Spins Wrangler description

[Bruno Wolff III]

In order to help recruit a spins wrangler I have written up a wiki page
describing what they do. Eventually some of the tasks should get more
detail, but I think this provides a general overview. Feel free to tweak
the page. The plan is to go over it at Monday's meeting.
https://fedoraproject.org/wiki/Spins_Wrangler

[Kevin Fenzi]

On Wed, 30 Jun 2010 19:08:27 -0500
Bruno Wolff III <bruno(a)wolff.to&gt; wrote:

>  In order to help recruit a spins wrangler I have written up a wiki
>  page describing what they do. Eventually some of the tasks should get
>  more detail, but I think this provides a general overview. Feel free
>  to tweak the page. The plan is to go over it at Monday's meeting.
>  https://fedoraproject.org/wiki/Spins_Wrangler 
Looks like at least a good start.

There might be specific dates they need to remind people of things or
take actions if deadlines are missed. 

Thanks for working on this Bruno. 

kevin

[Paul W. Frields]

On Wed, Jun 30, 2010 at 11:25:50PM -0600, Kevin Fenzi wrote:
>  On Wed, 30 Jun 2010 19:08:27 -0500
>  Bruno Wolff III <bruno(a)wolff.to&gt; wrote:
>  
>  > In order to help recruit a spins wrangler I have written up a wiki
>  > page describing what they do. Eventually some of the tasks should get
>  > more detail, but I think this provides a general overview. Feel free
>  > to tweak the page. The plan is to go over it at Monday's meeting.
>  > https://fedoraproject.org/wiki/Spins_Wrangler
>  
>  Looks like at least a good start.
>  
>  There might be specific dates they need to remind people of things or
>  take actions if deadlines are missed.  
Yes, excellent point.  The actual dates for testing and verification
to rel-eng that things are ready should really feed into this as
well.  So I'd recommend getting together with rel-eng to make sure
those dates jibe with the overall release schedule now.  I think that
would make it easier for a wrangler to understand the scope of the
responsibilities as well.

In all honesty, from what I've seen of feature and spin wrangling,
it's not so much time consuming, but rather requiring organization and
minding the calendar -- as long as the constituents (in this case the
Spins SIG) are also paying attention.  The wrangler and the SIG should
have a clear way to determine "Spin owners U, V, W, and Y have all
responded by <DATE>, but spin owner X hasn't, therefore <REMEDY>."
Being able to rely on a very objective standard keeps everyone's
expectations met.

>  Thanks for working on this Bruno.  
+10.

-- 
Paul W. Frields                                http://paul.frields.org/
  gpg fingerprint: 3DA6 A0AC 6D58 FEC4 0233  5906 ACDB C937 BD11 3717
  http://redhat.com/   -  -  -  -   http://pfrields.fedorapeople.org/
          Where open source multiplies: http://opensource.com

[Bruno Wolff III]

On Thu, Jul 01, 2010 at 08:18:53 -0400,
  "Paul W. Frields" <stickster(a)gmail.com&gt; wrote:
>  
>  Yes, excellent point.  The actual dates for testing and verification
>  to rel-eng that things are ready should really feed into this as
>  well.  So I'd recommend getting together with rel-eng to make sure
>  those dates jibe with the overall release schedule now.  I think that
>  would make it easier for a wrangler to understand the scope of the
>  responsibilities as well. 
I was thinking that if we can aggree that spin owners need to sign off
their spins for final at Monday's meeting, I'd try to make a releng
meeting to discuss how they would like stuff to flow. (Whether spin
owner to wrangler or direct to whoever is managing things on releng's
side.) Also what they think the timeline should be and what options are
practical if there are problems or someone doesn't reply in time.

>  In all honesty, from what I've seen of feature and spin
> wrangling,
>  it's not so much time consuming, but rather requiring organization and
>  minding the calendar -- as long as the constituents (in this case the
>  Spins SIG) are also paying attention.  The wrangler and the SIG should
>  have a clear way to determine "Spin owners U, V, W, and Y have all
>  responded by <DATE>, but spin owner X hasn't, therefore <REMEDY>."
>  Being able to rely on a very objective standard keeps everyone's
>  expectations met. 
For the most part. I think doing reviews of spins that do tricky stuff
(not just package lists) can be time consuming in indivual cases. However,
I think that is something that the wrangler should be able to delegate
if need be (for time or expertise).

I'd like to see a page just with easy tests for broken things. There
is a pointer to a validor somewhere in the documentation, but I'd rather
see the tests that one might used to check for cruft in existing spins
in one place under testing.

At one point we were going to do test plans for spins, but I think that
currently we aren't having QA do anything with spins other than Desktop
and maybe KDE. In the near term I think having the owners try out what
they feel is critical is reasonable.