openexr-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Openexr-devel] R6034 with OpenEXR and Visual Studio 2008


From: Thomas Binder
Subject: Re: [Openexr-devel] R6034 with OpenEXR and Visual Studio 2008
Date: Fri, 24 Jul 2009 12:12:15 +0200

It seems that my problem was solved while working on a clear separation
of debug and release binaries. I suspect this was the reason although
I have no clue about special details which leaves a bad taste. Due to impatience
I turned too many screws at once.

So, I guess your remark on having identical settings for all libs and
binaries was a match.
Thanks, Ger.

The current version of createDLL is lacking the call to the manifest tool.
Therefore, simply converting the Visual Studio 2005 project files to
Visual Studio 2008
project files is not sufficient.

Here are some remarks after having played with the ilmbase/ilmimf Windows
setup for quite a while:

At first sight I did not like createDLL at all. I would have preferred
seeing some conditional

#define OPENEXR_API __declspec(dllexport)

in the source code, where OPENEXR_API occurs next to each exported
class, function all over the place.
createDLL seemed to be too far off the beaten track.
However, I changed my mind when I used createDLL with different
command line options to build my current setup:
Half.lib, Iex.lib, Imath.lib, IlmThread.lib statically linked into IlmImf.dll.

Still I agree that using openEXR in a nonstandard way (like I do)
on Windows can be painful. I spent much time changing options in VS
project files via the properties windows. It would be nice to reduce
the number of projects. I was also confused with the Deploy directory
which doubles the number of binaries around, one must be really
disciplined that each one has the linker called twice plus the manifest tool.

On the other hand, calling the API to read/write exr images is easy
and transparent from the documentation.

Best,
Tom




reply via email to

[Prev in Thread] Current Thread [Next in Thread]