Sorry, I didn't dare to post to the list since I know very little about fonts and I don't want to spam the list, but here are a couple of my thoughts. You can forward to the list if you want to discuss anything further there.
I doubt there's a cleaner way (except that it's polite to start the definitions with something like \starttypescript [..?..] [computer-modern] [size])
Hello Mojca, I know it and tried it, but without success. This is what I did: - moved t-cmscbf.tex to type-cmscbf.tex - used "\starttypescript [cmscbf] [computer-modern] [size]" inside the file - replaced "\usemodule[cmscbf]" by "\usetypescriptfile[type-cmscbf] \usetypescript[cmscbf]"
That's the reason why I use modules now.
A whole lot could be wrong (anyway: I learnt a lot from your example). Yesterday I noticed misbehaviour because I loaded those definitions before the definitions in the core were loaded, so that they were overwritten But if I would judge from the definitions in type-size, I don't see the reason why \starttypescript [cmscbf] [computer-modern] [size] \usetypescript[cmscbf] would work. Typscript name is computer-modern in your case, not cmscbf. cmscbf is only the variant, which can be mono/sans/serif/... which is invoked with \tt, \ss, \rm, ..., I don't know what happens to your nam, cmscbf, probably it saves that definitions to memory, but you can't invoke them in an elegant way. It surely is better to use modules unless Hans ads the definitions to the core, but in my opinion there must be a way to surround the definitions with \starttypescript.Don't ask me how. (I guess that you loded that file manually.)
But there are at least two other problems with the file (so there must be a cleaner way): - I would like to use LM wherever possible, but now even for normal \sc the CM fonts are used.
I get lmr for normal caps in your example. You can use the (lmcsc10 instead of cmcsc10) directly or create a synonym from cmr to lmr (already defined somewhere), but since it works OK here, I cannot test the wrong behaviour.
- It would be better, to have just one command for all possible sizes. If you want 17.8pt, it wouldn't work until you add a \definebodyfont line...
PS to Peter: I would suggest you to write "author=Peter M\udiaeresis nster" instead of \enableregime[utf] in the header of your modules
I prefer to keep my source-code readable. I don't like my name written "M\udiaeresis nster", too ugly... ;-)
I agree that it's ugly (M\"unster being slightly more readable), but consider it from someone else's point of view. I work with vim which generally supports utf-8 (some well-known editors such as WinEdt don't even support it), but from some unknown reason it has problems on this windows machine. Now it's configured so that it opens the documents in utf-8 (if it thinks they are utf-8 - encoded) and in cp1250 otherwise. Now I see author=Peter MĂĽnster in the header and the sample text that is supposed to be in il9 as Äö¤˝ćĽĆ (don't ask me what you'll see at the other side of your mail client, but I can imagine that the sample code in il9 is not valid for utf-8, so editors may have general problems to view it) The most weird thing is that you mix two different encodings within the same file. Mojca (no need to answer my stupid remarks)