Proposed patch for the module-searching code

[mi]

As currently written, the modules will be searched first in the directories specified by the environment variables, and then -- if the desired module is not found -- in the compiled-in path.

The patch would modify that algorithm to only consider the compiled-in path, if the environment variables are not found.

An earlier buggy version of this patch is currently included in the FreeBSD port. The proposed version should be correct:

Code: Select all

--- magick/module.c     2007-09-29 20:39:15.000000000 -0400
+++ magick/module.c     2007-10-14 18:22:26.000000000 -0400
@@ -516,4 +516,5 @@
     }
 #if defined(UseInstalledMagick)
+  else
 #if defined(MagickImageCodersPath)
   {

[magick]

We applied your patch, released ImageMagick 6.3.6-6, and released libfpx-1.2.0.14. Feel free to post to this forum any and all bugs and problems you encounter with these releases. Thanks.

[mi]

We applied your patch, released ImageMagick 6.3.6-6

OpenEXR is still a problem... Can you please, state for the record, that ImageMagick-6.3.6-6 builds fine on all of the systems you use for testing (and list these systems)? Before I dig into investigating the failure on FreeBSD, I'd like to be reassured, that it is indeed a FreeBSD-specific problem...

and released libfpx-1.2.0.14.

Thank you. However:

• I was a little disappointed to not find myself mentioned anywhere in this new release. The AUTHORS file, for example, still dates from 2005 and the NEWS is from 2003.
• I think, the release number should've been 1.2.1.0 or even 1.3.0.0 - the internal changes are large and the external interface (API) is slightly modified, thus justifying a more significant version bump.
• The directory inside the tarball is "1.2.0" instead of the full release number as in the past. A minor inconsistency...

[magick]

We will release libFPX 1.3.0 sometime today with the version change as you suggested and an attribution to you in the AUTHORS file. You can edit it as you see fit and we will get the changes into the next point release.

[mi]

Thanks.

What about the EXR problem, though?