Comment 1 by jteh on 2009-06-24 01:42
Changes:
Milestone changed from 0.6 to None

Comment 2 by Bernd on 2011-03-06 14:09
What about asking the translators on the developers list? I think all translators could comment to this ticket. Maybe we could also add serveys to later questions.

Comment 3 by Bernd on 2012-06-11 17:14
As we now have the translators Mailinglist, we could ask the Translators to note the preffered table in a list so someboddy could do the job. Alternatively we could use Computerbraille which belongs to a language. For most languages we have almost one translation table.

Comment 4 by nvdakor on 2013-01-25 08:35
Hi,
Commenting a bit late...
If you want, I can ask around for suggestions on which table we should use. I see two problems:

• Some languages, such as English, provides two or more dialects and/or tables. In this case, it would not be a good idea to enforce one table per language unless if we can detect the country code.
• Not all languages have a braille table. If so, one option is to choose the last table used. Alternatively, we can set an "automatic" braille table so langs which doesn't have a braille table can fall back to this table.
  For putting this into source code, we could have:
• Pair NVDA interface language with the appropriate braille table. In this case, the interface lang value in settings file would also contain the braille table name. This allows future expansion to speech language setting.
• Add a field in braille tables list in braille.py - a third field which specifies whether the given braille table file should be loaded when a particular interface of NVDA is chosen. This requires concensus on which table to use if a language provides more than one table (such cases are Chinese and English).
  Thanks.

CC @dkager @jcsteh @bdorer @josephsl

Labeled as feature. Some thoughts.

It seems that @jcsteh doesn't want to rely on liblouis metadata alone, which I understand and agree with. @dkager, you might have a better overview on how the metadata thing is going?

Personally, I think we should allow language maintainers to set their prefered defaults with some restrictions (e.g. no grade 2). Such an automated selection system could be broader in scope. For speech, I belief selection of the desired language is done in the espeak driver, which is a bit ugly imo.

Related: #6952 (comment)

@dkager, you might have a better overview on how the metadata thing is going?

Slowly. :)
It's hard to standardize. One country's grade 1 is uncontracted, another country calls that grade 0. Yet others say grade 0 is computer braille.

The new brailleTables module introduced in #6449 makes our own table metadata a lot more extensible. I think we could add a defaultForLanguages argument to the addTable function which specifies a list of languages for which this is the default. To be future proof, I think we should allow language defaults to be specified for both contracted and uncontracted. For now, we can just choose uncontracted. In future, though, we may want to make them separately configurable to allow fast switching, uncontracted input outside of editable content, etc.

One concern I have is how we should deal with dialects. For example, there is a Dutch (Netherlands) and Dutch (Belgium) table, however there is only one single NVDA language for both dialects. Using the windows language for this could be sufficient, of course.

One result that should come out of the Braille Authority is that these two tables get merged. But I imagine some Asian countries have the dialect problem too. English countries had it, but UEB covers most or all of that.

one easy way would be the following:
Ask the user in the welcome dialog wheather he wants to use braille and provide him one of the table wich fits to his language.

Since this was asked on Twitter, would setting the Braille table for the current language be done just when installing, or when detecting changes of language in text (like the synthesizer changing language on the fly)?

Changing the braille table dynamically based on content language should probably not be the default behavior. You don't necessarily have knowledge of all the braille tables for the languages that you understand.

If there will be an option to dynamically switch braille tables it will be helpful if the user can also specify if she wants 6-dot/literary braille or 8-dot/computer braille.