Re: [O] [patch] race condition in org-babel R

[Dan Davison]

Christophe Rhodes <address@hidden> writes:

> Hi,
>
> A substantial fraction of times when exporting my org-babel document
> (with many R session code blocks), I get "code block produced no
> value".  I think this is because of a race condition between
> waiting for the transfer.file to exist and actually populating it with
> output; the scenario is: 1. write.table() creates its output file
> 2. emacs notices its existence, and reads it 3. write.table() gets round
> to writing .Last.value to the file.
>
> The attached patch has made exporting much more reliable for me.

Thanks, I agree that the race condition seems plausible, and I've
applied a slight modification of your patch (the initial evaluation of
the "object" argument looks pointless, but was there to ensure that
errors in user code are not smothered by the subsequent exception
handling.)

Dan

>
>
>
> Best,
>
> Christophe