Re: [Kma-anyg] KMA 0.3.2 released

[Erwin K. <er...@ew...>]

Jarl van Katwijk wrote:

>
> Did you ran ./autogen.sh after editing configure.in? Doing so rebuilds 
> configure, and will probably fix compilation on freebsd4.x. Maybe you 
> can check.. I commited to changes to cvs just now.
>
> Jarl
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: IBM Linux Tutorials
> Free Linux tutorial presented by Daniel Robbins, President and CEO of
> GenToo technologies. Learn everything from fundamentals to system
> administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
> _______________________________________________
> Kma-anyg mailing list
> Kma...@li...
> https://lists.sourceforge.net/lists/listinfo/kma-anyg

autogen.sh introduces ltconfig to the configure script, which is not 
found by default, so the ./configure will fail afterwards. I must admit 
I did not delve into this too deeply. ltconfig is found in 
/usr/local/share/libtool13, but not in the directory where install-sh is 
located (kma/src directory).
Option: include ltconfig in kma cvs?

Next to that, I did some digging into the getopt_long not being defined. 
A number of non-linux systems do not define getopt_long. NetBSD has it, 
FreeBSD5 also (I believe), Solaris has iberty, FreeBSD4 has gnugetopt. 
General consensus in the discussions I found was to include
getopt1.c, getopt.c and getopt.h with the source (and have it compiled) 
and not rely on a system provided version.
People have been struggling with this for a long time: 
http://sources.redhat.com/ml/automake/2000-08/msg00013.html :-)

Last detail, on line 43 in main.c
      static struct option long_options[3] = {
                   {"config", 1, 0, 'c'},
                   //              {"daemon", 0, 0, 'd'},
                   {0, 0, 0, 0}
               };
it says three options will follow long_options, but the middle one is 
commented out, leaving only two.