Adding operating system detection to configure for bug 1209

Umberto Nicoletti umberto.nicoletti at GMAIL.COM
Fri Apr 7 09:47:44 EDT 2006


On 4/7/06, Frank Warmerdam <warmerdam at pobox.com> wrote:
> Umberto Nicoletti wrote:
> > A month has passed without any objections so I believe no one is
> > opposing this change.
> >
> > After some thoughts I am going to call the configure option --with-os
> > (I dumped the name part which is redundant) and I will commit to cvs
> > HEAD next week (last call for comments).
> ...
> >> - configure option name and explanation text
> >> - os checking code: should we use the AC_CANONICAL_SYSTEM macro instead?
>
> Umberto,
>
> Sorry, I never noticed this the first time around.
>
> Does your patch use AC_CANONICAL_SYSTEM or does it require the configurer
> to supply the os name?  It seems unfortunate if we would always force
> builders to know the java include directory name for their system if
> we could avoid that somehow, even just with a good default.
>

I have tried AC_CANONICAL_SYSTEM but it requires install.sh so I
decided to go with a lighter implementation. If the option is not
explicitly set configure will try to detect it by using the uname
output ( WARNING: could this break Windows builds? ).

> I'm a little leery about calling the option --with-os as I can't see
> many non-Java applications.  In fact if there were another package with
> similar issues it seems likely that it would use a slightly different
> naming scheme.  I think I would prefer something like
> --with-java-include-sys-name if possible.
>

What about the shorter --with-java-include ?

Thanks for your reply,
Umberto

> All that said, my objections are just suggestions.  I can live with your
> approach.
>
> Best regards,
> --
> ---------------------------------------+--------------------------------------
> I set the clouds in motion - turn up   | Frank Warmerdam, warmerdam at pobox.com
> light and sound - activate the windows | http://pobox.com/~warmerdam
> and watch the world go round - Rush    | President OSGF, http://osgeo.org
>
>



More information about the mapserver-dev mailing list