bug-gv
[Top][All Lists]
Advanced

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

[bug-gv] RE: gv-3.6.1 / gv-3.6.0 on solaris -- setenv problem


From: jemarch
Subject: [bug-gv] RE: gv-3.6.1 / gv-3.6.0 on solaris -- setenv problem
Date: Wed, 12 Jan 2005 16:57:46 +0100

Hi Niki.

    On 11-Jan-2005 Niki Waibel wrote:
    > just figured out that 3.6.1 also uses
    > setenv and unsetenv. solaris has getenv and
    > putenv only.
    > 
    > pretty sure all this can be done easyly
    > using autoconf.
    
    regarding the setenv problem i've just seen that
    there is still code in gv-3.6.1 to make it working
    on solaris.

Yes. gv is shipped with an implementation of setenv functions.
    
    the only problem is that there is a little bug in
    configure.ac:
    ===
    AC_ARG_ENABLE(setenv-code, AC_HELP_STRING([--enable-setenv-code],[Use extra 
code for setenv() and getenv(). Defaults to `yes'])
    ,
    [case "${enableval}" in
       yes) opt_setenv_code=true ;;
       no)  opt_setenv_code=false ;;
       *) AC_MSG_ERROR(bad value ${enableval} for --enable-setenv-code) ;;
     esac],[opt_setenv_code=false])
    ===
    
    Defaults to `yes'
    =============^^^
    is wrong.


Ok. It should defaults to `no' (use the system setenv functions when
available).
    
    i think configure should check itself if
    the functions (setenv/unsetenv) are available.
    then you dont need the AC_ARG_ENABLE at all.

I fully agree. I am writing a patch for seek setenv/unsetenv on the
system, and only falling back on the shipped setenv implementation
when necesary. 

About the getopt trouble... see
http://lists.gnu.org/archive/html/bug-gv/2005-01/msg00001.html for a
similar problem on HP-UX 11.0, where the standard c library do not
provide getopt_long. On HP-UX 11.0 getopt_long seems to be provided by
'sharutils'. Perhaps you could install sharutils on solaris.






reply via email to

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