It seems to me that, in my current development environment, i have one primative for two seperate tasks tasks 1. naming some functionality to abstract its meaning 2. sharing that functionality. There are times when i need the second but not the first. e.g I need to pass a subsection of a vector to two different "places". However, naming a subsection of a vector doesn't really offer a good abstraction because the functions over the vector are as high level as possible. Has anyone tackled this problem directly? It seems like a UI issue, i want my editor to show the functions, but if i change one the other should udpate. with a UI interaction to signify im changing it in more then one place. It would probably be wise to assign a name as well, given that not everyone will be using this UI.

How can you talk about something without some symbolic representation though? Even if it’s autogenerated, don’t you need a name to signify the thing?

Yes, i'm almost positive the textual representation would need a shared symbol. What i'm considering is that could visualize that symbol in different ways. I might want to view the code instead of the symbol. I do a version of reverse all the time at my editor what i eval an expression to see the data directly.

I think I’ve heard someone suggest semantic rather than syntactic code highlighting (it was in that talk “Monads and Gonads”). If you could see based on color where a symbol was defined, would that help?

Yea, Those are both ways of dealing with the issue. Though, those solutions are external to the program. In part, this feels like wanting my program to contain information the editor can use. i can see some rather obvious draw backs to going down that route though.