[GRASS-dev] i.atcorr don't compile

Hi,

the module i.atcorr in 6.4svn and 7svn don't compile (debian wheezy - 64 bit).

With make in the mail folder I have not error.

With make in i.atcorr folder I've this log:

http://dpaste.com/hold/623443/

Thanks,

Alfredo

On Thu, Sep 29, 2011 at 4:57 PM, Alfredo Alessandrini
<alfreale74@gmail.com> wrote:

Hi,

the module i.atcorr in 6.4svn and 7svn don't compile (debian wheezy - 64 bit).

With make in the mail folder I have not error.

(for the list record, since Anne solved it offlist already)

You need to configure GRASS with --with-cxx to enable the compilation
of i.atcorr.

Markus

Markus Neteler wrote:

You need to configure GRASS with --with-cxx to enable the compilation
of i.atcorr.

Should this be enabled by default?

With libraries, there's at least a possibility that the user might
infer that a particular module will need the appropriate --with-*
flag. But there's no clue as to which modules are written in C++.

While we're on the subject, can we choose a preferred file extension
for C++? There's no valid reason for needing duplicate sets of make
rules for both .cc and .cpp (and also .cxx, .c++ and/or .C if someone
ever uses those).

--
Glynn Clements <glynn@gclements.plus.com>

2011/9/29 Glynn Clements <glynn@gclements.plus.com>:

You need to configure GRASS with --with-cxx to enable the compilation
of i.atcorr.

Should this be enabled by default?

I would say 'yes'.

Martin

--
Martin Landa <landa.martin gmail.com> * http://geo.fsv.cvut.cz/~landa

On Thu, Sep 29, 2011 at 10:28 PM, Glynn Clements
<glynn@gclements.plus.com> wrote:

Markus Neteler wrote:

You need to configure GRASS with --with-cxx to enable the compilation
of i.atcorr.

Should this be enabled by default?

Yes, why not.

With libraries, there's at least a possibility that the user might
infer that a particular module will need the appropriate --with-*
flag. But there's no clue as to which modules are written in C++.

Exactly. We were initially in the dark, why i.atcorr didn't come up
while no error was reported.

While we're on the subject, can we choose a preferred file extension
for C++? There's no valid reason for needing duplicate sets of make
rules for both .cc and .cpp (and also .cxx, .c++ and/or .C if someone
ever uses those).

GDAL uses .cpp. Also others - a quick local poll on my dev machine:

[neteler@north ~]$ locate '.cpp' | wc -l
14213
[neteler@north ~]$ locate '.cc' | wc -l
834
[neteler@north ~]$ locate '.cxx' | wc -l
672

I'd vote for .cc|.cxx -> .cpp

Markus

2011/9/29 Markus Neteler <neteler@osgeo.org>:

I'd vote for .cc|.cxx -> .cpp

personally I prefer `.cpp`.

Martin

--
Martin Landa <landa.martin gmail.com> * http://geo.fsv.cvut.cz/~landa

Markus Neteler wrote:

GDAL uses .cpp. Also others - a quick local poll on my dev machine:

[neteler@north ~]$ locate '.cpp' | wc -l
14213
[neteler@north ~]$ locate '.cc' | wc -l
834
[neteler@north ~]$ locate '.cxx' | wc -l
672

Probably influenced by MSVC/DevStudio using .cpp.

I'd vote for .cc|.cxx -> .cpp

Done in r48562 (modify libiostream, r.terraflow, wximgview, xganim to
use .cpp instead of .cc) and r48563 (remove the *.make rules for .cc).

--
Glynn Clements <glynn@gclements.plus.com>

Markus Neteler wrote:

>> You need to configure GRASS with --with-cxx to enable the compilation
>> of i.atcorr.
>
> Should this be enabled by default?

Yes, why not.

Done in r48564.

--
Glynn Clements <glynn@gclements.plus.com>