question about building mesa with the two scripts from Xorg wikis

[Dan Nicholson]

On Tue, Feb 3, 2009 at 6:18 AM, Joe Smith <stopads at yahoo.com> wrote:
> Hi,
>
> I noticed a discrepancy in the Xorg documentation that's been causing me some confusion:
>
> At http://wiki.x.org/wiki/ModularDevelopersGuide:
> When using the build.sh script, it says:
> "If you want to include Mesa/OpenGL support in the X server, you will need unpack the Mesa source code and pass the directory of the Mesa sources to the build.sh script."
> And you pass the mesa directory like this:
> -m path-to-mesa-sources-for-xserver : full path to Mesa sources
>
> Also, in the build.sh script, building the mesa sources is commented out, like this:  #build mesa mesa.
>
> While at http://www.x.org/wiki/Development/git:
> The mesa sources are built before the xserver, and the sources directory is *not* passed to the xserver build directives.
>
> So the questions are:
> 1) Does mesa need to be built separately, as demonstrated in the second link?  If so, why is it commented out in build.sh?
> 2) Should the --with-mesa-source parameter always be used, and if so, why is it not used in the second link?

Mesa should be built before the server, and it's no longer necessary
to use the -m/--with-mesa-source flag. The server will pick up the
installed headers it needs from mesa. The first page is out of date.

I have no idea why it's commented out in build.sh. You can certainly
uncomment it and give it a whirl. Building mesa should be about as
simple as ./configure. Development/git has a little more info about
that.

--
Dan