The documentation clearly states that bz2.open(mode='r') opens a file in binary mode. I would have to use 'rt' for text mode.

The basic Python open(mode='r') method opens a file in text mode.
This is how I would expect any open(mode='r') method to work, especially since the standard Python implementation of open() treats 'r' as a synonym to 'rt'.

IMHO the behavior of bz2.open() is a possible pitfall for many developers and should be in line with the standard open() behavior.

Sorry, if this is in the wrong place, I am happy to put this elsewhere. However, I do see that as a conceptual issue (in our case this bug was not found for years)

All of the compression modules (gzip, lzma) have this behavior, not just bz2; it's consistent in that sense. Changing it now, after literally decades with the old behavior, would needlessly break existing programs. As you say, it's documented clearly, I'm not seeing a gain to be had strong enough to violate the existing documentation.

I concur with Josh and will mark this as closed.

It is unfortunate, but the time to get an API right is before it lands, not years after people have come to depend on it.

I understand and accept that but I would recommend highlighting this difference in the documentation a bit more since this information can be easily skipped when reading the documentation.

I still find it important to hint to this stronger since I've seen countless code snippets and answers on Stackoverflow that either tell the user to use "r" for text mode and "rb" for binary or - if listing file open modes do not mention the "*t" modes at all.

If you can point me to where to suggest a change (or pull request) for the documentation, I am happy to propose a change myself :-)