Jean Christophe,
If we use PO, we use translate toolkit pot2po for upgrading, which works
quite well. No need for any TMX files from anybody.
For TM we can use Kbabel or poEdit. Both can generate internal TM and
use it.
It is extremelly simple, works like a clock, and data is consistent.
Should TMX contain single or double escaping?... I do not have the
faintest idea. What I do know is that what we do works, and that if I
generate TMX form PO, y should use it with POs, and if I generate from
XLIFF, I should use it for XLIFF... and then it works. I am afraid that
at this point we do not have such a thing as correct/universal TMX files.
... and that there is no truth on this, just opinions and systems that work.
Javier
Jean-Christophe Helary wrote
On 27 déc. 07, at 10:17, Javier SOLA wrote:
Jean-Christophe Helary wrote
We need a localization format (PO, XLIFF, key=value, anything) that
matches the localization data SUN provides us with (TMX).
I have read this 20 times in 2 days, and I dissagree. We need to use
TM, but we can perfectly well generate our own TM with the format
that we have.
At this point I believe that the best PO editor is KBabel, but it
runs only on Linux. It does quite good TM matching, enough for the
needs of OOo translation.
Do you see a connection between your first sentence and your second
sentence ?
I am talking from the perspective of a user who does not run Linux and
who uses multiplatform translation tools that are created to _support_
TMX based TMs.
KBabel/Linux users may be very lucky to benefit from already tested
workflows, but as far as Windows or Mac users are concerned such
workflows are not portable.
As I said before, and you have read this a number of times probably, I
don't care about the formats that we are provided with, as long as
their is _internal consistency_ in the data. Which is not the case
today with the PO/TMX pair provided by the coordinators/SUN.
Also, for your information, the TMX created from the Translate Toolkit
with po2tmx do not produce data that is consistent with the PO files
that are created with the same Translate Toolkit's oo2po.
I have discussed that in the past with the Translate Toolkits
developers and they argued that their was no problem with this
inconsistency...
In the end, I am arguing that besides for nice workflows, redundant
and accessible meta-data, what we need is tools that produce
consistent data from one side of the tool chain to the other side.
That is not the case with the current tool chain.
Can you explain how you do to get a "good enough" workflow with TM
re-use in Khmer ?
Jean-Christophe Helary
------------------------------------
http://mac4translators.blogspot.com/
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
