TestDUChain::testImportStructure() randomness

[Manuel Breugelmans]

On Thursday 05 June 2008 19:13:41 David Nolden wrote:
> On Thursday 05 June 2008 18:47:13 Manuel Breugelmans wrote:
> > I observed odd behaviour for this test command. It would pass when
> > running the full duchain test case but fail when run in isolation. Whats
> > more, always on the exact same cycle+sub-cycles. Not sure if this on
> > purpose (to make it a bit reproducable) but there's no call to srand in
> > there ... debian randomness ;)
>
> Yep I didn't call srand() so when something fails it is reproducible.
>

I see. It's probably useful to call it in init() or so anyway, with a fixed 
value. That way the above scenario (different behaviour when run alone as 
opposed to in batch) doesnt happen.

Do you know if rand is supposed to generate the same sequences across 
platforms/OS'es?


Manuel