Carine Pivoteau wrote:
While working on a proper complexity analysis of the algorithm, we realised that there was an error in the last paper reporting such a bug (http://envisage-project.eu/wp-content/uploads/2015/02/sorting.pdf). This implies that the correction implemented in the Java source code (changing Timsort stack size) is wrong and that it is still possible to make it break. This is explained in full details in our analysis: https://arxiv.org/pdf/1805.08612.pdf.
We understand that coming upon data that actually causes this error is very unlikely, but we thought you’d still like to know and do something about it. As the authors of the previous article advocated for, we strongly believe that you should consider modifying the algorithm as explained in their article (and as was done in Python) rather than trying to fix the stack size.