticed they related to physically separate ways. I hope you have looked at this very small segment and are now content.

A separate issue is to the east of Bristol Bridge and above Floating Harbour. The way here appears from the signage to be a shared path. And the available imagery shows many on foot and none on a bicycle, although a small number are parked, including several for hire.
Accordingly this way is more appropriately marked, in accordance with the wiki, as "=path". I would appreciate your observations.

Richard.
I use WayMarkedTrails.org and that normally gets eveything in order. But there was one route relation that not even josm's sort function would resolve. So did this manually so as to use all the features that 'Trails' offers.

I understand the principle you refer to, but within the bounds offered by relevant OSM wiki. One such is on naming things.

But here is the thing.

About 10 or so years ago and earlier there was a tag "highway=cycle and foot way" and its colour was blue. It was clear to me that this tag was meant to be inclusive of both means of getting about. Some route elements still have that tag. See Thames Path just west of the confluence of the Kennet with the Thames.

Then, about 10 years ago that (joint) tag was removed and replaced by "=footway", "=cycleway" and "=path".

In the current wiki it is only "=footway" that has the admonition to apply when the use is "mainly or exclusively" for walking. I think that admonition should apply to "=cycleway" also

What I suspect (guess) has happened (knowing human nature about change) is that some mappers have continued to use "cycleway" instead of "=path" as I believe was intended.

I have that belief as when saving additions to the OSM database recently I was asked to change "=cycle and foot way" tags that were proximate to my additions to ="path .

I have attempted to find the time line for these changes, without success. As I say based on that experience and what I can see and based on my understanding of human nature (I have had to introduce so much of it in my professional career) I come to these preliminary conclusions. It may be helpful in someone can give an authoritative chronology, but that may just be a rabbit hole.

I am happy to enter into a open discussion. But where the response is simply "that (cycleway) is just how we do it in the UK, then no thank you.

I am indebted to your information the focus of many cycle route mappers have a primary focus on that means of getting about. That may explain many of what I see as "funny" instances. For example: where a perfectly adequate suburban sidewalk without any signage about shared use is tagged as "=cycleway". I have no difficulty with the sidewalk elements being included in a cycling route relation (if that is actually appropriate) so it can be properly rendered in, say, cycling,waymarkedtrails.org. But it seems to me tagging a not very wide sidewalk as something else, I thinks, suggests a step too far.

I would be pleased to continue the discussion.

Richard, I have looked at the point that Phil (@trigpoint) makes. It seems his point has been corrected. But there are still many pre-existing discontinuities in the ordering of the elements elsewhere in this route.
kind regards, Alan

Phil, thank you and an important point part of my continuing learning.
I know I checked that the NCR4 relation was still present.
I have just now also looked at all the elements of NCR4 middle section Newbury to Severn Bridge and, at first blush, there are quite a number of discontinuities.
It was the around 12 discontinuities that got me started on the elements making up the route relation for K&A towpath.
Kind regards, Alan

Andy, hi and thank you for considered and considerate mention.

As a result of your mentions I have had another more detailed look at the last several kilometres within Reading: from Bridge Street to Thames River.

When first here, I encountered the name “Kennet Side cycle way” on sections that were not streets, to which I made the addition you note above.

My more detailed look shows much signage for a shared path (a rondel with a bicycle above a child holding the hand of an older person). Accordingly, I have deleted any name on sections between London Street and Sidmouth Street and “Kennet Side” from there to Thames River. I would appreciate your review.

I have also looked at the "atownsend" site and see the point you make. That site seems to apply only in the UK and Ireland. My long-distance walking takes me well beyond that area so I need a different solution. I carry an Android tablet with the relevant app being OSMAnd+: this works off-line. From experience I prefer to have the name of a recognised route at regular intervals when no other name is relevant, especially away from built up areas. I can expand on that point if you wish.

I have read the wiki section Name is the name only you refer to above. I take some refuge in the later part of the last paragraph starting “… however most names were invented at some point … “ and assert the many local users I have seen online will know it locally as the “canal tow path” or similar but may not know the name of the canal itself.
An analogy for my irregular usage is putting the names of train lines on the OSM website. For example, Great Western Main Line or South Wales Main Line. When at Paddington Station I do not recall signage for those names

Kind regards, Alan

PS: My current detailed interest in the K&A arises from an intention to walk from Avonmouth to Woolwich in 2020 (after making more progress / completing Via Francigena – Canterbury to Rome - started in 2018 !!!)

@trigpoint
I've had other stuff these past two days.
Thank you for your observations and you raise some interesting points for discussion. Rather than attempt to answer all at once I've made a first selection, so, with a slight alteration in sequence, here goes:
1) o) These changes are a bit out of step with the consensus of the UK community.
1) r) I believe my tags follow OSM documentation. And thus understood by readers wherever they live.
2) o) Highway=path is a bit of a late comer in OSM terms and is rarely used in the UK.
2 r) In my recent work I encountered the tag "Highway=Cycle and Foot Way" with a blue line when rendered. On saving my work OSM said these were redundant and should be replaced with "Highway=Path".

The practice of local territorial authorities where I live (city councils, for example) create shared paths in urban and rural settings and / or allow bicyles on certain footpaths. Some of these may be designated by a cycling club as a cycle route and maintain sihns (like the National Trail signs, but with a bike rather than an acorn). They are still shared paths and official signage is a walker above a bike. Often there is text below saying "Pedestrian Priority" and less often "Cyclists must give way".
It is clear to me, reading "official" web sites that the intention for ways such as Kennet and Avon Canal tow path is a shared path. To tag them 'cycleway' seems to not only an appropriation and, to an international reader, saying do not go here.
I would appreciate your observations. But to rely on UK customary use is not helpful. I can elaborate why I have that viewpoint in a latter discussion.
Phil, over to you.
Kind regards