[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

behavior differences compared to SUSv2

From: Stew Benedict
Subject: behavior differences compared to SUSv2
Date: Mon, 14 May 2007 10:37:26 -0400 (EDT)


I'm working on the LSB project, and we have a bug filed on the 
specification, with regards to the behavior of keyname() function.


The LSB spec references SUSv2 for this function, but in practice, most 
Linux distributions use ncurses.

The confusion seems to be coming from this bit of the SUSv2 spec:


    The string has a format according to the first applicable row in the 
following table:

    Input       Format of Returned String
    Visible character   The same character
    Control character   ^X
    Meta-character (keyname() only)     M-X
    Key value defined in <curses.h> (keyname() only)    KEY_name
    None of the above   UNKNOWN KEY

    The meta-character notation shown above is used only if 
meta-characters are enabled. 


The ncurses manpage makes no mention of needing meta-characters enabled 
and the sample code attached to the bug seems to reflect that indeed it
is not needed.

Thoughts? Is this the intended ncurses behavior? Should LSB point to the 
ncurses behavior or at least describe the difference from SUSv2 in the 

Stew Benedict

reply via email to

[Prev in Thread] Current Thread [Next in Thread]