On Tue, 2008-03-11 at 13:42 +0000, Peter Clifton wrote:
> On Tue, 2008-03-11 at 13:34 +0000, Peter Clifton wrote:
> > On Mon, 2008-03-10 at 23:57 -0700, Matt Ettus wrote:
> > > When I load the attached schematic, select the large group of
> > > components on the right, and then hit delete, gschem crashes without
> > > any messages.  I am using version 1.4.0.20080127, from fedora 8 rpms.
> > > 
> > > BTW, this version seems a lot slower when panning.
> > 
> > I've not got reproducing down to a simple set of operations, but I did
> > manage to get this...

Ok, a small test-case. Valgrind is very unhappy when you delete the SMC
connector.

Turns out (I think) that its because the SMC has four pins atop each
other, and the connection tracking code "knows" they are connected to
each other.

We're redrawing cues on "other" objects, which in the normal case
doesn't include other pins attached to the symbol. I've not dug deep
into this, but you'll hopefully see that the second testcase does not
upset gschem. (I separated the four pins from each other).

I'm not sure what the best fix is here, any thoughts?

An obvious workaround is to have a 2-pin SMC symbol, and name all the
grounds the same in the PCB footprint. Matt, I'm not going to "impose"
that on you if you prefer the "5"-pin variant, we still need to fix the
underlying bug.

Matt, is there any technical reason why the PCB footprint can't have its
four ground terminals named the same (I ask from a curiosity / stylistic
point of view).

-- 
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!)
(component-library ".")

Attachment: SMA-5.sym
Description: application/geda-symbol

Attachment: testcase.sch
Description: application/geda-schematic

Attachment: testcase2.sch
Description: application/geda-schematic

Attachment: testsym.sym
Description: application/geda-symbol


_______________________________________________
geda-dev mailing list
[email protected]
http://www.seul.org/cgi-bin/mailman/listinfo/geda-dev

Reply via email to