Changeset: 56127378
(no comment)
Closed by Verdy_p
Tags
created_by | RawEdit |
---|
Discussion
-
Comment from Verdy_p
This was done in rawedit (which still offers no way to add edit comments); this added more evident dictinctions in names, between land areas and admin boundaries
-
Comment from SomeoneElse
Hello,
What exactly is "Jersey (land mass)"? That looks like a description rather than a name. If you want to have a relation for that, fine - but please don't make up names for things like that. Also, this does look very much like an undiscussed mechanical edit...
Also, this is a very large changeset with no changeset comment. Even if you can't find a tool to make the edits that you want (which I doubt) there's absolutely no reason to include objects over such a large geographical area in it. In this changeset "France" is large because of the included overseas territories but "Jersey" (for example) is small.
Also I think the wiki data entries are wrong - https://www.wikidata.org/wiki/Q142?uselang=en-GB is the entry for France the country not the land area only for example.
Best Regards,
Andy -
Comment from Verdy_p
Raw edit is a set of individual edits on isolated tags. There's no way in it to create separate changesets, they are closed automatically but after long time (only made by the server, not by that editor).
The names for landmasses are not rendered on map, but they do no represent the whole area so yes "Jersey (land mess)" is accurate here, as it does not include the maritime areas of "Jersey" as a whole. -
Comment from Verdy_p
An no this was absolutely not "mechanical edit", they were done individually at very small speed, with raweditor, over a long perdiod of several hours using other editors.
If you know how to force "rawedit" to close its open changeset...
Why did I use rawedit, it's because for changing single tags, it was not necessary to load the full objects (which is dramatically slow here, even with JOSM partial doanload of relations) -
Comment from SomeoneElse
"Jersey (land mass)" is not a name, it is a description.
Re "RawEdit", you might find https://wiki.openstreetmap.org/wiki/Level0 more functional.
-
Comment from Verdy_p
Except that rawedit is integrated with other tools that propose to use either it (simple and fast for minor tag changes), or JOSM (too slow for these cases)
The names are consistant witrh other names for Landmass and were used because of confusion of admin boundaries and the fact that soime people created duplicate admin boundaries but actually not the same thing.
Nominatim does not even help to disambiguate the two, and for now adding the precision in the name has been completely harmless everywhere else it is used. -
Comment from Verdy_p
Note: "Jersey" alone would alsop be incorrect ! Jersey is larger than that.
The name is already fully descriptive of the intended use.
It will never cause any problem in searches of boundaries.... -
Comment from SomeoneElse
Hu Verdy_p,
Please don't be so argumentative - I'm trying to help you here!With regard to names, their use in OSM is well-established - see https://wiki.openstreetmap.org/wiki/Names#Name_is_the_name_only . By those criteria "Jersey (land mass)" is not a name. The actual name for both the individual island (the dry bit) and the administrative area that includes a wet bit and a dry bit is "Jersey". If you want to create a relation including the (dry) island of Jersey and the smaller islands administered by it then by all means create it, but please don't make up a name for it.
Best Regards,
Andy -
Comment from Verdy_p
This method of naming differently has been used everywhere else since the beginning, because there were confusions betwee ndistinct objects (and Nominatim was also confused and did not allow making the distinction between landmasses and admin boundaries, but also because various contributors changed landmasses into admin boundaries, creating duplicate)
This is NOT specific to Jersey, it has been used since long in Germany (the first to use it), and it was not by me but was decided because of these confusion problems between distinct objects. -
Comment from Verdy_p
Note: the OSM website (here). Does not allow seeing distinctions easily (the types of objects are NOT specified at all in list of members or list of parent relations).
So confusion is extremely frequent when this OSM site does not make any difference, but just displays the "name" tag, the object "id", and the object version.
When people see that, they want to remove what they think are duplicates, or select the wrong objects as references.
iD also does not offer any easy distinction. -
Comment from Verdy_p
Remember that these "names" found landmasses are never rendered at all with labels on maps because they are NOT boundaries. They are just used by reference from other objects or selected and found by specific tags.
May be iD and this site should display at least the "note" tag, or the "type" tag of relations. -
Comment from SomeoneElse
Another well-established principle is that we do not tag for the renderer, or how it appears in a search engine such as nominatim, or how it appears in an editor such as iD, or on this website.
"Jersey (land mass)" is not a name, and shouldn't be assigned the name tag.
Also, yes, I am just using Jersey as an example here.
Either you can go through these changesets and remove the "not actually names" that you added, or I can with a Data Working Group hat on. Obviously it would be a lot easier for all concerned if you did it.
Best Regards,
Andy Townsend, (reluctantly) on behalf of OSM's Data Working Group -
Comment from Verdy_p
So you want to reverse a decision that was taken long time ago by other people (not me) to use distinctive name without even asking them why they chose that convention? And you don't care at all about the confusions that were made (which was why I made this changeset to solve duplicates or incorrect tagging for landmasses which were confused with boundaries !)
-
Comment from SomeoneElse
> So you want to reverse a decision that was taken long time ago by other people (not me) to use distinctive name without even asking them why they chose that convention?
Can you please point to such a "decision"? The best description of how we map names in OSM includes https://wiki.openstreetmap.org/wiki/Names#Name_is_the_name_only . That text was I believe written by Harry Wood 3 or 4 years ago, and has stood the test of time.
> And you don't care at all about the confusions that were made
Sorry, I don't understand what you're saying there (maybe write it in French if that's clearer)?
May I respectfully suggest that you take a step back and ask why it's always _you_ that seems to be having these discussions? Also, why is it that your wiki user talk page includes comments from extremely mild-mannered and tolerant mappers such as this: https://wiki.openstreetmap.org/wiki/User_talk:Verdy_p#MapLesotho_Pages ? The common factor is that your behaviour is different. When there's a difference of opinion in OSM, what normally happens is that people put their point of view forward and then try and find some kind of compromise. You aren't doing this (here or elsewhere). What you have said instead is:
o Explained what RawEdit does (I didn't ask, and already knew).
o Said that it was OK to use made-up names because "The names for landmasses are not rendered on map".
o Said "Nominatim does not even help to disambiguate the two" (I suspect that this means that you think that Nominatim can't distiguish between one of your landmass relations and an admin boundary - which I find difficult to believe).
o Said "The name is already fully descriptive of the intended use. " - which is fine, but if that value is not a name but a description then you should be adding it to a description tag not a name tag.
o Said "This method of naming differently has been used everywhere else since the beginning" but provided no evidence of that.
o Said "the OSM website (here). Does not allow seeing distinctions easily (the types of objects are NOT specified at all in list of members or list of parent relations). ". This is true to some extent (because it does not recognise what these relations represent), but irrelevant. We don't tag things in a certain way just so they appear in a certain way in a certain renderer (of which this website is one).
o Said "iD also does not offer any easy distinction.". This is untrue. It's very easy in iD to view tags and relation memberships.
In other words, the only reason that you're using these made-up names is because you're tagging for the renderer (in this case this website). https://wiki.openstreetmap.org/wiki/Tagging_for_the_renderer makes it clear - "avoid using incorrect tags". -
Comment from Verdy_p
I've not tagged for the renderer. FULL STOP. I have used existing conventions (sused since long). FULL STOP. I have solved incorrect tagging caused by confusion. Names are accessory but still essential for editors. the iD editor still does not make any distrinction when selecting relations, you have to visit each one after selecting to know if this is the correct one by looking at the list of tags.
Many people are confused. And as long as this does not cause any rendering problem for any one there's no reason to change something that was in place since the begining: "Jersey (landmasses)" is NOT "Jersey", only a part of it and this makes sense to have this precision, just like we add disambiguation prefixes in wikis that need unique names for linking and especially because this also causes confusion (OSM still has no mechanism to signal ambiguities and help resolve them to perform the correct selection; yes we have tags, but unfortunately they are ignored by the tools, as well as by people that don't check them at all or have no idea about how to choose correctly).I repeat it becauser you don't want to read: I have used those names to help fixing incorrect tags (not names, they don't matter here, but tagging distinct landmasses as if they were boundaries, and they are not !)
The augmented names are not descriptive they are accurate because these landmasses are NOT the same as the place from which they borrow a part of their name and only a part of their semantic meaning. -
Comment from Verdy_p
In summary: I did not use any incorrect tags, I fixed them to remove duplicates or incorrect references between unrelated areas.
-
Comment from SomeoneElse
> In summary: I did not use any incorrect tags
So where (in Jersey or elsewhere) can I go to find the name "Jersey (land mass)" then? Or is it just a "name" that you made up? -
Comment from SomeoneElse
But the more serious question is: why it's always _you_ that seems to be having these discussions?
What is your theory as to why this happens? -
Comment from Verdy_p
The incorrect tags (notably using type=boundary and boundary=administrative) or references caused also severe rendering problems for the international borders or some regional borders: they were discontinuous because queries for borders retrieved two unrelated relations, and renderers tend to eliminate ways that are selected twice: these borders were NOT rendered at all.
This caused also severe problem in Nominatim while it tried to locate a place in relevant superboundaries: superboundaries where eliminated. It caused severe problems for geocoding places (finding places by name or address and retrieving coordinates: places were not found).These bugs were NOT caused by names of landmasses but the fact that landmasses were incorrectly tagged as boundaries.
Confusion between relations will continue to happen again as long as editors are not changed to display relation types (at least): landmasses are NOT tagged with boundary=administrative, or type=boundary. It is what I fixed, individually, area by area.
And this was absolutely NOT an automated edit but individual fixes to solve existing problems that would have not exited at all if names were more accurate and not so confusing.
I repeat: "Jersey (land mass)" is NOT "Jersey". (I can say it for other cases, this is of course an example, I did not invent it, this naming scheme was used since long and not invented by me, I have respected the existing conventions that worked everywhere else and did not cause any problem except for you).
-
Comment from SomeoneElse
It's happening again (see https://wiki.openstreetmap.org/wiki/User_talk:Verdy_p#Route_relation:_Roles_forward:stop:.3Cnumber.3E_etc._on_page_Relation:route ) so I'll ask the question again:
Why it's always _you_ that seems to be having these discussions?
What is your theory as to why this happens? -
Comment from Verdy_p
Why is it only you that inists on what is not an issue at all ? Stop this harassment.
-
Comment from SomeoneElse
It's a genuine question, honestly - most people don't have lots of people complaining on their wiki talk pages.
Why do you think it happens to you? -
Comment from Verdy_p
"lot of people is your interpretation" actually this is very small to the number of thank you I receive directly.
You are just harassing me for sometinh which is absolutely not an issue anywhere and don't want to recognize that I solved really incorrect tagging that caused real problems.
You are doing transforming this into a agressive personal war for no valid reason at all, based only on your own theoretical assumption which here is also false. No I did not tag anything for rendering, and yes the names are accurate, and the convention used does not come from me, it has been used since long without any problem to avoid similar issues, and it will remain as long as editor tools (or navigation tools) to not disambiguate things correctly. There's much more benefit in disambiguating this explicitly, and espacially there when the designated objects are clearly different. -
Comment from SomeoneElse
> "lot of people is your interpretation" actually this is very small
No - most people don't have multiple archives of complaints on their wiki talk page. I'm trying to understand why this is happening to you and if there's anything that can be changed that would avoid it.
> You are doing transforming this into a agressive personal war
Again no - quite the opposite, actually. I'm trying to understand (and to help you understand) the causes for the conflicts that you have with other OSM contributors that other people, by and large, don't have.
Why do you think it happens to you? -
Comment from Verdy_p
Only because I'm above average users, just means there is a higher chance of friction occuring with just a few (like you do) for things that are in fact not even a problem or just because of personal opinions. Landmasses could even have no name it all (they are never needed for rendering), but since they were used to disabmiguate things more easily only in editors and tools, a choice was made by others (since long) to use this disambiguation is a way that is immediately usable in editors and QA tools. These names don"t even need to be translated, they or only there as helpers for contributors editing the map and that's fine.
-
Comment from Verdy_p
This was documented since long first in Germany https://wiki.openstreetmap.org/wiki/DE:Relation:land_area
A name attribute is allowed, but the doc never says it is mandatory, its value is completely unspecified and it can be omitted completely. If it is there it is just a helper for contributors but it has no use at all in rendering or when searching data for these relations (which are also not necessarily "administrative" in the way each country define rules for their coastal area, given that legally they should be defined on the baseline and not the highwater level of OSM "coastlines". OSM still has almost nowhere data about baselines and OSM coastlines are approximations.
So take it seriously, these relations are not legal boundaries, they do not represent the administrative unit you want to reuse to abbreviate them, they are only the emerged part of these units. So qualifid names are complelely fine and less misleasing.All this changeset was not about changing these names but disambiguate and fix other tags and relation members, because they were confused and this caused severe rendering and usage problems on conflicting boundary relations with the same attributes where the attroibutes should have been clearly different. After I made it, the map here looks finally OK and shows all admin borders (before there were "holes" everywhere caused by conflicts between confused relations. That job is done, all fixes are now OK, geocoding now works properly. And I don't understand why you criticize that only on a minor detail (non mandatory names for these relations, just used as very useful hints for contributors to avoid they make new confusion later or break existing relations)
-
Comment from SomeoneElse
On the issue of "land area relations" (the original issue here, but less important than the way that we work together as mappers) the https://wiki.openstreetmap.org/wiki/DE:Relation:land_area page just says "name" - i.e. "add the name of the thing". It does _not_ say "make up a name to make it appear special on the OSM website", which is what you are doing here. There appears to be no documentation that backs up your approach here.
-
Comment from SomeoneElse
... and on the more important issue:
> Only because I'm above average users
By _quantity_ of _wiki_ edits, certainly.
However, looking at your wiki talk page and at http://resultmaps.neis-one.org/osm-discussion-comments?uid=90780 there does seem to be a consistent problem. Have you ever replied to a question in a way that is non-argumentative? Have you ever said to anyone "actually, you're correct". This is something that most OSM mappers (and most people in the wider world) do all the time, yet it's something that you do not seem to do at all. Do you genuinely believe that you are always correct and everyone else is always wrong? Wouldn't it be statistically be very unlikely for that to be the case?
OSM is a colloborative project; we need to work together to create the best map. We simply can't do that if everyone were to take the attitude that "I am right, everyone else is wrong".
Again - to reiterate - I'm trying to help here. We can achieve so much more as mappers if we work together, and not spend time arguing with each other.
Best Regards,
Andy -
Comment from Verdy_p
You're inventing interpretations.
Since the begining it has always included accurate disambiguation. These landmasses are ***not*** the countries/regions, it has alwyas been a ***part*** of them.
You're just continuing to fight against what is actually not a problem at all anywhere.
The diusambiguation could as well be "Archipelago", or other similar terms. Many of these landmasses have their own defined geographical name distinct from the admin boundary (this is just to demonstrate that they actualyl don't even have to match the name of the admin boundary in which they are located). And as these names are optional, they are actually not used except by contributors trying to edit them or reference them. -
Comment from Verdy_p
for your most recent reply: you continue making personal attacks. and asserting false things about what I do or think.
I replied, you refused and argumented, and you don't want me to argument ? Sorry but you are not trying to find any solution, your going to nowhere.
You retierate your arguments, I reitarate mines which are justified (your arguments are found nowhere in any discussion or decision, and you actually want to reduce the help we want to give to users).
Stop this fight: ask to developers of editors and tools to add the missing disambiguation. I won't help much for that, as I do not develop these tools.
Only JOSM offers a disambiguation but JOSM is used by a minority of users. Instead, please post a RFE to iD bug tracker and support if you want this to change... You're speaking to the wrong person as I cannot decide their priorities. -
Comment from SomeoneElse
> Many of these landmasses have their own defined geographical name distinct from the admin boundary
There are examples of this - https://www.openstreetmap.org/relation/375761 is genuinely named "Mainland" and is part of the Orkney Islands. However the "landmass" names are simply invented by you; they do not exist as names anywhere in the real world. Some may be valid _descriptions_ of what they represent but they are not _names_. Remember - https://wiki.openstreetmap.org/wiki/Names#Name_is_the_name_only .
> Stop this fight:
Repeating myself, but I am not trying to pick a fight here - I am trying to help you. OSM progresses in accordance with a number of key princples, including "map what is on the ground", "don't tag for the renderer" and https://wiki.openstreetmap.org/wiki/Verifiability . If you think that there's a problem with a tool used within OSM then you entirely at liberty to suggest an improvement, or (even better) to code that suggested improvement yourself. Personally I don't see that there is a problem to be "fixed", but you clearly do. -
Comment from SomeoneElse
> you continue making personal attacks
What in my comments above constitutes a "personal attack"? The fact that I've linked to your wiki talk page? Throughout I've tried to be positive (e.g. "https://wiki.openstreetmap.org/wiki/Names#Name_is_the_name_only is how OSM handles names") and tried to ask questions about the reasoning behind your changes here and elsewhere (and comments here and elsewhere). Does that constitute a "personal attack"?
If we're to work together then the only way to do that is to try and understand each others' views, and the only way to find those out is to ask.
Please do provide more detail about what you think the problem is - I'm genuinely trying to find out. -
Comment from Verdy_p
"Map what is on the ground" is not applicable. There's no such tyhing on the ground. So these names can ONLY be descriptive and should be made accurately so that it it is still distinct from what would be really on the ground...
This is a case where any choice (with or without the disambiguation) here is equally arbitrary, but then disambiguation is enough to take the priority to all other considerations.You're continuing to repeat yourself, you are truing to force to accept an argument that you have NEVER justified validly. I gave my reasons and they are valid and this does not contradict any past decision or common practices.
You just fight for a principle that I respect, but that is not even applicable here. And systematic principles in mapping are always wrong ! Maps are full of exceptions everywhere, you should know that, otherwise you should not use OSM at all!
-
Comment from SomeoneElse
I don't understand most of what you're trying to say here, but to be clear, I am simply trying to explain some of the key principles of OSM (in each case linking back to what documentation we have), and am asking questions about why you find yourself in conflict with so many people in OSM so much of the time. I'm trying to find the reasoning behind the conflict that occurs (here, and elsewhere - in fact everywhere you come into contact with other mappers). This is consistent with almost every reaction of yours on http://resultmaps.neis-one.org/osm-discussion-comments?uid=90780 and on https://wiki.openstreetmap.org/wiki/User_talk:Verdy_p . To be clear - most people in OSM manage to go through life in OSM without this level of conflict, yet it happens to you all the time. Could it be that you are reacting in a different way to comments than other people do?
OSM values the integrity of the mapping community above almost anything else, and anything that damaged that community can be dealt with harshly. As an example of that, see https://wiki.osmfoundation.org/wiki/Working_Group_Minutes/DWG_2014-07-10_Special_Romania (one of the very few long-term bans that have been placed on users in OSM).
Please think carefully about every response you make to other mappers in OSM. Every time you reply to someone you do not need to "argue" or "fight" with them - it is perfectly possible to state your case and listen to someone else's argument. Sometimes they'll accept what you are saying in its entirety; sometimes you'll accept that you got it wrong and most often a compromise will be reached whereby you both respect each other's point of view and ensure that both of your goals can be met. I can understand that for various reasons this level of understanding of another person's point of view can be difficult, and it sometimes requires a special effort, but it is really important so that we can all work together in OSM. -
Comment from Verdy_p
When I do make errors, I recognize them immediately, and correct fast. But you've not conveinced me that there was an error, so it is an unfruitful discussion. Nothing wxas decided, but you want absolutely from me that I reply to you that I was worng or made an "error" or that I'm "wrong"'. This is going nowhere ! You have an opinion, I have mine, you have absolutely NO right to force me to accept yours. We exposed our arguments , we just disagree, that's all.
But nothing at all was broken in the data, and that's you absoltuely don't want to recoignize it, and it is really an agressive attitude you have adopted.
We can perfectly have opposite opinions without having to force the others to accept one view.
And here I am strongly convinced that none of your arguments apply. You are seeing importance on things that in fact have no importance. you've not demonstrated any break in the logic or any problems this has caused.
Youy are still convinced that I took my own decision, but I only followed the existing practices adopted since many years and that you just want to change ALONE.
Sorry? I will no longer reply you because of your oppressive attitude, which is all but communatary if you still think that there must be ONLY ONE solution and when you still pretned that I really made a real error that you've never demonstrated anywhere. -
Comment from SomeoneElse
> But you've not conveinced me that there was an error, so it is an unfruitful discussion.
Unfortunately, that's not how "compromise" works.
Also to be clear it's not just "my opinion against yours" - I've quoted accepted OSM best practice (and linked to the relevant articles all the way through this process). Simply saying "you are wrong" will not convince people - two other things need to happen. One is that supporting evidence needs to be presented, and the other is that the other person in the discussion has to be willing to consider that evidence. The only evidence that you have presented is https://wiki.openstreetmap.org/wiki/DE:Relation:land_area , and that doesn't say anything about making up names to disambiguate them from any other relations. With regard to the second point you don't appear to have considered the links that I've used here at all (i.e. you haven't said "that's wrong, because..." or "that doesn't apply here, because..."). You've simply said "you're wrong". If we all behaved like that there would be no OSM "community" - there'd just be a bunch of people carrying out edit wars against each other in OSM itself, in the wiki, and on the help site.
If you take away one thing from this discussion please let it be this:
"OSM is a community project and we need to work together to create the best map. This means talking to people, and actually listening to them as well". -
Comment from Verdy_p
STOP STOP STOP!
You are saying "You're wrong since the begining", and have given NO proof at all that there was any error or problem.STOP THIS HARASSMENT WAR ! I absolutely don't have to ackowledge your different opinion, keep your opinion and I keep mine, but NO there has been NO demonstrated error at all.
You have repe"atedly refuisedf to acknowledge the fact that I ha
-
Comment from Verdy_p
that I had effectively corrected bugs, but you insist on something else that is a NON-ISSUE and NOT what I really corrected, pretending that this name has a semantic value, when in fact it has no value except for contributors using the editing tools as a useful hint (because this name has NEVER been formally assigned
the semantic you describe, and there has NEVER been any name used on the ground. It has always been descriptive just like "notes=*", except that it is imemdiately visislbe everywhere in efditors when you are about to make confusions with "similar objects" sharing most nodes and ways but not all.These landmass objects are not formally and legally defined they are almost always estimations of the legal baseline (but with differences because the OSM coastline is on the average high tide, whilst the baseline is normally defined on the average low tide level and include submersible grounds that attach costal rocks to the main land without creating islands, and also will also include water through narrow bays or between chains of islets with narrow passes, or through large estuaries, or through lagoon tide passes as well as all the inner lagoons). The landmasses also incldues riverbanks, lakes and some human constructions over water, which are on the opposite normally ignored by legal baselines used for international borders except after multilateral agreements and proper registration of claims on the seabed!)
Rethink about it: these landmasses have NO legal name at all on the ground or even in books or legal texts. These names are then ONLY descriptive and still have to be clearly distinctive from names used by legal objects if they don't cover exactly the same thing (and this is the case: landmass objects are always created with the purpose to be different from the legal boundary objects, they describe the physical ground, but not the name which does not exist at all on the ground).
-
Comment from Verdy_p
And you pretend that I said only "you're wrong" to you, this is an invention you do! You have constantly used only this argument yourself against me. And you've never presented any evidence. I showed you evidences that whhat you want has NEVER been decided and that it contradicts all past usages made by others since years !
Relations (1-20 of 24)
- 1
- 2
- Монгол улс (161033), v134
- France (terres) (11980), v710
- France (terres) (11980), v711
- Deutschland (62781), v2085
- Jersey (1711283), v26
- Polska (ląd) (936128), v753
- Guernsey (6571872), v4
- België / Belgique / Belgien (land mass) (937244), v467
- Lietuva (sausuma) (4474651), v125
- قطر (3832630), v17
- Jersey (1711283), v24
- Jersey (1711283), v25
- Slovenija (958693), v171
- Российская Федерация (суша) (4452839), v57
- Eesti (4463372), v52
- South Africa (land mass) (1252792), v330
- Polska (ląd) (936128), v754
- Guernsey (land mass) (6571872), v5
- Jersey (land mass) (1711283), v27
- قطر (3832630), v18
Welcome to OpenStreetMap!
OpenStreetMap is a map of the world, created by people like you and free to use under an open license.
Hosting is supported by Fastly, OSMF corporate members, and other partners.
https://openstreetmap.org/copyright | https://openstreetmap.org |
Copyright OpenStreetMap and contributors, under an open license |