Re: Deprecate cut()?

[forkandwait]

> Ok, that's slightly different--if the intent of cut() is to be similar
> to R then that should be added to the documentation and the bug-fix
> should be verified to match what R does. Perhaps we should have an
> r_compatibility package on octave-forge.

Who knows what the original author (Kurt Honick, I think) intended.... I am sure
it wasn't to emulate R. 

> I personally feel like cut() doesn't "fit in" with the rest of the
> language very well and doesn't add anything of value over histc()--at
> least not enough to justify it's inclusion in the base. Easing porting
> of algorithms is a good reason to keep it around somewhere--but only
> if it actually matches the behavior with R and it should be labeled as
> such. There are a few IDL-ish functions I've wanted to add, too.

I would drop it out of base.  But perhaps we should also think of starting a
guide to Octave for R users and IDL users.  It would guide R users familiar with
table and xtabs to accumarray, for another example.

Just curious -- what IDL functions?