[Tux4kids-discuss] [Tuxmath-devel] Further clarifications about maintainance

[Pere Pujal i Carabantes]

El dj 09 de 06 de 2016 a les 13:13 -0400, en/na LM va escriure:
> scottmc wrote:
> > 
> > We now the Tux4Kids github site up and running and have cloned most
> > of the
> > repositories over to it.  We have now started adding more team
> > members.  If
> > you are interested in being a team member or stepping up to be a
> > maintainer
> > for any of programs, let us know.  Once we have the maintainers in
> > place we
> > should be better able to handle Pull Requests and new issue
> > reports.
> > 
> > We have a few of the students who applied to work with Tux4Kids for
> > Google
> > Summer of Code express interest in working on their proposed
> > projects
> > outside of GSoC.  We have asked them to join the tux4kids-discuss
> > mailing
> > list and to ask questions here if they needed any help.
> I'm not a student, but I am looking at adding SDL2 support to
> t4k_common, tuxmath and tuxtype.

I saw your mail last month, sorry that I can not give more help than
testing. Anyway, if you want access to the tux4kids github, just let it
know.

> 
> At the moment, I'm trying to figure out the best way to keep
> internationalization support
> working properly with SDL2 (or to improve it).  I found out some
> interesting things
> regarding this.  I'm using SDL_ttf/SDL2_ttf in my builds instead of
> Cairo.  SDL_ttf and
> SDL2_ttf only support the UCS-2 character set, not the entire UTF-32
> character set.  Would
> think the UTF-8 functions would handle the entire UTF-8/UTF-32
> character set, but they only
> handle 2 bytes.  Some programs (such as sdl_unifontview) use a
> modified version of SDL_ttf
> so anything above 2 bytes isn't cut off.  I'm experimenting with some
> of those modifications
> now.
> 
> The way tuxmath and tuxtype are currently written, if one builds with
> SDL_ttf/SDL2_ttf
> instead of Cairo, internationalization support is limited to 2 bytes
> (UCS-2 character set).
> If it's built on Windows, wchar_t is 2 bytes, so even if it's built
> with Cairo,
> internationalization support is limited to 2 bytes.  Also, thought it
> was strange that
> t4k_tts.c uses espeak_Synth with espeakCHARS_AUTO but it's passing
> the
> variable text
> using tts_argument and text is wchar_t.  According to the espeak
> documentation,
> espeakCHARS_AUTO is "8 bit or UTF8".
> 
> Would be curious to know how many people are actually using the
> internationalization
> features and if anyone's willing to test them if modifications are
> made.  Thanks.
> 

I'd assume that any not native English speaker would use
internationalization, I for one would want to help by testing, maybe
even I could learn something from you to apply into Tux Paint.
Will need access to the modified source code :)

Best
Pere