On Thu, Nov 21, 2013 at 3:49 PM, Martin Thomson <martin.thomson@gmail.com>wrote: > On 21 November 2013 12:34, Roberto Peon <grmocg@gmail.com> wrote: > > And to be clear, I mean for #1 to happen only when the semantic-layer > cares > > about ordering. > > If it doesn't (e.g. potentially cookie), then it doesn't collapse the > values > > into a null-delimited-list-value > > This is where it gets interesting. That exact same argument can be > made for #3, and that doesn't require new semantics. > I think the new semantics are pretty simple.. and given that we've seen it br reliable I'm not strongly feeling like there is motivation to go another direction. But maybe I'm just overcome by pragmatism today :)
Received on Thursday, 21 November 2013 21:05:46 UTC