Message-Id: <199402111122.AA20401@mail.cs.tu-berlin.de>
Reply-To: Mailing list for the LaTeX3 project <LATEX-L@DHDURZ1.BITNET>
In-Reply-To:  <199402110936.KAA02421@hp5.iti.informatik.th-darmstadt.de> from
              "Mike Piff" at Feb 11, 94 09:27:43 am
Date:         Fri, 11 Feb 1994 12:20:39 +0100
From: Joachim Schrod <schrod@iti.informatik.th-darmstadt.de>
Sender: Mailing list for the LaTeX3 project <LATEX-L@DHDURZ1.BITNET>
To: Multiple recipients of list LATEX-L <LATEX-L@DHDURZ1.BITNET>
Subject:      Re: On compatibility in LaTeX2e [was: Re: keyed options lis
Status: R

(Yesterday, I promised myself not to react to Mike Piff any more --
but his statements are *******<censored>. Below is a general comment I
hope you find of interest, it is _not_ a reply to the mail, it was
simply triggered by its contents.)

He wrote:
>
> But the implication is that you are actively encouraging users to convert
> their styles to clopages.

It might be an implication for the LaTeX team -- for my work it will
be explicitely. I.e.: I'll be the moderator for the to-be-established
contrib/supported area on CTAN. The guidelines state that new modules
_must_ be usable either as a class or as a package. Modules that can
_only_ be used in compatibility mode will *not* be added.

> With hard-wired limited options?

(Such questions make clear that the writer either didn't read the
replies to his mails or that he didn't understand them.)

Options are the means to _configure_ a module, they are by definition
from a fixed set. (Def: `To configure s.th.' is the action to choose
a feature from a predefined, and therefore limited, set of
possibilities.) So the clause `hard-wired limited option' is a double
oxymoron, every hard-wired thing has limitations, and every option is
of course hard-wired.

_Extensions_, i.e., introducing new features, is done by packages, not
by options. They are the tool to use if one does not want `hard-wired
things'. (One should not try to use a hammer if one wants to fix a
screw, a screwdriver might be the better tool then...)

In CS, the difference between _configuration_ and _extension_ is well
known since decades. (_Configuration_ is sometimes also called
_customization_.) To a larger (non-CS) public the difference was
brought to attention in 1981, when RMS wrote about Emacs:

       "Extensible" means you can go beyond simple customization and
    write entirely new commands, programs in the Lisp language to be run
    by Emacs's own Lisp interpreter.  Emacs is an "on-line extensible"
    system: it is divided into many functions that call each other.  You
    can redefine any function in the middle of an editing session and
    replace any part of Emacs without making a separate copy of all of
    Emacs.  [...]  Only a programmer can write an extension to Emacs,
    but anybody can use it afterward.
                                                [Info node: emacs(Intro)]

If one replaces `Emacs' by `TeX', `Lisp' by `macro', and `function' by
`package', one has a description of (La)TeX.

        Joachim


PS: That has nothing to do with the actual concept of classes and
packages, it's a general principle we find in all kind of systems.

--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Joachim Schrod                  Email: schrod@iti.informatik.th-darmstadt.de
Computer Science Department
Technical University of Darmstadt, Germany