Comment 15 for bug 550300

poy (poy) wrote :

here is an updated patch, test please.

i have a few questions:
- DC++ 0.762 has been released with the wrong variant of resetPos, can it cause corruption when flushing fails? in that case, better not make it stable.
- this bug isn't fully fixed until some patch like this one i'm submitting is applied, right? or are compression / other kind of failures not a problem when flushing?