> -----Original Message-----
> From: Richard Harke
> Sent: 30 June 2004 6:04 am
> To: FlightGear developers discussions
> Subject: Re: [Flightgear-devel] Rebuild
> On Tuesday 29 June 2004 09:57 pm, Tony Peden wrote:
> > If you change an oft-included header in simgear, this is 
> pretty normal.
> Well, that would explain it. I did do a make install on simgear
> I guess I should have just cp'ed the relevant .a file.
> >
> > On Tue, 2004-06-29 at 20:02, Richard Harke wrote:
> > > I have found a bug in libGLcore.so from Nvidia for ia64. I think
> > > I have a work-around that involves a small change (temporary)
> > > in simgear. after making the changes, I deleted the .o 
> and corresponding
> > > .a and ran make. Seemed to work fine and just re-compiled 
> the one file
> > > and did the one link.
> > >
> > > Then I deleted fgfs and ran make on flightgear and it seems to
> > > be recompiling the whole world. I didn't change any flightgear
> > > files so I am rather puzzled.
> > >
> > > Is this normal bahavior for flightgear make?
> > >
> > > Richard Harke
> > >

AFAIK, by default (on cygwin at least), install sets the file date on the installed 
copy to the date of installation, not the date of compiliation, so running make 
install in SimGear makes it look as if the whole of SimGear is new. 
There is a flag you can pass to install (-p perhaps?) which tells it to use the date 
on the original file, not the date of installation, so the impact on subsequent makes 
of changing one SimGear file is reduced to the degree that it should be.

Does that make sense?


This e-mail has been scanned for Bede Scientific Instruments for all 
viruses by Star Internet. The service is powered by MessageLabs. For
more information on a proactive anti-virus service working around the
clock, around the globe, visit: http://www.star.net.uk

Flightgear-devel mailing list

Reply via email to