[osg-users] ShaderComposer::releaseGLObjects warning

[Andy Skinner]

Thanks for that fix.  There is a remaining issue, and I'm not sure if you would see it as a bug or a dodgy compiler warning.  :)

We get a similar message about osgUtil::CullVisitor::clone().

NodeVisitor uses META_Object, which brings in:
virtual osg::Object* clone(const osg::CopyOp& copyop) const { return new name (*this,copyop); }

But CullVisitor defines clone as:
        virtual CullVisitor* clone() const { return new CullVisitor(*this); }

They differ in whether they take an argument.  I believe we've discussed this kind of thing before when considering using "using", and it was, if I remember correctly, a dodgy compiler warning then.

thanks
andy


From: osg-users [mailto:osg-users-bounces at lists.openscenegraph.org] On Behalf Of Robert Osfield
Sent: Monday, April 18, 2016 4:05 PM
To: OpenSceneGraph Users <osg-users at lists.openscenegraph.org>
Subject: Re: [osg-users] ShaderComposer::releaseGLObjects warning

On 18 April 2016 at 20:30, Robert Osfield <robert.osfield at gmail.com<mailto:robert.osfield at gmail.com>> wrote:
Hi Jannik and Ulrich,
You are right it's a bug, the missing const is an error, rather than the missing virtual.  I've fixed this and rewritten the implementation of ShaderCompose::releaseGLObjects() const.
I need to do a build and test before checking it in.

Now checked into OpenSceneGraph-3.4 and master.
I am bit concerned that this fix breaks the ABI of 3.4 though...  Would changing const affect the ABI?  I'm thinking yes, but it's too late in the evening for me to go and search for an answer...
Robert.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openscenegraph.org/pipermail/osg-users-openscenegraph.org/attachments/20160426/13e7b0fe/attachment-0003.htm>