SPIRES interface and tag-suggestion

sebastian guttenberg guttenb at inp.demokritos.gr
Tue Oct 30 05:08:34 EDT 2007


Hi John

- One more comment to your question concerning the SPIRES-interface:

> I'm not sure what interfaces SPIRES provides for applications.  Do you
> know?  In the future, I'd like to have a general plugin system to
> support various backends.  At present, only crossref OpenURL and
> arxiv.org bibtex are there.

I still don't know about such an interface, but I have found a script
that generates a BiBTeX-file to a LaTeX file, where one only has to
provide the preprint-numbers. It is called findbib. See
ftp://cam.ctan.org/tex-archive/biblio/bibtex/utils/findbib/README

So suppose, I have a pdf-file that contains an arxive number (Like the
one I sent to you, which contains the number hep-th/0112160). With this
information one could build a very simple tex-File, let's call it
auxiliary.tex:

\documentclass{article}
\begin{document}
\cite{hep-th/0405007}
\end{document}

Then one can run the script ./findbib auxiliary.tex which produces an
auxiliary.bib which contains all the MetaData. (There will also be two
error messages because of missing bibliography-style and bibliography
file. One can extend the tex-file to avoid this, but the bib-file will
be the same anyway)
Would that be a possible solution to include SPIRES?

- My second comment is about the tags. I hoped that marking two tags at
the same time would lead to an intesection of the results for each
single tag. Instead it leads to the sum, which I find very unfortunate.
Let me explain why I think that the first would be much more useful.
Suppose I am interested in finding an article which has either the tag
"important" or "interesting", but I cannot really remember. In this
case, I can use the tool as it is at the moment and mark both tags. But
that does not really help, because I have to look through all of
"important" and all of "interesting" anyway, because both are now
displayed. 
In the case, however, where I know that an article was interesting AND
important at the same time, an intersection method would help a lot to
find the article.
Related to that: at the moment, it is only possible to mark tags with
shift-cursor, which are directly next to each other in the list. There
is no possibility to mark in addition others with the common
ctrl-Mouseclick , which is also a pity (the same is true for the files
themselves).
At the moment, I am a bit hesitating. I wanted to start to manage all of
my downloaded pdf-files. Assigning tags to all of them is quite some
work, but would be worth doing. But my strategy to assign tags would be
very different if I knew that there will be the possibility for an
"intersecting" tag-selection in near future...

Sorry again for the long comments. Best, Sebastian

P.S.: I am now confused what you meant with 'filing the pdf-file as a
bug'. I did not find a possibility to attach pdf-files to a bug report.
Or did I misunderstand s.th.?




-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.




More information about the referencer mailing list