Repeating an edit in which tab-completion was used
Here's an example text
... the measure is aggregated across subjects. ... the effect of the manipulation is expected to be the same in all subjects.
I wanted to replace "subjects" by "participants". I moved to cursor to the beginning of the first "subjects" and did
cw an typed
part<tab> which completed to "participants". Then I moved the cursor to the second "subjects" and pressen
. to repeat the edit. This time "part" was expanded to "particular" because the expansion mechanism that I'm using is context-sensitive. My expectation was of course that
. would result in the exact same substitution as before. I'm not 100% sure if this should be considered a bug or a feature but my impression is that it breaks the semantics of
. but you may disagree.