[Rd] testing on platforms, was Fortran error

[Simon Urbanek]

John,

On Nov 21, 2011, at 8:17 AM, John C Nash wrote:

> I think the poster is interested in being able to try the build/check on a Mac in the
> fashion that Winbuilder does. That is, rather than have CRAN do all the platform checks,
> is there a way to submit a package to be tested for just one platform?
> 
> It may be useful to have such a facility so package builders can test before submission,
> and from my own experience I believe would help to render packages more platform neutral.
> 

It has been discussed and there was a tentative plan (which involves getting a dedicated machine for this and R-forge builds), but nothing happened so far.


> I suspect that all the pieces are in place to do this, but perhaps a nice front end is
> needed. Is that something for a Google Summer of Code task?
> 

Software infrastructure is not a problem. For example I have full multi-platform submission system in place for my RForge.net (since I'm building Linux, OS X and Windows).

Personally, my view is that developers should be using the existing package-based build systems (like R-forge or RForge.net). I am paranoid and thus reluctant to perform builds on my own systems for packages that have no source provenance (i.e. if I don't have an accountable registered user) - for security reasons.

Cheers,
Simon



> Best,
> 
> JN
> 
> On 11/21/2011 06:00 AM, r-devel-request at r-project.org wrote:
>> What does 'submit a test mean'?  Any test you write can check the 
>> value of Sys.info()["sysname"]: it is "Darwin" on Mac OS X.
>> Or R.version$platform, which will be something like 
>> i386-apple-darwin9.8.0.
> 
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
> 
>