[Proj] Re: Transverse Mercator algorithm with good accuracy/speedtrade-off?

[Gerald I. Evenden]

On Saturday 24 May 2008 1:14 pm, you wrote:
> It's not purely an "academic" horse if you practice in the oil patch. 
> Reservoir engineers want a system recognizable to them and don't really
> care about practical geodetic applications, ... just solutions to their own
> problems and UTM is the way they want to go.

I do not think of UTM as simply a "practical geodetic application" especially 
since it was championed by the military as a mechanism to standardize mapping 
and help artillery FOs in their job.  Indeed, if we are to champion practical 
application we should educate the oil patch industry to find better 
solutions.

A practical problem with UTM and its longitude limit is that if one proceeds 
too far to the west the grid system goes negative---a fatal situation for 
those who can't handle signed numbers.  If you change the X offset to 
compensate please be careful to NOT use the term UTM as the name of the new 
system(s).

> You're not going to succeed in educating the entire world's Oil Patch
> industry.

It is a noble undertaking to educate the ignorant.

> There are practical applications to this even if a classical cartographer
> is ignorant of the reasons.

Please explain and justify.  I do not understand the use of the 
term "classical." I am just discussing practical, reasonable and proper 
usage.

> The horse ain't dead, just maybe the jockey.

:-)

> Cliff Mugnier

As an aside: I was never able to get Dozier's procedure to properly function 
in the nearly +/-90 longitude situation and communications with Dozier seemed 
to indicate that his interest in the problem had waned.
> ________________________________
>
> From: proj-bounces at lists.maptools.org on behalf of Gerald I. Evenden
> Sent: Sat 24-May-08 09:39
> To: PROJ.4 and general Projections Discussions
> Subject: Re: [Proj] Re: Transverse Mercator algorithm with good
> accuracy/speedtrade-off?
>
>
>
> Ahhh!  Sigh,  Deja vu all over again.  I thought we beat this dead horse
> and Dosier several years ago.
>
> It seems worthless to expand the longitudinal accuracy of the projection as
> long as the scale error and usefulness also goes to hell at the same time.
> The useful range for the tmerc is in the immediate vicinity of the central
> meridian so why bother worrying about extending computational accuracy in
> the longitudinal direction?
>
> Enough is enough.  Please.
...

-- 
The whole religious complexion of the modern world is due
to the absence from Jerusalem of a lunatic asylum.
-- Havelock Ellis (1859-1939)  British psychologist