2

BibTeX has many shortcomings but apparently there was (at some point) a plan to develop BibTeX 1.0 with the aim of overcoming at least some drawbacks.

See this paper by Oren Patashnik, the author of BibTeX. However, that paper is from 2003 (!!) and, unless I'm missing something, the plans never materialised. The latest version of BibTeX seems to be 0.99d from 2010.

Now I realise many TeXperts consider Biblatex+Biber the modern alternative but BibTeX still ships with every LaTeX distribution and many journals/publishers use it. Surely there cannot be much harm in improving the default bibliography system that ships with LaTeX.

Is anyone aware of why BibTeX 1.0 was abandoned? Or was it? Finally, and this may be naive, but could there be value in putting BibTeX on, say, github.com/latex3 for people to consider contributing to it?

3
  • 2
    you could ask on the texlive mailing list or to biblio@tug.org about the status and how to contribute. But it certainly doesn't belong into latex3. Jul 20, 2021 at 14:42
  • 3
    My interpretation is the version is tending asymptomatically to 1.0 ...
    – Joseph Wright
    Jul 20, 2021 at 14:50
  • 3
    The problem with 'improving' here is that means 'changing', and as BibTeX is extremely (also plain-TeX like) stable, that's a real issue. For example, if you make a Unicode-aware BibTeX, there's all sorts to think about in terms of locales. I suspect that really wouldn't ever 'take off'.
    – Joseph Wright
    Jul 20, 2021 at 14:51

1 Answer 1

5

There are simply too many downsides to BiBTeX for it to be worth developing: among them the lack of multibyte character support, the language it is written in, and the antiquity of many of the styles. I don't know if Oren considers it abandoned or not, but I think it needs abandoning. I see no merit at all in putting it on github.

The BIG difficulty is publishers' own styles and classes that use ancient BiBTeX styles. Many of the styles are available in biblatex, but IMHO TUG need to make a concerted effort to encourage publishers to update their offerings and join the 21st century.

Your kilometerage may, of course, vary.

[edit] I should add that this is not to ignore the work that Oren has put in, nor the many people who have done good things with BiBTeX over the years, and their contributions all need to be preserved whatever direction things go. Publishers will make their own decisions anyway, but my concern is that if biblatex does not gain traction with them, and BiBTeX cannot provide their needs, they will find less reason to support TeX.

7
  • 2
    Hmmm, if there is a way to make BibTeX Unicode aware in a largely backwards compatible way (for both .bib databases and .bst styles) that would be a step forward. I expect this to be highly non-trivial (especially if you want backwards compatibility) and you'd land yourself in the "does my coauthor have the same version of BibTeX as I do" malarkey, but it'd be useful nonetheless - especially because I don't really see biblatex gain traction with publishers.
    – moewe
    Jul 21, 2021 at 7:44
  • I absolutely agree about the styles. The BibTeX language is ... unusual nowadays and many styles that are out there have not been updated in ages or have issues. But again, updating without messing up backwards compatibility is hard.
    – moewe
    Jul 21, 2021 at 7:47
  • @moewe I think it would be nice to have CSL style to BibLaTeX converter. Or CSL processor included in TeX Live.
    – michal.h21
    Jul 21, 2021 at 8:26
  • Peter! Good to see you!
    – Don Hosek
    Jul 21, 2021 at 19:44
  • @michal.h21 I',m not at all convinced that's workable - a lot of the 'good stuff' in biblatex is outside-the-scope of such a style. Even if it is doable to some approximation, I wonder how many journals would care: I suspect very few 'trust' such styles.
    – Joseph Wright
    Jul 22, 2021 at 8:18

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.