Re: improved console interface (was: Re: some console code checked in

[Gaute B Strokkenes]

On Tue, 23 Apr 2002, Marcus.Brinkmann@ruhr-uni-bochum.de wrote:
> On Tue, Apr 23, 2002 at 02:18:33AM +0200, Marcus Brinkmann wrote:
>> There are probably more points that I forgot now,
> 
> Oh, yeah, the exact data format of the mmap'able area.
> 
> We need to store the character value and the attributes (bold, color
> etc).  We can settle on some common attributes, and leave it to the
> display driver to deal with them, maybe the attributes even fit into
> a byte.  However, for the character value, we should store the
> UCS-4, eg wide chars.
> 
> The basic questions here are: Interleave the data (as we don't need
> 32 bits for the attribute we face alignments issues)

Any Unicode character will fit in 21 bits, so you have plenty of bits
left over for attirbutes of various sorts.  Furthermore, the Unicode
standard guarantees that this will alwasy be so.

They yucky bit is when you consider characters with width not equal to
one.  CJK ideographs are twice as wide as Latin charactes, and
composing characters generally occupy the same cell as the base
character.  This breaks the model with one character per cell,
unfortunately.

> or keep it in seperate "planes" (a bit evil when
> extending/truncating the scrollback buffer).  And what about
> endianess of the data?

-- 
Gaute Strokkenes                        http://www.srcf.ucam.org/~gs234/
..  I think I'll KILL myself by leaping out of this
 14th STORY WINDOW while reading ERICA JONG'S poetry!!