On Monday, July 30, 2012 1:31:53 PM UTC-7, Nils Bruin wrote: > While I agree that tab completion is very useful as a low-cost > documentation tool, I find it a little worrisome if our design > decisions come to be dominated by having tab completion generate > appropriate indices. In lots of cases it will go hand in hand with > writing clean code to begin with (messy namespaces are often a sign of > messy code), but if it forces us to go through strange contortions > (like a separate index namespace module as above) perhaps we have to > rethink our use of tab completion. >
Thanks, Nils, for the explication - that is very helpful. And thanks, Mike, for the IPython pointer. I agree that we shouldn't be making messes to drive tab-completion. But as always, we have design decisions from the past, which we might want to change with hindsight. In this case, most groups are now available via imports into the global namespace. They are easy to find, if you already know they are there and what their names are. ;-) As usual, I'm trying to help the beginner get started quickly and easily with a large collection of examples. The current approach I've been taking (and perhaps the one originally proposed in this thread) means a lot of (unnecessary) duplication. And I have written new/additional/consistent/modern documentation that would be better added to improve the original classes. While perhaps galling, it sounds to me like layering a "separate index namespace module" on top of the current disorganization will be a small price to pay for a big improvement in usability and will be much more maintainable down the road. Thanks again, Rob -- -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
