Hello,
Andy Townsend from OSM's Data Working Group here. We were cc:ed on http://lists.openstreetmap.ch/pipermail/talk-ch/2022-April/011470.html and I've picked it up because I've been involved with a couple of related tickets elsewhere. Sorry about the out-of thread message; I've just joined this list to reply.
Mapbox have been creating these "projects" country by country at https://github.com/mapbox/mapping/issues and what's happened in each country has been broadly similar:
* Mapbox say that they will "resolve some issues" in that country, but don't say what they will fix or how they detected them. * The community in that country asks questions and gets only vague handwavy answers (e.g. https://lists.openstreetmap.org/pipermail/talk-ie/2022-April/003016.html ). * The DWG then gets complaints about poor editing in that country and has to try and deal with the mappers concerned (see e.g. https://www.openstreetmap.org/user/del_account/blocks ). * The local community and/or the DWG then has to go through a "bucket and shovel" exercise to tidy up.
It makes no sense for Mapbox to take this approach in every country - it just creates extra work for everyone.
A much better approach would be what (eventually, after much badgering by the local community) happened in the UK. At https://github.com/mapbox/mapping/issues/387#issuecomment-1015316614 Mapbox published their list of "detected problems". What then happened in the UK was that the local communityfixed a large number of the issues (where there were genuine issues), and Mapbox's "mappers" didn't need to do anything - something of a win/win for everyone. I suspect that the Swiss OSM community is even better qualified than the UK one for that approach to work here too.
Looking outside of Switzerland, I'm now worried about where there hasn't been feedback on issues at https://github.com/mapbox/mapping/issues/ , but that's offitopic for this list.
Best Regards,
Andy (from the DWG)
Hello,
Andy Townsend from OSM's Data Working Group here. We were cc:ed on http://lists.openstreetmap.ch/pipermail/talk-ch/2022-April/011470.html and I've picked it up because I've been involved with a couple of related tickets elsewhere. Sorry about the out-of thread message; I've just joined this list to reply.
Mapbox have been creating these "projects" country by country at https://github.com/mapbox/mapping/issues and what's happened in each country has been broadly similar:
* Mapbox say that they will "resolve some issues" in that country, but don't say what they will fix or how they detected them. * The community in that country asks questions and gets only vague handwavy answers (e.g. https://lists.openstreetmap.org/pipermail/talk-ie/2022-April/003016.html ). * The DWG then gets complaints about poor editing in that country and has to try and deal with the mappers concerned (see e.g. https://www.openstreetmap.org/user/del_account/blocks ). * The local community and/or the DWG then has to go through a "bucket and shovel" exercise to tidy up.
It makes no sense for Mapbox to take this approach in every country - it just creates extra work for everyone.
A much better approach would be what (eventually, after much badgering by the local community) happened in the UK. At https://github.com/mapbox/mapping/issues/387#issuecomment-1015316614 Mapbox published their list of "detected problems". What then happened in the UK was that the local communityfixed a large number of the issues (where there were genuine issues), and Mapbox's "mappers" didn't need to do anything - something of a win/win for everyone. I suspect that the Swiss OSM community is even better qualified than the UK one for that approach to work here too.
Looking outside of Switzerland, I'm now worried about where there hasn't been feedback on issues at https://github.com/mapbox/mapping/issues/ , but that's offitopic for this list.
Best Regards,
Andy (from the DWG)
(resending; original mail seems to have gone astray)
I want to chime in and say I am more than happy to put hours into fixing a list of issues if Mapbox can provide them.
Even if no one is putting the time in to fix them from the community (a very far hypothetical situation from what would actually happen), Mapbox should publish the list of issues, or at the very least the list of categories with some examples.
If the goal is good map data with minimal errors, then every reason to collaborate toward that end, no need to have competing objectives.
On Fri, Apr 15, 2022, 15:39 Andy Townsend ajt1047@gmail.com wrote:
Hello,
Andy Townsend from OSM's Data Working Group here. We were cc:ed on http://lists.openstreetmap.ch/pipermail/talk-ch/2022-April/011470.html and I've picked it up because I've been involved with a couple of related tickets elsewhere. Sorry about the out-of thread message; I've just joined this list to reply.
Mapbox have been creating these "projects" country by country at https://github.com/mapbox/mapping/issues and what's happened in each country has been broadly similar:
- Mapbox say that they will "resolve some issues" in that country, but
don't say what they will fix or how they detected them.
- The community in that country asks questions and gets only vague
handwavy answers (e.g. https://lists.openstreetmap.org/pipermail/talk-ie/2022-April/003016.html ).
- The DWG then gets complaints about poor editing in that country and
has to try and deal with the mappers concerned (see e.g. https://www.openstreetmap.org/user/del_account/blocks ).
- The local community and/or the DWG then has to go through a "bucket
and shovel" exercise to tidy up.
It makes no sense for Mapbox to take this approach in every country - it just creates extra work for everyone.
A much better approach would be what (eventually, after much badgering by the local community) happened in the UK. At https://github.com/mapbox/mapping/issues/387#issuecomment-1015316614 Mapbox published their list of "detected problems". What then happened in the UK was that the local communityfixed a large number of the issues (where there were genuine issues), and Mapbox's "mappers" didn't need to do anything - something of a win/win for everyone. I suspect that the Swiss OSM community is even better qualified than the UK one for that approach to work here too.
Looking outside of Switzerland, I'm now worried about where there hasn't been feedback on issues at https://github.com/mapbox/mapping/issues/ , but that's offitopic for this list.
Best Regards,
Andy (from the DWG)
(resending; original mail seems to have gone astray) _______________________________________________ talk-ch mailing list talk-ch@openstreetmap.ch http://lists.openstreetmap.ch/mailman/listinfo/talk-ch
For those wondering about examples, the wiki page linked in the initial e-mail contained a further link to this page https://wiki.openstreetmap.org/wiki/Mapbox_linters_types_description which describes in detail the kind of issues Mapbox is planning to fix.
On 2022-04-17 16:31, Christopher Beddow wrote:
I want to chime in and say I am more than happy to put hours into fixing a list of issues if Mapbox can provide them.
Even if no one is putting the time in to fix them from the community (a very far hypothetical situation from what would actually happen), Mapbox should publish the list of issues, or at the very least the list of categories with some examples.
If the goal is good map data with minimal errors, then every reason to collaborate toward that end, no need to have competing objectives.
On Fri, Apr 15, 2022, 15:39 Andy Townsend ajt1047@gmail.com wrote:
Hello, Andy Townsend from OSM's Data Working Group here. We were cc:ed on http://lists.openstreetmap.ch/pipermail/talk-ch/2022-April/011470.html and I've picked it up because I've been involved with a couple of related tickets elsewhere. Sorry about the out-of thread message; I've just joined this list to reply. Mapbox have been creating these "projects" country by country at https://github.com/mapbox/mapping/issues and what's happened in each country has been broadly similar: * Mapbox say that they will "resolve some issues" in that country, but don't say what they will fix or how they detected them. * The community in that country asks questions and gets only vague handwavy answers (e.g. https://lists.openstreetmap.org/pipermail/talk-ie/2022-April/003016.html ). * The DWG then gets complaints about poor editing in that country and has to try and deal with the mappers concerned (see e.g. https://www.openstreetmap.org/user/del_account/blocks ). * The local community and/or the DWG then has to go through a "bucket and shovel" exercise to tidy up. It makes no sense for Mapbox to take this approach in every country - it just creates extra work for everyone. A much better approach would be what (eventually, after much badgering by the local community) happened in the UK. At https://github.com/mapbox/mapping/issues/387#issuecomment-1015316614 Mapbox published their list of "detected problems". What then happened in the UK was that the local communityfixed a large number of the issues (where there were genuine issues), and Mapbox's "mappers" didn't need to do anything - something of a win/win for everyone. I suspect that the Swiss OSM community is even better qualified than the UK one for that approach to work here too. Looking outside of Switzerland, I'm now worried about where there hasn't been feedback on issues at https://github.com/mapbox/mapping/issues/ , but that's offitopic for this list. Best Regards, Andy (from the DWG) (resending; original mail seems to have gone astray) _______________________________________________ talk-ch mailing list talk-ch@openstreetmap.ch http://lists.openstreetmap.ch/mailman/listinfo/talk-ch
talk-ch mailing list talk-ch@openstreetmap.ch http://lists.openstreetmap.ch/mailman/listinfo/talk-ch
On 18/04/2022 14:11, Timeo Gut wrote:
For those wondering about examples, the wiki page linked in the initial e-mail contained a further link to this page https://wiki.openstreetmap.org/wiki/Mapbox_linters_types_description which describes in detail the kind of issues Mapbox is planning to fix.
Also, as mentioned at https://github.com/mapbox/mapping/issues/387#issuecomment-1015316614 , the json files at https://github.com/mapbox/mapping/issues/407#issue-1204668361 can be unzipped and dropped on a browser pointing at geojson.io . That way you can see the "actual detected problems" (and see if they really are problems).
Best Regards,
Andy
On 2022-04-17 16:31, Christopher Beddow wrote:
I want to chime in and say I am more than happy to put hours into fixing a list of issues if Mapbox can provide them.
Even if no one is putting the time in to fix them from the community (a very far hypothetical situation from what would actually happen), Mapbox should publish the list of issues, or at the very least the list of categories with some examples.
If the goal is good map data with minimal errors, then every reason to collaborate toward that end, no need to have competing objectives.
On Fri, Apr 15, 2022, 15:39 Andy Townsend ajt1047@gmail.com wrote:
Hello, Andy Townsend from OSM's Data Working Group here. We were cc:ed on http://lists.openstreetmap.ch/pipermail/talk-ch/2022-April/011470.html and I've picked it up because I've been involved with a couple of related tickets elsewhere. Sorry about the out-of thread message; I've just joined this list to reply. Mapbox have been creating these "projects" country by country at https://github.com/mapbox/mapping/issues and what's happened in each country has been broadly similar: * Mapbox say that they will "resolve some issues" in that country, but don't say what they will fix or how they detected them. * The community in that country asks questions and gets only vague handwavy answers (e.g. https://lists.openstreetmap.org/pipermail/talk-ie/2022-April/003016.html ). * The DWG then gets complaints about poor editing in that country and has to try and deal with the mappers concerned (see e.g. https://www.openstreetmap.org/user/del_account/blocks ). * The local community and/or the DWG then has to go through a "bucket and shovel" exercise to tidy up. It makes no sense for Mapbox to take this approach in every country - it just creates extra work for everyone. A much better approach would be what (eventually, after much badgering by the local community) happened in the UK. At https://github.com/mapbox/mapping/issues/387#issuecomment-1015316614 Mapbox published their list of "detected problems". What then happened in the UK was that the local communityfixed a large number of the issues (where there were genuine issues), and Mapbox's "mappers" didn't need to do anything - something of a win/win for everyone. I suspect that the Swiss OSM community is even better qualified than the UK one for that approach to work here too. Looking outside of Switzerland, I'm now worried about where there hasn't been feedback on issues at https://github.com/mapbox/mapping/issues/ , but that's offitopic for this list. Best Regards, Andy (from the DWG) (resending; original mail seems to have gone astray) _______________________________________________ talk-ch mailing list talk-ch@openstreetmap.ch http://lists.openstreetmap.ch/mailman/listinfo/talk-ch
talk-ch mailing list talk-ch@openstreetmap.ch http://lists.openstreetmap.ch/mailman/listinfo/talk-ch
talk-ch mailing list talk-ch@openstreetmap.ch http://lists.openstreetmap.ch/mailman/listinfo/talk-ch
Hoi Zäme
In meinem "Interessebereich" habe ich die Mapboxpendenzen soweit wie möglich abgearbeitet. Es sind einzelne Strasseninseln in Waldgebieten überiggeblieben, welche man schlicht vor Ort mal anschauen muss, wo diese nun wirklich an eine weitere Strasse angehängt sind. Ein Paar "idiotische Strassenkreuzungen" sind Autobahnbrücken über Strassen.
Danke an Simon, auf dein aufmerksames News-lesen
Viele Grüsse
Hubert
Am 18.04.22 um 15:25 schrieb Andy Townsend:
On 18/04/2022 14:11, Timeo Gut wrote:
For those wondering about examples, the wiki page linked in the initial e-mail contained a further link to this page https://wiki.openstreetmap.org/wiki/Mapbox_linters_types_description which describes in detail the kind of issues Mapbox is planning to fix.
Also, as mentioned at https://github.com/mapbox/mapping/issues/387#issuecomment-1015316614 , the json files at https://github.com/mapbox/mapping/issues/407#issue-1204668361 can be unzipped and dropped on a browser pointing at geojson.io . That way you can see the "actual detected problems" (and see if they really are problems).
Best Regards,
Andy
On 2022-04-17 16:31, Christopher Beddow wrote:
I want to chime in and say I am more than happy to put hours into fixing a list of issues if Mapbox can provide them.
Even if no one is putting the time in to fix them from the community (a very far hypothetical situation from what would actually happen), Mapbox should publish the list of issues, or at the very least the list of categories with some examples.
If the goal is good map data with minimal errors, then every reason to collaborate toward that end, no need to have competing objectives.
On Fri, Apr 15, 2022, 15:39 Andy Townsend ajt1047@gmail.com wrote:
Hello, Andy Townsend from OSM's Data Working Group here. We were cc:ed on http://lists.openstreetmap.ch/pipermail/talk-ch/2022-April/011470.html and I've picked it up because I've been involved with a couple of related tickets elsewhere. Sorry about the out-of thread message; I've just joined this list to reply. Mapbox have been creating these "projects" country by country at https://github.com/mapbox/mapping/issues and what's happened in each country has been broadly similar: * Mapbox say that they will "resolve some issues" in that country, but don't say what they will fix or how they detected them. * The community in that country asks questions and gets only vague handwavy answers (e.g. https://lists.openstreetmap.org/pipermail/talk-ie/2022-April/003016.html ). * The DWG then gets complaints about poor editing in that country and has to try and deal with the mappers concerned (see e.g. https://www.openstreetmap.org/user/del_account/blocks ). * The local community and/or the DWG then has to go through a "bucket and shovel" exercise to tidy up. It makes no sense for Mapbox to take this approach in every country - it just creates extra work for everyone. A much better approach would be what (eventually, after much badgering by the local community) happened in the UK. At https://github.com/mapbox/mapping/issues/387#issuecomment-1015316614 Mapbox published their list of "detected problems". What then happened in the UK was that the local communityfixed a large number of the issues (where there were genuine issues), and Mapbox's "mappers" didn't need to do anything - something of a win/win for everyone. I suspect that the Swiss OSM community is even better qualified than the UK one for that approach to work here too. Looking outside of Switzerland, I'm now worried about where there hasn't been feedback on issues at https://github.com/mapbox/mapping/issues/ , but that's offitopic for this list. Best Regards, Andy (from the DWG) (resending; original mail seems to have gone astray) _______________________________________________ talk-ch mailing list talk-ch@openstreetmap.ch http://lists.openstreetmap.ch/mailman/listinfo/talk-ch
talk-ch mailing list talk-ch@openstreetmap.ch http://lists.openstreetmap.ch/mailman/listinfo/talk-ch
talk-ch mailing list talk-ch@openstreetmap.ch http://lists.openstreetmap.ch/mailman/listinfo/talk-ch
talk-ch mailing list talk-ch@openstreetmap.ch http://lists.openstreetmap.ch/mailman/listinfo/talk-ch
-
Andy Townsend -
Christopher Beddow -
Hubert Rüttimann -
Timeo Gut