Crash test update

[Jan Holesovsky]

Hi Eike,

Eike Rathke píše v Po 13. 04. 2015 v 17:28 +0200:
 
> > So we have a root cause there now I guess.
> 
> Whatever the root cause in this case may be, SetZoom() may be indirectly
> called from the API and a 0 value or very high values are possible and
> passed down, so removing some "funny numbers" sanity check there should
> at least had introduced proper sanity checks at other places.

Sure - so these documents at least allow me to see what these other
places are :-)

> My other assumption is that when not having actual view data (e.g.
> because there is no view) then setting an initial zoom at ScViewData
> might fail with a 0 value or something. Additionally, imported documents
> may come with arbitrary view settings. Just wild guesses.

0 for zoom is definitely the worst default value that might be there; 1
is what I'd expect - I'll have a look what these documents have to say &
act accordingly - either find what is wrong there in the first place, or
revert.

All the best,
Kendy