community-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Raul Kripalani <ra...@apache.org>
Subject Re: Passion and vigilance in open source
Date Wed, 23 Sep 2015 17:55:36 GMT
To me, Github and the ASF are not comparable. GH is a service to facilitate
collaboration in the OSS world (with a private offering for managing
private code too); whereas the ASF is an organisation that promotes,
protects and mentors OSS projects.

One of our prime drivers is Community over Code, whereas I'd argue that
Github's priorities are the opposite: Code over Community. Just get your
code out there and see if it gains traction. Moreover, we have a principle
of meritocracy which is not even modelled at Github. Open Source
traditionally being a voluntary commitment, I would argue that a
meritocracy policy tends to drive people to get involved and absorbed into
the fun ;-)

About the organisations publishing on Github or on Apache... Companies
publishing to Github prioritise building a user base (not necessarily a
community), whilst *retaining their control on the project*. Conversely,
organisations contributing their projects to the ASF are actually
*donating* their codebase with the expectation of building a solid and
vibrant community around it – that's where the fun is IMHO.

In fact – for the sake of illustrating my thoughts – I *could* envision
part of the ASF's daily business/processes running on top of Github, why
not? Code management, issue management, knowledge management, mainly.
Obviously we'd lose a lot of our delicacy and uniqueness if we do that ;-)

The ASF has a prestige, whereas – like Alex said GH – can be thought of as
a marketplace: each project has its own reputation. In fact, the dev
communities I've dealt with tend to regard the ASF highly as a body;
whereas you can't ask them for an opinion about GH as a whole, you'd have
to ask for specific projects (unless you're asking about GH as a product).

The public recognises that being an ASF projects implies having passed many
filters, votes and approval of talented people. And they appreciate that.

*Raúl Kripalani*
PMC & Committer @ Apache Ignite, Apache Camel | Integration, Big Data and
Messaging Engineer
http://about.me/raulkripalani | http://www.linkedin.com/in/raulkripalani
http://blog.raulkr.net | twitter: @raulvk

On Wed, Sep 23, 2015 at 6:17 PM, Alex Harui <aharui@adobe.com> wrote:

> IMO, I believe history does repeat itself, and that there are pendulums
> and cycles, but I think the dynamic here is more linear.  Now I certainly
> haven’t been involved with open source and Apache for very long, but
> having met Roy and heard him describe why Apache was founded and how it
> was supposed to be like a potluck/party, I would say that JimJag is just
> witnessing the “natural” progression of large organizations in the US.
>
> I’ve seen other organizations have to become “more serious” as the
> customer base grows and the stakes get higher.  You sort of have to, in
> order to attempt to propagate the vision and messaging to more and more
> people.  The famous party games of “telephone” where one person tells
> something to another person and by the time the story has been relayed
> several times the story has changed is a true human dynamic.  So more
> process is put in place, things take longer because they have to be
> written and reviewed, etc.
>
> IIRC, there are people in the world who are serial entrepreneurs.  They
> start a company, it grows to a certain point, then they decide (or their
> board helps them decide) that it isn’t fun any more and they leave and
> start another company.  I know folks who are serial code project starters.
>  They take a good idea, develop the prototype, get folks excited, a team
> is formed, version 1.0 ships, and then during the drudgery work of making
> the product mature, they take off for another new idea.  They don’t find
> it fun to fix all of the bugs folks expect to be fixed for version 2.0.
>
> So yeah, being at Apache now probably isn’t nearly the fun it was many
> years ago when trying to create a legal entity to protect open source was
> new and different.  We now know it is possible and are now trying to fix
> all the small bugs.
>
> And meanwhile, while your organization matures, some other person starts
> up a similar company with a different angle and folks find it new and
> different and fun compared to your now-slower process machinery and people
> flock to the new and different thing.  There may be aspects of that new
> and different variant that will make it the winner for the next
> generation, or it may be that they will inevitably have to implement the
> same process machinery.  The legal system in the US tends to make that
> happen.  So does the US insurance industry.  Lots of young single people
> don’t buy health insurance in the US, then they get married and have kids
> (i.e., their organization grows and matures and the stakes get higher) and
> then they start buying insurance.
>
> So, I don’t know that Apache can ever swing back to being young and fast,
> there is just too many people and too much at stake.  But IMO, any project
> within GitHub that becomes as important as some of the ASF projects will
> have to go down this same road.
>
> To me, the question about ASF’s attractiveness vs GitHub is all about
> whether GitHub has something that will still shine through this additional
> process that their high profile projects with US corporate customers will
> have to take on at some point.  That’s probably worth a separate thread,
> and I don’t spend any time on GH to know for myself, but I think I’ve seen
> folks say that their UX is better integrated, which is something we could
> work on here at the ASF.  I posted a thread a couple of days ago about
> Apache FlexJS being used to build a more integrated UX for ASF projects.
>
> So, sorry Jim, the ASF may never be as much fun as it was, but we should
> definitely be discussing whether we need to be more popular to our target
> customers and how to go about doing that and whether we can learn anything
> from GH or other organizations.
>
> A higher-up executive at Starbucks once told me that he was taught to “try
> to imagine what the younger guy/gal competing for his job would do.”  Is
> GH that younger guy/gal?  GH could just be a flea market.  Corporations
> rarely shop at flea markets.  On occasion a vendor refines their product
> at a flea market and starts a million dollar company and leaves the flea
> market to try to attract corporate customers.  What can the ASF do to make
> that vendor want to come to the ASF at that point?  What are the barriers?
>
> -Alex
>
> On 9/23/15, 8:12 AM, "Alexei Fedotov" <alexei.fedotov@gmail.com> wrote:
>
> >... and thanks to the topic starter for the great topic.
> >23.09.2015 17:48 пользователь "Jim Jagielski" <jim@jagunet.com>
написал:
> >
> >> Yeah, that's pretty much the way I'm looking at it. To me though, just
> >> as gravity is what pulls the pendulum back down to its mid-point, for
> >> Open Source, it's the "true Open Source" community (or, if you prefer,
> >> the "real" one) which acts as gravity, and pulls the pendulum back
> >> to 'b'. But if that real community doesn't exist, then the pendulum
> >> never swings back.
> >>
> >> As long as we are talking mechanical analogies, one I like to use isn't
> >> the pendulum but rather the fly-ball governor on a Watt steam engine[1].
> >> In this case, when the internal temperature gets too high, the spin of
> >> the governor speeds up, which releases steam and the temp goes down;
> >>when
> >> the internal temperature goes down too low, the spin is slower, and it
> >> closes a valve which increases pressure and temp... In our case, the
> >> meritocratic open source governance model is the base operating mode (b)
> >> whereas a and c are the 2 extremes. Where I see a problem is when (b)
> >> is no longer considered the "right" or "optimal" mode, and instead
> >> the default "regulated" mode is set closer to 'a' or 'c'; My point is
> >>that
> >> this set-point can, and *is* controlled but all the players in the
> >> open source community, but we are in "danger" of 'b' no longer
> >> being the desired mode simply because those who favor 'b' are no
> >> longer active in wanting that... We need to ensure that 'b' being
> >> the correct/right/optimal set-point for "how open source should be"
> >> is always being "pushed", always being fostered, always being nurtured.
> >>
> >> BTW: Thx for all the comments, this is VERY VERY useful!
> >>
> >>
> >> 1. https://en.wikipedia.org/wiki/Centrifugal_governor
> >> > On Sep 23, 2015, at 2:27 AM, Ross Gardler <Ross.Gardler@microsoft.com
> >
> >> wrote:
> >> >
> >> > I reckon Jim is describing a different kind of pendulum (see my
> >>earlier
> >> essay - sorry I got on a roll with that one).
> >> >
> >> > Jim's pendulum is something like:
> >> >
> >> > Let a = autocratic open source governance (vendor owned/benevolent
> >> dictator)
> >> > Let b = meritocratic open source governance
> >> > Let c = fully distributed open source governance (GitHub style fork
> >>and
> >> forget - note not all GitHub projects are this style)
> >> >
> >> > The interesting thing is that I don't think we are really at point c,
> >>I
> >> think we are really at point a. The numbers point to c but many
> >>rock-star
> >> projects are at point a. I'd argue that this goes hand in hand with my
> >> argument that open source is currently more about the business model
> >>than
> >> the development model. As with the other pendulum I believe this one
> >>will
> >> swing back towards the center as those companies realize that there is a
> >> glass ceiling to their growth using that model (if you haven't read
> >>Henrik
> >> Ingo's paper [1] on this you should).
> >> >
> >> > Another interesting point about this spectrum is that while (if
> >>history
> >> repeats) there will be a swing past b and towards c this side of the
> >>swing
> >> is much shorter. I guess because any "fork and forget" projects that
> >> succeed will typically become either an autocratic or meritocratic
> >>project
> >> in order to scale.
> >> >
> >> > As with my other pendulum thought experiment I believe we sit at the
> >> "sensible" place on that spectrum (point b). That isn't today it's the
> >>only
> >> place that can work, but that it is where it works for the Apache Way. I
> >> think plenty of people still do this for the fun (and education).
> >>Speaking
> >> personally a recent change in my dayjob role means that I'm coding for
> >>fun
> >> again - so that's at least one person going in the opposite direction to
> >> the one Jim sees is the majority (lucky me!)
> >> >
> >> > Ross
> >> >
> >> > [1]
> >>
> >>
> http://openlife.cc/blogs/2010/november/how-grow-your-open-source-project-
> >>10x-and-revenues-5x
> >> >
> >> >
> >> > -----Original Message-----
> >> > From: Jim Jagielski [mailto:jim@jaguNET.com]
> >> > Sent: Tuesday, September 22, 2015 7:01 PM
> >> > To: dev@community.apache.org
> >> > Subject: Re: Passion and vigilance in open source
> >> >
> >> > I would be OK with us getting older and forgetting the child-like
> >>wonder
> >> (but I don't think that's the case; well, we *are* getting older, but
> >>not
> >> forgetting the wonder), IF we were seeing the child-like wonder being
> >> continued, esp by the next gen.
> >> >
> >> > Some see Github as "proof" that the wonder is still there; even if so,
> >> then it's a different kind of 'wonder' and one which is risky for the
> >> continuation of open source.
> >> >
> >> > Wonder is not being able to fork a project, make some patches, submit
> >>a
> >> bunch of pull requests and then get a handful of them committed
> >>upstream...
> >> That is so.... solitary. The wonder is working *with* and collaborating
> >> *with* and reaching consensus
> >> > *with* a group of similarly-minded individuals towards a common goal.
> >> The wonder is the community. And I think that that is something which
> >>is at
> >> risk.
> >> >
> >> > To me, Open Source provided an avenue that allowed coders (and other
> >> contributors) to finally work together, openly and honestly,
> >>transparently
> >> and meritocractically (if you get my meaning); it fostered sharing, but
> >>not
> >> by letting someone share our toys by playing with them by themselves in
> >> some corner of the sandbox. It was about us all sharing the toys to
> >>build a
> >> great sand castle all together in that sandbox, when before we couldn't.
> >> >
> >> > Are people doing it for fun? Are people seeing the joy and wonder in
> >>our
> >> eyes? Or are people doing it just because "that's what I get paid to
> >>do"?
> >> >
> >> > Good questions. Not simple answers :)
> >> >
> >> >> On Sep 22, 2015, at 4:35 PM, Ted Dunning <ted.dunning@gmail.com>
> >>wrote:
> >> >>
> >> >> Jim,
> >> >>
> >> >> Is that really happening?  Is the fun leaving?  Or is it we are all
> >> >> just getting old and are forgetting the child-like wonder?
> >> >>
> >> >>
> >> >>
> >> >> On Tue, Sep 22, 2015 at 12:58 PM, Jim Jagielski <jim@jagunet.com>
> >> wrote:
> >> >>
> >> >>> Some of you may know that I've started a Vlog series on Youtube
> >> >>> around some topics I find interesting, mostly around open source.
> >> >>>
> >> >>> My latest is about the risks around open source today where the
fun
> >> >>> and passion that used to exist around open source is drying up
or
> >> >>> being discounted. Since Apache is one of the still remaining oasis
> >>of
> >> >>> open source being all about community and fun whilst still changing
> >> >>> the world, I'd like to ask for some thoughts from the membership
> >> >>> about their concerns, etc... that I can fold into the 2nd part
of
> >> >>> this mini-series.
> >> >>>
> >> >>> If so, please contact me directly. I have set the Reply-To header
> >> >>> accordingly.
> >> >>>
> >> >>> Thx!
> >> >>>
> >> >
> >>
> >>
>
>

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