Changeset: 136557861
fixed old landuse=reservoir with replace with natural=water and water=reservoir
Closed by avinet_ua
Tags
created_by | OSM Editor UA |
---|
Discussion
-
Comment from silversurfer83
Hi there :)
I just checked the community and couldn't find any topic related to your large scale edit going on. So I'd like to ask you here: Did you discuss this thing you're doing?
cheers
silversurfer83 -
Comment from avinet_ua
do we need to discuss fixes? its basic, old tag, and pair of replace for it
-
Comment from silversurfer83
I believe, according to this [1], we do.
[1] https://wiki.openstreetmap.org/wiki/Automated_Edits_code_of_conduct
-
Comment from elveril
34 622 05 12 67 is the real phone , El Veril is a famous restaurant in Graciosa island , at second time , give a second chance us a corner in the center , of service of bicycles , the other phone is a fake , phone this .. the real El Veril
-
Comment from avinet_ua
what you want to discuss?
-
Comment from silversurfer83
I believe you have to act according to the code of conduct laid out in the wiki before starting something like this, don't you think?
-
Comment from avinet_ua
maybe, but if this just fix of broken tags (!), nothing bad happens. there are lots of fixes, it is very nice, is not it?
-
Comment from silversurfer83
Sure, it's nice.
But this project (OSM) is huge and there are rules for this kind of thing (see my link).
And the tags aren't broken, they are just old.
Please read the wiki page and think about that before you embark on a mission like this.Gotta get to bed now :)
cheers
-
Comment from avinet_ua
thanks, bye
-
Comment from InsertUser
Where was this mechanical edit discussed?
---
Published using OSMCha: https://osmcha.org/changesets/136557861
-
Comment from aTarom
Fully automated? For example,
https://www.openstreetmap.org/way/174925279/history
lavoirs have not been taken into account to properly label? By size or by name... -
Comment from silversurfer83
That's why we have the rules, dear avinet_ua.
Please refrain from further large scale activities without prior discussion with the affected community. In this case, it would be the world wide community which you can find at the tagging mailing list and / or general talk at community.openstreetmap.org -
Comment from Thiskal
https://wiki.openstreetmap.org/wiki/What%27s_the_problem_with_mechanical_edits%3F
I'd also like to offer this wiki page in addition to the one Silversurfer already shared. It provides a good reasoning and explanation for why mechanical edits, even ones that seem harmless or obvious can have bad side effects.
-
Comment from InsertUser
Is someone reverting this?
-
Comment from tux67
There is more like that https://www.openstreetmap.org/changeset/136557767 If the user is not willing to revert I suggest to pass that to the DWG
-
Comment from MxxCon
It's sad that the principle is more important than the usable data. :/
-
Comment from SekeRob
And more
https://www.openstreetmap.org/changeset/136560492Makes it running total of 15K+
-
Comment from avinet_ua
revert just for principle? this is straightforward fixes that do a lots of good things, is not it?
-
Comment from silversurfer83
As others, who invested some time, in this discussion pointed out: these edits, when not done carefully (with prior discussion) usually make a lot of the data worse than before.
So you see, @mxxcon and avinet_ua, the rules are there for a reason.Once again: please discuss these things before embarking on missions that will surely be in vain if undiscussed
- Comment from SomeoneElse
-
Comment from SomeoneElse
Revert started in https://www.openstreetmap.org/changeset/136598791 . It may take some time.
-
Comment from DaveF
@aTarom
What was wrong with this edit by
https://www.openstreetmap.org/way/174925279/historyWhat do you mean by "By size or by name"?
-----------------
OSM should be assessing edits based on quality, not quantity. These edits by avinet_ua fix n tag approved for deprecation.It appears a few contributors are replying purely to try & look important rather than improve the OSM database, which just a waste of time & effort.
-
Comment from SekeRob
hmmm, this mapper did 15K plus mass edits of reservoirs with the title"fixed old landuse=reservoir with replace with natural=water and water=reservoir. If then 1 or more obfuscated in there were not doing that but something else legit then it seems the CS comment was incomplete as well. No one would go seeking out single items in the 15K changes and exclude these from the revert with what was described as a single type change. Tough luck me thinketh.
-
Comment from MxxCon
If 1 out of 15000 changes were wrong, I'll take that accuracy any day. Every single day I see dozens of edits around my city with 100% wrong changes. 🙄
Ok, we had wrong data in OSM. Somebody fixed it. It got reverted because it didn't follow the process. Let's see how long it'll take for somebody to step up and follow that process. In the meantime we have wrong data in OSM.
-
Comment from MxxCon
And where's guarantee that a changeset that through the "proper" process won't have wrong change in it sneak in?
-
Comment from HellMap
Where is the 1 out of 15000 number coming from anyway? Someone mentioned a mistake and someone mentioned 15K and we jumped to the 1 in 15K conclusion?
So I clicked 20 random ways. 2 were natural ponds. 2 were river areas. 2 were gone/dry ground. 2 were swimming pools. 1 was industrial emergency water pool. 1 was waste water plant basin. So that's (unexpectedly) 50% wrong in this arbitrary tiny sample. Sure, it's not technically a wrong change based on swapping deprecated tags. But at the same time, who knows when the feature was added - possibly before a tagging scheme for such features existed.
Personally, I think this should really be a MapComplete task instead and each case manually reviewed.
-
Comment from MxxCon
But such mapcomplete task doesn't exist and didn't exist and who knows when it'll be created and completed. So instead status quo is better?
Were the tags worse before or after the change? -
Comment from HellMap
One example. Someone added `landuse=reservoir` as a waste water basin in 2012, then `water=wastewater` was introduced around 2013. Now this edit changes it to `water=reservoir`. So, yes, the tag is now worse as it implies the original reservoir designation indeed meant a reservoir and not something else.
Anyway, this kind of discussion is exactly why such edits need to be discussed.
-
Comment from tux67
Fully agee with HellMap ... as a community project we have deleloped guidelines as a result of tons of those discussions. Instead of debating here the time could be spend to set up the Mapcomplete task mentioned or find a solution that widely agreed. .. following the process .
-
Comment from ZeLonewolf
Look, some communities are fine with these kinds of edits and some aren't. So when you make a huge edit like this, it will inevitably make a bunch of people mad in a place where they hate bulk editing and cause a revert. I would suggest familiarizing yourself a bit more with OSM culture in that regard and I wish you success in communities that are more amenable to your preferred contribution style.
-
Comment from Carnildo
In the United States, the GNIS import (itself a bad idea) brought in many different types of water-storage objects as "landuse=reservoir". Those need to be properly sorted out, not blindly re-tagged.
-
Comment from DaveF
Auto/bot/mechanical edits amend one tag on one type of object. If (& it's a big if) there are amendments to other objects they stand out like sore thumbs. edits are much more likely to be erroneously missed where a contributor changes multiple tags on multiple different objects, even if it's only over a small area.
It's disappointing to see OSM is attracting people who are not really interested in its primary purpose of providing *accurate* data to create maps, but just here to 'manage the process'.
Please remember this anti mass edit ruling was created, in a display of blatant 'do as we say, not as we do' arrogance, by those who oversaw the crap Tiger import in the US & yet couldn't be bothered to fix it.
-
Comment from DaveF
@HellMap (and @tux67 & Carnildo)
So, by your own admission Your argument is straw man. *All* of the errors were in existence prior to these edits. These edits created no errors. The only reason you've become aware of these errors & thnk MapComplete is a solution to fixing them is due to edits occurring. Discussing them before hand wouldn't have enlightened you. If there are errors, fix them; instead of moaning about other edits which improve the quality of the OSM database.@ZeLonewolf
As the edits haven't created errors, only tidied up deprecated tags that have already been discussed at length, what's your logic for "getting mad"? Communities should assess on quality, not quantity. -
Comment from ZeLonewolf
@DaveF, I live in a pragmatic world where I'm concerned about actually getting stuff done. No amount of whinging on a changeset comment will change the fact that some communities and individuals are heavy on rule-lawyering and perfectionism. With this reality in mind, any edit that will get reverted is a waste of everyone's time. Despite what you seem to imply above, the US is one of the more amenable places to bulk and mechanical editing. However, it still requires a base understanding of community expectations.
I similarly bulk-edited 30,000 or so riverbank tags in the US, with community support and a good understanding of where the data came from and how it came to be. So I am absolutely a fan of bulk and mechanical
See:
https://wiki.openstreetmap.org/wiki/WikiProject_Waterways/River_modernizationI largely agree with Carnildo's description of reservoirs tagged from the GNIS import. Tens of thousands of farmer's cow ponds in North America are tagged as reservoirs, and the presence of the old tag gives us something to go off of in sorting these out into "reservoir" and "not reservoir". But, if you've never looked at the data or talked to people who map these areas, you'd never know about them. Merely upgrading the tag makes finding and fixing problem data harder.
Meanwhile, after a brief discussion on the Czech mailing list, every landuse=reservoir in the Czech Republic was upgraded to the new tagging. The Czech mappers understood that no underlying data issue made this problematic.
In the Czech case, it was a quick note on their mailing list, and then the edit a few days later. In the riverbank case in the US, it was quick notes on Slack once it was clear there was no opposition. Bulk editing doesn't need to be excessively process-driven, but it does need a minimum level of understanding of the data that's being edited and at least a courtesy interaction with the community involved.
-
Comment from SomeoneElse_Revert
This changeset has been reverted fully or in part by changesets 136609773, 136639484, 136633089, 136645524, 136654047, 136623452, 136614651, 136672560, 136641952, 136607063, 136663965, 136598791 where the changeset comment is: Revert undiscussed landuse=reservoir edit. See https://www.openstreetmap.org/changeset/136557861 and https://www.openstreetmap.org/user_blocks/7146
-
Comment from SomeoneElse
Well, that took a while :)
It'd be great if some of the excellent points above were discussed a but more widely. Perhaps https://community.openstreetmap.org/c/help-and-support/tagging/71 might be a suitable place?
Also, I'm sure that there's some "low hanging fruit" among the reverted data that "obviously isn't any sort of reservoir". I've done a couple of the more obvious ones from earlier in the series of series of reverts, but the 394 nodes (including at least one mobile phone shop!_ on https://www.openstreetmap.org/changeset/136672560 are surely worth a look.
Best Regards,
AndyPS: Other reverts of changes by this user are still ongoing (currently https://www.openstreetmap.org/changeset/136671982 - there will likely be a couple more in that series too), as we have had what might be a record number of separate individual complaints about their edits.
- 10277626, v2
- 10277632, v2
- 10277634, v2
- 23884784, v5
- 23884785, v5
- 23967501, v8
- 24161855, v3
- 24254770, v4
- 24259023, v3
- 24607904, v7
- 24940322, v4
- 25696642, v4
- 25787212, v3
- 26109071, v4
- Étang Bull (26498245), v19
- Reservoir de Pomery (27683331), v3
- 29078468, v3
- 29078472, v3
- 29078473, v5
- Étang de Cravanche (29263039), v8
- Bassin de Balizy (110572), v3
- Plan d'eau (178452), v3
- Lac de Vinça (1321991), v4
- 1627236, v3
- 1627237, v3
- 1627239, v3
- 1627240, v3
- 1627308, v3
- Étang du Chalam Bief (1690605), v3
- 1696173, v2
- 1696174, v2
- 1696816, v2
- 1696817, v2
- 2006885, v3
- 2064720, v3
- 2189228, v3
- 2534540, v3
- 2733594, v2
- 2798191, v3
- 2869378, v3
Nodes (7)
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 |