[R] Error with 1.9.0 - Invalid HOMEDRIVE
Prof Brian Ripley
ripley at stats.ox.ac.uk
Wed Apr 21 11:21:36 CEST 2004
Thanks for the clarification. We have several machines which have all the
critical updates on (and one XP machine with *all* the updates) and both
1.8.1 and 1.9.0 work properly on all the machines I tested this morning.
So there is some additional factor.
If anyone who has the problem knows which critical update caused it (David
Scott), it might help others with the problem.
It might be worth trying setting the environment variable R_USER or HOME
(which I have set, but I have also tested after unsetting it).
For the record, it looks like the C entry point getenv() has been broken,
and since 1.9.0 introduced a length check on the result of getenv it
terminates `cleanly' whereas 1.8.1 segfaults. Or possibly the setting of
environment variables (HOMEDRIVE and HOMEPATH) has been broken, which
would amount to the same thing.
On Wed, 21 Apr 2004, Brett Melbourne wrote:
> Hi Brian,
> R1.9.0 fails with the message "Fatal error: Invalid HOMEDRIVE" and no
> prompt to send an error report to Microsoft.
> R1.8.1 fails with the message "R for Windows GUI front-end has encountered a
> problem and needs to close. We are sorry for the inconvenience", included
> in a prompt to send an error report to Microsoft.
> For both versions, R fails to start at all. I'm running Windows XP Pro, and
> like David Scott I'm supposed to install the XP critical updates as soon as
> they come out, which I did last week.
> I suppose, given the other reports of this behaviour, the subject line
> should be "R fails after installing Windows XP updates".
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
More information about the R-help