[R] "all connections are in use" error during lazyload stage of packa ge installation

[Tuszynski, Jaroslaw W.]

 I found the problem, by doing comparison of directories and files of
working and not working versions, and applying changes one by one until one
caused install to fail. It was a case of having a call to a "source"
function somewhere in my code, that I forgot about. 

I was definitely doing something silly , but it was not a meaningful error
message.

Jarek
====================================================\=======

 Jarek Tuszynski, PhD.                           o / \ 
 Science Applications International Corporation  <\__,|  
 (703) 676-4192                                   ">   \
 Jaroslaw.W.Tuszynski at saic.com                     `    \



-----Original Message-----
From: ggrothendieck at gmail.com [mailto:ggrothendieck at gmail.com] 
Sent: Wednesday, June 29, 2005 10:07 AM
To: Tuszynski, Jaroslaw W.
Cc: r help
Subject: Re: [R] "all connections are in use" error during lazyload stage of
packa ge installation

On 6/29/05, Tuszynski, Jaroslaw W. <JAROSLAW.W.TUSZYNSKI at saic.com> wrote:
> Hi,
> 
> I suddenly started getting strange errors while working on my caTools
> package:
> 
>     >RCMD install C:/programs/R/rw2011/src/library/caTools
>      ......
>        preparing package caTools for lazy loading
>        Error in file(file, "r", encoding = encoding) :
>                all connections are in use
>        Execution halted
>        make: *** [lazyload] Error 1
>        *** Installation of caTools failed ***
> 
> I searched through R website and mailing lists but did not found many
clues.
> I am not running R environment so showConnections and 
> CloseAllConnections functions are not helpful. Did I run into some OS 
> specific limit on number of files allowed? I closed all other programs 
> and rebooted my machine but it did not help.

I don't know what the problem is but you could see if adding

LazyLoad: no

to the DESCRIPTION file causes it to go away.  Even if it does it would
probably be best to track down what the cause is and in the past when I have
had problems running R CMD on a package of mine I repeatedly removed half my
package and reran R CMD until I found the offending portion.