>>>>> Henrik Bengtsson <h...@biostat.ucsf.edu>
>>>>> on Mon, 7 Apr 2014 08:20:27 -0700 writes:
> On Mon, Apr 7, 2014 at 1:40 AM, Martin Maechler
> <maech...@stat.math.ethz.ch> wrote:
>>>>>>> Henrik Bengtsson <h...@biostat.ucsf.edu> on Sun, 6 Apr
>>>>>>> 2014 21:33:15 -0700 writes:
>>
>> > Contrary to other functions in 'base', attach() output
>> > messages to stdout instead of stdout, e.g.
>>
>> >> a <- 1 capture.output(attach(list(a=1))) > [1] "The
>> following object is masked _by_ .GlobalEnv:" > [2] "" >
>> [3] " a"
>>
>> > Shouldn't this message go to stderr?
>>
>> well, it this is changed... it should be changed to use
>> message() really -- as library() does.
>>
>> Then, it will not only go to stderr, but also be
>> something you can e.g. use suppressMessages( . ) with.
>>
>> I tend to agree to change this from cat() to message(),
>> and will do so after a bit of waiting..
> Thanks - sounds good.
> I proposed cat(..., file=stderr()) rather that message(),
> because I saw it was used in several other place in 'base'
> - left overs from a earlier era?
partly, probably. In principle you / we should not only look in
'base' but all "base packages", i.e. all in
installed.packages(priority = "base")
Yesterday, I have committed the change -- to use message() --
to R-devel, i.e., of course too late to make it into
R 3.1.0, released tomorrow.
Martin
______________________________________________
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
