The "Membership" section of the governance charter draft

Máirín Duffy duffy at redhat.com
Thu Oct 31 19:36:08 UTC 2013


Hi,

> On 10/31/2013 12:55 AM, Máirín Duffy wrote:
>> can you maybe provide what you were specifically thinking of when you
>> mentioned 'server community,'

On 10/31/2013 01:30 PM, "Jóhann B. Guðmundsson" wrote:
> The role of server administrators

>>   in your draft? Who were you thinking of /
>> how would you define such a person?

> Person that is active in his field but in essence you cannot apply any
> particular role to a community of volunteers since the role they play is
> what matters to *them* and what *they* decide at the moment they
> contribute *their free time* to it.

I think this is the point at which I can't follow. We appear to agree
that being a member of the server community requires at least some
experience in the role of being a server administrator. You go on to
say, however, that we can't require that necessary experience for
working group members because working group members are volunteers?

Or by referring to volunteers, do you mean the members of the community
who aren't necessarily on the working group who are doing the work? I am
not referring to that body of people at all; this section of the
governance charter specifically refers to membership in the working
group itself so I thought that is what we are talking about.

That seems at odds to me to say that we should have a certain number of
people from the 'server community' as members of the working group, but
to also say that we cannot require the experience necessary to be
considered a member of the 'server community.'

How can we make sure we have a composition with the correct background
to make informed decisions about what to do if we have no requirements
about said background?

>>   But certainly, I think it would help to have people
>> who have this experience make up the majority (5) of the group because
>> don't you need to understand about how servers are used to understand
>> how to develop a successful server product?
> 
> Yes to develop a server product, for PRD no you dont.

I think you might have this backwards? I've been involved in the writing
of multiple PRDs, many specifically for enterprise systems management
software. In my experience, I've found that you do actually need to
understand how servers are used to create a PRD for a server product.
The PRD is essentially the blueprint for that product or feature.

You need to be a developer to develop the product and that specifically
does not require experience in system administration.

>> How do you think the requirement would drive away potential
>> contributors?
> 
> Simple because by doing that you are adding a barrier of entry and at
> the same time forming a group of elites.

We are writing on this mailing list in English. Aren't we creating a
barrier of entry by requiring the ability to understand, read, and write
English to participate?

>>>   ( those contributors might make up them not meeting the said
>>> requirements by pure enthusiasm and the more experience would just help
>>> guide that energy into the right places )
>> Couldn't they take one of the four seats that doesn't require the
>> background?
>>
>>> Actually after going through an PDR
>> Which one?
> 
> Actual there where several I went through to see the common denominator
> in them so I could start writing a sample draft for us to start working on.

It would be nice if we could share the specific examples to look
through? There are a lot of bad PRDs out there and some good ones. Maybe
we could talk about which ones we like and why and figure out what
sections we think we should have based on that discussion?

> PRD are very specific in their nature and function and can only be
> applied to a single product not community of products let alone on a
> community of volunteers.

We're not proposing to apply a PRD to a community of products or a
community of volunteers. We've been tasked to create a PRD for a
specific product: Fedora Server. Isn't that right, or am I confused?

> Applying it to these 3 groups will fail in the same manner as you tried
> to apply it to Red Hat as a company or Fedora in whole as in making a
> single product out of both of these.

I don't understand why. Can you explain? Where is a PRD being applied to
people instead of the specific product?
> 
> And afaik I know the other half and the presiding one required to make
> it work is the MRD which the community should have done first and that
> part has been conveniently left out by the individuals driving this.

An MRD is not required to create a PRD. It's a nice-to-have. I've
proposed in another thread that, in lieu of an MRD, we develop a set of
personas. It would provide us a similar kind of information. What do you
think about this?

> The function of the voting member is more of steering the ship then
> dictate the direction as in ensuring that we get from point a) to point
> b) and stay on course

That makes sense and sounds similar to how the Fedora Board is meant to
work -

> In the end of the day only thing we actually would need to vote upon (
> as I see it ) is which server application we make the first product out
> of and that's a vote everyone should be a part of anyway.

I'm not sure, having spent 4 hours of today combing through the previous
discussions on this list from the past week, that other working group
members are on board with creating multiple (I believe you said
500-550?) server products out of server applications. I don't think that
approach is something we have consensus on.

Would it make sense to hash this '500 product' proposal out at the next
meeting? Or in a different thread perhaps?

>> I don't follow; can you explain a little bit more what you mean here? I
>> don't think the server working group is pushing 3 products; isn't the
>> server one of three products? (The three being server, cloud,
>> workstation?) Or am I missing a reference here?
> 
> You cannot apply PRD to those product again PRD are very specific in
> their nature and function

>> In your opinion, is the current composition of the working group is
>> temporary and just to set up the framework,
> 
> Not in my opinion afaik there is supposed to be election in January in
> every WG to let each group re-elect or replace the initial pointed
> members but maybe I misunderstood something.

Oh okay, that might be correct. I simply don't know. Can Stephen or Matt
clear this one up? I don't see it written anywhere but maybe I'm looking
in the wrong places. I think this is something we should make clear?

> Attempting to apply PRD on a community of volunteers is rather bold move
> and we dont even have proper tools to do PRD and track them and their
> progress properly and so efficiently in the project.

What tools do you think we need?

Also, as a summary which might make this easier to follow since we're
covering a lot of things not necessarily specific to the membership
section of the governance charter, here's some of the loose threads I
think we should take up either in other threads, in IRC, or at a meeting:

- What exactly is a PRD? What are some examples of good PRDs? What are
some examples of bad PRDs? What sections does a PRD typically have? What
sections do we want our PRD to have? How detailed / high-level should
our PRD be?

- Do we need an MRD or would an alternative set of related data (such as
Personas, which I proposed in a separate thread) be sufficient in
helping us understand our target users well enough to construct a
sensible / reasonable PRD?

- Is the Fedora Server one product, or is it a set of 500+
application-based products under a Fedora Server brand? (Is the latter a
correct assessment of your proposal, Jóhann?)

- Is there an election planned in January to re-elect/replace all of the
working group members?

- What tools are necessary to create a PRD?


~m



More information about the server mailing list