[ESS-bugs] ess-mode 5.3.5; M-x R messes up RET in new shells

[Tim Hesterberg]

It did leak.  The variable isn't buffer-local, 
so (setq) changes the value globally.

It also doesn't seem necessary.  Both my R buffer and .R files
seem to work fine using the default value for default-process-coding-system
of (undecided-dos . undecided-unix).

Perhaps default-process-coding-system was buffer-local or had a different
default when you did your work?  I'm using 22.1.1

Tim

>That line you point in essd-r.el out is absolutely necessary for R
>in an ESS buffer.  Unix coding does not work with Rterm.  I remember
>that when I put it there, my intent was to limit it to the *R* buffer.
>Did it leak?
>
>Rich