[ESS-bugs] ess-mode 5.3.7; ess-eval-visibly-p nil leaves artifacts in process buffer

[tyler]

On Mon, Apr 07, 2008 at 12:46:52PM +0200, Martin Maechler wrote:
> 
> Yes.  We (our group) have 
> ess-eval-visibly-p (sic!) set to nil ``since ever'' ( > 15 years).
> So I love it and am entirely used to using it.
> 
> 
>     TS> For some reason, this plays havoc with the command-line history.
>     TS> C-<up> and C-<down> a few times, and those pluses seem to propagate
>     TS> themselves, and it also interferes with sending commands from further
>     TS> up the transcript on occasion.
> 
> Well, if you are used to work with *.R files, you are not so
> much interested in the command-line history.
> But in the other cases (and some of "my" group users are
> different than me here as well),
> I recall that I haven't really suffered from any "havoc"
> 
> As I try now, I don't see real problems.
> Of course, the invisibly sent commands are not in the command
> line history,
> but I see no problems with the "+".
> 
> I use  M-p (and M-n) instead of your C-<up>  {they are faster
> once you have re-trained your fingers!}
> but these are absolutelly equivalent.

I have had one reply from another ESS user that the + symbols
propagate themselves when you use the command line after invisibly
sending a function to the R process. I continue to run into problems
with my usual workflow, which involves sending good code from an .R
file into the process, then switching to the process to work out some
ideas, which eventually get copied back into the .R file when I know
what I want to do and how to do it. I can't yet predict when the
"+ + >" symbols will interfere with sending commands, but I do
regularly run into variations of the following situation:

> if(!exists("baseenv", mode="function")) baseenv <- function() NULL
> options(STERM='iESS', editor='emacsclient')
> + + + + + + + + + + + + > 3 + 3
[1] 6
> 2 + 2
[1] 4
> + + + + + + + + + + + + > 2 + 2
[1] 4
> + + + + + + + + + + + + > 2 + 2
Error: unexpected '>' in "+ + + + + + + + + + + + >"
> 

Sometimes it comes up when I try and send a command from further up
the transcript, sometimes when passing a command from my .R file. I
have spent all the time I can afford on this this week, I'll have
another go at it next week, and at least try and identify exactly when
it does and does not happen.

The possible work-around I sent previously doesn't appear to work at
all consistently. Even the admittedly bad approach of sleeping for a
second following sending a command doesn't work - the function appears
never to sleep, even if I set the sleep-for to a very high number. So
I'm missing something fundamental somewhere. I've asked about this on the
gnu.emacs.help list, but failed to attract any attention. I'm getting
used to it.

Cheers,

Tyler

-- 
Support standardized open formats and control your own data -
	Reject Microsoft OOXML

http://noooxml.org