Re: Update on Localisation Style Guides and Glossaries

[E.A.Tacão <tacao conectiva com br>]

Aoife Dunne - Sunsoft ELC wrote:
> 
> Once the English is 95% cleaned up we can send a mail to all notifying it is
> available for translation. 

Ok, I'll be waiting. 8^)

> Thereafter, the glossary can be updated.  Does
> anyone have any ideas on what is the best processes for updating the glossary
> with new terms?.

Ideas? Sure! The best process? Hmm... Most probably the next lines don't
describe the 'best' process, but I hope that we can discuss a little on
that.

Let's take at first that an English list of words is our starting point
_only_ when we're creating glossaries. Please note that an "English list
of words" is not a "glossary".

I see that "updating" can be executed for two main reasons: due to a new
entry on the existing list, or because there's a need on changing an
existing entry because someone made some mistake at some time during the
glossary creation (or even during the glossary updating!).

However, new entries can be apllied by, for example, a French group of
translators. I mean, new entries can be inserted on glossaries
themselves, not only in the existing list of English words. So that
every time a new pair of words (English/French) is inserted in an
existing French glossary, all other existing glossaries (Portuguese,
German, Catalan, etc) should be updated.

And depending on how the correction of some word in an existing glossary
is executed, perhaps also the other glossaries should be updated - this
is the second case on "updating". It's really a more obscure case but
still possible.

Thus, in order to maintain consistency as long as glossaries are
updated, there should be people (humans indeed - not message hacking
scripts) responsible for updating and checking if new proposed entries
are really valid _before_ adding them. And the responsibles for each
glossary (Portuguese, German, Catalan, etc., and also the maintainer of
the English list of words) should be somehow 'connected' between
themselves (discussion list) to make all this possible and synchronous.

A CVS tree somewhere containing all glossaries is needed. This tree
should be arranged in a way that only glossary maintainers are given
access to commit new glossaries. (And commits should be done after
discussion of terms between translators, but this is a 'local' point.)

Some branching would be needed to keep relations between the several
languages and their commited glossaries. This way, all possible pairs
not involving English are instantaneously generated.

Furthermore, another important point is to use the same glossaries to
translate Gnome apps, OpenOffice, Kde, whatever. Simple reason: since
words contained in an app shall feed the glossary that will be used to
aid new translations (and this obviously is an "application-independent"
process), this could be "environment-independent" or even
"platform-independent". 

> Would you be interested in working to help clean up the English, with other team
> members?.

I'm really really sorry but I think English native speakers can do this
work much better than non native (like me) could! But as I said before,
I'd really appreciate to join efforts on making and maintain the
Brazilian Portuguese version of the glossary.

All the best
-- 
Tacão