[Rd] pure R code package for Windows

[Duncan Murdoch]

On 4/29/2007 9:53 AM, Petr Savicky wrote:
>> You can also use the Windows binary build and check service at 
>> http://129.217.207.166 after reading its disclaimer...
> 
> Thank you very much. This worked including generating .chm file. The resulting
> package is installable under R GUI on Windows XP, which is sufficient for me.
> 
> Let me point out that I did not succeed to install any package
> (including the standard ones like rpart) on Windows using R CMD INSTALL <package>,
> although I have Rtools and perl installed, running and in the path.

On Windows, INSTALL is for installing source packages and is documented 
this way, but you may not have read the Windows-specific docs.  The 
normal way to install binary packages in Windows is to use the 
install.packages() function (or the menu item, which calls it).

It should be a simple modification of the INSTALL script to recognize a 
.zip file, and do a binary install.  I wouldn't put this at a high 
priority:  people who are using R CMD INSTALL tend to be package 
developers, and want to install the source; people who want to install 
the binaries tend not to have the necessary tools to use R CMD INSTALL.

Duncan Murdoch

> 
> The error message for windows binary package rpart from CRAN is:
>   Error in Rdinfo(RdFiles[i]) : missing/empty \name field in 'E:/R/rpart/man/rpart.Rd.gz
>   Rd files must have a non-empty \name.
> 
>> >4. Install the package under Linux and zip the directory
>> >  library/<package> and unzip it in the library directory
>> >  on Windows machine. This works. The package behaves
>> >  correctly. However, I do not think that this is a suggested
>> >  method.
>> 
>> Well, it _is_ a suggested method: see README.packages.
> 
> This also produced a package installable by the standard procedure under
> Windows GUI. Thank you for pointing this out. (Again, R CMD INSTALL does
> not work for the resulting package.)
> 
> All the best, Petr.
> 
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel