[Xastir-dev] Blame Canada for xastir segfaults

[Curt, WE7U]

On Tue, 23 Mar 2004, Owen DeLong wrote:

> I may be naive here, but, doesn't that mean that we should improve the
> error reporting in Xastir so that it detects this problem and informs
> the user instead of segfaulting?

You're not naive.  Well, not in this case anyway... hi hi

Problem here is that Shapelib LOVES to segfault all on it's own.
It's not Xastir that's doing the segfaulting here.


> Again, I'm not an expert, but, I was always taught that a segfault indicated
> a bug in the software.  Usually a failure to adequately check some form of
> input before depending on it conforming to assertions.

Correct.  In this case Shapelib.


> If we're using a library to parse these files and that library is what is
> actually segfaulting, then, I think we have a few different options
> (which aren't necessarily mutually exclusive):
>
> 	1.	File a bug with the library development team.
> 	2.	Fix the library and submit appropriate patches to the
> 		development team (yeah, I know, resource issue, but,
> 		it's an option).
> 	3.	Trap the SIGSEGV from the library and produce a more
> 		helpful error message.

Shapelib is another open project, so yes, it'd be good to fix up
Shapelib (and other libraries that we use from the same site).

Frank Warmerdam is the guy to contact, and most of the projects
there have Bugzilla bug-tracking I believe.

Anyone want to take this on?

--
Curt, WE7U			    archer at eskimo dot com
Arlington, WA, USA		http://www.eskimo.com/~archer
"Lotto:    A tax on people who are bad at math." -- unknown
"Windows:  Microsoft's tax on computer illiterates." -- WE7U
"The world DOES revolve around me:  I picked the coordinate system!"