[Imports] Worldwide fuel stations import, 59k objects

[Frederik Ramm]

Hi,

   the following might be a bit confusing to the casual reader because
I'm arguing in favour of Ilya's import even though I'd really prefer us
to have less imports, not more. But I think Ilya's proposal is being
criticised for the wrong reasons.

On 08.03.2018 17:07, Martin Koppenhoefer wrote:
> 2018-03-08 15:03 GMT+01:00 Ilya Zverev <ilya at zverev.info
> <mailto:ilya at zverev.info>>:
> 
>     .... Import Guidelines points to make me do tons of extra community
>     work, like finding communication channels for ~20 countries and
>     convincing every single mapper on these, in their local languages.
> 
> maybe we can improve the documentation for the primary comunity
> comunication channels per country, on a single (hopefully up to date) page?
> 
> I think it is generally a good provision of the guidelines to ask the
> local community for comments.

I think there's certainly a practical limit to diversity, especially for
a small project/organisation like ours.

If you don't speak English, then there are areas of OpenStreetMap where
it is difficult for you to make your voice heard. You cannot join the
OSMF board if you don't speak English; and even for passing a meaningful
vote on who should be on the board, you will have to rely on trusted
third parties to explain to you what the candidates stand for. You will
have difficulty participating in most of the international working
groups, in tagging discussions, or international OSM software
development. That's a fact and it would require a very large number of
volunteers or a very big amount of money to change that. Requesting that
everything that could possibly affect the OSM community in a country is
also accessible to them when they don't speak English might sound great
but it would definitely make most of the project grind to a halt.

So even when we talk about desirable diversity goals, we have to remain
practical; the goal of "a working OSM" is more important than the goal
of "a diverse OSM" and diversity must take the back seat when it would
make working in the project impossible.

We generally request that imports and automated edits are discussed
before they are executed. The main reason for this is that we want to
have a chance to discover flaws in the process, the licensing, or clean
up misunderstandings. The import guidelines also say that "community
buy-in" should be sought. Now if someone ran an import in Panama, then
it would be a good idea to discuss this with the Panama community, and
out of courtesy do in in Spanish. Ideally, the importer would be from
Panama.

But I can feel Ilya's exasperation at the suggestion of discussing a
world-wide import on every local mailing list / forum / facebook group,
in the language appropriate for each. This is not practical, and would
kill the import in its tracks if we were to demand that.

Five or ten years in the future, when we have established local chapters
around the globe, maybe then we'll have a mechanism to submit an import
proposal and then have each country say yes or no after consulting with
their folks in their language, but we're not there yet.

My practical suggestion would be to:

* discuss the import here, in English;
* make a list of countries most affected by the import (perhaps all with
more than 1000 edits - choose a practical threshold) and if they have a
talk-xx mailing list, make a posting there ("I am planning to import
1000 fuel stations in your country, discussion over on imports@")
* if there's significant opposition overall, scrap the import altogether
* else, split up the import by country, and run the import in all
countries potentially excluding those where people have objected
* if, after the import, voices pop up complaining because they were
unaware of the discussion and the import somehow breaks something in
their country - simply revert for that particular country

Bye
Frederik

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