cmusphinx
Reintegrate pocketsphinx-python code to main source tree
It's difficult to understand which version of the Python module is the most functional and least buggy, but a good place to start would be whatever is used by https://github.com/Uberi/speech_recognition - as this, in turn, is used by a variety of other packages.
Honestly I do not understand why anyone would use PocketSphinx for LVCSR, and I hope people won't do it, but I suppose for compatibility reasons we should continue to support this.
Funnily enough, I'm actually using Uberi's speech recognition. Mainly using google's API for when I have a connection and was hoping to use PocketSphinx for offline usage. It'd be really handy to know which version is the most stable for long-term deployment in my project
Good to know! The goal is to release PocketSphinx Python packages from the main source branch. Sadly due to namespace pollution we have to use a new name, which will be "pocketsphinx5".
I will be making a new release candidate quite soon which will get uploaded to PyPI under that name - the APIs from pocketsphinx-python are quite easy to recreate so they should be present there, unless there are some serious problems.
I'll keep an eye on that, it'll be really nice to fix the mildly infuriating(even more so because it's not maintained) namespace issue. Let me know if there's anything I can do to help!
Done since https://github.com/cmusphinx/pocketsphinx/releases/tag/v5.0.0rc3