OpenAFS RT changes (was Re: [OpenAFS-devel] Re: Moving Forwards)

[Simon Wilkinson]

On 10 Sep 2012, at 20:37, Dave Botsch wrote:

> Can you clarify "a bug" below? Does this qualify as "any bug" or "any
> bug not in the security queue"?

Sorry, all of this applies only to bugs in the openafs-bugs queue. I'm =
not (currently) proposing changing access to bugs in the =
openafs-security queue. I'm not say that openafs-security is perfect as =
is, just that it's a discussion for another time.


> On Mon, Sep 10, 2012 at 08:33:54PM +0100, Simon Wilkinson wrote:
>>=20
>> On 10 Sep 2012, at 16:30, Andrew Deason wrote:
>>=20
>>> Aside from the spam/abuse angle, which people have covered ("it's a
>>> problem, but tractable")... many extant RT accounts are pretty
>>> restricted. IIRC, I can't 'steal' ownership of anything, or comment =
on
>>> any ticket I'm not explicitly included on, which makes it pretty =
darn
>>> difficult to use.
>>>=20
>>> I think that means I can't 'resolve' things owned by other people =
unless
>>> I get them to explicitly reassign ownership. I also cannot 'delete' =
spam
>>> tickets and such.
>>>=20
>>> Opening _that_ much up I believe is simpler.
>>=20
>> Okaty, so as a first step, let's fix this. I believe that this is =
just a case of changing the permission set, and possibly the list of =
users with a given permissions.
>>=20
>> I believe a while back, the offer was made to do this, providing that =
we had some agreement on what those permission sets should be.
>>=20
>> So, I'm proposing the following sets:
>>=20
>> guest:
>> 	As present, can view the openafs-bugs queue, but can't update =
them
>>=20
>> commenter:
>> 	Can do anything to a bug, but can't delete it, or merge it with =
another bug, (these are potentially irreversible steps in RT)
>>=20
>> developer:
>> 	Can do anything to a bug
>>=20
>> And then the existing set of permissions for the security queue. All =
active accounts would be by default in the 'commentor' state. Anyone who =
meets my earlier criteria for +2 access to gerrit would also be in the =
developer set.
>>=20
>> And then we solve the issue of how to do automatic signup as a =
separate step.
>>=20
>> Any comments?
>>=20
>> Cheers,
>>=20
>> Simon.
>>=20
>>=20
>> _______________________________________________
>> OpenAFS-devel mailing list
>> OpenAFS-devel@openafs.org
>> https://lists.openafs.org/mailman/listinfo/openafs-devel
>>=20
>=20
> --=20
> ********************************
> David William Botsch
> Programmer/Analyst
> CNF Computing
> botsch@cnf.cornell.edu
> ********************************
> _______________________________________________
> OpenAFS-devel mailing list
> OpenAFS-devel@openafs.org
> https://lists.openafs.org/mailman/listinfo/openafs-devel