Check if file exists during hash check

[mb300sd]

I am getting a bunch of Readfile errors during hash checking on a 70+GB torrent. Would it be possible to have the client check weather a file exists before trying to read it? It is currently taking over 3 hours for the hashing to finish, and utorrent is using 25% CPU on my quad core. I am using build 3.0.21663

[paintball9]

It already does. If the file doesn't exist it just marks it as 0% and moves on.

Read errors would imply that something changed on that file after it finished downloading.

And BTW, hashing is a very CPU intensive process, It will compute as fast as it can read it off the HD.

[Firon]

Does 2.2 have this behavior as well?

We'll get this fixed soon in any case. Thanks for the report!

[mb300sd]

It skips the non-existent files properly until it encounters a file, then will try to read all files after.

For example, a torrent containing 6 files, around 12GB each, where I only want to download file 4

if only file 4 exists, it will instantly go to 50% completed on the hash check, check file 4, and then continue trying to read each piece of files 5 and 6, both of which do not exist.

Also, I'd like to suggest adding the option of not trying to check files that are skipped, maybe as an advanced configuration option.

[paintball9]

If you don't check files that are skipped, you would end up losing the pieces at either end of the one's you do want. Requiring that they be re-downloaded. Optimally it should save the data inside a temp file in that directory, (unless you have pre-allocate all files checked off)