Mailing-List: contact labs-help@labs.apache.org; run by ezmlm
Precedence: bulk
Reply-To: labs@labs.apache.org
From: Wade Chandler <wadechandler@apache.org>
Content-Type: multipart/alternative; boundary="Apple-Mail=_3E426442-FFAA-4C98-8B01-832B85310E90"
Message-Id: <35D36EF8-7A0F-456F-AD23-CA1FF53614B7@apache.org>
Mime-Version: 1.0 (Mac OS X Mail 9.3 \(3124\))
Subject: Re: On Labs identity (Blue Sky thoughts)
Date: Tue, 7 Mar 2017 20:54:46 -0500
References: <CAEFAe-HsxdzRCRbynd8yEKLKH9Lhqo7B=eZ91rg=JRmxcJ-JiA@mail.gmail.com>
To: labs@labs.apache.org
In-Reply-To: <CAEFAe-HsxdzRCRbynd8yEKLKH9Lhqo7B=eZ91rg=JRmxcJ-JiA@mail.gmail.com>
archived-at: Wed, 08 Mar 2017 13:39:43 -0000

--Apple-Mail=_3E426442-FFAA-4C98-8B01-832B85310E90
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=utf-8

> On Mar 7, 2017, at 16:26, Alexandre Rafalovitch <arafalov@gmail.com> =
wrote:
>=20
> (Just joined due to the Committer list's call to action).
>=20

Yes, I must say, I had not noticed the =E2=80=9CLabs=E2=80=9D link under =
=E2=80=9CInnovation=E2=80=9D at the bottom of the home page before, nor =
did I know about it. I am curious in success stories (or even past =
feedback on it), and too, how is this different than a current project =
possibly making a separate repository for something, and then later =
branching it off into a separate project at ASF?

> I think the question of Lab's identity is where it fits in the
> _modern_ ecosystem. Its original offering (VM, etc) as mentioned on
> some of the messages in the archive is not sustainable.
>=20
> Github/Gitlab is more visible, offers more control and is integrated
> more with additional awesome services, never mind being a
> social-development destination.

Would what you are saying here differ or does it differ from the same =
concept of repositories and mirror on GitHub as the rest of Apache? I =
think GitHub is certainly a perfect place to host repos, but it seems we =
do that now with mirroring etc, but we can=E2=80=99t just create repos =
there easily.

> VMs are kind of cheap. I am putting a
> lot more than $5/month into my Open Source Project of Choice. Running
> a $5/month VM is just one of those things. And setting up VM with Labs
> is probably as - or more - difficult than doing it on Digital Ocean or
> whatever.
>=20

Maybe this is a different topic if we are talking about infrastructure. =
What infra would be provided or should be provided by Apache; even =
Apache Labs? I know there are some things that can be provided, but my =
understanding is internally that is fairly limited to some build tools, =
and occasionally project specific items such as a system to run a plugin =
portal or something specific to a project (negotiated or otherwise). I =
may be missing something, and will (re-)read the labs pages.

> So, is Labs then a sub-project level of offering to Apache Committers?
> Something that is smaller than a project but is still somehow
> "Apache"? Again, that does not seem to be enough of an incentive for
> technical reasons.
>=20
> What about for the community reasons? Can the identity be reformulated
> (or reinforced) to be a lot more about community than about technical
> offering? In fact, reading the By-laws, I get all warm and fuzzy and
> excited, especially about the last part of the last para:
> "Apache Labs are the place where ASF committers can work on
> innovative, blue-sky and off-the-wall ideas, without having to worry
> about fitting in an existing project bylaw or building a community
> around it, but unlike other external venues that can offer similar
> hosting services, as a place where fellow committers can offer
> suggestions and help.=E2=80=9D
>=20

To me that suggests the idea is to be lighter than incubator or dev =
proper, and to then allow committers a space to play in Apache, and get =
others involved, and then to possibly move items to the incubator and =
beyond, or not even. I like this interpretation a lot. People and =
committers can see what we are doing or possibly thinking about doing, =
and let=E2=80=99s ideas flow without having to have consensus first, but =
still has association with the organization.

> So, if _I_ was interpreting these ByLaws, I would sort of focus them
> inwards on a lot more dogfooding _across_ Apache projects with Labs
> being that inner space where different communities can meet. Ideally,
> it would give growth to tiger teams which brings together people from
> different projects who then, as a group, help out other Apache
> projects.
>=20
> To take a semi-technical example, it would be great to have somebody
> who has Technical Documentation skills, somebody who is good sysadmin
> and somebody who has good technical/demo building skills and them
> going around multiple Apache projects and helping them to build their
> "hello world" examples that make it easier for the beginners to get
> going. I bet after 3-5 runs through "take a project, build it, make a
> basic dockerized example for it, push the examples and bugs discovered
> back to the community", it will become an easy thing to do well. And
> most communities would love it. And Labs could be the space that sets
> up an infrastructure to help those people do it. Even if that
> infrastructure is not actually running on Apache Infra.
>=20

I=E2=80=99m not sure that would be labs per my view point, but it =
certainly sounds like something that could be useful in general =
especially for esoteric or more complex projects, but I guess if that is =
what someone wanted to use =E2=80=9Clabs space=E2=80=9D for, it fits the =
=E2=80=9Cit=E2=80=99s light, do what you want to do=E2=80=9D model or =
vision. I see it more of a place where I =E2=80=9Cwant=E2=80=9D to do =
something, and even if others don=E2=80=99t yet think it =E2=80=9Cis=E2=80=
=9D or =E2=80=9Cshould=E2=80=9D be an Apache project, I can at least =
exercise it, and possibly change some minds by way of real code and =
demonstration. So, I guess that could be the above too for some people.

> Labs could also be the one with the strong Call to Action. I have
> search expertise for example (Apache Solr). I do a number of
> mini-projects around Solr that are not direct code contribution. I
> would be happy to step into other projects that are downstream from
> Solr and provide my Solr expertise to their committers by reviewing
> their setup or helping to explain to them why they have troubles
> migrating to latest Solr. But they don't really call and if they do
> call, most of the time, you are then expected to find your own path
> from the start to the level of contribution. If there was an umbrella
> of people helping to mentor people already in ASF into helping them to
> cross-contribute to other projects, this could be a lot easier. Labs
> could be one collecting such willing experts and connecting/enabling
> them.
>=20
> Could Labs be the place where people with different first language
> hang out and help to translate "Hello World" tutorials into other
> languages, using common infrastructure they don't have to reinvent.
> Discovery of educational resources for open-source projects is
> actually a big issue, giving a centralized space (basically a
> well-referenced blog) is a step in a right direction. Again, this
> could happen spontaneously by just individual efforts, but having a
> coordinated effort in removing all barriers towards that happening
> consistently and in repeated fashion, could encourage contributions.
>=20
> We already have http://community.apache.org/, which focuses on less
> technical aspects and on on-boarding (for those who find it, similar
> problem really....). Could Labs be complimentary to that, by focusing
> on technical aspects of building the community and focusing a lot more
> on getting more senior members of community to contribute back in in
> the mentoring ways?
>=20

More of the same as what I wrote before for me, but I definitely don=E2=80=
=99t think the above would be labs. To me labs should be something like:
https://github.com/awslabs <https://github.com/awslabs>

Which are possibly ways of using technology or showing how to do it or =
totally concept projects which are not yet even trying to become proper =
projects where the cost of entry as a committer is super low simply =
because we are committers, and what we do there may even be throw away =
code. I wouldn=E2=80=99t choose for labs to be about mentoring etc. To =
me that isn=E2=80=99t innovation or experimental, but is more in line =
with learning the Apache way or something else depending on where that =
is going, and to me just seems like =E2=80=9Csomething else=E2=80=9D =
from my point of view, but that=E2=80=99s just me.

Thanks,

Wade=

--Apple-Mail=_3E426442-FFAA-4C98-8B01-832B85310E90--