Received: by nummer-3.proteosys id <01C19443.A21AFA64@nummer-3.proteosys>; Thu, 3 Jan 2002 11:44:34 +0100 Return-Path: MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----_=_NextPart_001_01C19443.A21AFA64" x-vm-v5-data: ([nil nil nil nil nil nil nil nil nil][nil nil nil nil nil nil nil nil nil nil nil nil nil nil nil nil]) X-MimeOLE: Produced By Microsoft Exchange V6.5 Content-class: urn:content-classes:message Subject: {1} RE: {1} Size options (and alternatives) Date: Thu, 19 Mar 1992 17:32:09 +0100 Message-ID: X-MS-Has-Attach: X-MS-TNEF-Correlator: From: Sender: "LaTeX-L Mailing list" To: "Multiple recipients of" Reply-To: "LaTeX-L Mailing list" Status: R X-Status: X-Keywords: X-UID: 628 This is a multi-part message in MIME format. ------_=_NextPart_001_01C19443.A21AFA64 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable David Rhead has just reminded me that I haven't responded to his earlier mailing in re `Size options (and alternatives)': >>> This is a comment about Phil's comment (about my comment about Rocky >>> Bernstein's request). >>> Is >>> `generic style-file that, when it's been given a certain minimum >>> number of dimensions (and perhaps other parameters), yields a >>> good design' >>> attainable? I hope that it is. In other words, I hope that we can factor out the metanotions of `book', `article', `thesis', etc into a small number of generic files which implement the basic meta-concepts. Layered onto these generic files would then be design-specific variants. One would hope that the initial distribution would have at least one instance of a design-specific variant for each meta-concept. David's proposed alternative, of three levels of support, would require each implementor to go through the hassle of writing a complete style file (admittedly based on the `canonical design-specific variant' referred to above). But this is exactly the unsatisfactory situation which obtains in = LaTeX-2: each new style file is just that, requiring major intellectual = investment, rather than a layered product which takes full advantage of the = meta-construct of which it is an instance. I would strongly urge that we take maximum advantage of the concept of `factoring out the metaness', thus allowing users of whatever ability to write the minumum necessary code to achieve their ends. Thus the totally na\"\ive user would write LaTeX-3, pure and simple, and would take advantage of whatever specific style files are available to him/her; the slightly more sophisticated user, with some design awareness, but by no means a fully-fledged designer, would be able to write pure LaTeX-3, plus small `add-on' style files which would modify minor elements of the design. More sophisticated = users still would write more serious style files, re-implementing major = portions of the design but still based (and using, and nested below) a specific = style file, while the ultimate designer would implement a top-level specific style = file, based on (and nested under) the generic style file, using (perhaps) the pre-supplied `canonical specific style file' as a model. In summary, I think it very important that each level of user need only concern himself with, and write, code relevant to his/her level of = expertise; re-inventing of the wheel should never be necessary. ** Phil ------_=_NextPart_001_01C19443.A21AFA64 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable {1} RE: {1} Size options (and alternatives)

David Rhead has just reminded me that I haven't = responded to his earlier
mailing in re `Size options (and = alternatives)':

>>> This is a comment about Phil's comment = (about my comment about Rocky
>>> Bernstein's request).

>>> Is
>>>    `generic style-file = that, when it's been given a certain minimum
>>>     number of = dimensions (and perhaps other parameters), yields a
>>>     good = design'
>>> attainable?

I hope that it is.  In other words, I hope that = we can factor out the
metanotions of `book', `article', `thesis', etc into = a small number of
generic files which implement the basic = meta-concepts.  Layered onto
these generic files would then be design-specific = variants.  One would
hope that the initial distribution would have at = least one instance of
a design-specific variant for each = meta-concept.  David's proposed
alternative, of three levels of support, would = require each implementor
to go through the hassle of writing a complete style = file (admittedly
based on the `canonical design-specific variant' = referred to above).
But this is exactly the unsatisfactory situation = which obtains in LaTeX-2:
each new style file is just that, requiring major = intellectual investment,
rather than a layered product which takes full = advantage of the meta-construct
of which it is an instance.  I would strongly = urge that we take
maximum advantage of the concept of `factoring out = the metaness',
thus allowing users of whatever ability to write the = minumum necessary
code to achieve their ends.

Thus the totally na\"\ive user would write = LaTeX-3, pure and simple,
and would take advantage of whatever specific style = files
are available to him/her; the slightly more = sophisticated user,
with some design awareness, but by no means a = fully-fledged designer,
would be able to write pure LaTeX-3, plus small = `add-on' style files
which would modify minor elements of the = design.  More sophisticated users
still would write more serious style files, = re-implementing major portions of
the design but still based (and using, and nested = below) a specific style file,
while the ultimate designer would implement a = top-level specific style file,
based on (and nested under) the generic style file, = using (perhaps) the
pre-supplied `canonical specific style file' as a = model.

In summary, I think it very important that each level = of user need only
concern himself with, and write, code relevant to = his/her level of expertise;
re-inventing of the wheel should never be = necessary.

        =         =         =         =         ** Phil

------_=_NextPart_001_01C19443.A21AFA64--