As I understand it, the JIS X 0213 kanji standard was revised in 2004 with the effect that the recommended forms of several characters changed from simplified primitives back to the traditional forms. This does not affect daily use characters whose simplified primitives are firmly entrenched but it affects rarer characters with the same primitives, which makes things inconsistent.
An example is "crossing" in Heisig which reverts to a two-dotted road/movement primitive: 辻. I think this character is commonly used in the simplified one-dot form but since the official standard reverts to the old form, the fonts in Vista show it that way. Another is "briar": 茨. The "ice" part of the "next" primitive changes to something like an equals-sign with the bottom line slanted up, which corresponds to the old form of "next".
This is probably all good for historical purposes for Japanese scholars, but for my point of view, this is a bit of a pain. I wish standardization went in the direction of simplified forms rather than old ones but the trend seems to be going the other way. I think that paradoxically, the use of computers eases the pressure on simplification. Who cares about being consistent and making it easy to relate primitives? Just encode every odd form out there!
You can check a page such as
this for more details on such changes.
I suppose the upshot is it makes the study of kanji more interesting!
