Closed stelfrich closed 6 years ago
should the
scijava.*
properties take precedence overimagej.*
ones if both are set simultaneously?
That seems the most intuitive to me, yeah. I would suggest not formally specifying behavior in that circumstance, though. Then if some unintuitive circumstance arises where it would be convenient for the precedence to go the other way, we can flip it. But I doubt it will be important.
should the
scijava.*
properties take precedence overimagej.*
ones if both are set simultaneously?
How about ignoring all imagej
properties if at least one scijava
property is set, and issuing a warning to make it transparent to the caller what's happening?
How about ignoring all
imagej
properties if at least onescijava
property is set, and issuing a warning to make it transparent to the caller what's happening?
I like that and will start implementing it!
@ctrueden I have implemented @imagejan's suggestion and have only "defined" the behavior in a log message. Should we get rid of that as well?
This PR migrates the functionality of
imagej-maven-plugin
as of9240da2
toscijava-maven-plugin
.It keeps backwards compatibility as far as it's possible. This means, supporting the legacy properties
imagej.app.directory
,imagej.app.subdirectory
,imagej.deleteOtherVersions
, anddelete.other.versions
. This means that projects that have usedimagej-maven-plugin
implicitly by makingpom-scijava
the parent for their project, should see no difference once, we switch toscijava-maven-plugin
inpom-scijava
.There is one point that I would like some input on: should the
scijava.*
properties take precedence overimagej.*
ones if both are set simultaneously? Do you have an opinion, @ctrueden?TODOs:
AbstractCopyJarsMojo