Boggis, I appreciate that you're trying to find bugs and you probably have found several.
Confirmed several, found the video line bug. (Difficult to check without a calibration source.)
However, in the case of not displaying a line when you go to AC-coupled triggering, I am with rf-loop that this is not a bug - it should show a numeric level but no line on the screen. Watch Dave's video above which is absolutely on point, particularly after 17:17 if you want to save time.
The way I have seen other 'scopes handle this is to show the level position as you adjust it. (I don't see why you'd want to set it, with AC coupling, really.) We had a Tektronix in today (DPO 2024B) so I had a quick look at this and it does allow trigger adjustment when AC coupled, and also indicates the level on the display.
I think that Siglent should stick to displaying a trigger level position on the main display if you are able to adjust it. If you know of other 'scopes that don't indicate the level, but still allow you to set it (using the numeric indication only), then perhaps this is a convention -- but it just seems broken to me.
I'm waiting for the new firmware but meanwhile, I'm using the SDS1202X-E and it is doing just fine; I have yet to experience a bug that's stopped me doing anything I have wanted to so far.
Some people have obviously been experiencing more problems than others. I haven't had to use the 'scope on anything yet and have simply been familiarising with it. Nothing I have planned should present a problem for it.
This isn't a high-priced instrument, so you can't expect it to be as polished as similar models costing multiples of what this one does.
I have one other thought/theory about the phantom blip you see in the center of the screen when nothing is connected to the input; digital cell phones pulse RF energy out (try holding one near your speakers) and, even with nothing connected, even with maybe a scope probe connected where the tip is shorted to the ground croc-clip (which forms a small loop antenna), a real signal pulse is being detected which is causing the triggering. Putting a shorted BNC connector on the input might not even make this go away as the PCB trace inside the scope between the BNC connector and the ADC front end might be enough to pick up a signal. However, I see this as an exercise in futility if you're looking for bugs with no input connected! My Agilent and Fluke multimeters show fluctuating voltage readings when I have nothing connected to their probes
No, that isn't the case, and isn't what I am concerned about (a noise source isn't a repetitive signal, so I think I see why you're thinking about interference from a signal source).
It seems that I am not getting across my main points about this behaviour. There is the change in behaviour, suggesting some parameters change from default, and there is the rather odd signal positioning. Only Siglent know why the behaviour shifts (some sort of retriggering with new parameters would be my guess) -- and it is probably not going to be an issue for real use, particularly if you do set the trigger. (Poor coding is still something worth fixing.)
The signal issue seems to be an artefact of how the 'scope samples. If you look at the 'dot' display then you can see that the samples are not uniformly distributed, and you end up with a division at the centre (trigger location) where the triggered samples are stacked together. My present assumption is that this is just an aliasing effect, but I do wonder if this would present problems for real signals. (The Tektronix DPO 2024B didn't behave like this. There is no 'dot' mode available, though, and the behaviour is generally different. Apples and oranges, perhaps.)
I have a lot of experience working on 'scopes for verification of specs, but that is quite different from poking about in odd corners and uncovering behaviour that may not be documented. For many things, unless something is documented you can't really tell if it is a bug, or intentional behaviour.