Please try this ECW 3.1-rc2 patch for linux/posix

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Please try this ECW 3.1-rc2 patch for linux/posix

Bill Binko

Hi everyone.

I found what I think is a problem in the ECW libraries (version 3.1-rc2).  
Basically, they are not initializing a boolean attribute of a class to
zero (FALSE).  This means that if you test with the debug version of the
library, it works (GCC initializes it) and if you work with the optimized
version, it doesn't.

So, I put together a patch that seems to fix the problem.  I'm attaching
it to this message.

Please let me know if this helps, or if there are more occurrences of
this.  I have posted this to the ERMapper forum, so hopefully, it will
make it into the 3.1 release.

Bill

_______________________________________________
Gdal-dev mailing list
[hidden email]
http://xserve.flids.com/mailman/listinfo/gdal-dev

ecwhang.patch (2K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Please try this ECW 3.1-rc2 patch for linux/posix

Bill Binko

Sorry, I forgot to say what the problem it fixes is ....

When you run gdalinfo (or anything else that uses an ECWDataset), it hangs
when the destructor gets called about 80% of the time (on my box -- YMMV).

Bill

On Wed, 15 Jun 2005, Bill Binko wrote:

>
> Hi everyone.
>
> I found what I think is a problem in the ECW libraries (version 3.1-rc2).  
> Basically, they are not initializing a boolean attribute of a class to
> zero (FALSE).  This means that if you test with the debug version of the
> library, it works (GCC initializes it) and if you work with the optimized
> version, it doesn't.
>
> So, I put together a patch that seems to fix the problem.  I'm attaching
> it to this message.
>
> Please let me know if this helps, or if there are more occurrences of
> this.  I have posted this to the ERMapper forum, so hopefully, it will
> make it into the 3.1 release.
>
> Bill
>
_______________________________________________
Gdal-dev mailing list
[hidden email]
http://xserve.flids.com/mailman/listinfo/gdal-dev