> Putting the motions in the diversion was how I went about it first. > When that didn't pan out, I went the string route, which was when I > discovered the diversion needed an actual glyph in it in order to > work as I hoped.
Just out of curiosity, was displaying the image the only purpose of the newly defined character, or did this come up as a by-product of trying something else? If the former, then it seems a pretty roundabout way of achieving the purpose. For example, using the ps device, the following is sufficient: .char \[gnu] \X'ps: import GNU-head-small.ps 0 0 25 22 10000'\h'10p' A GNU head \[gnu] image. Can something similarly simple be done in pdf, too?
