This is a discussion on Re: building perl outside srcdir (was: Re: CVS: net-snmp configure, - SNMP ; On 25/09/06, Thomas Anders wrote: > Wes Hardaker wrote: > > Log Message: > > tests to see if we're in the main source tree so we can install the perl modules or not > > I strongly object. I'm ...
On 25/09/06, Thomas Anders
> Wes Hardaker wrote:
> > Log Message:
> > tests to see if we're in the main source tree so we can install the perl modules or not
> I strongly object. I'm making use of the mentioned symlink tricks ("shadow
> trees") to successfully build outside the source tree *with* perl enabled in all
> of my nightly builds. I don't think we can afford to lose that aspect of build
> testing, especially since embedded perl is now the default.
I haven't checked Wes' patch, but does this *force* the disabling of
the perl stuff? Or just affect the default build.
> What can we do? Offer an override flag? Make it a warning only?
$ cd ~/MAIN-SOURCE-TREE; ./configure ; make ; make install
// builds and installs perl modules
$ cd ~/SHADOW-TREE; ./configure ; make ; make install
// omits embedded perl and perl modules
$ cd ~/SHADOW-TREE; ./configure --with-embedded-perl ; make ; make install
// builds and installs perl modules, etc - as requested
The details may not be quite correct, but you should get the idea.
I've no idea how this matches against the current state of play.
I'm also a little concerned about how much tinkering this might
involve to get right.
We're into pre-release mode now, so need to be concentrating on that.
"Unusual" configurations (such as shadow tree builds) may have to take
something of a back seat until we've released 5.4. Not inevitably,
but if there's a conflict between fixing stuff for such a
configuration, and getting 5.4 out of the door, I don't think we
should delay things unduly.
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
Net-snmp-coders mailing list