I would add - the objective problems. In the 20th century, when these
legacy databases were created, there was no GPS, no satellite imagery as
we know it. Officials had to actually walk and use manual tools to map
properties what is certainly prone to human errors. Databases themselves
were unreliable, the data was often got corrupted or partially corrupted
at the drop of a hat, the first versions of MySQL and PostgreSQL
appeared by the end of 90s only.
Nowadays, when the e-commerce requires imagery processing on an
industrial scale, a lot of automatic optical shape & patterns
recognition software is being developed, because a product has to be
extracted from a background to get a "flying on air" style. But in the
product photography a photo-studio with the controlled conditions could
be used, besides there are billions of new products each year, and this
process will never end.
It is quite different with the Earth surface. We've got a finite amount
of land to map. There will be no new land (neither on Earth nor on other
planets, at least not in a foreseeable future). Besides new technologies
arrive, like better satellite imagery quality, more GPS-style systems,
affordable industrial grade RPAS, faster databases, better editing
tools, better displays, etc.
There is place for legacy maps. I myself like to view historical maps,
but I am not sure that it makes sense to import massively this
historical data into the main OSM map.
Best regards,
Oleksiy
On 8/16/2017 1:44 PM, john whelan wrote:
...There have been problems in the past with the data quality of
imports ...
Cheerio John
_______________________________________________
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk
