[mdp]

Finally the new device being out, Max2, new possibilities arise and old ones improve.

For reference my experience of the HDMI connection as written in the parallel thread about the device in general:

I tried the monitor, HDMI connected, since the kind folks at Onyx are so kind to distribute the HDMI -> HDMI-mini cable with the box.

Now the question was: how well do you find your way using the mouse. And the answer to me is, well not as well as with traditional systems, but it is usable and you get used to it soon. The discriminator is if you forget the issue during the use, and I yes did forget the non-usual experience with the mouse and started using it naturally. Like managing the voice delays in a Voice over IP call with non perfect connection, better than many VoIP calls with imperfect connection.

Some lag is there, I did not compute it: but like a call it is not a problem if both parties agree to respect turns.
So, Cuphead looks very nice but you will not beat the boss ; )
Many other contexts are certainly adequate.

The "monitor" experience
-- either requires an extra setting that updates a screen cell only if the changes are above some threshold, or
-- requires insuring that changes in some area of the screen do not involve dithering changes in areas that did remain unchanged in the original rendering. In other words, it seems that if on the screen an area is changed, other graphics in the EPD is changed, in areas which are instead static, because the computation of their dithering changed.
The problem with this, is that the battery would be drained uselessly.

[...]

Onyx's monitor solution (software + hardware) works very, very well.
You plug the cable, you see a "Toshiba" monitor in your park - I tried on several OS, no problem. There, you can extend, mirror etc. from the source desktop OS.

Yes, the experience on HDMI + monitor software surely, after experience, beats VNC /on the weak MaxCarta/. But to use remote desktop solutions on the Max2 - VNC, RDP, NX, TeamViewer - will be very interesting and there is ample chance to obtain productive environments.
So, remote desktop solutions will work in more cases (network distance etc.) and at the moment could reveal more flexible.

[...]

Hello, I am writing this message from the Max2 used as a monitor through HDMI (I needed light to test). Again confirm it works well, actually.
Now, latency is a potential for problems, so I tried Braid. Looks beautiful, and works well, and it is usable (you can actually play it with some degree of control)! Not sure I would finish the game though. The test is to see the accuracy of the reactions. So, taking Cuphead as a "impossible", Braid looks doable. To those who do not know what the two things are: I do this to verify how well you control an interaction.

The mouse is, in real experience, quite usable. Other tricks I earlier suggested, like the "mouse cursor trail", may enhance the experience (cannot try, I am on Linux) - but they are not required. You just get used to it. The annoyance is (for those who never worked with laggy monitors, typical of remote contexts), precision when pointing to small items, such as the corner of the text edit area in this real time case - but even in these cases, really, it's fully doable.

I can identify three lacks in the current implementation, that as shipped with the device:
- the Monitor software only uses A2 mode (of course A2 with dithering), but Normal mode would be really, really welcome, and there is no use to do without it. It is like in all other situations: A2 is good in case of large moving parts on the screen, normal is good to properly eye details on the screen. The ability to switch to Normal mode should be added.
- The dithering is dynamic: a static screen continuously changes its dots, because a grey bar is rendered with always differently placed dots. It's like a bee swarm. This is why it consumes battery! It does, but there is no need to do it! This must be just an intermediate implementation. If a bar in frame 0 is rendered as "1 0 0 1 1 0 1 0 0 1 0 1 1 0...", that dithering should remain identical in frame 1, frame 2, frame 3 etc. until that bar is modified on the screen (translated, scaled etc.). "Static screen" should imply "identical rendering".
This also gives the effect of a pulsating screen.
- it would be useful, although not strictly necessary, to have a software contrast enhancer in the Monitor application itsefl. Of course, similar settings are also available in the desktop OS, so it's not really needed.