Hi all,

sorry to reopen this thread,  but this does not work for me (GNU Emacs
23.2.1, Org almost HEAD c7700d7bbee4f7596feb199b1ec1bc7750d4fb48)

For me the image in the attached example is always created during
(LaTeX)export.  Even if I do the manual evaluation once (and the hash is
created).

The manual re-evaluation respects the :cache yes and does not evaluate
again.

What am I missing?

- Andreas

================================

* Test
  Some text.
  #+srcname: some_code
  #+begin_src R :session :exports code :eval never
    mean <- 0
  #+end_src

  Some text.
  #+srcname: plot_code
  #+begin_src R :session :exports code :eval never
    plot(x=rnorm(1000000, mean=mean, sd=sd),
         y=rnorm(1000000, mean=mean, sd=sd))
  #+end_src


* Plots
  #+LABEL:      fig:test
  #+ATTR_LaTeX: height=.29\textheight
  #+srcname: plot_1
  #+begin_src R :session :noweb yes :file test.pdf  :exports results
:results graphics :width 8 :height 4 :cache yes
    sd <- 4
    <<some_code>>
   
    <<plot_code>>
  #+end_src

  #+results[17681e8e450f1f4e1d8f55fa57f9be1b24dae4db]: plot_1
  [[file:test.pdf]]
================================


Am 03.02.2011 15:34, schrieb Eric Schulte:
> Dan Davison <dandavison7@gmail.com> writes:
>
>   
>> Loris Bennett <loris.bennett@fu-berlin.de> writes:
>>
>>     
>>> Erik Iverson <eriki@ccbr.umn.edu> writes:
>>>
>>>       
>>>> Loris Bennett wrote:
>>>>         
>>>>> Hi,
>>>>>
>>>>> I have an org file containing several src blocks which generate images
>>>>> using ditaa. When I publish to PDF via LaTeX, the images are all
>>>>> generated every time, which makes publishing rather slow.
>>>>>
>>>>> Is there some way to toggle the evaluation of the src blocks on and off
>>>>> when the file is published?
>>>>>
>>>>>           
>>>> You could try the :cache header argument, http://orgmode.org/org.html#cache
>>>>
>>>>         
>>> Ah, thanks. There is a slight gotcha here, though.
>>>
>>> I added :cache yes to the source headers and exported again, but nothing
>>> changed; all the images were generated again. Also, no SHA1 hash was
>>> added to the +results header.
>>>
>>> After some fruitless fiddling I was about to write to the list again and
>>> moan, when I did a slightly random C-c C-c in the begin_src line and,
>>> hey presto, the hash was added to the results header. I then did this
>>> for all the images and found that the image were no longer regenerated
>>> on export, as advertised.
>>>       
>> Hi Loris,
>>
>> Yes. It does seem that it would be nice if in this situation, the first
>> export added the SHA1s, and subsequent exports recognized that
>> evaluation wasn't required. I think the reason this does not happen is
>> that behind-the-scenes Org makes a copy of the buffer for export
>> preprocessing (including src block evaluation). But Eric S is the expert
>> -- he may have more to say here.
>>
>>     
> Yes, this is exactly the case.  Org-mode is very careful that the
> process of exporting does not make any permanent changes to the original
> org-mode file.  I agree this should be mentioned in the :cache
> documentation.
>
> Best -- Eric
>
> _______________________________________________
> Emacs-orgmode mailing list
> Please use `Reply All' to send replies to the list.
> Emacs-orgmode@gnu.org
> http://lists.gnu.org/mailman/listinfo/emacs-orgmode
>