[Imports] Imports Review

[Jo]

I'll add my €0.02. The imports I like best, are the ones where we figure
out a way to provide local users with the data, such that it becomes one of
many sources they can use to improve the map. For those imports it doesn't
make sense to require the use of a separate import account, though, as that
data will be integrated right away with surveyed data, preexisting data
imagery and Mapillary. This works well in places where there is an active
community.

The import I'm involved in at the moment is the one organised through the
tasking manager to import schools across Uganda. We did find locals to
perform this import, but even then, I don't think this mattered much in the
end. They live in Uganda, but they also don't have local knowledge all
across Uganda.

I'm validating those tiles and I do find a fair share of spelling errors to
correct, sometimes positions can be improved upon. There are times when
they behave like robots too, adding schools 2 or even 3 times, or not
merging them with already existing data pre import. It goes without saying,
I always solve those problems before validating a task. It's a learning
process for them as well.

In general, I think the result of this import will be good data, that can
be used to build further upon. (The data contains 5 levels of
administrative division of the country, I created a MapCSS style to
visualize this at varying zoom levels)

Now I'm looking at making a proposal for importing schools across Tanzania.
Jumping through the required hoops seems a daunting task in itself. I'm not
looking forward to it, so I didn't do it yet. Too many other stuff on my
plate.

At the moment I'm mostly exploring (and tidying) the data. The data is less
'rich' than the one for Uganda. Only place names are part of the names,
never the affiliation to organisations, like churches.

I've been involved with other imports as well. All bus and tram stops in
Belgium, for example. That's not a mere data dump either. It took a lot of
work to get it right. I do remain convinced it makes OpenStreetMap better
to have this data. I'm using it myself, when I render maps where I want to
combine public transport with hiking node networks, for example.

Polyglot

2016-03-15 18:51 GMT+01:00 Christoph Hormann <chris_hormann at gmx.de>:

> On Tuesday 15 March 2016, Frederik Ramm wrote:
> >
> > Also, proponents of an import often say things like "this is going to
> > encourage community participation in the local area" etc., but does
> > it? Nobody ever bothers to check.
> >
> > It would be great if we had a process by which we would review
> > imports *after* the fact and perhaps even a few years later, to judge
> > what went well and what didn't, to learn from mistakes, or in extreme
> > cases perhaps even revert the whole thing when it turns out that it
> > didn't work out or that promises made during the initial discussion
> > were not kept.
>
> Some time ago i had a look at the current import guidelines to see what
> could be improved for the purpose of ensuring better results.  My
> motivation was that while most mappers think that imports can have
> value most also agree that the current import procedure is not really
> working that well (see for example the statements in that direction in
> the last OSMF board election Q&A).
>
> Ultimately i did not get very far in that direction because it quickly
> occured to me that effective changes in the import procedure would
> mostly mean adding bureaucracy and even if done in a well-meaning way
> this would probably primarily mean making life more difficult for
> responsible mappers and it would not significantly limit the less
> responsible ones who do not care much about the project and just want
> to dump some data in the OSM database for some selfish or misguided
> reason and who tend to follow rules as necessary by the letter but not
> in their spirit.
>
> Among the ideas i contemplated were:
>
> - the requirement of an obligatory post-import review which would then
> be discussed by the community just like the import plan before and
> which could include requiring the import organizer to fix things that
> did not go acconding to the pre-announced plan.
> - the requirement for each import to have an experienced mapper as
> mentor who supervises the whole process while not being involved in the
> import itself.
>
> My reasoning with these was that for something like this to work in
> reality it needs to be the responsibility of the import manager to
> organize review and supervision work that is considered necessary
> because otherwise it will likely not happen.
>
> --
> Christoph Hormann
> http://www.imagico.de/
>
> _______________________________________________
> Imports mailing list
> Imports at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/imports
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/imports/attachments/20160315/2d1960da/attachment.html>