incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Greg Stein <>
Subject Re: Starting from the other end
Date Tue, 20 Oct 2015 05:33:56 GMT
On Mon, Oct 19, 2015 at 11:22 PM, Marvin Humphrey <>

> The Maturity Model defines evaluation criteria.  It is probably the most
> useful for the task we're focused on now, which is how to decide when a
> proposed TLP is ready, regardless of the path it took towards readiness.

As I've mentioned else-thread, that criteria has had no impact on the
Board's evaluation of incoming direct-to-TLP projects. Instead, the Board
took faith that *while a TLP* the community would reform itself to the
desired goal. There was no requirement for the TLP to be fully compatible
on Day One, but simply that the TLP knew *what* was needed, and *how* to
reach that goal.

I don't know about Apache Zest, but the (now) Apache Serf project always
operated itself very closely to an ASF community. But, as an example, we
did not use KEYS to sign releases. As a responsible TLP, we know the
various gaps, and are making the appropriate changes as we fit ourselves
into the ASF and produce our first ASF-branded release.

This is why I suggest to avoid comparing Incubation projects against
direct-to-TLP. By most measures that I can think of, they are quite


ps. it could be argued that a podling can become a TLP once it was
self-aware enough to know how to fix/mend the gaps, much like Serf is
doing. As Sam noted else-thread, this boils down to trust of those in the
PMC. Serf gets it via a 10 ASF Members on Day One; a podling has a very
high bar to reach comparative trust. [from the Board's perspective, IMO]

> The incubation checklist is what we use to track a podling's incremental
> progress through the Incubator.  You could see it as an Incubator-specific
> concrete realization of the Maturity Model, though as it is implemented
> today
> it only covers a fraction of the Model's bullet points.
> The Project Requirements document encompasses only policy, and thus only
> covers a subset of the curriculum.  However, because the Foundation's
> current
> policy documentation is unbounded, irregular, and bloated, mastering this
> part
> of the curriculum requires disproportionate effort.
> I believe that if we can streamline all aspects of the curriculum,
> incubation
> can take far less time and effort than it does today.  And if we can reduce
> the average time-to-incubate, many problems will simply solve themselves.
> Marvin Humphrey
> [0] The "curriculum" originally referred to the knowledge gained by a
> podling
>     release manager while guiding a release candidate through the
> Incubator.
> [1]
> [2]
> [3]
> ---------------------------------------------------------------------
> To unsubscribe, e-mail:
> For additional commands, e-mail:

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