bug in fontinst 1.801: duplicated kernings (was: ae fonts Q)

[Thierry Bouche]

Concernant « Re: bug in fontinst 1.801: duplicated kernings (was: ae fonts Q) », Alan Jeffrey écrit : «
» Vladimir Volovich wrote:
» 
» > BTW: what do you mean here? Why/how are these kern tables
» > compressable?
» 
» If a font has the same kern pairs for M and N (which, let's face it is
» pretty likely) then fontinst will just generate the duplicate ligtable
» for both, rather than combine them into one ligtable.  This makes a big
» difference for some commercial type 1 fonts, which seem to have kern
» entries for just about every pair of characters.  

which is typical of computer generated pairs (like URW's kernus).

» It's especially
» noticable in small caps fonts.

but here most of the pairs are created by fontinst itself, no ?