[ESS] ess-remote trouble

[Dan Davison]

"RICHARD M. HEIBERGER" <rmh at temple.edu> writes:

> I haven't looked at this for a long time.
> My memory and the documentation say that M-x ess-remote should
> do all the buffer naming magic.
>
> What happens if you change the
> name of the buffer from *shell* to *R2*?  Does it work?

Hi Rich,

Yes, using rename-buffer on an M-x shell buffer does make
ess-switch-to-end-of-ESS behaviour consistent (splits frame) (although
ess-switch-process still refers to the buffer as "shell" even if
rename-buffer is done before R, which wasn't obvious to me).

But no, that doesn't solve the problems with an inactive prompt when
using ess-remote with M-x shell. However, I see that the ESS
documentation suggests using M-x ssh, not M-x shell, and this seems to
work fine. So perhaps I'm reporting problems with something that's not
encouraged usage.

Also, it was suggested to me off-list that an alternative is to open a
file on the remote machine using tramp (C-x C-f /user at host:file), and
then to issue M-x R in which case R executes remotely, and ess-remote is
thus bypassed. Both the M-x ssh and tramp methods seems to be working
well for me so far. I see myself using ESS with remote R processes a
lot.

Dan

>
> Rich