<html>
<head>
<meta content="text/html; charset=windows-1252"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<div class="moz-cite-prefix">Op 27-01-17 om 12:05 schreef Thomas
Friedrichsmeier:<br>
</div>
<blockquote cite="mid:20170127120505.5faa24d0@edge" type="cite">
<pre wrap="">Hi!
Thanks a lot for your feedback, and sorry about the grief we're
causing! I was in fact totally unaware of this issue, so far (we had
worried quite a bit about making it easier to translate rkward plugins,
but never spent a thought on i18n of the technical documentation).
In essence, the (most) problematic areas appear to the
<programlisting/> elements, and we use those quite a lot. Much of the
content of those areas is actually not translatable, of course, but
there are comments (and some labels) in between, which should be
translated. As a matter of fact, I think(TM), that almost all of those
comments and labels could reasonably be treated as separate translation
units, totally ignoring all the XML surrounding them in the
<programlisting/>s.
Hmm. Do you happen to know just how I could change the markup to
achieve just that?
Regards
Thomas</pre>
</blockquote>
<p>Hi, Thomas, thanks for this reply.</p>
<p>The problem that you mention is another one, which indeed ís a
problem.</p>
<p>I guess most translators are not programmers themselves, but know
their native language and English well enough to be able to
translate from English into their own, satisfactorily, for normal
texts.<br>
</p>
<p>Personally, as a retired math teacher, I am not really a
programmer myself, but to a certain extend have dabbled with some,
also modern, programming languages, of which Python is the latest
one. And programming RPN calculators, of course, such as HP 67, 48
and 50.</p>
<p>I leave the program listings themselves as they are, right now,
and while translating the texts I have to account for that, and
explain. This means that I have to make at least a second pass
through the whole document, just to get this right. Making an
error in the program texts (except in the comments) may ruin the
experience of the users. I am very cautious when translating names
of variables, as I am not always quite sure whether they are
standard names or not. Therefore:<br>
</p>
<p>I did think of this problem, but I don't think it's feasible (?)
to print the texts, in a .po file, which should <font size="+1">not</font>
be translated, in a very distinct letter, or colour (pity if a
translator is colour blind :) ). But then one is still left with
the necessity for explaining these untranslated texts, in the
other messages. These I usually explain in one or two places
between parentheses, when appropriate. But then: a handbook is
never read straight from page 1 to page last, I assume.<br>
</p>
<p>It pleases me very much, reading in your answer, that you have
understood this problem yourselves ... hope that you'll find a
workable solution for this.</p>
<p>Regards,<br>
</p>
<p>Jaap Woldringh<br>
</p>
<p><br>
</p>
</body>
</html>