[Bioc-devel] Tips for using R-2.15 with BioC 2.10 and 2.11

[Nicolas Delhomme]

Makes sense, thanks,

Nico

---------------------------------------------------------------
Nicolas Delhomme

Genome Biology Computational Support

European Molecular Biology Laboratory

Tel: +49 6221 387 8310
Email: nicolas.delhomme at embl.de
Meyerhofstrasse 1 - Postfach 10.2209
69102 Heidelberg, Germany
---------------------------------------------------------------





On 17 Apr 2012, at 18:52, Dan Tenenbaum wrote:

> On Tue, Apr 17, 2012 at 9:46 AM, Nicolas Delhomme <delhomme at embl.de> wrote:
>> Hi Dan,
>> 
>> I, indeed, have a different setup, therefore the script. I just pasted it unchanged.
>> 
>> Now I'm curious about what your recommendations are for the current development version (2.11). I have read your previous email about this topic, but I still need some clarifications. Is it important wether we use R-2.16 or R-2.15 for developing packages for Bioc 2.11? Or do we have to use R-2.15 for Bioc 2.11 and switch to R-2.16 for Bioc 2.12 in six months?
> 
> We recommend using R-2.15 for developing packages for BioC 2.11.
> After BioC 2.11 is released, you can switch to using R 2.16 (R-devel)
> for developing BioC 2.12 packages.
> 
> We don't recommend using R-2.16 before then because it could introduce
> changes that will not be present in R-2.15, which is the R version
> that BioC 2.11 is intended to run with, and is what our build machines
> are running (for that matter, R-2.15 is also the R version that BioC
> 2.10 is intended to run with).
> 
> Thanks,
> Dan
> 
> 
> 
>> 
>> Cheers,
>> 
>> Nico
>> 
>> On 17 Apr 2012, at 18:36, Dan Tenenbaum wrote:
>> 
>>> On Tue, Apr 17, 2012 at 6:00 AM, Robert Castelo <robert.castelo at upf.edu> wrote:
>>>> Dear Dan, and specially those using mac osx for developing packages,
>>>> 
>>>> i've tried to follow the instructions below for Mac OSX, concretely
>>>> 
>>>> On 04/03/2012 01:04 AM, Dan Tenenbaum wrote:
>>>>> 
>>>>> Hello BioC Developers,
>>>>> 
>>>> [...]
>>>> 
>>>> 
>>>>> On Linux and Mac OS X, You can create bash aliases (called e.g. Rdevel
>>>>> and Rrelease) that will invoke the command line above. (Note that the
>>>>> "devel" in "Rdevel" refers not to the development version of R, but to
>>>>> that of Bioconductor; slightly confusing). Put the following in
>>>>> ~/.bash_profile:
>>>>> 
>>>>> # Mac OS X
>>>>> alias Rdevel='R_LIBS_USER=~/Library/R/2.15-bioc-devel/library R'
>>>>> alias Rrelease='R_LIBS_USER=~/Library/R/2.15-bioc-release/library R'
>>>> 
>>>> 
>>>> i've created this ~/.bash_profile in my laptop running Mac OSX 10.6.8 and
>>>> when i start with
>>>> 
>>>> $ Rdevel
>>>> 
>>>> my .libPaths() always reports the release directory and never the devel
>>>> directory, i.e.,
>>>> 
>>>>> .libPaths()
>>>> [1] "/Users/rcastelo/Library/R/2.15-bioc-release/library"
>>>> [2] "/Library/Frameworks/R.framework/Versions/2.15/Resources/library"
>>>> 
>>>> if i rename '~/Library/R/2.15-bioc-release' to 'tmp' or i comment the alias
>>>> line about the release in the .bash_profile (i.e., trying to force the
>>>> system to see only ~/Library/R/2.15-bioc-devel/library) i only get the
>>>> default path, i.e.,
>>>> 
>>>>> .libPaths()
>>>> [1] "/Library/Frameworks/R.framework/Versions/2.15/Resources/library"
>>>> 
>>>> i've checked permissions and both directories
>>>> ~/Library/R/2.15-bioc-release/library and
>>>> ~/Library/R/2.15-bioc-devel/library are exactly the same in this regard
>>>> (drwxr-xr-x).
>>>> 
>>>> can any Mac OSX user replicate this problem?
>>>> 
>>> 
>>> I was having a similar problem until I realized that I hadn't created
>>> both directories (I had only created one of them). After I did that,
>>> my aliases worked just fine.
>>> 
>>> Creating them in .bash_profile has also not been a problem for me.
>>> 
>>> Nico, thanks for sharing your script, but if you are working with
>>> R-devel (2.16) I'm not sure you need it. R 2.16 will keep all its
>>> packages in a separate directory from 2.15 automatically. The
>>> R_LIBS_USER trick is intended for using R 2.15 with two different
>>> versions of Bioconductor.
>>> 
>>> Thanks,
>>> Dan
>>> 
>>> 
>>>> if not, i guess i must be doing something wrong, but i have no clue what it
>>>> could be, any hint to a possible diagnostic/QC will be very much
>>>> appreciated.
>>>> 
>>>> thanks,
>>>> robert.
>>> 
>>> _______________________________________________
>>> Bioc-devel at r-project.org mailing list
>>> https://stat.ethz.ch/mailman/listinfo/bioc-devel
>>