Possible Characters issue

PS: the wf was released 1 year ago with no issue, even on the sim when I worked on it.

hi Jim, I use Eclipse and UTF8 option is checked

Should I open a ticket or did I miss something ? 

I'm afraid I don't fully understand the problem you're describing.

var font = WatchUi.loadResource(Rez.Fonts.MyFont);

var cx = dc.getWidth() / 2;
var cy = dc.getHeight() / 2;
var fy = dc.getFontHeight(font);

// I put manually in dc.drawText => "říjen", no issue
dc.drawText(cx, cy, font, "říjen", Graphics.TEXT_JUSTIFY_CENTER);

cy += fy;

// I assume you are doing something like this, but where does `text` come from?
dc.drawText(cx, cy, font, text, Graphics.TEXT_JUSTIFY_CENTER);

Could describe how you are getting the string that fails to draw properly?

sorry for the bad explanations,

// IF I DO THAT, NO ISSUE, NO SQUARES
var text = "říjen"

dc.drawText(cx, cy, font, text, Graphics.TEXT_JUSTIFY_CENTER);


// BUT IF I DO THAT: SQUARES, WHERE MONTH IS SUPPOSED TO BE "říjen"
var text = Date.info(Time.now(), Time.FORMAT_LONG).month;
dc.drawText(cx, cy, font, text, Graphics.TEXT_JUSTIFY_CENTER);

Seems the strings are different, though the chars may look the same.

Try using toCharArray() on both and see what the value for each character is in both.  In your .fnt file you can just have both values pointed to the same thing in the png.

Here's an example in the sim with the degree symbol for a 245.  The lat/lon is generated with a call to toGeoString()  and the degree symbol shows as a diamond with a ?, yet when I display the temperature using 

var degree="°"; //utf-8, b0 hex, 176 dec

it's fine (bottom of the screen).  This is with a native font, and I see the degree symbol on the real device in both cases.

many thanks,

but is is odd, the last 8 months, no issue and now I have it (sorry can't tell from when)

an other test I did:

var M = "октябрь";Sys.println(M.toUpper().toUtf8Array()+"   "+M);
var M2 = date.month;Sys.println(M2.toUpper().toUtf8Array()+"   "+M2);

I don't have the same value at all, an utf8 code is supposed to be always the same isn't it?

if I do the same test with M="octobre" and simulator language in French, no issue.

It seems that the issue come from Garmin Translation.

On devices, in the Garmin\text folder are the translation files for languages, and the do get updated at times.

Maybe there was a new one for a languages with alternate characters.  But with a custom font. as I said you can just do some minor editing to the .fnt file to handle them.

Many thanks, it is what I'm doing ! only eddit the .fnt,

but something I don't get (probably because I'm new in dev world)

October will always be called October

an O will always be an O so the char code will always be the same won't it?

So I guess it is the same for "special characters" language?

so for exemple říjen will remain říjen and the ř will always be 345?