As someone that's managed localization for a few projects, I would say that we should try Transifex. It looks like it has all the tools needed for the job.

By the way, I can't stress enough the need for a way to see the strings in an emulated UI as mentioned earlier in the thread. A lot of people that have never worked within localization on smaller formats such as cell phones don't realize how much MORE screen real estate is used by translations in comparison to the original English strings.

I've worked on 10 different games for cell phone platforms which were translated into about 15 different languages and 6 or 7 encodings. Just because English fits there doesn't mean other languages will.

As a general guideline, French strings will be anywhere from 50% to 80% longer per string and German is between 80% and 120% longer per string. (Germans seem to enjoy joining together multiple words into one long 20 character word for some reason.) Most of the other languages shouldn't be as long.

None of my work projects had the ability to directly view strings in our game UI, unfortunately.. so it meant we had long periods of going back and forth from testers to localization until each string worked. Let's not allow that to happen in Maemo, please.

Maemo Summit? Woohoo, I'd be thrilled to be there and discuss all our localization needs (new features, Maemo-special ones, integration with developer and desktop tools).

On a sidenote, we just released 0.7 Release Candidate 1 with some nifty features:

• Online Translation Editor
• More control to Project Maintainers (register and edit your projects, control who can submit files)
• Submissions via email
• User Profiles w/ Google Maps integration
• Integrated support for Publican-like (multi-POT) projects

For more information, including links to pre-built and preconfigured ISO, VMware or EC2 images (x86 and x86_64), take a look at the 0.7 release notes.

About integration with other tools, we bootstrapped the base for an API, as the base for such support.

So is there anything happening on this front yet? Is it still too early to make a decision? Has anything advanced at all?

I hate to be a pain, but I'm just sick of things being discussed, decided upon and then forgotten. (This line is just a general reference to things I see in my life, not anything having to do with Maemo in particular.)

If nothing has changed since this topic started, what needs to be done on my/our end to make this happen? What else can we do to push this along and make community-based translation/string fixing a reality for Harmattan?

I'm also interested how this is going. I have done also some localizations and I would like to see this going forward.

A good scenario would be:

- The Maemo community decides that Transifex is the way to go. Council + maemo.org dev team explicitly agree.

- Planning phase starts: what concrete objectives do we have? what needs to be done in maemo.org? does Transifex need any customization or extra development? does this project need any extra funding from Nokia?

- A plain Transifex instance is available in a maemo.org URL so we all can start playing with it. For instance Extras apps can see how to use it. We would get the Harmattan program to play with it as well. Maybe something can be done with Fremantle at a test/unofficial level.

- The Maemo Summit would be a place to meet and make some concrete progress.

Guess I'll put in for sponsorship, then.

I've decided to split this up into several task proposals for Maemo's sprints. I've created the first task proposal and will be writing up an overarching Wiki page covering all the tasks that will need to be pushed/done to make Community Translation a reality.

You can find the first proposal here: http://wiki.maemo.org/Task:Evaluate_Transifex

Please feel free to let me know what you think or if there's anything else that you feel needs to be added to it.

Glezos, I would really appreciate your help and some comments on this so we can push this forward.

What's the reason for the need for an upfront decision? I've not read every post in this thread in-depth, is there a licence cost for Transifex that Nokia'll be paying for? Is there an enormous investment in maemo.org infrastructure which will distract the gang-of-four from their other tasks?

What we need is someone passionate to pull together enough interested people and technology to provide a mechanism for community translations of their app, which gets used by someone else, which gets used by someone else. Example apps would include Conboy (where conny's been very good at including translations from the community submitted through tmo).

I can't see a dictat coming down that "Transifex is the community's preferred l10n/i18n solution" working, when it's not been trialled by the community.

Come to me with actual results and the council'll make a decision for the community. However, unless we can't avoid it; we're not in the business of making decisions for the whole community (or pretending to).

I think you might have jumped the gun a bit here. Quim just skipped the "community tests Transifex" stage which is kind of implied.

No, there's no fees or anything for Transifex... and yes, we've already got a test subject: Mer. (And if any other smaller projects out there wish to be a part of the testing and evaluation phase, by all means let us know.)

All of this is outlined in the Task page already... unless I completely misunderstood what Quim was saying.