On 06/17/2015 01:15 PM, Dennis Gilmore wrote:
On Wednesday, June 17, 2015 01:03:23 PM Mike McLean wrote:
> ---
> builder/kojid | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/builder/kojid b/builder/kojid
> index b6b79c1..8243434 100755
> --- a/builder/kojid
> +++ b/builder/kojid
> @@ -4460,7 +4460,7 @@ def get_options():
> 'topdir': '/mnt/koji',
> 'topurl': None,
> 'workdir': '/var/tmp/koji',
> - 'pluginpath': '/usr/lib/koji-builder-plugins',
> + 'pluginpath':
> '/usr/lib/koji-builder-plugins:/usr/lib64/koji-builder-plugins',
'mockdir':
> '/var/lib/mock',
> 'mockuser': 'kojibuilder',
> 'packager': 'Koji',
What is the use case here? do we expect to have architecture specific plugins
(compiled c code etc)?
That is certainly possible.
or is it just masking over a packaging bug?
also that. I encountered a plugin that used %{_libdir}. Incorrect, I
know, but given that compiled plugins are theoretically possible, I
figured it wouldn't hurt to check both places.
-shrug- I guess it is probably better to leave this out. Compiled
plugins should probably place their binaries somewhere in the normal
import path and place a noarch wrapper in the plugin dir.
Thanks for the feedback
Dennis
_______________________________________________
koji-devel mailing list
koji-devel(a)lists.fedorahosted.org
https://lists.fedorahosted.org/mailman/listinfo/koji-devel