User-agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:78.0) Gecko/20100101 Thunderbird/78.4.1
Hi David,
I said it on the Spatial4J list and I'll say it here... I'm
comfortable with Spatial4J bumping up its minimum Java version to
8. (Basically, I'm +1 on it if you feel that you need a second
round of approval from someone else!)
Cheers,
Jim
On 11/19/2020 9:58 AM, Wayne Beaton
wrote:
I defer to the wisdom of the PMC.
The EMO has no absolute rules other than that project teams
should have some sort of ramp-down policy/plan. Ideally, that
ramp-down policy/plan should be documented, but the EMO
doesn't rigorously enforce this (some PMCs do). For some
projects, this means feature freeze some number of weeks in
advance of the release; other projects allow themselves last
minute changes.
Ultimately, it's about not breaking your communities. Last
minute changes run the risk of introducing entropy that might
adversely impact user experience, or break adopter products.
HTH,
Wayne
On Thu, Nov 19, 2020 at 8:56
AM David Smiley <dsmiley@xxxxxxxxxx> wrote:
Thanks Wayne & Jim!
I see in Bugzilla that the release review was
approved. I filled out the release date to be about one
week after the release review, and thus it's on Thursday
the 26th.
I observe that the release process allows contributions
that still affect the release (FAQ entry #8). From my
involvement on a Apache Software Foundation project, we
don't do that -- it'd necessitate a new vote. I confess
I'm tempted to exercise this right as a contributor to
bump the minimum Java version from 1.7 to 1.8 to take
advantage of default methods on interfaces (amongst other
things). That would help Spatial4j's primary consumer
(Lucene). Yet from my years of ASF experience, it feels
wrong to "shove this in" (wearing my ASF hat/perspective)
post-vote. WDYT? I will suggest this on the Spatial4j
dev-list ASAP. Even if it doesn't happen for 0.8, it
could follow thereafter.
On Wed, Nov 18, 2020 at
6:06 PM Jim Hughes <jhughes@xxxxxxxx>
wrote:
Hi all,
I'm also +1 on the Spatial4J release!
Cheers,
Jim
On 11/17/20 11:12 PM, Wayne Beaton wrote:
I'll review the language in the
handbook to ensure that it is clear that a project
committer needs to use the PMC mailing list to
request (and receive) PMC approval.
In terms of timing, the process says that we're
supposed to have that PMC approval before we start
the one-week period of community review. The EMO
generally waits until all of the requirements (the
IP Log and PMC approval) are in hand before
scheduling the review. Lately, however, we've been
taking a bit more of a relaxed approach of
optimistically scheduling the review and accepting
PMC approval before we declare success on that
review. Note that PMC approval and IP Log
review/approval are separate/independent things
that can be executed in parallel.
Whether or not a PMC member can approve their
own requests is left to the PMC to decide. The EMO
will accept a +1 from any PMC member; but tends to
wait before recording the +1 for a bit to give
other PMC members a chance to argue (since
arguments occur rarely, we've found that the "EMO
dragging its feet a bit" method is lowest impact
on everybody; i.e. way easier than calling a
formal vote).
FWIW, the PMC +1 requirement serves a few
purposes. The EMO depends on the PMC to ensure
that the project is operating in an open,
transparent, and meritocratic manner as defined by
the EDP's Open Source Rules of Engagement. The PMC
is closer to the project than the EMO, and so is
better positioned to know whether or not the
release is within the bounds of the project's
scope and operating in accordance with the mission
and scope of the top-level project's charter.
Further, the PMC, being closer to the project than
the EMO, is better positioned to know if there are
"shenanigans" (related to intellectual property;
operating in an open, transparent, and vendor
neutral manner; etc.) that the EMO needs to know
about. Finally, the PMC approval requirement,
makes sure that at least one committer from the
project is subscribed to the PMC mailing list and
able to represent the project's interests in
conversations that occur on that list.
Anyway... the EMO will accept this note as
a +1... first thing tomorrow morning.
Regarding "Type A or B"... we changed the IP
Policy in late 2019 to remove the notion of Type
B. All third party due diligence is now license
certification only (what we had formally referred
to as "Type A"). The field needs to be removed.
I've opened a long-overdue issue to take care of
this.
Wayne
On Tue, Nov 17,
2020 at 5:04 PM David Smiley <dsmiley@xxxxxxxxxx>
wrote:
I reviewed the Eclipse Project
Handbook, and I *think* I'm supposed to ask
for an approval by the PMC:
The
review materials must be approved by
the PMC; the project lead or a
designated project committer should
send an email to the PMC’s mailing
list with a request for approval. The
PMC will respond with feedback or a
simple +1 indicating
approval.
Here's my +1.
It's a little unclear as to the
relative timing of this relative to the
release review, which is scheduled for
tomorrow (Wednesday).
I've gone about deploying Spatial4j
0.8 into Apache Lucene/Solr master to
run tests to see that it works. I can
see some additional interface methods
cause the need to write some trivial
convenience methods (Spatial4j is not
yet on Java 8, so doesn't benefit from
default methods) but most Spatial4j
users wouldn't implement their own
implementations of these interfaces any
way.
On Fri, Nov
13, 2020 at 9:02 AM David Smiley <dsmiley@xxxxxxxxxx>
wrote:
Hello LocationTech PMC,
I know you haven't
heard from me in a long while.
Spatial4j is in maintenance mode but
it's time now to do a release. It's
been 3 years since I last did this
so please bear with me if I screw
up. The release record is here:
The IP Log was
submitted and approved. The release
records contain the list of
changes. When the form asked me for
Type A vs Type B of IP diligence, I
chose "A" because there are no
entirely new dependencies, only some
version bumps. Furthermore,
Spatial4j only has works-with and
build dependencies; nothing
fundamentally required.
The release review is
on the 18th, a Wednesday.
If you are using Spatial4j, then
you can try out v0.8 SNAPSHOT here: