Re: More /.{I->-1} craziness
- To: mathgroup at smc.vnet.net
- Subject: [mg106420] Re: More /.{I->-1} craziness
- From: Bill Rowe <readnews at sbcglobal.net>
- Date: Tue, 12 Jan 2010 04:47:55 -0500 (EST)
On 1/11/10 at 6:54 PM, btreat1 at austin.rr.com (DrMajorBob) wrote: >WRI has blithely broken user code in the past, so Bill's argument >that they shouldn't in THIS case rings hollow. My point wasn't that WRI should not do things that break existing code. But rather, they should not do this without good reason. For example, the change in the way graphics works between version 5 and 6 certainly caused broke some existing code. But I believe the enhanced capabilities for graphics provided by version 6 more than justified the effect on existing code. OTOH making a change simply so that some subset of new users is less confused and provides no other benefit would not be sufficient reason for WRI to break existing code. On the contrary, I would argue the need not to break existing code out weighs the need to make Mathematica more intuitive for some subset of users. But, as you point out WRI has broken code in the past and will certainly do so in the future. In order to grow Mathematica with more functionality, it is inevitable some changes will break existing code. Additionally, my opinion as to what justifies breaking existing code is just that, my opinion. And that has essentially no impact on what WRI does in the future.