[Octave-bug-tracker] [bug #45528] mapping: missing function roundn (implementation supplied)

[Philip Nienhuis]

Follow-up Comment #3, bug #45528 (project octave):

The mapping package release got a bit delayed, so maybe....

But even then I'm not sure if the mapping package is the right place for a
fairly general function like roundn that has no direct relation with mapping
other than that by some opaque and probably irreproducable process it ended up
in a mapping toolbox.
I'm not against it, not at all, but we can do better than Matlab with respect
to categorizing and assigning functions to packages. Maybe the miscellaneous
package is a better home.
I'll ask our octave-forge guardian for his opinion (cc'd).

As to your second remark:
AFAICS it is especially Matlab that keeps old functions around, maybe related
to that motive you gave. E.g., a while ago in old but still valid Matlab code
I stumbled over "setstring" that turned out to be an ancient "char"
predecessor. Octave doesn't even have it.  At my work we have lots of 10-15
years old Matlab scripts that still run in 2015b containing more of such
ancient functions.
Looks like TMW changed their policy - these days on their site I often see
that function descriptions that used to show a warning "will be removed in a
future Matlab release" now rather carry "not recommended".


    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/bugs/?45528>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.gnu.org/