[El] GDAL 1.7.3 in elgis-testing

[Mathieu Baudier]

>> we would need to decide whether to have only one soname pointing either
>> to the "normal" or to the iioext "real" file or to have different
>> sonames (in different packages, of course +1). If we use differing
>> sonames, the iomageio-ext classes will have to be patched to look for
>> that differing .so
>
> I would suggest different sonames in different package. This will be
> safer and we can have a separate package.
>
> Mathieu, what do u think?

I don't have much experience with JNI packaging but this indeed seems safer.
I think that this very unlikely that people would use both at the same time.

The question also is: which one would have as a dependency a package
like, say, geoserver?
Event though the one without imageio-ext is "purer", esp. from a FLOSS
point of view, most people would probably aim at the best
performances.
That is were the alternatives mechanism would be handy.

As Ralph suggested, let's try and test.
An important part of the test would be to make sure that there is no
side-effect on non-java stuff so that we can push GDAL 1.7.3 to stable
quickly (that fact that we are doing this work now is independent from
this maintenance release, so it should not be kept "hostage").

Many thanks to you both for your patience!
I'm very excited to see a proper packaging for GIS Java apps around
the corner...