W3C home > Mailing lists > Public > www-international@w3.org > October to December 2003

Re: Translation Memory (TM) and text-transform

From: Chris Lilley <chris@w3.org>
Date: Fri, 24 Oct 2003 02:04:36 +0200
Message-ID: <1429148314.20031024020436@w3.org>
To: "RICHARD,FRANCOIS (HP-France,ex1)" <francois.richard@hp.com>
Cc: "'www-international@w3.org'" <www-international@w3.org>, "'www-style@w3.org'" <www-style@w3.org>

On Thursday, October 23, 2003, 10:15:23 AM, ex1) wrote:


>> RI> Francois wrote:
>> RI> I have been looking around for more info on the CSS
>> RI> 'text-transform', its purpose and  usage. I have the 
>> feeling that it
>> RI> might make the processing of text more complex since it actually
>> RI> transforms characters.
>> 
>> It doesn't transform characters, and is thus designed to make

RFHFe> It does transform the glyph.

Yes. Characters and glyphs are not the same.

>> text processing in general (including use of TM) *more* efficient.

RFHFe> I do not think so. Visual aspect is important information as
RFHFe> input for a translator. And could be classified under the
RFHFe> infamous "contextual information" that is so important to
RFHFe> translators.

Please see my example in the previous mail. Visual aspect is
important, but restyling the page does not require it to be
retranslated.


RFHFe> I agree on the principle that TM should be cleared of any
RFHFe> style, layout and essentially concentrate on pure linguistic
RFHFe> content. But it fails in practice. First simply because of lack
RFHFe> of support of these rendering mechanisms in most of
RFHFe> translation/localisation tools.

Why would a translation tool require a rendering mechanism?

RFHFe> And secondly, because the
RFHFe> actual style does impact the translation. You do not translate
RFHFe> "Products And Services catalog" the same way you translate
RFHFe> "Products and services catalog"

You would note that my example did use consistent capitalization to
indicate that sort of thing.

RFHFe> or "products and services
RFHFe> catalog". It is in my mind the same for tags such as bold or
RFHFe> emphasis. You do not want to eliminate them from the
RFHFe> translation process, but keep them in a smart way so that they
RFHFe> can be restored to the translators without messing the content
RFHFe> analysis.

You are talking about tags and I am talking about style, so I don't
see any disagreement here.

>> This will, with two lines of CSS, display identically to the
>> first example. However, by using a consistent capitalisation 
>> throughout the text, the efficiency of Translation Memory is 
>> improved.

RFHFe> Again, not true when different sytles impact the translation
RFHFe> and require differences in the translated content.

Which they will not, in this example.


-- 
 Chris                            mailto:chris@w3.org
Received on Thursday, 23 October 2003 20:05:01 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 19:17:03 GMT