[texhax] Bilbliography problem

[Uwe Lück]

\iffalse meta-comment ... just on fairness and policy [continued below]

At 12:47 01.11.05, Robin Fairbairns wrote:

> > i Then
> > bibtex it again with no problems. now when i latex the document i get
> > the following error.
> >
> > [13.1] [14.1]) [15.1] (zthesis.bbl
> > ! Undefined control sequence.
> > <argument> \@listctr
> >
> > l.3 \bibitem{str98}
> >
> > ? (press return then the following)
> >
> > ! You can't use `\relax' after \the.
> > <recently read> \c@
> >
> > l.3 \bibitem{str98}
> >
> > ?
> >
> > It does this for all the references written up into the .bbl file form
> > .bib.
>
>it's fairly obvious you've got something completely broken up there,
>yet you tell us *nothing* about it other than "it doesn't work".

I have received mere "it doesn't work" reports for packages
I am maintaining and would have been glad if they had been
accompanied with any hint on what error report resulted --
here at least the messages have been reported precisely.
If one takes the time, a diagnosis seems to be available by
a search of CTAN for `\@listctr'. One also sees that a LaTeX
counter with an `empty' name has been referred to by \the
-- I consider this strange, but it may refer to some private
package involved.

"Bug" reports of this kind usually may be unanswerable for
"average" readers; yet as a maintainer of certain packages
I am aware of weaknesses, so a message of this kind
may suffice ... (see below)

I have just typed a section containing hints for bug reports
for a package that I maintain. My personal conclusions were:
(a) one shouldn't discourage seeming bug reports in any way;
(b) there are some important guidelines for bug reports, yet
one shouldn't expect that users can find out exactly what
one package writer needs for diagnosing. The usually
procedure might rather be: (i) a failure message from a user
(ii) a request (from package maintainers) on which informations
on the problem might be helpful ... this often makes clear
how important \listfiles may be ...

It seems to me that Robin is quite sensitive wrt RTMF/FAQ
postings -- may be understandably concerning someone
maintaining an FAQ which doesn't seem to be used properly
(from my experience a problem with these is that wordings of
problems may differ very much).
(This statement has been influenced
by reminding an Indian request on `doc' etc. this year.)
Yet I'd like to make clear the user's situation ...

>produce a minimal example, and there's a good chance someone will
>help.
>
>as i said before:
>
>http://www.tex.ac.uk/cgi-bin/texfaq2html?label=minxampl

I have experienced situations where requiring "a minimal example"
was much more than could be expected from the user (and where I
would have been unable for days as well). One example (with which
I'm dealing these days again) is drawing footnotes down one or
more pages (\@reinserts; my `tamefloat' package may help, but I
little remember it myself). Another -- quite trivial example -- was:
editorial notes generated thousands of control sequences.
Ten thousand of them generated a "TeX capacity exceeded ..."
-- i.e.: 9999 notes on 400 pages worked, while some 11.000 on
450 pages didn't. The user reported a number of problems with my
macros, and told me something like "BTW, ... doesn't work any
more", thus incidentally referring to this detail -- which I had been
afraid of anyway for many months. "Any more" of course meant
that only 450 or so pages with 11.000 or so notes were a
"minimal example". lineno.sty had a similar problem in its
"birth year" 1995.

Best,
   Uwe.