[R] Build error on Windows

Duncan Murdoch murdoch at stats.uwo.ca
Sat Sep 30 15:13:36 CEST 2006


On 9/29/2006 5:41 PM, Pankaj Savdekar wrote:
> Thanks for the quick reply.
> 
>> On 9/29/2006 8:53 AM, Pankaj Savdekar wrote:
>>> Hi,
>>>
>>> I'm trying to build R-2.3.1 on windows, but make gives me following error 
>>> while building pkg-base:
>>> ---------- Making package base ------------
>>>   adding build stamp to DESCRIPTION
>>> make[4]: *** [frontmatter] Error 1
>>> make[3]: *** [all] Error 2
>>> make[2]: *** [pkg-base] Error 2
>>> make[1]: *** [rpackage] Error 2
>>> make: *** [all] Error 2
>>>
>>> Please note that R.exe, Rterm.exe, Rgui.exe, RCmd.exe are build without 
>>> any errors.
>>>
>>> I have three questions, can anyone please help me to resolve it?
>>> 1. How to solve (or get more details) of the above mentioned error?
>> You need to look through the make files, to see what was happening. Reading 
>> the messages in reverse order:  "make all" called "make rpackage" and so on 
>> to "make frontmatter".  The errors don't tell you which makefiles these are 
>> in, but the "frontmatter" target only occurs in src/gnuwin32/MakePkg.  You 
>> could try deleting the "@" signs from the lines for that target to see 
>> exactly what was happening when the error was generated.
> 
> Yes I could figure out the source of 'frontmatter', but my problem is, there 
> is no error message. I tried 'make -d' too. I tried removing '@', but no 
> change.
> 
>> I'd guess that this is happening because your build is messed up:  the base 
>> package is used in later build steps.  If you start from a clean checkout 
>> and just call "make", you probably won't see this.
> 
> Is there any way to check what could have been wrong in building base 
> package?

You could look at whatever command in the makefile failed, and try it 
outside of the makefile, try variations on it, etc.  I can't give more 
specific advice without more specific information on where the error 
happened.

If you're not used to working in Windows, remember that it is not like 
Unix in several ways.  In particular, you can't delete an open file, 
because it's considered an error for a file to exist unless it has a 
valid directory entry.  If you try to replace a file that is open, the 
replacement will fail, and that may lead to other errors later.

Duncan Murdoch



More information about the R-help mailing list