On 2011-03-04 10:09, Yury Tarasievich wrote:
On 03/04/2011 09:40 AM, Dwayne Bailey wrote:
On 2011-03-04 09:14, Yury Tarasievich wrote:
...
description of the .PO based L10N process. Pootle isn't acceptable for
the languages with the comparatively weak terminological base . In
such cases it's common for everybody to translate "just as one sees
...
We have many African languages who also have weak terminology that
benefit from using Pootle.
The difference is that they include their terminology in the terminology
project on Pootle. You can do that also, in which case people giving
suggestions and translations are guided by the teams terminology lists.
The big problem is this procedure isn't well visible or integrated.
E.g., one may easily bypass it or ignore it.
It's quite visible the terms show up on the left of the translation.
But you are correct you can easily ignore it. Enforcing terminology is
difficult though, especially in language where words transform depending
on their context. A social agreement is in most cases a better approach
I think.
Also, the terminology matching isn't as quite clear-cut process as
many tools make it to look (Pootle, too, as far as I can understand --
I know Pootle only superficially). E.g., is there a support for the
context variants (translate as A1 in context C1, as A2 in C2 etc.)?
Terminology matching is ultimately a human decision, a machine can only
do so much. Context itself is difficult for a machine to determine.
The skills needed to understand the issues of translation and
terminology use and development are the reasons why I prefer to have all
translators go through a vetting stage (suggestions only) before being
given full translate rights.
--
Dwayne
--
Unsubscribe instructions: E-mail to l10n+h...@libreoffice.org
List archive: http://listarchives.libreoffice.org/www/l10n/
*** All posts to this list are publicly archived for eternity ***
