[mythtv-users] Undefined symbol error when running CVS

[Jarod Wilson]

On Thursday 11 November 2004 09:06, Shawn wrote:
> > > Likely a library it's linking to has been compiled with a compiler not
> > > ABI compatible with what myth was compiled with (2.96 v 3.X?), or some
> > > psycho-babble sounding like that. (The "_ZNBlahHorkBeeble" bullshit is
> > > a clue)
> >
> > No, its not likely. The last Red Hat distribution with a 2.96 compiler
> > was Red Hat Linux 7.3, over three years and six releases ago. He said
> > he's running Fedora Core 2, using packages from ATrpms, compiled on a
> > Fedora Core 2 system. (Well, in a FC2 chroot using FC2's native tool
> > chain, anyhow).
>
> While I may not be correct in the specific sense, I am correct in the
> general sense. That link error crap with the underscore garbage is
> binary incompatible shit lying around. Old RPMs, whatever. If it were
> anything else it would have more traditional non-underscore errors.

I believe "That link error crap" as you so eloquently put it, is CVS myth 
being built against 0.16 Myth libraries. It has nothing to do with the 
compiler. You're closer this time around with the (again eloquent) "binary 
incompatible shit lying around", since it is a library from 0.16 that isn't 
compatible with the latest CVS.

> > > Try looking at all the libraries myth links to (ldd), and recompiling
> > > those libraries with the same gcc (or g++ as it were).
> >
> > Don't bother. Way off base, like Isaac said.
>
> Way off base? Geez, people really like to try and act superior around
> here.

Because I know they guy's system a helluva lot better than you do. He put it 
together following a document *I* wrote, thank you. And Isaac most definitely 
has a right to act superior, given the fact he's the lead developer on the 
project.

> I've fought with c++ compiles for 9 years as an admin and a user 
> on every popular linux distribution since 95.

Half the people on this list have done the same.

> I think I'm right, and I 
> may even be wrong, but way off base?

I say way off, because you don't understand the guy's system. Your initial 
supposition was that it was because he had something compiled with a 
different compiler on his system. You were closer with your follow-up.

> No, it would have led to the correct answer, showing runtime links to
> libraries he did not intend to link to, like the RPMs that ought not be
> there...

I suppose he might have got there eventually.

> > # rpm -qa | grep myth
> >
> > You probably still have a bunch of rpms installed, apt-get remove mythtv
> > only removes the mythtv package, there are several packages that have to
> > be removed (namely, all the ones in the output from the above command).
>
> Different way of coming to the same conclusion.

A much more direct way, and much easier for non-dev types to understand.

> Also, I suggest recompiling your CVS after removing any old RPMs, as it
> may have compiled against older headers.

Definitely.

I'm going to preemptively remove myself from this thread now.

-- 
Jarod C. Wilson, RHCE
jcw at wilsonet.com

Got a question? Read this first...
     http://catb.org/~esr/faqs/smart-questions.html
MythTV, Fedora Core & ATrpms documentation:
     http://wilsonet.com/mythtv/
MythTV Searchable Mailing List Archive
     http://www.gossamer-threads.com/lists/mythtv/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://mythtv.org/pipermail/mythtv-users/attachments/20041111/eb6b5119/attachment.pgp