[Talk-us] Do City Labels look funny to you?

[Richard Weait]

On Tue, Oct 19, 2010 at 6:00 PM, Ian Dees <ian.dees at gmail.com> wrote:

> On the contrary: I say all these critiques indicate the need for a

> - More granular city name placement based closer on population than on the
> place tag

Place looks broken in some parts of US because some mappers choose to
ignore existing practice for the basing the place tag on population.

One reason that OSM has used place=city/town/village/hamlet as
shorthand for population, if i remember the conversation correctly
from 3-4 years ago, is that the UK Ordnance Survey claimed copyright
on the population on signs in UK.  So we couldn't just plunk that data
into OSM.  What remained, that a city is a city, and larger than a
town, was what was available.  So we agreed on arbitrary, and
imperfect, guidelines of 1,000, 10,000, 100,000 to divide them.  And
that we would use that relative scale as a good guideline for
prominence.

US doesn't have the same restriction on the use of population data.
So population should be entered from good data where we have it.  No
question there.  You want to render prominence based on population,
absolutely, do so.

That US also has "legal incorporation status" which uses some of the
same values the OSM uses for place is just a "false friend".  Legal
incorporation status should be accurately recorded in a legal
incorporation status tag.  Misusing place= because the values match
leg_inc_stat= is just silly.

Just because building=yes and oneway=yes are well used tags does not
mean that all buildings are oneway.