Hello Everyone,

As a file is being downloaded, due to various causes (upload host crash, power failure) the last few KB downloaded will get corrupted. When this happens it assures this file will not be resumable. Gnucleus compares the last few bytes of your previously downloaded file with the first few bytes of the incoming file, making sure they match exactly, thus assuring it is the exact same file being downloaded and no corruption has occured during the disconnection. Obviously, if the last few KB of your file is corrupted it will not match the new uncorrupted incoming file and the transfer will be cancelled. If you have ever had the "Extended Info" window open and seen "Remote File Corrupted", chances are, it's your file that is corrupted and not the "remote file". This is not always the case however, it could be totally different file that fits your keywords, and happens to have the same byte size.

If you think you have one of these corrupted non-resumable files, I have found a freeware app. that allows you to "tear" the last few KB off the end of your file, leaving you with a file that is resumable again (but a bit smaller). I have tried it, it works. Make sure you tear off the end of your file and not tear a chunk out of the middle. Use Explorer on your file to right click=> properties and get the byte size of your file to determine where and how much to tear off.

Download "Japos Tear File" at http://www.japos.com/

Swabby, would it be possible to have Gnucleus show us an exact byte size in searches, instead of just KB. Is this info not available until a download is actually started? With this info wouldn't it be possible for us to create 0 byte resumable files using "File Name (size in bytes,keywords ext).ext"?

These corrupted partials do happen, but rarely.

One method I have used to find these corrupted files is to check the file size. Win2k reports the file size and size on disk. If these two values match exactly, it most likely is a corrupted file. I use fat32, this may not work on NTFS.

I wish Gnucleus would flag, as suspicious, all files that have been resumed and cancelled due to corruption issues.

I have a large file which I think is corrupted under properties the file size in brackets matches exactly the size out side the brackets.

Also many for many hosts I get remote file corrupted.

Could you go through exactly how to fix this so that I can continue the download (I am on a slow connection and the file is a biggy)

I have tried using the file tearing utility before on other files with no luck.

Hi

Create 2 new directorys on your drive, "Corrupted" and "Scraps".
Cut and Paste or Drag and Drop your corrupted file into "Corrupted".
Copy the entire file name to the clipboard.
Open the Properties dialog and write down the file size ( in bytes).
Start the file tear utility
Click the button under "Source File", browse to "Corrupted", select your file, and click "Open"
Click the button under "Save Extraction...", browse to "Scraps", paste file name into "File name" field and click "Open".
Click the button under "Destination ...", browse to whatever directory your Gnucleus partials are in, paste file name into "File name" field and click "Open".
In the "Tear from..." field enter the file size, in bytes that you want to keep and try to resume from.
In the "Count" field enter the file size, in bytes that you want to remove.
The "tear from..." field and the "Count" field together, should equal the file size of your corrupted file.
Maybe someone else can tell you what the minimum amount you should remove would be. I always just trim back to an even number.
Example, corrupted file size- 123,456,789 bytes "Tear from"- 123400000 and "Count"- 56789
Click "Tear file!". When it finishes creating the new files you can delete the original corrupted file and the scrap file.

Questions? Just ask away. Good Luck.