fixing lang/spidermonkey

[Dan Rench]

On Tue, Aug 09, 2005 at 01:39:05AM -0400, Joe Marcus Clarke wrote:
> On Mon, 2005-08-08 at 15:18 -0400, Mikhail Teterin wrote:
> > Hello!
> > 
> > The attachment contains an upgrade to lang/spidermonkey. The benefits are:
> > 
> > 	. build on amd64 at all and _properly_ on all other platforms
> > 	  (-fPIC -DPIC should be used for shared objects everywhere);
> > 	. do not build vendor's own version of libm (fdlibm) -- FreeBSD's
> > 	  -lm provides sufficient functionality (mozilla, firefox contain
> > 	  the same foolishness, BTW);
> > 	. uses -lreadline instead of building vendor's own "editline";
> > 	. unless WITHOUT_TESTS is set, download, patch and run the vendor's
> > 	  set of self-tests;
> > 	. patch to compile cleanly with '-Wall -Werror'.
> > 
> > Mozilla people indicate, that they plan to make a new release of spidermonkey, 
> > when Firefox-1.5 ships (circa September):
> > 
> > 	https://bugzilla.mozilla.org/show_bug.cgi?id=303857
> > 
> > When that takes place, spidermonkey can be used _instead_ of browsers' own 
> > snapshots of js thus reducing the porting effort, cutting build-times, and 
> > providing a single place for patches.
> > 
> > Considering that, on one hand, the port is marked broken on amd64 since March 
> > 2004 (a tier-1 platform), and, on the other, nothing depends on it (no new 
> > possible new breakage), I ask portmgr to also allow me to merge these fixes 
> > before release -- unless the maintainer objects in a few days.
> 
> Approved assuming the maintainer accepts.

Yes. Looks fine to me. Go ahead.

As for contacting the developers about these fixes, I'm going to ask
Mikhail to do that as my working knowledge of C is (we'll say) somewhat
'limited.'