Created attachment 660 [details]
one zoom level

Protocol decoders can provide a set of text variants which differ in their 
length, to have the application pick one of them which fits the available 
screen space for an annotation. It's a known contraint that decoders should 
provide the list of text variants in strict order of decreasing length. The 
application may rely on that order.

It was observed that pulseview can pick a text variant which is slightly 
longer than the available space, and then will truncate the text and add a 
"continuation marker" (a sequence of periods). Which results in the absence 
of data values after their explanatory prefix. In unfortunate circumstances 
several zoom levels can suffer from that issue, and several adjacent zoom 
levels keep hiding information from the user.

See the four attached screenshots for an illustration.