Chemical structures with plain TeX

William F Hammond hmwlfsr at
Sat Jul 6 18:38:11 CEST 2019

"Taylor, P" <P.Taylor at> writes:

> . . .    Peter's formulation, while 
> arguably elegant, is cryptic beyond belief.  And if Peter's were to 
> fail,for some unknown reason, how would one go about debugging it ?  By 
> deliberately piping the output of one program into the input of another, 
> eschewing temporary files, no trace is left of the intermediate results, 
> so one cannot go back to inspect these in an attempt to identify the 
> source of the error.
> Yes, I do not deny for one second that I loathe and detest Unix, but 
> that was not the motivation for posting my message — I was genuinely 
> saddened (OK, I didn't actually weep, that was a slight exaggeration) 
> that such a potentially useful solution to such an important (within the 
> universe of discourse) problem should be transmitted as such (sorry, I 
> cannot find a better phrase) complete and utter gibberish.

For seeming gibberish compare with David Carlisle's famous
xii.tex.  Even non-obfuscated TeX code can be more
impenetrable than Peter's example.

I quite commonly do on-the-fly tasks that pipe the results
of a search into a command line invocation of a stream
editor with that output, in turn, piped into a command line
gawk strip or a command line perl script.  For example, I
might use something like that to generate a LaTeX tabular
environment from column-delineated data in a text/plain file
sitting base 64-encoded in a mail folder.

It can save lots of time for small things.

Of course, for *ix systems, it helps to have looked at the
book by Kernighan and Pike (even if it is now rather quaint),
just as for TeX it helps to have looked at The TeXbook.

                              -- Bill

More information about the texhax mailing list