<p></p><div>Hello everyone,<br></div><div>I've been a contributor of OpenStreetMap for a few year, with a
couple of different accounts. I got them deleted today and I though it
could be worthwhile talking about this here.<br></div><div> <br></div><div> In our current era of big data, I have been more and more concerned
about having all my osm edits publicly linked to my profiles, and these
profiles publicly listing all these positions and places where I've
been, also with somehow time information and sometimes comments, etc...
visible forever by anyone, or any bot. I've looked into making the link
between all that data not publicly visible, but it seems the
functionality there use to be for that (anonymous editing) is not
possible since 2007/2009.<br></div><p></p><p>I've read (a good few of the) related e-mails from that time [1], and
I understand that there was an important ground and a general consensus
for that decision, despite a minority of voice disappointed by this
"security rather than freedom" direction being taken.<br></p><p>My humble point of view is that with the important evolution that OSM
has experienced since back then, it would be a very good thing, if not
done already, to revisit this issue and find a middle point which
possibly was not easily feasible at the time but would now be more
achievable. For instance, if I had a tick box "do not publicly link
changes to my account", either at account level or at changeset level,
but that every user still had the possibility to send a message to the
author of such edits, and to roll them back (even potentially with a
procedure for banning users with too much anonymous changes rolled back
by the community, as the edits-author link is not lost, it's just not
visible to users, whether registered or not). Then, I think, everyone
would be happy, or close to? I mean, I think this would address both the
concerns that led to the decision of disabling anonymous editing back
in 2007 and the privacy concerns I summarised above.<br></p><p>(At the time it was also mentioned that OSM is all about being a
community project, and that it was probably inconsistent to allow
anonymous contributions in that context. In particular, a comparison
with Wikipedia was made. My opinion on this is that location-related
information are in general much more privacy-critical than Wikipedia
edits are, in particular now that you have user-friendly mobile apps for
mapping on the go, and therefore the comparison is inappropriate.)<br></p><p>I've been looking a bit around to see if there was a plan for
developing something like that anytime soon, or if it had been
implemented already, but I couldn't find. I've mailed the support team,
who confirmed there is currently no way (other than closing an account)
to
make edits become anonymous.<br></p><p></p><div>Therefore I'm afraid the only way forward I see to address my concerns is the following:<br></div><div> 1) on the one hand having my past accounts deleted, for the
corresponding change-sets not to be linked any more to my name or
pseudonym. I got that done today.<br></div><div> 2) on the other hand, from now on, to periodically create and abandon
accounts for keeping editing without a massive correlation of data being
too easily possible (but even like that it's an unperfect tradeoff).<br></div><p></p><p>I'm not sure how much of the community is having concerns similar to
mine, but I would guess that these can only have gone bigger and bigger
since back in 2007. As said above, I believe it would be worth having a
think about it again. (But maybe it has already been discussed again
recently, and I didn't find out?)<br></p><p>What are your thoughts?<br></p><p>Looking forward,<br></p><p>Jibix<br></p><p>[1]: <a href="https://lists.openstreetmap.org/pipermail/talk/2007-October/thread.html#18853">https://lists.openstreetmap.org/pipermail/talk/2007-October/thread.html#18853</a><br></p><p></p><div>--<br></div><div> Jibix<br></div><div> favourite webpage of the moment: <a href="https://degooglisons-internet.org/alternatives?l=en">https://degooglisons-internet.org/alternatives?l=en</a><br></div><p></p><div><br></div>