[Osmf-talk] Anonymous notes on osm.org

[Kai Krueger]

Hello everyone,

I presume you are all familiar with OpenStreetBugs and I suspect you 
also have heard that there are efforts underway to bring a similar 
concept to the front page of osm.org to increase the exposure of map 
bugs and notes to a wider audience of mappers and problem reporters.

The current efforts can be seen at

http://notes.apis.dev.openstreetmap.org/



One of the outstanding issues needing to be addressed before deployment 
is whether to allow anonymous (i.e. non logged in users) notes and 
comments. Although part of the issue is technical, it is equally a 
question of policy and if OSM wants to allow it or not. That is the 
reason I bring it up here on this list to see if a consensus can be found.

The question has already to a degree been discussed on the rails-dev 
list [1] and on talk-de [2] (probably amongst other places).

I'll try and summarize the points raised pro and contra so far:

- A question of legality: Some people have raised the concern of legal 
issues. One hope for the notes system is that people provide good enough 
bug reports for (experienced) mappers to enact on these bug reports 
without needing to have own local knowledge of the issue. However, with 
anonymous reports, the knowledge of the issues (e.g. the name of a road) 
might be copied from proprietary sources like Google maps. As a non 
logged-in user will not have confirmed the contributor terms and 
therefore has not vouched for non infringement of third party rights of 
their contributions, can this information be used directly?

- A question of spam: Having an open API to create notes on a (sort of) 
high profile page like osm.org probably invites the miss use of the 
feature in particular in form of spam. Obviously requiring a log-in 
won't completely prevent spam, but it might reduce its volume? An 
alternative would be to add e.g. captcha's to the posting process of 
notes. However, once you do that, is it really much easier than to 
create an account? Especially with the simplified sign-up process via 
third party authentication of e.g. "login with gmail"? It might however 
be a significantly higher psychological barrier though.

- A question of communication: A key component of OSM is the community 
and the discussions of local knowledge and how best to map it. With 
anonymous users it might be more likely that a two way communication is 
not possible, as a bug reporter may not respond to clarifying questions. 
This might leave a lot more bug reports open that could be solved with a 
quick answer to a short question of clarification.

This also brings up a secondary question, assuming the decision is to 
support anonymous users. Should the system allow people to add an 
arbitrary name to their report / comment or should they just be listed 
as "by anonymous"?

- A question of simplicity of use: The main argument for anonymous users 
is obviously the simplicity of use. One of the key points of the notes 
system is to make it as simple as possible for people to contribute 
their local knowledge even if they find actually editing OSM data to 
challenging or daunting and don't want to put in the effort to learn the 
complexities of OSM and its tagging schemata. Adding a signup / login 
process before being able to contribute valuable knowledge might be too 
high of a hurdle, scaring away people who might other wise contribute.

- A question of third party use: A number of third party sites using 
data and showing maps of OSM have included interfaces to add "report a 
problem" links to the old OpenStreetBugs. It would obviously be 
encouraged to do the same again with the new system. However, how would 
a login requirement to OSM interact with this third party use of OSM 
(whereby third party might be an editor like JOSM)? On the one hand it 
would probably make implementing this on third party sites more 
complicated as the site would possibly have to implement something like 
OAuth to osm.org. On the other hand it might increase the awareness of 
users of what osm is and what they are actually commenting on. For 
example one major flaw with the other wise nicely done bug reporting 
system of the navigation app Skobbler (MapDust) was the far to low 
signal to noise ratio. Quite a number of reports were of the form that 
e.g sound in skobbler didn't work or other software interface issues 
that had nothing to do with OSM's map quality data. Will requiring a 
signup to an openstreetmap account help educate people about what OSM is 
if they click the "report a bug" link on some unrelated third party site?

- A question of capacity: From  a more practical point of view, some 
have raised the issue of whether the mapping community has the capacity 
to deal with a larger influx of bug reports. If a lot more reports get 
submitted than the community can deal with, the notes system will 
perhaps cause more frustration than actually help solve problems. 
Requiring a log-in (at least initially until the capacity of the system 
and community is better understood) might reduce the volume to more 
manageable levels and favor those committed enough to be willing to sign 
up to an account, increasing the chance of providing enough information 
to solve the bug. But that is somewhat speculative at the moment.


As I am sure the discussion will show, there are many more pro and 
contra points for anonymous notes. But perhaps the more important point 
is to come to a workable conclusion to be able move forward on finally 
deploying the notes branch and not let the perfect be the enemy of the 
good. After all things aren't set in stone in the first iteration of 
deployment.

Kai




[1] 
http://lists.openstreetmap.org/pipermail/rails-dev/2012-October/001080.html
[2] 
http://lists.openstreetmap.org/pipermail/talk-de/2012-October/099091.html