1999.06.18 21:35 "Possible bug found in tool "tiffcmp"", by Eric B. Middlecamp

Hello everyone. I've just recently joined the mailing list..

I've been making some modifications to the libtiff tool "tiffcmp," and I think I've found a bug. When the tool compares images scanline by scanline, it first does a memcmp over the whole scanline, and if different, goes into a loop to find the specific difference. Some of the time, I've noticed, the tool will find a difference in the memory step but not in the loop step. I believe that the loop is only (1/samples per pixel) as long as it should be. The loop is currently over the number of pixels in the scanline, but it is actually incrementing in bytes.

I was just wondering if anyone else out there has used, tested, or modified "tiffcmp" and noticed similar problems.

Thanks,
Eric Middlecamp
Intern, Imation Corp