> On Jun 21, 2018, at 6:16 AM, Christoph Hormann <o...@imagico.de> wrote:
> 
> Then why do you object to Frederik's idea of separating the tagging 
> presets from editor development and give up control over the decisions? 

I offered to do exactly this a few years ago:
https://github.com/osmlab/editor-presets/pull/2 
<https://github.com/osmlab/editor-presets/pull/2>
Nobody cared, so it sat and eventually went stale over a few months.

I wouldn’t do it today.  

Looking back on the iD changelogs 
<https://github.com/openstreetmap/iD/blob/master/CHANGELOG.md> (and scanning 
through the still open issues) there are lots of usability items that affect 
the preset system.
Some recent examples of things I couldn’t do (quickly) if the presets were an 
external project:

* We changed all the icons to support multiple iconsets 
<https://github.com/openstreetmap/iD/issues/3025>  and this meant that I needed 
to change all the icon names in all the presets.
* We implemented a check to support min and max field values 
<https://github.com/openstreetmap/iD/issues/5043>, and it meant that I needed 
to go through all the presets and add some properties to certain ones.
* We renamed the field label from “Phone” to “Telephone” 
<https://github.com/openstreetmap/iD/issues/5019> so that a user can type 
either value in the Add field dropdown.

Usability and speed of development are very important to me, and these are 
things that would suffer if I split the presets off into a separate project.


> I am glad you work on improving possibilities for choice of presets and 
> this could over time be used to allow alternatives - like converting 
> the JOSM presets (which already includes a lot of specialized add on 
> preset collections) or managing diverse independent preset collections.  
> This is IMO the best way to go ahead here.

Yes allowing people to override the presets at runtime is something I want to 
add.  (I almost built it at the hackathon.)

For the curious, the somewhat tricky part about this is the bootstrap process.  
Currently, presets in iD are ready at startup and translations are loaded in 
later.  If a user specifies a replacement preset file, we need to delay some 
code until the browser has fetched everything.


Just going to cut my reply off here.  There was some vague stuff in your 
message about “poisoning OSM” that didn’t seem to be a serious question.  Also 
for the record, I’m fine with either method of mapping runways (I prefer to map 
them as a 2 node way myself).  iD supports both methods (drawn as a linear way 
or as a closed way area).  

If you don’t use iD, I encourage you to try it out..
Please report any specific bugs for feature requests on our issue tracker:  
https://github.com/openstreetmap/iD <https://github.com/openstreetmap/iD>  


Thanks, Bryan

_______________________________________________
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Reply via email to