This is an important feature, which avoids significant degradation of system performance for large fast downloads and therefore I think that it should have its own column too. Petr Matas (talk) 17:18, 16 April 2013 (UTC)
This doesn't seem to me to be a distinctive feature at all, not to speak of a separate column. It's more like task-specific disk I/O optimization which is virtually essential for any good program designed to work with large amounts of data. Program authors think the same: from the 4 client feature lists I checked at random - uTorrent, qBittorrent, Free Download Manager and BitComet, only the last one even mentions this topic. — Vano23:07, 16 April 2013 (UTC)
Another argument against this is that OS cache disabling isn't the only I/O optimization technique out there. Large read-ahead and write-behind buffers, for instance, are even more common. — Vano23:20, 16 April 2013 (UTC)
One would expect, that any good program with its own buffers (covered by the Cache column) disables the OS cache, but many do not. From my experience I can say that it is a very prevalent programming flaw present even in uTorrent (v3.2.3), which you mentioned. That's why I see it so important. Should we mention this flaw (if present) in the Notes or Cache column? — Petr Matas (talk) 02:09, 17 April 2013 (UTC)
I wouldn't say it's an "important flaw". Large-scale reads and writes effectively bypass the OS cache anyway. And trying to outwit the OS in other cases as well is almost always unwise.
In any case, discussing the impact of the feature on system performance isn't relevant to the matter in question. What is relevant is whether reliable sources consider the feature major enough to be comparing on it. — Vano21:12, 17 April 2013 (UTC)
