Hopefully someone can provide an answer to my question.

I wanted to know about SHA-1 hash values on the gnutella network, specifically with regard to partial file downloads. As I currently understand, if I have a partial download of a full file (let's call it file x), I can share that partial piece of file x on the network to other peers.

If file x has a unique SHA-1 hash value, and I have a portion of file x, (which itself would have a different SHA-1 hash value?), how does my computer broadcast to other peers that I have the SHA-1 value of file x, and not the SHA-1 value of the portion of file x? Is this even possible or am I just misinformed about how things work on the gnutella network?

Thanks

There are several hashes:


1. The file hash, its the "id" of the file, its the sha1 what you get when you search files.

2. Tiger Tree Hash. For detecting errornous data and localize where in the file. The aim is to minimize the amout of data for retransmitting when a part of the file is damaged/wrong.