> >well, restricted or not this is right now what a few million papers do > >use :-) anyway, if you look closely at the draft proposal (p33) then > >you see that MSP contains a full upper and lowercase script/cal > >alphabet > > In fact, one can write mathematical manucripts using only a typewriter, > and in fact some very good mathematicians do (or did) just that. So what > people actually do use and get along with is not a good indicator for what > to include in a typesetting program of the future. have you overlooked the smiley? what's the saying: a million flies can't err? all i was commenting on here is that it isn't as bad as in the days of typewriter and i think we can agree on that. > So why not add bold/bold-slanted versions of those script fonts then, > regardless whether the stuff ends up in the AMS-fonts or whatever? If this because bold and slanted do not normally have to do with encodings. just as we don't have bold and normal weight chars in T1 but we use two T1 encoded fonts for that. if you would like a bold script then you would use a second MSP encoded font and your done. granted you then waste some slots because also the other glyphs of that font are bold and you might not want all of them but then one can have only one alphabet per font if the glyphs are to be accessible directly so putting more than one into a font is not a good idea for that reason > works for a first working wersion of the proposal, it is probably no match > adding a more scripty font in the final proposal. so the proposal already allows for having bold/bold/slanted versions and there is no need to *add* them to the encoding proposal > It would be nice to add a lower-case blackboardbold letters too (these > are in AMS-fonts). but they are in the encoding proposal: MS1 has 64 glyphs bbb glyphs (p.34) and MS2 a fraktur alphabet > >once we have gained that experience i think going back once more to > >the drawing board might be fruitful > > If the idea is to try out how the current work, it would perhaps be good > idea putting in an extensible arrows package. This sounds like a rather > conservative addition to me. :-) if you read carefully you find MS1 (point 14 page 35) Alan's arrow construction set as a possible suggestion for inclusion > However, I felt one should look it over and make sure it need not be > changed by such things that people want to develop complete sets of fonts > (with the bold/leaning/bold-leaning versions bundled to the plain shape). > > The reason is this upright-for-constants-leaning-for-nonconstants > principle, that seems to not have been explicitly brought up before. but it has --- during the discussion on the math font list that resulted in the proposal --- but it is conveniently be solved by additional fonts in the same encoding (and yes kerning is a problem but that was looked at as well and within TeX you are limited to kern only within one font so there is unfortunately no way to get *all* kerns you ideally want. for this we have to wait for omega) frank