Uhm, they exist?!?

Please note that "railway=abandoned" means that a RR was once there and there are remaining traces (=it is visible that a RR used to be there). It does not mean that there are disused tracks. There are obvious traces, so this was correctly mapped. Before you delete things make you sure that you really understand the meaning of the tags you're deleting. (by reading the wiki).

Sure they don't exist? They were added very recently and there is recent construction. Best to ask the one who added the roads.

"If you disagree with the guidelines, then I suggest you get those guidelines changed" - If it's only my time .....
Guidelines are not rules, and I can use my time here better than dealing with bureaucratic nonsense, because a few people get upset about a few lines somewhere for very low reasons.

"Your argument [s] doesn't convince me." - Why not, exactly? Would be intresting. I also wrote why I am not convinced by your arguments.

You can't compare 50 year old RRs with 50 year old buildings, borders or even landuse. The point about RRs is to form a network - current and former ones. Forming an network is impossible and unnessescary with your examples (because they're usually independent from each other) and pretty much everything else (except roads, but i think abandoned roads are generally very rare). Most abandoned railways are still visible in some way, so they can be mapped with railway=abandoned. It is a good possibility to add the remaining <10% without traces to have a very good dataset on old RR routes, which dosen't exist elsewhere (except on OSM). And that is a kind of data in which a lot of people are intrested in - thats also why OSM has so much of those RRs mapped. They don't come from one or three mappers, they come from 100s of mappers.

"There is no "out of principle", it is all in accordance to the guidelines." - Then i change my wording to "out of guideline", which is basically the same, but sounds dumb. If i present you a lot of reasons why i think the mapping makes sense, and pretty much all you can tell me is "the guideline! the guideline!" then something is wrong with that guideline.

Stop telling me about "tagging for renderer" if you don't even understand it. None of the abandoned rails I add are "incorrect information" or even "fantasy mapping" if they're correctly tagged. That I want to have them shown on ORM does NOT mean tagging for the renderer. I guess 80% of mappers are motivated by the fact that their work is shown somewhere. I certainly would not have mapped 20k miles power line in GA and SC in such detail if they were not rendered on openinframap or elsewhere.

"This isn't just "a single line", these are many lines that are go through people's bedrooms, are snapped to roadways (incorrectly), and otherwise just do not exist anymore."
The expectation that a normal mapper will constantly encounter poorly mapped tracks is nonsense. However, if you regularly follow around old RRs in order to delete them, it is no wonder that it looks to you like they're all bad. I followed over thousand GA road miles in order to improve things around them, and you have no idea how much trash data there is that needs fixing. But i can't remember having a single problem with old RRs (good, i'm not so sensitive there), but encounters with them are even when i have a very 'dynamic editing area' quite rare.
"this is no different than an old city layout from the 1940s." - i said the exact opposite. Again, we are talking about a single line in most cases. Not hundreds.

Also, i had 4 more arguments where no one bothered to reply yet.

Yes, my data. I mapped most of the stuff which i restored here & invested many, many hours of work. E.g. relation relation/14022349 took at least 3+ hours and now you cutted it into tons of small pieces which makes it into ... data trash. Well, at least you didn't delete it entirely. But why didn't you tag the now still deleted sections as razed? This would be the perfect way to preserve the relation and the data about the old RR itself. Also my mistake to use too much rw=abandoned there and i will use less of that in favor of razed, but still, i don't think deletion is the way to do it here.

Sorry for the wall of text, but sometimes it is needed so you can understand my point of view.

railway=abandoned means "it is in some way visible that an RR used to be there". I explained that to you, so you know it. (Let alone that that's in the wiki (#Actual use: "The Wiki-Definition: Is a key to map former railways, where the rails have been removed but the route is still visible in some way."), and you are here since 2017 and expected to read the page in detail if you're doing such mass deletions).
Many deleted tracks like the Atlanta Beltline tracks in Bankhead, West Atlanta have VERY obvious embarkments. If you don't see that, you're blind, and shouldn't delete anything, and if you see them but delete 'em anyway, thats nothing else than vandalism.

Also, how do you have the mapping experience needed to conduct such mass deletions? If I check your editing stats at https://hdyc.neis-one.org/?CurlingMan13, you have edited 50 railway nodes and 300 ways, which is nothing (most of the ways must be from the recent deletion spree anyway). What this tells me is that you have no experience in rail mapping. And without such experience, you might not be able to see traces or you aren't aware of US-local mapping rules. And you shouldn't conduct any mass edits related to railways, especially not mass deletions.
osm.wiki/United_States/Railroads encourages to map old RRs.
BTW, i edited 15k railway nodes and 30k ways.

Why map them? My arguments, copied from another discussion:
- "The point of having them in OSM is to render them in openrailwaymap.org. So if you're deleting them, you're damaging this project"
- "Abandoned railroads have with at least 90% probability traces, which are mappable. So it makes sense to add the remaining 10% without to have a !very! good dataset regarding old routes. " [Which exists nowhere else than on OSM]
-"...You can't move things from OSM to OHM. It's literally illegal due to different licencing... properly merging that data type from two seperate databases is very hard"
-"You're expecting me to spend thousands and thousands of hours of work, if I want good data, even though the same result could be achieved with SIGNIFICANTLY less work here."
- and after all, why do you even care? In most cases, we're talking about a single line which goes thru a town. NOT an old city layout from 1940 on top of current data. How can you get so upset about a single line somewhere?
Look at osm.org/edit#map=18/33.80943/-84.45441
An old railyard, tons of railway=abandoned next to each other. If these are now being built over with a new neighborhood, then I can understand if you delete the old tracks. Because then it becomes too confusing. But in this CS and almost all other uses of rw=razed, it's just a single line or maybe, rarely two. Why can't you just let people map it and be not so persnickety? I don't understand all this drama.

Don't me come with DWG. The choice you have is to just let me do my thing and you continue closing notes which was way more helpful to the data quality than going from city to city and mass deleting valid data.
If you need DWG to prove me why those deletions are right, it shows me that you don't have good arguments to delete my data.

Your arguments & what could be interpreted as argument:
- "you are creating an invalid data model with features that do no represent the real world" -> Some features do not represent the real world, yes, but whats the problem?
- "Drawing railways which DO NOT EXIST is incorrect information" -> If it's tagged correctly, it is correct information. If i tagged them with railway=rail and name=Old Railroad, then it would be incorrect information.
-"If you want map features from the past, that's what OHM is for. Not OSM" -> Boils down to "out of principle"
-"That is outright wrong for OSM and is not allowed." -> Out of principle. OSM dosen't has rules like "You are not allowed to map that kind of stuff".
-"These tracks DO NOT EXIST and no combination of tags will make them appear in the real world or satellite imagery." -> The goal is to have them in a database, not in the real world.
-"The fact that some map can not function properly without these non-existing features means there's a problem with THAT map that needs to be fixed." -> That 'fix' would be to spend thousands of hours re-mapping the old RRs from scratch in another DB, as i described.
-"Too many of the features you mapped HAVE NO TRACES left in the world world and impossible to confirm." -> That they were there can be (in 99% of cases) confirmed with USGS topo maps or other sources. Which disproves "100% imaginary mapping".
-"features that you added have no traces whatsoever in the real world, therefore can not be mapped." -> "Out of principle".
-"The problem is that they do not exist now and therefore do not belong on OSM. " -> "Out of principle".
Hope I haven't forgot anything.

Of course you're supposed to accept my reasons. And you're supposing that I accept your reasons, that's how a discussion works. Your arguments just don't convince me at all.

My arguments:
- "The point of having them in OSM is to render them in openrailwaymap.org. So if you're deleting them, you're damaging this project"
- "Abandoned railroads have with at least 90% probability traces, which are mappable. So it makes sense to add the remaining 10% to have a !very! good dataset regarding old routes. "
-"...You can't move things from OSM to OHM. It's literally illegal ... properly merging that data type from two seperate databases is very hard"
-"You're expecting me to spend thousands and thousands of hours of work, if I want good data, even though the same result could be achieved with SIGNIFICANTLY less work here."
- AND AFTER ALL, why do you even care? We're talking about a single line which goes thru a town. NOT an old city layout from 1940 on top of current data. How can you get so upset about a single line somewhere?
Look at osm.org/edit#map=18/33.80943/-84.45441
An old railyard, tons of railway=abandoned next to each other. If these are now being built over with a new neighborhood, then I can understand if you delete the old tracks. Because then it becomes too confusing. But in this CS and almost all other uses of rw=razed, it's just a single line or maybe, rarely two. Why can't you just let people map it and be not so persnickety? I don't understand all this drama.

Most exist with obvious traces, reverted.
I tolerated your latest deletion spree a few days ago, but now that's straight & simple vandalism.

I told you why i want to map them here and and why it is the best way.

"My main question:
What is the problem... What exactly is so annoying about the ways that justifies deleting them?"

I want to have good data on the subject, and you don't. Ok. But if you expect me to basically redo the complete old rail mapping on OHM from scratch if i want good data, then I have the right to expect a good reason for why I should do that.
You're expecting me to spend thousands and thousands of hours of work, if I want good data, even though the same result could be achieved with SIGNIFICANTLY less work here. And your reason for this is "out of principle, this doesn't belong here".

No, im not accepting this. I want real reasons, good reasons why I should invest MY time for this. Your feelings being hurt by a few railway=razed ways is NOT a good reason.

How many times do I have to say that "razed" objects have no traces until you understand? I never said "razed" tracks have traces. Only "abandoned" have (if they're correctly tagged).
The example was already tagged razed, therefore i can't give you traces on the ground, but verifying that the track was there (and thus 100% correctly mapped) can be done via old maps: https://i.imgur.com/8ZkQbAe.png
This one is created by the USGS and therefore in public domain (usable for OSM, accessible at https://ngmdb.usgs.gov/topoview/viewer/)
"shitnitpicking" - if someone is doing this here, then you, because you are constantly avoiding my question. "What's the problem?"

All you're saying is "Uhh it violates my mapping principles because it is not visible on the ground, we should delete it". If you have no other reasons to delete, that is just as dumb as if i would say "Uhh we should delete everything in New Jersey because i don't like the state".

"You are insisting on mapping features THAT DO NOT EXIST IN THE REAL WORLD." - YES I DO! WHAT'S THE PROBLEM?

"In this changeset you created many railroad tracks that DO NOT EXIST at all!" - I guess you just don't want to understand. NONE of the re-added tracks here exists 'as track'. 90% of them are railway=abandoned, which implies existing traces (traces ≠ tracks) and are thus completely fine to be mapped, and the other 10% are railway=razed, which were removed entirely (on the ground). And you have still failed to provide ONE SINGLE good reason why those 10% should be deleted. "Out of principle" ain't a reason for me.

"In this changeset you created many railroad tracks that DO NOT EXIST" -
Just no. If I had tagged them with railway=rail, I would have done what you say here. But I did not.
razed/abandoned are NOT rail tracks currently on the ground.

"why don't we map every single building" - Because you can't compare this. Abandoned railroads have with at least 90% probability traces, which are mappable. So it makes sense to add the remaining 10% to have a !very! good dataset regarding old routes.
Unlike buildings, this data is as good as complete.

"If you want map features from the past, that's what OHM is for." - OHM in its current form is nonsense and unusable. I could write a long story with many arguments now, but to make it short: My two main problems: You can't move things from OSM to OHM. It's literally illegal, different licences. But we have the data here. In addition, you often need the data of the old routes in connection with the current ones. They are also here, and properly merging that data type from two seperate databases is very hard.
My main question:
What is the problem with the railway=razed ways here? What exactly is so annoying about the ways that justifies deleting them? (Remember, they're other people's work) What justifies damaging the completeness of projects like ORM?

If I and the other people who map them would use railway=rail for non-existing railroads, then it would be incorrect information. But we use railway=abandoned/razed and both tags mean that the RR does no longer exist. Read the definitions at osm.wiki/Demolished_Railway#Actual_use
And you still haven't answered my main question "Why is this invalid, although it is used very usefully? What is the problem with this?"

"Tagging for the render" means adding INCORRECT data to draw a certain render result on the map. Those railway=abandoned/razed ways are NOT incorrect information.

"because you are creating an invalid data model with features that do no represent the real world"
Why is this invalid, although it is used very usefully? What is the problem with this?

"10% error rate" - You're acting like we're talking about the distinction of active/removed rail. No, it is the distinction of removed rail - whether it has traces or not. When most of that rail got mapped, railway=razed didn't really exist, so people used =abandoned. Thats why the tagging is sometimes not perfect.

80% of your examples are already tagged railway=razed, which means that they are removed without traces.
The point of having them in OSM is to render them in openrailwaymap.org. So if you're deleting them, you're damaging this project, and what's your reason to to do so, in other words, how are you improving OSM data by deleting such RRs? Because as i see it, you are damaging OSM data by deleting those ways.
"removing disused" - Disused features are definitely on the ground, so deleting those is definitely vandalism.
"creating fake features in the name of restoring deleted disused ones." - I don't get that.
way/11603038 and 509892541 are the only ones which should be tagged with railway=razed instead, the northern part of way/283123430 has a clear "shrubline" along it's path.
3 out of let's say, 30 ways you've checked, are imperfectly tagged. Wow, congratulations.

"they do not have any railways or rails" - Yes - if they had, it would be railway=disused.
"are constantly being attached to buildings, roads and other relations" - Compared to landuses/borders, it is veery rare and not a reason to delete. Select the one shared node, press D and the problem is solved.

"There are no longer rails, and OSM guidance is to map what is physically on the ground." - If used correctly, railway=abandoned is physically on the ground because it implies that there are traces/remmnants of an old RR.

I don't care over wiki-principles. If you want to convince me that you are improving data quality by deleting the ways, then explain to me exactly where the improvement is. My point with the link of the abandoned yard in atlanta was to give an example of when deletion would improve the overall data quality; because it would get too cluttered otherwise.
"with no trace remaining, it does not meet the definition of "razed" - 'No trace remaining' is the definition of razed. 'abandoned' means there are some remmnants.

"I encourage these "historical" railways with no "on the ground" presence to be added to OHM (OpenHistoricalMaps) instead. " - Sounds good in theory, but it is nonsense currently. First of all, you cannot just move things from OSM to OHM because OHM uses a different licence. So everything in OHM has to be remapped completely from scratch, you can't reuse anything from OSM. Then lets suppose i do that and add an old route. I would still have more or less data garbage, because i can't properly view it anywhere. The OHM database is empty, so if i want to view it on the main site i would have an random line somewhere with zero relation to the current world i.e. i have no idea where the line actually ran.
An very important aspect of the abandoned RRs is the relation to the current lines. Unfortunality, also none of them exists in OHM, and i don't think anyone of those who invested millions of hours in the world's rail system on OSM (including me) wants to do the exact same work again (remember, copying is illegal).
Renderers like openrailwaymap don't display anything from OHM. Firstly because there is nothing to display, secondly, why?? OSM has very good data on old RRs (+450,000 km, data holes due to speedy deleters are fortunately rare)

TLDR; OHM is useless IMO.
Using OSM for abandoned railroads is the easiest and best way to maintain good data on the subject. If you disagree, then -as i said above- then give me specifc reasons and not an "itS AgaInEsT tHe wIkI!11!".

Because the orginal deletions blindly deleted all railway=abandoned in the area. It is not my responsibility to check every single track if i revert such vandalism. (As is said above, i checked a lot on a lot of tracks if railway=razed would fit better)
Would be nice if you could provide a few examples.

If i check the CS at https://osmcha.org/changesets/123215313/ and take a look at the green ways, I see traces on almost all ways. I tagged a few ones without traces as "razed".

railway=abandoned does not mean that there are still tracks.
railway=abandoned