Re: Discovery: libproj4 stmerc = French Gauss-Laborde

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
12 messages Options
Reply | Threaded
Open this post in threaded view
|

Re: Discovery: libproj4 stmerc = French Gauss-Laborde

Melita Kennedy
> Cliff Mugnier wrote:
> Melita Kennedy of ESRI informed me a couple years ago that they
> successfully programmed Jeff Dozier's TM equations and have incorporated
> that routine into one of the ESRI packages, I don't remember which one.

David Burrows implemented a complex math version of transverse Mercator
working from Dozier's equations. On a brief stop in the office today (he's
on vacation), he couldn't remember any errors in Dozier's equations
off-hand. He did mention that he could not make it work out to 90 degrees
from the central meridian as the paper states. I think we ended up limiting
it to 85 or 87 degrees. Next week he'll take a look at  his copy of the
paper to see if he wrote any notes down.

Melita

--
Melita Kennedy
[hidden email]

_______________________________________________
Proj mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/proj
Reply | Threaded
Open this post in threaded view
|

Re: Discovery: libproj4 stmerc = French Gauss-Laborde

Gerald I. Evenden-2
On Thursday 15 June 2006 3:10 am, Melita Kennedy wrote:

> > Cliff Mugnier wrote:
> > Melita Kennedy of ESRI informed me a couple years ago that they
> > successfully programmed Jeff Dozier's TM equations and have incorporated
> > that routine into one of the ESRI packages, I don't remember which one.
>
> David Burrows implemented a complex math version of transverse Mercator
> working from Dozier's equations. On a brief stop in the office today (he's
> on vacation), he couldn't remember any errors in Dozier's equations
> off-hand. He did mention that he could not make it work out to 90 degrees
> from the central meridian as the paper states. I think we ended up limiting
> it to 85 or 87 degrees. Next week he'll take a look at  his copy of the
> paper to see if he wrote any notes down.
>
> Melita

I managed to dig out my old work with the Dozier package---dates on the files
are 2003.

I AGAIN compared my original copy of his original paper code and noted a few  
problems I detected several years ago:

1. I had to add a header file to all files to define the complex declaration:
typedef struct {
        double re, im;
} complex;

2. Three files did not declare complex functions---which were added.

3. Around line 70 in file tmw.c I did the following change:
/* old
     70     ms.re *= m;
     71     ms.im *= m;
     72   new */
     73     ms.re *= k_tran;
     74     ms.im *= k_tran;
     75 /* end change */
At the moment I would have to dig for my reason.

I did get the code to run but could not get it to agree with tmerc to better
than about .1%.  Obviously, not very good.  (also in the lon < 3degree
range).

There are a number of questionable code practices but most are unlikely to
produce errors.  The code is VERY early K&R and would be GREATLY improved
with updating to modern declarative standards.  Most, if not all, of the
basic complex routines are available in either C99 library or GSL (Gnu
Scientific Library).

After I hopefully solve my problems with the French version I will try to
regress to my 2003 problems with Dozier.  If anyone has any news of Dozier's
reexamining this material I would be greatly appreciative of learning of any
errata.
--
Jerry and the low-riders: Daisy Mae and Joshua
"Cogito cogito ergo cogito sum"
   Ambrose Bierce, The Devil's Dictionary
_______________________________________________
Proj mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/proj
Reply | Threaded
Open this post in threaded view
|

Re: Discovery: libproj4 stmerc = French Gauss-Laborde

Gerald I. Evenden-2
In reply to this post by Melita Kennedy
The French Transverse Mercator (NT/G 76 publication) is now in libproj4 on my
desk---+proj=ftmerc .  It extends well beyond the 5 degree limit of tmerc but
starts to fail with low latitudes and extended longitudes beyond 40 degrees.  
It compares with the German online site within +-0.2mm.  Unfortunately the
German site does not display major axis and ellipsoid parameters so it is
possible that errors in this range are due to minor variations of these
figures and/or how they are specified.

ftmerc has forward and inverse mode.

This release of libproj4 will not be on my web site until the documentation is
brought up to date and the new output formating routine is installed.

If anyone wants a prelim copy of proj_ftmerc.c and proj_psi.c, I can email
these files.

I am not going to do anything at the moment with the Dozier method.

--
Jerry and the low-riders: Daisy Mae and Joshua
"Cogito cogito ergo cogito sum"
   Ambrose Bierce, The Devil's Dictionary
_______________________________________________
Proj mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/proj
Reply | Threaded
Open this post in threaded view
|

Re: Discovery: libproj4 stmerc = French Gauss-Laborde

OvV_HN
In reply to this post by Gerald I. Evenden-2
> From: Strebe-aol.com

> The "trick" is this: Wallis uses the polar stereographic because it's the
> simplest way to get a conformal mapping to the plane. Once the ellipsoid is
> mapped, he treats the plane as the complex plane and looks for a complex
> "co-latitude" which can be used with the polar stereographic, but this time
> treating the polar stereographic as function of a complex variable.

Now where have I seen this again?
Ah yes:

"Transverse Mercator Map Projection of the Spheroid Using Transformation of
the Elliptic Integral software (NPO-18086)."
"... Using the colatitude (complement of latitude) and the longitude
(departure), the initial step is to map the point to the North Polar
Stereographic Projection. The closed-form, analytic function that coincides
with the North Polar Stereographic Projection of the spheroid along the
principal meridian is put into a Newton-Raphson iteration that solves for
the tangent of one half the parametric colatitude, generalized to the
complex plane. ..."
Etcetera.

There is documentation available ($) and Fortran software ($). Export out of
the USA is forbidden.
"The program was developed in 1989."
<http://www.openchannelfoundation.org/projects/Transverse_Mercator_Map_Proje
c/>

Texts between "" are quotes from web pages found at the Open Channel
Foundation.


_______________________________________________
Proj mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/proj
Reply | Threaded
Open this post in threaded view
|

Re: Discovery: libproj4 stmerc = French Gauss-Laborde

Gerald I. Evenden-2
On Monday 19 June 2006 8:09 am, Oscar van Vlijmen wrote:
        ...
>
> There is documentation available ($) and Fortran software ($). Export out
> of the USA is forbidden.
> "The program was developed in 1989."
> <http://www.openchannelfoundation.org/projects/Transverse_Mercator_Map_Proj
>e c/>
>
> Texts between "" are quotes from web pages found at the Open Channel
> Foundation.

Sounds like a private clone of NTIS where a 18 minute VHS video tape on charts
and projections sells for $110US.  I believe the above goes for $11US for the
documentation and $50US for the FORTRAN code.

At least NTIS does not forbid export.

What BS and complete nonsense!!

Interesting: I could not find an author associated with the product--maybe it
is Dr. Wallis.

--
Jerry and the low-riders: Daisy Mae and Joshua
"Cogito cogito ergo cogito sum"
   Ambrose Bierce, The Devil's Dictionary
_______________________________________________
Proj mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/proj
Reply | Threaded
Open this post in threaded view
|

Re: Discovery: libproj4 stmerc = French Gauss-Laborde

strebe
In reply to this post by OvV_HN

I doubt the Open Channel Foundation source code has anything to do with
Wallis unless the program was written by someone else he communicated
with without his knowledge. Dr. Wallis has never tried to sell me
anything; never mentioned the existence of commercial source code; and
has communicated with me about his own implementation, including
algorithmic subtleties and requests for independent confirmation of
accuracy. He also considers the method to be "obvious" and not worth
publication on its own. On the other hand, either he believes he
originated the method, or else wanted me to believe he did.

I can't say whether the product mentioned before came first, or whether
Dr. Wallis's method came first. I made Dr. Wallis's acquaintance in
1996, and I know he had written his implementation years before that,
though I don't know how many. I'll contact him about the Open Channel
Foundation program.

$61 for source code and documentation seems modest enough, but the
inane export restrictions won't help Mr. van Vlijmen. While I will not
supply source code, I'm happy to answer specific questions about
implementing the method.

Regards,
-- daan Strebe


-----Original Message-----
From: Gerald I. Evenden <[hidden email]>
To: PROJ.4 and general Projections Discussions <[hidden email]>
Sent: Mon, 19 Jun 2006 10:42:51 -0400
Subject: Re: [Proj] Discovery: libproj4 stmerc = French Gauss-Laborde

...

Interesting: I could not find an author associated with the
product--maybe it is Dr. Wallis.

-----

and

-----

-----Original Message-----
From: Oscar van Vlijmen <[hidden email]>
To: PROJ.4 and general Projections Discussions <[hidden email]>
Sent: Mon, 19 Jun 2006 14:09:30 +0200
Subject: Re: [Proj] Discovery: libproj4 stmerc = French Gauss-Laborde

   > From: Strebe-aol.com

> The "trick" is this: Wallis uses the polar stereographic because it's
the
> simplest way to get a conformal mapping to the plane. Once the
ellipsoid is
> mapped, he treats the plane as the complex plane and looks for a
complex
> "co-latitude" which can be used with the polar stereographic, but
this time
> treating the polar stereographic as function of a complex variable.

Now where have I seen this again?
Ah yes:

"Transverse Mercator Map Projection of the Spheroid Using
Transformation of
the Elliptic Integral software (NPO-18086)."
"... Using the colatitude (complement of latitude) and the longitude
(departure), the initial step is to map the point to the North Polar
Stereographic Projection. The closed-form, analytic function that
coincides
with the North Polar Stereographic Projection of the spheroid along the
principal meridian is put into a Newton-Raphson iteration that solves
for
the tangent of one half the parametric colatitude, generalized to the
complex plane. ..."
Etcetera.

There is documentation available ($) and Fortran software ($). Export
out of
the USA is forbidden.
"The program was developed in 1989."
<http://www.openchannelfoundation.org/projects/Transverse_Mercator_Map_Pr
oje
c/>

Texts between "" are quotes from web pages found at the Open Channel
Foundation.


_______________________________________________
Proj mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/proj


________________________________________________________________________
Check out AOL.com today. Breaking news, video search, pictures, email
and IM. All on demand. Always Free.

_______________________________________________
Proj mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/proj
Reply | Threaded
Open this post in threaded view
|

Re: Re: Discovery: libproj4 stmerc = French Gauss-Laborde

Gerald I. Evenden-2
On Monday 19 June 2006 12:59 pm, [hidden email] wrote:
> I doubt the Open Channel Foundation source code has anything to do with
> Wallis unless the program was written by someone else he communicated
> with without his knowledge. Dr. Wallis has never tried to sell me
> anything; never mentioned the existence of commercial source code; and
> has communicated with me about his own implementation, including
> algorithmic subtleties and requests for independent confirmation of
> accuracy. He also considers the method to be "obvious" and not worth
> publication on its own. On the other hand, either he believes he
> originated the method, or else wanted me to believe he did.

Sorry, my statement was only idle speculation because of the similarity of the
subject and the fact Dr. Wallis promised to publish the code several years
ago.  It could have been an outlet for him and probably easier than going
thought the pains of a professional publication---peer review, etc..

> I can't say whether the product mentioned before came first, or whether
> Dr. Wallis's method came first. I made Dr. Wallis's acquaintance in
> 1996, and I know he had written his implementation years before that,
> though I don't know how many. I'll contact him about the Open Channel
> Foundation program.

I did post a letter to Dr. Wallis a few days ago.  I explained my interests
and promised that I would not release any material related to his work
without his prior permission.  The reason for a letter was to communicate
information related to libproj4 web site that is difficult to convey
verbally.  I hope he has not moved.

> $61 for source code and documentation seems modest enough, but the
> inane export restrictions won't help Mr. van Vlijmen. While I will not
> supply source code, I'm happy to answer specific questions about
> implementing the method.

My objection to the ten bucks for the documentation reminds me of similar
costs for NTIS material and only getting one or two pages of a totally
worthless dissertation.  If the source code is commercially useful than it is
worth more than $50 but seems steep to me for just academic interest.  
Especially considering the amount and often high quality of material
available on the net for nothing.  There also seem to be chance of hidden
strings to seeing the material that might compromise my being able to use and
document derived material.

--
Jerry and the low-riders: Daisy Mae and Joshua
"Cogito cogito ergo cogito sum"
   Ambrose Bierce, The Devil's Dictionary
_______________________________________________
Proj mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/proj
Reply | Threaded
Open this post in threaded view
|

Re: Re: Discovery: libproj4 stmerc = French Gauss-Laborde

Melita Kennedy
In reply to this post by Melita Kennedy
Here's a website that does list the author as Dr. Wallis:

http://www.nttc.edu/cosmic/pub/authors.html

Yet another website mentioned that "international" prices were $22/doc, $100/code.

http://www.nttc.edu/cosmic/abstracts/npo-18086.html

Melita

>------------------------------
>
>Date: Mon, 19 Jun 2006 13:43:21 -0400
>From: "Gerald I. Evenden" <[hidden email]>
>Subject: Re: [Proj] Re: Discovery: libproj4 stmerc = French
> Gauss-Laborde
>To: "PROJ.4 and general Projections Discussions"
> <[hidden email]>
>Message-ID: <[hidden email]>
>
>On Monday 19 June 2006 12:59 pm, [hidden email] wrote:
>> I doubt the Open Channel Foundation source code has anything to do with
>> Wallis unless the program was written by someone else he communicated
>> with without his knowledge. Dr. Wallis has never tried to sell me
>> anything; never mentioned the existence of commercial source code; and
>> has communicated with me about his own implementation, including
>> algorithmic subtleties and requests for independent confirmation of
>> accuracy. He also considers the method to be "obvious" and not worth
>> publication on its own. On the other hand, either he believes he
>> originated the method, or else wanted me to believe he did.
>
>Sorry, my statement was only idle speculation because of the similarity of the
>subject and the fact Dr. Wallis promised to publish the code several years
>ago.  It could have been an outlet for him and probably easier than going
>thought the pains of a professional publication---peer review, etc..
>
>> I can't say whether the product mentioned before came first, or whether
>> Dr. Wallis's method came first. I made Dr. Wallis's acquaintance in
>> 1996, and I know he had written his implementation years before that,
>> though I don't know how many. I'll contact him about the Open Channel
>> Foundation program.
>
>I did post a letter to Dr. Wallis a few days ago.  I explained my interests
>and promised that I would not release any material related to his work
>without his prior permission.  The reason for a letter was to communicate
>information related to libproj4 web site that is difficult to convey
>verbally.  I hope he has not moved.
>
>> $61 for source code and documentation seems modest enough, but the
>> inane export restrictions won't help Mr. van Vlijmen. While I will not
>> supply source code, I'm happy to answer specific questions about
>> implementing the method.
>
>My objection to the ten bucks for the documentation reminds me of similar
>costs for NTIS material and only getting one or two pages of a totally
>worthless dissertation.  If the source code is commercially useful than it is
>worth more than $50 but seems steep to me for just academic interest.  
>Especially considering the amount and often high quality of material
>available on the net for nothing.  There also seem to be chance of hidden
>strings to seeing the material that might compromise my being able to use and
>document derived material.
>
>--
>Jerry and the low-riders: Daisy Mae and Joshua
>"Cogito cogito ergo cogito sum"
>   Ambrose Bierce, The Devil's Dictionary
>
>
>------------------------------

_______________________________________________
Proj mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/proj
Reply | Threaded
Open this post in threaded view
|

Re: Re: Re: Discovery: libproj4 stmerc = French Gauss-Laborde

Gerald I. Evenden-2
On Tuesday 20 June 2006 2:02 pm, Melita Kennedy wrote:

> Here's a website that does list the author as Dr. Wallis:
>
> http://www.nttc.edu/cosmic/pub/authors.html
>
> Yet another website mentioned that "international" prices were $22/doc,
> $100/code.
>
> http://www.nttc.edu/cosmic/abstracts/npo-18086.html
>
> Melita
>
> >------------------------------
> >
> >Date: Mon, 19 Jun 2006 13:43:21 -0400
> >From: "Gerald I. Evenden" <[hidden email]>
> >Subject: Re: [Proj] Re: Discovery: libproj4 stmerc = French
> > Gauss-Laborde
> >To: "PROJ.4 and general Projections Discussions"
> > <[hidden email]>
> >Message-ID: <[hidden email]>
> >
> >On Monday 19 June 2006 12:59 pm, [hidden email] wrote:
> >> I doubt the Open Channel Foundation source code has anything to do with
> >> Wallis unless the program was written by someone else he communicated
> >> with without his knowledge. Dr. Wallis has never tried to sell me
> >> anything; never mentioned the existence of commercial source code; and
> >> has communicated with me about his own implementation, including
> >> algorithmic subtleties and requests for independent confirmation of
> >> accuracy. He also considers the method to be "obvious" and not worth
> >> publication on its own. On the other hand, either he believes he
> >> originated the method, or else wanted me to believe he did.
> >
> >Sorry, my statement was only idle speculation because of the similarity of
> > the subject and the fact Dr. Wallis promised to publish the code several
> > years ago.  It could have been an outlet for him and probably easier than
> > going thought the pains of a professional publication---peer review,
> > etc..

Well, well.  I guessed right.

NASA putting a copyright on the material---what bull shit!  Sorry, but there
is no other appropriate description.  US tax dollars at work!!

Presumably there is no patent on the idea then redeveloping the software in
another language is no problems---especially if one does not view the source
code.

Sorry, but my blood pressure is starting to get out of hand so ...

If I get a reply from Dr. Wallis I wonder what it will say?

--
Jerry and the low-riders: Daisy Mae and Joshua
"Cogito cogito ergo cogito sum"
   Ambrose Bierce, The Devil's Dictionary
_______________________________________________
Proj mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/proj
Reply | Threaded
Open this post in threaded view
|

Re: Re: Re: Discovery: libproj4 stmerc = French Gauss-Laborde

Gerald I. Evenden-2
In reply to this post by Melita Kennedy
On Tuesday 20 June 2006 2:02 pm, Melita Kennedy wrote:

> Here's a website that does list the author as Dr. Wallis:
>
> http://www.nttc.edu/cosmic/pub/authors.html
>
> Yet another website mentioned that "international" prices were $22/doc,
> $100/code.
>
> http://www.nttc.edu/cosmic/abstracts/npo-18086.html
>
> Melita
       
Another item from the above:

"DISTRIBUTION MEDIA: 5.25 Inch IBM PC DOS Format Diskette"

I have three computers in my house and not a one of them has a floppy
disk---5.25 nor 3.5---especially 5.25.  Threw all that stuff out a long time
ago.  So I am safe from seeing the source code.

The distribution outfit must be in the dark age.
--
Jerry and the low-riders: Daisy Mae and Joshua
"Cogito cogito ergo cogito sum"
   Ambrose Bierce, The Devil's Dictionary
_______________________________________________
Proj mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/proj
Reply | Threaded
Open this post in threaded view
|

Re: Re: Re: Discovery: libproj4 stmerc = French Gauss-Laborde

HamishB
In reply to this post by Gerald I. Evenden-2
Gerald I. Evenden wrote:

> NASA putting a copyright on the material---what bull shit!  Sorry, but
> there is no other appropriate description.  US tax dollars at work!!
>
> Presumably there is no patent on the idea then redeveloping the
> software in another language is no problems---especially if one does
> not view the source code.


US Code title 17, ch 1, section 105:
"Copyright protection under this title is not available for any work of
the United States Government, but the United States Government is not
precluded from receiving and holding copyrights transferred to it by
assignment, bequest, or otherwise."

http://www.law.cornell.edu/uscode/html/uscode17/usc_sec_17_00000105----000-.html
http://www.law.cornell.edu/uscode/html/uscode17/usc_sec_17_00000105----000-notes.html
http://en.wikipedia.org/wiki/Work_of_the_United_States_Government
http://en.wikipedia.org/wiki/United_States_copyright_law


re. the "limited exception for national technical information service"
in the notes section- if we are talking 5.25" floppies I think it is
pretty safe to say that any 5 yr copyright period will have expired by
now!


Hamish
_______________________________________________
Proj mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/proj
Reply | Threaded
Open this post in threaded view
|

Re: Re: Re: Discovery: libproj4 stmerc = French Gauss-Laborde

Eric Miller-4
In reply to this post by Melita Kennedy
Well, the NASA Technical Reports Server has a reference, so maybe you can get it from them (or CASI actually).  But, it'll still probably cost money for the reproduction and you may only get a copy of a source print out of the Fortran code.  But, there shouldn't be any license restriction if NASA is the clear copyright holder (which the reference claims is the case).

NTRS REFERENCE:

http://tinyurl.com/rpj37 (Document Id: 19940003196, Publication Date: Jan. 1, 1994)

GENERAL ORDERING CONTACT INFO:

http://ntrs.nasa.gov/index.cgi?method=ordering

PRICE LIST:

http://www.sti.nasa.gov/price1.pdf 

>>> [hidden email] 6/22/2006 2:56:50 AM >>>
Gerald I. Evenden wrote:

> NASA putting a copyright on the material---what bull shit!  Sorry, but
> there is no other appropriate description.  US tax dollars at work!!
>
> Presumably there is no patent on the idea then redeveloping the
> software in another language is no problems---especially if one does
> not view the source code.


US Code title 17, ch 1, section 105:
"Copyright protection under this title is not available for any work of
the United States Government, but the United States Government is not
precluded from receiving and holding copyrights transferred to it by
assignment, bequest, or otherwise."

http://www.law.cornell.edu/uscode/html/uscode17/usc_sec_17_00000105----000-.html 
http://www.law.cornell.edu/uscode/html/uscode17/usc_sec_17_00000105----000-notes.html 
http://en.wikipedia.org/wiki/Work_of_the_United_States_Government 
http://en.wikipedia.org/wiki/United_States_copyright_law 


re. the "limited exception for national technical information service"
in the notes section- if we are talking 5.25" floppies I think it is
pretty safe to say that any 5 yr copyright period will have expired by
now!


Hamish
_______________________________________________
Proj mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/proj


_______________________________________________
Proj mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/proj