[syslinux] submenus and menu title

[Ady Ady]

> Date: Fri, 3 Aug 2012 11:58:27 -0700
> From: hpa at zytor.com
> To: syslinux at zytor.com
> CC: ady-sf at hotmail.com
> Subject: Re: [syslinux] submenus and menu title
> 
> 
> That looks like it should work. We might have a bug there.
> 
> -hpa
> 
Hello HPA,
 
Here is a resume of the worst cases and what SHOULD had happened.
________
#CASE 4
MENU BEGIN tag4
LABEL lab4
COM32 hello.c32
MENU END
________
________
The (sub)menu title:
________________
Should be          : parent menu title (inherited)
Result using v.4.05: blank (no subtitle)
 
________
#CASE 8
MENU BEGIN
LABEL lab8
COM32 hello.c32
MENU END
________
________
The (sub)menu title:
________________
Should be          : parent menu title (inherited)
Result using v.4.05: blank (no subtitle)
 
 
Conclusion:
___________
 
If no MENU LABEL nor MENU TITLE directives are used in the submenu, then no submenu title is displayed at all
(the submenu title is not inherited from the parent menu, as it should be).
 
This points to at least one BUG: the MENU LABEL directive (and/or tagname) inside the submenu always takes precedence over the inherited submenu title regarding the displayed submenu title. Although someone could advocate for this situation, it is certainly incorrect when no MENU LABEL is used for the submenu entry in the parent menu, and it goes against the general logic of all other directives in submenus.
 
There are at least 3 more cases of submenus where the precedence (which directive overrides the other) is wrong, but the 2 examples here are the ones that contradict the normal expected result (what SHOULD had happened) the most. The (incorrect and unwanted) result is that there is no case (at all) where the parent menu title is inherited.
 
Correcting the directive's precedence in those cases is completely safe, and will not affect any current menu result in any negative way.
 
I can provide a table showing all the cases, with the current result against what should had happened, and I also have popular ISO images using submenus, so I can test and compare those results too, once a bugfix is out.
 
HPA, is there any chance to resolve this issue, please?