On Sun, Jul 23, 2023 at 11:17:38PM +1000, Mark Jamsek wrote:
> But if we keep the author colour scheme all the way up to the '*' (as
> in the below diff), it looks a bit better:
> 
>   https://ss.jamsek.net/20230723T230108-ss.png

I prefer the above. Having the * appear in blue visually ties it to the
author, which seems wrong to me. It should be tied to the commit.

> Full disclosure: I haven't tested on a dark colour scheme so this could
> be entirely the fault of my environment.

My patch looked fine in a dark scheme but I agree that yours looks
much better on a white background. It also looks fine in a dark
terminal so your patch is ok by me, if you don't mind :)

> diff /home/mark/src/got
> commit - 6ed9ffec8b7318ac6f69a138854afefc48bca2c0
> path + /home/mark/src/got
> blob - 7850a8a50fe8ff23c3f0d68ab713639f97d85090
> file + tog/tog.c
> --- tog/tog.c
> +++ tog/tog.c
> @@ -2484,9 +2484,16 @@ draw_commit(struct tog_view *view, struct commit_queue
>  	while (col < avail && author_width < author_display_cols + 2) {
>  		if (tog_base_commit.id != NULL &&
>  		    author_width == marker_column &&
> -		    entry->idx == tog_base_commit.idx)
> +		    entry->idx == tog_base_commit.idx) {
> +			tc = get_color(&s->colors, TOG_COLOR_COMMIT);
> +			if (tc)
> +				wattr_on(view->window,
> +				    COLOR_PAIR(tc->colorpair), NULL);
>  			waddch(view->window, tog_base_commit.marker);
> -		else
> +			if (tc)
> +				wattr_off(view->window,
> +				    COLOR_PAIR(tc->colorpair), NULL);
> +		} else
>  			waddch(view->window, ' ');
>  		col++;
>  		author_width++;
>