[Osmf-talk] Alternative Strategic Plan

[Brian M. Sperlongano]

On Sat, May 13, 2023 at 1:32 PM Steve Coast <steve at stevecoast.com> wrote:

> Do we want to complete the map or not? That’s the question.
>

I would say no, or at the very least this is the wrong way of looking at
the problem.

I contribute to OSM because the applications and services that use OSM data
make people's lives better. Whether it's finding a faster way to get to
work, locating a hiking trail in your local nature reserve, or compiling a
list of ice cream shops on your quest to visit every one of them in your
city, the end goal of all of this is that somebody benefits from it.

A better way to frame the problem that I think you're describing is: "Are
there ways we can motivate mappers to contribute data that has the biggest
impact on making people's lives better?"

>From there you can step down and ask what gaps in data are the most
important. I view this in a way that's agnostic to OSM being the solution
in all cases. In many cases, a certain class's "complete" data exists in a
government database that isn't fully incorporated into OSM. For example,
names of places in other languages are widely present in wikidata.
Geo-located addresses in the United States are available in a US government
database. Canada has a national land cover database, as does France. I
don't feel that competing with or incorporating data sources that already
exist externally is a particularly time-sensitive priority for us.
Application developers handle conflating these data sets on a routine
basis. So "completing" these data classes doesn't do much to improve
people's lives other than perhaps making things slightly easier for the
application developer. Yes, a complete map is "good" but that's a different
question from "where should we direct people's limited time and energy to
be the most impactful?"

I'm not poo-pooing the idea of the map driving mappers towards particular
mapping goals, and you're hardly the first to suggest it[1]. We just need a
far more realistic understanding of why we care about completeness and what
it enables that can't be done today. I think we also need to understand
that the patchwork of available government databases means that what's
important is likely different in different places so a one-size-fits-all
approach will be poorly suited. Also, recognize that the team behind
osm-carto may not share these goals, and they aren't beholden to any ideas
that a hypothetical OSMF may have regarding what should be displayed. Thus,
you may have trouble attracting volunteers with sufficient skills and
motivation to produce an alternative map that meets the goals described
here.

[1] https://github.com/gravitystorm/openstreetmap-carto/issues/4723
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/osmf-talk/attachments/20230513/45fb6ec8/attachment.htm>