Dermot McCluskey wrote:
> As a convenience, eg the Approver spots a small, obvious
> mistake and decides to fix it themselves, rather than
> go back-and-forth with the submitter.
> 
> I know Christian (and others?) do this frequently, so I'll
> let them speak up for how useful this functionality is.  I
> think it's also been useful for them to touch a file in order
> to kick off a build.

I'm really uncomfortable with that functionality.  It puts the approver 
in the place of changing someone else's work (possibly) without that 
person's knowledge or agreement.  As convenient as it might be, I'm not 
really comfortable with this.

The little things that people need to fix are part of the education 
process and getting used to performing project work.  Also, reviewers 
need to get use to delegating that work to the submitter so that they 
can *learn* about the correct process for submitting packages.

Allowing the reviewer to edit the and approve materials violates the 
fundamental idea of a peer review process.  There's just too much room 
for accidents in my view with letting the approvers be reviewer, editor, 
and approver no matter how good their intentions are.

There are a few possible solutions to me:

* The submitter gets to control over whether others are allowed to edit 
the submission.

* Email notifications are sent and a comment is logged against the 
submission with the exact differences any time a submission is changed 
so that its clear who has changed what and hopefully why.

Cheers,
-- 
Shawn Walker

Reply via email to