I'm thinking that Sprites are going to be old-hat, anyway, by the time the jump is going to be made.
Procedurally-generated variations of vertex-defined body-shapes, with the hints from the generated body description (or the data stored that does output "...its front legs are cyan, its second pair of rear legs are puce") augmenting the bodyplan (from full describing that creature, or from the components grabbed from the body_rcp.txt for the weirder constructs), and perhaps with a little extra description by the time this aspect is developed. And augmented by the "inventory" and "wounds" details.
This means that thin and fat dwarves can be seen as thin and fat dwarves. The one carrying the silver pick will be obvious next to the one carrying the iron one. The one with a cut on his finger will have (at sufficient zoom) a stain or mark there... or even periodically suck the wound! Unless it has been dealt with by a doctor in which case a dressing might be visible.
Pregnant creatures might even be visually detectable, starving creatures will show their ribs more, a creature (or dwarf!) that has just run away from something might be seen taking some deep breaths.
But not yet. It's not necessary. But at some point we will be each working with three primary main processor cores, cross-linked with redundant melacortz-ramistat 14-kiloquad interface modules whose core elements are based on FTL nanoprocessors with 25 bilateral kelilactirals, 20 of those being slaved into the primary Heisenfram terminals...[/nerd]
What year is projected for that type of technology?
And that is what I'm talking about. Getting a program to visually show details. As long as someone can create a uniform detail for all the changes being output it can work really well. And you only have to modify it so it only reads output you place for it to read. This way it can be easily modified by fans and made a group project.
For instance, at the start you don't need to have all the dwarves shown, or even some of the clothing. Perhaps most clothing would use a place holder. The same as other images. I realize that part, you'd need to show detail for large amounts of activity change. Or have place holders to use.
Well I'm going to start my arduos task into the world of C++ and see what can be done. Maybe one year I'll do something. Anyone know what stonesense was made with? Is that with C++?
EDIT: One Idea I have, is how to illustrate the various effects in combat.
I don't know how you would show a human, dwarf, or something wrestling another entity. The combinations would be numerous. But let's say you did all of the work. If you're attacking something that is on an adjacent tile how would you best go on about visualizing a fight that includes wrestling? At first you wouldn't be able to show all of the output effects. Like using right hand to strangle. So it would be a matter of showing a different image based on the highest tier of output. So in combat I would have the image show you strangling or holding up your hands if there was any hold on the creatures neck. Then I would need to change the image for the creature to its being-choked-by-one-hand animation. So if you tried it with the left arm, it wouldn't show. But that is something that could be made a placeholder for any choking animation until the rest of the work is done. So it would still work. Then I'd have one for 2 hands. Now the problem is attacking while wrestling. I'm unsure about how you could show attacking while wrestling. I would guess when you have the output for any wrestling command, it changes the image. But how would you make it so when you change one image, if it has a certain output, you would fill the requirements for a different image showing. Or not showing wether or not that output is there?
For instance, you show the character change to a grappling image. Then you show the entity being grappled in the adjacent tile. So now you stab the enemy with a weapon. If I wanted to show you stabbing the enemy while still grappling, and have a different one for when you are not grappling. Could I tell a program to an image based on those requirements? Special images?
Also, is it possible to have stance change based on the activation and deactivation of some events? So your images don't get stuck.
Of course I'm thinking about all of this in the way that each image is shown based off of information in the game. Not in how you can go about certain images. You could in fact have image colors change based off of a sprite palette and have colors change to match the data in the colors. Such as Jagged Alliance 2. Then you could just sprite the image one way, in many different ways, but only once. And the coloring can be changed to whatever colors certain items are represented as. So you could have colors for cheetah print, have colors for all sorts of things.
But that's out of my league in terms of sprite animation and sprite programming. Or Image manipulation on any scale.