Thomas S. Binns

The code for annotations works fine, but changing `_convert_times()` has some knock on effects for converting data objects to dataframes. This is an alternative approach to fix the current errors:...

Hey @drammock, please could I get your input on the proposed changes. The original annotations issue is resolved, but there are some implementation details that weren't discussed in the issue...

> The code for annotations works fine, but changing `_convert_times()` has some knock on effects for converting data objects to dataframes. > > This is an alternative approach to fix...

@larsoner I'll have a look through this again tomorrow and refamiliarise myself/start trying to fix the failures. Then will see if I have some more concrete questions then!

Okay, so everything is green now. This PR changes it so that nanoseconds in times will be dropped on reading annotations. As noted here (https://github.com/mne-tools/mne-python/issues/13108#issuecomment-2648331187), sanitising on read will prevent...

Yeah, I can do that. But at the moment we would only set `drop_nano=True` for `Annotations -> DataFrame` conversions at this time, correct?

Okay. I haven't seen any issues reported with nanoseconds in dataframe conversions of other classes, so I'll keep it simple and stick with enabling this only for annotations!

Just pushed a commit that drops nanoseconds only for `Annotations -> DataFrame` conversions. Should be everything sorted once tests are green.

That's @larsoner's comments addressed. Here's an example of what the warning about bad `orig_time` as input to `Annotations` looks like now: ``` RuntimeWarning: The format of the `orig_time` string is...

I have some time to look into this now; wanted to get some thoughts on the best way to proceed. So currently, the max image width in `sg_gallery.css` for the...