labs-labs mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Shane Curcuru <>
Subject Re: Releases
Date Tue, 14 Mar 2017 21:52:23 GMT
(Not subscribed, reading on lists.a.o)

Excellent discussion, and kudos to Danny for 1) starting it, and 2)
keeping discussions going here.

One problem that seems easier to solve is awareness.  Few people know
about labs, few people talk about them, and they're not really easy for
newcomers to find.  So any sort of blogs, tweets, and updates to places
like the community.a.o website that help point to and explain labs are a
good start to at least see if newcomers are interested.

The other big value I think Labs could have is awareness for individual
labs.  Github is certainly easy for hosting projects; but it's easy to
get lost or overlooked on github.  If there were some way that Labs
could make the repos committers can ask for be more visible, that might
help.  I.e. not just "some random github repo from a random coder", but
"here's an experimental project by an Apache Committer from projects X,
Y, Z".  No idea how to do this though.

On 2017-03-08 10:27 (-0400), Danny Angus <> wrote: > Hi
> I would love to be the kind of guy who would set out a series of topics
> that we need to discuss, contextualise them and initiate each one,
> collating the output at the end, but I'm not.
> So I have a question for the people who have mentioned lack of releases as
> a problem..
> The idea behind "no releases" was that a release implies that certain
> standards common to the ASF are being met, standards which Labs don't
> apply, and that if a lab needs to cut a release it probably has enough
> users and maintainers to become a project of its own.

The policy on Apache projects making software releases is clear, if a
little wordy and finicky:

This is important because of the phrase 'an official Apache release is
one which has been endorsed as an "act of the Foundation" by a PMC.'
Documenting this process - and having the board ensure that PMCs follow
it - is a necessary step to ensure that software releases used by other
people are acts of the ASF as an organization - and not an act of the
release manager as an individual.

This would become important if someone $evil ever wanted to sue the
release manager; extremely rare, but if it ever happens, in theory they
could go after your house, your car, your savings.  But if a PMC follows
the policy, the $evil person would have to sue the ASF, which has
insurance, lawyers, and could shield the individual release manager or
developers on the project from the lawsuit.

Labs could create finished software packages and make them available, as
long as it's crystal clear to any possible user (and their lawyers) that
the software package isn't an "Apache release".

Brainstorm: one concept might be to have the Labs act like a PMC, and
when someone wants to do a release, if they can get two other Lab PMC
members to vote on it (by downloading, checking, and testing it), that
could qualify as an Apache release.  If there's a lot of interest,
that's a possibility to look into.  But requires each lab make it easy
for other PMC members to vote +1 on their code.

> So my question to those people is this; Is it the lack of the formality and
> assurance that is a problem or the lack of a build? Would "pre-release
> downloads" be enough?
> d.


- Shane

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

View raw message