[Talk-in] Classifying places - cities, towns and villages

I Chengappa imchengappa at gmail.com
Thu May 17 08:35:49 BST 2012


>
> City: Any UA with more than 100,000 population
> Town: 10,000 - 100,000 people
> Village: Below 10,000
>

Yes, the above is what I would use, because it is the generic guideline and
hence it is most likely to correspond to the rest of the world.  This is an
important point; India is a country and can be treated in isolation, but
there are also other countries and it is useful to have the map cover a
wider area and still mean the same thing. When looking at a border area,
say India and Nepal, it would be useful not to have the same tag mean
different things on different sides of the border. Similarly, it will be
useful to be able to use the Cloudmade downloads with the knowledge that
things have the same meaning wherever you are. And most users will not be
especially interested in the formal government designations unless they
become widespread in normal use (and as with names it is common in India to
ignore such formal names in favour of past practice e.g. Bangalore /
Bengaluru).

Again, to go back to the US as an example, there any 'city' that is
actually a small town of 10k population is tagged in the OSM as place=town,
because users understand that and it allows the same rules files to be used
as the rest of the world. There is nothing to stop them also adding the
government classification and then adaptng the osmarender rules files so as
to show towns  with formal 'city' tags for an administrative map .
Meanwhile the base map still retains its comparability with other
countries, and the straightforward renderings function as now for the more
common uses such as a driving map.

Hamlet: ?
>

The common English usage of hamlet is of course pretty much that of a tiny
village; the OSM wiki says typically 100 people. At this size in the India
map there are few such even marked. This may be considered to leave a gap
to village, but it seems to me that village suffices for most small
settlements. I've just seen that in the US parts of the OSM, place=hamlet
is sometimes used instead of place=suburb.

Suburb: ?
>

It has a useful meaning as a named and identifiable village sized part of a
larger urban area.

As to coming up with a specific proposal, changing the existing default
definitions was not my idea (and I will anyway be busy with real life for
some weeks).  I am not arguing against the classification of
http://wiki.openstreetmap.org/wiki/India:Places but only against trying to
shoehorn it into the existing tags/values.  Ideally  a new namespace prefix
would be agreed for the new scheme; this would also allow for further
differing policies to be recorded, e.g. if a state had a classification
that was different from the central government's. I've already suggested
census2011:town as one possible prefix but others based on the policy might
be more suitable e.g. mdds:town.

Thanks, indigomc

On 14 May 2012 20:55, Arun Ganesh <arun.planemad at gmail.com> wrote:

>
> On Sun, May 13, 2012 at 2:08 PM, I Chengappa <imchengappa at gmail.com>wrote:
>>
>> Can you point me to the definition of a city in the MDDS designations
>> because I cannot find it?
>
> This is what the http://wiki.openstreetmap.org/wiki/India:Places proposal
> is about. It is the local body type > place= value mapping that we need to
> discuss and decide. And that will solve any further confusion.
>
>
>> Can we agree that if any such designation is changed from an existing
>> one, that the justification should be included. Again I would ask that you
>> consider using a tag. Census2011;town or similar, which would deal with all
>> such issues without a problem.
>>
> I still do not see how using a census tag will solve the primary problem
> of which place= tag to use. If I understand correctly this is the
> guidelines you want to use?
> City: Any UA with more than 100,000 population
> Town: 10,000 - 100,000 people
> Village: Below 10,000
> Hamlet: ?
> Suburb: ?
>
> If you have proper definitions for how all the tags should be used for
> India please make it clear. There is no point debating over an incomplete
> tagging proposal.
>
>>
>> Also, I see that Mandya has a new tag/value of admin_level=5?
>>
> This is for indicating that it is an district headquarter.
>
>
>>
>> Thanks, indigomc
>>
>>
>> On 11 May 2012 06:22, Arun Ganesh <arun.planemad at gmail.com> wrote:
>>
>>> Indigomc, let me recap, because we are going in circles:
>>>
>>>
>>> What should not be done is to divert a generally understood global
>>>> designation such as 'city' that already has the neutral population
>>>> definition.
>>>
>>>
>>> The so called global definition of a city on the osm wiki has been
>>> arbitrarily defined and has no basis or credibility.
>>> http://wiki.openstreetmap.org/wiki/Tag:place%3Dcity
>>>
>>>
>>> Especially to change it to correspond to a mixture of regional
>>>> government designations that may change.  Do people in Mandya believe that
>>>> Cambridge is more important than their 'town'?
>>>
>>>
>>> There will arise cases in future where a person from a village, may want
>>> his place to be marked as a town or a city because he feels its important.
>>> It is therefore important to have a clear definition which links official
>>> classification of the place to osm tags and is not ambiguous.
>>>
>>>
>>>> As to the suggestion that the government has an 'official system',
>>>> please document where this policy is stated and how it is being implemented.
>>>
>>>
>>> The census 2011 has an official and exhaustive list of populated places
>>> with the classification of the local governing body. It has also
>>> established a system of unique e-governance code for every place
>>> (urban+rural) which is now the standard for future e governance services.
>>> This is the best reference document available to date which records the
>>> nature and size of every place in the country.
>>>  http://censusindia.gov.in/2011census/MDDS_Directory.aspx
>>> http://en.wikipedia.org/wiki/MDDS
>>>
>>>
>>>>  A standard used by the census and not used by the state governments or
>>>> in general usage is just that, of use for the census and nothing else.
>>>>
>>>> The census is a collection of recorded facts, if it says Pune is a
>>> municipal corporation, then it is a municipal corporation. It is a central
>>> document which plays a vital role in the planning of essential services for
>>> the next ten years. I would rather have places classified according to the
>>> census, than an arbitrary definition of a mapper sitting in Europe who
>>> thinks that 100,000 is a good number for a place to be called a city.
>>>
>>>
>>> --
>>> j.mp/ArunGanesh
>>>
>>> _______________________________________________
>>> Talk-in mailing list
>>> Talk-in at openstreetmap.org
>>> http://lists.openstreetmap.org/listinfo/talk-in
>>>
>>>
>>
>> _______________________________________________
>> Talk-in mailing list
>> Talk-in at openstreetmap.org
>> http://lists.openstreetmap.org/listinfo/talk-in
>>
>>
>
>
> --
> j.mp/ArunGanesh <http://j.mp/ArunGanesh>
>
> _______________________________________________
> Talk-in mailing list
> Talk-in at openstreetmap.org
> http://lists.openstreetmap.org/listinfo/talk-in
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-in/attachments/20120517/63a3c490/attachment.html>


More information about the Talk-in mailing list