Re: [Nano-devel] about support for more than 16 colors

[Brand Huntsman]

On Tue, 9 Jan 2018 21:47:30 +0100
Benno Schulenberg <address@hidden> wrote:

> Thanks.  Looking at the output of ./a.out 256, on my terminal #333333
> looks darker than its surrounding colors.  The same for #1A1A1A and
> #4D4D4D.  But #66666 and #808080 and above blend in fine with their
> neighbors.

That is the mapping of RGB format colors with identical channels to gray 
colors. There are 6 grayscale codes in the cube map, but 24+6 in the gray map. 
This allows someone to use gray0-100 or RGB formats to get grayscale colors. It 
looks like a glitch in the test output but will be good in practice. So 
technically, the gray0-100 format could be dropped as well.

Run a.out with 16 argument and then 8, see how the black in upper left corner 
of each color block got bigger. I'm thinking of adding a fake point in the 
8/16/88-color tables that would reduce the number of dark colors that get 
mapped to black. And probably also need a fake point to reduce number of bright 
colors that get mapped to white. This would reduce mapping issues with dark 
colors on a black background and bright colors on a white background.

Should I start patching this into nano?