Tree list x coordinate bug

After around this push the x coordinate is getting offset when more than one tree list is open. (#2)

Before:

After:

[dumblob]

Does this commit 75c7179 solve it for you (i.e. do you use the latest nuklear.h from the master branch)?

No, it's still the same, and yes i used the latest branch.

[dumblob]

@Hejsil do you still have some bits of it in your head to take a look?

[exchg]

Seems that the problem with restoration of x offset at this point:

void nk_tree_state_pop(struct nk_context *ctx)
{
    struct nk_window *win = 0;
    struct nk_panel *layout = 0;

    assert(ctx);
    assert(ctx->current);
    assert(ctx->current->layout);
    if (!current_layout_exist(ctx))
        return;

    win = ctx->current;
    layout = win->layout;

    /* Look here !!!  */
    /* layout->at_x -= ctx->style.tab.indent + ctx->style.window.padding.x; */
    layout->at_x -= ctx->style.tab.indent + (float)*layout->offset_x;

    layout->bounds.w += ctx->style.tab.indent + ctx->style.window.padding.x;
    assert(layout->row.tree_depth);
    layout->row.tree_depth--;
}

I am not familiar with Nuklear internals, so this is just a hint.

[Hejsil]

Hmm. It seems that #2 was not as trivial of a fix as I thought it would be. I'm on vacation right now, so can't take a look at it until Friday maybe.

[dumblob]

@Hejsil enjoy your vacation until then!

[Hejsil]

@exchg I've look into this a bit, and it seems you where strait on the mark. The nk_tree_state_pop function restores the layout, which means it should revert the layout and bounds changes that nk_tree_state_base setup, which it wasn't doing correctly.

Do you want to make a pull request so you get credit for this fix, or should I just do it?

[exchg]

@Hejsil Great! Do it please.