[Libreoffice] gravity of raised assertions?

[Terrence Enger]

My cross-posting to dev and qa lists is deliberate; if you
find that it is the wrong thing to do, I shall have to
apologize.

On the dev list,
On Mon, 2011-10-31 at 10:31 -0400, Terrence Enger wrote:
> There has been a lot of discussion in the past about raised
> assertions and how seriously they should be treated.  Is
> there a current concensus?
> 
> I raise the question again for no better reason than that
> even a newbie like me can see a raised assertion and collect
> a backtrace.  Guidance welcome.
> 

And on the dev list, 
On Tue, 2011-11-01 at 10:12 -0700, julien2412 wrote:
> I know that it comes from Apache but what about
> http://logging.apache.org/log4cxx/ ?
> log4j is quite used in Java world, I don't know the state of log4cxx
but it
> could help.

Oh my, what a lot of different dimensions there are to this
question!  I was hoping that a consensus had formed since
the last time I asked the question, but that hope seems to
be frustrated.


Thoughts arising ...

(*) We are *far* away from a code contribution.  (Sorry,
    Stephan.  I had a couple of raised assertions "in hand"
    when I asked, and I thought I might be able to clarify
    the conditions leading to one of them.  After your
    answer Monday, I was not able to make them happen
    again under gdb.  Sigh.)  So, it would be good to move
    the discussion elsewhere.  Perhaps ...

     -  A wiki page.  Is there an obvious place to put one?
        But considering how far we are moving from the
        existing code, perhaps a personal page would be
        better.  I *think* I am allowed to create one.

     -  The qa list.  I assume, subject to correction, that
        assertions are mostly for the benefit of quality
        assurance.

(*) There is an ongoing discussion

<http://lists.freedesktop.org/archives/libreoffice-qa/2011-October/000240.html>
    on the qa lilst about automated testing of LibreOffice.
    (Thirty-nine messages so far.  Again, I thought I was
    asking a small, modest question.)  Making either
    detection or diagnosis of problems either faster or more
    effective would justify a lot of effort.  Is there any
    place outside of automated testing that would make the
    effort worthwhile?


Thank you for your attention,
Terry.