fix the summary.

I guess I'm curious why you'd like such a behavior.

This actually fits with some things I'd been thinking about - namely, the
ability to turn off the auto-archiving of artifacts/javadoc/static
analysis/coverage/etc for a project. For example, I've got a project which, in
its nightly build, generates javadoc and then zips it up with an assembly. But
Hudson (specifically the Maven plugin) sees the javadoc mojo fire, and
automatically archives the Javadoc, whether I want it to do so or not. Or I'm
running Cobertura in the project, but I don't have the Cobertura plugin enabled
for the project in Hudson - and yet it still archives the Cobertura reports,
because the mojo fired.

I'm not sure if there's an elegant way to handle this, and it's probably not
worth doing anything about unless there's an elegant way to handle it. But
it's something to consider.

Well... I have several "heavy" jobs running on hudson, jobs that take hours and
that produces several gigabytes of data which is published as the artifacts.

From time to time I have to stop some job and restart it - e.g. when I have
some fix available and need to test it quickly - without waiting when the
running job finishes. So I abort the build and start it again.
Skipping archiving artifacts will serve for 2 purposes - first, it makes the
"restart" process be faster (archiving can take several minutes) and it also
saves some disk space (a few GB). In such scenario I`m not interested in the
artifacts that were produced in the not finished build.

I can certain imagine the cases when archiving artifacts for the manually
aborted job is necessary - e.g. if the job hanged (for some reasons) and the
artifacts can help to diagnose the problem after the job in totally finished.

• • • Issue 4014 has been marked as a duplicate of this issue. ***

@dl198383104 ... and even IF you want to check the artifacts of an aborted
build, you have them in the workspace until you start the next build. So you
don't need to archive them, as well.

Essentially duplicates JENKINS-22699 (the newer one has a PR already)