community-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Myrle Krantz <>
Subject Re: Why are large code drops damaging to a community?
Date Thu, 25 Oct 2018 08:56:34 GMT
Thank you for sharing your experiences with offlist development Alex.

Even single-developer large code drops can be damaging to the
community in some cases.  Malcolm Upayavira gave a good example.  But
single developers may find it easier to remedy the situation in the
same way Upayavira did: by cutting the code up into reviewable-sized
pieces, and courageously accepting community feedback to improve the
code.  Responding constructively to criticism is how we grow, right?
By developing alone for a long period, we deprive ourselves of that
criticism.  Criticism outside of our in-group is all the better
because it raises issues we wouldn't discover on our own.  That's one
of the major strengths of open source.

My goal here is *not* to find the exact shade of grey at which we
prohibit off-list development.  You're right, Alex, almost all
decisions on this topic should be made by the project community and
should be focused on the effect on that project community.  If people
who want to participate are being excluded by a particular development
practice (even if it's a cool hackathon at ApacheCon), that is
something that a community should at very least be aware of and be
ready to mitigate.  Each community can weigh downsides and upsides for

My goal here is to spark awareness, by helping to discover and
communicate the spectrum.  An inventory can be a diagnostic tool by
which communities can determine where they are on the spectrum and
decide whether they are comfortable with their current shade of grey.

Best Regards,

On Wed, Oct 24, 2018 at 6:42 PM Alex Harui <> wrote:
> This is just a nitpick, but it is the subject line that is bothering me.  Having a "no
large code drops" mantra is not the same as a "limit off-list collaboration" mantra which
is different from a "no off-list development" mantra.
> For Flex, Adobe made something like 5 large code drops.  It simply took so long to clear
the various chunks of code being donated to Apache through Adobe's approval process that it
was done in stages.  In fact, if I had time, the Flex/Royale community wanted it, and I could
get Adobe to support it, I'd bestow a couple of other large code drops to Apache.  It is all
pre-existing code, but again, the subject line makes it sound like any large code drop is
bad, which is not true.
> To say "no off-list development" could be construed as some limit on how many lines of
code you can write as an individual before making it available to others to review.
> I think the real key here is "off-list collaboration".  There will always be off-list
collaboration, and I'll bet some really significant things can happen at a hackathon.  That
should not be prohibited.  I think you are simply trying to express the notion that groups,
especially groups defined by having a common employer, don't do too much collaboration off-list
before inviting others into the conversation.
> My 2 cents,
> -Alex
> ´╗┐On 10/24/18, 7:17 AM, "Myrle Krantz" <> wrote:
>     Hey all,
>     I'd like to invite anyone with relevant positive or negative
>     experiences with off-list development or large code drops to share
>     those experiences.  The ASF policy of "no off-list development" is
>     implemented in a wide variety of ways in various communities, but
>     there are still may be some things that we can agree are absolute
>     no-goes.  I'd like to figure out what things we see as:
>     * You might get away with that once, but don't try it twice.  That's
>     damaging to the community.
>     * Avoid doing that unless you have a really good reason that the
>     community has accepted.
>     * That has a bit of a smell to it.  Have you discussed that with your community?
>     * That's fine.  You're helping your community.
>     * What a wonderful idea! Absolutely do that if you and your community want to!
>     I'm hoping to put together a diagnostic tool for offlist development
>     that can help communities help themselves.  Something similar to
>     Beck's Depression Inventory.  Because like mental health, community
>     health is complex, and sometimes it is not clearly 'good' or 'bad'.
>     In order to do that though, I'd like to read your stories.  Especially
>     from the people who've been around a few projects and seen a few
>     things.  The stories Malcolm, Chris, and Jim already shared are
>     exactly the sort of thing I'm looking for, so if y'all would like to
>     elaborate that'd be really cool too.
>     Best Regards,
>     Myrle

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

View raw message