🚀 go-pugleaf

In article <3ajl12$q6f@sun.lclark.edu>,
John Miller <miller@sun.lclark.edu> wrote:
>
>SHEN> A mathematic coordinate system is the most efficient way to express a
>point and to search the point on a surface, that has been promoting all the
>development of whole modern science and technology.  Don't argue!  It has
>been the proved fact.
>
>ME> On a SURFACE, o.k., but the proof does not apply to a PLANET.
>Shen claims that his system is low cost, but then goes on to explain that
>it really needs this extensive computer database at every sorting station
>that knows all about the surrounding terrain, etc:

I am really puzzled why you always mention some planets.  Are there such demandings
now or in the forseeable future?  If you you want to include this kind of demandings,
then you have to consider how to include the moving of Autralia continent. You
have to include the function that a geocode system can automatically adapt to the
new communities created and old communities destroyed by war etc.  Finally you
will not reach any conclusions.  Everything once formed, is not perfect.

Including the database for sorting machine can not be said that it needs much 
cost.  In fact, an international sorting database can be copied for all post
sorting machines.

>
>SHEN> A computer sorting machine should include lots of information, such as
>the local geographical information, road and street distribution, and post
>offices locations, etc, which will determine the optimized transport and
>delivery route.  For sorting international mails, a sortng program will take
>into account all the information about the countries borders, main airports,
>airlines, and some post agreements between countries, etc;
>
>ME> This is a lot of information to maintain and use.  Other systems could
>be developed that do not depend on augmentation by detailed information to
>get the mail delivered AND to route emergency vehicles, etc.

Now, the pc can already deal with databases of hundreds magabites efficiently.
We can estimate the size of the database for the sorting machine.  As I know
that the main contribution of the data is from the borders of countries.  The 
world has less than 100 million km borders.  If you use polynomials to represent
them, you need to save the coordinates of one point in each one hundred km in
average, i.e., you need about 16 magabyte memory to save the information.  It is
not a significant database.  You do not need to worry aboout it.

If you have better systems, why don't you post it?  If you do not have any idea
yet, please not be bothered by the unsure future, just as you need a car, you
would better pick up one of the exisiting cars but not wait for an unsure perfect 
car to come.
>
>What would a NAC map really look like?  Why would a single scale map not be
>completely covered with NAC codes?  In order to be useful without the common
>address, a map would need to have every NAC code for every 50x25 m parcel.

A NAC map is exactly the existing maps simply with NAC grids. A small scale 
map can tell you exact 8 character code, for example a 1:10,000 scale map.
I never say that a NAC will replace the common address but as an complementary
information together with the common address which help people and computer
to locate an address.


>Without all digits of the code (8 for every parcel) I could only get to the
>general area.  Then I would need either another scale map to get to the
>actual house unless there are NAC signs posted virtually every where and I
>had some sense of how to navigate NAC signs!

You still have to go through streets and roads with common names.

>
>You couldn't locate places in the west hills of my city unless you had
>complete resolution down to each house because of the way streets go through
>or do not go through.  As explained above, I don't think that it is
>practical to print such detailed maps to support the use of NAC.
>
A NAC map does not necessary to be 1:10,000 scale.  It can be in any scale
depending on the needs.

>Such a set of maps for my own city would be quite thick, and would have more
>information than practically needed in order to explicitly locate every
>single home or business.  I don't think I should need a satellite or a
>portable computer tapped into a data base somewhere in order to locate an
>address via NAC code.

A 1:10,000 map can be a single piece of paper with a size about 1x2 meters.
I never ask people to use any equipment to locate an address.  A NAC can
be located by a GPS but not necessary.  People simply use the information
of a NAC to find an address on a map, then follw the streets to reach it.


>
>Further, houses, buildings and roads are not built on exact NAC points. I
>can't see this working for the various size parcels and the overlap
>situations that might come up.  Who gets what NAC Code!!?!?!

I never tell that you should follow the NAC grid lines which usually are
not exactly on the streets.  A NAC simply help you to find your friend's
home easily, let emergent agents to locate needed houses easily and let
the post corporations to send your mails faster.



>
>Shen seems to think that NAC codes will help locate folks in need of rescue
>in the rural areas.  Unlikely.  Let's say that I am out in the country and
>just went off the road.  I have a cell phone with me.  What do I do?  Walk
>to the nearest NAC sign and call it in?  If we have NAC signs every 25 or 50
>meters -- no problem.. or is it?  Do I not understand where people would
>come up with NAC codes out of thin air.

If you were really off the road, maybe it was the best way to find a nearest
NAC and told it to some agents that would help you.  Don't always go to the
planets or thin air where you would be in problem all the time.

>
>This proposal looks as if it were produced by a second-year computer science
>student who has learned about compression and quadrature.  Just because
>Latitude and Longitude can be compressed into 8 or so digits does not mean
>that we must try to use them to locate places on a fractal like surface --
>too much overhead in translating from code to the terrain.  Other naming
>systems can deal with the real nature of the earth -- millions of wonderful
>nooks and crannies, homes down dead-end lanes, places in the country.  NAC
>codes are a technical nightmare, and they have no chance of being used for
>the purposes stated.  It sounds useful but it really isn't.
>
>John
>-- 
>"What's ahead?"
>

I reaaly can not see why you want to curse the proposal.  Does it hurt you
or blame you?  You try to say that I am a second year student.  If you
were doctors or professors, why don't you present your proposal that
really looks like from an expert?

I hope that you will not be hurt this time simply because I explain my
idea.  I do not expect your blames either.  If you want, please present
your reasonings.  Thanks.


Xinhang