Hi Marc.  If you can, please direct inquiries like these to the
ECJ-INTEREST mailing list; they'd be of some interest there I think.

The steady-state facility underwent *extreme* revision from 12 to 13, part
of a long-overdue cleanup to make possible a lot of the new stuff that 13
has in it.  #1 on the list: pseudogenerations are gone in the sense you're
used to seeing them.  See the revised SteadyStateStatisticsForm interface
and the SteadyStateEvolutionState documentation: you probably want
postEvaluationStatistics(EvolutionState) instead.

It's also possible that v13 has bugs with regard to steady-state as it
underwent such significant revision.  But it sounds like the problem
is in porting instead.  Let me know if you need further help.


On Fri, 21 Oct 2005, Marc Richards wrote:

> Hi ECJ team,
> I have some steady-state statistics classes that I'm pretty sure used to
> work (maybe with version 12) but they aren't outputting anything with
> version 13. The problem seems to be that my overridden
> _nextPseudoGenerationStatistics() method is never being called. I know you
> guys did away with pseudogenerations in there another method I
> should be overriding instead now? As far as I can tell,
> SteadyStateStatisticsForm.nextPseudogenerationStatistics is never invoked
> by SteadyStateEvolutionState (and it was in v12), which seems like it
> could be the problem. I tried to hack up the source and call it myself,
> but that just lead to a bunch of NullPointerExceptions that I couldn't
> figure out :)
> I actually upgraded to v13 a while ago and this is the first time I've
> noticed this problem, but I'm not certain if I did any steady state work
> since I switched versions. I did a quick check on the mailing list
> archives and noticed that Sean had sent a CVS snapshot to someone else
> that was working with the steady state stuff. Does this sound like a
> problem that could be fixed with the newest build?
> Thanks a bunch,
> Marc Richards
> Colorado State University
> Dept. of Computer Science