[R] Re: [RMySQL] unable to establish connection since R-1.8.0 upgrade

David James dj at research.bell-labs.com
Tue Nov 25 00:29:26 CET 2003


Hi Laurent,

The issue is a very simple one to correct --- you only need to
re-install RMySQL with the R version you're currently using.

The issue is that the internal representation of objects in the
methods package changed in 1.8.0, and thus packages that use S4
classes and methods and save their images (like RMySQL) and were
installed prior to 1.8.0 need to be re-installed to be used with
newer R versions.

Hope this helps,

--
David

Laurent Faisnel wrote:
> Hi David,
> 
> I have problems with RMySQL since I have upgraded to R-1.8.0 (now I have 
> R-1.8.1). I have been using this package for quite a long time (don't 
> konw if you remember, you helped me out with a function lean.exec which 
> runs queries much faster than the standard dbGetQuery), without trouble, 
> but there seems to be something incompatible with R >= R-1.8.0.
> 
> The error comes up when I try to establish a connection :
> 
>  > library(DBI)
>  > library(RMySQL)
>  > dbConnect("MySQL")
> Error in initialize(value, ...) : Invalid names for slots of class 
> MySQLDriver: Id
> 
> This is the error I have after having added to the NAMESPACE file of the 
> "methods" package the line :
> 				export(.valueClassTest)
> 
> following John Chambers' piece of advice (see 
> http://maths.newcastle.edu.au/~rking/R/devel/03b/0861.html).
> 
> Before adding that line I had the same problem Fernando mentionned (see 
> previous link).
> 
> My settings are :
> 
> Red Hat Linux 8.0
> R-1.8.1
> RMySQL-0.5.2
> 
> I don't have a sufficient knowledge of the way R is programmed to fix 
> the problem by myself. What's your opinion about that ? Should RMySQL be 
> updated ? Are you working on that ? In the previous debate you seemed to 
> believe the problem lies in R (>= 1.8.0), and not in RMySQL. I'm 
> enclined to think you may be right, since I had a few difficulties with 
> R >= 1.8.0 (new library to be added for X11, eg).
> 
> If you think this can be useful to others you can cc your answer to R-devel.
> 
> Thanks in advance for any help,
> Best regards,
> Laurent




More information about the R-help mailing list