On Thu, 23 May 2013, omd wrote: > On Thu, May 23, 2013 at 10:27 AM, Kerim Aydin <ke...@u.washington.edu> wrote: > > If a player can take back (by cashing) a promise e posts on the tree, it > > removes the whole purpose of the tree (if a pledge can be taken back, it's > > not a pledge). > > > > How about "in eir possession, or in the Tree's possession if e is not the > > author". > > Good point. I'm not sure this is needed in all cases, since the > author cashing it is not always a no-op (and in the recent past it > would often be possible to get around this by having a golem cash it), > but it doesn't hurt.
I think whichever we use as the default, the author can set up the conditions the other way (e.g. the conditions could say "and the author has not announced the Expiry of this promise", or, for the opposite direction "and the cashier is not the author"). The default assumption that fits the common definition of "promise" (and the "pledges" the tree was meant to replace) would be "no take backs". You're right we never plugged the golem hole... never came up... Thinking of defaults, I was also wondering about default expirations, (as these promises tend to lose relevance over time). -G.
