Karl Lehenbauer <karl at flightaware.com> writes:
> Thanks for the reply, Filip.  I don't get that it's a bug either,
> insofar as I can't even speculate on the amount of work to "fix" it,
> but it is a gotcha and could be a good thing to document in the
> FAQ.  In any case, hopefully someone will find even this email
> exchange by googling and will save themselves some time when it
> happens to them.

It doesn't strike me as a terribly resolvable matter; yes, indeed, you
may readily capture, with lastval(), a value from one of the
Slony-I-internal sequences.

If you know which sequence whose value you want, I'd imagine you'd
want to "name that sequence."  To get "the last sequence updated"
seems *way* too nondeterministic for comfort!  If I have several
sequences in play, it could get pretty unpredictable which one I'd get
a value for, and the notion of grabbing whichever was the "last
updated" one is the sort of thing I'd want to smack our developers for
if they tried it ;-).

The very thought of using lastval() leaves me mighty queasy; lastval()
may be expected to be affected by *any* trigger that might get fired
along the way, and Slony-I's not the only thing one might introduce
like that.

I'll certainly see about noting this issue in the FAQ.  I'm not sure
it's quite "frequently asked," but there you go ;-).
-- 
(reverse (concatenate 'string "moc.enworbbc" "@" "enworbbc"))
http://www3.sympatico.ca/cbbrowne/languages.html
If a hole in the street is a manhole, is a hole in a man a streethole?