To contribute to a feasibility study of including translations in RTL languages in Slint installers, I have uploaded 'bogl' build stuff and packages for Slackware version 14.1
here.
Basically, the packages ship the 'bterm' terminal used by Debian's polyglot text installer, alongside an 'unifont' font, in two formats:
- unifont.bdf, found in the the source tarball[1]
- unifont.bgf used by bterm (the utility bdftobogl used to make the conversion is also shipped in the package)
Although the 'unifont' font be not specially well looking IMHO, this offers the advantage of using one single font for a lot of languages without X, in a framebuffer (needed by bterm to draw the characters on the screen): instead of being limited to 256 or 512 glyphs as in a raw Linux console, you get no less than 57086 characters at no additional cost.
If you'd like to try it, become root and in a (framebuffered) tty type as root "bterm.sh" (this wrapper scripts sets the font to unicode.bgf, see "bterm.sh --help").
You'll be dropped in a login shell and will have to log in as a regular user (but you may "su" after that).
While we are at it, if you want to get a wide range of characters on the console but prefer a better looking and scalable at will TTF font, like those included in the DejaVu collection, you could use fbterm, available @
http://slackbuilds.org.
For obvious reasons, to use either 'bterm' or 'fbterm' you will need an UTF8 encoded locale.
HTH some.
PS If you prefer to use another font than unifont.bgf in bterm, you can convert any to BDF with fontforge then convert it to BGF with:
Code:
bdftobogl -b font.bdf > font.bgf
I have uploaded a few ones
here. They don't look so good, that's my fault: I am not used to fontforge yet and probably used bad settings or options for rasterization. But you get the idea.
You can then start bterm like that from a framebuffer:
[1]See the
Glyph Bitmap Distribution Format (BDF) Specification from Adobe. BDF is a rather simple text format (see an example
here) so you can use the "reduce-font" utility shipped in 'bogl' packages if you just need a subset of the glyphs. Writing instead a simple sed script to reduce the font to a set of ranges of Unicode code points is left to the reader as an exercise