[OSM-talk] Suggestion: fallback tag
Joseph Reeves
iknowjoseph at gmail.com
Thu Dec 17 13:11:24 GMT 2009
>to know that at least *something* will render.
This seems like a confusing proposal to ensure that a tiny fraction of
a percentage of the whitespace on OSM.org gets *something* rendered in
it.
-1 from me.
2009/12/17 Steve Bennett <stevagewp at gmail.com>:
> On Thu, Dec 17, 2009 at 11:55 PM, Shalabh <shalabh.w at gmail.com> wrote:
>>
>> 1. And what if even the 'fallback tag' is not rendered by the renderer?
>
> That's the current situation. So, the worst case scenario, with this tag,
> is...what we have now. Every other case is an improvement.
>
>>
>> 2. Some things may have just no fallback e.g. Mountain passes are not
>> rendered by most renderers. What would be the fallback tag in such a case.
>
> Not sure which tag you're referring to. If the tag is just a bit of
> decoration, maybe you don't need a fallback. But in lots of other cases, it
> would be good to be able to fall back to at least a landuse=industrial, or
> building=yes or something.
>
>>
>> In my opinion, this will only confuse the mappers more when they tag a
>> POI.
>
> Nah. Only the smart mappers would even think of using it.
>
>>
>> A few days down the line, we would have many more questions on what should
>> be the fallback of what and why?
>
> Nah. It doesn't matter enough. I would see it mostly as a personal thing for
> the mapper, to know that at least *something* will render.
>
>> I would think a better idea is to have atleast one renderer and the main
>> OSM map to render a superset of all tags.
>>
>
> There are hundreds of thousands of distinct key=value pairs. And even if
> that was one renderer could do it...how does that help? Say I want Mapnik to
> look nice, but you propose that I use Osmarender instead because it supports
> every single tag.... This isn't a solution at all.
>
> Steve
>
> _______________________________________________
> talk mailing list
> talk at openstreetmap.org
> http://lists.openstreetmap.org/listinfo/talk
>
>
More information about the talk
mailing list