hbase-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Daniel Vimont <dan...@commonvox.org>
Subject Re: [DISCUSS] Using the new 'help wanted' tool from comdev
Date Mon, 16 May 2016 07:49:14 GMT
Stack & Sean -- Thanks for your positive responses to all this (and the
intriguing Mother Jones link).

Will any of the HBaseCon2016 events be live-streamed? (I can't make it --
an Osaka/SF round-trip was not in the budget this spring!)

On Sun, May 15, 2016 at 2:05 PM, Stack <stack@duboce.net> wrote:

> On Sat, May 14, 2016 at 7:18 PM, Daniel Vimont <daniel@commonvox.org>
> wrote:
>
> > The first step in working through a problem is acknowledging that there
> IS
> > a problem, so in that respect, this conversation is going well so far!!
> > Here are some thoughts (some of which are barely half-baked, so pardon
> the
> > "raw dough").
> >
> > SHORT-TERM BRAINSTORMING
> > I know that HBaseCon2016 is very close, and all the speaking slots and
> > topic sessions are locked into place, but I do have a few suggestions:
> >
> > (1) At *several* points (not just once!) throughout the plenary sessions,
> > make a statement that this conversation has come up on hbase-dev, and
> it's
> > an issue that you think is worthy of our attention and discussion. (I
> think
> > we're just at the beginnings of a very long conversation, but just
> > announcing that the conversation is underway and inviting people into it
> > seems a good way to start.)
> >
>
> We can do that.
>
>
> > (2) I see that Intel is sending four representatives to make
> presentations
> > (two women and two men!). Intel seems to be at the forefront of
> > diversification efforts in the broader IT community, having set its goal
> > for "full representation" in its workforce by 2020. If these Intel
> > presenters feel comfortable, it would be great to devote a few minutes in
> > one of the plenary sessions to allow them to informally talk about how
> the
> > Intel z is going from their
> > perspective, and whether that offers any lessons that might be applied in
> > the Hadoop/HBase ecosystems.
> >
>
> Let me ask the Intel'ers who are coming to the conference.
>
>
> > (3) For HBaseCon2017, devote complete sessions to the issue of
> > diversification. (Perhaps invite Intel's Chief Diversity Officer,
> Danielle
> > Brown, to lead up some proactive discussions.)
> >
> >
> Yes. I like this suggestion.
>
>
>
> > LONGER-TERM STRATEGIZING
> > Stepping back to the generic questions of how and why to "recruit" new
> > contributors, perhaps this is where my "newbie" eyes can assist in
> taking a
> > fresh look at things, beginning with a big lead-in question:
> >
> > The current "staff" of the HBase project: how were we "recruited", and
> why
> > are we here?
> >
> > When I look at the contributors list, I see what I expect to see: Almost
> > all of the contributors are employees of companies that are the biggest
> > stakeholders in the HBase ecosystem, with the biggest number coming from
> > those companies that are existentially bound to the HBase/Hadoop
> > ecosystems, and a smaller number of contributors coming from companies
> that
> > make strategically vital use of HBase in their operations. My assumption
> is
> > that these companies are not magnanimous philanthropists, they're just
> > (quite appropriately) looking after their own vital interests. Bottom
> line:
> > this majority of contributors are *paid* to be here; their employers
> > "allow" them to spend a percentage (or maybe in some cases, all) of their
> > working hours tending to the care and feeding of the open-source golden
> > goose that keeps their proprietary enterprises going.
> >
> > And then there are people like me (likely in much smaller numbers) who
> are
> > here to beef up their expertise in current technologies in a way that
> also
> > gives them a publicly-displayed "portfolio" of work with which to
> > subsequently seek out paid engagements (of either the contract or
> > employment variety). I'm sure I don't fully buy into the prevailing
> wisdom
> > that "code is the new resume" and "just put your stuff out there on
> GitHub
> > -- WE'LL find YOU!!"; nonetheless, for the last several months I've been
> > setting aside my doubts and building just such a "portfolio".
> >
> >
> The above is a fair assessment I'd say.
>
>
> > So, if my take on the current "staffing" situation is approximately
> > correct, it suggests to me that the quickest path to "full
> representation"
> > in the HBase project is initially through conversations with the Human
> > Resources chiefs and the CEOs of the corporations that provide the
> majority
> > of this project's "staff". What initiatives are underway in these firms
> (1)
> > to recruit a diverse workforce and (2) to assure that once members of
> > underrepresented minorities are brought in, that they find themselves in
> a
> > safe, supportive environment in which they want to stay? Finally, if good
> > initiatives are already underway, what might be done to hasten the
> "trickle
> > down" from these initiatives into a more diversely-staffed HBase project?
> >
> >
>
> There was a call-to-arms a while back that echoed around the valley (Here
> is one origin story [1] that references this interesting looking project
> [2]). Where I work there is a special project afoot to work on problem #1,
> at least, from your list above. Ultimately it should translate into better
> numbers out here in projects like ours but I'm not sure improving presence
> in open source is a captured metric. Let me ask.
>
> St.Ack
>
> 1.
>
> http://www.motherjones.com/politics/2015/05/silicon-valley-gender-sexism-women-engineers-tracy-chou
> 2. https://github.com/triketora/women-in-software-eng
>
>
> > On Sat, May 14, 2016 at 12:39 AM, Sean Busbey <busbey@apache.org> wrote:
> >
> > > On Thu, May 12, 2016 at 10:10 PM, Stack <stack@duboce.net> wrote:
> > > > On Thu, May 12, 2016 at 8:37 PM, Daniel Vimont <daniel@commonvox.org
> >
> > > wrote:
> > > >
> > > >
> > > >
> > > >> With regard to recruiting new contributors, I'll just toss out a
> > > question
> > > >> which I hope does not give offense: What can be done to encourage
> more
> > > >> gender-diversity in this project (and projects like it)?
> > > >>
> > > >>
> > > > Thank you for raising this 'awkward' question up here on the dev
> list.
> > > >
> > > > Our ratio is awful. I'm ashamed to cite numbers.
> > > >
> > > > I don't know what we can do to encourage participation. I'd be
> > interested
> > > > in any ideas others might have and would be up for acting them to try
> > and
> > > > make redress.
> > > >
> > >
> > >
> > > Yes, thank you Daniel. This is an issue that weighs on me. The ASF in
> > > general has a demographic problem, and our particular corner of it
> > > seems particularly homogenous.
> > >
> > > I'm a firm believer in "can't fix what you don't measure." Stack do
> > > you have numbers from somewhere? I was looking at ways we could run an
> > > opt-in poll to get an idea of how our community looks outside of what
> > > I can already see in the committer  and PMC ranks from meetups.
> > >
> > > I don't think there are any fast answers to this issue, but I think
> > > there are some things we could try doing that would help the project
> > > generally:
> > >
> > > 1) We could use better student outreach. Presuming we come up with
> > > some materials for trying to get students involved overall, we should
> > > get a few folks go out of our way to present those materials to
> > > student groups that try to provide space for folks who aren't in our
> > > dominant demographic.
> > >
> > > 2) Similar to #1, there are meetups (at least in my area) that try to
> > > make things accessible and comfortable for e.g. women. We could make a
> > > habit of presenting to these meetups in addition to our normal "big
> > > data" themed groups.
> > >
> > > 3) I've noticed that none of the meetups or conferences where I see
> > > HBase stuff have child care options. This doesn't only impact women,
> > > but it disproportionately impacts them due to societal expectations.
> > > We the community could start pushing folks to have something and we
> > > the PMC could perhaps push this a little harder, like we do the need
> > > for a Code of Conduct.
> > >
> >
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message