Re: [Bibulus-dev] Re: Many authors

[Thomas M. Widmann]

"Torsten Bronger" <address@hidden> writes:

> address@hidden (Thomas M. Widmann) writes:
> 
> > address@hidden (Palle Jørgensen) writes:
> >
> >> "Torsten Bronger" <address@hidden> writes:
> >> 
> >> > Arne Jørgensen <address@hidden> writes:
> >> >
> >> >> [...]
> >> >>
> >> >> Another issue with authors.
> >> >>
> >> >> Should it be (is it already?) possible to define an author
> >> >> outside the book/article/whatever and then crossref it?
> >> >>
> >> >> If the same author appears in more than one publication it
> >> >> will probably be convenient to define him/her only once.
> >
> > I've actually been thinking about the same.  [...]
> >
> >> > This is not a good idea in my opinion.  It makes the file hard
> >> > to read, it is very uncommon, and you don't win much.
> >
> > Depends.  It's certainly not worth it just to avoid a nametype
> > every now and then, but what with authors publishing under many
> > different names, it could be nice to have a way of building an
> > author database.
> 
> I don't think that a significantly large number of users may want to
> do that, and even those people can do that separately.

I'm not so sure.  I'd really wish we had some people from all over the
world on this mailing list, because I'm not really sure what the needs
are.  But it's my impression that a surprising number of people have
multiple (and unpredictable) forms of their names in other languages.
I've seen it here in Scotland (English/Gaelic), Georgia (Georgian/
Russian/Western), East Asia, and I suspect there are many more.

> I like the idea of context independent entries.  If I extract a part
> of my bibliography for a friend, I needn't pick all necessary
> authors, too.

It would be very easy to implement an extract function in Bibulus.

Besides, in the syntax I proposed, the entries would still make sense
on their own, because the author would still be there -- there would
just be an informative link to extra information about him/her.

> As far as language dependent names are concerned: If you refer to
> something, you must give the original data.  So even if you cite the
> TeXbook in a Chinese journal, you must write Knuth with Latin
> letters.  The script is part of the bibliography entry and therefore
> I see no need for this sort of "language switchboard" that you
> propose.

Sorry, no.  Let me try to reverse what you said:

    So even if you cite his "little red book" in a English journal,
    you must write Mao Zedong with Chinese characters.

See? ;-)  (You might argue this is nonsense, because you'd never cite
the original in an English journal, but rather the translation.  But
perhaps there is no translation.  Would you then expect the author to
be written in Chinese characters?)

What should in general be done, I think, is that the local form of the
name should be used in text (and first in the bibliography if it's
author-date), while the original form should appear in the
bibliographic entry.  However, the should be configurable, of course.

I know all of this sounds pretty theoretical, but it is an important
goal (for me) that Bibulus is able to support multilingual
bibliographies.  And I think one of the most key insights of BibTeX
was "write once, use many times"; by this I mean that there should
only be a need for one bibliograhic entry for a book, no matter which
language the text it is cited in is in.

> >> >  And, you can use entities for it.
> >
> > How much can you put into an entity?
> 
> A whole XML document.  But no tag must be left open.

OK, so you could pretty easily make an entity per author.  That might
be a better solution.  Or it might not.  That is the question...

/Thomas
-- 
\author{Thomas Widmann\thanks{3/2,  54 Mavisbank Gardens,  Glasgow  G51\,1HL,
Scotland, address@hidden Tel.~+44 (141) 419\,9872.}\\{\tt address@hidden