Hi Ruben, > Well, I'm new to this list, so my comments on templates might already have > been discussed, but anyway, here goes no they haven't been (at least not on the list); while writing template.dtx David and I had some discussions on the topics you raised but we decided to defer them (and any potential interface for it) until after we got some more real templates/instances available. To get a feeling for what is needed in practice. will only deal with the (in my opinion) easier part of your message tonight: > * xinitials > > 1) You should use `totalheight', not `height' --- just try Palatino J, or > just about any Q. (BTW, this mistake is also made by the dropping package). i beg to disagree. point is that for certain letters \height may not be right but in my opinion \totalheight would be worse. In most circumstances alinging the baseline of the initial with a bseline of the text is a good starting point (even if the initial has decenders) --- and if it is not adequate the template offers ample possibilities to adjust the placement. > 2) Accented characters can be used for initial if you just use > \meaning/\@onelevel@sanitize or a plain \string (compare the code the > accents use to check for existing `template instances' [=composites]). True, but then either \DeclareInstance would need to do the same or the whole thing would not work either. And I'm not that keen on allowing anything in template/instance names but would rather prefer a lean name space there. It kind of feels wrong to me to open up that name space just to be able to support \"A as an instance name --- but convince me otherwise :-) > Neither English nor Dutch use accented initials/caps, but languages like > German, Danish, Swedish, French, Portuguese, etc. do (well, I think so ;-) > So I think support SHOULD be part of a standard template. Don't force people > to constantly work around such idiosyncracies when salvation is so nigh! I agree with your argument, I just don't see a good implementation/interface yet (though this might come one day) > Maybe another hint could be to add a feature to the accent commands of > \DeclareTextAccent/\DeclareTextCommand, that they can revert to unaccented > behaviour under certain circumstances, i.e., > \"a -> \@firstofone a (thus practically a mere `a') > There could be several uses for this behaviour (initials being one of > them). an interesting thought. but i think that \"A producing just A is not a good idea in case of initials. Im sure an initial \"A would need a completely different treatment to \textogonek{A} and both would return A in your scheme. > Well, that's all for my input today; as for more positive comments: I think > those should be obvious. thanks for them good night frank