I like #3 best. (Naturally
- - #3 does not commit to what activity the user intends with the Compare. After all, XnView Compare is also useful for quickly tiling two or more images into the display a la Lightroom's Survey mode.
- The Eye means "look at" or "eye see you". Since it is already a "verb" symbol in the icon set, for me Mezich has firmly established the former meaning. Therefore #3 is easily interpreted: "look at two images side-by-side".
- The Question mark means "Ask the user something" and implies that XnView wants the user's input, such as in the new "Save As" icon. XnView does not ask the user anything during Compare. The question mark is in the user's mind!
- The Equals sign is ambiguous, both visually and for meaning. Is it an action ("make these equal")? Or a declaration ("these are equal")? This ambiguity is the reason several of us here want to pair the Equals sign with the Question Mark. It is not enough by itself.
The Mezich icon syntax is well planned. There is the action/verb vocabulary (Question Mark, Eye, Green Arrow) and the context/subject/predicate vocabulary (Orange Document: before/source, Blue Document: after/destination, etc.
Some times a symbol is naturally monadic
(undo, redo, save) but most of the time it needs a context and an action (Save As = Document + Question Mark).
But a symbol is NOT a full narrative. So the action and context components (sub-symbols) themselves need also to be monadic if possible.
One reason for this is that the human mind is supposed to identify
the symbol, not "read it". "Reading" slows you down. If one had to read
a Stop sign at every corner, consider the consequences...
The second reason for this is simplicity of appearance. The more visually complex the symbol is, the longer it takes for the brain to interpret the distinctive features. If our icons were 48x48, then the 2-part action components (Question Mark + Equals, for instance) could be more easily distinguished by the eye. But on my monitor (1920x1200 24"), they merge.
Question Mark + Equals isn't too
narrative-like, but the visual distinctiveness of the entire icon (#2) is compromised by having too many "action" components.
I have more to say but...