[OSM-talk] Zero tolerance on imports

[Serge Wroclawski]

On Mon, Feb 21, 2011 at 10:03 AM, Peter Budny <peterb at gatech.edu> wrote:
> I find this discussion very distasteful.

Peter,

I'm sorry that you feel that way. Can you tell us what specifically
you find distasteful? Is it the issue of automated edits or is there
something else?

>
> I really like OSM's goals: a complete map of the Earth with more-or-less
> unlimited detail.  But I don't understand why people think that this
> 500+ GIGABYTE map should be managed using 19th-century methods,
> i.e. manual labor.

I the issue, which Wikipedia also faces, is that automated edits have
a history of being problematic for the project.

A very small number of people in the community are against all
automated edits, but a majority of us are concerned that current
automated edits have been so problematic as to turn away users and
contributors.

What we're discussing is how to get good automated edits while still
allowing for 20th century methods of input, which include on the
ground surveying.

> Waze is a much newer project than OSM.  I don't really care for its data
> model, because I think it's much too limited in the amount of detail
> it's capable of capturing; OSM's model is much better IMO.

Waze is proprietary. Is that right?

>  BUT, Waze
> has captured traces of a much larger portion of the US than OSM has.
> Waze has both average and real-time speed data, whereas OSM has no
> provision for this whatsoever.

That's true. I think some of this data could have come in through
Transiki, but that project appears permanently on hold.

Maybe y0u can be one of the people to revive it?

> At some point, OSM will reach a size -- either size of database, or
> number of users/contributors -- where it will become totally infeasible
> to manage with the tools we have (or rather, the tools we don't have).

That's the issue we're discussing, and await your suggestions on ways
to move forward with better tools.

> Those of you who think all automated or semi-automated data
> contributions are harmful to OSM are dooming this project to never be
> able to grow to become a leading source of mapping data.

Peter, how does this mail help move the conversation for better tools forward?

> Do you think that when MapQuest started using OSM data to generate their
> maps, they performed all the necessary data transformations BY HAND?

They didn't make many changes to OSM AFAICT. According to the talk
they gave, they did indeed find many problems by hand, and had to make
manual fixes.

> Last year, as part of a school project, I built a robot that will
> automatically create route relations for all the state highways in the
> US, being careful not to change or duplicate existing data.  I haven't
> shared it with the community because a handful of users were so
> terrified of the prospect of automated edits, they insisted I do a
> large-scale trial run on a local copy of the database, and I haven't had
> the time to compile the results of those trial runs for review.  The
> code would be in use already if not for a few people running around
> panicking about my devil-robot and its witchcraft.

How could those user's concerns be addressed while still making things
easy for you to do?

- Serge