Order Tray | Contact Us | Home | SIG Lists

Position Ambiguity [was] Re: [aprssig] findu.com....

Bill Diaz william.diaz at comcast.net
Mon Jun 12 02:57:43 UTC 2006


Jared,
  See below:

>-----Original Message-----
>From: aprssig-bounces at lists.tapr.org 
>[mailto:aprssig-bounces at lists.tapr.org] On Behalf Of KC2MMI (Jared)
>Sent: Sunday, June 11, 2006 17:44
>To: aprssig at lists.tapr.org
>Subject: RE: Position Ambiguity [was] Re: [aprssig] findu.com....
>

>Bill, are you being disingenious, or did you not understand 
>the context of James
>complaint?

Your editing of my response to your post omits all of YOUR comments.  Lets
keep this thread in context.  You are criticizing James for using US Census
Bureau Tiger files.  You feel they are inaccurate and James / APRS World
should include a disclaimer every time APRS World supplies a map to a user.
  
>"Will you do the same for every position you supply to the 
>APRS system?  Just include a note with every beacon indicating 
>that your position may be inaccurate.  Wouldn't be necessary if 
>you could certify the accuracy of your
>position in every packet."

I am just asking you to do the same thing you are demanding of James and
APRS World.

>  The position generated from a GPS is known to be accurate 
>within the norms of the GPS system, typically 3-10 meters all things being 
>considered. 

FYI there is NO guarantee that a GPS will provide 3-10 meter accuracy.  It
can if it is WAAS capable and is receiving the WAAS differential
corrections.  Otherwise the accuracy can be as bad as 25 meters from your
actual position.  Perhaps you should take the time to learn about GPS, its
capabilities, and accuracy before commenting in public.

I said nothing about a GPS in my post.  I asked you to include a note
indicating YOUR POSITION may be inaccurate.  Based on your comments above,
it appears that you may not be sure what the accuracy of positions you are
supplying in APRS packets.  Difficult to estimate the amount of error in a
map when you don't know how much error you have in your own measurements.

>The problem is, APRSWorld, and any software using the Tiger maps of the 
>NavTeq maps (or several others as a base) is displaying those geographic 
>coordinates INCORRECTLY on the street maps they use.

Interesting.  Exactly how are they displaying the Tiger map data
incorrectly?  Are they modifying the supplied data to introduce the errors
you are claiming?  If so, please explain exactly how James is introducing
the inaccuracy you are claiming.  I believe he is faithfully reproducing the
data supplied by the Census Bureau.  Same for WinAPRS, APRS-SCS, and Xastir
Tiger maps.  

> As GPS users, it is easy for us to know the accuracy inherent 
>to the system, therefor when we send a position that is designated as a GPS

>position, using the designated default datum, we know the accuracy inherent
to 
>that position.

You have demonstrated above, you do not know the what accuracy GPS receivers
are capable of.  Same for most casual APRS users.  

> I'd be glad to send EPE but as you should certainly know, APRS does 
>not provide for sending that. APRS and GPS both are fixed and well known 
>constraints. The errors from bad data in mapping systems are not, 
>they are unknowns to the ordinary user.

Can you correlate an EPE value to the amount of error in Degrees, Minutes,
Seconds.  No you can't since Garmin has never released the proprietary
alogrithym they use to compute it.  Therefore EPE does not provide much
usuable information about the accuracy of a position.

>  The norms of mapping and information graphics all say that 
>using a start which covers perhaps a 50-foot radius, when it is being used
to plot 
>a 200-foot radius, is simply wrong. 

Huh?  What is a start?  Who is plotting a 50' radius as a 200' radius?  

If you are claiming that APRS World is doing this with Tiger data, you need
to spend some time learning about vector based maps and ESRI shape files as
well as cartography in general.  When the Tiger street files are in shape
format (what APRS World uses) the file contains only polylines.  Each street
segment is a series of line segments given as discrete points.  A shape file
can only contain one type (Point, Polygon, Polyline etc).  There is no way
to supply a radius in polyline shape files.

If you are referring to polygons in other Tiger Shape files, you have the
same situation.  The vertices of the polygons are supplied as discrete
points more or less.  There is no provision for Radius.

>And easily corrected.

Really.  Explain how you can correct something that does not exist?  The
shape files do not contain radius!

> Unlike your request to send data that is not supported by the 
>systems, cannot easily be sent, and has a known uncertainty in it, among
other 
>differences.

You claim a "known uncertainty".  Simply include your "known uncertainty" in
every position you supply to APRS.  This is what you are asking James / APRS
World to do with every map he supplies to users.

>"If you supplied the necessary information with every position 
>packet you send, this could be done for you too." What could be done for 
>me too? Correcting the display? 

What is wrong with the way APRS World displays the data?  IMO, he is
faithfully displaying it as presented by the Census Bureau.  Of course, feel
free to correct me if you can tell me EXACTLY what is being displayed
incorrectly.  Just supply the file name and record number of the street
segments you claim are being displayed incorrectly.  Oh, have you bothered
to compare this same street segment using Xastir, WinAPRS, and APRS-SCS.
I'll bet they display it the same exact way.

>As I said, there's nothing wrong with the data 
>being sent, there's no reason (or easy means) to provide 
>more data. The problem at 
>hand is that the
>existing data is being displayed incorrectly. Yes, it is a 
>simple matter of objective right and wrong, the data is 
being displayed wrong and it is easily
>displayed "right" by changing the display icon.

Huh?  You originally claimed the streets were in the wrong position.  You
said it was because the Tiger files were using the wrong datum in some
areas.  

>"FYI, the census bureau data is supplied in vector format, not 
>as an image." I'm
>aware of that.
>"APRS World simply renders the vectors on an image." 

>Are you sure about that? Is it Jim who is rendering those Tiger Maps, 
>or is he getting his maps from another source? 

Geeezzzz.  You really need to take the time to learn something about
mapping.  Most APRS programs use vector mapping data since they are
scalable.  IOW, you can zoom in and out without loosing much detail.
Raster maps on the other hand, will pixellate when you zoom in too far or
the pixels will loose detail as you zoom out too far.  Anyone with the least
bit of knowledge of mapping can EASILY determine if a map is vector or
raster based.  

>And again, that doesn't matter. I'm asking where he 
>gets his data--not just the map, but the position indication and the
default star 
>marker. 

I have explained where he gets the mapping data and it does matter.  You
should know by now that the position indication comes from APRS packets.
Not a big surprise, is it?

>Is he generating the map? 

It is not magic.  It is not something he finds under a rock.  You know he is
using Census Bureau Tiger files for the street data.  Yes, he is generating
the entire map using the Tiger files.  You have castigated him for using
these files previously.  Did you forget?

>Is he generating the marker icon? As long as he is
>generating the *icon*, that's enough, all he has to do is 
>generate it in a
>larger size. It would be even better if the size was related 
>to the display
>scale of the map, so the star represented a set range circle 
>matching the scale
>of the map.

How big?  If you supplied your "unknown uncertainty" it may be possible.
Why don't you try sending some APRS data with that information.  I can't
seem to find any data from you on FindU.  Have you ever used APRS on the air
or via the APRS-IS?

> And as Bob indicated, it is also normal and "right" to 
>display a mileage scale
>with every map.

Ok, in which plane?  

>"Yep, James should do that for every position packet he 
>displays on a map too<g>."

If you ever send any packets on APRS, you should provide him with your
"unknown uncertainty" so he can accommodate your request accurately.

>  You really don't get it, do you? Or are you looking for a 
>flame war? James doesn't need to present the error notice for *any*
position 
>packet, much less for *every* packet. The error warning can be a fixed 
>slug that is on every page, the same way that his map copyright notice 
>or any other boilerplate is already there. The extra time/load on the 
>server is zero, the bits for the screen are
>already being sent regardless of what the boilerplate says.

Have you demanded the same of WinAprs, XAstir, APRS-SCS, Delorme, Navteq,
TeleAtlas etc?  If they agree to do the same I can see no reason for James
not to follow suit.

>"Interesting.  Since virtually every street mapping product 
>for the US was originally based on the TIGER files,"
> You are misinformed. The source of many maps are USGS topo 
>maps and state planimetric maps. The state planimetric maps are usually of 
>extremely high accuracy but you have to PAY to obtain that data from most 
>states for commercial use, which is why some wanna-be map makers grabbed
the TIGER 
>maps instead.

You are showing your lack of cartograhpic knowledge again.  You obviously
need to take the time to learn mapping.

>DeLorme's famous paper map books were, for many years, simply 
>not available for some states because they couldn't afford to buy the good
data. 
>ESRI and other
>commercial vendors have not used the TIGER maps for their bases either.

You are totally, absolutely incorrect.  Show me a topo map which includes
street address ranges.  Virtually every electronic map has this information
and I can assure you that this came directly from Tiger files.  To be sure,
most commercial mapping products have corrected known errors and have added
value by incorporating 3rd party data to the original files, but they all
started with Tiger files.  
>"How can you say anyone is misappropriating files the Census Bureau has
>placed in the Public Domain?  "

>How I can say it, as exactly as I did say it. The files are meant for
>demographic studies. Using them for navigation, when they are 
>expressly not issued for navigation, is misappropriating them. 
>Misappropriate means "to use wrongly", not just "to use illegally". 
>The primary definition is to use something WRONGLY and using the 
>TIGER maps for navigation purposes, is wrong.
>According to the source.

There you go again, accusing every APRS and commercial author of illegality.
You need to get real.

I don't use APRS for navigation and never have and never will.  I can't
think of any APRS user who does.  We all use APRS maps (derived from Tiger
or wherever) for reference.   Pretty simple, when I am driving down the
street, I don't look at maps.  Stupid thing to do while driving.  I look at
street signs.  You should try it some time.

Most APRS users are home based.  They don't need the maps for navigation
either, since they generally know how to find the bathroom and other useful
places in their own home.  Besides, the Tiger maps are not that detailed<g>.

>"You are basically accusing every mapping vendor in the US, 
>and all other users of Census Bureau Tiger file of misappropriating 
>government property.


>Get real." You can apologize for that after you realize you 
>didn't understand the word "misappropriate" and you didn't understand where
map 
>makers got their base maps from, as noted above.

And you have repeatedly demonstrated that you know NOTHING about mapping.
Therefore your views need to be ignored on this subject.

>"If you can get all vendors of US mapping data to do this, 
>than APRS World should do it as well.  "

>  Ah, that's a worthless point to raise in this venue. James 
>made a complaint, we're addressing HIS complaint and HIS software here.
What his 
>software should or should not do, has nothing to do with what some vendors 
>(not the "all" you proclaim) choose to do. "All" vendors aren't doing
anything 
>uniformly, and this is not the venue to argue about what they should be
doing.

Lemme see if I get this straight.  You want James to put a notice on his
maps about inaccurate data but you don't think this should be required of
other authors using the same data?  

And you are the one complaining about his maps.  I don't recall him
complaining about his maps, other than to acknowledge what any knowledgable
mapping application developer already knows:  Tiger maps are not very
accurate.  

>"You certainly need to learn something about cartography 
>before you criticize someone who is EXTREMELY well versed in the science
and 
>application of same."

> Ah, the infamous ad hominem attack when you run of facts to 
>discuss. James is extremely well versed? And I'm not? You know both of
those 
>things for facts?

Yes I do since you demonstrated you don't know the difference between vector
and raster maps, don't have a clue how he creates his maps etc.  Read your
own words above.

>Then he also would have known his display is incorrect. 
>Correct me if I'm wrong, I thought James started this as a *student* 
>project in college. I have no idea what his credentials in your 
>unspecified "science and application" are, but
>unless he's been working in the field for a while, you'd be 
>presuming too much.

Visit his web site.  Some of his papers are available there.  I have known
him for about 6 years now and have had several very interesting face to face
conversations with him.  Talked to him briefly at Dayton this year.  Those
of us who are APRS authors have the utmost respect for his knowledge and
expertise.  He has contributed quite a bit to APRS and we all have learned
from him.

>And if he was an expert? Experts will be the first to tell 
>you, sometimes they still get things wrong.

IMO, your expertise is not unknown, based on what you have posted in this
thread.

> As James knows, I had to pull teeth and do extensive 
>research--including some with academics that he knows--in order to find the
base 
>problem in the TIGER maps. It isn't readily known or available to the
public, 
>including "experts".

You are kidding right?  It is WELL known how inaccurate Tiger files are.
You may have just learned this, but I have worked with them and derivatives
for over 10 years developing custom AVL applications and commerical mapping
software.  We had to use them years ago, because there was simply no
affordable alternative.  I still use them on occasion when I need to develop
a mapping application for users who cannot afford the licensing fees for
commercial mapping data.   Mostly Pro-Bono work.  I would not use TIGER for
commercial application development.  But I certainly use the commerical
derivatives.

>I've done my research, and he knows it. I've traced datum 
>problems back to Dutch surveys of the 1600's, and the 
>state engineers who maintain modern planimetrics,
>and the USGS, are aware of them. It's the HAMS, the APRS 
>users, and the folks who naively misappropriated the 
>TIGER maps, who are in error  there. All of the
>real experts--the state survey offices and academic 
>professionals--know it.

Do you really think you are telling the APRS authors something they don't
know?

Yes, I know it, James knows it, Gerry Creager knows it, and all of the APRS
authors know it.  You have just learned it.  

>The only question is, should hams continue to be the "ham 
>fisted amateurs" that the Marconi operators belittled them 
>as? Or can we do better, 
>and make our presentation reflect its limits?

IMO, your presentation in this thread has FAR exceeded your limits.

I resent the fact that you are accusing me of "naively misappropiating the
TIGER maps".  I resent the fact that you have accused me of illegal
activity.  I resent the fact that you don't have the slightest idea of what
you are talking about, yet you continue to criticize others who are much
more knowledgeable about the subjects at hand.

As I said in my earlier post, "Those who code, rule".  If you don't like the
way maps are presented, then write your own mapping program.  You have said
you are not a programmer.  I have little trouble believing that.  You can do
the same as many hams before you who either didn't have the tools they
needed or knew they could create a better program.  Learn to program.  In
the process you will learn something about logic.  You could certainly
benefit from that aspect of your programming education.  There have been
quite a few hams who have learned to program to fulfill a perceived need,
and have supplied the rest of us with some absolutely wonderful software.
James is but one of them.

The bottom line is:  If you can do better, than do it.  If you can't, then
we don't need to discuss this further.

Bill KC9XG





More information about the aprssig mailing list