[Olpc-Haiti] professional translation resources for OLPC/SugarLabs

[Alexander Dupuy]

Chris Leonard wrote:
> I believe that all translations are from volunteers. Each language 
> group has the opportunity to organize itself and control the process 
> as they see fit.  For example, Nepali translations have been provided 
> by the very organized team from OLE Nepal.  A number of the languages 
> (pidgins) in Oceania share some of the challenges of Haitian Kreyol in 
> that they are not richly described as written languages and they are 
> controlled by designees from OLPC Oceania for harmonization of the 
> written forms.
>
> One of the beauty of open source tools like Pootle is that while they 
> are well-designed to accept input from a broad community of 
> participants, they still have significant quality control mechanisms 
> that can be employed to achieve high quality.  Differing levels of 
> privileges, review with some automated error checking to identify 
> possible problems, contributions to the Pootle server are not 
> "published" to the software repo until a language administrator 
> "commits" the particular PO file.  Of course any individual string 
> translation can be reviewed and corrected easily by changing it and 
> re-committing the PO file.

Michel DeGraff writes:

> Hopefully the current language administrators are certifiably competent
> professional translators who are duly remunerated for their much needed
> expertise and their hard work, on a par with other employees and
> consultants working for OLPC.

Unfortunately, given the financial structure of OLPC and SugarLabs (the 
latter of which actually administers the Pootle translation/localization 
system used for the XO software), I would guess that none of the current 
language administrators for the SugarLabs Pootle system are remunerated, 
and I suspect that this is unlikely to change anytime soon.

That doesn't mean that there may not be professional translators or OLPC 
employees acting as administrators, just that they are unlikely to be 
getting paid for doing this.  It is more possible that some translations 
have been done (for pay) by professional translators (this may be the 
case with the Kreyòl translations Guy-Serge Pompilus reports, I don't 
know) but I would guess that these would have been delivered to the 
administrators in bulk form and incorporated with little if any review 
or modification.

In most cases, the administrator's role in a Pootle 
translation/localization system involves coordination between and among 
translators and developers, along with some amount of automated 
consistency checking to handle technical (non-language-specific) issues 
with the translations.  As such, IT and software development experience 
is probably more relevant than language proficiency although of course 
both are helpful.

As Chris noted, Pootle has a permissions system that can be configured 
by an administrator (again - IT experience helpful here) to give 
different users different levels of access to the translation: 
"suggestion" (noted as a comment on the current translations, for later 
review), "submission" (a change to the current translation), "review" 
(incorporating and/or deleting suggestions), and "commit" (actually 
delivering the set of translations to the software development/build 
system).  By limiting non-expert translators to suggestions, expert 
translators can review any translations before changes are made, or an 
administrator performs a commit.

However, that is the theory, and in practice, most professional 
translators are much more skilled in languages than in computer 
technology - they probably already use their own preferred set of 
translation tools (this may just be Microsoft Word), and are less likely 
to be as interested in learning the moderately complex Pootle interface 
to review translations.  So the ability to effectively utilize 
professional translators (even when money is available to pay them) can 
be limited, and will often require volunteer administrators familiar 
with the Pootle system to "bridge the gap" and provide export and import 
to and from the formats the professional translators use.

That said, I think that professional translators do have something to 
contribute to this effort, and what's more, in the specific context of 
Haiti there is currently some opportunity to engage them even without 
any money to lay on the table.  In particular, I would note the 
following two blog postings 
http://renatobeninatto.blogspot.com/2010/01/what-translation-industry-can-do-for.html 
and especially 
http://www.milengo.com/what-the-translation-industry-is-doing-for-haiti-relief/ 
as providing a lot of information about professional translation 
resources being mobilized in support of Haiti.  Many of these, I expect, 
may have very limited ability to support Kreyòl translation, but there 
are at least some.  It certainly would make sense for this group to make 
contact with some of these others to see what joint efforts and/or 
coordination is possible.

@alex

-- 
mailto:alex.dupuy at mac.com