<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META content="text/html; charset=us-ascii" http-equiv=Content-Type>
<META name=GENERATOR content="MSHTML 8.00.6001.18928"></HEAD>
<BODY>
<DIV><FONT size=2 face=Arial></FONT> </DIV>
<DIV><SPAN class=811110609-28032010><FONT size=2 face=Arial>Dear
list,</FONT></SPAN></DIV>
<DIV><SPAN class=811110609-28032010><FONT size=2
face=Arial></FONT></SPAN> </DIV>
<DIV><SPAN class=811110609-28032010><FONT size=2 face=Arial>Has there been any
discussion about a tighter integration of the GDAL and OGR class
hierarchies?</FONT></SPAN></DIV>
<DIV><SPAN class=811110609-28032010><FONT size=2
face=Arial></FONT></SPAN> </DIV>
<DIV><SPAN class=811110609-28032010><FONT size=2 face=Arial>Why I ask:
</FONT></SPAN><SPAN class=811110609-28032010><FONT size=2 face=Arial>Over the
last years, we have been developing a framework for implementing spatio-temporal
simulation models. This has been motivated by distributed hydrological
modelling, but is also appliccable to other fields. All the variables in these
models are GIS data sets that change through time, but there might be both
raster and vector data sets, even mixed with scalars.</FONT></SPAN></DIV>
<DIV><SPAN class=811110609-28032010><FONT size=2
face=Arial></FONT></SPAN> </DIV>
<DIV><SPAN class=811110609-28032010><FONT size=2 face=Arial>We started this
before knowing of GDAL, and have a data structure where the different GIS data
classes (at present raster, point-vector and scalar) inherits a common
superclass containing metadata not specific to any one GIS data type. This is
advantageous for the model framework, which can list and handle all
the model variables as one common type.</FONT></SPAN></DIV>
<DIV><SPAN class=811110609-28032010><FONT size=2
face=Arial></FONT></SPAN> </DIV>
<DIV><FONT face=Arial><FONT size=2><SPAN class=811110609-28032010>Now Statkraft
(the hydropower company that financed the development) has decided to
go Open Source with this project. We have not started the process, but I am
guessing that getting rid of MFC and other Windows/Microsoft specifics will
increase the number of potential users.</SPAN></FONT></FONT></DIV>
<DIV><FONT face=Arial><FONT size=2><SPAN
class=811110609-28032010></SPAN></FONT></FONT> </DIV>
<DIV><FONT face=Arial><FONT size=2><SPAN class=811110609-28032010>Finally coming
to the point: My existing CGisData superclass is an MFC derivative, and thus
needs a new parent class. It is tempting to look for a possible
solution in the well established GDAL/OGR libraries, which might also
provide a shortcut to adding line and polygon covers to the list of
acceptable data types. But the ability to implement a model or subroutine
for "any" data type is something I would like to
preserve.</SPAN></FONT></FONT></DIV>
<DIV><FONT face=Arial><FONT size=2><SPAN
class=811110609-28032010></SPAN></FONT></FONT> </DIV>
<DIV><FONT face=Arial><FONT size=2><SPAN class=811110609-28032010>So: Does
anybody else see any advantage of making GDALMajorObject a subclass
of OGRGeometry; or giving the two a common
superclass?</SPAN></FONT></FONT></DIV>
<DIV><FONT face=Arial><FONT size=2><SPAN
class=811110609-28032010></SPAN></FONT></FONT> </DIV>
<DIV><FONT face=Arial><FONT size=2><SPAN class=811110609-28032010>Best
regards,</SPAN></FONT></FONT></DIV>
<DIV><FONT face=Arial><FONT size=2><SPAN
class=811110609-28032010></SPAN></FONT></FONT> </DIV>
<DIV><FONT face=Arial><FONT size=2><SPAN class=811110609-28032010>Sjur K
:-)</SPAN></FONT></FONT></DIV>
<DIV><FONT face=Arial><FONT size=2><SPAN
class=811110609-28032010></SPAN></FONT></FONT> </DIV></BODY></HTML>