tag:blogger.com,1999:blog-3062430.post7059158001433788514..comments2008-12-29T11:51:52.755-08:00Vichttp://www.blogger.com/profile/03074915000996247481noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-3062430.post-88913130131251084932008-12-29T11:51:52.755-08:002008-12-29T11:51:52.755-08:00Brendan, are you saying the results were different for different runs on the same data?<BR/><BR/>Would it be possible for you to send me some of the files that were reported duplicates?<BR/><BR/>The script actually returns whether the md5 hash of files match. I suppose with diverse enough data there could be some false positives, but it's pretty unlikely. I can add the final comparison check to eliminate these, if this is what is actually happening.Vichttp://www.blogger.com/profile/03074915000996247481noreply@blogger.comtag:blogger.com,1999:blog-3062430.post-42269608363846330582008-12-29T06:00:48.985-08:002008-12-29T06:00:48.985-08:00Hi,<BR/><BR/>I applaud (and appreciate) your efforts, but this did not function as expected on my WinXP system. I ran it under Python 2.4 on a directory full of tiffs (and some other file types) that were band images from IKONOS. It detected many duplicates among bands, i.e., for a given tile, there are 4 bands, and it would detect three as duplicates of the first. Not all the time, but much of the time.<BR/><BR/>I checked the results with the DOS comp command, and it says they are, in fact, different. In short, it doesn't seem to work.Brendan Hemenshttp://www.blogger.com/profile/00721568107945654992noreply@blogger.com