Message-ID: <Pine.LNX.4.61.0504190931170.14838@gannet.stats>
Date: 2005-04-19T08:43:46Z
From: Brian Ripley
Subject: error due to locales again
In-Reply-To: <0IF600CSDNFDZD@mail.fudan.edu.cn>
On Tue, 19 Apr 2005, ronggui wrote:
> Prof Brian Ripley,
>
> I found the error by chance. the error is found in R2.1.0 too.
As far as I can see, if anyone had run any example() in R on a Windows
machine in your locale it would have failed. But no one reported a
problem. I don't even have that locale on my Windows machine, and it is
in a format I have never seen before.
>> Yes. Please do beta-test *before* release: R 2.1.0 is released.
>>
>> If no one actually tests R in your language, we will not discover that
>> there are problems in that language. Over to the users to beta-test when
>> asked to.
My point is that we rely on users to test things we cannot test. Unless a
user participated in the beta testing (s)he should regard his/her locale
as unsupported.
I have attempted to fix this in R-patched, so please try a build of that
in a day or so (I am not sure automated builds for Windows are running
yet).
--
Brian D. Ripley, ripley at stats.ox.ac.uk
Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/
University of Oxford, Tel: +44 1865 272861 (self)
1 South Parks Road, +44 1865 272866 (PA)
Oxford OX1 3TG, UK Fax: +44 1865 272595