Re: [Mldonkey-users] Re: wrong md4 digest

[MLdonkey]

>  I assume the reason for doing it this way is that edonkey works using
>  these 9M chunks for verification, etc (if you look in the
>  shared_files.ini for mldonkey you'll see the MD4 hashes for each chunk)
>  so it has to calculate the hashes for each block as it downloads.  It's
>  a lot quicker then to generate an overall hash from these rather than
>  having to go through the entire file again.

It can also be used to verify that the Md4s of chunks are those
corresponding to the file you want to download. When you start
downloading a file, you only have the md4 of the file, not of the
chunks. Therefore, someone could try to corrupt this data, so that you
never manage to get the chunks, since that, even when they are
correct, their md4 is different from the (fake) one. Hashing the hashs
is thus necessary.

- MLDonkey