Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
On 2020-08-24 00:18, Martin Koppenhoefer wrote: > I have recently found a lot of highway=path which clearly were tracks > according to aerial imagery. A tool which would allow to filter for “paths by > this mapper” (maybe in a similar timeframe) could speed up finding and fixing > them. Probably this could be answered by an Overpass query. If you happen to have more details, we can tell you how. -- ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
sent from a phone > On 23. Aug 2020, at 21:41, mmd wrote: > > That's a pretty dystopian view on the OSM future, if you ask me... I did not mean to callout mappers, but it could help to highlight potentially weak parts of the map where a resurvey could make more sense than in other parts or also to interpret map information according to mapping styles of the mappers. It is not necessarily about “good“ and “bad“ mapping. Another example: I have recently found a lot of highway=path which clearly were tracks according to aerial imagery. A tool which would allow to filter for “paths by this mapper” (maybe in a similar timeframe) could speed up finding and fixing them. Cheers Martin ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
Hi mmd mmd skrev: (23 augusti 2020 21:38:45 CEST) >On 2020-08-23 18:27, Martin Koppenhoefer wrote: >> There is a lot of stuff that could be analyzed, immense. All the >history is still available with all the user information... > >What's next? Do we want to invite "unreliable" mappers to an exciting >two hours training course to improve their railway mapping skills? Upon >successful completion of the final test, they can earn 5 extra mapping >days that count towards their 42-day threshold for an OSMF active >contributor membership. LOL. I don't believe in punishment or restrictions based on past performance. I believe that people do the best they can as much as the time they can. OSMF have lowered the bar of entry to people with little money which is good. More engaged members in OSMF is a good sign. Of course this new policy could also backfire and lead to 42 days of crap edits for all our millions of users in the worst case, but I hope not. > >That's a pretty dystopian view on the OSM future, if you ask me... I hope this reliability index will be used for good. I have no intention of using it to pick on others. But maybe someone we ill be curious about to he results. We could create a page that suggests what a particular user can do to get a higher score. E.g. take extra care to avoid integrity errors on your multipolygon edits. Read this page x, contact your local mapping community to find a mentor ask in a forum for help with reviewing an edit. We already have a light version here: https://hdyc.neis-one.org/ where you can profile any user after login. I never used user statistics in conversations with others because they are judgements best used to evaluate a whole. I dislike judgements about humans. Ranking and competitions are not my favorites either, but I love learning with others. So when I write other contributors I only refer to facts about a concrete edit and ask what they tried to do. I have never seen anyone anywhere in OSM use data to try to control others which is a good sign. Cheers ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
Hi Martin Martin Koppenhoefer skrev: (23 augusti 2020 18:27:58 CEST) > > >sent from a phone > >> On 23. Aug 2020, at 13:55, pangoSE wrote: >> >> We could e.g. set a verification-needed >> flag on objects edited in a changeset with "please review". > > >while you can (already) add a fixme tag, I fear that creating a special >feature for less reliable information could lead to people being >encouraged to adding more “guess work” because they “set the unreliable >flag so what’s the problem?“ Yeah, that's a good point. We are social animals. > >I just had an idea: You could calculate a reliability index for each >and every object in OpenStreetMap (and maybe for each of their tags, by >looking at the mapping experience of the person that added it. Beautiful idea! I'm gonna try that with a small country. Working on a small excerpt of the planet could be problematic because the whole user history will not be available. Hmm that means big data is the only viable way forward. >In a >more complex iteration, it could also take the reliability of specific >mappers into account by analyzing whether things they add or modify are >kept or changed by following mappers (and it would probably have to >take time into account, because if something is changed after a long >time it is more probable that it was because of a change in the real >life and not because of bad representation, and maybe also the kind of >change). I love this idea too. This is what I have done in my head editing in Sweden for multiple years. I have a very short list of editors that frequently map in a way I don't like or make errors e.g. things showing up in keep right and they were the last editor. Usual suspects . I always try communicating with the them and most respond and we find a way forward, but some never react to changeset comments and just keep on what they are doing. Not reacting to changesets comments is another red flag. >It could also be done according to the field of thing (e.g. >this mapper does reliable work with buildings or this mapper is an >expert for outdoor routes but does poor work in cities, or is an expert >for railways, etc. etc.) Yes this is a good observation. OSM is hard. It takes time to learn all the long ropes. > >There is a lot of stuff that could be analyzed, immense. All the >history is still available with all the user information... I get your point. You could also flag changesets with huge BBOXes and filter away those done by experienced mappers and those concerning one big relation. Using to this search https://duckduckgo.com/?q=osm+history+analysis I just found https://heigit.org/big-spatial-data-analytics-en/ohsome/ which seem very promising I will contact them and see if I can use and contribute to their platform to get the information I want. A good algorithm for finding and rating experienced mappers is crucial. If anyone already has made one or ideas for improvements please share Feel free to add to this wikipage https://wiki.openstreetmap.org/wiki/Algorithms_for_QA I just signed up for Fuga cloud and I'm gonna start playing with the history data in python and postgresql to crunch the numbers for a small country if ohsome turns out not to be suitable. Thanks for sharing your ideas Cheers pangoSE ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
On 2020-08-23 18:27, Martin Koppenhoefer wrote: > There is a lot of stuff that could be analyzed, immense. All the history is > still available with all the user information... What's next? Do we want to invite "unreliable" mappers to an exciting two hours training course to improve their railway mapping skills? Upon successful completion of the final test, they can earn 5 extra mapping days that count towards their 42-day threshold for an OSMF active contributor membership. That's a pretty dystopian view on the OSM future, if you ask me... -- ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
On 8/22/20 03:12, pangoSE wrote: > Maybe we should have some kind of system flagging objects that has not > been edited for x number of years and rate all objects in the database > according to this? Even if something is edited, not everything on the object will necessarily have been verified at the time of that edit. Especially with armchair mappers fixing errors found by QA tools like Keep Right or doing single-purpose cleanups on opening_hours/service_times/collection_times tags, etc. > This would mean that a data consumer can decide based on the score if > they want to include the information or not. > > E.g. a high quality map should perhaps not contain objects with a > revision older than 3 years (and no references or sources) Some things just don't change over a period of three years. Had it been added when I started mapping (2012-ish), the house I'm in now (actually, most houses in this neighborhood) would have had no reason to be edited over that time. > Or even better: we could implement a verification system with a log that > can be queried easily. > > IMPLEMENTATION SUGGESTION: > > GET Openstreetmap.org/api/verifications/ > Lists latest added verifications (outputs 10 entries, can be > used to get more, can be used to output up to 300 entries) > > GET Openstreetmap.org/api/verifications/1234 > Outputs verifications for osmid 1234 with the newest first (outputs 10 > entries, can be used to get more, can be used to output up > to 300 entries) > > POST Openstreetmap.org/api/verifications/1234 > Add a new verification for osmid 1234 > > On openstreetmap.org we have a new button for every object "Verify this > object exists and is correct" which stores the date and userid in the > database. > > In JOSM we could add the possibility to download verification data for > all selected objects or from a new option in the download dialog. > > The latest verification date and count of verifications could be made > available in a separate dump. > > If we had such a system I believe the map data quality could increase > considerably by making it dead simple to hide hide old unverified data > from e.g. openstreetmap.org. A high-quality map we can be proud of could > also give an impetus to local mappers to revisit trails and verify them. > > WDYT? How big will this database need to be? Who's going to store it and maintain it? -- Shawn K. Quinn http://www.rantroulette.com http://www.skqrecordquest.com ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
On Aug 22, 2020, at 11:38 PM, pangoSE wrote: > Shawn K. Quinn" skrev: (23 augusti 2020 00:31:28 CEST) >> >> The big, huge difference between Wikipedia and OSM is that Wikipedia >> does not allow original research at all, whereas OSM thrives on the >> original research of everyone who contributes and in fact it is the >> stuff that comes from third parties that has to be vetted more closely >> for license compliance and copyright issues. Very well stated by Shawn. >> I agree we could do better in the quality control department but a lot >> of things added to OSM will be added there first before any third >> parties pick them up. That makes references a bit problematic, IMO. Not very well stated by Shawn, as it doesn't specify a problem. Problematic, yes, but ambiguously so. What problem? > All edits in OSM must be verifyable on the ground if I understood this > correctly: https://wiki.openstreetmap.org/wiki/Verifiability Verify-ABLE isn't the same as MUST-verify. The former has a relatively low bar, the latter, a high one. > Problem is to really make this easy to review without visiting the same spot > we would in many cases need a good photo or perhaps multiple photos from > different angles. That's one solution, but not the only solution. What problem are you trying to solve? > Unfortunately we neither encourage nor support image uploading anywhere > hosted by ourselves or others (we could probably easily integrate mapillary > uploading in the website and in our mobile tools. I take photos with > osmtracker sometimes but cannot upload them to mapillary from inside JOSM). > I'm not saying it should be a demand, but I think we would gain a lot in many > changeset discussions if adding images to the chat and changesets is made > possible or if images in mapillary in the area were visible and referencable > on the changeset discussion page. Whenever I hear somebody say about OSM "it can't" I immediately think "well, it could." It might be a lot of work and can often happen outside and around OSM (with a wrapper, with an API, with a layer of spaghetti-to-spaghetti translation...), precluding the necessity that a wholesale tagging change take place within OSM (as it appears this proto-proposal would, but remains too vague for me to be sure). > Alternatively we could cook our own image storage service if we want. We got > the money for it now and commercial persistent object storage solutions are > available from multiple providers releasing the burdon of infrastructure > maintenance on our operations working group. WDYT? With all these bubbling ingredients, I'm still unclear what it is you are trying to brew. Will you cook first and taste along the way? Will you develop the recipe first before cooking anything? Will it be a cake, a beverage, a repository, a translator, a fast storage exchange mechanism, a portal between other naming / semantic identity hives, what, exactly? Spec it out! > This and my proposal to mark features as verified at this point in time could > potentially make it much easier to judge the overall quality of our data and > map. Now it sounds like we get closer to what you (or you and others might be aiming to do): judge the work / data of others. Judges are made, not born. I feel OK judging certain software and data, this is after decades of software development and quality assurance engineering at Silicon Valley giants and startups alike. May I ask pangoSE to offer qualifications and / or a portfolio of work by which we might elevate him / her to such an important position? > We would still be lacking a REAL granular referencing system where every > statement (tag) is references individually with a date, author and optionally > a photo. That would be really awesome, but it would require additions to the > main database model and ruby website to support (this is perhaps a perfect > GSoC project). Being able to browse to a specific tag on an object and > discuss that would be a crucial addition to the website because now we are > forced to comment on the changeset (or sending pms) and I think its really > cumbersome to manually reference which one of the sometimes hundreds of > objects I'm talking about. Please stop cheerleading this unclear concept, instead, spec it out. This list and the wider OSM community will read that and see if it might have merit. Such things in OSM merit their way in, they don't force or crowd their way in by a single, vocal individual. Unless and until they are well-presented. So, make a presentation rather than a list of complaints with a wish list. > Andy Allen (he runs http://www.thunderforest.com/ which has a nice vector > map service by the way on a free limited tier) a former member of the > operations working group and current co-maintainer of the rails website > posted this a year ago: > https://gravitystorm.github.io/osmf-infra-plans/ and this july the OSMF and > the
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
sent from a phone > On 23. Aug 2020, at 13:55, pangoSE wrote: > > We could e.g. set a verification-needed > flag on objects edited in a changeset with "please review". while you can (already) add a fixme tag, I fear that creating a special feature for less reliable information could lead to people being encouraged to adding more “guess work” because they “set the unreliable flag so what’s the problem?“ I just had an idea: You could calculate a reliability index for each and every object in OpenStreetMap (and maybe for each of their tags, by looking at the mapping experience of the person that added it. In a more complex iteration, it could also take the reliability of specific mappers into account by analyzing whether things they add or modify are kept or changed by following mappers (and it would probably have to take time into account, because if something is changed after a long time it is more probable that it was because of a change in the real life and not because of bad representation, and maybe also the kind of change). It could also be done according to the field of thing (e.g. this mapper does reliable work with buildings or this mapper is an expert for outdoor routes but does poor work in cities, or is an expert for railways, etc. etc.) There is a lot of stuff that could be analyzed, immense. All the history is still available with all the user information... Cheers Martin ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
Hi Martin Den Sat, 22 Aug 2020 19:30:23 +0200 Martin skrev Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !): > sent from a phone > > > On 22. Aug 2020, at 10:15, pangoSE wrote: > > > > Here is yet another example of bad data in our database: > > > fix it ;-) Yeah! But to fix "it" (it being the overall low or unknown quality of the map) we need good tools that encourage reviewing and fixing. We have a discoverability and usability problem IMO in this area. I fixed loads of errors during my time and I like it, but I have a poor grasp of the overall quality of the map in the area and OSM is not making it easy for me to find the less good quality spots with stale/old/non-reviewed data. > > Of course OpenStreetMap contains errors, just like any other source, > and probably more, given that most contributors are laymen and have > very few experience (few total edits, often just 1). > > On the other hand, we may be very fast when something changes, very > flexible in emergencies (think Haiti), and have interesting niche > data that commercial and public data providers don’t care for. Yes, that really nice. I would like to find a middle ground between fast and poor/unknown and slow and high degree of verification. > > It all depends on the local community in the end. If you have reached > a critical mass to have locals everywhere, it will work great and > bugs will wash out. Otherwise the data might get stale just like any > other data. Also using the data is essential to find the problems, > for example the 212 story garage is likely fixed now ;-) Yeah, I agree. Lets make it easy for a local community to keep the map verified and up to date. We could e.g. set a verification-needed flag on objects edited in a changeset with "please review". That would make it easy create an overview of all things todo in your local area based on the objects - not the changesets that touched them (they can easily be found in todays interface from the object). /pangoSE ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
pangoSE skrev: (23 augusti 2020 08:38:45 CEST) > >Andy Allen (he runs http://www.thunderforest.com/ which has a nice >vector map service by the way on a free limited tier) a former member >of the operations working group and current co-maintainer of the rails >website posted this a year ago: >https://gravitystorm.github.io/osmf-infra-plans/ and this july the OSMF >and the operations working group announced hiring of a Senior Site >Reliability Engineer: >https://mobile.twitter.com/OSM_Tech/status/1287395222847139846 > >This seems like a good move. We would benefit a lot from being able to >easily load balance and adjust VMs on our own or someone elses >openstack infrastructure where we can easily provision new servers for >development or testing when needed instead of having dedicated physical >hardware servers that causes availability issues if they break because >of single point of failures. Speaking of free software like openstack: here are a few companies that contribute to and use openstack: * the swedish company City Cloud seems to be the only one hosting their cloud on free software with multiple re. They are not the cheapest, but they goe datacenters in many regions and have a lot of ISO certifications and are used to dealing with GDPR compliance. https://citycloudng.com/ * the dutch company Fuga. They do not disclose how many locations they got so I'm guessing only one in the Netherlands. They are way cheaper than City Cloud it seems and are ISO certified and GDPR compliant. https://fuga.cloud/about-fuga/ /pangoSE -- Skickat från min Android-enhet med k9.___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
Hi Shawn "Shawn K. Quinn" skrev: (23 augusti 2020 00:31:28 CEST) >On 8/22/20 03:26, pangoSE wrote: >> I meant that a verification system does exist in Wikipedia and they >> now require references on all statements to keep up the quality of >> the articles which is sane IMO. We have no such system. > >The big, huge difference between Wikipedia and OSM is that Wikipedia >does not allow original research at all, whereas OSM thrives on the >original research of everyone who contributes and in fact it is the >stuff that comes from third parties that has to be vetted more closely >for license compliance and copyright issues. > >I agree we could do better in the quality control department but a lot >of things added to OSM will be added there first before any third >parties pick them up. That makes references a bit problematic, IMO. All edits in OSM must be verifyable on the ground if I understood this correctly: https://wiki.openstreetmap.org/wiki/Verifiability Problem is to really make this easy to review without visiting the same spot we would in many cases need a good photo or perhaps multiple photos from different angles. Unfortunately we neither encourage nor support image uploading anywhere hosted by ourselves or others (we could probably easily integrate mapillary uploading in the website and in our mobile tools. I take photos with osmtracker sometimes but cannot upload them to mapillary from inside JOSM). I'm not saying it should be a demand, but I think we would gain a lot in many changeset discussions if adding images to the chat and changesets is made possible or if images in mapillary in the area were visible and referencable on the changeset discussion page. Alternatively we could cook our own image storage service if we want. We got the money for it now and commercial persistent object storage solutions are available from multiple providers releasing the burdon of infrastructure maintenance on our operations working group. WDYT? This and my proposal to mark features as verified at this point in time could potentially make it much easier to judge the overall quality of our data and map. We would still be lacking a REAL granular referencing system where every statement (tag) is references individually with a date, author and optionally a photo. That would be really awesome, but it would require additions to the main database model and ruby website to support (this is perhaps a perfect GSoC project). Being able to browse to a specific tag on an object and discuss that would be a crucial addition to the website because now we are forced to comment on the changeset (or sending pms) and I think its really cumbersome to manually reference which one of the sometimes hundreds of objects I'm talking about. Andy Allen (he runs http://www.thunderforest.com/ which has a nice vector map service by the way on a free limited tier) a former member of the operations working group and current co-maintainer of the rails website posted this a year ago: https://gravitystorm.github.io/osmf-infra-plans/ and this july the OSMF and the operations working group announced hiring of a Senior Site Reliability Engineer: https://mobile.twitter.com/OSM_Tech/status/1287395222847139846 This seems like a good move. We would benefit a lot from being able to easily load balance and adjust VMs on our own or someone elses openstack infrastructure where we can easily provision new servers for development or testing when needed instead of having dedicated physical hardware servers that causes availability issues if they break because of single point of failures. See also https://operations.osmfoundation.org/ BTW osm-fr already made this move and is mostly running VMs now and has moved some of their VMs (heavy tile rendering) into the OVH cloud to manage their hardware more efficiently. See https://wiki.openstreetmap.org/wiki/FR:Serveurs_OpenStreetMap_France Cheers PangoSE ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
On 8/22/20 03:26, pangoSE wrote: > I meant that a verification system does exist in Wikipedia and they > now require references on all statements to keep up the quality of > the articles which is sane IMO. We have no such system. The big, huge difference between Wikipedia and OSM is that Wikipedia does not allow original research at all, whereas OSM thrives on the original research of everyone who contributes and in fact it is the stuff that comes from third parties that has to be vetted more closely for license compliance and copyright issues. I agree we could do better in the quality control department but a lot of things added to OSM will be added there first before any third parties pick them up. That makes references a bit problematic, IMO. -- Shawn K. Quinn http://www.rantroulette.com http://www.skqrecordquest.com ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
On 8/22/20 03:20, Mateusz Konieczny via talk wrote: > Nobody claims OpenStreetMap data contains no mistakes. There are a lot of cases where OSM data is better than that in Google Maps, Mapquest, Bing Maps, etc. Unfortunately there are also a lot of cases where the converse is true; in particular, we have almost no addressing data save for the few places where dedicated mappers have added it via exhausting on-foot surveys (not to be confused with exhaust*ive* surveys, speaking from experience here) or gotten lucky enough to score a compatible import. To its credit, Vespucci at least tells mappers "object may be out of date" when it has sat unedited for over a year. I have missed out-of-date data sitting right under my nose, the best example of this being the Whitehall Hotel in downtown Houston (finally noticed and fixed a while back). > https://wiki.openstreetmap.org/wiki/Notes are designed as tool allowing to > describe incorrect data that someone is unable or unwilling to fix (and > yes, we have thousands of reports of mistakes) I have also used notes (and seen the notes feature used by others) to quickly note business information that I can't add in Vespucci or another app right then and there. Yes, I do close a lot of my own notes, and I suspect I'm not the only one. -- Shawn K. Quinn http://www.rantroulette.com http://www.skqrecordquest.com ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
>it was one person in CA adding 400 unverified tags to rail service in chicago. > >one just 818 m, away from my home. > >>Saturday, August 22, 2020 12:32 PM -05:00 from Martin Koppenhoefer < >>dieterdre...@gmail.com >: >> >>sent from a phone >> >>> On 22. Aug 2020, at 10:15, pangoSE < pang...@riseup.net > wrote: >>> >>> Here is yet another example of bad data in our database: >>fix it ;-) >> >>Of course OpenStreetMap contains errors, just like any other source, and >>probably more, given that most contributors are laymen and have very few >>experience (few total edits, often just 1). >> >>On the other hand, we may be very fast when something changes, very flexible >>in emergencies (think Haiti), and have interesting niche data that commercial >>and public data providers don’t care for. >> >>It all depends on the local community in the end. If you have reached a >>critical mass to have locals everywhere, it will work great and bugs will >>wash out. Otherwise the data might get stale just like any other data. Also >>using the data is essential to find the problems, for example the 212 story >>garage is likely fixed now ;-) >> >>I tend to agree with Steve A. >> >>Cheers Martin >>___ >>talk mailing list >>talk@openstreetmap.org >>https://lists.openstreetmap.org/listinfo/talk > > > > ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
sent from a phone > On 22. Aug 2020, at 10:15, pangoSE wrote: > > Here is yet another example of bad data in our database: fix it ;-) Of course OpenStreetMap contains errors, just like any other source, and probably more, given that most contributors are laymen and have very few experience (few total edits, often just 1). On the other hand, we may be very fast when something changes, very flexible in emergencies (think Haiti), and have interesting niche data that commercial and public data providers don’t care for. It all depends on the local community in the end. If you have reached a critical mass to have locals everywhere, it will work great and bugs will wash out. Otherwise the data might get stale just like any other data. Also using the data is essential to find the problems, for example the 212 story garage is likely fixed now ;-) I tend to agree with Steve A. Cheers Martin ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
On Sat, 22 Aug 2020 at 11:02, pangoSE wrote: > Hi > > Jo skrev: (22 augusti 2020 11:44:49 CEST) > >On Sat, Aug 22, 2020, 11:30 pangoSE wrote: > > > >> Hi > >> > >> Mateusz Konieczny skrev: (22 augusti 2020 > >> 10:51:49 CEST) > >> >(1) Wikipedia may strongly encourage or mandate it in theory, but > >there > >> >are > >> >still edits being made without any citations > >> > >> Yeah I know, but the point is its really hard to create a new article > >in > >> WP without references without it being flagged for deletion. So by > >> "threatening" with deletion they raise the bar for inclusion and > >hence > >> hopefully raise the quality too. We have no system to flag for > >deletion, > >> nor to verify an object. > >> > > > >I find this highly annoying on Wikipedia and it is the reason I don't > >contribute there anymore. > > Interesting. I guess you are not the only because > https://en.m.wikipedia.org/wiki/Deletionpedia exist. > I don't propose we annoy our users the same way, because the downside is > fewer editors. > > I guess its a choice on an continuum between general low quality edits and > many editors and generally higher quality edits and fewer editors. > > Right now OSM accepts almost any crap edit you can throw at it with a big > thank you and we have no really good way of measuring the quality of what > remains after our sometimes spotty QA. > > I would like to help change that by providing better tools for > verification and follow up of things you added/edited in the past. > > I would very much love a telegram bot flagging a new user making an edit > to an object I help curate, but no such tool exist to my knowledge today. > OSMCha tags new users and offers RSS feeds for saved filters. I'm not aware of a way to do this for "ways once touched by [username]" though. > > WDYT? Would such a tool be nice to have? > > Cheers > pangoSE > > ___ > talk mailing list > talk@openstreetmap.org > https://lists.openstreetmap.org/listinfo/talk > ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
Hi Jo skrev: (22 augusti 2020 11:44:49 CEST) >On Sat, Aug 22, 2020, 11:30 pangoSE wrote: > >> Hi >> >> Mateusz Konieczny skrev: (22 augusti 2020 >> 10:51:49 CEST) >> >(1) Wikipedia may strongly encourage or mandate it in theory, but >there >> >are >> >still edits being made without any citations >> >> Yeah I know, but the point is its really hard to create a new article >in >> WP without references without it being flagged for deletion. So by >> "threatening" with deletion they raise the bar for inclusion and >hence >> hopefully raise the quality too. We have no system to flag for >deletion, >> nor to verify an object. >> > >I find this highly annoying on Wikipedia and it is the reason I don't >contribute there anymore. Interesting. I guess you are not the only because https://en.m.wikipedia.org/wiki/Deletionpedia exist. I don't propose we annoy our users the same way, because the downside is fewer editors. I guess its a choice on an continuum between general low quality edits and many editors and generally higher quality edits and fewer editors. Right now OSM accepts almost any crap edit you can throw at it with a big thank you and we have no really good way of measuring the quality of what remains after our sometimes spotty QA. I would like to help change that by providing better tools for verification and follow up of things you added/edited in the past. I would very much love a telegram bot flagging a new user making an edit to an object I help curate, but no such tool exist to my knowledge today. WDYT? Would such a tool be nice to have? Cheers pangoSE ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
Aug 22, 2020, 11:28 by pang...@riseup.net: > Hi > > Mateusz Konieczny skrev: (22 augusti 2020 10:51:49 > CEST) > >(1) Wikipedia may strongly encourage or mandate it in theory, but there > >are > >still edits being made without any citations > > Yeah I know, but the point is its really hard to create a new article in WP > without references without it being flagged for deletion. So by "threatening" > with deletion they raise the bar for inclusion and hence hopefully raise the > quality too. We have no system to flag for deletion, nor to verify an object. > We have. https://wiki.openstreetmap.org/wiki/Notes ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
On Sat, Aug 22, 2020, 11:30 pangoSE wrote: > Hi > > Mateusz Konieczny skrev: (22 augusti 2020 > 10:51:49 CEST) > >(1) Wikipedia may strongly encourage or mandate it in theory, but there > >are > >still edits being made without any citations > > Yeah I know, but the point is its really hard to create a new article in > WP without references without it being flagged for deletion. So by > "threatening" with deletion they raise the bar for inclusion and hence > hopefully raise the quality too. We have no system to flag for deletion, > nor to verify an object. > I find this highly annoying on Wikipedia and it is the reason I don't contribute there anymore. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
On Sat, 22 Aug 2020 at 18:28, pangoSE wrote: > Hi > > Mateusz Konieczny skrev: (22 augusti 2020 > 09:55:10 CEST) > >"It a playground with half-ass quality more than an authoritative and > >verified source of information (like e.g. Wikipedia)" > > > >I am not sure whatever you claim that > >Wikipedia is > >"playground with half-ass quality" or > >"authoritative and verified source of information". > > I meant that a verification system does exist in Wikipedia and they now > require references on all statements to keep up the quality of the articles > which is sane IMO. We have no such system. > I do think OSM is slowly moving to such a system, at least in areas that have an active community and are well mapped. I try to collect Mapillary imagery and one of the reasons is it provides a reference for my change, other's who do the same make the data quality a bit higher because others can verify remotely from the imagery. If I see a suspicious change, I'll post a changeset comment asking if they are sure and if it changed recently especially when I have visited there recently and seems unlikely it would have changed since. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
Hi Mateusz Konieczny skrev: (22 augusti 2020 10:51:49 CEST) >(1) Wikipedia may strongly encourage or mandate it in theory, but there >are >still edits being made without any citations Yeah I know, but the point is its really hard to create a new article in WP without references without it being flagged for deletion. So by "threatening" with deletion they raise the bar for inclusion and hence hopefully raise the quality too. We have no system to flag for deletion, nor to verify an object. > >(2) Wikipedia is explicitly forbidding original research, OSM is >explicitly encouraging it >The best edits are where people map things not mapped anywhere else, >or at least not mapped in any other open data source. Is this relevant to the discussion? I proposed a button that makes it easy for a user to state 1) I attest this is correct (no proof or anything required) NOTE: a malicious user could of course mark all objects in the database as verified, so we probably need a way to handle vandalism, but my implementation is a first draft so feel free to suggest improvements. > >It makes impossible to require citations for everything and requiring >people >to contribute to Mapillary or equivalent would be an unreasonable >burden. I'm not suggesting requiring that, but we should motivate the user to reference a source and make it dead simple to do so. But this is off topic for this thread IMO and we probably need a new system for that too because our current changeset references does not add much value IMO. > >(3) Yes, better verification tools would be likely better. So what do you think about the proposed system? > >(4) Have you seen >https://wiki.openstreetmap.org/wiki/Microgrants/Microgrants_2020/Proposal/Map_Maintenance_with_StreetComplete >(BTW, I really need to finish my resurvey opening hours quest for >StreetComplete). No. Thanks for the link Cheers pangoSE ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
On Sat, 22 Aug 2020, 09:28 pangoSE, wrote: > Hi > > Mateusz Konieczny skrev: (22 augusti 2020 > 09:55:10 CEST) > >"It a playground with half-ass quality more than an authoritative and > >verified source of information (like e.g. Wikipedia)" > > > >I am not sure whatever you claim that > >Wikipedia is > >"playground with half-ass quality" or > >"authoritative and verified source of information". > > I meant that a verification system does exist in Wikipedia and they now > require references on all statements to keep up the quality of the articles > which is sane IMO. We have no such system. > We have a method of referencing sources on the changeset. Although I do think it lacks granularity sometimes. I am not keen on repeated references to "authoritative and verified sources of information" as this starts to walk back the primary principle of on the ground verifiability that OSM relies on. Additional supporting evidence via Mapillary, OpenStreetCam etc. should always be welcome, but in OSM the highest grade of source is "=survey" and I think that is as it should be. OSM has a scope that goes beyond areas with well funded GIS Departments and friendly licenses. > > >OSM would benefit from better verification > >tools and so on but insult-laden post > >filed with misunderstandings will not > >lead towards them. > > Sorry if it came across as harsh, I get your point and will try to > moderate my criticism a little more. > > I love OSM and have contributed a lot over the years and recommend it to > everyone I meet who uses maps. > > I just sent a follow up email with a suggestion for implementing such a > verification system. > > I still believe we have data with bad quality in many places (in Sweden). > I have to fix stuff often when I visit new places apart from all the stuff > we are missing. We are basically trying to keep up with an ever changing > surrounding without a good way to indicate our data quality. > Data going out if date is just the nature of the beast. It's a perennial problem with official maps too. > We can do better, but we need a new system that make it easy for > contributors to verify our precious data (see my previous email). > I think StreetComplete is currently working on a project that helps with this. > Cheers > pangoSE > > ___ > talk mailing list > talk@openstreetmap.org > https://lists.openstreetmap.org/listinfo/talk > ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
(1) Wikipedia may strongly encourage or mandate it in theory, but there are still edits being made without any citations (2) Wikipedia is explicitly forbidding original research, OSM is explicitly encouraging it The best edits are where people map things not mapped anywhere else, or at least not mapped in any other open data source. It makes impossible to require citations for everything and requiring people to contribute to Mapillary or equivalent would be an unreasonable burden. (3) Yes, better verification tools would be likely better. (4) Have you seen https://wiki.openstreetmap.org/wiki/Microgrants/Microgrants_2020/Proposal/Map_Maintenance_with_StreetComplete (BTW, I really need to finish my resurvey opening hours quest for StreetComplete). (5) Anyone with any real knowledge of OSM is already aware that it has mistakes "Hey, here is proof that OSM have mistakes" is not interesting in any way. Aug 22, 2020, 10:26 by pang...@riseup.net: > Hi > > Mateusz Konieczny skrev: (22 augusti 2020 09:55:10 > CEST) > >"It a playground with half-ass quality more than an authoritative and > >verified source of information (like e.g. Wikipedia)" > >> >> > >I am not sure whatever you claim that > >Wikipedia is > >"playground with half-ass quality" or > >"authoritative and verified source of information". > > I meant that a verification system does exist in Wikipedia and they now > require references on all statements to keep up the quality of the articles > which is sane IMO. We have no such system. > > >OSM would benefit from better verification > >tools and so on but insult-laden post > >filed with misunderstandings will not > >lead towards them. > > Sorry if it came across as harsh, I get your point and will try to moderate > my criticism a little more. > > I love OSM and have contributed a lot over the years and recommend it to > everyone I meet who uses maps. > > I just sent a follow up email with a suggestion for implementing such a > verification system. > > I still believe we have data with bad quality in many places (in Sweden). I > have to fix stuff often when I visit new places apart from all the stuff we > are missing. We are basically trying to keep up with an ever changing > surrounding without a good way to indicate our data quality. > > We can do better, but we need a new system that make it easy for contributors > to verify our precious data (see my previous email). > > Cheers > pangoSE > ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
Hi again Mateusz Konieczny via talk skrev: (22 augusti 2020 10:20:51 CEST) >Nobody claims OpenStreetMap data contains no mistakes. > >Are you really expecting that we will be shocked by proof that >some data somewhere is wrong? No. Are you shocked by my constructive criticism and constructive suggestions for improvement of what I perceive as a problem? >Please stop posting about every single inaccurate data in OSM that you >managed to notice. You seem to have the impression that I'm going to spam the list with examples of errors. Where did you get that idea from? I only included Marjins example because it showed 2 things related to my call for verification: 1) people capable of mapping correctly can produce map errors because our subject changes over time. This means that what I map today correctly might be incorrect next month or next year. 2) Martijn had the intention of producing high quality data but we as a community did not help him archive that over time because we have no system e.g. that could alert him if an object he created has not been verified/changed for the last x years. I would love to have a telegram bot notify me when one of my objects has passed a certain threshold of staleness. Cheers pangoSE ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
Hi Mateusz Konieczny skrev: (22 augusti 2020 09:55:10 CEST) >"It a playground with half-ass quality more than an authoritative and >verified source of information (like e.g. Wikipedia)" > >I am not sure whatever you claim that >Wikipedia is >"playground with half-ass quality" or >"authoritative and verified source of information". I meant that a verification system does exist in Wikipedia and they now require references on all statements to keep up the quality of the articles which is sane IMO. We have no such system. >OSM would benefit from better verification >tools and so on but insult-laden post >filed with misunderstandings will not >lead towards them. Sorry if it came across as harsh, I get your point and will try to moderate my criticism a little more. I love OSM and have contributed a lot over the years and recommend it to everyone I meet who uses maps. I just sent a follow up email with a suggestion for implementing such a verification system. I still believe we have data with bad quality in many places (in Sweden). I have to fix stuff often when I visit new places apart from all the stuff we are missing. We are basically trying to keep up with an ever changing surrounding without a good way to indicate our data quality. We can do better, but we need a new system that make it easy for contributors to verify our precious data (see my previous email). Cheers pangoSE ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
Nobody claims OpenStreetMap data contains no mistakes. Are you really expecting that we will be shocked by proof that some data somewhere is wrong? I would be able to post one mail per minute with examples of serious mistakes, forever - even after my death, as it would be fairly easy to automate. If something is wrong then fix it or create note or switch to other data source (that WILL contain wrong/inaccurate data - or will be very expensive and cover some limited area). Or ignore it. Please stop posting about every single inaccurate data in OSM that you managed to notice. https://wiki.openstreetmap.org/wiki/Notes are designed as tool allowing to describe incorrect data that someone is unable or unwilling to fix (and yes, we have thousands of reports of mistakes) Aug 22, 2020, 10:12 by pang...@riseup.net: > Here is yet another example of bad data in our database: > > Originalmeddelande > Från: Martijn van Exel > Skickat: 22 augusti 2020 00:33:24 CEST > Till: talk@openstreetmap.org > Ämne: Re: [OSM-talk] Use of OSM data without attribution > > Curious anecdote: some AllTrails user apparently looked up a phone > number for OSM US and called up Maggie. Turns out the complaint was > about a trail that I originally mapped *blush*. In my defense, that was > 9 years ago, I haven't been to that part of town much since I moved, and > nobody else updated the trail, which has since disappeared. > > Here is the changeset in case you're interested: > https://www.openstreetmap.org/changeset/89419938 > > Martijn > -- > > Maybe we should have some kind of system flagging objects that has not been > edited for x number of years and rate all objects in the database according > to this? > > This would mean that a data consumer can decide based on the score if they > want to include the information or not. > > E.g. a high quality map should perhaps not contain objects with a revision > older than 3 years (and no references or sources) > > Or even better: we could implement a verification system with a log that can > be queried easily. > > IMPLEMENTATION SUGGESTION: > > GET Openstreetmap.org/api/verifications/ > Lists latest added verifications (outputs 10 entries, can be used to > get more, can be used to output up to 300 entries) > > GET Openstreetmap.org/api/verifications/1234 > Outputs verifications for osmid 1234 with the newest first (outputs 10 > entries, can be used to get more, can be used to output up to > 300 entries) > > POST Openstreetmap.org/api/verifications/1234 > Add a new verification for osmid 1234 > > On openstreetmap.org we have a new button for every object "Verify this > object exists and is correct" which stores the date and userid in the > database. > > In JOSM we could add the possibility to download verification data for all > selected objects or from a new option in the download dialog. > > The latest verification date and count of verifications could be made > available in a separate dump. > > If we had such a system I believe the map data quality could increase > considerably by making it dead simple to hide hide old unverified data from > e.g. openstreetmap.org. A high-quality map we can be proud of could also give > an impetus to local mappers to revisit trails and verify them. > > WDYT? > > Cheers > pangoSE > > > pangoSE skrev: (22 augusti 2020 09:32:09 CEST) > >> Hi >> >> 80hnhtv4agou--- via talk skrev: (22 augusti 2020 >> 03:06:37 CEST) >> >> >>> >>> Also there is no wiki on unverified edits. >>> >>> >> >> In OSM we don't yet have an established system for verification or accurate >> machine readable references for the data to my knowledge. >> >> This means the whole database is basically just a mess of biased data that >> one of our millions of editors thought should be included. Most objects have >> very few revisions and we have no idea about the overall quality or >> correctness. It a playground with half-ass quality more than an >> authoritative and verified source of information (like e.g. Wikipedia). >> Building upon it can lead to strange things. E.g. >> >> https://www.nyteknik.se/popularteknik/mystisk-jatteskrapa-dok-upp-i-flygsimulator-6999771>> >> (building:levels=212 was entered erroneously and committed to the database >> without any kind of QA follow-up. If someone knows the osmid I would like to >> know how long this error was present in OSM) >> >> We should really fix this and start a verification effort after implementing >> a sane verification model. >> talk mailing list >> talk@openstreetmap.org >> https://lists.openstreetmap.org/listinfo/talk >> ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
Here is yet another example of bad data in our database: Originalmeddelande Från: Martijn van Exel Skickat: 22 augusti 2020 00:33:24 CEST Till: talk@openstreetmap.org Ämne: Re: [OSM-talk] Use of OSM data without attribution Curious anecdote: some AllTrails user apparently looked up a phone number for OSM US and called up Maggie. Turns out the complaint was about a trail that I originally mapped *blush*. In my defense, that was 9 years ago, I haven't been to that part of town much since I moved, and nobody else updated the trail, which has since disappeared. Here is the changeset in case you're interested: https://www.openstreetmap.org/changeset/89419938 Martijn -- Maybe we should have some kind of system flagging objects that has not been edited for x number of years and rate all objects in the database according to this? This would mean that a data consumer can decide based on the score if they want to include the information or not. E.g. a high quality map should perhaps not contain objects with a revision older than 3 years (and no references or sources) Or even better: we could implement a verification system with a log that can be queried easily. IMPLEMENTATION SUGGESTION: GET Openstreetmap.org/api/verifications/ Lists latest added verifications (outputs 10 entries, can be used to get more, can be used to output up to 300 entries) GET Openstreetmap.org/api/verifications/1234 Outputs verifications for osmid 1234 with the newest first (outputs 10 entries, can be used to get more, can be used to output up to 300 entries) POST Openstreetmap.org/api/verifications/1234 Add a new verification for osmid 1234 On openstreetmap.org we have a new button for every object "Verify this object exists and is correct" which stores the date and userid in the database. In JOSM we could add the possibility to download verification data for all selected objects or from a new option in the download dialog. The latest verification date and count of verifications could be made available in a separate dump. If we had such a system I believe the map data quality could increase considerably by making it dead simple to hide hide old unverified data from e.g. openstreetmap.org. A high-quality map we can be proud of could also give an impetus to local mappers to revisit trails and verify them. WDYT? Cheers pangoSE pangoSE skrev: (22 augusti 2020 09:32:09 CEST) >Hi > >80hnhtv4agou--- via talk skrev: (22 augusti >2020 03:06:37 CEST) > >> >>Also there is no wiki on unverified edits. >> > >In OSM we don't yet have an established system for verification or >accurate machine readable references for the data to my knowledge. > >This means the whole database is basically just a mess of biased data >that one of our millions of editors thought should be included. Most >objects have very few revisions and we have no idea about the overall >quality or correctness. It a playground with half-ass quality more than >an authoritative and verified source of information (like e.g. >Wikipedia). Building upon it can lead to strange things. E.g. >https://www.nyteknik.se/popularteknik/mystisk-jatteskrapa-dok-upp-i-flygsimulator-6999771 >(building:levels=212 was entered erroneously and committed to the >database without any kind of QA follow-up. If someone knows the osmid I >would like to know how long this error was present in OSM) > >We should really fix this and start a verification effort after >implementing a sane verification model. > >___ >talk mailing list >talk@openstreetmap.org >https://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Call for verification (Was: Re: VANDALISM !)
"It a playground with half-ass quality more than an authoritative and verified source of information (like e.g. Wikipedia)" I am not sure whatever you claim that Wikipedia is "playground with half-ass quality" or "authoritative and verified source of information". Though any of this claims would demonstrate that you are wrong and uninformed. Like with your "deprecate name tag" there are so many wrong things here. OSM would benefit from better verification tools and so on but insult-laden post filed with misunderstandings will not lead towards them. 22 Aug 2020, 09:32 by pang...@riseup.net: > Hi > > 80hnhtv4agou--- via talk skrev: (22 augusti 2020 > 03:06:37 CEST) > > > > >Also there is no wiki on unverified edits. > > > > In OSM we don't yet have an established system for verification or accurate > machine readable references for the data to my knowledge. > > This means the whole database is basically just a mess of biased data that > one of our millions of editors thought should be included. Most objects have > very few revisions and we have no idea about the overall quality or > correctness. It a playground with half-ass quality more than an authoritative > and verified source of information (like e.g. Wikipedia). Building upon it > can lead to strange things. E.g. > https://www.nyteknik.se/popularteknik/mystisk-jatteskrapa-dok-upp-i-flygsimulator-6999771 > (building:levels=212 was entered erroneously and committed to the database > without any kind of QA follow-up. If someone knows the osmid I would like to > know how long this error was present in OSM) > > We should really fix this and start a verification effort after implementing > a sane verification model. > > ___ > talk mailing list > talk@openstreetmap.org > https://lists.openstreetmap.org/listinfo/talk > ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
