|Re: How to tell buckminster to build from a concrete plugin [message #910801 is a reply to message #910768]
||Mon, 10 September 2012 12:08
|| Thomas Hallgren
Registered: July 2009
On 2012-09-10 13:20, Jan Krakora wrote:|
> Hi Thomas,
> thanks for your response.
> 1) Why is there useMaterialization="false"?
Because only p2 aware components will end up in your target platform. Components fetched from Maven or using URL's will
end up in a special project called .buckminster. You can also use an mspec to make them end up elsewhere. Those
components are considered materialized (and can thus be found) although they are not in your TP. If you want to force
buckminster to bypass the and search the net during resolution, then you must use userMaterialization="false".
> 2) I use headless Buckminster, when I call buckminster import ".../bucminster.cquery" it resolves and materializes the
> required components to the workspace, right?
> So the cquery is used for materialization. And then, when I call buckminster perform "some build action", it operates on
> the materialized components. Does it mean that the buckminster's workspace is always regenerated when it's materialized?
No, not unless you use useWorkspace="false" in your cquery. That would force Buckminster to regenerate the workspace. I
would recommend starting with an empty workspace instead when running unattended build. In your IDE, you generally
"regenerate" your workspace using other means such as "git pull" or "svn update".
Powered by FUDForum
. Page generated in 0.20377 seconds