[ESS] SAS. Is Ess 12.09 working?

Paul Johnson pauljohn32 at gmail.com
Sun Nov 18 08:44:09 CET 2012


I'm not much of a SAS user anymore, rarely use Emacs for that. On
Friday I got some student questions about it and I can't make it work
right, and I started to suspect I've run into one of the bugs you were
discussing in this group last week.

I'm have Emacs  23.4-ESS 12.09 on Debian wheezy. I want to use SAS
both in batch and interactive modes. If these are related of the known
problems you were discussing last week about associations between
buffers and sessions, then I will feel relieved. Otherwise, I'm just
plain frustrated.

Here's where I stand.

In emacs.d/init.el, my only SAS change is:

(setq ess-sas-local-unix-keys t)

That's what you do if you want the SAS keys, but only for SAS, not
other programs. Yes?

My test file is /tmp/flop.sas/flop.sas, I cd into /tmp/flop.sas and open.

Emacs SAS notices it is a SAS file

First try the batch. Hit F3 button. The *shell* buffer says:

pauljohn at pjlap-124:flop.sas$
cd /tmp/flop.sas/
nohup nice sas *shell* -rsasuser &
pauljohn at pjlap-124:flop.sas$ pauljohn at pjlap-124:flop.sas$ [1] 28364
pauljohn at pjlap-124:flop.sas$ nohup: ignoring input and appending
output to `nohup.out'

Nothing happens. The content of nohup.out:

ERROR: Invalid file, /tmp/flop.sas/*shell*.sas.

Humphf. Well, the name of the SAS file is seen as *shell*.


I can get an iSAS session started.

M-x SAS causes the Emacs frame to divide into 4 very narrow windows.
However, I can't get commands to go from the SAS code window to the
others.

Highlight some SAS code.

In the ESS menu, I choose "Eval and Go" and nothing happens, the
minibuffer says only:

Text is read-only

I banged on this quite a while, tried quite a few hints from 2004 -
2007 in this list's archives.

I've verified I can run sas in the command line of an ordinary terminal,

$ cd /tmp/flop.sas
$ sas flop.sas

that generates the lst and log files I'm looking for.

If you tell me that ESS 12.09 is working for SAS, then I suppose my
next step is to go to a completely clean system and see if it works
for me. There's always a chance I've broken something with endless
tinkering.

pj

-- 
Paul E. Johnson
Professor, Political Science      Assoc. Director
1541 Lilac Lane, Room 504      Center for Research Methods
University of Kansas                 University of Kansas
http://pj.freefaculty.org               http://quant.ku.edu



More information about the ESS-help mailing list