kerning in slanted/unslanted founts

[Vladimir Volovich]

"AJ" == Alan Jeffrey writes:

 AJ> OK, point taken.  But, as Lars also pointed out, there's not much
 AJ> fontinst can do here, since it only `knows' about bounding boxes.

I noticed that charwd values in cmu10 are ~1.085 times wider than in
cmti10. So, maybe the kern differences are be partially due to this
"xscaling". But again, should fontinst change kern values when scaling
transformations (e.g. \xcalefont) are made?

	Best regards, -- Vladimir.