[Bioc-devel] Failing to installing prada in 1.9 release

[Florian Hahne]

Hi Wolfgang, Seth, all,

I tracked down the build error to a namespace problem as Seth has 
already pointed out to me, however, I'm not sure if my solution is 
merely a temporary fix:
I assumed that if I want to extend the eSet class in my package it is 
sufficient to do a
importClassesFrom("Biobase", "eSet")

but this produced the following (with R-2.4.1 release)
Error in setIs(class2, cli, extensionObject = obji, doComplete = FALSE, :
Unable to find package environment for class "VersionedBiobase" to 
revise subclass information

So I went ahead and added
importClassesFrom("Biobase", "eSet", "VersionedBiobase" )

which lead me to
Error in setIs(class2, cli, extensionObject = obji, doComplete = FALSE, :
Unable to find package environment for class "Versioned" to revise 
subclass information

Finaly, after adding "Versioned" to the import everthing built ok
importClassesFrom("Biobase", "eSet", "VersionedBiobase", "Versioned")

Is this the necessary procedure for extending the eSet class. Shouldn't 
R be able to deal with the dependecies that arise from importing a 
class, after all these are pretty quite opaque

Best wishes,
Florian




Wolfgang Huber schrieb:
>
> Hi Florian & Biocore,
>
> I just installed an R-2.4.1 from scratch,  then ran
>
>  source("http://www.bioconductor.org/biocLite.R")
>  biocLite("cellHTS")
>
> I get errors that seem to do with the usage of classes from Biobase in 
> prada that hasn't kept up with the moving target that these classes 
> have been...
>