gEDA-dev: gEDA-announce: gEDA/gaf stable version 1.2.1-20071231 released!

[Peter Clifton]

On Wed, 2008-01-02 at 08:42 +1100, Hamish Moffatt wrote:

> > This should "just work" if it is called libgeda31. I just checked
> > packages.debianorg. The 1.2.0 released gEDA packages depend on libgeda31
> > (>= 1:1.2.0) which will be satisfied by this newer version of libgeda31.
> > If installing this new libgeda31 broke any of the 1.2.0 versions of the
> > apps, then yes, we would have a problem. In this case, I don't think we
> > do. geda-gschem-1.2.1 will just need to depend on libgeda31 (>= 1:1.2.1)
> 
> Your reasoning is correct. However, installing new libgeda31 (1.2.1)
> does break the installed 1.2.0 applications. gschem segfaults on
> startup for me. 
> 
> Now I'm a little stuck. I have started work on the libgeda31.1 rename
> but not finished all of it yet. I have not yet tried Peter Brett's patch.
> Ales says that this version is not expected to be compatible with 1.2.0
> anyway (despite the major version number indicating otherwise to the
> dynamic linker). I would appreciate some guidance on this issue :-)

Ok, I'll look into it. You're finding gschem 1.2.0 seg-faults with
libgeda 1.2.1. Apologies for the misunderstanding, I hadn't read the
original post properly.

I'll see what I can discover. I'd rather not see the package versions
upset in Debian unless we have to. I don't think Peter's patch is such a
good idea as is, since this will re-introduce a crash bug (albeit an
obscure one). If we had to apply something like this, there would be a
corresponding fix moved into gschem too.

Could you send me the sources for the 1.2.1 packages you're building to
save me having to re-work them when testing this out?

I wonder if we haven't in-fact brought on a crash elsewhere in the
program due to file-version checks. libgeda and gschem know which
versions they come from, and it is possible we're looking at a segfault
unrelated to ABI changes. Could you post or email me a gdb backtrace, or
strace output.

Best wishes,

-- 
Peter Clifton

Electrical Engineering Division,
Engineering Department,
University of Cambridge,
9, JJ Thomson Avenue,
Cambridge
CB3 0FA

Tel: +44 (0)7729 980173 - (No signal in the lab!)