In this version, the processing of concatenated files has been modified. This is potentially a breaking change, so check that you queries are still producing the expected outcome (see details below). The job names, and hence the cache file naming system has also been modified to reduce the risk of collision, use shorter filenames, and ventilate the files into subfolders to ease their listing. As a result, it is highly recommended that you cleanse your entire cache. To do so, execute: vt_server_cache.py -l 1996 The code in vt_server_brain.py has been largely cleaned and simplified to reduce code duplication. Finally, the channel_patch module has been added (see https://egaudrain.github.io/VTServer/vt_modules.html#channel-patch). Sub-queries Prior to v2.3, when an array of files was provided, the stack was applied to each file before concatenation. In principle the stack had to be an array of arrays, and if the stack was a simple array, it was duplicated to match the number of files. From v2.3, the file attribute can be array of strings (sound filenames) and sub-queries {...}. The stack is now applied to the concatenated result. Roadmap Some groundwork preparations have been made for the introduction of generators that will happen in v3.0.

If you use this software, please cite it as below.