Re: [OSM-talk-be] Industrial & port areas

[rodeo .be]

Hello Vincent,

adding "internal features" sounds a great idea to me. Will increase the
utility of the map.

One point of attention, I would avoid adding security measures (fences,
access buildings, gates, security cameras etc).
Depending on the site (Doel, BASF), this information is protected by law.

Kind regards
Maarten

Op do 15 okt. 2020 om 20:09 schreef Vincent Van Eyken <
vincent.vaney...@gmail.com>:

> Following some open notes [1] in the Antwerp port area, I have some
> questions about mapping in industrial environments:
>
>1. In general, how should we handle features within (semi)public to
>private industrial zones: e.g. port areas, big company plants etc. ? Those
>features (ways, buildings, parkings, infrastructure) often have varying
>degrees of “public” access or visibility (aerial imagery and/or public
>signposts, media…), and range of usefulness for a specific or general
>public, e.g. orientation, routing, emergency situations…
>I suppose it is basically up to the local mapper’s judgment if and to
>what extent features get mapped. But maybe some of you have more experience
>in this matter or can point to clear guidelines or “best practice” examples
>for future reference?
>
>2. For example this specific note [2] about how to map terrain
>blocks/sectors (“blokvelden”) within a major industrial plant (BASF).
>Questions arise:
>
>
>- Should these be included in the OSM database in the first place?
>- If so: only after asking and informing the company about permission,
>licensing, etc.?
>- If so: with which tags; e.g. in this case “ref=” and
>“place=plot”?
>
>
>
>1. A related case is the Port of Antwerp’s system for port/quay
>numbering (haven-/kaainummers). Mind you this is an old topic [3], which
>was never really resolved, I guess? However, the added value of these
>numbers is quite obvious: they are useful for land and waterway navigation,
>often better-known (at least within the “port community”) and used more
>frequently/easily than the conventional (street+housenumber) addresses,
>often visible on highway destination signs, warehouses, gates, quaysides…
>and mentioned in virtually every company’s contact details.
>So despite the potential usefulness, the complete number set is not
>(yet) present in OSM and in any case not as consistently tagged nodes.
>Currently the data are publicly accessible to download, e.g. to be
>used as satnav POIs [4]. I assume this might facilitate a straightforward
>import, if they are released as open data.
>
> Similar questions:
>
>- Should they be in the OSM database?
>- If so: how: through a documented import?
>- If so: how to tag?
>
> Regards
> Vincent / QuercE
>
> [1] e.g.: https://www.openstreetmap.org/note/582619
>
> [2] https://www.openstreetmap.org/note/2380581
>
> [3]
> https://lists.openstreetmap.org/pipermail/talk-be/2010-November/001600.html
> and: https://www.openstreetmap.org/user/dancelis/diary/12282
>
> [4]
> https://www.portofantwerp.com/nl/my-poa/services/gps-module-antwerpport?fbclid=IwAR0OqZUkYOiSdPu8aMsZHkzAgIKqJoxEJcZf4HV1lb3bFBP84J0IP6UpAiQ
> and:
> https://portaal-stadantwerpen.opendata.arcgis.com/datasets/574c21e24b9b4fd8b36a52df5d8e4293_590
>
>
>
>
>
>
> ___
> Talk-be mailing list
> Talk-be@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-be
>
___
Talk-be mailing list
Talk-be@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-be

Re: [OSM-talk-be] waymarked or not?

[Tim Couwelier]

'The future is virtual'. Yes and no.

Personally, when I go walking - I prefer to.. walk. Not keep an eye on my
phone for the route all the time.
I do love having them available virtually - the majority of my signposted
walks I've been able to find on RouteYou.

But when do I use the app?
1) When we're uncertain we're on the right track
2) If the kids start nagging 'how far do we have to go still'.

I can see cycling routes (with a phone mounted on a steering wheel) go this
route much faster. But for walking - I'd like to think signposted routes
very much have their merit.
Also, could be epic drama if cell phone battery dies along the way.



Op ma 19 okt. 2020 om 23:06 schreef Wouter Hamelinck <
wouter.hameli...@gmail.com>:

> Totally agree with that.
>
> I also think that the future will be virtual, but don't think that it will
> come from a website with a predefined network. I think that in the future
> you will just insert some parameters (or the parameters have been deduced
> from other routes that you liked) and a personal route will be generated
> automatically. Hopefully based on OSM data.
> For me it is a lot more important to have the underlying path right, than
> to copy routes from all kinds of websites. Especially in forested regions,
> it is incredible how many paths are still missing. Or how many that don't
> exist have been mapped by armchair mappers. We should really focus on the
> basis in my opinion.
>
> wouter
>
> On Mon, Oct 19, 2020 at 10:38 PM Sander Deryckere 
> wrote:
>
>> If the virtual routes are available under a strict copyright, there's
>> nothing we can map. And if they are available under a free copyright, we
>> add very little value by adding them to OSM.
>>
>> So I believe they don't belong in the main OSM db, but rather in a side
>> project (a project made for routes, prrhaps something umap like?).
>>
>> Op ma 19 okt. 2020 21:38 schreef Stijn Rombauts via Talk-be <
>> talk-be@openstreetmap.org>:
>>
>>>
>>> Hi,
>>>
>>> That's also what I would expect: virtual is the future. Installing all
>>> those signposts and keeping them in order takes a lot of time and money. If
>>> the tourism agencies see that they can virtualize them away without losing
>>> tourists, they will. We will indeed lose relevance if we don't go along.
>>> By the way, if we stick to ground truth, we'll also have to remove most
>>> of the cycle highways because a lot of them haven't been waymarked yet and
>>> are still virtual. We just copied the information available on
>>> https://fietssnelwegen.be/ (and went even a lot further with those so
>>> called 'alternatives' which are still just somebody's fantasy in my
>>> opinion). So, in fact we already did decide that there is a place for
>>> virtual routes in OSM...
>>> But indeed: we will have to make a thorough choice in the official
>>> operators AND their choices.
>>>
>>> Some further comments on other reactions:
>>>
>>> No, it's not harder to keep the virtual routes up to date. It's even
>>> easier. You don't have to go out to check if there are still signposts or
>>> you don't have to buy a map or check if it's still for sale. If the route
>>> is available on the 'source-website', it exists, otherwise not. We only
>>> need to know which is the 'source-website', so we don't rely on a
>>> (outdated) copy. For routes like the Randonnées en Boucle which are only
>>> available in a book, it's as dubious as a map: is the book still in print
>>> or not?
>>>
>>> Adding virtual routes won't make it more 'messy' than it already is. Who
>>> checks regularly (every few years) whether the hiking/cycle/... routes in
>>> OSM haven't changed in the meantime or still exist? E.g. how long did it
>>> take before the outdated LF-routes got removed?
>>>
>>> To Pierre and company: adding waymarked routes to OSM by using only
>>> gpx-tracks (if that is what you're doing) is even worse than adding virtual
>>> routes, because you have no guarantee that those gpx-tracks correspond to
>>> the ground truth. I know from experience. Also maps which correspond to the
>>> ground truth are rare. (But go ahead, I don't mind what you're doing.)
>>> And indeed, we can't even keep up with the waymarked routes, but we
>>> could as well use that as an argument to give up mapping routes completely.
>>>
>>> "A route, right now, is something you can expect to see waymarked." I
>>> feel we'll have to let go of this. "If someone starts mapping virtual
>>> routes, they should definitely be put in their own data model." They're
>>> still local/regional/... hiking/cycle/... routes. Adding some tag like
>>> 'virtual=yes" on the route relations and nodes should suffice. (It will be
>>> a bit more complicated because a node can be both a virtual hiking node and
>>> a real cycle node.)
>>>
>>> Regards,
>>>
>>> StijnRR
>>>
>>> On Monday, October 19, 2020, 07:34:48 PM GMT+2, Steven Clays <
>>> steven.cl...@gmail.com> wrote:
>>>
>>>
>>> Tendency in Toerisme Vlaanderen > ALL hiking nodes will go virtual
>

Re: [OSM-talk-be] waymarked or not?

[Wouter Hamelinck]

Totally agree with that.

I also think that the future will be virtual, but don't think that it will
come from a website with a predefined network. I think that in the future
you will just insert some parameters (or the parameters have been deduced
from other routes that you liked) and a personal route will be generated
automatically. Hopefully based on OSM data.
For me it is a lot more important to have the underlying path right, than
to copy routes from all kinds of websites. Especially in forested regions,
it is incredible how many paths are still missing. Or how many that don't
exist have been mapped by armchair mappers. We should really focus on the
basis in my opinion.

wouter

On Mon, Oct 19, 2020 at 10:38 PM Sander Deryckere 
wrote:

> If the virtual routes are available under a strict copyright, there's
> nothing we can map. And if they are available under a free copyright, we
> add very little value by adding them to OSM.
>
> So I believe they don't belong in the main OSM db, but rather in a side
> project (a project made for routes, prrhaps something umap like?).
>
> Op ma 19 okt. 2020 21:38 schreef Stijn Rombauts via Talk-be <
> talk-be@openstreetmap.org>:
>
>>
>> Hi,
>>
>> That's also what I would expect: virtual is the future. Installing all
>> those signposts and keeping them in order takes a lot of time and money. If
>> the tourism agencies see that they can virtualize them away without losing
>> tourists, they will. We will indeed lose relevance if we don't go along.
>> By the way, if we stick to ground truth, we'll also have to remove most
>> of the cycle highways because a lot of them haven't been waymarked yet and
>> are still virtual. We just copied the information available on
>> https://fietssnelwegen.be/ (and went even a lot further with those so
>> called 'alternatives' which are still just somebody's fantasy in my
>> opinion). So, in fact we already did decide that there is a place for
>> virtual routes in OSM...
>> But indeed: we will have to make a thorough choice in the official
>> operators AND their choices.
>>
>> Some further comments on other reactions:
>>
>> No, it's not harder to keep the virtual routes up to date. It's even
>> easier. You don't have to go out to check if there are still signposts or
>> you don't have to buy a map or check if it's still for sale. If the route
>> is available on the 'source-website', it exists, otherwise not. We only
>> need to know which is the 'source-website', so we don't rely on a
>> (outdated) copy. For routes like the Randonnées en Boucle which are only
>> available in a book, it's as dubious as a map: is the book still in print
>> or not?
>>
>> Adding virtual routes won't make it more 'messy' than it already is. Who
>> checks regularly (every few years) whether the hiking/cycle/... routes in
>> OSM haven't changed in the meantime or still exist? E.g. how long did it
>> take before the outdated LF-routes got removed?
>>
>> To Pierre and company: adding waymarked routes to OSM by using only
>> gpx-tracks (if that is what you're doing) is even worse than adding virtual
>> routes, because you have no guarantee that those gpx-tracks correspond to
>> the ground truth. I know from experience. Also maps which correspond to the
>> ground truth are rare. (But go ahead, I don't mind what you're doing.)
>> And indeed, we can't even keep up with the waymarked routes, but we could
>> as well use that as an argument to give up mapping routes completely.
>>
>> "A route, right now, is something you can expect to see waymarked." I
>> feel we'll have to let go of this. "If someone starts mapping virtual
>> routes, they should definitely be put in their own data model." They're
>> still local/regional/... hiking/cycle/... routes. Adding some tag like
>> 'virtual=yes" on the route relations and nodes should suffice. (It will be
>> a bit more complicated because a node can be both a virtual hiking node and
>> a real cycle node.)
>>
>> Regards,
>>
>> StijnRR
>>
>> On Monday, October 19, 2020, 07:34:48 PM GMT+2, Steven Clays <
>> steven.cl...@gmail.com> wrote:
>>
>>
>> Tendency in Toerisme Vlaanderen > ALL hiking nodes will go virtual within
>> 10 years or so. (At least, that is their vision) So if you do not follow
>> this tendency, you make OSM irrelevant for routes. I'd make a thorough
>> choice in the official operators AND their choices. Eg. Natuurpunt DOES
>> stick to signposting AFAIK.
>>
>> Op ma 19 okt. 2020 om 14:47 schreef Matthieu Gaillet > >:
>>
>>
>> Wether they are following another route is not relevant since it’s a
>> separate relation.
>>
>> Matthieu Gaillet
>>
>> On 19 Oct 2020, at 14:33, Wouter Hamelinck 
>> wrote:
>>
>> Are there any EV routes in Belgium that are not also LF or RV?
>>
>> Wouter
>>
>> On Mon, 19 Oct 2020, 12:29 Matthieu Gaillet,  wrote:
>>
>> Things are actually much less obvious and deserve a real second thought
>> before taking position : it just came up to my mind that much of the
>> Eurovelo network is still current

Re: [OSM-talk-be] waymarked or not?

[Sander Deryckere]

If the virtual routes are available under a strict copyright, there's
nothing we can map. And if they are available under a free copyright, we
add very little value by adding them to OSM.

So I believe they don't belong in the main OSM db, but rather in a side
project (a project made for routes, prrhaps something umap like?).

Op ma 19 okt. 2020 21:38 schreef Stijn Rombauts via Talk-be <
talk-be@openstreetmap.org>:

>
> Hi,
>
> That's also what I would expect: virtual is the future. Installing all
> those signposts and keeping them in order takes a lot of time and money. If
> the tourism agencies see that they can virtualize them away without losing
> tourists, they will. We will indeed lose relevance if we don't go along.
> By the way, if we stick to ground truth, we'll also have to remove most of
> the cycle highways because a lot of them haven't been waymarked yet and are
> still virtual. We just copied the information available on
> https://fietssnelwegen.be/ (and went even a lot further with those so
> called 'alternatives' which are still just somebody's fantasy in my
> opinion). So, in fact we already did decide that there is a place for
> virtual routes in OSM...
> But indeed: we will have to make a thorough choice in the official
> operators AND their choices.
>
> Some further comments on other reactions:
>
> No, it's not harder to keep the virtual routes up to date. It's even
> easier. You don't have to go out to check if there are still signposts or
> you don't have to buy a map or check if it's still for sale. If the route
> is available on the 'source-website', it exists, otherwise not. We only
> need to know which is the 'source-website', so we don't rely on a
> (outdated) copy. For routes like the Randonnées en Boucle which are only
> available in a book, it's as dubious as a map: is the book still in print
> or not?
>
> Adding virtual routes won't make it more 'messy' than it already is. Who
> checks regularly (every few years) whether the hiking/cycle/... routes in
> OSM haven't changed in the meantime or still exist? E.g. how long did it
> take before the outdated LF-routes got removed?
>
> To Pierre and company: adding waymarked routes to OSM by using only
> gpx-tracks (if that is what you're doing) is even worse than adding virtual
> routes, because you have no guarantee that those gpx-tracks correspond to
> the ground truth. I know from experience. Also maps which correspond to the
> ground truth are rare. (But go ahead, I don't mind what you're doing.)
> And indeed, we can't even keep up with the waymarked routes, but we could
> as well use that as an argument to give up mapping routes completely.
>
> "A route, right now, is something you can expect to see waymarked." I feel
> we'll have to let go of this. "If someone starts mapping virtual routes,
> they should definitely be put in their own data model." They're still
> local/regional/... hiking/cycle/... routes. Adding some tag like
> 'virtual=yes" on the route relations and nodes should suffice. (It will be
> a bit more complicated because a node can be both a virtual hiking node and
> a real cycle node.)
>
> Regards,
>
> StijnRR
>
> On Monday, October 19, 2020, 07:34:48 PM GMT+2, Steven Clays <
> steven.cl...@gmail.com> wrote:
>
>
> Tendency in Toerisme Vlaanderen > ALL hiking nodes will go virtual within
> 10 years or so. (At least, that is their vision) So if you do not follow
> this tendency, you make OSM irrelevant for routes. I'd make a thorough
> choice in the official operators AND their choices. Eg. Natuurpunt DOES
> stick to signposting AFAIK.
>
> Op ma 19 okt. 2020 om 14:47 schreef Matthieu Gaillet  >:
>
>
> Wether they are following another route is not relevant since it’s a
> separate relation.
>
> Matthieu Gaillet
>
> On 19 Oct 2020, at 14:33, Wouter Hamelinck 
> wrote:
>
> Are there any EV routes in Belgium that are not also LF or RV?
>
> Wouter
>
> On Mon, 19 Oct 2020, 12:29 Matthieu Gaillet,  wrote:
>
> Things are actually much less obvious and deserve a real second thought
> before taking position : it just came up to my mind that much of the
> Eurovelo network is still currently completely virtual (work in progress),
> yet deleting in from our map would be totally irrelevant since this routes
> are actually existing by the simple fact that thousands of users are using
> it.
>
> Matthieu Gaillet
>
> On 13 Oct 2020, at 19:21, joost schouppe  wrote:
>
> I think we shouldn't actively map purely virtual routes. But there's a lot
> of info that only lives on paper and still is relevant to OSM. So I find it
> hard to give it a hard no. What is essential though, is that we don't make
> a mess of the tagging. A route, right now, is something you can expect to
> see waymarked. If someone starts mapping virtual routes, they should
> definitely be put in their own data model.
>
> Op di 13 okt. 2020 om 13:27 schreef Matthieu Gaillet  >:
>
>
> That might be true but apply as well to signposted trails on the fled… I’m
> not ful

Re: [OSM-talk-be] waymarked or not?

[Stijn Rombauts via Talk-be]

 
Hi,

That's also what I would expect: virtual is the future. Installing all those 
signposts and keeping them in order takes a lot of time and money. If the 
tourism agencies see that they can virtualize them away without losing 
tourists, they will. We will indeed lose relevance if we don't go along.
By the way, if we stick to ground truth, we'll also have to remove most of the 
cycle highways because a lot of them haven't been waymarked yet and are still 
virtual. We just copied the information available on https://fietssnelwegen.be/ 
(and went even a lot further with those so called 'alternatives' which are 
still just somebody's fantasy in my opinion). So, in fact we already did decide 
that there is a place for virtual routes in OSM...
But indeed: we will have to make a thorough choice in the official operators 
AND their choices.

Some further comments on other reactions:

No, it's not harder to keep the virtual routes up to date. It's even easier. 
You don't have to go out to check if there are still signposts or you don't 
have to buy a map or check if it's still for sale. If the route is available on 
the 'source-website', it exists, otherwise not. We only need to know which is 
the 'source-website', so we don't rely on a (outdated) copy. For routes like 
the Randonnées en Boucle which are only available in a book, it's as dubious as 
a map: is the book still in print or not?

Adding virtual routes won't make it more 'messy' than it already is. Who checks 
regularly (every few years) whether the hiking/cycle/... routes in OSM haven't 
changed in the meantime or still exist? E.g. how long did it take before the 
outdated LF-routes got removed?

To Pierre and company: adding waymarked routes to OSM by using only gpx-tracks 
(if that is what you're doing) is even worse than adding virtual routes, 
because you have no guarantee that those gpx-tracks correspond to the ground 
truth. I know from experience. Also maps which correspond to the ground truth 
are rare. (But go ahead, I don't mind what you're doing.)
And indeed, we can't even keep up with the waymarked routes, but we could as 
well use that as an argument to give up mapping routes completely.

"A route, right now, is something you can expect to see waymarked." I feel 
we'll have to let go of this. "If someone starts mapping virtual routes, they 
should definitely be put in their own data model." They're still 
local/regional/... hiking/cycle/... routes. Adding some tag like 'virtual=yes" 
on the route relations and nodes should suffice. (It will be a bit more 
complicated because a node can be both a virtual hiking node and a real cycle 
node.)

Regards,

StijnRR

 On Monday, October 19, 2020, 07:34:48 PM GMT+2, Steven Clays 
 wrote:  
 
 Tendency in Toerisme Vlaanderen > ALL hiking nodes will go virtual within 10 
years or so. (At least, that is their vision) So if you do not follow this 
tendency, you make OSM irrelevant for routes. I'd make a thorough choice in the 
official operators AND their choices. Eg. Natuurpunt DOES stick to signposting 
AFAIK. 

Op ma 19 okt. 2020 om 14:47 schreef Matthieu Gaillet :


Wether they are following another route is not relevant since it’s a separate 
relation.
Matthieu Gaillet

On 19 Oct 2020, at 14:33, Wouter Hamelinck  wrote:
Are there any EV routes in Belgium that are not also LF or RV?
Wouter
On Mon, 19 Oct 2020, 12:29 Matthieu Gaillet,  wrote:

Things are actually much less obvious and deserve a real second thought before 
taking position : it just came up to my mind that much of the Eurovelo network 
is still currently completely virtual (work in progress), yet deleting in from 
our map would be totally irrelevant since this routes are actually existing by 
the simple fact that thousands of users are using it.
Matthieu Gaillet

On 13 Oct 2020, at 19:21, joost schouppe  wrote:
I think we shouldn't actively map purely virtual routes. But there's a lot of 
info that only lives on paper and still is relevant to OSM. So I find it hard 
to give it a hard no. What is essential though, is that we don't make a mess of 
the tagging. A route, right now, is something you can expect to see waymarked. 
If someone starts mapping virtual routes, they should definitely be put in 
their own data model.

Op di 13 okt. 2020 om 13:27 schreef Matthieu Gaillet :


That might be true but apply as well to signposted trails on the fled… I’m not 
fully convinced. 
But it is true that other websites or apps are specialised into publishing 
“virtual" trails and that might be something pertaining to the OSM project.
Matthieu Gaillet

On 13 Oct 2020, at 13:20, Wouter Hamelinck  wrote:
Hi all,
I follow those who propose to limit ourselves for the mapping purposes to what 
is waymarked on the ground.Taking routes from other sources (be they official 
or not) makes everything so fluid that we will end up with a huge mixed bag of 
gpx files that were at some point in time on some website of an authority, 
routes that are acti

Re: [OSM-talk-be] waymarked or not?

[Steven Clays]

Tendency in Toerisme Vlaanderen > ALL hiking nodes will go virtual within
10 years or so. (At least, that is their vision) So if you do not follow
this tendency, you make OSM irrelevant for routes. I'd make a thorough
choice in the official operators AND their choices. Eg. Natuurpunt DOES
stick to signposting AFAIK.

Op ma 19 okt. 2020 om 14:47 schreef Matthieu Gaillet :

>
> Wether they are following another route is not relevant since it’s a
> separate relation.
>
> Matthieu Gaillet
>
> On 19 Oct 2020, at 14:33, Wouter Hamelinck 
> wrote:
>
> Are there any EV routes in Belgium that are not also LF or RV?
>
> Wouter
>
> On Mon, 19 Oct 2020, 12:29 Matthieu Gaillet,  wrote:
>
>> Things are actually much less obvious and deserve a real second thought
>> before taking position : it just came up to my mind that much of the
>> Eurovelo network is still currently completely virtual (work in progress),
>> yet deleting in from our map would be totally irrelevant since this routes
>> are actually existing by the simple fact that thousands of users are using
>> it.
>>
>> Matthieu Gaillet
>>
>> On 13 Oct 2020, at 19:21, joost schouppe 
>> wrote:
>>
>> I think we shouldn't actively map purely virtual routes. But there's a
>> lot of info that only lives on paper and still is relevant to OSM. So I
>> find it hard to give it a hard no. What is essential though, is that we
>> don't make a mess of the tagging. A route, right now, is something you can
>> expect to see waymarked. If someone starts mapping virtual routes, they
>> should definitely be put in their own data model.
>>
>> Op di 13 okt. 2020 om 13:27 schreef Matthieu Gaillet > >:
>>
>>>
>>> That might be true but apply as well to signposted trails on the fled…
>>> I’m not fully convinced.
>>>
>>> But it is true that other websites or apps are specialised into
>>> publishing “virtual" trails and that might be something pertaining to the
>>> OSM project.
>>>
>>> Matthieu Gaillet
>>>
>>> On 13 Oct 2020, at 13:20, Wouter Hamelinck 
>>> wrote:
>>>
>>> Hi all,
>>>
>>> I follow those who propose to limit ourselves for the mapping purposes
>>> to what is waymarked on the ground.
>>> Taking routes from other sources (be they official or not) makes
>>> everything so fluid that we will end up with a huge mixed bag of gpx files
>>> that were at some point in time on some website of an authority, routes
>>> that are actively promoted, routes that were actively promoted for some
>>> event a few years ago and still can be found somewhere but are no longer
>>> maintained, routes where nobody really knows where they come from but they
>>> sound kind of official...
>>> It will get messy...
>>>
>>> Wouter
>>>
>>> On Tue, 13 Oct 2020, 09:51 Francois Gerin, 
>>> wrote:
>>>
 +1 for the "end user's perspective".

 From my point of view, two key rules make the ground for OSM as pointed
 out in several places of the documentation:

 1. Think to end users

 2. Map what really exists

 "Map what really exists" is visible in many places in the docs, and
 this is indeed important, up to some "threshold".
 "Think to the end users" is much less visible, but is visible anyway.

 I'm afraid that, being driven mostly by technical profiles/mappers, the
 "Map what exists" rule seems to take the precedence because it is more
 visible.

 According to me, "Think to the end users" should be the first rule, in
 terms of priorities.
 Followed by "Map what really exists", at the very same priority as "Use
 your common sense" which is also very visible in the docs...

 => My 2 cents.



 On 13/10/20 09:37, Matthieu Gaillet wrote:

 At first I was going to agree with Tim and s8evq but hey, the world is
 changing and from an user perspective, having itineraries on the map is a
 plus, wether they are signposted or not. I personally never follow sign
 posts, I just follow ‘a' route on my OSM-sourced GPS.

 Regarding the question "what should be mapped or not", I believe the
 itineraries should appear in OSM only if their are proposed or designed by
 an official operator, not mr nobody. That’s enough to keep quality, not
 staying aside nice initiatives (even if virtual), and stay close to
 exhaustive when it comes to official itineraries.

 After all, a route, sign posted or not, is in a sense always virtual.

 Matthieu

 On 13 Oct 2020, at 08:49, Tim Couwelier 
 wrote:

 I'm inclined to go by 'mapping verifiable ground truth'. Which means no
 - don't add them unless signposted along the way.

 Op di 13 okt. 2020 om 08:45 schreef s8evq :

> I do not think they should be in OSM, and I wouldn't mind deleting
> them. :)
>
> First of all, they are harder to keep up to date and verify.
> Secondly, like you said, where do you draw the line. Who's routes do
> we add and who's not?
>
> Fo

Re: [OSM-talk-be] waymarked or not?

[Matthieu Gaillet]

Wether they are following another route is not relevant since it’s a separate 
relation.

Matthieu Gaillet

> On 19 Oct 2020, at 14:33, Wouter Hamelinck  wrote:
> 
> Are there any EV routes in Belgium that are not also LF or RV?
> 
> Wouter
> 
> On Mon, 19 Oct 2020, 12:29 Matthieu Gaillet,  > wrote:
> Things are actually much less obvious and deserve a real second thought 
> before taking position : it just came up to my mind that much of the Eurovelo 
> network is still currently completely virtual (work in progress), yet 
> deleting in from our map would be totally irrelevant since this routes are 
> actually existing by the simple fact that thousands of users are using it.
> 
> Matthieu Gaillet
> 
>> On 13 Oct 2020, at 19:21, joost schouppe > > wrote:
>> 
>> I think we shouldn't actively map purely virtual routes. But there's a lot 
>> of info that only lives on paper and still is relevant to OSM. So I find it 
>> hard to give it a hard no. What is essential though, is that we don't make a 
>> mess of the tagging. A route, right now, is something you can expect to see 
>> waymarked. If someone starts mapping virtual routes, they should definitely 
>> be put in their own data model.
>> 
>> Op di 13 okt. 2020 om 13:27 schreef Matthieu Gaillet > >:
>> 
>> That might be true but apply as well to signposted trails on the fled… I’m 
>> not fully convinced. 
>> 
>> But it is true that other websites or apps are specialised into publishing 
>> “virtual" trails and that might be something pertaining to the OSM project.
>> 
>> Matthieu Gaillet
>> 
>>> On 13 Oct 2020, at 13:20, Wouter Hamelinck >> > wrote:
>>> 
>>> Hi all,
>>> 
>>> I follow those who propose to limit ourselves for the mapping purposes to 
>>> what is waymarked on the ground.
>>> Taking routes from other sources (be they official or not) makes everything 
>>> so fluid that we will end up with a huge mixed bag of gpx files that were 
>>> at some point in time on some website of an authority, routes that are 
>>> actively promoted, routes that were actively promoted for some event a few 
>>> years ago and still can be found somewhere but are no longer maintained, 
>>> routes where nobody really knows where they come from but they sound kind 
>>> of official...
>>> It will get messy...
>>> 
>>> Wouter
>>> 
>>> On Tue, 13 Oct 2020, 09:51 Francois Gerin, >> > wrote:
>>> +1 for the "end user's perspective".
>>> 
>>> From my point of view, two key rules make the ground for OSM as pointed out 
>>> in several places of the documentation:
>>> 
>>> 1. Think to end users
>>> 
>>> 2. Map what really exists
>>> 
>>> "Map what really exists" is visible in many places in the docs, and this is 
>>> indeed important, up to some "threshold".
>>> "Think to the end users" is much less visible, but is visible anyway.
>>> 
>>> I'm afraid that, being driven mostly by technical profiles/mappers, the 
>>> "Map what exists" rule seems to take the precedence because it is more 
>>> visible.
>>> 
>>> According to me, "Think to the end users" should be the first rule, in 
>>> terms of priorities.
>>> Followed by "Map what really exists", at the very same priority as "Use 
>>> your common sense" which is also very visible in the docs...
>>> 
>>> => My 2 cents.
>>> 
>>> 
>>> 
>>> 
>>> 
>>> On 13/10/20 09:37, Matthieu Gaillet wrote:
 At first I was going to agree with Tim and s8evq but hey, the world is 
 changing and from an user perspective, having itineraries on the map is a 
 plus, wether they are signposted or not. I personally never follow sign 
 posts, I just follow ‘a' route on my OSM-sourced GPS.
 
 Regarding the question "what should be mapped or not", I believe the 
 itineraries should appear in OSM only if their are proposed or designed by 
 an official operator, not mr nobody. That’s enough to keep quality, not 
 staying aside nice initiatives (even if virtual), and stay close to 
 exhaustive when it comes to official itineraries.
 
 After all, a route, sign posted or not, is in a sense always virtual.
 
 Matthieu
 
> On 13 Oct 2020, at 08:49, Tim Couwelier  > wrote:
> 
> I'm inclined to go by 'mapping verifiable ground truth'. Which means no - 
> don't add them unless signposted along the way.
> 
> Op di 13 okt. 2020 om 08:45 schreef s8evq  >:
> I do not think they should be in OSM, and I wouldn't mind deleting them. 
> :)
> 
> First of all, they are harder to keep up to date and verify.
> Secondly, like you said, where do you draw the line. Who's routes do we 
> add and who's not? 
> 
> For example, Natuurpunt and some of the local tourism offices already 
> have 'virtual' hikes, where they only suggest which node numbers to 
>>

Re: [OSM-talk-be] waymarked or not?

[Wouter Hamelinck]

Are there any EV routes in Belgium that are not also LF or RV?

Wouter

On Mon, 19 Oct 2020, 12:29 Matthieu Gaillet,  wrote:

> Things are actually much less obvious and deserve a real second thought
> before taking position : it just came up to my mind that much of the
> Eurovelo network is still currently completely virtual (work in progress),
> yet deleting in from our map would be totally irrelevant since this routes
> are actually existing by the simple fact that thousands of users are using
> it.
>
> Matthieu Gaillet
>
> On 13 Oct 2020, at 19:21, joost schouppe  wrote:
>
> I think we shouldn't actively map purely virtual routes. But there's a lot
> of info that only lives on paper and still is relevant to OSM. So I find it
> hard to give it a hard no. What is essential though, is that we don't make
> a mess of the tagging. A route, right now, is something you can expect to
> see waymarked. If someone starts mapping virtual routes, they should
> definitely be put in their own data model.
>
> Op di 13 okt. 2020 om 13:27 schreef Matthieu Gaillet  >:
>
>>
>> That might be true but apply as well to signposted trails on the fled…
>> I’m not fully convinced.
>>
>> But it is true that other websites or apps are specialised into
>> publishing “virtual" trails and that might be something pertaining to the
>> OSM project.
>>
>> Matthieu Gaillet
>>
>> On 13 Oct 2020, at 13:20, Wouter Hamelinck 
>> wrote:
>>
>> Hi all,
>>
>> I follow those who propose to limit ourselves for the mapping purposes to
>> what is waymarked on the ground.
>> Taking routes from other sources (be they official or not) makes
>> everything so fluid that we will end up with a huge mixed bag of gpx files
>> that were at some point in time on some website of an authority, routes
>> that are actively promoted, routes that were actively promoted for some
>> event a few years ago and still can be found somewhere but are no longer
>> maintained, routes where nobody really knows where they come from but they
>> sound kind of official...
>> It will get messy...
>>
>> Wouter
>>
>> On Tue, 13 Oct 2020, 09:51 Francois Gerin, 
>> wrote:
>>
>>> +1 for the "end user's perspective".
>>>
>>> From my point of view, two key rules make the ground for OSM as pointed
>>> out in several places of the documentation:
>>>
>>> 1. Think to end users
>>>
>>> 2. Map what really exists
>>>
>>> "Map what really exists" is visible in many places in the docs, and this
>>> is indeed important, up to some "threshold".
>>> "Think to the end users" is much less visible, but is visible anyway.
>>>
>>> I'm afraid that, being driven mostly by technical profiles/mappers, the
>>> "Map what exists" rule seems to take the precedence because it is more
>>> visible.
>>>
>>> According to me, "Think to the end users" should be the first rule, in
>>> terms of priorities.
>>> Followed by "Map what really exists", at the very same priority as "Use
>>> your common sense" which is also very visible in the docs...
>>>
>>> => My 2 cents.
>>>
>>>
>>>
>>> On 13/10/20 09:37, Matthieu Gaillet wrote:
>>>
>>> At first I was going to agree with Tim and s8evq but hey, the world is
>>> changing and from an user perspective, having itineraries on the map is a
>>> plus, wether they are signposted or not. I personally never follow sign
>>> posts, I just follow ‘a' route on my OSM-sourced GPS.
>>>
>>> Regarding the question "what should be mapped or not", I believe the
>>> itineraries should appear in OSM only if their are proposed or designed by
>>> an official operator, not mr nobody. That’s enough to keep quality, not
>>> staying aside nice initiatives (even if virtual), and stay close to
>>> exhaustive when it comes to official itineraries.
>>>
>>> After all, a route, sign posted or not, is in a sense always virtual.
>>>
>>> Matthieu
>>>
>>> On 13 Oct 2020, at 08:49, Tim Couwelier  wrote:
>>>
>>> I'm inclined to go by 'mapping verifiable ground truth'. Which means no
>>> - don't add them unless signposted along the way.
>>>
>>> Op di 13 okt. 2020 om 08:45 schreef s8evq :
>>>
 I do not think they should be in OSM, and I wouldn't mind deleting
 them. :)

 First of all, they are harder to keep up to date and verify.
 Secondly, like you said, where do you draw the line. Who's routes do we
 add and who's not?

 For example, Natuurpunt and some of the local tourism offices already
 have 'virtual' hikes, where they only suggest which node numbers to
 combine. On the ground, nothing is marked. I don't think this should be in
 OSM.

 If I get this correctly, 'Randonnées en Boucle' (SGR) are hikes made
 out of parts of existing GR trails? I wouldn't add that. The possibilities
 are just endless...

 On Mon, 12 Oct 2020 19:57:59 + (UTC), Stijn Rombauts via Talk-be <
 talk-be@openstreetmap.org> wrote:

 > Hi,
 >
 > There is a guideline or rule that only waymarked hiking/cycle/...
 routes should be added to OSM. Not every

Re: [OSM-talk-be] waymarked or not?

[Matthieu Gaillet]

Things are actually much less obvious and deserve a real second thought before 
taking position : it just came up to my mind that much of the Eurovelo network 
is still currently completely virtual (work in progress), yet deleting in from 
our map would be totally irrelevant since this routes are actually existing by 
the simple fact that thousands of users are using it.

Matthieu Gaillet

> On 13 Oct 2020, at 19:21, joost schouppe  wrote:
> 
> I think we shouldn't actively map purely virtual routes. But there's a lot of 
> info that only lives on paper and still is relevant to OSM. So I find it hard 
> to give it a hard no. What is essential though, is that we don't make a mess 
> of the tagging. A route, right now, is something you can expect to see 
> waymarked. If someone starts mapping virtual routes, they should definitely 
> be put in their own data model.
> 
> Op di 13 okt. 2020 om 13:27 schreef Matthieu Gaillet  >:
> 
> That might be true but apply as well to signposted trails on the fled… I’m 
> not fully convinced. 
> 
> But it is true that other websites or apps are specialised into publishing 
> “virtual" trails and that might be something pertaining to the OSM project.
> 
> Matthieu Gaillet
> 
>> On 13 Oct 2020, at 13:20, Wouter Hamelinck > > wrote:
>> 
>> Hi all,
>> 
>> I follow those who propose to limit ourselves for the mapping purposes to 
>> what is waymarked on the ground.
>> Taking routes from other sources (be they official or not) makes everything 
>> so fluid that we will end up with a huge mixed bag of gpx files that were at 
>> some point in time on some website of an authority, routes that are actively 
>> promoted, routes that were actively promoted for some event a few years ago 
>> and still can be found somewhere but are no longer maintained, routes where 
>> nobody really knows where they come from but they sound kind of official...
>> It will get messy...
>> 
>> Wouter
>> 
>> On Tue, 13 Oct 2020, 09:51 Francois Gerin, > > wrote:
>> +1 for the "end user's perspective".
>> 
>> From my point of view, two key rules make the ground for OSM as pointed out 
>> in several places of the documentation:
>> 
>> 1. Think to end users
>> 
>> 2. Map what really exists
>> 
>> "Map what really exists" is visible in many places in the docs, and this is 
>> indeed important, up to some "threshold".
>> "Think to the end users" is much less visible, but is visible anyway.
>> 
>> I'm afraid that, being driven mostly by technical profiles/mappers, the "Map 
>> what exists" rule seems to take the precedence because it is more visible.
>> 
>> According to me, "Think to the end users" should be the first rule, in terms 
>> of priorities.
>> Followed by "Map what really exists", at the very same priority as "Use your 
>> common sense" which is also very visible in the docs...
>> 
>> => My 2 cents.
>> 
>> 
>> 
>> 
>> 
>> On 13/10/20 09:37, Matthieu Gaillet wrote:
>>> At first I was going to agree with Tim and s8evq but hey, the world is 
>>> changing and from an user perspective, having itineraries on the map is a 
>>> plus, wether they are signposted or not. I personally never follow sign 
>>> posts, I just follow ‘a' route on my OSM-sourced GPS.
>>> 
>>> Regarding the question "what should be mapped or not", I believe the 
>>> itineraries should appear in OSM only if their are proposed or designed by 
>>> an official operator, not mr nobody. That’s enough to keep quality, not 
>>> staying aside nice initiatives (even if virtual), and stay close to 
>>> exhaustive when it comes to official itineraries.
>>> 
>>> After all, a route, sign posted or not, is in a sense always virtual.
>>> 
>>> Matthieu
>>> 
 On 13 Oct 2020, at 08:49, Tim Couwelier >>> > wrote:
 
 I'm inclined to go by 'mapping verifiable ground truth'. Which means no - 
 don't add them unless signposted along the way.
 
 Op di 13 okt. 2020 om 08:45 schreef s8evq >>> >:
 I do not think they should be in OSM, and I wouldn't mind deleting them. :)
 
 First of all, they are harder to keep up to date and verify.
 Secondly, like you said, where do you draw the line. Who's routes do we 
 add and who's not? 
 
 For example, Natuurpunt and some of the local tourism offices already have 
 'virtual' hikes, where they only suggest which node numbers to combine. On 
 the ground, nothing is marked. I don't think this should be in OSM.
 
 If I get this correctly, 'Randonnées en Boucle' (SGR) are hikes made out 
 of parts of existing GR trails? I wouldn't add that. The possibilities are 
 just endless...
 
 On Mon, 12 Oct 2020 19:57:59 + (UTC), Stijn Rombauts via Talk-be 
 mailto:talk-be@openstreetmap.org>> wrote:
 
 > Hi,
 > 
 > There is a guideline or rule that only waymarked hiking/cycle/... r