| Summary: | smart: --yaml option doesn't work | ||
|---|---|---|---|
| Product: | [Runtime] Package Management Issues | Reporter: | Yi Zhao <yi.zhao> |
| Component: | package-management-issues | Assignee: | Paul Eggleton <bluelightning> |
| Status: | RESOLVED FIXED | QA Contact: | |
| Severity: | normal | ||
| Priority: | Medium | CC: | bluelightning, bogdan.a.marinescu, hongxu.jia, liang.li2, mark.hatle, Qi.Chen |
| Version: | 1.4 | ||
| Target Milestone: | 1.4 M4 | ||
| Hardware: | x86 | ||
| OS: | Multiple | ||
| Whiteboard: | |||
| OS type for building Yocto: | --- | Type of Regression: | --- |
| Verified: | Documentation change: | --- | |
|
Description
Yi Zhao
2013-01-22 07:45:09 UTC
This is simply because the yaml python module isn't installed. I wouldn't consider this core functionality, so the question is do we want to add a dependency just for the sake of enabling this option? Changing to NEEDINFO, see my above comment yaml is a pretty common way of exporting information from an RPM repository. We likely should package up the yaml python module.. and then add a test in smart to see if it is installed. That way it works if the user needs it, but it's disabled (to save space) for the normal case. Sounds like a sensible approach to me - thanks. It turns out there's a python-pyyaml recipe in meta-oe, it just needed an update; I have sent a patch to update it and the libyaml recipe that is also there. I've also sent an OE-Core patch to ask the user to install PyYAML if it can't be imported instead of printing a python backtrace. The OE-Core/Poky part of this patch is merged: http://git.yoctoproject.org/cgit/cgit.cgi/poky/commit/?id=0d85fa531b03f0e7b3ad3d2a5a4d92293b43b8ff The meta-oe part is still pending. libyaml/python-pyyaml changes have now been merged into meta-oe: http://cgit.openembedded.org/meta-openembedded/commit/?id=9c824550785d96dda0e6aeb4a2a37bb2c30dc125 http://cgit.openembedded.org/meta-openembedded/commit/?id=f976a44387f45ac5ede7d74d511bef1228599597 |