Re: [OSM-talk] Revert requests in general
I wanted to revive the discussion inspired by NearMap, regarding making an easier way to on-ramp new users, since it seemed to have died. Nick wrote: So let's bring this back to people who want to create tools to make it easier for everyone to participate in OSM. How can we get past the problems and make it easy for people to map? Auth and new mappers workshop ++ Nick +1 here. So - how do we make that workshop happen? whats the usual way of going about it? I'm a n00b, so educate me on what to do to help move this to the next (or first) step. Thanks, Ant ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On Fri, Aug 6, 2010 at 8:48 AM, Kai Krueger wrote: > You can associate an OpenID to an existing account. You can also switch your > associated OpenID at any time (provided you are logged in) just like you can > change your password. (The OpenID is never revealed to anyone other than the > account holder). But as Tom sais, for the moment you can only associate a > single OpenID with your account. If there is demand for linking multiple > OpenIDs to a single account, it should be reasonably easy to change that in > the future though too. Also, if you choose, you can always continue to use a > standard password instead or in addition to the OpenID. I'd like to see multiple OpenIDs linked to the one osm account, since I'm one (of many) people who has many OpenIDs. But I realise that it will complicate a few things so I wouldn't mind that being postponed. Cheers, Andy ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
A couple of days ago I wanted to register with some site... ah, yes, it was Hutch. I was okay about setting up a username and password, but they offered me the ability to authenticate via Facebook - three clicks and I was done. Very, very easy and didn't trigger my personal (admittedly quite high) privacy issue paranoia, since I got to make decisions about the areas of my profile to which I was granting Hutch access. Given that what the OSM wants (if I've understood this correctly) is: 1. That a new user walks through some process that shows them the terms and conditions so that they can make an informed decision to accept, 2. That the OSM has a clear an unambiguous way to identify and contact that user in the event of vandalism (or for other important, non-spam needs), ...then might FB and/or Twitter authentication be an option (and note that I say "option", not The One True Way to register)? Just a thought :) b On 7 August 2010 08:46, Nick Black wrote: > So let's bring this back to people who want to create tools to make it > easier for everyone to participate in OSM. How can we get past the problems > and make it easy for people to map? > > Auth and new mappers workshop ++ > > Nick > > On 7 Aug 2010, at 01:03, Eugene Alvin Villar wrote: > > On Sat, Aug 7, 2010 at 7:43 AM, John Smith < > deltafoxtrot...@gmail.com> wrote: > >> On 7 August 2010 08:56, Frederik Ramm < >> frede...@remote.org> wrote: >> > Hi, >> > >> > John Smith wrote: >> >> >> >> Even wikipedia doesn't take that attitude, they're currently being >> >> threatened by the FBI over a SVG image. >> > >> > Nothing to do with copyright, and thus completely irrelevant in this >> > discussion. >> >> Wikimedia is claiming fair use, sounds like a copyright argument to me. >> > > Nope. FBI's problem with Wikipedia has nothing to do with copyright. The > work in question, the FBI seal, is the work of the U.S. Federal Government > which would make it public domain and thus there is no copyright in the > first place. FBI's problem is that people might make fake FBI badges and > stuff like that because Wikipedia provides a high-quality SVG image of the > seal. > > ___ > > talk mailing list > talk@openstreetmap.org > http://lists.openstreetmap.org/listinfo/talk > > > ___ > talk mailing list > talk@openstreetmap.org > http://lists.openstreetmap.org/listinfo/talk > > -- Ben Last Development Manager (HyperWeb) NearMap Pty Ltd ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 7 August 2010 07:57, Eugene Alvin Villar wrote: > From how I understand it, NearMap's editor is so simple (can't do complex > edits) and is often used for one-off edits that I think how NearMap does it > is pretty spot on in trying to get the most number of contributions. What I > guess NearMap should do more is to explain about OSM more clearly if a > NearMap user wants to get much more involved with contributing map data such > as saying that the user should go straight to OSM and create an account. > That's a good point; we already do send people to OSM to fix edits; that's what the Edit button on the current site does, it's our standard response by email and in our forums. When we roll out the simpler edit, OSM (essentially, potlatch on OpenStreetMap.org) will be the "advanced" edit option. Cheers b -- Ben Last Development Manager (HyperWeb) NearMap Pty Ltd ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
So let's bring this back to people who want to create tools to make it easier for everyone to participate in OSM. How can we get past the problems and make it easy for people to map? Auth and new mappers workshop ++ Nick On 7 Aug 2010, at 01:03, Eugene Alvin Villar wrote: > On Sat, Aug 7, 2010 at 7:43 AM, John Smith wrote: > On 7 August 2010 08:56, Frederik Ramm wrote: > > Hi, > > > > John Smith wrote: > >> > >> Even wikipedia doesn't take that attitude, they're currently being > >> threatened by the FBI over a SVG image. > > > > Nothing to do with copyright, and thus completely irrelevant in this > > discussion. > > Wikimedia is claiming fair use, sounds like a copyright argument to me. > > Nope. FBI's problem with Wikipedia has nothing to do with copyright. The work > in question, the FBI seal, is the work of the U.S. Federal Government which > would make it public domain and thus there is no copyright in the first > place. FBI's problem is that people might make fake FBI badges and stuff like > that because Wikipedia provides a high-quality SVG image of the seal. > ___ > 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] Revert requests in general
On Sat, Aug 7, 2010 at 7:43 AM, John Smith wrote: > On 7 August 2010 08:56, Frederik Ramm wrote: > > Hi, > > > > John Smith wrote: > >> > >> Even wikipedia doesn't take that attitude, they're currently being > >> threatened by the FBI over a SVG image. > > > > Nothing to do with copyright, and thus completely irrelevant in this > > discussion. > > Wikimedia is claiming fair use, sounds like a copyright argument to me. > Nope. FBI's problem with Wikipedia has nothing to do with copyright. The work in question, the FBI seal, is the work of the U.S. Federal Government which would make it public domain and thus there is no copyright in the first place. FBI's problem is that people might make fake FBI badges and stuff like that because Wikipedia provides a high-quality SVG image of the seal. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
>From how I understand it, NearMap's editor is so simple (can't do complex edits) and is often used for one-off edits that I think how NearMap does it is pretty spot on in trying to get the most number of contributions. What I guess NearMap should do more is to explain about OSM more clearly if a NearMap user wants to get much more involved with contributing map data such as saying that the user should go straight to OSM and create an account. On Fri, Aug 6, 2010 at 8:27 PM, Ben Last wrote: > On 6 August 2010 20:04, Nick Black wrote: > >> How would the people voicing opinions in this thread feel about a hack >> / planning day when editor developers, OSM-F and the OSM server admin >> team can get together to talk through each side's concerns and come up >> with a plan that is good for everyone - OSM-F, OSM admin team, editor >> developers, and most importantly for mappers? >> > From the NearMap point of view, I'd welcome that :) > Cheers > b > > -- > Ben Last > Development Manager (HyperWeb) > NearMap Pty Ltd > > > ___ > talk mailing list > talk@openstreetmap.org > http://lists.openstreetmap.org/listinfo/talk > > -- http://vaes9.codedgraphic.com ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 7 August 2010 08:27, Ian Dees wrote: > That's different because the FBI is quite obviously wrong. There is a law > that says they are wrong. The FBI are asserting they're right, and wikimedia are asserting they're right, it's up to a court to be the adjudicator. > Almost any complaint that someone might bring against OSM would be a brand > spanking new, precedent-less, law-less (there aren't clear laws about data > rights) suit. We don't want to be groundbreaking when it comes to data > copyright court cases, I don't think. I'm not saying we should, but it seems to me Steve's intentions to concede at the first sign of trouble would make OSM(F) appear weak and so anyone could abuse OSM's license because we don't have the ability to defend or protect ourselves. This is the sort of thing that might lead to death by 1000 cuts type situation. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 7 August 2010 08:56, Frederik Ramm wrote: > Hi, > > John Smith wrote: >> >> Even wikipedia doesn't take that attitude, they're currently being >> threatened by the FBI over a SVG image. > > Nothing to do with copyright, and thus completely irrelevant in this > discussion. Wikimedia is claiming fair use, sounds like a copyright argument to me. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
Hi, John Smith wrote: Even wikipedia doesn't take that attitude, they're currently being threatened by the FBI over a SVG image. Nothing to do with copyright, and thus completely irrelevant in this discussion. Bye Frederik -- Frederik Ramm ## eMail frede...@remote.org ## N49°00'09" E008°23'33" ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On Fri, Aug 6, 2010 at 5:21 PM, John Smith wrote: > On 7 August 2010 03:14, SteveC wrote: > > If they have several orders of magnitude more money then probably the > cost/benefit tradeoff would suggest throwing out the data is the better > option. > > Even wikipedia doesn't take that attitude, they're currently being > threatened by the FBI over a SVG image. > > That's different because the FBI is quite obviously wrong. There is a law that says they are wrong. Almost any complaint that someone might bring against OSM would be a brand spanking new, precedent-less, law-less (there aren't clear laws about data rights) suit. We don't want to be groundbreaking when it comes to data copyright court cases, I don't think. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On Aug 6, 2010, at 4:21 PM, John Smith wrote: > On 7 August 2010 03:14, SteveC wrote: >> If they have several orders of magnitude more money then probably the >> cost/benefit tradeoff would suggest throwing out the data is the better >> option. > > Even wikipedia doesn't take that attitude, they're currently being > threatened by the FBI over a SVG image. How much money does the wikimedia foundation have? How much money does OSMF have? At a guess, they have approximately 1,000 times our resources. Therefore, they have more hope in a fight like that. I'm not saying that's how it should be forever, or it's a wonderful situation, I'm just pointing out the realities of where we are right now. Steve stevecoast.com ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 7 August 2010 03:14, SteveC wrote: > If they have several orders of magnitude more money then probably the > cost/benefit tradeoff would suggest throwing out the data is the better > option. Even wikipedia doesn't take that attitude, they're currently being threatened by the FBI over a SVG image. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On Fri, Aug 6, 2010 at 12:14 PM, SteveC wrote: > > On Aug 6, 2010, at 11:09 AM, John Smith wrote: > > > On 7 August 2010 03:04, SteveC wrote: > >> Sounds like you've never been to court. Who's right or wrong is a > secondary consideration here, the first order of magnitude issue is who has > more money. We lose on that one. > > > > So basically anyone can make any copyright claim they like and OSM > > will throw out data rather than risk going to court over the matter? > > If they have several orders of magnitude more money then probably the > cost/benefit tradeoff would suggest throwing out the data is the better > option. > > Of course you could envisage entirely frivolous claims or losing huge > amounts of data, but I suspect it would more be a list of 10-100 users and a > relatively small set of data. Losing that compared to an injunction shutting > down OSM (which would be an early step if we didn't comply, as we're the > publisher of the data and safe harbor would be argued against) I'd pick lose > the data. Slightly offtopic (hah!), but I'm curious: is OSMF setup to respond to this sort of situation quickly while maintaining communication and input from the OSMF members and greater community (where "OSMF members" is a set of people contained within the set "greater community"). What procedure would be followed if OSM received a legal threat? To whom would the threat go? (I think this is important for more than legal-talk@ to know) ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On Aug 6, 2010, at 11:09 AM, John Smith wrote: > On 7 August 2010 03:04, SteveC wrote: >> Sounds like you've never been to court. Who's right or wrong is a secondary >> consideration here, the first order of magnitude issue is who has more >> money. We lose on that one. > > So basically anyone can make any copyright claim they like and OSM > will throw out data rather than risk going to court over the matter? If they have several orders of magnitude more money then probably the cost/benefit tradeoff would suggest throwing out the data is the better option. Of course you could envisage entirely frivolous claims or losing huge amounts of data, but I suspect it would more be a list of 10-100 users and a relatively small set of data. Losing that compared to an injunction shutting down OSM (which would be an early step if we didn't comply, as we're the publisher of the data and safe harbor would be argued against) I'd pick lose the data. Of course, IANAL. I've just taken people to court for copyright infringement in the past. Steve stevecoast.com ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 7 August 2010 03:04, SteveC wrote: > Sounds like you've never been to court. Who's right or wrong is a secondary > consideration here, the first order of magnitude issue is who has more money. > We lose on that one. So basically anyone can make any copyright claim they like and OSM will throw out data rather than risk going to court over the matter? ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On Aug 6, 2010, at 10:38 AM, Ed Avis wrote: > SteveC asklater.com> writes: > >> The reason is pretty simple - the first line of copyright defense if we get >> an >> email from TeleAtlas Legal saying 'user NearMap copied our data' is that we >> will remove _all_ NearMap data. > > Wouldn't you tell them to get lost, since copyright doesn't apply to map data, > etc etc? Sounds like you've never been to court. Who's right or wrong is a secondary consideration here, the first order of magnitude issue is who has more money. We lose on that one. Steve stevecoast.com ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
SteveC asklater.com> writes: >The reason is pretty simple - the first line of copyright defense if we get an >email from TeleAtlas Legal saying 'user NearMap copied our data' is that we >will remove _all_ NearMap data. Wouldn't you tell them to get lost, since copyright doesn't apply to map data, etc etc? -- Ed Avis ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
Richard Fairhurst writes: > Nick Black wrote: >> The current mechanism by which Mapzen and Mapzen POI Collector >> users authenticate against OSM is horrible for users. > > At the risk of being really hand-wavy and imprecise, I'd just say: Twitter's > OAuth UI is really exemplary. It's a great demonstration of how to get it > right. Every time I use an OAuth app with Twitter I think "wow, is that all > there is to it?". > > There's a slight difference in that the usual Twitter OAuth experience is > with a user who already has a Twitter account, whereas what we're talking > about here is setting up the account in the first place. But that's not > insurmountable. I downloaded mapzen poi collector early on. I was already an active osm conributor. So all i had to do was type my osm username and password into the app's preferences, and it's worked fine ever since. I can see that making it really easy for some exising oauth creds to register and agree to terms would help. pgpVCK16865E1.pgp Description: PGP signature ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
Nick Black wrote: > The current mechanism by which Mapzen and Mapzen POI Collector > users authenticate against OSM is horrible for users. At the risk of being really hand-wavy and imprecise, I'd just say: Twitter's OAuth UI is really exemplary. It's a great demonstration of how to get it right. Every time I use an OAuth app with Twitter I think "wow, is that all there is to it?". There's a slight difference in that the usual Twitter OAuth experience is with a user who already has a Twitter account, whereas what we're talking about here is setting up the account in the first place. But that's not insurmountable. cheers Richard -- View this message in context: http://gis.638310.n2.nabble.com/Revert-requests-in-general-tp5370633p5380468.html Sent from the General Discussion mailing list archive at Nabble.com. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 6 August 2010 20:04, Nick Black wrote: > How would the people voicing opinions in this thread feel about a hack > / planning day when editor developers, OSM-F and the OSM server admin > team can get together to talk through each side's concerns and come up > with a plan that is good for everyone - OSM-F, OSM admin team, editor > developers, and most importantly for mappers? > >From the NearMap point of view, I'd welcome that :) Cheers b -- Ben Last Development Manager (HyperWeb) NearMap Pty Ltd ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
Hi Guys, The current mechanism by which Mapzen and Mapzen POI Collector users authenticate against OSM is horrible for users. In each user test we do this is the main area where users fall down. We have several one star reviews on the App Store for Mapzen POI-C where users have got lost half way through the Auth process and ended up on the OSM site, thinking its Mapzen. We do user tests of Mapzen web every couple of months - we test on a range of people, from engineers to people off the street - and they all get confused about the account creation and auth process. There are a few things that CloudMade can do to make the process easier for the user and there are also things that could be done from the OSM side to make things easier. Please don't hold Mapzen up as a good example here - we have a technically spot on implementation of something that makes new users' lives hell and limits the ability of people to contribute to OSM. IMO, this is one of the most pressing issues that needs to be solved in order to reverse the stagnation of OSM monthly active contributors. The % of users actively contributing to OSM each month has fallen from 5.7% in March 2010 (13,675 / 238,985) to 4.7% in June (297,041 / 14,018). Of the 7,000 Mapzen POI Collector downloads so far, we see only a fraction get through the OAuth procedure (we have around 350 active contributors each month using POI-C). That equates to a massive missed opportunity in my book. I also appreciate the concerns and problems associated with using a single OSM account - that's why we supported OAuth from the beginning. How would the people voicing opinions in this thread feel about a hack / planning day when editor developers, OSM-F and the OSM server admin team can get together to talk through each side's concerns and come up with a plan that is good for everyone - OSM-F, OSM admin team, editor developers, and most importantly for mappers? -- Nick n...@cloudmade.com On Fri, Aug 6, 2010 at 8:48 AM, Kai Krueger wrote: > > > David Earl wrote: >> >> Are you going to take the email address on trust? It is really very easy >> to set up an OpenID provider which supplies any old email address on >> request. (There are some I think you can trust in principle - we know >> for example that Google and Yahoo provide verified email addresses, but >> in general I think it needs the round trip with the verification link in >> the email to be trustworthy). >> > Yes, it does still require the round trip to verify the email address. > Technically, it would be no problem to remove the need for that step, even > to selectively "white list" a few openID providers if they were deemed > trustworthy, but both presumably would need a wider agreement and discussion > in the community about the social implications. Hence for the moment, the > implementation is only really a replacement for having to remember yet > another username and password combo. > > Get the technical side and user interface right and deployed, and then one > can later still see if removing additional barrieres is a good next step. So > for the moment it is trying to solve a related, but not quite the same issue > as the NearMap, but perhaps it would help them too. > > > > David Earl wrote: >> >> Also, are you able to link OpenID logins together and with existing OSM >> accounts (i.e. keeping login identity separate from OSM identity and >> allowing OSM identities to have multiple ways of logging in)? >> > You can associate an OpenID to an existing account. You can also switch your > associated OpenID at any time (provided you are logged in) just like you can > change your password. (The OpenID is never revealed to anyone other than the > account holder). But as Tom sais, for the moment you can only associate a > single OpenID with your account. If there is demand for linking multiple > OpenIDs to a single account, it should be reasonably easy to change that in > the future though too. Also, if you choose, you can always continue to use a > standard password instead or in addition to the OpenID. > > Kai > -- > View this message in context: > http://gis.638310.n2.nabble.com/Revert-requests-in-general-tp5370633p5379621.html > Sent from the General Discussion mailing list archive at Nabble.com. > > ___ > talk mailing list > talk@openstreetmap.org > http://lists.openstreetmap.org/listinfo/talk > -- -- Nick Black twitter.com/nick_b ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
David Earl wrote: > > Are you going to take the email address on trust? It is really very easy > to set up an OpenID provider which supplies any old email address on > request. (There are some I think you can trust in principle - we know > for example that Google and Yahoo provide verified email addresses, but > in general I think it needs the round trip with the verification link in > the email to be trustworthy). > Yes, it does still require the round trip to verify the email address. Technically, it would be no problem to remove the need for that step, even to selectively "white list" a few openID providers if they were deemed trustworthy, but both presumably would need a wider agreement and discussion in the community about the social implications. Hence for the moment, the implementation is only really a replacement for having to remember yet another username and password combo. Get the technical side and user interface right and deployed, and then one can later still see if removing additional barrieres is a good next step. So for the moment it is trying to solve a related, but not quite the same issue as the NearMap, but perhaps it would help them too. David Earl wrote: > > Also, are you able to link OpenID logins together and with existing OSM > accounts (i.e. keeping login identity separate from OSM identity and > allowing OSM identities to have multiple ways of logging in)? > You can associate an OpenID to an existing account. You can also switch your associated OpenID at any time (provided you are logged in) just like you can change your password. (The OpenID is never revealed to anyone other than the account holder). But as Tom sais, for the moment you can only associate a single OpenID with your account. If there is demand for linking multiple OpenIDs to a single account, it should be reasonably easy to change that in the future though too. Also, if you choose, you can always continue to use a standard password instead or in addition to the OpenID. Kai -- View this message in context: http://gis.638310.n2.nabble.com/Revert-requests-in-general-tp5370633p5379621.html Sent from the General Discussion mailing list archive at Nabble.com. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 05/08/10 20:35, David Earl wrote: Are you going to take the email address on trust? It is really very easy to set up an OpenID provider which supplies any old email address on request. (There are some I think you can trust in principle - we know for example that Google and Yahoo provide verified email addresses, but in general I think it needs the round trip with the verification link in the email to be trustworthy). I assume it still verifies it, but Kai is the expert. Also, are you able to link OpenID logins together and with existing OSM accounts (i.e. keeping login identity separate from OSM identity and allowing OSM identities to have multiple ways of logging in)? Currently it only allows one openid per account, but you can still have a password as well and use that if you want. Tom -- Tom Hughes (t...@compton.nu) http://compton.nu/ ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 05/08/2010 14:44, Tom Hughes wrote: If the OpenID provider supplies sufficient data (basically an email address and nickname) then they need do little more than click OK to accept the details and then accept the terms. Are you going to take the email address on trust? It is really very easy to set up an OpenID provider which supplies any old email address on request. (There are some I think you can trust in principle - we know for example that Google and Yahoo provide verified email addresses, but in general I think it needs the round trip with the verification link in the email to be trustworthy). Also, are you able to link OpenID logins together and with existing OSM accounts (i.e. keeping login identity separate from OSM identity and allowing OSM identities to have multiple ways of logging in)? David ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
Hi, On 5 August 2010 17:09, Andy Allan wrote: > Let's imagine nearmap have been running their new editor and > 'cloaking' all their users under the one account for a couple of > years, and that their editor is great and everyone wants to use it. > > * I want to run a mapping party in Sydney - who's been editing in the > area? Ah, "nearmap". How many other people are there beyond just me? I > can't find out. But (as I mentioned in this thread) you usually look just at the last user editing every feature because the OSM XML for some reason includes this little bit of redundancy by putting the last editor's user name (historical data actually) in non-history extracts. This is not ideal, there may have been a big edit in the area removing created_by tags for example so the last editor's name tells you nothing. Or the last editor may have just changed their user name (so you need user id instead). So you need to look at full history anyway, at which point you can identify the users behind the nearmap account because Ben said changesets are tagged. It's not a cloak, it's a different way to store user identities. And IMHO it's completely justified considering most users will just have a single or a couple of simple edits (assuming what Ben said about the _simple_ editor is true), I imagine anyone who wants to start contributing regularly will be motivated enough to find out about the OSM project behind the map, and perhaps register directly. Cheers ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On Thu, Aug 5, 2010 at 1:42 PM, SteveC wrote: > Ben if I read this right then you're hiding the users from OSM and we'll see > a stream of edits from NearMap which are actually from multiple users. This > is why CM/matt/others built the OAuth code so that mapzen etc didn't do that, > because it's horrific. I agree, but for non-legal reasons. I know it's fashionable to only consider the licensing nowadays :-) Let's imagine nearmap have been running their new editor and 'cloaking' all their users under the one account for a couple of years, and that their editor is great and everyone wants to use it. * I want to run a mapping party in Sydney - who's been editing in the area? Ah, "nearmap". How many other people are there beyond just me? I can't find out. * Someone makes a change and I'm suspicious. What else has that person been doing? Oh, it's "nearmap" with about 1000 edits per hour. Hard to examine. * I'm looking at OWL and it looks like there's an edit war going on with the amenity key on the local restaurant/cafe. Who's doing that? "nearmap" and "nearmap", it seems. * I'm reading through the diary entries for OSM Things like this is what concerns me more than the legal aspects (which can be made bulletproof, but see below*) or the technical aspects around signing up. Cloaking the "nearmap users" from the rest of the community strikes a stake right into the heart of the community by separating it into two parts and putting nearmap as the gatekeepers. That's something I don't want to see - it's why I make sure OAuth existed, and why the small-screen work (for the iPhone) was implemented - every excuse I've seen for avoiding user-signups I've made sure it's been removed. And some other points: * If many edits are channelled through the one account we won't be able to ban it, it simple won't be allowed by the rest of the community. If we had vandals in Germany and the only way to stop them was to ban everyone else in Germany for 72hours there would be outrage. So lets not kid ourselves that we would have the moral authority to ban such an account. * Those of us who remember when anonymous editors made up a big chunk of activity remember that it was a really bad idea, and one that we changed our minds on for good reason. This is recreating the anonymous users, just calling them "nearmap" instead of anonymous. Cheers, Andy * Oh, and we should all be very, very suspicious of any entity who tries to build a community of OSM contributors where that entity has more rights over the contributions than OSM itself does. That's an obvious opportunity for a bait-and-switch if ever there was one. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
JohnSmitty wrote: > > Will they still need to register with OSM? > Have a look at the link to the source code I posted earlier (I know you are a coder, so I can send you that way...). You can also have a look at http://openid.dev.openstreetmap.org/ although that is by now outdated, has it hasn't been updated to reflect the git branch since the rails_port move from svn. The current version should be more user friendly, but the gist has more or less stayed the same. But the short answere (as also mentioned in my previous post) is yes. You still need to register for all the obviouse reasons. So it isn't ideal, but at least imho a step in the right direction. Kai -- View this message in context: http://gis.638310.n2.nabble.com/Revert-requests-in-general-tp5370633p5376560.html Sent from the General Discussion mailing list archive at Nabble.com. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 5 August 2010 23:44, Tom Hughes wrote: > If the OpenID provider supplies sufficient data (basically an email address > and nickname) then they need do little more than click OK to accept the > details and then accept the terms. That would probably satisfy Nearmap and others trying to minimise the barrier to entry... ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 05/08/10 14:42, John Smith wrote: On 5 August 2010 23:34, Tom Hughes wrote: On 05/08/10 14:33, John Smith wrote: On 5 August 2010 23:27, Tom Hughes wrote: Err actually, please don't start OpenID support from scratch as we already have a branch with more or less complete OpenID support! Is that OpenID support from other sites, like Nearmap, or is that OpenID support from OSM? Relying party - ie from other sites. Will they still need to register with OSM? Well of course - the OpenID URL has to be associated with some sort of local identifier that can be attached to any changes they make. If the OpenID provider supplies sufficient data (basically an email address and nickname) then they need do little more than click OK to accept the details and then accept the terms. Tom -- Tom Hughes (t...@compton.nu) http://compton.nu/ ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 5 August 2010 23:34, Tom Hughes wrote: > On 05/08/10 14:33, John Smith wrote: >> >> On 5 August 2010 23:27, Tom Hughes wrote: >>> >>> Err actually, please don't start OpenID support from scratch as we >>> already >>> have a branch with more or less complete OpenID support! >> >> Is that OpenID support from other sites, like Nearmap, or is that >> OpenID support from OSM? > > Relying party - ie from other sites. Will they still need to register with OSM? ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 05/08/10 14:33, John Smith wrote: On 5 August 2010 23:27, Tom Hughes wrote: Err actually, please don't start OpenID support from scratch as we already have a branch with more or less complete OpenID support! Is that OpenID support from other sites, like Nearmap, or is that OpenID support from OSM? Relying party - ie from other sites. Tom -- Tom Hughes (t...@compton.nu) http://compton.nu/ ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 5 August 2010 23:27, Tom Hughes wrote: > Err actually, please don't start OpenID support from scratch as we already > have a branch with more or less complete OpenID support! Is that OpenID support from other sites, like Nearmap, or is that OpenID support from OSM? ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 05/08/10 14:23, Serge Wroclawski wrote: Ben, why not look at the Rails code and offer an OpenID authentication mechanism. I can't speak for the administrators, but it seems like if some simple solution could be created that solves this ongoing issue with OpenID, that it would solve your problem entirely and benefit OSM directly. Err actually, please don't start OpenID support from scratch as we already have a branch with more or less complete OpenID support! Tom -- Tom Hughes (t...@compton.nu) http://compton.nu/ ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 5 August 2010 12:44, Kai Krueger wrote: > Frederik Ramm wrote: >> One signup page, one E-Mail >> confirmation, and then click "ok" for the OAuth page. How often does the >> modern Internet user do that every day? >> > Exactly that is the problem! I have to sign-up to far too many accounts per > day already. Risking new spam on the email, having to think of new Passwords > and usernames, as the "security rules" make passwords incompatible and thus > increase the chance of forgetting them. > > Quite often I then just say "Oh sod off! I can't be asked to go through yet > another process of creating a new password and forgetting it again." and > just not contribute. It is not worth going through that hassle of creating a > new password that by the time I will use this site again I will have > forgotten the password anyway. Bugmenot to the rescue! :) Oh too bad openstreetmap.org is blocked from Bugmenot, http://www.bugmenot.com/view/openstreetmap.org But osm.org lists five logins. Cheers ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
It seems to me we have two sides trying to reach the same end point. Ben and NearMap want to make it easy for people to use and contribute to OSM. Steve and Frederik want to ensure for technical and legal reasons that the changes from NearMap users doesn't cause problems in the OSM database. It seems like the solution to both is staring everyone in the face, since it's been mentioned. Ben, why not look at the Rails code and offer an OpenID authentication mechanism. I can't speak for the administrators, but it seems like if some simple solution could be created that solves this ongoing issue with OpenID, that it would solve your problem entirely and benefit OSM directly. Maybe start here: http://wiki.openstreetmap.org/wiki/Committing_to_the_rails_port - Serge ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 5 August 2010 11:27, Erik Johansson wrote: > On Thu, Aug 5, 2010 at 9:13 AM, Ben Last wrote: >> Actually... I'm not sure you would :) My reasoning is thus; OSM members are >> interested in mapping, and relish the power of JOSM or Potlatch (I do >> myself). You don't want a simpler editor, you want one that helps you do >> OSM mapping. The motivations and interests of the "average user" community > >> Or build a Facebook app so that both OSM and >> NearMap could let a user sign in with their Facebook credentials :) I can >> hear some people cringing, but there's a much bigger percentage of our users >> who are on Facebook than are on OSM. >> > > Is there any special reason why this should only be on nearmap.com? Yes, this is because, as Ben said the editor issue and the login issue are thightly related. You don't want to let users sign up to OSM too easily and make nearly anonymous edits. If they're to be allowed to do that, it should be coupled with a really simple editor that lets you change exactly the trivial things like fixing a street name or adding a POI, and even better if it's coupled with paid staff who either reviews the edits or is there ready to make reverts and take responsibility for it. (For such simple edits IMHO it is even be better to avoid having a new user account on OSM for every single edit, it won't help the DWG or anyone in anything now that I think of it) So I believe this can work really well, *if* the two things are connected and work together, the issues 1 and 2 Frederik listed. Cheers ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
Ben if I read this right then you're hiding the users from OSM and we'll see a stream of edits from NearMap which are actually from multiple users. This is why CM/matt/others built the OAuth code so that mapzen etc didn't do that, because it's horrific. The reason is pretty simple - the first line of copyright defense if we get an email from TeleAtlas Legal saying 'user NearMap copied our data' is that we will remove _all_ NearMap data. We don't have many other options as we don't have any money to fight and we are the infringer as the publisher. If you don't believe me, go ask a lawyer. There are lots of other reasons. I suggest you rethink. Steve stevecoast.com On Aug 5, 2010, at 1:13 AM, Ben Last wrote: > On 5 August 2010 14:44, Frederik Ramm wrote: > You're trying to remove two "barriers" at the same time, both quite unrelated: > 1. The barrier of users having to sign up to OSM; > 2. The barrier of a (supposedly) complicated editing process. > An interesting take on it :) But I disagree that these are not related. > Right now, you need to do 1 in order to do any edits, so it's not negotiable. > So you can't implement a better editor without users facing the barrier of > signup. We could build a better editor for users who are already signed up > with OSM, but that's a very small subset of the user population we're aimed > at, and I'm not sure that most OSM users want a simpler editor. > > Think of it as a use case; someone is happily using NearMap and finds that a > street they know well isn't named (this happens a lot). All they want is to > be able to quickly fix that. At that point in time, from their point of > view, given that this is almost certainly a side-issue to whatever they're > trying to do, the signup barrier is a really big deal. They don't care about > OSM, they don't care about mapping, they don't want to join an OSM community. > We have a small window of opportunity to have them help out before they lose > interest and motivation. So our starting point is that it has to be as easy > as possible for them to contribute. > > If we at OSM had an editor available that was easier than everything else we > can offer, we'd surely have put it up on the web site some time ago - but we > don't have one. So your effort and money on that front are surely welcome. > Actually... I'm not sure you would :) My reasoning is thus; OSM members are > interested in mapping, and relish the power of JOSM or Potlatch (I do > myself). You don't want a simpler editor, you want one that helps you do OSM > mapping. The motivations and interests of the "average user" community are > very different, and that drives the definition of "easier". > > I think the problem with your suggestion is that you're offering your help > only in the form of a package (1+2). > That's true; we do have valid reasons for doing that (well, we think they're > valid). We can't solve 1, because we don't run the OSM website, nor is there > a defined way in which we can help users sign up with some degree of > assurance that someone won't rework openstreetmap.org and break integration > with our site. We can do something to solve 2, but as expressed above, we > see 1 as a big barrier. If, as you suggest, there were a way to use openid > so that the OSM site could authenticate against our user database (or any > other openid one), then it wouldn't even be an issue; we'd just submit edits > with openid authentication. Or build a Facebook app so that both OSM and > NearMap could let a user sign in with their Facebook credentials :) I can > hear some people cringing, but there's a much bigger percentage of our users > who are on Facebook than are on OSM. > > Yes there would be an added burden for your users if you dropped "1", but > would that really be such a problem? One signup page, one E-Mail > confirmation, and then click "ok" for the OAuth page. How often does the > modern Internet user do that every day? > Given enough motivation, sure, people will sign up. But if the only reason > for signing up is to fix something that they think should be right in the > first place... not so much. > > If you were to decide to actually send your users to create an account with > OSM, you'd also be saving time because you would no longer have to be the > middle man in community communication. And if this is a factor for you, you > could still retain whatever rights you want on the content submitted by the > user, by way of their agreement with NearMap. > We save ourselves time at the expense of making it more work for our users. > Not really what we want to do. We're not interested in rights in the edits > (in fact, we have some rights anyway because those edits are derived from our > PhotoMaps and therefore we must be able to use them under CC-BY-SA). The > primary motivation here is to make the OSM data more usable, as fast as > possible. > > But having said th
Re: [OSM-talk] Revert requests in general
Frederik Ramm wrote: > > You're trying to remove two "barriers" at the same time, both quite > unrelated: > > 1. The barrier of users having to sign up to OSM; > 2. The barrier of a (supposedly) complicated editing process. > No, they are not really unrelated. If 1 is prerequisite of 2 (which it is) and 1 is the larger of the two "barriers" (which imho for the given target audience it is), then the two are very much related and offering 2 without 1 make much less sense Frederik Ramm wrote: > > On the other hand, doing "1" in the above, is relatively cheap; we could > do that ourselves at any time by, say, allowing users to log in to OSM > with any OpenID credentials (just like we do on help.openstreetmap.org). > I guess we might even do that one day if we get the messaging and > license stuff sorted out but we're not ready for that. > Well, imho having an OpenID password would be very useful (http://git.openstreetmap.org/?p=rails.git;a=shortlog;h=refs/heads/openID hint hint...), but the current implementation actually solves a slightly different issue and not the signup one, precisely because "1" is not cheap in terms of social and legal implications! The above OpenID implementation would still require you to sign-up, provide an email and agree to the CT, because OSM needs a contact, just that on top of all that you don't also have to come up and remember yet another username and password. The NearMap case would be very different though. NearMap would be fully responsible for its user account, thus from OSMs point of view all data is fully accountable for. For example OSM doesn't prevent me e.g from proxying edits for a colleague of mine who wants to change something but can't be bothered to sign up. Even more so, OSM doesn't prevent me from doing imports if I take the full responsibility of the data I add under my account. Indeed, if I am not mistaken, you have (or at least helped) imported data your self. So how is this different from a OSM point of view? If NearMap want to take responsibility for their users fine, that is their problem. If it turns out in the future that the NearMap account causes to much troubles and NearMap isn't living up to its responsibility, OSM can still block the entire NearMap account. (Blocking one account rather than thousands is much easier anyway) Again, that is more NearMap's problem then OSM's to deal with a potential fall out. So if NearMap is willing to take on this responsibility, and so far all indications are that they are and are fully aware of what they are getting into, why would OSM care? In fact it should be very happy that someone is investing the effort in providing these easy casual options, as OSM can't cater for these themselves. Frederik Ramm wrote: > One signup page, one E-Mail > confirmation, and then click "ok" for the OAuth page. How often does the > modern Internet user do that every day? > Exactly that is the problem! I have to sign-up to far too many accounts per day already. Risking new spam on the email, having to think of new Passwords and usernames, as the "security rules" make passwords incompatible and thus increase the chance of forgetting them. Quite often I then just say "Oh sod off! I can't be asked to go through yet another process of creating a new password and forgetting it again." and just not contribute. It is not worth going through that hassle of creating a new password that by the time I will use this site again I will have forgotten the password anyway. Unless of cause I want to become a regular contributor, but not for just fixing e.g. a spelling mistake! So if there are middle men like NearMap who are prepared to put in this effort to attract these people, then great. OAuth was designed for those people (site owners) who aren't prepared to act as middle man and therefore should be cut out as a "Middle man" with respect to passwords too, quite a different situation. Kai -- View this message in context: http://gis.638310.n2.nabble.com/Revert-requests-in-general-tp5370633p5376017.html Sent from the General Discussion mailing list archive at Nabble.com. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On Thu, Aug 5, 2010 at 9:13 AM, Ben Last wrote: > Actually... I'm not sure you would :) My reasoning is thus; OSM members are > interested in mapping, and relish the power of JOSM or Potlatch (I do > myself). You don't want a simpler editor, you want one that helps you do > OSM mapping. The motivations and interests of the "average user" community > Or build a Facebook app so that both OSM and > NearMap could let a user sign in with their Facebook credentials :) I can > hear some people cringing, but there's a much bigger percentage of our users > who are on Facebook than are on OSM. > Is there any special reason why this should only be on nearmap.com? I want simple anonymous data adding, with optional Facebook^W Open ID logon, on osm.org. Lets just keep on discussing all this in one big thread. 1. anonymous editing 2. 3rd party submission (proxy) 3. web apps for editing 4. how to ban users 5. what to log about anonymous users. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 5 August 2010 16:44, Frederik Ramm wrote: > On the other hand, doing "1" in the above, is relatively cheap; we could do > that ourselves at any time by, say, allowing users to log in to OSM with any > OpenID credentials (just like we do on help.openstreetmap.org). I guess we > might even do that one day if we get the messaging and license stuff sorted > out but we're not ready for that. So why do you do this, then Nearmap could setup an openid system at their end and problem solved... > If you were to decide to actually send your users to create an account with > OSM, you'd also be saving time because you would no longer have to be the It's obvious that they don't care about saving their time, they want to save their users time, so they don't see this as a barrier, where as sending users to OSM is a barrier, and as for people signing up for new accounts every day, there was studies produced in the past about how this can result in lost sales, so people aren't willing to do this as much as you think they will. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 5 August 2010 14:44, Frederik Ramm wrote: > You're trying to remove two "barriers" at the same time, both quite > unrelated: > 1. The barrier of users having to sign up to OSM; > 2. The barrier of a (supposedly) complicated editing process. > An interesting take on it :) But I disagree that these are not related. Right now, you need to do 1 in order to do *any* edits, so it's not negotiable. So you can't implement a better editor without users facing the barrier of signup. We could build a better editor for users *who are already signed up with OSM*, but that's a very small subset of the user population we're aimed at, and I'm not sure that most OSM users want a simpler editor. Think of it as a use case; someone is happily using NearMap and finds that a street they know well isn't named (this happens a *lot*). All they want is to be able to quickly fix that. At that point in time, from their point of view, given that this is almost certainly a side-issue to whatever they're trying to do, the signup barrier is a *really big deal*. They don't care about OSM, they don't care about mapping, they don't want to join an OSM community. We have a small window of opportunity to have them help out before they lose interest and motivation. So our starting point is that it has to be as easy as possible for them to contribute. > If we at OSM had an editor available that was easier than everything else > we can offer, we'd surely have put it up on the web site some time ago - but > we don't have one. So your effort and money on that front are surely > welcome. > Actually... I'm not sure you would :) My reasoning is thus; OSM members are interested in mapping, and relish the power of JOSM or Potlatch (I do myself). You don't want a simpler editor, you want one that helps you do OSM mapping. The motivations and interests of the "average user" community are very different, and that drives the definition of "easier". > I think the problem with your suggestion is that you're offering your help > only in the form of a package (1+2). That's true; we do have valid reasons for doing that (well, we think they're valid). We can't solve 1, because we don't run the OSM website, nor is there a defined way in which we can help users sign up with some degree of assurance that someone won't rework openstreetmap.org and break integration with our site. We can do something to solve 2, but as expressed above, we see 1 as a big barrier. If, as you suggest, there were a way to use openid so that the OSM site could authenticate against our user database (or any other openid one), then it wouldn't even be an issue; we'd just submit edits with openid authentication. Or build a Facebook app so that both OSM and NearMap could let a user sign in with their Facebook credentials :) I can hear some people cringing, but there's a much bigger percentage of our users who are on Facebook than are on OSM. > Yes there would be an added burden for your users if you dropped "1", but > would that really be such a problem? One signup page, one E-Mail > confirmation, and then click "ok" for the OAuth page. How often does the > modern Internet user do that every day? > Given enough motivation, sure, people will sign up. But if the only reason for signing up is to fix something that they think should be right in the first place... not so much. > If you were to decide to actually send your users to create an account with > OSM, you'd also be saving time because you would no longer have to be the > middle man in community communication. And if this is a factor for you, you > could still retain whatever rights you want on the content submitted by the > user, by way of their agreement with NearMap. We save ourselves time at the expense of making it more work for our users. Not really what we want to do. We're not interested in rights in the edits (in fact, we have some rights anyway because those edits are derived from our PhotoMaps and therefore we must be able to use them under CC-BY-SA). The primary motivation here is to make the OSM data more usable, as fast as possible. But having said that, if the response from OSM is "you need to make your users sign up", then maybe we have no choice and we'll have to rework what we've built. The problem I have right now is that I see conflicting advice from people who are all part of the OSM community - there is no single answer here. Cheers b -- Ben Last Development Manager (HyperWeb) NearMap Pty Ltd ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
Ben, Ben Last wrote: Actually, you can message them, since "they" are us (NearMap). Which is my point; the edits come from us, and we're the ones taking on the necessary responsibility. This is us, as a company aiming to support OSM, trying to remove barriers from contributions; that's going to involve us in spending effort and money in doing so. It's somewhat frustrating to find this being immediately classified in the same box as anonymous editing and/or vandalism. You're trying to remove two "barriers" at the same time, both quite unrelated: 1. The barrier of users having to sign up to OSM; 2. The barrier of a (supposedly) complicated editing process. Writing a good and simple to use editor, aka doing "2" in the above, is surely complicated, and takes a lot of effort. If we at OSM had an editor available that was easier than everything else we can offer, we'd surely have put it up on the web site some time ago - but we don't have one. So your effort and money on that front are surely welcome. On the other hand, doing "1" in the above, is relatively cheap; we could do that ourselves at any time by, say, allowing users to log in to OSM with any OpenID credentials (just like we do on help.openstreetmap.org). I guess we might even do that one day if we get the messaging and license stuff sorted out but we're not ready for that. I think the problem with your suggestion is that you're offering your help only in the form of a package (1+2). If you were to do only "2", and link the Nearmap account with the OSM account using OAuth, nobody would have any problem with that; your editor would be unreservedly welcome. People are critical of your package only because of "1", not because of "2". There is no technical necessity to package them. Yes there would be an added burden for your users if you dropped "1", but would that really be such a problem? One signup page, one E-Mail confirmation, and then click "ok" for the OAuth page. How often does the modern Internet user do that every day? If it should turn out that all the talk of "making OSM easier for a wider range of contributors" etc.etc. is indeed just "people cannot be bothered to go through a signup process", then it is not NearMap we need to fix this; then we need to discuss if we, as a project, can afford to drop the signup and use 3rd party ID in general. But this, I think, would a decision we'd have to take on a general level - are we happy with 3rd party ID - instead of doing that on a case by case basis - are we happy with NearMap relaying edits under a collective account. If you were to decide to actually send your users to create an account with OSM, you'd also be saving time because you would no longer have to be the middle man in community communication. And if this is a factor for you, you could still retain whatever rights you want on the content submitted by the user, by way of their agreement with NearMap. Bye Frederik -- Frederik Ramm ## eMail frede...@remote.org ## N49°00'09" E008°23'33" ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 5 August 2010 09:02, Ian Dees wrote: > So let's talk about making that process easier instead of using the current > broken system. Here we have Nearmap willing to spend time, money and other resources to address the issue and you want to waste further resources to discuss something no one else really seems to want to spend time and effort doing something about it. > Licensing is an important part of OSM, but I don't think I mentioned > anything about licensing in my response... You didn't even read the part of your email that I quoted? > 2) So we can communicate with the end mapper (regarding license changes, > community events, etc.). That looks like licensing to me... As I said, the people likely to make one off changes want to fix 1 maybe 2 errors, they don't care about licenses and they don't want to become part of a community. So this point is completely irrelevant. The alternative is like someone else pointed out, Nearmap adopt an OSB style system and then someone else has to update both the map data and the bug to achieve the same goal, I did this earlier in the week for some OSB reports and it's not much fun, in fact if there were a lot of them I wouldn't bother and I doubt anyone else will long term either. > Again, let's fix that problem instead of trying to wedge it into the current > system. I'm pretty sure this is inline with some of SteveC's ideas that everyone was shouting down earlier this year, so what happened after Steve stopped talking about it, pretty much nothing as far as I can see. If we want to keep growing I doubt we can exclude the efforts of others, like Nearmap, we just don't have the resources, or the foresight, to do something like this, there is too many nay sayers like yourself that shouts down anything remotely beyond the status quo and the existing user base. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On Wed, Aug 4, 2010 at 10:20 PM, Ben Last wrote: > It's somewhat > frustrating to find this being immediately classified in the same box > as anonymous editing and/or vandalism. > > I wanted to make it clear that I'm ecstatic to finally see a simple map editor coming out. I look forward to seeing how it works. The point I was trying to make is that we should figure out how to make this whole process easier rather than try and shoehorn it in to the current system. SteveC has talked about this several times: we should do everything we can to speed up immediate feedback to users' edits while maintaining quality and blocking spammy stuff. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 4 August 2010 16:27, Richard Fairhurst wrote: > The major problem arises when, for example, a NearMap user starts correcting > 300 street names using Google Maps as a source > (http://www.openstreetmap.org/blocks/10); or they make a whole bunch of > fictitious "corrections" a la Charlie Sheen Highway > (http://www.openstreetmap.org/blocks/19). > > There are procedures for when an actual OSM user does this - both > organisational procedures and code written specially for the purpose. But > it's a whole lot harder when there's an intermediary involved. We can't > message the user to say "this is against the terms & conditions you signed > up to" when they signed up to something else, and besides, they're behind a > proxy so they're not messageable anyway. Actually, you can message them, since "they" are us (NearMap). Which is my point; the edits come from us, and we're the ones taking on the necessary responsibility. This is us, as a company aiming to support OSM, trying to remove barriers from contributions; that's going to involve us in spending effort and money in doing so. It's somewhat frustrating to find this being immediately classified in the same box as anonymous editing and/or vandalism. Anyway; whatever the reasoning, it's clear that there are some who object to this. From our point of view, we need clarity on whether something is acceptable or not (and unfortunately clarity is not always possible to get from a mailing list). I'm going to email the Data Working Group, as per your (Richard's) helpful suggestion and see if I can get a clear response on this. Just noting the additional points made by Emilie and Andrzej; the system we have built tags all edits with information identifying the originating user (in our user database), exactly so that it's easy to find all related edits. We'd happily add any other necessary data to changesets or edits. We are also happy to provide ways to contact users (though we've accepted that would be our responsibility to do that if needed) and we're also able to (and prepared to) block users from editing if need be. Cheers b -- Ben Last Development Manager (HyperWeb) NearMap Pty Ltd ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On Wed, Aug 4, 2010 at 5:20 PM, John Smith wrote:
> On 5 August 2010 08:02, Ian Dees wrote:
> > I think the point that Frederik was trying to make was that this model
> > ("bulk imported in real time") is not ideal. Ideally, we want the users
> > interacting directly with the OSM API rather than going through some
> > intermediary service.
>
> It's obvious that Nearmap, and others, want something simpler than
> potlatch to allow people to add their home address or even just a
> missing street name as a one off, sure this might get abused and it
> will be up to Nearmap or others running these services to deal with
> abuse or face the problem of having their account blocked until they
> can. Making this process unnecessarily complicated is exactly the
> reason why Nearmap is attempting this in the first place.
So let's talk about making that process easier instead of using the current
broken system.
> These users don't give a toss about licenses, they just want to fix a
> mistake, such as a missing street name, why make things more
> complicated than that?
>
Licensing is an important part of OSM, but I don't think I mentioned
anything about licensing in my response...
>
> > OAuth was implemented for exactly this purpose. The user creates an
> account
> > on OSM.org, NearMap's client authenticates with OAuth, and the user can
> make
> > edits. It sounds like NearMap has an issue with sending the user off to
> > OSM.org to generate a user account and trying to draw them back in to
> > complete the OAuth process.
>
> It might have been, but that's authentication, not account creation,
> which is the whole point Ben made in the first place, they don't want
> to subject their users to multiple sets of terms and conditions and
> confirming account creation and so on and so forth just to add a
> street name, no wonder OSM is only for the geeks when the process has
> to be so convoluted and overly engineered just to fix a simple mistake
> like a missing street name.
>
Again, let's fix that problem instead of trying to wedge it into the current
system.
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
As long as the user is traceable, contactable and blockable (by
Nearmap), and that user is clearly reminded not to copy data off other
maps, then I'd let them get on with it.
Richard
On Wed, Aug 4, 2010 at 11:20 PM, John Smith wrote:
> On 5 August 2010 08:02, Ian Dees wrote:
>> I think the point that Frederik was trying to make was that this model
>> ("bulk imported in real time") is not ideal. Ideally, we want the users
>> interacting directly with the OSM API rather than going through some
>> intermediary service.
>
> It's obvious that Nearmap, and others, want something simpler than
> potlatch to allow people to add their home address or even just a
> missing street name as a one off, sure this might get abused and it
> will be up to Nearmap or others running these services to deal with
> abuse or face the problem of having their account blocked until they
> can. Making this process unnecessarily complicated is exactly the
> reason why Nearmap is attempting this in the first place.
>
>> We want this for at least two reasons:
>> 1) So we can follow our standard procedure for blocking users that perform
>> unwanted edits (whether they be vandals, inappropriate imports, or unusable
>> sources).
>
> As above, this will be up to Nearmap to police, and to some extent
> this should shift some burden from the OSM community onto others, with
> paid staff, to monitor so others can get on and do the mapping, I see
> this as a good thing!
>
>> 2) So we can communicate with the end mapper (regarding license changes,
>> community events, etc.).
>
> These users don't give a toss about licenses, they just want to fix a
> mistake, such as a missing street name, why make things more
> complicated than that?
>
>> OAuth was implemented for exactly this purpose. The user creates an account
>> on OSM.org, NearMap's client authenticates with OAuth, and the user can make
>> edits. It sounds like NearMap has an issue with sending the user off to
>> OSM.org to generate a user account and trying to draw them back in to
>> complete the OAuth process.
>
> It might have been, but that's authentication, not account creation,
> which is the whole point Ben made in the first place, they don't want
> to subject their users to multiple sets of terms and conditions and
> confirming account creation and so on and so forth just to add a
> street name, no wonder OSM is only for the geeks when the process has
> to be so convoluted and overly engineered just to fix a simple mistake
> like a missing street name.
>
> ___
> 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] Revert requests in general
On 5 August 2010 08:02, Ian Dees wrote:
> I think the point that Frederik was trying to make was that this model
> ("bulk imported in real time") is not ideal. Ideally, we want the users
> interacting directly with the OSM API rather than going through some
> intermediary service.
It's obvious that Nearmap, and others, want something simpler than
potlatch to allow people to add their home address or even just a
missing street name as a one off, sure this might get abused and it
will be up to Nearmap or others running these services to deal with
abuse or face the problem of having their account blocked until they
can. Making this process unnecessarily complicated is exactly the
reason why Nearmap is attempting this in the first place.
> We want this for at least two reasons:
> 1) So we can follow our standard procedure for blocking users that perform
> unwanted edits (whether they be vandals, inappropriate imports, or unusable
> sources).
As above, this will be up to Nearmap to police, and to some extent
this should shift some burden from the OSM community onto others, with
paid staff, to monitor so others can get on and do the mapping, I see
this as a good thing!
> 2) So we can communicate with the end mapper (regarding license changes,
> community events, etc.).
These users don't give a toss about licenses, they just want to fix a
mistake, such as a missing street name, why make things more
complicated than that?
> OAuth was implemented for exactly this purpose. The user creates an account
> on OSM.org, NearMap's client authenticates with OAuth, and the user can make
> edits. It sounds like NearMap has an issue with sending the user off to
> OSM.org to generate a user account and trying to draw them back in to
> complete the OAuth process.
It might have been, but that's authentication, not account creation,
which is the whole point Ben made in the first place, they don't want
to subject their users to multiple sets of terms and conditions and
confirming account creation and so on and so forth just to add a
street name, no wonder OSM is only for the geeks when the process has
to be so convoluted and overly engineered just to fix a simple mistake
like a missing street name.
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On Wed, Aug 4, 2010 at 4:53 PM, John Smith wrote:
> I'm slightly confused by all this talk about needing contractual
> agreements with all the end users and the OSM-F, or needing to
> identify Nearmap users to OSM-F.
>
> OSM already has data in the database from other projects, which was
> community sourced and licensed under various cc-by style licenses,
> sure it was bulk imported, possibly only once, and the only difference
> here is Nearmap will be bulk importing in real time, frankly they
> should be applauded for taking a pro-active approach to try to deal
> with faulty data themselves, rather than leaving it up to the OSM
> community to deal with it later like other bulk imports have.
>
>
I think the point that Frederik was trying to make was that this model
("bulk imported in real time") is not ideal. Ideally, we want the users
interacting directly with the OSM API rather than going through some
intermediary service.
We want this for at least two reasons:
1) So we can follow our standard procedure for blocking users that perform
unwanted edits (whether they be vandals, inappropriate imports, or unusable
sources).
2) So we can communicate with the end mapper (regarding license changes,
community events, etc.).
OAuth was implemented for exactly this purpose. The user creates an account
on OSM.org, NearMap's client authenticates with OAuth, and the user can make
edits. It sounds like NearMap has an issue with sending the user off to
OSM.org to generate a user account and trying to draw them back in to
complete the OAuth process.
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
I'm slightly confused by all this talk about needing contractual agreements with all the end users and the OSM-F, or needing to identify Nearmap users to OSM-F. OSM already has data in the database from other projects, which was community sourced and licensed under various cc-by style licenses, sure it was bulk imported, possibly only once, and the only difference here is Nearmap will be bulk importing in real time, frankly they should be applauded for taking a pro-active approach to try to deal with faulty data themselves, rather than leaving it up to the OSM community to deal with it later like other bulk imports have. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 4 August 2010 15:13, andrzej zaborowski wrote: > > Let's look at it practically. If a proxy (e.g. nearmap) user commits > vandalism, there are several things OSM may want to do: 1. undo the > vandalism, 2. contact the user, 3. block the user. > > For 1. it's actually better that the edits are logically grouped into > changesets, rather than imported by a 3rd party in 5 element > changesets. Obviously it would be even better if all the proxy user's > changesets were grouped in an individual user account. But Ben > mentioned that changes were going to be tagged, so I suppose it will > be possible to locate all the individual human editor's edits. > > For 2. again Ben mentioned that there would be a way to do that, and > for 3. he hasn't said anything but I expect they have thought of it > too. So considering this, blocking the entire account would be > overzealous. But then if it is eventually determined that nearmap.com > were the "bad guys", that would be useful. > > Yes, it would require support in editors like JOSM to see who edited a > given feature last.. on the other hand most of the times if you have > doubts about the quality of some change, you have to see the full > history of the object, because the interesting edit may have been > before last edit. > What nearmap could do is provide some kind of hash in the changeset that could help identify someone. Hopefully, that would allow us to point out to them when someone is behaving very badly. How that hash is defined is of course to be defined (it would probably be a composite key, and they would be the only who knows what it means). Emilie Laffray ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
Hi, On 4 August 2010 15:47, Frederik Ramm wrote: >>> You kind of have a point there with addresses and all; assume you'd just >>> produce your very own database of house numbers built by your users, then >>> release that, say, as PD or CC0. It would only be days until someone in >>> OSM >>> came along and proposed to import your database into OSM, which would >>> effectively end up the same (all data being contributed under one user >>> id). >>> >>> But in that scenario, the importing user would take full responsibility, >>> and >>> if it turned out that a significant portion of the import was in some way >>> faulty, the whole import would be rolled back. >> >> That's possibly the worst way to handle it > > I was just using that as an example which would lead to all data being under > the same account, and in a way that nobody on OSM's side would complain. I > wasn't suggesting they actually do that. Sorry then, good that it's clear. It did sound like a suggestion. > >> Note that nearmap.com is taking it pretty seriously about taking >> responsibility. > > I was trying to explain that this responsibility might mean that OSM takes > measures - such as reverting all contributions - against a particular user, > even if that one user happens to be a concentration point for contributions > of many human beings. Let's look at it practically. If a proxy (e.g. nearmap) user commits vandalism, there are several things OSM may want to do: 1. undo the vandalism, 2. contact the user, 3. block the user. For 1. it's actually better that the edits are logically grouped into changesets, rather than imported by a 3rd party in 5 element changesets. Obviously it would be even better if all the proxy user's changesets were grouped in an individual user account. But Ben mentioned that changes were going to be tagged, so I suppose it will be possible to locate all the individual human editor's edits. For 2. again Ben mentioned that there would be a way to do that, and for 3. he hasn't said anything but I expect they have thought of it too. So considering this, blocking the entire account would be overzealous. But then if it is eventually determined that nearmap.com were the "bad guys", that would be useful. Yes, it would require support in editors like JOSM to see who edited a given feature last.. on the other hand most of the times if you have doubts about the quality of some change, you have to see the full history of the object, because the interesting edit may have been before last edit. Cheers ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
Andrzej, You kind of have a point there with addresses and all; assume you'd just produce your very own database of house numbers built by your users, then release that, say, as PD or CC0. It would only be days until someone in OSM came along and proposed to import your database into OSM, which would effectively end up the same (all data being contributed under one user id). But in that scenario, the importing user would take full responsibility, and if it turned out that a significant portion of the import was in some way faulty, the whole import would be rolled back. That's possibly the worst way to handle it I was just using that as an example which would lead to all data being under the same account, and in a way that nobody on OSM's side would complain. I wasn't suggesting they actually do that. Note that nearmap.com is taking it pretty seriously about taking responsibility. I was trying to explain that this responsibility might mean that OSM takes measures - such as reverting all contributions - against a particular user, even if that one user happens to be a concentration point for contributions of many human beings. Also note that they mentioned in another thread that they want to contribute under share-alike licenses (including ODbL) so that they can use improvements made to data that they release or that is based on the imagery. The repeated asking for releasing as PD is amounting to trolling. As I said above, that was just a theoretical situation in which someone else's data would be concentrated under one account. I said "PD or CC0" in the example as a placeholder for "a license that is guaranteed to be compatible with anything OSM uses at the time". Bye Frederik ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
Hi, On 4 August 2010 11:10, Frederik Ramm wrote: > Ben, > Ben Last wrote: >> Interesting idea, but one aim of this whole effort is to increase the >> number of people who can contribute to OSM and help bring it to the >> point where OSM data is a usable way to do geocoding or address-search >> (which it isn't at the moment). Using OSB doesn't really meet that >> aim. > > I was thinking more along the lines of using OSB as an entry drug. > > You kind of have a point there with addresses and all; assume you'd just > produce your very own database of house numbers built by your users, then > release that, say, as PD or CC0. It would only be days until someone in OSM > came along and proposed to import your database into OSM, which would > effectively end up the same (all data being contributed under one user id). > > But in that scenario, the importing user would take full responsibility, and > if it turned out that a significant portion of the import was in some way > faulty, the whole import would be rolled back. That's possibly the worst way to handle it, and it has been done a couple of times already and I really hope we're moving away from this model to better collaboration with other projects, and so as to avoid multiple diverging databases. It's a *huge* amount of work trying to "sync" databases, a work that could really be avoided if a model like nearmap.com or that openaddress (was that the name?) are introducing (it wasn't discussed much when that OA site was announced..). I know first hand about this kind of imports. Note that nearmap.com is taking it pretty seriously about taking responsibility. Also note that they mentioned in another thread that they want to contribute under share-alike licenses (including ODbL) so that they can use improvements made to data that they release or that is based on the imagery. The repeated asking for releasing as PD is amounting to trolling. Cheers ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
Ben, Ben Last wrote: I'm pretty certain there was some kind of web-based tag editor just before OAuth was finally set up but I cannot find the mailing list references. There wasn't a huge discussion back then - it was clear to everyone that what that editor was doing could be a proof of concept at most because the account would soon be banned otherwise. Hmm. On what grounds would such an account be banned? I think at the time we assumed that either it would be banned because it invited unaccounted vandalism, or it would be banned after the first such vandalism occurred, which was just a question of time. I kind of understand your situation but I think the way forward would be to either use OpenStreetBugs or set up an OpenStreetBugs like system yourself, maybe integrate that in your editor - so that users without an OSM account can only place OSB markers, and those (the slightly more advanced users) who have an OSM account can then pick these up and fix OSM data properly. Maybe you can even do that in a way that lets people "start easy" in your application and then progress if they feel more comfortable with it. Interesting idea, but one aim of this whole effort is to increase the number of people who can contribute to OSM and help bring it to the point where OSM data is a usable way to do geocoding or address-search (which it isn't at the moment). Using OSB doesn't really meet that aim. I was thinking more along the lines of using OSB as an entry drug. You kind of have a point there with addresses and all; assume you'd just produce your very own database of house numbers built by your users, then release that, say, as PD or CC0. It would only be days until someone in OSM came along and proposed to import your database into OSM, which would effectively end up the same (all data being contributed under one user id). But in that scenario, the importing user would take full responsibility, and if it turned out that a significant portion of the import was in some way faulty, the whole import would be rolled back. I would think that the OSM position would be that it's not worth the *risk* to trace without it being clear that the licence allows it That is indeed generally the OSM position. Which is one reason we make it clear that you can derive data from our images under CC-BY-SA, to remove that risk. I'll post something about CC-BY-SA datasources in a separate thread. Bye Frederik ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
Ben Last wrote: In particular ODbL+CT will require a contractual relationship (i.e. the contributor terms) between OSMF and the user. If >> you are not exposing the user to the sign-up process, they >> are not agreeing to this contract. No, they're agreeing to terms and conditions with us. We (NearMap) are agreed to the terms and conditions with OSM, and submit the edits, as NearMap, to OSM. We're not trying to do some sort of back-to-back legal framework; that would never work. The edits come from NearMap. The major problem arises when, for example, a NearMap user starts correcting 300 street names using Google Maps as a source (http://www.openstreetmap.org/blocks/10); or they make a whole bunch of fictitious "corrections" a la Charlie Sheen Highway (http://www.openstreetmap.org/blocks/19). There are procedures for when an actual OSM user does this - both organisational procedures and code written specially for the purpose. But it's a whole lot harder when there's an intermediary involved. We can't message the user to say "this is against the terms & conditions you signed up to" when they signed up to something else, and besides, they're behind a proxy so they're not messageable anyway. The Data Working Group could in theory block the NearMap account every time this happens, and this would indeed be the standard way of sorting this out, given that it's the responsibility of the user (i.e. NearMap) to modify their behaviour. But I guess that wouldn't be something that appeals to you. :) [e-mailing OSMF] Glad to. Can you provide a way to contact someone there who'd be willing to have the conversation? http://www.osmfoundation.org/wiki/Working_Groups I'd suggest you e-mail the Data Working Group (d...@osmfoundation.org) as a first port of call. [tracing] I would think that the OSM position would be that it's not worth the *risk* to trace without it being clear that the licence allows it, since if it turned out that your opinion is wrong, that would lead to data loss. Yes, absolutely. cf the "Big Important Disclaimer" bit at the start of the blog post (in red and everything!). cheers Richard ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 4 August 2010 15:06, Frederik Ramm wrote: > Maybe worth taking a clue from Cloudmade here, who have a similar situation > with their Mapzen editor - they go through some effort to make the process > as painless as possible for their users while still requiring them to > register with OSM *as well as* with CM. Yes, we've looked quite closely at what Cloudmade do. They're a site aimed more at the mapping community than general user/media website use, so I'd argue that the cases are different. Certainly their editor is still way too complex for the average user (for our definition of "average user"). > I'm pretty certain there was some kind of web-based tag editor just before > OAuth was finally set up but I cannot find the mailing list references. > There wasn't a huge discussion back then - it was clear to everyone that > what that editor was doing could be a proof of concept at most because the > account would soon be banned otherwise. Hmm. On what grounds would such an account be banned? > One reason why we disallowed anonymous editing is to make sure that > community members can always be contacted by other community members if > their edits are worthy of discussion in one way or another. Do you have a > strategy of how do deal with incoming messages for the "nearmap" user? Will > you assign staff to forward these messages to the appropriate individual? Yes, and yes. There is actually another angle to this; the edits are being submitted by NearMap to OSM. The edits are given to NearMap by NearMap users, under an appropriate licence (that's part of the terms and conditions under which they'll register with us), and NearMap then submit them "separately" (in a legal sense) to OSM. So the edits are NearMap's, in a legal sense, and submitted by NearMap. From that point of view, it makes most sense for them to be submitted under the NearMap account. We have to balance legality and ease of use. > As I said, without knowing the internal Cloudmade procedures, I am pretty > sure there will have been a number of people in that organisation who'd have > said "are you mad, every additional signup button loses us 50% of people..." > but still they do what they do. For a reason, I guess. Indeed, and we have our reasons also; this isn't an arbitrary choice, it's been the subject of internal debate :) > I kind of understand your situation but I think the way forward would be to > either use OpenStreetBugs or set up an OpenStreetBugs like system yourself, > maybe integrate that in your editor - so that users without an OSM account > can only place OSB markers, and those (the slightly more advanced users) who > have an OSM account can then pick these up and fix OSM data properly. Maybe > you can even do that in a way that lets people "start easy" in your > application and then progress if they feel more comfortable with it. Interesting idea, but one aim of this whole effort is to increase the number of people who can contribute to OSM and help bring it to the point where OSM data is a usable way to do geocoding or address-search (which it isn't at the moment). Using OSB doesn't really meet that aim. Richard Fairhurst wrote: > I see your pain, but ease of getting map data into OSM doesn't trump concerns > of legality and ownership of data. Um... and we're not arguing that it does! It should be fairly clear from our website that we take legality and ownership of data at least as seriously as OSM :) >In particular ODbL+CT will require a contractual relationship (i.e. the >contributor terms) between OSMF and the user. If you are not exposing the user >to the sign-up process, they are not agreeing to this contract. No, they're agreeing to terms and conditions with us. We (NearMap) are agreed to the terms and conditions with OSM, and submit the edits, as NearMap, to OSM. We're not trying to do some sort of back-to-back legal framework; that would never work. The edits come from NearMap. >Your lawyers can of course find a way which satisfies them (and you) that >there is sufficient agreement between your user terms and CC-BY-SA/ODbl+CT, >but for any >novel way of getting data into OSM, the onus is on the importer to satisfy >_OSM_, not just themselves. That's the conversation we need to have here, and >potentially >also that you need to have with OSMF. (I would suggest that, as a courtesy, >you drop OSMF a line and ask them to consider the matter.) Glad to. Can you provide a way to contact someone there who'd be willing to have the conversation? >I'm of the opinion that tracing from aerial imagery does not carry through any >IP from the photography. Our lawyers, looking at more than just English law, would beg to differ :) But like you say, if that were true, people would be frantically tracing from Google imagery... and they're not. I would think that the OSM position would be that it's not worth the *risk* to trace without it being clear that the licence allows it, since if it turned out tha
Re: [OSM-talk] Revert requests in general
Frederik Ramm wrote: I kind of understand your situation but I think the way forward would be to either use OpenStreetBugs or set up an OpenStreetBugs like system yourself, maybe integrate that in your editor - so that users without an OSM account can only place OSB markers, and those (the slightly more advanced users) who have an OSM account can then pick these up and fix OSM data properly. Which is what Skobbler does, I think. http://www.skobbler.co.uk/osmbugs cheers Richard ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
Ben Last wrote: I'm not sure I agree. We don't want to put barriers in the way of an average user (and I use that term to explicitly distinguish between the average map site user and a mapping enthusiast) making simple corrections such as adding address information or naming un-named streets. In particular, we don't want to bounce them to the OSM site to register (and face yet another set of terms and conditions), when they're already registered on our site. I see your pain, but ease of getting map data into OSM doesn't trump concerns of legality and ownership of data. Otherwise I'd have introduced a Google aerial background into Potlatch like a shot. ;) As Frederik says, Mapzen - designed, like your editor, to lower the barrier to entry - is an instructive example. The OAuth support was introduced exactly so that other sites could provide OSM editors, whether Mapzen, the mooted OpenCycleMap editor, or whatever. In particular ODbL+CT will require a contractual relationship (i.e. the contributor terms) between OSMF and the user. If you are not exposing the user to the sign-up process, they are not agreeing to this contract. Your lawyers can of course find a way which satisfies them (and you) that there is sufficient agreement between your user terms and CC-BY-SA/ODbl+CT, but for any novel way of getting data into OSM, the onus is on the importer to satisfy _OSM_, not just themselves. That's the conversation we need to have here, and potentially also that you need to have with OSMF. (I would suggest that, as a courtesy, you drop OSMF a line and ask them to consider the matter.) My contention is that the only fair way to do it without imposing any risk on OSM is to require an explicit PD/CC0-type waiver from your users. For trivial edits made by a simple editor, this is probably good practice as they're unlikely to be substantial anyway. As per previously cited blog post (http://www.systemeD.net/blog/?p=100) I'm of the opinion that tracing from aerial imagery does not carry through any IP from the photography. It's up to the provider of the imagery whether they want to impose contractual restrictions. So the ball's in your court, really. :) > I hope by now that many OSMers will appreciate that we continue > to do a lot of support OSM, and that we do take the integrity and > reliability of the data very seriously. Absolutely. cheers Richard ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On Wed, Aug 4, 2010 at 3:45 AM, Richard Fairhurst wrote: > > Ben Last wrote: >> the edits that we're submitting all come from one user >> (that represents NearMap) since we don't (and can't) require >> users of our site to all be registered with OSM. > > Whenever it has been raised in the past, the opinion of the community has > generally been that proxy edits like this are strongly discouraged. You should go ahead with nearmap editing, even if it might become an anonymous map doodle tool. It's a good thing that we are getting more anonymous editting tools. It would be nice to beable to at least track in the same way as wikipedia does (publicly showing IP addresses). Amenity Editor, the web based anonymous editor, hasn't posed a problem for me yet, but the edits come in pretty low volume. http://ae.osmsurround.org/ http://www.openstreetmap.org/user/amenityeditor/edits /emj ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
Hi, Ben Last wrote: And I've tried to discuss it on a few occasions, and not had much of a response :) We've (at this point) ruled out asking all our users to register separately with OSM Maybe worth taking a clue from Cloudmade here, who have a similar situation with their Mapzen editor - they go through some effort to make the process as painless as possible for their users while still requiring them to register with OSM *as well as* with CM. Maybe my Google skills are sadly lacking, but I spent a while looking for information on this. There's some discussion of the freietonne.de site back in January, but nothing else significant. I'm pretty certain there was some kind of web-based tag editor just before OAuth was finally set up but I cannot find the mailing list references. There wasn't a huge discussion back then - it was clear to everyone that what that editor was doing could be a proof of concept at most because the account would soon be banned otherwise. Nor is there anything on the wiki; the OAuth section discusses technical usage, but there is no statement about proxy editing, the section of automated edits doesn't cover this. Note also that this is not anonymous editing; the identity of the user submitting the change to us is tagged with every edit. One reason why we disallowed anonymous editing is to make sure that community members can always be contacted by other community members if their edits are worthy of discussion in one way or another. Do you have a strategy of how do deal with incoming messages for the "nearmap" user? Will you assign staff to forward these messages to the appropriate individual? Previous discussions, IIRC, took place in the days before OAuth. Now that we have OAuth on OSM there is even less reason to allow editing without an OSM account. I'm not sure I agree. We don't want to put barriers in the way of an average user (and I use that term to explicitly distinguish between the average map site user and a mapping enthusiast) making simple corrections such as adding address information or naming un-named streets. In particular, we don't want to bounce them to the OSM site to register (and face yet another set of terms and conditions), when they're already registered on our site. As I said, without knowing the internal Cloudmade procedures, I am pretty sure there will have been a number of people in that organisation who'd have said "are you mad, every additional signup button loses us 50% of people..." but still they do what they do. For a reason, I guess. I'm not saying that we have to implement it this way, but up until this point we've not been aware of any great community disapproval of proxy editing as a principle. I can absolutely understand that it's not something that should be undertaken lightly, but I hope by now that many OSMers will appreciate that we continue to do a lot of support OSM, and that we do take the integrity and reliability of the data very seriously. I kind of understand your situation but I think the way forward would be to either use OpenStreetBugs or set up an OpenStreetBugs like system yourself, maybe integrate that in your editor - so that users without an OSM account can only place OSB markers, and those (the slightly more advanced users) who have an OSM account can then pick these up and fix OSM data properly. Maybe you can even do that in a way that lets people "start easy" in your application and then progress if they feel more comfortable with it. Bye Frederik -- Frederik Ramm ## eMail frede...@remote.org ## N49°00'09" E008°23'33" ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On 4 August 2010 09:45, Richard Fairhurst wrote: > Ben Last wrote: >> the edits that we're submitting all come from one user >> (that represents NearMap) since we don't (and can't) require >> users of our site to all be registered with OSM. > Um... this is the sort of stuff that really, really needs to be discussed > first. And I've tried to discuss it on a few occasions, and not had much of a response :) We've (at this point) ruled out asking all our users to register separately with OSM, and we can't register with OSM on their behalf (complex both technically and legally). So we don't have many options. > Whenever it has been raised in the past, the opinion of the community has > generally been that proxy edits like this are strongly discouraged. Maybe my Google skills are sadly lacking, but I spent a while looking for information on this. There's some discussion of the freietonne.de site back in January, but nothing else significant. Nor is there anything on the wiki; the OAuth section discusses technical usage, but there is no statement about proxy editing, the section of automated edits doesn't cover this. Note also that this is not anonymous editing; the identity of the user submitting the change to us is tagged with every edit. > At the > very least, you need to make sure your contributor agreement is widely > discussed so that others (not least OSMF, which may be held liable for any > copyright violations on the part of your users) are happy that they are > compatible with current and future OSM licensing. Our signup terms and conditions are being rewritten (by lawyers) to explicitly handle this issue. > Previous discussions, IIRC, took place in the days before OAuth. Now that we > have OAuth on OSM there is even less reason to allow editing without an OSM > account. I'm not sure I agree. We don't want to put barriers in the way of an average user (and I use that term to explicitly distinguish between the average map site user and a mapping enthusiast) making simple corrections such as adding address information or naming un-named streets. In particular, we don't want to bounce them to the OSM site to register (and face yet another set of terms and conditions), when they're already registered on our site. I'm not saying that we have to implement it this way, but up until this point we've not been aware of any great community disapproval of proxy editing as a principle. I can absolutely understand that it's not something that should be undertaken lightly, but I hope by now that many OSMers will appreciate that we continue to do a lot of support OSM, and that we do take the integrity and reliability of the data very seriously. Cheers Ben -- Ben Last Development Manager (HyperWeb) NearMap Pty Ltd ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
Ben Last wrote: > the edits that we're submitting all come from one user > (that represents NearMap) since we don't (and can't) require > users of our site to all be registered with OSM. Um... this is the sort of stuff that really, really needs to be discussed first. Whenever it has been raised in the past, the opinion of the community has generally been that proxy edits like this are strongly discouraged. At the very least, you need to make sure your contributor agreement is widely discussed so that others (not least OSMF, which may be held liable for any copyright violations on the part of your users) are happy that they are compatible with current and future OSM licensing. (My personal opinion is that the only circumstance under which such contributions are acceptable is under a PD-like waiver. If you are indeed requiring that then please disregard this message. http://www.systemeD.net/blog/?p=100 may be relevant here.) Previous discussions, IIRC, took place in the days before OAuth. Now that we have OAuth on OSM there is even less reason to allow editing without an OSM account. cheers Richard -- View this message in context: http://gis.638310.n2.nabble.com/Revert-requests-in-general-tp5370633p5370865.html Sent from the General Discussion mailing list archive at Nabble.com. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On Tue, Aug 3, 2010 at 7:37 PM, Ben Last wrote: > I think I remember something like that; however, the edits that we're > submitting all come from one user (that represents NearMap) since we > don't (and can't) require users of our site to all be registered with > OSM. So we obviously don't want that single username to get banned. > What I'm after is some process by which we can flag/raise changeset > numbers ourselves and get them requested for reversion asap. If we > could revert our own changesets that'd be good, but I don't have > enough detail about the api 0.6 revert facilities to be sure that > they'll do what we need, especially if the changeset is a couple of > weeks old at the time that an issue is spotted. Anyone can run the revert script if they want to. I think Frederik wrote it in Perl and it's sitting somewhere in SVN. I think the community has decided (so far, anyway) that we'd rather secure that functionality under a mild layer of obscurity for now. Perhaps we should start that topic again: should we have an automatic revert on OSM? It's technically feasible (but difficult). The deeper issue, though, is not the revert tool, but the lack of history viewing tools that would help make the decision about what to revert. How do we know what changesets are bad? We don't currently have a good way to access the complete history of a way, for example (including where it's nodes were in previous revisions so that we can draw what it looked like in the past). Perhaps that tool needs to be thought about before we think about how to do a revert. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
I think I remember something like that; however, the edits that we're submitting all come from one user (that represents NearMap) since we don't (and can't) require users of our site to all be registered with OSM. So we obviously don't want that single username to get banned. What I'm after is some process by which we can flag/raise changeset numbers ourselves and get them requested for reversion asap. If we could revert our own changesets that'd be good, but I don't have enough detail about the api 0.6 revert facilities to be sure that they'll do what we need, especially if the changeset is a couple of weeks old at the time that an issue is spotted. Cheers b On 4 August 2010 08:32, Ian Dees wrote: > On Tue, Aug 3, 2010 at 7:11 PM, Ben Last wrote: >> >> On 3 August 2010 18:13, Dave F. wrote: >>> >>> Can somebody revert this edit ASAP? >> >> Which reminds me... we at NearMap are preparing to begin rollout of OSM >> editing on our site; simple operations like adding house numbers and >> adding/correcting street names. Edits can be made by registered users of >> our site through a simple interface that's not as scary as Potlatch :) Our >> aim is therefore to encourage non-mappers to contribute to OSM, since they >> won't have to worry about the complexities of tagging rules, etc. Whilst we >> have a number of things in place to guard against vandalism and incorrect >> edits, we're also looking at the best ways to (a) tag uploaded edits so that >> they can be clearly extracted from the data and (b) work with the OSM >> community to revert bad edits. We could do (b) by emailing the various >> lists, but is there a more efficient way to arrange to get edits reverted? >> I'm looking for whatever way is least load on us *and* those in OSM who >> maintain the data. > > Wasn't there a discussion a long while ago about a system of flagging > changesets (and users, too, probably) as bad/inappropriate/spammy? Enough > bad changesets = a nomination for revert or temp. ban/kick? -- Ben Last Development Manager (HyperWeb) NearMap Pty Ltd ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Revert requests in general
On Tue, Aug 3, 2010 at 7:11 PM, Ben Last wrote: > On 3 August 2010 18:13, Dave F. wrote: > >> Can somebody revert this edit ASAP? >> > > Which reminds me... we at NearMap are preparing to begin rollout of OSM > editing on our site; simple operations like adding house numbers and > adding/correcting street names. Edits can be made by registered users of > our site through a simple interface that's not as scary as Potlatch :) Our > aim is therefore to encourage non-mappers to contribute to OSM, since they > won't have to worry about the complexities of tagging rules, etc. Whilst we > have a number of things in place to guard against vandalism and incorrect > edits, we're also looking at the best ways to (a) tag uploaded edits so that > they can be clearly extracted from the data and (b) work with the OSM > community to revert bad edits. We could do (b) by emailing the various > lists, but is there a more efficient way to arrange to get edits reverted? > I'm looking for whatever way is least load on us *and* those in OSM who > maintain the data. > Wasn't there a discussion a long while ago about a system of flagging changesets (and users, too, probably) as bad/inappropriate/spammy? Enough bad changesets = a nomination for revert or temp. ban/kick? ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
