Re: Restarting W3C Volunteer Translation Tracking

Am 11.10.2018 um 21:15 schrieb Jean-Christophe Helary:
>> On Oct 11, 2018, at 22:44, Dominique Hazael-Massieux <dom@w3.org
>> <mailto:dom@w3.org>> wrote:
>> Le 11/10/2018 à 14:38, Jens Oliver Meiert a écrit :
>>> a) a process that allows for rapid translations that could be marked
>>> prominently as “unofficial/unreviewed translations” (and which may
>>> certainly be handed off for further refinement), and
>> I can understand the potential value of allowing for this, but I'm not
>> sure how to limit the abuse of such an approach we've seen happening in
>> practice in the previous program. Do you have any idea I could explore
>> in this space?
> That's where "tools" (or processes) come into action.
> A translation hosted on a collaborative translation site (a pootle
> server for ex) would allow for a number of people to review the document
> until a "translation manager" deems it is good enough for official
> publication, meanwhile it is available but only a "draft translation".

Well, looking at close to 20 years of history of translations I
followed, most good translations were a one person effort. At least I
can say that for the translations into my native language. I did not
only publish a lot on my own I also read most of the translations of
other translators into my native language "quietly".
The ones from Jens Oliver Meiert too. Which I think was a single person
approach, and some others when the translation of W3C documents was
still more popular.

In comparison I wasn't that happy about the "authorized" translation of
WCAG 2.0. It was authorized, but in some parts not even as good as some
single person efforts I had seen before.

I also saw some translations that were Google translations for the
benefit of traffic. I reported these, I think, two incidents. That was a
long time ago.

So overall the easy effort to announce a translation, maybe having it
started already or almost ready to publish, then to announce the
availability brought out the majority of translations of good quality.

Having a quality management system is a good idea (I did some ISO 9001
certification work), but we had that already before, not so obviously,
but it worked, as bad translations got reported.
It doubt it will become better by making every step visible at Github
and having a hundred "smart" people watch and "help" with the
translation. Very often there won't be any helpers, sometimes you find
just talkers who actually hold you back from doing work.

I am helping on a PHP project for many years now where the hurdle to
contribute is actually quite high, not really high, but you have to show
some effort before you can commit to the repository and you got to show
some clean code.
The number of participants is not necessarily raising the quality of the
work, in this case the hurdles were risen at a time to actually avoid an
too easy entry for people messing things up.

"Translating" this into W3C translations I was asked one time to
translate a specification for a book. The specification was already
translated by a team of 7 translators before. It was a mess I was told.
So only two people were asked who had some good translations out already
and we did it again. The only translation I did because someone else
asked me to do it. All the others were only for fun or to learn.

So with all the eyes watching your daily effort I am not sure if I had
put out any translations. I remember lying at an indoor pool in winter
with everything printed on white paper and correcting the translation
the fourth or fifth time before showing anything to the public.

After that I felt that I had done some good work, I got my ego polished
by some responses later. Not the thumbs ups on social media that you
would get now. After publication there was still room for helping hands
to fix the one or other typo.

So redefining the whole process into some group effort without the
possible way to do it on your own and produce quick unauthorized
translations, like Jens Oliver Meiert said, will not help.
This works very well if you put 10.000 $ on the table and ask people to
do the work though. And that is where most of the people talking about
"professional" translation tools come from. Most W3C translators are not
professional translators.

Jens Oliver Meierts comment on "show some appreciation" for the
translators is right on. Nobody does this for a long time when you don't
even get proper recognition. I remember being invited to some talks and
some events, that helped a lot with the motivation because you could
actually talk to people working on the specs or somewhere around it.

But that was in one or two years and that effort to bring translators to
one table was lost very soon.
Right now I would like to hear how many people writing here asking for a
huge process for translations are volunteers who do that for fun and how
many do this here to make a living?

So if you want a complicate process, please translate a medium sized
specification, not in your paid working hours though, just do it for
free, and then please help us with your valuable experience.

I am just cruising into my 20th year of self employment this November,
seeing the time I can spend on Open Source or voluntary projects I am
quite lucky to have had such a luxury. Not everyone has that and giving
some good motivation is a good idea to squeeze some time out.

Remembering the words of Steven Pemberton (on another topic) many years
back as if was yesterday: "IT's A MESS."
So let's do something, but not too much, otherwise the waggon get's stuck.

Yeah, I started a short translation today, but I won't tell you which
one until it is ready. Because my last announced one got stuck and will
probably never get finished, because I wasn't rich enough to pull it
through.

Ok that was a bit long now
Stefan


>>> b) some solid ways of recognition, at the least including names of
>>> translators and links to their personal sites and portfolios
>>
>> Thanks for that feedback, this is extremely useful. I had been wondering
>> what would provide the right incentive and the much-deserved recognition
>> to translators in this new approach, and had indeed imagined providing
>> some kind of public profile for translators - I would be happy to hear
>> from more translators on this, but will already look into increasing the
>> priority of that aspect of the program based on your input.
> 
> If the translation project is hosted on a collaborative site like
> github, seeing who has contributed is trivial (although contributors
> will need a github account). Putting their names on the project readme
> file should be enough to give them proper recognition and since the
> information is on the github site new contributions do not require the
> document to be republished to update the contributors list.
> 
> 
> Jean-Christophe Helary
> -----------------------------------------------
> http://mac4translators.blogspot.com @brandelune
> 
>

Received on Thursday, 11 October 2018 17:59:35 UTC