I had, and continue to have difficulty downloading similar files. What appears to happen is that if I attempt to use HTTP in a browser (either Mozilla or Konquerer) to download a file, the file size ends up being about 3 or 4 times as large as it should be, and what's supposed to be a, say, 500K file ends up being 2M. This behavior unfortunately seems to be fairly consistent, and is especially disappointing because it generally makes any d/l attempts from Sourceforge an exercise in futility, at least for me. Heck, for big files, if I keep an eye on Download Manager, I can see the byte count start zooming upward as soon as the expected download should stop. Not too long ago I tried downloading a file that was supposed to be something like 9M, and much to my surprise once the progress bar had reached 100%, the byte count immediately started to increase all the way up to 28M, as if it were somehow decompressing itself on the fly. If anyone can explain why this is happening, or if I'm making some kind of obvious bonehead move, I'd be very interested.
Otherwise, as the previous poster mentioned, this behavior does not seem to affect FTP downloads. One good source for packages that seems to work pretty well for me is:
The files are in .tar.gz format which is a little more of a hassle to deal with, but otherwise this works reasonably well for me. -- J.W.