LibreRouter: new project with software and HowTo booklets

Jump to navigation Jump to search

You do not have permission to edit this page, for the following reason:

The action you have requested is limited to users in the group: Users.


You can view and copy the source of this page.

Return to Thread:Support/LibreRouter: new project with software and HowTo booklets/reply.

The JSON format looks okay except we don't include the language code as a key in the file (it's already there in the filename). Would you consider changing that format?

Is the repository in GitHub your main repository?

I would request to add some message documentation to guide the translators (see https://www.mediawiki.org/wiki/Localisation#Message_documentation).

Nike (talk)20:45, 6 October 2017

Does taking the language code as a key out of the file just mean removing the line   "es": { As well as the closing bracket } ?

Yes, the repository in GitHub is our main repository.

We will add some message documentation to guide the translators based on the advice at https://www.mediawiki.org/wiki/Localisation#Message_documentation. It might take us a week or so to start adding that documentation.

Pato (talk)21:03, 14 October 2017

Yeah if the file looks like following, then we support the format.

{
  "key": "value",
  ...
}
Nike (talk)14:10, 15 October 2017
 

I'm talking with coders... we're moving forward on changes to be more compatible with TranslateWiki, and we're coming up with questions about TranslateWiki:

1) We're looking at 4-7 different pieces of software to translate. Each piece has it's own practices of what branch development happens in, so we're getting all that info in one place. Arranging direct push commit access for TranslateWiki to the development branch of each software is part of our conversations.

2) If one or two pieces of software are ready to connect with TW.net before the others, can we go ahead and connect those? If yes, how?

3) Can the TW.net platform translate the Readme.md file for each software? Or maybe we'll have to setup a separate help folder with a file in each language, like at https://github.com/LonamiWebs/Stringlate/tree/master/help -- would that work? Is there a TW.net parser-generator for MarkDown .md files?

4) Is there a smartphone interface for TW.net, either via a web browser or a special app? (It seems that using a computer with a keyboard and at least a 10-inch screen is easier for translating, but sometimes a smartphone is the only device available to a translator.)

5) How do we include screenshots in the translator notes (the qqq messages)?

Pato (talk)03:29, 11 November 2017

1) Are those pieces in different git repositories?

2) Yes we can. I need the basic info about the repository, push access, file format.

3) We do not have MarkDown parser at the moment. It could work as translatable page.

4) Amir is working on a Telegram app. There is also unmaintained android app. The web interface should accommodate tablets pretty well, but mobile phones are not well supported.

5) Uploading them here in translatewiki.net and using the wikitext syntax to include them. Can also put direct links to images, if they are too large, but then translators have to click.

Nike (talk)15:24, 12 November 2017

1) Yes, each piece in a different repo.

2) The first two pieces that I think are ready:

3) "It could work as translatable page." Do you mean it could work without a parser?

4) Cool! I hope Amir and crew get that Telegram app to a usable release.

5) Where in TranslateWiki would we upload the screenshots? When you say "too large", what are the maximum dimensions of the image (like 80px by 200px)? What is the maximum file size?

Pato (talk)23:22, 13 November 2017
 
 

6) Where is the code for the parser-generators that are currently in use on TW.net? We'd like to look at them to get a sense of what it would take to write another.

7) Any suggestions on how to internationalize this relatively simple web page interface, https://github.com/libremesh/chef ?

Pato (talk)14:36, 12 November 2017
Nike (talk)15:26, 12 November 2017

6) Thanks, we'll look at those.

7) The maintainer of the software to internationalize (Chef) says: "I'm currently not using jQuery in Chef but plain JavaScript. I'd rather keep it that way - if possible. Some quick searches resulted in https://github.com/eligrey/l10n.js/ - maybe thats already enough?"

Pato (talk)23:23, 13 November 2017
 
 
 

Also, what are the howto booklets you mention in the subject? Our primary focus is on software user interfaces. [edit] We don't currently support ODT files. Some kind string extraction to another format would be needed. Are the booklets created so that they can be automatically generated from the translated texts (i.e. no hardcoded positionings with length restrictions)?

Nike (talk)20:46, 6 October 2017

The HowTo booklets are a series of guides about the social, technical, and economic aspects of how to create and maintain a community network, with the LibreRouter as the example hardware and software. The idea is that each booklet will be laid out by hand after translating. This means that an automated workflow (as is used in software) is not our plan for the first three languages, and we would join TranslateWiki.net only for the software interfaces.

However, if anyone can suggest a workflow to create the booklets that aligns with the way TranslateWiki.net functions, we're interested to try it out since that could facilitate the translation of the booklets into many more languages. Perhaps the ODT parser from the Zanata translation management platform could be included in TranslateWiki.net, or perhaps we could use a different format. Or perhaps a bit of the BookType software used to create the books at Flossmanuals.net could be reused.

Pato (talk)21:22, 14 October 2017

I haven't done that before. I am sure there are string extraction plugins for ODT. The problem is that usually booklets have quite tight positioning and size constraints. Doing the layout manually for many dozens of languages (and dealing with possibly frequent updates) doesn't scale. But a flexible or simple ODT file we could probably support if there is a suitable parser-generator we can integrate.

Nike (talk)14:29, 15 October 2017
 

Hi Nike, any chance you can setup the repo for LibreMesh document translation this week? Following the thread at https://github.com/libremesh/lime-packages/issues/257#issuecomment-423850005

Pato (talk)19:49, 4 November 2018

This has been continued in chat discussions.

Nike (talk)15:23, 6 November 2018