Hi Michael, Sophie, Kinshuk, all,
@Michael: you're quite right that the English documentation team
hasn't chosen its tools and workflow yet, and I know you're currently
working on ideas for us all to examine together. I'm looking forward
to that.
@Kinshuk: It's true that TDF and the LibreOffice project were only
recently launched *as a separate entity* from OpenOffice.org. But the
actual community *behind* LibreOffice (formerly OpenOffice.org) and
the software have existed for 10 years, and many of the people now
contributing to the LibreOffice community are veterans with many years
of experience.
With regard to NL teams: NL (native language) teams and communities
are not in any way obliged to adopt the same workflow or tools as the
English documentation team. They are totally free to decide on their
own approach to localization of the software, documentation production
and other things, too.
For instance, the libreoffice.org website is set up so that NL teams
can develop their own NL site with total independence from the English
site, as regards structure ("information architecture") and page
content.
NL teams are not at all obliged to follow a "translation-based"
approach to either documentation or their NL site. If they feel that
their particular community is better served by developing their own
content, they are free to do so.
As regards localization of the software, obviously NL teams are going
to be translating the English strings into their own language. But
there are various tools and workflows one could use for that
translation process and, here again, NL teams have total freedom of
choice: you could do it online with Pootle, or offline with OmegaT or
Poedit, or simply using an appropriate text editor, to name but *some*
of the possibilities.
For instance, in the English documentation team, we're about to start
evaluating Alfresco (I've now got a sandbox site set up, by the way,
and we can start on assessing that anytime from this week onwards).
But that doesn't impose any obligations on NL teams to follow in our
footsteps.
Of course, some NL teams might adopt a translation-based approach for
their resources (documentation, NL site, LibreOffice help, wiki
content, ...). In this case, TDF has all the facilities you need for
that, and the English docs team already has a quantity of high-quality
documentation that you can translate (we're working on producing much
more, too).
My personal experience with machine translation is that it *can* be a
time saver in certain very particular cases. But it does not work at
all well with most of the content we have to deal with. The quality of
results also varies according to the target language, and the effort
that has been put into developing the translation memory. Current
translation systems usually use a statistical approach rather than a
rule-based approach, and I can tell you with mirth and glee that human
translators are *not yet* close to being replaced by computers.
To summarize about machine translation: it can end up giving you more
work than if you translate manually, especially if you care about
*quality*. Usually best avoided.
When you're thinking "internationalization" in the LibreOffice
community, we have a "native language" paradigm rather than a
nationality-based approach. For instance, the French NL community
services all the French-speaking countries and communities (parts of
Belgium, parts of Canada, France, various countries in Africa, ...).
For English, we service Australia, New Zealand, the UK, the US and
various other places and communities (we don't yet have the resources
to manage different varieties of English).
For India, you can well understand the reasons for this thinking.
Only in marketing are things somewhat different. Marketing is much
more country- and region-based. This is because there are notable
differences between the context in, for instance, the UK and the
context in the US. Indeed, marketing can address markets that englobe
other national/community markets - that's the case in Europe, for
example, with the EU being one marketing target in addition to the
national states it contains.
Well, that turned into quite an essay!
HTH.
David Nelson