[OSRS-PROJ] Weirdnesses in the source code

[Gerald I. Evenden]

First of all, can we change the name of this thread.  "Weirdnesses"
is not that meaningful.

One problem with projection parameters is that several projections
have different means of expressing their control  For example,
Lam. Conformal Conic can use two parallels or one parallel and
a scale factor.

Another item is that many control factors are not explicitly
mentioned in the -lP line because they are common to all projections.
For example, ellipsoid/sphere parameters, false easting/northing, ... .

Earlier versions attempted to address this problem with the
ancillary files that contained definitions for default
parameters but was later dropped.

What I see here is the never ending debate as to how much documentation
should be placed in the library itself.  What is in there now is
minimal and what I believe is appropriate for the current library.
To add more detail and user assistance then a complex addenda
is added that would have to be tailored to various
GUI systems.  I do not feel that is part of the basic library that
tries to restrict itself to merely doing the numbers.

The -lP option is only a quick reminder to the knowledgeable user
about the parameters required and *not* an instructional tool.
To add a meaningful interface for the novice user would require
software orders of magnitude more complex than the current libproj
library.

On Sat, 2003-07-19 at 13:00, Paul Kelly wrote:
> On Tue, 8 Jul 2003, Gerald I. Evenden wrote:
> 
> > Thank-you for all the bug reports.  Imbedded are some comments prior
> > to looking in to them in detail:
> >
> > On Tue, 2003-07-08 at 11:23, proj at ton.iguana.be wrote:
> > > Hi,
> > >
> > > I'm busy transforming some of the proj code to perl, and ran into a
> > > number of strange things.
> >
> > On the HEAD stuff, the second quote set is informational and does
> > not affect execution, but they will be fixed.
> >
> > > PJ_aeqd.c
> > >   in PROJ_HEAD, lat_0 should be lat_0=
> > >
> > > PJ_aitoff.c
> > >   in PROJ_HEAD, lat_1 should be lat_1=
> >
> 
> This is interesting to me as a while ago I was wondering if it was possible
> to use the information in the 'second quote set' for each projection (i.e.
> the output of the 'proj -lP' command) to automatically generate a list of
> which parameters were required for each projection.
> 
> The motivation for this was that I was doing some work on tidying up the
> way proj is used in the GRASS GIS.
> 
> Previously GRASS had its own private copy of proj4.3.3 inside the
> source tree but we are now swiching over to using an external shared
> library version of Frank's/remotesensing.org PROJ.4 library and would like
> to abandon the internal proj code in GRASS.
> 
> The only part I am finding it difficult to deal with is the part that has
> been added to GRASS that associates a list of parameters with each
> projection. This is needed to allow the user to easily interactively
> define all the projection parameters for the location they are working in.
> The GRASS module g.setproj asks for the projection name and then
> interactively prompts for each parameter, suggesting default values where
> appropriate (though we could always get rid of that part; most of them are
> zero anyway).
> 
> As far as I can see (though I haven't done a totally exhaustive search)
> there does not seem to be anything like this internal
> to libproj, except the information meant for human interpretation that Ton
> is trying to parse.
> 
> I would be very interested to see this parsing code, and perhaps a C
> version could be added to PROJ as a function in the future?
> 
> Paul Kelly
> 
> ----------------------------------------
> PROJ.4 Discussion List
> See http://www.remotesensing.org/proj for subscription, unsubscription
> and other information.
-- 
Gerald I. Evenden <gerald.evenden at verizon.net>

----------------------------------------
PROJ.4 Discussion List
See http://www.remotesensing.org/proj for subscription, unsubscription
and other information.