Bonjour,

Je vous ai expliqué hier sur votre changeset osm.org/changeset/106977603 que si un chemin autorise un accès pour différents types de véhicules, dans ce cas, véhicules agricoles, vélos et chevaux, il faut lui attribuer la valeur la plus pertinente. Je viens de vérifier sur le terrain, j'ai pris des photos de la signalétique, les véhicules agricoles on bien un accès (logique à la vue des images aériennes). J'ai même demandé a une fermière habitant n. 89 ...
Donc comme je vous l'ai dit hier, un chemin agricole avec vélos autorisés c'est en higwhay=track + bicycle=yes ou =designated.
Une telle erreur de valeur contribue à désinformer les utilisateurs de la carte qui pourraient penser qu'ils ont priorité sur ce chemin. S'il vous plait, veuillez en prendre la considération car je ne vous ferai pas la remarque à chaque fois.

Cordialement,

Salut Nanard! Je crois qu'il vaut mieux ne pas effacer des chemins uniquement en se basant sur les activités Strava surtout s'ils sont encore visibles sur les photos aériennes récentes. Par expérience, j'ai remarqué que ce sont souvent des chemins délaissés, des grade5, rarement il y a même des grade4 que les gens n'osent pas emprunter.

J'ai regardé les chemins que tu as a effacé et j'ai remarqué que tous étaient crées par pat176. Tu as peut-être aussi remarqué qu'il manquait énormément de précision. A cette époque là il n'avait pas autant de sources pour OSM et les GPS étaient encore chers. Il mettait presque tous les tracks en grade3 et si ce n'était pas le cas il faisait des aberrations comme grade4 + surface=gravel. De toute façon, ça fait des lustres qu'il n'est plus actif.

Observations:
osm.org/way/80777755/history est encore visible sur les photos et il y a tout de même quelques rares traces GPS.
osm.org/way/80777695/history n'a rien OK, mais ses intersections sont mal positionnés. Vu sa forme au N, à mon avis c'est un ancien chemin creux. Perso, j'ai commencé à les tagger comme cutting=hollow_way. Il y a des gens qui utilisent historic=hollow_way mais ils n'ont pas forcément une véritable valeur historique dans le sens "patrimoine culturel", donc c'est faux. C'est pour ça que sa proposition de tag avait été abandonné.
osm.org/way/80777639/history a aussi quelques rares traces GPS et très bien visible au SE.

Si des chemins ne sont vraiment plus praticables, je crois que ce serait mieux de les déactiver avec le préfixe abandoned:highway=track ou si intentionnel avec razed: (travaux forestiers) pour informer les autres contributeurs qui utilisent le LIDAR, les autres tags (surface, tracktype, etc) qui risquent d'influencer des rendus de cartes vectorielles sont à effacer, les notes ça va.

T'en penses quoi?

S'il te plait réponds-ici au cas où d'autres voudraient donner leurs avis. Merci!

Hi,

The "landuse=farmland" in your example osm.org/way/974591622 and any other merged one is still shown. No portion on the map has been deleted geospatially. So I don't see where the issue is.

All in all, the following arguments explain why I find those adjacent multiple fields tagged equally not being relevant, not useful, even deceiving. What I do is certainly not vandalism and in fact good practice.

What's the point of separating/keeping a big farmland in smaller parcels if there is no tag difference? The elements i merged didn't have any. If they had different crops detailed, i would find that relevant and very useful but there were none. Let's even assume someone wants to add the crop=* tag, if you check Geoportail.lu aerial images from 2020, they are actually the same crop.

If someone gives me the argument for parcels, sorry I don't find that neither valid since OSM is not a land registry. As analogy, OSM contributors neither split residential plots, even if there is a retail store on the ground floor of a big building but the rest of them are apartments.

In terms of data, it's the same situation if several portions of highways are split with the same tags. It is useless to keep them separated for nothing and even considered a bad practice, so merging is good practice for future edits. It's multiple nodes on a straight way, I would delete them as suggested in osm.wiki/Good_practice#Keep_straight_ways_straight I consider this useless data duplication. Deleting/merging them results with the same map rendering, just less data to process (ways id, nodes id and connections) that has been to be taken into account on rendering for vectorised maps. Small portable devices can't handle countless polygons. It is not only less data for devices or OSM's servers, which have each year issues of infrastructure needs, but also less data to handle for contributors who will update. Subsequently...

Mostly why I do this it is because fields are not fixed entities. Farmers have agricultural techniques to rest and develop soil's nutrients and they also adapt fields to demand depending how the year sold. So accordingly they change the plot sizes and crops time to time (yearly or bi-annual). Many contributors leave so many fields without any difference for so many years. In the meantime, the field structure has changed. So if people care about this separation without checking reality, they will rather be misled.

I don't expect you to believe me blindly, normal. So here is an example with a field I know and left untouched despite it has been merged for the same crop in real: This field was plotted at v1 a year ago osm.org/way/873210020/history It existed in real until 2018 (go wonder why tomolobla did so in 2020, it's one of the reasons why I went in conflict with him). Then on Geoportail.lu summer 2019 photos (careful, not photos from winter in early 2019) and until now, it has been merged in a big monoculture with osm.org/way/873210015 osm.org/way/873210022 and osm.org/way/873210010 . Even the fence on N between the present meadow in OSM and the farmland doesn't exist anymore. So if I follow the "do not merge field" inexistent rule, I can't merge the monoculture... Absurd!

I'm not forbidding anyone to map, I'm totally leaving room for other mappers and I welcome them. What woodpeck wrote in that block is irrelevant as I explained already. I was not bad-mouthing, but denunciating with facts how that user was making the map deceptive. I gave the proofs for that. As a user of the map too, it is normal for me to enhance or correct whatever I cross, as suggested in osm.wiki/Good_practice#Do_correct_errors . As you can see, I am very communicative with others. Not like some contributors who report someone else for futile reasons. Sorry to say this but this is childish. They assume OSM is their property. There are far bigger issues which in real mislead people. For example the imports from the EU - SOeS, CORINE Land Cover, 2006 polygons in FR or BE! There were some of those in LU too suggesting there is a certain landuse somewhere but it was misleading. Some even overlap conflicting areas. For example
- osm.org/way/208529088 cemetery overlapping with inexistent forest or osm.org/way/248373788 a building over what seems not being at all residential area osm.org/way/67369293 expand even over meadow fields, scrubs, woods... This is the type of inconsistency that you, as part of the DWG, should care and fight for in order to avoid misleading users who check the deceptive map. I never reported those because I know how things are with other contributors, so I just correct as I've been doing on the FR side recently. But if you really start on that path, be prepared to receive countless reports from me. I will expect you to fulfil them equally since they will certainly be considered to be more serious and problematic.

Hi! Sorry for the late reply.
Hmm? How would there be a road sign on the S side explicitly allowing normal traffic, even giving priority to agricultural if there is no road for normal traffic at all. There are tracks W + E and the cycleway coming from Elzange. Unless that way should be an unclassified road too? It is weird, no?

Bonjour! Ces chemins ont également le panneau C3, càd interdiction au trafic routier normal dans les deux sens, sauf ayant droits, exception, etc... Donc unclassified ce n'est pas assez pertinent. Normalement il faut les mettre en "highway=track" (+ "tracktype=gradeX" si connu). C'est le plus pertinent étant donné que les vélos etc peuvent rouler sur un track selon la définition du Wiki d'OSM mais un véhicule agricole ne peut pas rouler sur une piste cyclable.

+ forest/grass updated

Bonjour! Sur cet élément en v3 osm.org/way/460847775/history vous avez mis "friche" dans le nom. D'abord ce n'est pas le nom à proprement dire mais je comprend la démarche, puis dans l'idéal lorsque un terrain industriel/commercial/etc n'est plus exploité, il faut le mettre en "landuse=brownfield".
Bonne journée!

+ highways split/detailed, sidewalk added

Great, thank you! Yeah it's possible with time and depending on the season that it's less used and/or fewer traces are visible, its normal in nature.

Hi! I noticed you've reactivated these ways:
osm.org/way/542671228
osm.org/way/779440696
Long story short: I've been in conflict with the person who previously deactivated those, because he was doing so on countless highways which i had surveyed again and turned to be still existing. Though i can't reactivate all of them since OSM's DWG didn't believe me, blocked me, etc... even though i gave lots of proofs (photos, videos, GPS data). You know how people are when they don't admit they are wrong. We are between good hands in OSM, HAHA! Unfortunately, because of them, lots of work to redo!

I haven't been on this region in a long while though and i noticed that you reactivated them as "isolated island", meaning they are not reachable from the highways network either from W or E (where there is a block, you've left them deactivated "razed:highway=path". So can you confirm me if you've surveyed these ways on terrain and if so how from which end did you reach them, please?

OK thank you!

Also please, be careful when you remove a highway. For this one osm.org/way/812607194/history there is a reason why people added it. Please read my changeset comment osm.org/changeset/122835115 and check the highway's tags.

Hi! You added a set of restrictive access tags for this way osm.org/way/1009088592/history except foot=yes. This is ambiguous for coding as it exists different ways to interpret combinations. As you can see, the original standard OSM Carto directly grays up the way no matter extra specific access tags you add. In the end, it just misleads users. So please do not add a general access=* + specific exceptions as this is a interpretative combination. By definition, the access=* is for all means of mobility, no exceptions. It's either all or the specific tags only. Some people from other regions started to make the combo exceptions because they don't have the patience to add properly other specific tags (specially in Germany from my observations). You did it though (with a mistake for motor_vehicles, otherwise how can forestry vehicles exploit the land), so i guess you just added the tag just by calking from others. Thanks for your understanding!

Hi! You've added a ford here osm.org/node/9754846810/history on v1 but i know that the track is embanked up, at least on the spot where it is now. So unless that spot has been damaged some time ago or possibly the real location is a bit further N, it is very unlikely having a ford. Did you check on terrain?

And the track you've changed as cycleway, if agricultural still use it to access the fields, it should also stay as track + bicycle designated

Hi! About this way osm.org/way/1071345207 you've changed from track to unclassified, was the sign forbidding normal traffic removed on the N side? If not it should stay as track grade1

Bonjour! Vous voudriez bien éviter d'utiliser toujours le même un descriptif aussi général, svp. Sur ce changeset vous n'avez modifié aucun landuse, uniquement des highways... Une contribution sur OSM implique toujours une amélioration donc donner "améliorations diverses" est inutile.

Bonjour! "Fuzzy boundaries", pas vraiment! Par exemple ici vous avez la liste des communes incluses pour le Pays-Haut https://fr.wikipedia.org/wiki/Pays_Haut Il faudrait respecter les limites si vous insérez un tel élément sinon je crois qu'on peut considérer cela comme de la désinformation.

+ deleted obsolete/simplified crossings

+ tracktype added ¬¬