Re: Engineering requests
- To: mathgroup at smc.vnet.net
- Subject: [mg125331] Re: Engineering requests
- From: danl at wolfram.com
- Date: Wed, 7 Mar 2012 05:36:03 -0500 (EST)
- Delivered-to: l-mathgroup@mail-archive0.wolfram.com
- References: <jj4qka$ib2$1@smc.vnet.net>
> As a side note, there were discussions back before Math 7 that the interval function > was going to be updated to work with these dependence issues. > However, it appears not to have happened yet as of ver 8. > I think it's a complex task. That would not begin to describe it. Indeed, I am not recalling any discussion to the effect that Interval would be changed along the lines you describe. It would be, in my opinion, a bad idea. The reason is that it would take reasonably fast functionality and make it quite slow, and would require techniques that are not necessarily reliable. That's not what interval arithmetic is meant to do. Daniel Lichtblau Wolfram Research