Can't delete branches from the API that can be deleted from the web UI

Per a discussion with other developers, not having a branch delete over the webservice delete the involved artifacts was a conscious decision in the creation of the method.

That said, this method shouldn't return a 500, as shown in the OOPS.

So the problem is this:
 - the method cannot be used unless it deletes artifacts that there
*is no other way to delete*.

> *is no other way to delete*.

I don't think that's the case though--the artifacts that block (that I see) are stacked_on branches, which can be deleted, and merge proposals, which also can be deleted. All the other linkages should be removed when you use lp_delete or similar, since the exported delete method is destroySelfBreakReferences.

I agree with James:

> It is possible to manually delete all artefacts first, but that's a headache.

but I think in a scripting environment it's not too much to ask so that we don't have accidental deletions across the webservice, where it's much harder (and weirder) to pop up a message to the effect of "you have delete a bunch of other stuff! are you sure?"

On Thu, Jan 27, 2011 at 3:56 AM, j.c.sackett <email address hidden> wrote:
>> *is no other way to delete*.
>
> I don't think that's the case though--the artifacts that block (that I
> see) are stacked_on branches, which can be deleted, and merge proposals,
> which also can be deleted. All the other linkages should be removed when
> you use lp_delete or similar, since the exported delete method is
> destroySelfBreakReferences.

stacked on branches cannot be deleted if they are owned by someone
else; ditto merge proposals.

>> It is possible to manually delete all artefacts first, but that's a
> headache.
>
> but I think in a scripting environment it's not too much to ask so that
> we don't have accidental deletions across the webservice, where it's
> much harder (and weirder) to pop up a message to the effect of "you have
> delete a bunch of other stuff! are you sure?"

And yet if someone writes a manual iterator, isn't the end effect identical?

p-Rob

Fixed in stable r12297 <http://bazaar.launchpad.net/~launchpad-pqm/launchpad/stable/revision/12297>.