[R-pkg-devel] GSL update on CRAN

Raphael Hartmann r@ph@e|@h@rtm@nn @end|ng |rom p@ycho|og|e@un|-|re|burg@de
Fri Sep 13 15:17:21 CEST 2019


Hi Ege

Thank you for your answer.

I also think it won't be easy to build it in my package. The hint with
the description file is a good one. Does that get read automatically?
So far, I had the following line 'SystemRequirements: GNU Scientific
Library version >= 2.5' (at that time I was not aware that it needs
only version 2.3). Of course my package got rejected on CRAN when I
tried to upload it.

I can imagine that they all don't have much time. That means I can only
wait and see what happens.

Maybe at some point I will try to upload the package with a kind
message, as you suggested.

Thank you very much for all the hints.


Best wishes
Raphael

On Fri, 13 Sep 2019 13:54:00 +0200
 Ege Rubak <rubak using math.aau.dk> wrote:
>Hi,
>
>It may be a big task to ship all of GSL with your R package, and since
>it is a large standard library which is "easy" to install on most
>systems, I doubt that it's worth the effort. I would recommend putting
>`GSL (>=2.3)` in `SystemRequirements` in your DESCRIPTION file.
>
>As you point out, this requires that the CRAN team can find the time
>to update GSL on their machines. Uwe often sees things on this list,
>so you may be lucky that he replies directly when he has time (the
>CRAN people tend to have busy day jobs also).
>
>If you don't hear anything on this list in the near future you could
>also try to submit your package to CRAN with the message that you are
>aware that it fails on their Windows machines due to an old version of
>GSL which you would like to be updated if possible. If done in a
>respectful manner I think this should be OK and not wasting their
>time, but others on this list may have better suggestions of how to
>approach this issue.
>
>Best wishes,
>Ege
>
>On 11/09/2019 11.49, Raphael Hartmann wrote:
>> Hi,
>> 
>> I hope you don't mind me asking again:
>> Does anybody have an idea on how to implement the whole GSL library
>in
>> my package? I was not able to find sth. on Google.
>> 
>> Whom should I contact for asking whether it is possible to update
>GSL
>> for Windows on CRAN? There were no warnings or errors in the CRAN
>> checks for debian. For Windows on CRAN, the compiler gives only an
>> error where a GSL routine of version 2.3 is used.
>> 
>> Best wishes
>> Raphael
>> 
>> 
>> 
>> On Fri, 06 Sep 2019 08:19:37 +0200
>>   "Raphael Hartmann" <raphael.hartmann using psychologie.uni-freiburg.de>
>> wrote:
>>>
>>> Hello,
>>>
>>> I developed an R package that requires at least version 2.3 of GSL
>>> (ftp://ftp.gnu.org/gnu/gsl/). The newest version is 2.6.
>>>
>>> Version 2.3 is now available for over two years, and I was
>wondering
>>> whether it was possible to update GSL on the win-builder? I would
>>> really appreciate it. Without it my package will be rejected on
>CRAN.
>>>
>>> Or is there an option to include the whole GSL library in the
>package?
>>>
>>> On my machines (two debian based and one Windows server 16 and one
>>> Windows 10) the package passes the checks of "R CMD check
>--as-cran".
>>> On builder.r-hub.io with Windows server 12 it is also successful.
>>>
>>> I would be glad for any advice / suggestions / hints.
>>>
>>>
>>> Best wishes
>>> Raphael
>>>
>>> ______________________________________________
>>> R-package-devel using r-project.org mailing list
>>> https://stat.ethz.ch/mailman/listinfo/r-package-devel
>> 
>> --------------------------------------------------------------
>> Albert-Ludwigs-Universität Freiburg
>> Institute of Psychology, Cognition, Action, and Sustainability
>> Engelbergerstrasse 41
>> D-79085 Freiburg
>> Phone: +49 761 203 2465
>> 
>> ______________________________________________
>> R-package-devel using r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-package-devel
>>
>
>______________________________________________
>R-package-devel using r-project.org mailing list
>https://stat.ethz.ch/mailman/listinfo/r-package-devel

--------------------------------------------------------------
Albert-Ludwigs-Universität Freiburg
Institute of Psychology, Cognition, Action, and Sustainability
Engelbergerstrasse 41
D-79085 Freiburg
Phone: +49 761 203 2465



More information about the R-package-devel mailing list