Problem with ordered factors in 2/3 variable summaries

[rdstern]

I'm really liking the new 2/3 variable summarise dialog.
But I stmbled over a problem in the dialog while working on the help over christmas. I think it has been there for ever, and we just didn't notic it. It might be a bit more general.

You can check with our usual diamonds data. Then either 2 variable or 3 variable. Then try to put the 3 factors into the multiple receiver. Ordinary factors are fine, but it doesn't like more than 1 ordered factor in the multiple receiver. And if you try an ordered factor into the single receivers it thinks they are numeric. Ideally it just treats ordered factors as factors.