Hi all,

I encountered a problem with Fast Downward in problems involving conditional 
effects and universal quantification, playing with translations from conformant 
to classical planning. 
The issue is a (IMO unjustified) computational time spent by FD in the 
translation/grounding when universal quantification is done over a large (but not 
huge) set of objects. I am attaching the instance (domain+problem) where I 
observed this behaviour. On my machine in fact, the preprocessing time takes 
about 50 secs, and the search just 0.5 seconds. The same instance is solved with 
FF in 1.66 seconds. 

There seems to be some kind of blow up in the representation, but (as observed by 
Malte) could also only be an issue related to the poor performance of python.

Anyway, it would be wonderful if this could be fixed, or at least put in the 
known caveats

Cheers 

Enrico Scala

Thanks for registering the issue, Enrico!

Unfortunately, it may take a while until we have time to attend to this because
we currently have a heavy-duty teaching period, followed immediately by the
ICAPS deadline. I'll try to looking into this after the ICAPS deadline. If you
see no further activity on this until December, please feel free to remind me at
any time. :-)