Talk:Undefined values

From Rosetta Code

After discussion on IRC, it was noted that "undefined" as this task is not significantly distinct from NULL. However, I prefer this task's "identify and exercise the mechanisms" approach, as opposed to Null's more specific "check" approach. I'm certainly open to others' thoughts on how to effect improvement of the NULL task, make this task sufficiently distinct, or possibly deprecate that task. --Michael Mol 03:04, 28 November 2009 (UTC)

Tangentially, if this task is to be distinct from Null by way of being "talk about this" rather than "do this specific thing" (which is IMO an excellent split to have; cf. Eval and Eval in environment, created by me), then they should use the same terminology. —Kevin Reid 03:24, 28 November 2009 (UTC)
It is distinct from Null, at least in the context of Ada and other [strongly typed] languages where null is a legal defined value, as opposed to an illegal undefined value. I presume that some other solutions are wrong in this sense, i.e. when defined = any value from the type's domain set. In a properly typed language an object may bot have an undefined value unless something was badly wrong. --Dmitry-kazakov 15:53, 15 July 2010 (UTC)