[texhax] Puzzle: bibtex working when it *shouldn't*!!! ???

[Jerry]

On Wed, 9 May 2018 01:20:35 +0200, Reinhard Kotucha stated:

>On 2018-05-08 at 14:32:55 -0700, Tomas Rokicki wrote:
>
> > Don't forget, TeX was written in an era when each system had its
> > own file system with its own idiosyncrasies when it came to
> > extensions, upper/lowercase in file names, directory separators,
> > etc.  (Even today some filesystems confuse Makefile and makefile,
> > for instance [OSX].)  Windows has some very strange naming
> > conventions and issues today as well.
> > 
> > Short names where the extensions could be inferred were
> > felt to be more portable.
> > 
> > The file name conventions are located in what is typically the
> > system-dependent part of the code; MikTeX is likely making
> > different decisions there than TeXLive.  They should be
> > brought into conformance if possible, but the fact that they are
> > different should not be a shock.  
>
>Sure.  TeX Live works exactly the same way on all supported systems
>nowadays, including OS X and Windows.  There might still be slightly
>different behavior in respect of portability of files dependent on
>whether a file system is case sensitive or not.  With Karl's recent
>changes to Kpathsea I suppose that it's very unlikely that problems
>occur in the future.
>
>TeX Live proved that at user level everything can behave on Windows
>exactly as on Unix.  It would be nice if differences between TeX Live
>and MiKTeX can be avoided.
>
>Internally both distros work differently and thus have to provide
>their own documentation.  But at user level there shouldn't be any
>differences at all.  In other words: an author of a book about TeX or
>LaTeX should never have to explain the differences between TeX Live
>and MiKTeX.  But this requires a minimal amount of communication and
>more feedback from Christian is welcome.

I appreciate your opinion; however, I question if complete uniformity is
indeed a realistic or even advantageous goal. Perhaps there should not be any
differences at the user level, but is that really possible? There are numerous
differences at the user lever between *.nix and MS Windows. Worse, there is a
cornucopia of differences between different distros of all *.nix and *BSD OS’s.

While I can appreciate your desire for complete uniformity, I sort of like the
fact that I can choose an application or OS that works, or hopefully works, in
a way I find desirable.

>It's a pleasure to support students writing their theses when
>everything works the same way for me and them even if they use
>Windows.

By the way, I would prefer using, in my case MiKTex with TeXstudio 2.12.8 on a
MS Windows 10 system any day over using it on a FreeBSD system. For the
record, I use FreeBSD for several applications, but LaTeX is not one of them.
I have several friends who have expressed their opinion that MiKTeX works
better on MS Windows and is easier to maintain that TeX Live is on a *.nix or
*BSD system. Obviously, I have not verified everyone of those claims.

In any case, if teaching were easy, anyone could do it.

Just my 2¢.

-- 
Jerry