[texhax] typesetting a critical text edition [parallel]
uwe.lueck at web.de
Sat Jun 20 12:06:48 CEST 2009
I know some people who worried about parallel typesetting for translations
and am distributing Pierre MacKay's advice among them. I am focussing
here on the practical aspect of how you could obtain the "parallel" feature.
At 01:41 20.06.09, Pierre MacKay wrote:
>Uwe Lueck wrote:
>>3. For "original" even pages, "translation" odd pages, there is always
>>the "brute" method: typeset original and translation as different LaTeX
>>documents, augmenting the page number by 2 each, print them, sort the
>>pages at a table using your hands, eyes, brain ... Roy Flechner told me
>>that the fancyhdr package is helpful for this.
>For facing translations, I set verso pages in one file, using an
>increment of 2 for pagination, and recto pages in a separate file. I can
>then do fine control of page ends with two instances of xdvi. To blend
>the two files, I use dviselect and dviconcat, driven by a fairly simple
Apart from dviselect dviconcat There also is dvitodvi. Some may
need to know that xdvi is a UNIX previewer and that the same can be done
with YAP etc.
After some thinking it seems to me that this method isn't just a
workaround, but the preferable way. This way you move manual page breaks,
otherwise you must move text around in the file.
An aspect somewhat intermediate between the practical and philosophical one
is that some workers *do not know* how to use dviselect and dviconcat and
bash -- e.g., sitting at their PC or notebook equipped with Windows.
However, they could ask an IT group for the humanities or at the math or
physics department of their university. Either somebody skilled does the
dvi things, or you can get Windows versions of these utilities. (?)
I even know some who stick to DOS -- not so bad! -- they might learn that
Linux -- bash! -- is smarter -- and you get better support for utilities
and programs for Linux! This is about what Pierre MacKay indicates in the
more philosophical passages of his posting:
>The argument for the chain of programs is that each utility is doing what
>it does best, cleanly, and in a way that can easily be traced if anything
>goes wrong. The Microsoft black-box solution, cramming every task into
>the same gargantuan utility, no matter how inappropriate, produces
>constant frustration, because the resultant programs become so complex
>that even their authors can't trace the errors.
I go on in a second posting, because some people who get the present one
certainly don't want to see the rest.
-- Uwe Lueck.
 <pierre.mackay at comcast.net>
?? -- or rather:
 See README at
More information about the texhax