incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Stephen McConnell <>
Subject Re: Another cut at roles and responsibilities
Date Tue, 23 Sep 2003 00:59:03 GMT


Have just read though your email and I feel that I have very strong 
empathy with the position your raising - but all the same I'm going to 
disagree with you!  I'm confident that if we were in a cafe down in the 
14e we would tie this up nicely in less that a couple of hours.  But 
that isn't the case so I'll try my best to present the issues I see in 
this email.

Zut ... Australia really is at the end of the earth relative to France!
(Zut translated into Australian is B***** H***).

Berin Lautenbach wrote:

>>From: Stephen McConnell <>
>>1.  Entities (Board, Parent, Incubator PMC) should not assigned actional
>>    responsibilities - only decision responsibility.  Actional reposibility
>>    should be assigned to roles that are represented by accountable
>>    individuals.  There were a couple of places in the document that
>>    needed to be tightened up in this respect.
>Personally Not sure I fully agree with this,
>having seen XMLBeans.  If the XML Project wants
>to have the Incubator take on something on its
>behalf, then there is a two way accountability.  I
>fully believe that the XML Project has to take
>some accountability for assisting the podling.
>That accountability (in the case of XMLBeans) is
>discharged by the Shepherd, who is a member of the
>XML PMC, but can call on others in the XML project
>for assistance at any time.

Consider the following assertion.  The XML Project PMC, the Incubator 
PMC, the Avalon PMC, the Apache Board, ... all of these are basically 
dysfunctional entities when it comes down to doing something 
actionable.  These entities are only good for taking decisions (although 
I must confess that the Board does break this logic from time to time).  
Let me get to the point - the XML Project PMC can make a decision to 
sponsor something. In doing so the XML Project PMC Chair has a 
responsibility concerning the implementation of the decision of the 
PMC.  Now we all know that PMC chairs are gods, and as gods, they are 
surrounded by angels, and gods like playing golf, so gods, being 
responsible, delegate things to angels. Fortunately we can associate 
names with angels, we can hold them responsible and through them we hold 
the gods accountable.  What this means is that the XML Project is doing 
what you want - but me - as a outsider, I can point a finger at an angel 
and I can "hey - this needs to be done - and you (angel) personally are 
responsible".  If that doesn't get done - I can go to god and say "hey 
god, something is wrong - your angel isn't doing what he/she/it? should 
be doing and your responsible.  God gets kind of annoyed - goes to the 
council of angels (the XML Project PMC) and says - hey guys - we have a 
problem (meaning hey guys I have a problem).  God, using his immortal 
powers sends down another angel to fix the problem.

Have you ever seen the movie Dogma - at the end of the day *she* is 

Bottom line - we are always dealing with individuals. The individual may 
change over time, but there is an individual that is responsible and 
therefore accountable.

>Otherwise this is throwing all the responsibility
>back on a couple of people.  To me the whole
>Apache concept is about community, so lets
>demonstrate what that means to the podlings.
>If Ted stops doing his role as Shepherd, then I
>would see it as the responsibility of the XML
>project to step in and find someone else.

Small change in wording.  "If Ted stops doing his role as Shepherd, then 
I would see it as the responsibility of the XML Project PMC Chair" to 
step in and find someone else."

>>My impression is that we are actually aiming towards the same thing but 
>>that what you thinking of as Sheperd is what I'm thinking of as 
>>Sponsor.  There are a few other little things but I thought it best to 
>>get these two items clarified first.
>I think you are correct, that we are heading to
>the same end, but I think it important to 
>separate the sponsor of the original proposal
>away from the incubation.
>There are people who are visionaries.  "I can see
>why this is a great project and why it will be
>a good fit for Apache".  They can help a
>candidate "sell" a proposal to Apache.  Are they
>necessarily the best person to help a project
>through Incubation?  Not so sure.  

Absolutely 100% agree. 

But hang in there for a moment and thing about separation of these 
roles.  One role "A" is about responsible representation and guidance 
with a engagement that is implicit for the duration of incubation - for 
better or worse.  Another role "B" is about vision, excitement, 
opportunity, enterprise.  What the policies and procedures of incubation 
need is "A".  What the project needs initially and on re-occurring 
occasions is a brilliant "B".  But "B" is not the subject of concern of 
an incubation policy.  I think "A" needs to be on the PMC and to 
represent the project and I think "B" needs to in the public face making 
sure that the value proposition is communicated.  Tying "B" to a set of 
policies and procedures is the last thing you want.  But it does mean 
you need to establish an "A" for the long haul.

"A" == Respected and Recognized Sponsor
"B" == Director of Marketing

>To me, that's what the very notion of a shepherd is - someone
>who guards and protects the flock.  


Substitute you idea of Shepard for Sponsor.  Assume you have a Marketing 
Director in the wings and that you Sponsor and Marketing Directory are 
secretly working together on a plan titled "72 hour Incubator Exit 
Strategy".  Also assume that the Shepherd is the one to overcome (kind 
of like a VC Investor).  He has final say - do you get the green light 
or not - so everything your Sponsor and your Marking Director do is to 
move the Shepherd along the path you would like.  If you do this right - 
you have the ingredients backing you (a good project with a clean 
profile) then getting passed the Shepherd should not be a problem.  Keep 
in mind that Shepherds are simple minded people that know a lot about 
sheep but don't know anything about what sheep actually think. Also keep 
in mind that the Shepherd can kill the sheep with reasonable cause. But 
if you have a Marketing Manager in the wings - and if the project is OK 
- you exit, the Shepherd gets sent home with a pat on the back and a 
round of cheese - and the sheep run around looking happy and content - 
the Marketing Manager drives off looking or a new challenge, and you 
lean back in you chair, look at the screen, smile, and say to yourself 
.... "it works".

Cheers, Steve.

p.s. Who is someone terrible disappointed that you, an Australin 
to-boot, removed the so carefully and selectively prepared phrase 
"Incubator PMC meanderings" in you last wiki edit!

Will I ever forgive this guy?


Stephen J. McConnell

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message