Hello.

On 6/3/08, José Luis Redrejo jredrejo@gmail.com wrote:

2008/6/2 Yoshiki Ohshima yoshiki@vpri.org:

...

Hello,

We are still trying to package the Etoys for non-XO environments, along with, of course, for XO.

Currently, end of June is the *real* target of release day of non-XO packages. The idea is to put relatively small set of languages in it, and it looks like English, German and French are most complete ones. (Thank you!)

But, for the US audience, (new world's) Spanish is also important. And Portugese (br and pt) seems to be fairly good to me, but of course I can't tell the quality right away.

Can you help us giving the information of the translation quality and completeness for these languages? For the non-XO environment, we will not have good font rendering mechanism other than what we have now, so still the main target is the Latin languages.

The completeness doesn't have to be so close to 100%, as now many of the phrases are not for Etoys (rather for the Smalltalk development). But we would like to hear the opinions of the native speakers.

Please visit

https://dev.laptop.org/translate/

and see how far the translation are going in terms of quality and coverages. Bert told us that there is an old style .translation file for Spanish. We can convert it to .po file, but not sure again if we should merge them to the site, as it might wipe out newly translated phrases. Does anybody has good ideas how to proceed?

Yoshiki, as you know, in LinEx we have all the strings translated for the etoys image (plus some other strings from some packages as drgeo, skeleton, etc.). Antonio Moreno has contacted with some of the etoys people, but the answers are not really clear about the way you want us to give you the translation. Our current image is totally based on the current olpc image , with changesets added from us, but we have all the classes and objects the etoys image has, so if you give the exact way you want us to extract the translation you can have the spanish translation finished (and tested & checked) right now.

Thank you for sharing your work with us ! And sorry, I wasn't enough clear in my recent mails in list.

For OLPC etoys we are using gettext (emulation) and that have changed how translation dictionary is managed. In legacy translation engine, there could be obsolete or unused string in dictionary. Now we extract strings that are receivers of #translated or #translatedNoop as POT and populate translations into that. There can't be obsolete or unused strings in dictionary, as long as dictionary and source code are in sync . So legacy translation data needs to be "finalized" before it will be merged into the working version on Pootle. I described the finalization process in http://lists.laptop.org/pipermail/etoys/2008-June/002379.html (see step (1) and (2)) What I attached to http://lists.laptop.org/pipermail/etoys/2008-May/002362.html was the results and I suggested people to try it.

Some strings are missing in extracted POT while that are successfully translated in the legacy image. Because the strings aren't the receiver of #translated and gettext exporter can't extract them. There are missing strings also from migrated your translations. we have been trying to fix the problem (see https://dev.laptop.org/ticket/6704 for example) and we will fix also for yours.

About the newly translated phrases, if our image has been used and tested for several years, with teachers fixing translations when some mistake was found, what's the problem on wiping out newly translated phrases that has not been checked for so many people?

I (and Yoshiki, I think) couldn't discuss about quality of the translation language not familiar to me. I could just point out the fact that there are conflicts between yours and the one on Pootle (attached to http://lists.laptop.org/pipermail/etoys/2008-June/002379.html ) .

I think it should be no problem to wipe out the working version on Pootle if people like you think yours is authoritative. (but you might want to review the conflicts anyway). We would like just to see Spanish translation team agree on this, though I don't know who they are.

Thank you! /Korakurider

Regards. José L.

2008/6/5 Yoshiki Ohshima yoshiki@vpri.org:

José,

...

Yoshiki, as you know, in LinEx we have all the strings translated for the

etoys image (plus some other strings from some

...

packages as drgeo, skeleton, etc.). Antonio Moreno has contacted with

some of the etoys people, but the answers are not

...

really clear about the way you want us to give you the translation. Our

current image is totally based on the current

...

olpc image , with changesets added from us, but we have all the classes

and objects the etoys image has, so if you give

...

the exact way you want us to extract the translation you can have the

spanish translation finished (and tested &

...

checked) right now. About the newly translated phrases, if our image has been used and tested

for several years, with teachers fixing

...

translations when some mistake was found, what's the problem on wiping

out newly translated phrases that has not been

...

checked for so many people?

Sorry for not being clear. Korakurider has responded in more detail and we definitely proceed in the right way. As he wrote, if you think it is fine (and perhaps it is more consitent to use phrases from one team) to "wipe" the existing one (as probably most of them are derivatives of LinEx work). Some people may get a bit of bad feelings... So, please speak up before that.

Yes, I know, it's now on Antonio Moreno's shoulder. He's checking all the conflicts, I also took a look to them and some of them are clearly bad translations because of being literal dictionary translations, not using the context.

Another thing is the difference of european Spanish and new world Spanish. I don't know how much it really matters but the terminology of new stuff like around computer can be quite different (ordenador <=> computer^^;)

Yes, but the LinEx translation begun in the smallland group (in 2002/03) where all these words were discussed between people from different countries, mainly Argentina and Spain, but also Peru, Mexico, etc. so we reached consensus on most of them. And, really, the doubts were very few words that (as far as I seen) are not in the conflict list that Korakurider has sent.

What we can do is to save the current version just as back up, upload the version from LinEx, and make sub language group later.

I hope that won't be needed, let's wait until Antonio Moreno check all the conflicts, but as I told you above, I guess there won't be much differences when the errors are fixed.