4

I've been using BibTeX (with the plain.bst bibliography style) for years. Today I noticed something odd: Publisher information is laid out in two quite inconsistent styles, depending on whether an address is given. To be more precise, consider the entries

@inproceedings{kasher/approx-a,
  author = {Kasher, Roy and Kempe, Julia},
  title = {Two-source extractors secure against quantum adversaries},
  booktitle = {Proceedings of the 13th international conference on Approximation,
    and the 14th International conference on Randomization, and combinatorial
    optimization: algorithms and techniques},
  series = {APPROX/RANDOM'10},
  year = {2010},
  pages = {656--669},
  publisher = {Springer},
  address = {Berlin}}

@inproceedings{kasher/approx-b,
  author = {Kasher, Roy and Kempe, Julia},
  title = {Two-source extractors secure against quantum adversaries},
  booktitle = {Proceedings of the 13th international conference on Approximation,
    and the 14th International conference on Randomization, and combinatorial
    optimization: algorithms and techniques},
  series = {APPROX/RANDOM'10},
  year = {2010},
  pages = {656--669},
  publisher = {Springer}
 }

These entries are identical, except that the "-a" entry has an address. BibTeX, with the plain.bst style, will lay these out as follows:

BibTeX plain inproceedings

Note that when there is no address, the publisher data (name and publication year) are laid out together, after a period; when there is a location, the publisher data is broken up so that the publisher name is after the period, but date and address are before the period.

Question: Can anyone defend this behavior? If not, is BibTeX actively developed any more or, more to the point, is there any hope of having this addressed?

By the way, this nonsense does not exist with the @incollection entry type. The two entries

@incollection{kohayakawa2-a,
  author={Y. Kohayakawa},
  title={{Szemer\'{e}di's} regularity lemma for sparse graphs},
  booktitle={Foundations of Computational Mathematics},
  editor={F. Cucker and M. Shub},
  pages={216--230},
  address={Berlin},
  publisher={Springer},
  year={1997}
}

@incollection{kohayakawa2-b,
  author={Y. Kohayakawa},
  title={{Szemer\'{e}di's} regularity lemma for sparse graphs},
  booktitle={Foundations of Computational Mathematics},
  editor={F. Cucker and M. Shub},
  pages={216--230},
  publisher={Springer},
  year={1997}
}

produce the satisfactory output: BibTeX plain incollection

  • Looks like an error of plain.bst to me. biblatex would write in the first entry "Berlin: Springer, 2010". – Ulrike Fischer Aug 16 '13 at 15:27
  • 2
    @UlrikeFischer -- not an error of plain.bst, but a (misguided) decision of the implementor -- documented. – barbara beeton Aug 16 '13 at 15:58
6

from the file btxdoc.pdf ("bibtexing"), p.7:

10 . The PROCEEDINGS and INPROCEEDINGS entry types now use the address fi eld to tell where a conference was held, rather than to give the address of the publisher or organization. If you want to include the publisher's or organization's address, put it in the publisher or organization fi eld.

i won't try to justify this; in fact, i think it's terribly confusing. but it does explain the result that you're seeing.

  • Is nothing sacred? This means that the same field location refers to different things in different entry types? But what about the year--shouldn't that be typeset with the publisher in both cases? – acr Aug 16 '13 at 15:48
  • bibtex is long overdue for an overhaul. i've given up hoping for it. it's still listed as "maintained", and under the control of the original designer. that's where you have to ask. sigh. – barbara beeton Aug 16 '13 at 15:57
  • I understand that the location data--with this interpretation--should not appear with the publisher data, but I'm still at a loss for why the year should move away from the publisher area. Would you call this a bug in plain? – acr Aug 16 '13 at 17:48
  • 1
    @acr -- the documentation is definitely silent on this, but it's my understanding (from discussions ages and ages ago) that for "inproceedings" the year is the year of the event for which the proceedings are published, and thus attached to the location ("address"). it's somewhat assumed (but nowhere stated) that if an "inproceedings" entry appears, there will be an associated (and x-ref'ed) "proceedings" entry, and the year for that would be the year of publication; this is, of course, more often honored in the breach. this whole area is, in my opinion, a mess. – barbara beeton Aug 16 '13 at 19:07

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.