On 11/15/2011 12:20 PM, ais523 wrote: > On Tue, 2011-11-15 at 12:14 -0600, Pavitra wrote: >> On 11/15/2011 12:01 PM, Kerim Aydin wrote: >>> On Mon, 14 Nov 2011, Sean Hunt wrote: >>>>> This falls afoul of the precedent in CFJ 2737 (itself an extension of >>>>> the precedent in CFJ 1584), for the same reason that CFJ 3121 did. >>>>> >>>>> http://zenith.homelinux.net/cotc/viewcase.php?cfj=1584 >>>>> http://zenith.homelinux.net/cotc/viewcase.php?cfj=2737 >>>> >>>> I do not believe that the precedent established in CFJ 1584 applies >>>> here, and accordingly I find fault with the reasoning in CFJ 2737. In >>>> particular, I do not believe it's possible to read an additional >>>> requirement of finititude into the rules where none exists. Rule 217 >>>> is clear; the text of the rules takes precedence, and the text of the >>>> rules says that me cashing an infinite promise is equivalent to me >>>> cashing it; it specifies no termination condition. >>> >>> Except for voting, there is no text in any rule that explicitly allows for >>> any sort of conditional or multiplied action. It's *all* precedence and >>> meta-rules surrounding the administrative convenience. I hate to say it, >>> but your reasoning is a classic case of the fallacy of (etc. etc.). >> >> Promises are very much rules-supported. The rules don't *explicitly* >> allow infinite chains, but they're written in such a way that you have >> to infer an unwritten extra rule to *suppress* infinite chains. > > Why are people arguing about this, when my judgement was that the > infinite chain works, but messages containing infinitely many > conditional actions don't for much the same reason a message containing > a million conditional actions wouldn't?
I'm arguing that conditional actions are fundamentally different-in-kind to promised actions: the former are precedent-based, and the latter are rules-based.
