FIG may have at one time been meant to repleace IMG (I certainly liked
the idea), but that's not currently the case.
> But i might just be missing the obvious. Is there a solid rationale for this?
IMG is meant for "inline" graphics, in the sense that they go in the
line of text. For this case, the limitations of ALT for non-GUI
browsers is bearable, though you wouldn't do it that way if you
weren't shackled by bad prior design.
FIG is meant for graphics that are not part of a text block (which is
why they aren't allowed in a <P>), but stand alone. Since there are no
documents to break by doing it right, it can be done right. This means
you get captions, client-side image maps, and such goodies as well.
For those who haven't noticed, the latest version of Lynx implements
FIG. You get the alternate text (complete with anchors), the caption,
and a link to the image the figure references. Works quite nicely.
Now, if only popular GUI browsers would implement it, so we could take
really advantage of it.
<mike