Re: [OSM-talk] OSM TrustPoints
Fixing subway stations in Rome wouldn't be prohibited to newcomers if they limit themselves for that day to Rome, encouraging them to focus on an area rather than jumping around the globe changing things (True value of OSM is _local_ knowledge!). They could be first only allowed to map within some kilometers around their home, but that might be too restrictive and demotivating. As for number of daily affected nodes the number shouldn't be too low, but low enough to prevent moving or deleting whole cities (whether accidental or malicious) or importing some data without consulting the community (about legal and technical issues) and having some help from experienced members. I personally often travel along some long route (100-200 km in my case), then I start adding POIs and other stuff I saw along the route once I get home. And there are people travelling much longer routes (long road routes across US/australia or even flights - one user in one day can easily add some details on road to some european airport and then another details on road from one airport in south america to nearest city centre for example) I don't think we should put any hard limits on the bbox. Martin ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] OSM TrustPoints
--- On Sun, 12/7/09, MP singular...@gmail.com wrote: (long road routes across US/australia or even flights - one user in Even the current limits make it time consuming downloading multiple areas to cover drives in rural areas I've taken in the last month alone. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] OSM TrustPoints
Hi! First, some clarifications: The possible measures and use limitations were only provided as an example what can be observed and measured and what could be limited before earning enough trust in a wat that would be much more effective and less annoying than captchas. It was not a specific proposal, but listing possibilities to spark the debate. Probably not all measures and all limitations are needed and proposals for improvements are welcome. We don't need to implement every feature from every example site, but we have the chance to review those tried examples and see how their system would work for us, and pick only the best features. Now some specific issues: Fixing subway stations in Rome wouldn't be prohibited to newcomers if they limit themselves for that day to Rome, encouraging them to focus on an area rather than jumping around the globe changing things (True value of OSM is _local_ knowledge!). They could be first only allowed to map within some kilometers around their home, but that might be too restrictive and demotivating. As for number of daily affected nodes the number shouldn't be too low, but low enough to prevent moving or deleting whole cities (whether accidental or malicious) or importing some data without consulting the community (about legal and technical issues) and having some help from experienced members. Let's admit that adding Arabic names to all countries isn't something a newcomer would be doing in his first months. Due to seasonal mapping (eg. only during summer) we could prolong the time after which points start to decay, or make it significantly slower or drop decaying altogether. Although after some consecutive years of no activity it might be safe to pronounce someone dead (in terms of further OSM contributions, at least for statistical reasons). For simplicity's sake and easier understanding points could be summarized in at most 5 discrete levels, and additional rights given to users according to those levels. Levels could be stated in place where that is needed, eg. mappers near me, so i know who i can ask some local mapping question, probably also shown in profiles (publicly or jut to account owners). Better support for mentoring could also be very good thing. Involving other systems wouldn't mean a user has to use all of them, but perhaps 2 of them constructively and not doing extremely annoying things in others to be considered more trusted. This would also allow people to shift focus (eg editing wiki in winter) while maintaining their points or even gaining new ones. Here even map editing api, track uploads and diaries could be considered as separate systems. Randomly checked changesets also seem nice. But it might be hard to check the first mapper in a blan area Spam in diaries is existing fact. Volume is orders of magnitude lower than in email but even that could be eliminated or filtering can be crowdsourced without annoying admins (while still annoying those that encounter it before it is tagged by enough people). Overall, there should be very low threshold before user is given almost the same permissions as now (curently no limits,so even a new sock puppet account could be used to import some commercial maps in hopes of legalizing them this way for their use), starting with a limited set of rights, and very trusted ones would get additional powers. There were also some comments in wiki [1] and additional example site [2] was given there, so please let us not disregard those in further discussion. Also we should give people at SOTM some time to digest their mails and give some opinions. They are hopefully using their time there for better things than reading email :) Stefan [1] http://wiki.openstreetmap.org/wiki/User_talk:Stefanb/TrustPoints [2] http://stackoverflow.com/faq#reputation ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] OSM TrustPoints
What about not adding restrictions to the rights, but implement some warning system? We can use some trust points that are awarded for good mapping (objects that last on the map) Then warning points in each upload for possibly bad things (deleted nodes, removed tags, accidentally moving many nodes by same offset, etc ) Then publish some feed or list of such edits above certain threshold - so they can be spotted and either marked as false positives (by trusted users) which will cause them to lose warning points (the more trusted user mark it, the more point lost), ultimately moving them off the list, or dealt with (reverted, etc ... - which will remove trust points from the user) Newcomer that deletes even single node can be spotted that way - and even older trusted people could be spotted if they start some massive vandalism or big error. As for importing google maps issue, we can mark some places that are known to have no coverage from allowed sources (Yahoo, french cadastre, ) and every node there can trigger some warning points Point is, that suspicious edits (caused by lack of skill, human error or vandalism...) gets listed for people to review, while not limiting anybody. Martin ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] OSM TrustPoints
--- On Sat, 11/7/09, Stefan Baebler stefan.baeb...@gmail.com wrote: Fixing subway stations in Rome wouldn't be prohibited to newcomers if they limit themselves for that day to Rome, encouraging them to focus on an area rather than jumping around the globe changing things (True Hopefully the area isn't made small without considering sparse populations at the same time, there are areas on the global that are vast but the number of people per 100s of square km is almost nil. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] OSM TrustPoints
Hello, Goal Prevent creation of new sock puppet accounts for potential acts of vandalism on larger scale and spam. Gradually give trust to users, and give them additional privileges. It should not be in the way when new users want to contribute normally. It should not encourage competition so that itself doesn't become an abuse target. Well, in the way described, it will conflict with a couple of legitimate use cases. Is there a real vandalism problem on the map? Although I've been working for almost a year with a lot of the map data, I have not seen any piece of intentional vandalism. * allow larger daily bbox for changes A couple of weeks ago, I discovered that the subway stations in Roma are incomplete, so I added stations to my best knowledge. Or, an even more opposed case: A couple of day ago somebody added arabic names to all the coutries. I would consider this as a legitimate use, but it invokes almost the entire planet. * allow more daily edits (number of affected nodes, ways...) Even something simple like the bus route I've added yesterday might easily touch more than a hundred elements. On the other hand, you could damage significantly the map with less than a hundred destructive edits. * Regular editing activity I personally do a lot of mapping in my holidays, so a pattern like massive activity from time to time will appear from the system's point of view. On the other hand, it is easy to tune the amount of activity of a malicious bot to any pattern expected by the server as regular. * Track uploads There might be a lot of use cases, e.g. naming things, adding POIs, bus routes and so on that require no GPX data at all. * Regular activity in other systems? (mailing lists, forums, wiki, svn repository, diary/blog, trac...) Perhaps totally different systems shouln't be mixed - one who can program or is very vocal doesn't necessarily yet know how to map well and shouldn't be trusted with enormous imports and vice versa) This makes contributing even more complicated. Why should I be forced to use these frills? There might be users who can't contribute something useful in one or more of the above systems (What should a non-programmer contribute to the SVN Repository? Should every user write wiki pages that nobody wants to read, just to display honest acitivity?) or just aren't interested in certain tools (not everybody wants to blog). * Rated by other users (manualy: reverted changesets, reported spam in diary, getting comments to a diary entry while not being flagged as spam...) ? Oh, sounds like the eBay reputation system. Do we really want to have an intervention by lawyers which comments are admissible and which aren't? An implementation of the whole idea would not only take a lot of ressources of the implementers, but also affect significanty normal users. The way mappers and other users use the system in a legitimate way is so widespread that any reputation system would end up bullying a smaller or bigger fraction of the users. Remember that the API has intentionally been kept small to make usage simple. On the other hand, there is no strong need to prevent vandalism because there is only sparse vandalism. So I would suggest to keep the whole concept away from the map. Cheers, Roland ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] OSM TrustPoints
As osm grows the chances that somone will try to damage the map grows. Maybe a new user should have their edits checked when they first join and then build trust that way. Make it a random check but put the priority on the new users. If somone does a large edit then it can be flagged to be checked etc. A photogallery system called fotopic.net uses this kind of system to stop new users uploading pornographic images. It should only take a few trusted users to check on things. Jack Stringer On Jul 10, 2009 10:10 AM, Roland Olbricht roland.olbri...@gmx.de wrote: Hello, Goal Prevent creation of new sock puppet accounts for potential acts of vandalism on larger... Well, in the way described, it will conflict with a couple of legitimate use cases. Is there a real vandalism problem on the map? Although I've been working for almost a year with a lot of the map data, I have not seen any piece of intentional vandalism. * allow larger daily bbox for changes A couple of weeks ago, I discovered that the subway stations in Roma are incomplete, so I added stations to my best knowledge. Or, an even more opposed case: A couple of day ago somebody added arabic names to all the coutries. I would consider this as a legitimate use, but it invokes almost the entire planet. * allow more daily edits (number of affected nodes, ways...) Even something simple like the bus route I've added yesterday might easily touch more than a hundred elements. On the other hand, you could damage significantly the map with less than a hundred destructive edits. * Regular editing activity I personally do a lot of mapping in my holidays, so a pattern like massive activity from time to time will appear from the system's point of view. On the other hand, it is easy to tune the amount of activity of a malicious bot to any pattern expected by the server as regular. * Track uploads There might be a lot of use cases, e.g. naming things, adding POIs, bus routes and so on that require no GPX data at all. * Regular activity in other systems? (mailing lists, forums, wiki, svn repository, diary/bl... This makes contributing even more complicated. Why should I be forced to use these frills? There might be users who can't contribute something useful in one or more of the above systems (What should a non-programmer contribute to the SVN Repository? Should every user write wiki pages that nobody wants to read, just to display honest acitivity?) or just aren't interested in certain tools (not everybody wants to blog). * Rated by other users (manualy: reverted changesets, reported spam in diary, getting comme... Oh, sounds like the eBay reputation system. Do we really want to have an intervention by lawyers which comments are admissible and which aren't? An implementation of the whole idea would not only take a lot of ressources of the implementers, but also affect significanty normal users. The way mappers and other users use the system in a legitimate way is so widespread that any reputation system would end up bullying a smaller or bigger fraction of the users. Remember that the API has intentionally been kept small to make usage simple. On the other hand, there is no strong need to prevent vandalism because there is only sparse vandalism. So I would suggest to keep the whole concept away from the map. Cheers, Roland ___ talk mailing list talk@openstreetmap.org http://lis... ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] OSM TrustPoints
Agreed, Trust points not needed. The system in place now, which prevents vandalism, is of users all the time always have their eyes on the map. Unlike other websites, which bots might manage. .. we have actual people always watching it. The instant that new edits are made, everyone is aware of it. So as soon as a error (weather it's a bot, or a human-error) it's found, and it gets fixed. Ie. in Montreal someone accidentally moved a bunch of stuff (a newbie) it was reported on the talk-ca list, they messaged that user so to see if it was a bot or a person. And it was a person, who made an error, then work begins to learn how to un-do the edit. (Personally i don't know how, but it's easy to find out) ... just ask :-) and when i added a bunch of river names and accently had them tagged as towns, as soon as it got rendered someone told me about it. (knowing its a test area... but it still got fixed) I also notest there hasn't been any change on the 'history tab' where i see all edits that happened. So it's fine the way it is, as even if it's bulk_uploads, i like to know whats going on. Right now i see that new language name tags were added. Awesome! (it's easy enough for me to scroll down a little to find local peoples edits). But having that option to filter down to the area would help, but AFTER seeing all the edits. .. as the big edits would have local impact also. So that safegueard is what croudsourcing is all about. If for my example, if that user didn't respond, the next step is to ask on the talk list about it report it on the user diaries. Then almost instantly (within a short period of time) all efforts are made to contact that user, and those changesets are recorded. and can be reverted if it's found to be a bot that was malitious. So far (in the last year) it's only been xy bot that likes to make edits, but that's been fixed, as we have openstreetbugs, ... and that darn smoothness tag, that everyone likes to edit on the wiki :) ... but people still use that tag for some strange reasons. ... it probably becaue they want people to jump in and help further define the tags. Maybe some admins have seen lots? But get caught faster than we can see? So that said, although a nice idea. .. just 'cause other sites like to give thumbs up and use different forms of veri-sign and stuff. .. doesn't mean openstreetmap.org has to. I see ALL edits, no matter how big or how small as just as an important contribution. It's all done to make the map better. .. all edits get a changeset number, and open for everyone to see. So in short, IMO croudsourcing with thousands of eyes on the map at all times, is the best method of preventing spam, and vandels. Therefore having trustpoints isn't needed. Cheers, Sam Vekemans Across Canada Trails Twitter: @Acrosscanada Facebook: http://www.facebook.com/sam.vekemans ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] OSM TrustPoints
--- On Fri, 10/7/09, Jack Stringer jack.ix...@googlemail.com wrote: As osm grows the chances that somone will try to damage the map grows. Maybe a new user should have their edits checked when they first join and then build trust that way. Make it a random check but put the priority on the new users. If somone does a large edit then it can be flagged to be checked etc. This sounds more like a slashdot type comment rating/anti-spam system, and why limit it to new users, edits could be randomly rated by other users for quality control in general, not just in the prevention of abuse. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] OSM TrustPoints
On Fri, Jul 10, 2009 at 12:01 PM, Sam Vekemans acrosscanadatra...@gmail.com wrote: Ie. in Montreal someone accidentally moved a bunch of stuff (a newbie) it was reported on the talk-ca list, they messaged that user so to see if it was a bot or a person. And it was a person, who made an error, then work begins to learn how to un-do the edit. (Personally i don't know how, but it's easy to find out) ... just ask :-) and when i added a bunch of river names and accently had them tagged as towns, as soon as it got rendered someone told me about it. (knowing its a test area... but it still got fixed) Unfortunately not all of the places in the world are blessed with a large community and not all of them follow talk-* lists (not even the local ones). In fact I would say that (from my experience) most of them do not. And I would expect that most newbies that want to contribute to OSM start by clicking on the edit tab, not by subscribing to any particular mailing list. So errors (deliberate or not) can be left unnoticed for quite some time in these places. I'm not saying we need a complex guarding system against changing stuff, but some kind of a simple mechanism for protecting against common mistakes (e.g. moving whole ways etc) wouldn't be bad. And if that's somehow related to experience points, so what? You cannot drive a bus with an ordinary driver's license (at least not in my country) ;) Igor ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] OSM TrustPoints
I don't realistically see an automated reputation system working given the community we have but how about a mentoring type approach where experienced mappers adopt one or more newbies. When someone signs-up for the project (or maybe when they make their first edit) they could be given a list of available mentors preferably in their local area but failing that you only really need to speak the same language. The mentor could then informally review their first few edits, answer questions, give them tips and point them in the direction of useful resources, mailing lists, etc... Might be something useful for people to do who have completed mapping their own areas. Kevin On Fri, Jul 10, 2009 at 11:03 AM, John Smith delta_foxt...@yahoo.comwrote: --- On Fri, 10/7/09, Jack Stringer jack.ix...@googlemail.com wrote: As osm grows the chances that somone will try to damage the map grows. Maybe a new user should have their edits checked when they first join and then build trust that way. Make it a random check but put the priority on the new users. If somone does a large edit then it can be flagged to be checked etc. This sounds more like a slashdot type comment rating/anti-spam system, and why limit it to new users, edits could be randomly rated by other users for quality control in general, not just in the prevention of abuse. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] OSM TrustPoints
--- On Fri, 10/7/09, Kevin Peat ke...@kevinpeat.com wrote: I don't realistically see an automated reputation system working given the community we have but how about a mentoring type approach where experienced mappers adopt one or more newbies. Well it would have saved a lot of newbie questions hitting the talk-au list from me if such a programme had been in place before hand :) ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] OSM TrustPoints
2009/7/10 Peter Dörrie peter.doer...@googlemail.com Well, as the OSM community grows exponentially at the moment we will run out of expereinced mappers pretty soon ... There might be exponential growth in accounts opened by lurkers, bots, and spammers but I don't think the growth in people actually editing the map is quite so huge that it is unmanagable. By the way: what is an experienced mapper? Can one self-appoint to this position? ;-) Why not...I dont think you would get too many people wanting to be mentors that weren't experienced mappers. Anyway, you could always let people choose another mentor if things didn't work out. Kevin ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] OSM TrustPoints
When someone signs-up for the project (or maybe when they make their first
edit) they could be given a list of available mentors preferably in their
local area but failing that you
Cool, perhaps there could be a way that on the list of the 10 local mappers
in the area, to sort those who have made alot of edits, or declare
themselves as the 'area's main mapper', and post that on the wiki page,
could be made more clear on sign-up?
Similar to on the wiki there is the {{welcome}} notice, perhaps there could
be a way for users to receive notice via. automagical message, that new
mappers are in the area? So then there is a chance for the local area
mappers to welcome the newbie. (expanding on the 'add as friend' which IMO
hasn't been expanded much. Even incorporating facebook/OpenID, having a
spot to fill that in. I know the wiki lets users put what they want,
however, its a way for those users who dont plan on editing the wiki, but
are building the map. Why would they have to create a new user just for it?
Meet-up.com uses a similar approach where only those people in the area and
with the same common interests are filtered down to just a few. Perhaps a
set of interests can be checked of, as to the newbies level of interest,
programming / rendering / mapping / casual mapping / promotion etc. (so you
can contact those who have the specific expertise your looking for).
only really need to speak the same language. The mentor could then
informally review their first few edits, answer questions, give them tips
and point them in the direction of useful resources, mailing lists, etc...
When i discover new mappers in an area, im contacting them asking them to
post on the wiki that they are working on the page, and to see what there
level of interest is. As well as find out if there is a chance to meet in
person sometime soon.
But some of the above can also be provided in the initial sign-up
confirmation welcome message.
And perhaps it can be further listed on the page when the user is editing
their profile. and maybe a link to the talk-* list could be made available
just after the list of the 10 local area mappers. As a question Have more
questions? Why not join the (region) list to keep informed of mapping
parties region specific discussions?
Might be something useful for people to do who have completed mapping their
own areas.
Kevin
Cheers,
Brainstorming ideas,
Sam
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] OSM TrustPoints
I agree that the mentor process makes more sense than trying to implement any kind of raw points system. Project DP (http://www.pgdp.net/c/) has a good framework in place that allows people to get their feet wet without causing major damage to the project. User access is divided into several roles. Once a user reaches a threshold contribution level at one role, they are granted access to the next higher role. Throughout the progression, users with higher roles are encouraged/required to QC the work done by the lower roles. I'm not sure how many tags would be needed to implement this, but the way the editors handle the tiger:reviewed=no tag would probably be a good framework to build off of. Chris Hunter chunter...@gmail.com From: Kevin Peat ke...@kevinpeat.com To: John Smith delta_foxt...@yahoo.com Date: Fri, 10 Jul 2009 13:00:26 +0100 Subject: Re: [OSM-talk] OSM TrustPoints I don't realistically see an automated reputation system working given the community we have but how about a mentoring type approach where experienced mappers adopt one or more newbies. When someone signs-up for the project (or maybe when they make their first edit) they could be given a list of available mentors preferably in their local area but failing that you only really need to speak the same language. The mentor could then informally review their first few edits, answer questions, give them tips and point them in the direction of useful resources, mailing lists, etc... Might be something useful for people to do who have completed mapping their own areas. Kevin ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
