(assuming this was meant for Discussion and not me personally...)
---------- Forwarded message ---------- Date: Wed, 7 Feb 2018 10:35:48 +1100 From: Madeline <j...@iinet.net.au> To: Kerim Aydin <ke...@u.washington.edu> Subject: Re: DIS: About Paradoxes If a value becomes indeterminate, that in itself clarifies the situation - and nowhere does it say that PARADOXICAL is appropriate only because of something going indeterminate, it's only if it's undecidable. I don't think one suggests the other (otherwise they'd use the same language in the first place), and I'd suggest that even if it wasn't irrelevant, DISMISS or even FALSE would be more appropriate. On 2018-02-07 10:11, Kerim Aydin wrote: > On Wed, 7 Feb 2018, Madeline wrote: > > From the Ruleset: > > > > Rule 2518/0 (Power=3.0) > > Determinacy > > > > If a value CANNOT be reasonably determined (without circularity or > > paradox) from information reasonably available, or if it > > alternates indefinitely between values, then the value is > > considered to be indeterminate, otherwise it is determinate. > > > > That's Power 3, which overrides all the nonsense about winning by paradox > > ANYWAY by cleanly resolving any paradoxes that do happen to occur. > > Can we just repeal winning by paradox? It's literally ONLY there so people > > can > > scam it, it doesn't serve any real purpose. > This of it this way. R2518 does some error-trapping by setting > undefined values to INDETERMINATE. But you still have to actually > handle the error. In many cases, we've done this - for example > indeterminate switches take on their default value. But if you > find somewhere where it's not handled after being trapped - it's > good to offer an incentive to find it for a win instead of waiting > until it breaks something unexpectedly. > > That said, (1) I usually find 95% of paradox attempts pretty tiring/ > pointless, and (2) paradox wins don't mix well with contracts, > where anyone can give arbitrary text some regulatory force - that > combo knocks the 95% up to about 99%. :P
