R devel - Jun 2005 - 2.1.0a for Macs (PR#7952)

I am writing regarding serious problems that a number of 
researchers at a workshop held by the Smithsonian Tropical 
Research Institute are encountering with the most recent R 
version for Macs (the patched version of 2.1.0 (2.1.0a) that 
you indicate should be used over 2.1.0).

The following problems are not specific to my computer but are 
being experienced by ALL Mac users at this workshop running 
2.1.0a to varying degrees. I installed the program about a week 
ago and immediately I started getting a warning messages 
particularly when browsing my files using ls(), search(), gc(), 
etc.  2.1.0a automatically adds an open parentheses at the end 
of any command requiring () (only intermittently, however) and 
then generates a warning in red.  I am afraid that I do not 
have the warning because R no longer will open on my computer 
without reinstallation so I am forced to use a PC.  The same 
warning message has been posted by one other on the help email 
list.  More recently, R has begun to crash for no particular 
reason (no large program running).  Often when I then try to 
restart R 2.1.0a crashes during start up and in order to get it 
functioning again I have to reinstall!  The first time this 
happened I decided to go back to an earlier version of R that I 
knew would be stable but following binary compilation no gui is 
installed on my computer and I can't run the program.  The only 
version my computer can now install properly is 2.1.0a but as 
described above the problems with the program prohibit me from 
using it.

If you could please advise on how to deal with these problems 
(particularly the corruption of my system by 2.1.0a) all the 
Mac users at this conference would be appreciative.

Jennifer Baltzer

On Jun 16, 2005, at 5:30 PM, jbaltzer at oeb.harvard.edu wrote:
> I installed the program about a week ago and immediately I started  
> getting a warning messages particularly when browsing my files  
> using ls(), search(), gc(),
This was fixed in R.app build 1564 and was mentioned several times on  
the Mac list. It is recommended to update your R.app. If you have  
problems with the most recent version (as of today it's build 1596),  
please let us know.
> More recently, R has begun to crash for no particular reason (no  
> large program running).  Often when I then try to restart R 2.1.0a  
> crashes during start up and in order to get it functioning again I  
> have to reinstall!
Can you, please, send us the crash report and/or the console output?  
Without any evidence reports like yours are pretty much useless  
(please read the documentation and FAQ!). Also note that R 2.1.1 will  
be released really soon so it's important to report bugs using R  
2.1.1 beta and not old versions.
> The first time this happened I decided to go back to an earlier  
> version of R that I knew would be stable but following binary  
> compilation no gui is installed on my computer and I can't run the  
> program.
What do you mean by "binary compilation" - if you feel like  
installing an older version all you have to do is to get the 2.1.0  
image and install it. However, I doubt that this will solve any  
problems, but it's completely legal to do so. You can also safely use  
older versions of R.app if you want - 2.1.0 and 2.1.0a are by  
definition binary compatible. R itself didn't change substantially  
between 2.1.0 and 2.1.0a (only a couple of bugs were fixed, most  
notably package installation). What did change was R.app and as I  
already mentioned there are good reasons for using the most recent  
version.
> If you could please advise on how to deal with these problems  
> (particularly the corruption of my system by 2.1.0a)
You didn't mention any system corruption so far - you'll have to  
explain this a bit ...

Cheers,
Simon