We don't need anonymous notes

But then one should not complain that anonymous note writers do not respond to queries.

Unfortunately, this was necessary. Thus a few idiots have ruined a low-threshold offer for the improvement of open geodata :frowning:

Even if anonymous notes get removed, it won’t push those anonymous users that much to create an account in order to create a note, as the anonymous notes will still be possible through third-party apps. In this case, you already have almost no chance of that anonymous user to reply, as they are possibly never notified for comments on their notes, only the OSM accounts through which those notes are being created.

There are few occasions of anonymous notes that are actually useful, no one can deny that, usually in places with extremely low mapping activity (at least in the case of Greece), which makes it even harder for a typical mapper to verify the information.
I wouldn’t want anonymous notes to be stopped, but also I believe changes should happen, just not sure what else can occur, especially since the anonymous replies are stopped now.

I had an Anon Note last week, where they were complaining that routing wasn’t working properly. Couldn’t replicate the problem so resolved it with that comment.

The same “anonymous” user has then raised another Note to ask me if I looked at the OSMAnd error he’d mentioned, & pointing out what the problem is with reference to satellite imagery & so on (& it turns out one of the roundabouts in question wasn’t connected properly, so there may have been an error?).

From that response, you’re pretty obviously NOT just some person who’s spotted a problem, so why don’t you just fix it yourself? :roll_eyes: :face_with_raised_eyebrow:

this offloads filtering effort on others, and mostly avoids this type of notes:

Notes submitted or commented on by Mateusz Konieczny - bot account | OpenStreetMap

Notes submitted or commented on by Mateusz Konieczny | OpenStreetMap

1 Like

onosm.org is looking rather good in this snapshot by comparison. :sweat_smile:

I don’t see them that often in the area I’m working on.

I just look if there’s any helpful information - that I can use and if possibly the source is illegal or incompatible with the ODbL.

For the source issue - I always follow the golden rule “innocent until proven guilty” so as long as I don’t clearly see that the source is illegal or incompatible - I assume legit source.

and then the third I close them… simple as that!
If I can use the note I do so
and if not, or if the note is something only the anonymous contributor would understand, I always close with the comment - feel free to reopen and add valuable information for us mappers so we can map whatever it is you wish us to map for you - or feel free to register to OSM and contribute yourself.

it doesn’t take much of my time to be honest - so I’m not worried much about anonymous notes.

And if anonymous note is open 10+ years - I simply close with the remark:“Can’t verify if outdated or not, feel free to reopen with the information if it’s still valid or not then I can process it”

with 5yo notes - I sometimes do a bit of digging myself to see if there’s anything still actual on that note.

But it depends, some notes have some more explicit valuable information and that I can use then as it is.

But anyway - I don’t feel troubles by anonymous votes… closing them with a comment of my own isn’t really a big issue now is it?

2 Likes

It’s possible to remember anonymous note id client-side and check its state periodically.

1 Like

I recently added a business from an onosm.org note, tracking it past a business name change and several other details (the note was from 2020), so I think those are valuable side-quests and I’ll be pursuing them more often when I see them.

Missing/closed business notes can be used safely in places where street-level imagery is available and updated regularly.

1 Like

Just leaving these nice new anonymous notes here

https://antonkhorev.github.io/osm-note-viewer/#mode=bbox&bbox=20.962%2C57.463%2C28.740%2C59.679&closed=0&map=8/58.405/23.824

(I did report this)

1 Like

Is it possible to rate-limit them first? As decided on changesets. While this doesn’t prevent switching IP, it’s a start to make it more difficult.
The closing of gibberish ones could be automated (?) now @mavl , but it’s not always catching all. Needed to help finish closing at least 40 notes (relatively low for spam?) recently.
May be more important to limit exactly overlapping notes at the same coords.

vast majority of gibberish notes are hidden by DWG minutes/hours/days after being created so it is partially happening

if DWG spends significant amount of time on that, then it could be good enough reason to drop anonymous notes completely

I also occasionally run closing blatantly pointless notes like “SITE” or “start” or “.” as entire content of note (closures are human reviewed). So if you are familiar with some repeated note content that cannot be ever valid, and there are old unclosed notes like this feel free to ping me.

I assumed that it is happening, could be wrong (if there is indicator of no limit or it being too low and there is no issue at GitHub - openstreetmap/openstreetmap-website: The Rails application that powers OpenStreetMap then it could be a good idea to create it)

If it is already in force, maybe it should be stricter. Tens of them is still too many without some dedication or knowledge, and clutters other services showing them. I don’t remember if the limit is kept secret to avoid workaround, but it could be disclosed now, following the transparency in changeset limits. Aside from changesets, only saw info about accounts and messaging on Github.
As how this forum limits too short or repetitive titles, it may not be far-fetched to consider implementing a gibberish detector on the website to prevent the creation of these spam notes in the first place?

notes are internally comments like changeset comments

I proposed rate-limiting changeset comments in rate-limit changeset comments · Issue #4196 · openstreetmap/openstreetmap-website · GitHub

This was implemented by Tom Hughes in Add rate limiting for changeset comments · openstreetmap/openstreetmap-website@a274726 · GitHub

sadly I lack understanding of code and have not run it: but I see no special handling of anonymous comments. So likely either each note is treated as first one or anonymous user is treated as long standing account with maxed-out rate limit.

Notes are notes and changesets are changesets and the two are not related in any way.

But from rate-limit changeset comments · Issue #4196 · openstreetmap/openstreetmap-website · GitHub and rate-limit changeset comments · Issue #4196 · openstreetmap/openstreetmap-website · GitHub changeset comments and note comments are the same thing in data model. And close/open actions are also comments and also rate-limited by changeset comment rate limits.

And initial note appears to be just first comment from my understanding.

Or is changeset comment and note comment the same thing, but note is a separate thing and is not rate-limited by comment rate limit?

I think Mateusz alluded to the fact that, internally, notes don’t distinguish between the actual note text and comments.Note text is a comment with action="opened" in the XML.

No that is just wrong - note comments are in the note_comments table which is reflected in the NoteComment model and changeset comments are in the changeset_comments table which is reflected in the ChangesetComment model.

As such there is no rate limiting of note comments either for logged in users or for anonymous users.

I think you misinterpreted my reply to your comment on that ticket. I wasn’t trying to say that note and changeset comments are the same, just that notes don’t maintain any record of open/close events other than implicitly via the comments.

1 Like

Anonymous notes are not a problem at all.

There are many notes from registered users with useless text (opinions, personal data, temporal stuff, ads) and lousy location, even when they have photos or GPX trace (a new StreetComplete option). Therefore, the problem is not whether the note is anonymous or from a registered user; the problem is about its content and how helpful it is to perform a map change to be considered feedback.

Note location cannot be granted as exact. The GPS from mobile devices is approximate, and the mapper who solves notes should be aware of this.

Anonymous notes from a form like OnOSM are good because the map user should have filled in the fields before submitting it. In this case, there was a filter before creating the notes, and it takes time to place the note and then fill in all the fields. And I know the Iranian fork issue (It was me who identified that issue while solving notes in the middle of the ocean), but this is a side issue for OnOSM.

Instead, identifying the application that created the note could give an idea of whether the note was created on the field or remotely (for example, from the armchair mapper computer). Knowing the application could provide an idea about the precision of the note and if the note was just an aerial analysis or a real validation on the ground. I have created GitHub issues on many applications, and they have started to include a hashtag as part of the note creation. However, the bureaucracy of the OSM website and its slow response have yet to include the hashtag. This is a list of notes that its source application can be identified.

  • 1016 - EveryDoor
  • 21 - LocusMap
  • 236 - MapComplete
  • 451041 - Maps.me
  • 24447 - Mapy.cz
  • 452 - msftopenmaps
  • 124273 - OnOSM.org
  • 158 - OnOSM.OSMiranorg
  • 9004 - OrganicMaps
  • 1665 - OsmAnd
  • 446525 - StreetComplete

Forcing a map user to register in OSM to write a note will lead to a drop in the user’s motivation to give us feedback.

One thing in favor of notes from registered users is the ability to write a message to its creator. However, I have yet to see this especially useful. In LatAm, we have prepared a good message as part of the closing notes process, and we have closed thousands of notes, but we have seen just a few reactions to this message. So, this is not a persuasive argument for the non-anonymous notes. OSM internal messaging system is important for the mapper community (all of us who are reading this message), but not the people who use the map (OSM messages are redirected by default to SPAM in main email providers). I have written hundreds of messages to different users for several reasons and received few responses, only from real mappers.

Also, let us remember that creating an OSM account is simple; it just needs an unused username and a real email, but anyone can have multiple email accounts. In any case, this is not a strong filter, and it is prone to vandalism.

The OSM database has the IP of each comment including including the opening one, but this value is not available via API or daily Planet dump. This could help us to find the origin of those wrong notes. However, I wonder why IP should remain secret for all notes. (If there is secrecy, the note can be reported and hidden from the public).

OSM is a flexible map, unlike other commercial or official maps. Anyone can add something to the map; OSM has no contribution levels. This is another reason there are many fundamental issues on OSM, but that is also part of the beauty of this project, and that is why many people have started to contribute (not like the rigorous process of Wikipedia, where contributing is no longer fun) without cartography knowledge.

Another problem is the old notes, no matter if a registered user created the note 10 years ago or if it is anonymous (the registered user could have forgotten the note or does not have access to the email). The problem is if the note is still valid. This isn’t easy to figure out, and each community must make decisions about them; it could be like analyzing them and closing them mechanically. Let us remember that the OSM database also holds POIs dating back ten years ago that are not valid, and no one has complained that much about them (StreetComplete is trying to solve this by adding an extra tag – check_date).

The real issue behind this is the community of mappers (many of us) that still need to process notes of our territories, and many countries have hundreds of thousands of open notes. And we are trying to find a way to get rid of them, even by hiding or removing them.

The real solution to this problem is to try to keep ZERO open notes. That is why in LatAm, we have Notathon sessions, when we gather online to close the notes of our countries, and this has had a beneficial effect; Ecuador and Cuba are significantly reducing their number of open notes with just a few mappers. Another important thing is to monitor its area: Check new notes constantly, check recent changesets, welcome new users, etc.

Let us follow the examples of Chile, Taiwan, or Australia, where a note is processed in less than a month. In Colombia, we are trying to follow their example, but it has not been easy with a couple of mappers, and that is why we always invite to close notes as part of the mapping learning process for new mappers

Location of open notes as of 2023-12-21. Bright red are recent notes; dark red are old notes

6 Likes

We’re trying to save cartographers’ time so that they are engaged in really useful notes.

I can give you an example. Many StreetComplete users leave notes about new shops, while StreetComplete has a Shop overlay where you can add them. Due to the non-anonymity of the users, I was able to contact them and specify this method.


With anonymous users, however, I see the opposite. You can’t ban a user, nor can you suggest using the editor. I’ve already mentioned a user leaving incorrect data from dubious sources and notes from onosm, which if there are a lot of them, there are not enough hands to handle.

Here's another one:

Anonymous user left a hundred of similar notes about missed POIs around the city. Useful information? Yes. Is it convenient to enter them? No. Will anyone want to process them? Unlikely. But the user could just open a normal editor and enter the POIs himself.


By the way, it brought to mind an interesting idea: we can strongly limit the number of notes per IP (up to five for example) and offer the user to log in.

However, I have doubts that all such mass notes are made from osm.org and not, for example, from OsmAnd.


Speaking of new mappers for solving notes. We don’t have normal editors working with notes.