More out-of-the-box thinking

I was thinking about rotating parts on the slide scanner over the past few days, as I was writing here. That got me thinking about the hardest part of the whole mechanism: the slide feed apparatus. The purpose of this is to take a single slide from a stack of slides and dispense that into the rest of the works.The key point here is a single slide. This is challenging on a couple of different fronts. If you're purely sliding the slides past each other there will be occasions where the slides will bind up with each other; you have each slide with a large opening that needs to slide past another one. Any distortion of the slide will cause these two edges to basically interlock and jam the machine. Either that or you'll have a double-feed of two slides that will similarly mess things up. Another difficulty is that having a long sliding interface where you need to move a shuttle that moves the slides at least four inches is problematic from an automation perspective.Which is why I started thinking today.I decided to game out other ways that I can manipulate these slides. What are the rotational axes that I can work with?X, Y, and Z. Two of these are effectively equivalent since the slide is square. At first I was thinking about making something works like the scanning assembly does things, with a roller mech that tips one slide over. But this is actually really hard to do with the gear I have. You'd wind up with tipping over a few slides.Then I had my latest brainstorm: rotating the slide around its axis!From some initial tests, this will have no binding at all! I just need to construct a slide carrier that's thinner than the thinnest slide, I'm figuring around 0.04" or 1mm thick carrier, that rotates a slide 45º into someplace it can drop into. Only think I need to do is prototype that and I'll be set.As an added bonus it's a lot smaller and I can use a simple rotational servo to automate that if and when I want to automate it!  :-D