Closed josephwb closed 3 years ago
Trying to get feedback from the original/NCL authors on what is best to do. I think ideally NCL would support both forms, but they might be hesitant to do something that was not expressed in the original paper (which is a simple switch).
FWIW on ingest phyx supports both flavours. So it is only export that we need to worry about.
Hrm, to be on safe side just removed INTERLEAVED=NO
from write_nexus_alignment
with eea8dfc.
The keyword INTERLEAVE is a switch to turn on. INTERLEAVE=NO is thus not supported, and any program that uses the NCL (e.g., iqtree) will error. MrBayes, however, allows interleave=yes and interleave=no, and this doc suggests that it is valid usage. FWIW PAUP* opens such files fine.
I guess we should remove any
=yes/no
from this keyword.