Ha, understood - indeed I don't remember it happening during earlier beta periods, true.
- - - - -
Back to comments !
I find myself using the Image feature a lot as I mostly use Moi for 2D drafting these days. It works great overall but there are a number of little details about it that could be smoother.
- I think it would be great to be able to interact with images present in the scene without having to go back to the Image tool manually. Maybe some key combo allowing to select the image planes just like regular objects (key+click), or some clever mouse input (triple or quadruple click), or some UI widget which would enter Image mode from there.
- Once in the Image tool there should be a way to duplicate a current image. And even better, doing so in the scene state it is currently in. Either by somehow interacting with the image object itself by copy/paste, or through a duplicate button. Without that, making a duplicate of a current image involves importing it again and placing it in the viewport again, which is annoying and breaks the flow of work. At the moment if I need to have, say, 5 duplicates of a rough draft to make 5 clean variants from (which, as you might guess, is precisely what I need to do for some design work at the moment :D), I need to import and place the rough 5 times. I should really only have to import once and then duplicate from there. It seems minor but in practice it is quite a blocker, especially since each time the image needs to be "drawn in" from corner to corner meaning that for a perfect duplicate the user needs to remember to enable object snap too. It really does pile up :)
- The "Transparency" value (with 90 being very transparent and 10 being near opaque) is counter-intuitive. It should be labelled Opacity, with 10 being very transparent and 90 being near opaque.
- The numerical input field is error-prone, especially for fast users used to keyboard input. I understand that it is the same as everywhere else in the program and that many users are probably used to it, but imho it could behave much better (of course I don't know which part is tech limitation from the UI API and which part is intentional behavior here.)
First of all, there is really no need for the big keypad to show up for anyone using a keyboard (which I assume is 99% of the user base). It seems good in theory but in practice it is laggy (shows up a few milliseconds late) and visually distracting ; and any power user will just type with the physical keys of their keyboard anyways (either top row or keypad). I believe the visual keypad should be an option, probably best turned off by default. It is also stealing the cursor (making the caret disappear) which breaks the visual confirmation that the field is ready for keyboard entry so that's another way it goes against the habits of power users used to the subtle but universal visual clues of any OS.
Also, the visual keypad shows up when a field is highlighted and disappears when it isn't. It "makes sense", but it is strange for such a big UI element to be turned on by a positive input that is usually not a toggle (click), but *also* turned off by the same input (another click, removing the field highlighting). It gives the impression of a UI element that is always "just about to break" so to speak, which is problematic UX-wise.
An alternative (on top of the option to have it show up by itself or not) would simply to have a button that opens it up (see mockup). This would have the advantage of being more practical for tablet users, since hitting big buttons on screen is easier than hitting text entry fields.
Lastly in the case of the Image tool in particular, the entry field sometimes fails at displaying the current Transparency value of the currently selected image in the list.
I hope this helps !