a symref sometimes needs to reference more than one symbol

[kohlhase]

I am just realizing that sometimes \symref does not fully do the job as a "translation help". Take the case of the word Deschiffrierschlüssel from smlgom/cs/source/public-key-encryption.de.tex. In English we have \symname{decryption} \symname{key} so there is no problem. But if we do the same trick in German with \symref{decryption}{Dechiffrier}\symref{key}{schlüssels}, then we get

Note the very small space? Also this will not work if there is an entirely new (noncompositional) name for "Dechiffrierschlüssel".

The Problem here, is that at the time we make the symbol decisions in e.g. English, we do not know whether there are non-compositional composita in other languages.

To get around this, we could do two things:

• reference both symbols decryption and key for Dechiffrierschlüssel e.g. allowing \definiendum and \symref to take multiple symbols, or
• have general mechanism for concept composition, e.g. \compositum[gloss=2 for 1]{\symref{decryption}{Deschiffrier}\symref{key}{schlüssel}} and \conceptcombine[gloss=2 for 1]{\symref{decryption}{Deschiffrier}\symref{key}{schlüssel}}{FOO}in the case of a non-compositional combination. I really like the idea of the gloss here, which would give us a way to unraveling (e.g. on hover) a compositum or combination.

I am sure that there could be more ways of doing this. Unfortunately I do not have a good/natural example for a \combine yet.

[kohlhase]

Hmmm, I see that we could use the gloss for the "allow multiple sysmbols" syntax above as well.

[kohlhase]

Also, the \compositum[gloss=]{...} would also be nice to have in English as well to unravel composita that are made by juxtaposition.

[kohlhase]

But maybe that is too much work in practice.

[Jazzpirate]

I would argue that pulic-key and private-key should be saparate symbols anyway ;) (or encryption-key and decryption-key). How both relate to a more general "key"-symbol is still a question, however. Probably something-something-subtyping...

[kohlhase]

But this is not the point I am trying to make here, there is a general representation problem about composita.

[Jazzpirate]

But this is not the point I am trying to make here, there is a general representation problem about composita.

Fair, but I'm not entirely convinced that every such compositum does not "deserve" a dedicated symbol in general.... do you have any examples in mind where the compositum does not it fact represent a distinct concept?