Summary: | dev-lang/R-2.13.1-r1: potentially using java without the proper eclasses | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Ralph Sennhauser (RETIRED) <sera> |
Component: | Current packages | Assignee: | Gentoo Science Mathematics related packages <sci-mathematics> |
Status: | RESOLVED INVALID | ||
Severity: | normal | ||
Priority: | Normal | ||
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Bug Depends on: | |||
Bug Blocks: | 388937 | ||
Attachments: |
build.log with sandbox violations
patch agains R-2.13.1-r1.ebuild |
Description
Ralph Sennhauser (RETIRED)
2011-09-27 20:27:25 UTC
Created attachment 287979 [details]
build.log with sandbox violations
Created attachment 287983 [details]
patch agains R-2.13.1-r1.ebuild
Since the original sandbox issue was workaround system-wide, the remaining question is what is R using java for, and what e.g. java jars, or JNI bindings it installs ? If there is really some java support, and not just spurious configure checks, java eclasses should be used. It seems that Java is only needed for CRAN packages which are written in Java, meaning once R is installed. During configuration, it is only needed to get the java path for proper conafiguration of the R package installer. @java team: please apply the patch to >= dev-lang/R-2.14.1 if you feel it is necessary. What's the status of this bug? Is it still reproducible with recent versions of R? The version listed in this bug is no more (dev-lang/R-2.13.1-r1). I have managed to install =dev-lang/R-3.0.1 (which is the only stabilised version available) without a hitch (no sandbox violations nor warnings occured). @sci-maths: feel free to close it. Taking @java off CC. Closing this bug as per comment #5. |