Ensuring Quality in Ubuntu Translations

Tim Morley t_morley at argonet.co.uk
Tue Apr 11 09:48:35 BST 2006


> From: Matthew East <mdke at ubuntu.com>
> Subject: Ensuring Quality in Ubuntu Translations
> To: rosetta-users at lists.ubuntu.com
>
> An interesting discussion on the #launchpad irc channel yesterday has
> been making me think about the question of quality assurance for  
> Ubuntu
> translation teams.

Interesting to see this discussion. This was one of my main worries  
that held me back from using Rosetta for six months -- that the  
quality of translation would suffer. In the end, it was a pragmatic  
decision -- I simply didn't have time to proof-read every submission,  
as I had done for a long time, so now the core translators can submit  
directly to Rosetta.

However, as I've said recently on this list, the issue of quality  
control in intimately linked with adding comments to translations,  
but apparently this *isn't* a high priority.  :o(

> Date: Mon, 10 Apr 2006 17:53:17 +0200
> From: Jordi Mallach <jordi at sindominio.net>
> Subject: Re: removing/updating translator notes in Rosetta
>
> Yes, there are plans to add UI to edit [comments], but it isn't a  
> priority
> right now, as there are more important features being developed for
> rosetta.

If QA is a priority, then I would politely suggest that the ability  
to add/edit comments is also a priority. As I said before, before we  
used Rosetta, there were comments added absolutely everywhere  
throughout our po files, often half a dozen on a single translation,  
all of which permanently recorded the train of thought that had led  
to a particular translation, our level of confidence in the current  
translation, and references to uses on websites/in other software  
(free or non)/in dictionaries/etc.

Using Rosetta, none of this is possible, and the quality of  
translation is undoubtedly suffering. The only advantage that Rosetta  
currently gives us is the easy interface that pleases non-techies,  
and the instant gratification of seeing one's work immediately  
available to all, without having to wait for a team leader to find  
the time to upload submitted work (and, possibly, proof-read it).  
This alone has revitalised the team, and we're making good progress  
now, but to get this, we've had to trade in the ability to record our  
thoughts about the work we're doing, and, I repeat, the quality of  
the translation has suffered.  :o(

I intend all of the above to be constructive, helpful commentary,  
offered in the hope that it might make future iterations of Rosetta  
even better. If it sounds like whining about stuff that I'm getting  
for free, go back and read it again in a different tone of voice,  
please.  :o)

Cheers all.


Tim




More information about the rosetta-users mailing list