R devel - Jul 2006 - Making a patch

Dear R developers,

is there a preferred format or strategy for making a patch to
contribute to a package that is maintained by R-core?  Berwin Turlach
and I have written a very minor extension to lmeControl to allow it to
pass an argument to nlminb for the maximum number of evaluations of
the objective function.  I've edited the nlme/R/lme.R and
nlme/man/lmeControl.Rd files.  I can diff the files, but the options
seem numerous.  I haven't been able to find any hinst on the preferred
approach.  Any advice would be welcomed.

Cheers

Andrew

-- 
Andrew Robinson  
Department of Mathematics and Statistics            Tel: +61-3-8344-9763
University of Melbourne, VIC 3010 Australia         Fax: +61-3-8344-4599
Email: a.robinson at ms.unimelb.edu.au         http://www.ms.unimelb.edu.au

Hi,

Andrew Robinson <A.Robinson <at> ms.unimelb.edu.au>
writes:
> 
> Dear R developers,
> 
> is there a preferred format or strategy for making a patch to
> contribute to a package that is maintained by R-core?  Berwin Turlach
> and I have written a very minor extension to lmeControl to allow it to
> pass an argument to nlminb for the maximum number of evaluations of
> the objective function.  I've edited the nlme/R/lme.R and
> nlme/man/lmeControl.Rd files.  I can diff the files, but the options
> seem numerous.  I haven't been able to find any hinst on the preferred
> approach.  Any advice would be welcomed.
> 
look at 

http://article.gmane.org/gmane.comp.lang.r.devel/8858

perhaps this (with modifications if necesarry) could be added to developer page 

http://developer.r-project.org/

Afterall, there is a good guide on posting mails to r-mailing lists and it would
be good to have the same for pacthes.

Gregor

On 7/23/2006 6:04 PM, Andrew Robinson wrote:
> Dear R developers,
> 
> is there a preferred format or strategy for making a patch to
> contribute to a package that is maintained by R-core?  Berwin Turlach
> and I have written a very minor extension to lmeControl to allow it to
> pass an argument to nlminb for the maximum number of evaluations of
> the objective function.  I've edited the nlme/R/lme.R and
> nlme/man/lmeControl.Rd files.  I can diff the files, but the options
> seem numerous.  I haven't been able to find any hinst on the preferred
> approach.  Any advice would be welcomed.
Gregor pointed out a description of the suggested format.  But the 
strategy is also important:  in order for your patch to be incorporated, 
you need to convince someone to accept it.  To do that, you should post 
the following sort of information along with your patch:

  - what the patch does
  - what problem it solves
  - what problems it may cause

You should also make sure your patch is complete, e.g. if it changes 
lmeControl, then you should also include a patch to the documentation 
for lmeControl.

Duncan Murdoch

Gregor Gorjanc <gregor.gorjanc <at> bfro.uni-lj.si> writes:
> 
> Hi,
> 
> Andrew Robinson <A.Robinson <at> ms.unimelb.edu.au> writes:
> > 
> > Dear R developers,
> > 
> > is there a preferred format or strategy for making a patch to
> > contribute to a package that is maintained by R-core?  Berwin Turlach
...
> 
> look at 
> 
> http://article.gmane.org/gmane.comp.lang.r.devel/8858
Is there also any strategy in accepting patches in plain text or "somehow
compressed" format? The later due to line wrapping in some mailers, etc.

Thanks!

Gregor