Según la FAQ de
gtranslator
[38]
«UMTF» is a
quite effective translation memory format currently used by
gtranslator.
Sigue explicando la FAQ:
- P:.
What is this learn buffer you're talking so often?
- P:.
Will gtranslator support enhanced translation memories
(in formats like
TMX/OpenTag)?
P:. |
What is this learn buffer you're talking so often?
|
R:. |
The «learn buffer» is the
gtranslator specific
implementation of a «translation» to help
translators easyfying their work by translating always
re-occuring messages with already learned and saved
translations. gtranslator's learn buffer is a
quite simple but effective kind of translation memory in
the UMTF XML format.
After learning some of your already
well translated po files you can simply use the auto
translating capabilities of gtranslator to perform a good
level of auto translation for already in the learn buffer
saved message & translation pairs. A learn buffer just
makes your life as a translator much easier as if you
did build up a quite normal-sized (effectively 300+ KB)
learn buffer gtranslator performs an auto translation
level of ca. 15 %.
|
P:. |
Will gtranslator support enhanced translation memories
(in formats like
TMX/OpenTag)?
|
R:. |
Surely it will; support of translation memories is one
of the biggest TODO's for future
gtranslator
releases.
OpenTag support will occur surely one day,
TMX also, but
I think the order might be so that we would integrate
first OpenTag and then TMX —
but we will do it.
|
Este punto podemos actualizarlo con una referencia al archivo
de la lista
<lucas-desarrollo@listas.hispalinux.es>,
hilo ‘gtranslator, learn buffer etc.’ que se
inicia el 16 de enero de 2003 (y que continúa una conversación
privada entre kabalak, Ismael Olea y Juan Rafael Fernández).
Puede consultarse en
https://listas.hispalinux.es/pipermail/lucas-desarrollo/2003-January/000256.html
. En él kabalak pide consejo sobre qué formatos de TM
implementar y si hacerlo de forma nativa o simplemente la
capacidad de importar y exportar los formatos
estándar. Nosotros le hablamos de la necesidad de adoptar los
estándares abiertos, del mismo modo que la industria ha
comprendido que son necesarios.
La conclusión de kabalak es esta:
I think I'll code for a UMTF-backboned gtranslator further
on but with support for Opentag and TMX which is also
somehow included in gtranslator's code (good news in this
field, as the TMX and OpenTag fields are "very easy" to
handle as XML files…)
Nota
Ismael Olea
ha puesto en
http://www.olea.org/gtranslator-es/
«una memoria de traducción de español construida
con gtranslator
con las actuales traducciones del proyecto Gnome».