setting up ess-source-directory

Brett Presnell presnell at stat.ufl.edu
Thu Feb 3 01:15:40 CET 2000


This is exactly what I asked about a few weeks ago.  I think the
answer is that it's broken by design. The relevant variables interact
significantly with "customize" and yet the integration of ESS is
fragmented and incomplete.  It would be nice to see things go either
one way or the other, or for someone to hack together a temporary fix.
Unfortunately, with my somewhat limited elisp abilities, I don't have
the time to pursue the hack (I already spent more time than I had
trying to figure out if there was any way to make the current system
work).

The thing that really surprises me about this is that no one else
seems to complain about it (you're the first I know of besides me).
I'm sure that this problem has existed for a very long time (since the
first time I tried ESS several years ago), and I find that ESS is
practically useless to me mainly because of this one thing.  I still
use S-mode 4.8mmxe2 whenever possible, though I would love to have a
system that interacted seemlessly with both S and R.

I really have no place to comment on this, since I'm not a contributor
to the ESS effort, but I have always felt that ESS was a bit too
ambitious, or perhaps more to the point, that perhaps too much effort
is spent on supporting myriad systems without first getting any one of
them right.  For me personally, and I suspect a lot of other people,
ESS would be much more useful if just the S and R stuff all worked as
expected (at least reliably reproduce the functionality of S-mode
under GNU Emacs on UNIX systems) and were reasonably well documented
(reading the documentation leads one to believe that what you and I
were trying to do would work, as it does in S-mode).  Then perhaps a
working lispstat and/or SAS mode would be a nice second step (I've
never tried these, perhaps they work well).  Similar comments apply to
things like the integration with customize.  Getting halfway there
really just makes things worse.

As long as I'm offering my unsolicited opinion, I'll add that I
personally think that supporting GNU Emacs 19.28 is beyond the pale
(this was an issue in the discussion surrounding the integration with
customize), and I think we're getting to the point where 19.34 can be
ignored too.  I know that some people say that GNU Emacs 20 is still
seriously flawed, but how many people are likely to be simultaneously
using ESS and Emacs 19.28, which must be 8 years old by now?  They
need to upgrade and free the rest of us from their tyranny. ;-)

Anyway, I don't mean any of this as a flame.  Even though my
frustration shows in my comments, they are intended to be
constructive.  I am also aware that the main problem is no time to
spare on the part of anyone (that's certainly my problem at present),
and I don't mean to be disrespectful or ungrateful for the efforts of
Tony and others.  It's just that I find ESS in its present state is
something I avoid dealing with whenever I can, and I wish it were
otherwise.

-- 
Brett Presnell
Department of Statistics
University of Florida
(presnell at stat.ufl.edu)
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
ess-help mailing list -- To (un)subscribe, send
subscribe	or	unsubscribe
(in the "body", not the subject !)  To: ess-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._



More information about the ESS-help mailing list