[OSM-talk] Keep Mapnik relevant

Thomas Wood grand.edgemaster at gmail.com
Thu Jan 8 12:23:26 GMT 2009


2009/1/8 Andreas Fritsche <andreas.fritsche at gmail.com>:
> Hi!
>
>>> I don't get it.
>>
>> Really? It's pretty straightforward.
>
>>[..]
>> Right, so have a look at the following.
>>
>> highway = primary => I have a primary road
>> name = Foo Street => I have a primary road that's called Foo Street
>> ref = 58 => I have a primary road with ref 58 that's called Foo Street
>> abandoned = yes => I don't actually have a primary road at all.
>>
>> Do you see how the last one is completely different? If we start
>> [..]
>
> Actually: No. If interpreted straightforward, your example would end
> with the line
> abandoned = yes => I have an abandoned primary road with ref 58 that
> was called Foo Street
> The tag just adds another property. A client might chose 5px-width and
> red color because it's a primary road. It might not interpret the name
> - because for its application the name doesn't matter and the client
> may decide whether to display it or not because it is abandoned. No
> difference, just another choice.
>
> Nevertheless I do understand the "*actually*"- and espacially the
> "PITA"-part. So I am willing to accept the OSM-mission and join the No
> History Club.
>
> /Andreas

A historical version of OSM using tags probably is possible, as long
as we differentiate the tags so they are not seen as current features.
For example we could just 'namespace' them all to
historical:tagname=value.

It would still be a bit of a PITA to edit the raw data on editors that
do not allow different sets of features to be hidden. (Only really an
issue with potlatch at the moment, since both JOSM and Merkaartor have
their relevant theming options).

-- 
Regards,
Thomas Wood
(Edgemaster)




More information about the talk mailing list