This is the reason the corners are nice and square compared to the other image that was posted.
And in my first message with these images I emphasised that they are not for compare...
Only what can compare are my two images with each others and note how things change when samplinc period time/risetime ratio change.
Also you told
(and to be difficult to reproduce at home)
How this can be difficult. There is nothing difficult at all. If this is difficult then...<advance censored>.
Just simple basic simple generator and basic simple scope, one cable and turn knobs.
Btw, trigger jitter (when signal violate Nyquist rules etc) is other question but is was not shown with these two images. It start seriously rise when signal risetime what ADC can see is around equal or more if clearly less than sample period. Also this is natural (and makes full digital side trigger engine different if compare good analog side pathway trigger system in DSO/DPO it can nearly say (least if we talk today entry and middle level realtime digital scopes)
This can say is nearly like some kind of "achilles heel" of this kind digital side trigger system, not problem if user understand and can avoid this kind of situations in use or least that he recognize this situation)
In some special cases true dots mode can help, depending how its is implemented in system. But these are perhaps somehow explained in future in "lesson 2" if nothing more important is in "to do" list.
I want oscilloscope where is "brick wall style" filter in analog side (what is only place for it. It can not be in digital side) so that corner frequency is automatically (or manually selectable) adjusted with current sample rate and what user can switch on or off and stay with common "gaussian type" BW.
example about undersampling and Sinc horizontal time jitter (one kind of trigger jitter)
and other image same signal same trig but dots (randomly interleaving acquisitions, not ETS) and waveform shape and positioning without huge jitter)
This is why it is good to have opportunity to always turn all interpolations off (exept trigger engine internal "what ever"...). Just because only real data what we have are true sample points level and time and nothing else. Independent of what Keysight or who ever talks. There money talks.
In images of course they are still images so they do not looks same (specially dots mode) as it looks when human brain+eye handle sequential TFT frames. (also intensity gradation process have less position collisions in dots mode so it can not produce same as Sinc or Linear. But scope is tool, not for beautiful images generator.)