It does happen anyway. I had that in to get around this issue a while ago (long story, but it is why we still run 8.2.10), and I still get the issue without it. 

I think I have figured it out. The order the packages are installed is important. The code below illustrates the issue for me. If I install org-plus-contrib first, the issue goes away. Note the first thing the pydoc package does is:

(require 'org), which I guess interferes with subsequent building of org. (I am using GNU Emacs 25.1.50.1).


Also, you need to have (R . t) in there. that specifically triggers the error. 


(setq user-emacs-directory "./sandbox/")

(require 'package)

(setq package-archives
      '(("org" . "http://orgmode.org/elpa/")
("gnu" . "http://elpa.gnu.org/packages/")
("melpa" . "http://melpa.org/packages/")))

(package-initialize)
(package-refresh-contents)


(package-install 'pydoc)
(package-install 'org-plus-contrib)


(org-babel-do-load-languages
 'org-babel-load-languages
 '((emacs-lisp . t)   
   (R . t)
   ))


(message "done")


John

-----------------------------------
Professor John Kitchin 
Doherty Hall A207F
Department of Chemical Engineering
Carnegie Mellon University
Pittsburgh, PA 15213
412-268-7803

On Sun, Feb 14, 2016 at 11:36 AM, Achim Gratz <Stromeko@nexgo.de> wrote:
Aaron Ecay writes:
> This looks like the kind of problem that results when org is byte-compiled
> in an emacs with a different version of org loaded.  This used to be a
> problem when installing via package.el (and not just for org), but I think
> there has been some work put into making sure it doesn’t happen any more.
> (I know emacs developers were discussing them and then they seemed to go
> away so I assumed they were fixed...it’s possible I was mistaken though.)

No, this is bug #10125 and the others that it was merged with (most
recently #21267, with the exact same error message as John got).

> If I had to guess, I’d say that line 63 of packages.el in your jmax
> setup defeats these fixes.  Try commenting it out.  It should not be
> necessary to manually byte-compile packages: package.el should do that
> for you.  And in this case, I hypothesize that it is actually harmful.

Yes it is.  Although I'd venture to guess that it would have happened
anyway.


Regards,
Achim.
--
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+

Wavetables for the Waldorf Blofeld:
http://Synth.Stromeko.net/Downloads.html#BlofeldUserWavetables