Stefan Monnier <monn...@iro.umontreal.ca> writes: >> And against the additional variable to make this more >> backward-compatible? > > Yup. The var would be my second-best choice (and I assume it's > immediately declared obsolete).
I tend to agree with Stefan M here.