Fatal I/O issue in 6.22/04

Hi there,
In the latest announce for 6.22/06, Axel wrote:

The previous patch release 6.22/04 was not announced due to a fatal issue in the I/O subsystem discovered right after tagging. Please don’t use it, but move directly to 6.22/06.

Our team and me have already processed some data with 22/04 so we are wondering what’s the IO issue you’ve been talking about, @Axel? Currently it’s a bit unclear for us should we reprocess our data or not. Some details about that IO issue could hopefully clarify the point.

Thank you!


Hi Timur,

The problem was with std::pair where in some cases it was getting the offsets incorrect which could results in incorrect read/write for the std::pair and potentially std::map.


This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.