[Talk-us] NYC Name Vandalism

Frederik Ramm frederik at remote.org
Wed Sep 5 07:45:39 UTC 2018


On 05.09.2018 03:36, Alan Brown wrote:
> Granted, it would be nearly impossible to make this criteria perfect:

I think it would already be nearly impossible to make these criteria
even *good*. It is easy to come up with a knee-jerk "nobody should be
allowed to change the name tag of New York", and many will nod in
approval. It seems obvious, doesn't it. Who, then, makes the catalogue
of such places? Is only their name tag "protected"? Or also their
location? Can the node be moved by a mile, 10 miles, 100 miles? Can the
population be changed, and if so, by what amount?

> I'd have no idea what would be
> offensive in Hungarian, much less Thai; someone could draw something
> offensive (like a peeing Android) that would be very hard to catch;
> there are places like "Dildo, Newfoundland" that are legitimate.

All this is true, and simple regular expression matching will never fix
things (the village of Fucking in Austria is a well-known example but
the number of names that are legit in one language and offensive in
another is high).

> But I
> don't think it would be all that hard to flag a changelist like this
> last vandalism,

If you prohibit me from changing the name of New York to "Jewtropolis",
I'll just create a city node one block away from it with a slightly
higher population, causing it to be rendered with priority.

If you start down this road, you will end up not using OSM place names
at all but instead relying on a curated data set like Natural Earth,
which is a valid decision to make for a cartographer but means taking
control away from mappers and giving it to a hand-picked circle of data

> At very least, you can force your vandals to be clever to succeed.

But is this really what we want - ever more clever vandalism that is
ever less likely to be detected? Is it not even *better* to have
"obvious" vandalism that we can fix easily? Today, getting "Jewtropolis"
written large across OSM for an hour or two is no big deal, nothing to
brag about before your cool hacker friends. "So what" is the answer. Do
we want to make this into a trophy? Today, the headline is "some asshole
put 'Jewtropolis' on OSM" - tomorrow, "clever hacker penetrates OSM

> In our usage, we will scan the names of significant objects for
> potentially offensive changes.  But it would be good to have some sort
> of gateway in the OSM database itself. 

It is ok for a data consumer to do that. Nobody is hurt if your filters
wrongly reject a valid contribution in Africa. It would also be ok to
build something that prioritizes things for review. But trying to build
some kind of "protection" into the data ingestion at OSM would

* impact performance negatively
* disenfranchise mappers
* bind resources for the constant maintenance of block lists
* encourage clever(er) vandalism

and hence not be worth it.


Frederik Ramm  ##  eMail frederik at remote.org  ##  N49°00'09" E008°23'33"

More information about the Talk-us mailing list