►
From YouTube: 2020-06-23 Apache Cassandra Contributor Meeting
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
And
we
are
recording,
so
everything
you
say
can
and
will
be
used
against
you
in
a
court
of
public
opinion.
Is
your
Miranda
rights
of
zoom
recorded
meetings?
Well,
it's
been
a
month
and
there's
a
lot
of
activity
on
the
project.
There
was
a
couple
of
things
on
the
agenda.
If
I
bring
it
up,
bring
up
today's
actually
one
thing
on
the
agenda:
it
was
just
an
update
on
the
cassandra
kubernetes
sig.
I
can
give
that
real,
quick
that
that
has
gone
through
some
interesting
evolution
over
time.
A
A
Is
there's
a
different
language
on
how
kubernetes
does
things
it
was
better
to
tackle
it
in
a
kubernetes
kind
of
way,
and
the
way
that
you
make
changes
in
a
kubernetes
project
is
by
starting
with
the
CRD
and
the
cid
is
that
you
know
the
customer
resource
definition
and
thankfully
john-san
de
who's
got
a
ton
of
experience
with
kubernetes
and
cassandra.
He
used
to
work
a
red
hat
on
OpenShift
has
some
really
good
background.
A
He
said
I
will
I,
will
take
the
mantle
and
run
with
this
and
try
to
take
all
the
different
operators
out
there
and
do
more
or
less
like
a
greatest
hits,
mixtape
CRD
for
lack
of
a
better
word,
and
he
presented
that
last
week
at
our
meeting.
Here's
here's
a
common
CRD
with
like
some
of
the
things
that
you
know
like
the
best
of
and
then
maybe
some
changes
in
between
awesome.
This
is
where
the
best
part
of
it
is
now.
A
The
the
current
state
is
lots
of
really
good
discussions.
The
people
building
operators
are
showing
up
at
these
meetings
and
we're
having
some
good
like
more
technical
discussions
unless
less
you're
like
well.
Why
do
you
want
to
do
that
versus?
This
is
like
this
is
the
way
I
think
we
should
do
it,
and
then
we
debate
the
merits
that,
and
so
those
are
big
all
in
a
I
send
this
on
an
email.
This
is
all
on
a
github
account
right
now
this
week
we're
we
have
another
round
of
that.
A
We're
meeting
basically
every
week
to
go
through
this
trying
to
keep
the
pace
going,
but
I'm
cautiously
optimistic
that
we're
gonna
have
something
we
can
work
with
the
end
state
of
this
will
be
that
I
think
will
have
a
CR
D
that
is
different
from
all
the
other
operators
out
there
in,
but
also
one
that
expresses
some
of
the
best
common
thinking.
If
you
look
at
all
of
the
operators
that
are
currently
in
existence,
they've,
they
scratch
a
particular
itch,
but
they
don't
have
a
hundred
percent
coverage
as
much.
A
So
this
is
an
attempt
to
do
that.
You
know
the
I
think
the
feeling
is
for
the
community.
This
is
a
good
thing
where
it's
more
generalized
operator
and
not
so
specific.
This
is
one
of
the
mornings
we
got
from
Netflix
from
the
guys
at
Netflix
and
Joe,
even
named
when
they
were
talking
about
Priam
I.
A
Don't
know
if
you
remember
Priam,
but
that
was
a
super
opinionated
stack
and
if
you
wanted
to
use
Priam
and
you
had
to
drag
everything
in
with
it
so
trying
to
keep
from
that,
make
it
more
useful
for
the
general
populace
but
yeah.
So
that's
the
current
update.
If
you
want
to
learn
all
about
kubernetes
I've
learned
more
about
kubernetes
working
on
this
project
and
any
other
time.
B
A
Because
an
operator
an
operator
does
not
have
is
not
supposed
to
have
a
lot
of
smarts
about
like
internals
of
Cassandra.
It
knows
about
hey
service
that
we
want
to
use
with
Cassandra
or
they
we
want
to
use
with
kubernetes.
Tell
them
tell
me
about
your
state.
I
want
to
send
you
things,
I
want
to
change
your
state
and
do
it
in
a
kubernetes
kind
of
way.
A
The
api's,
you
know,
kubernetes
the
cloud
native
foundation
says
you:
will
thou
shalt
communicate
via
api's
HTTP,
api
s--,
and
so
that
you're
not
gonna,
be
you're,
not
gonna,
connect
an
operator
to
gmx.
Think
please!
No,
so
you
need
an
HTTP
api
and
sidecar
provides
that
there's
a
lot
of
smarts
built
into
a
sidecar
in
that
case.
So
in
order
for
the
kubernetes
cluster
to
maintain
a
certain
state,
let's
say
a
physical
know
goes
down.
How
do
you
manage
all
right?
So
what?
What
does
the
controller?
A
You
know
this
operator
gonna
do
at
this
point,
because
there's
two
state
managers
there's
the
Cassandras
managing
replication
state
in
its
own
way
and
then
there's
replication
controller
on
kubernetes,
just
like
having
a
mild
mild
panic.
Whenever
it
loses
a
node,
so
the
controller
has
to
be
the
go-between
and
it
has
to
be
able
to
control
what
happened
like
in
a
kubernetes
way.
A
Control
of
the
way
cassandra
wants
to
be
controlled
instead
of
just
throwing
up
a
new
node
and
saying
here's
a
brand
new
node,
sorry
about
the
other
one
that
broke,
and
then
you
probably
get
a
bootstrap
in
something
which
really
isn't
what
needed
to
happen.
That
site
card
is
really
important
to
have
all
that
business
logic
or
the
sorry
cluster
logic
on
the
region
of
closed-door
logic
in
the
cassandra
yeah.
B
Has
some
rough
edges
shall
we
say
in
terms
of
like
security
implications,
and
I
think
nobody,
injecting
things
and
I
suspect
that,
like
part
of
the
pain
that
we
run
into,
is
that
our
original
view
of
the
sidecar
as
a
project
was
a
API
intersection
point
for
operators
to
actually
run
clusters,
not
necessarily
for
kubernetes
integration
and
in
that
timeframe,
since
the
sidecar
was
originally
envisioned.
All
of
a
sudden
there's
just
this
massive
massive,
like
technical
market
pull
towards
kubernetes
is
just
being
like
the
de-facto
orchestration
mechanism
for
things
at
scale.
B
A
No
I,
this
is
a
is
a
far-reaching
project.
Like
I
said,
this
is
forcing
Cassandra
to
think
more.
The
Cassandra
community.
Do
you
think
more
about
kubernetes
in
a
lot
of
ways?
It's
not
just
it's
not
like
making
Cassandra
work
and
chef
or
ansible
or
even
terraform.
You
know
that
it
is
so
much
more
than
something.
That's
just
gonna
be
like
a
scripted
setup
and
teardown
tool.
A
B
B
C
A
Because
the
CR
DS
are
much
more
around
resource
management
like
setting
up
a
disk.
How
many
nodes
you
have
data
center
placement
is
actually
the
hardest
one
right
now,
because
kubernetes
doesn't
have
any
concept
of
multi
cluster.
It's
like
the
cluster.
Is
it
it's
an
island?
It
doesn't
communicate
with
the
outside
world
unless
you
have
ingress
set
up
so
tough
disk.
This
is
another
one
of
those
things
where
there's
a
kind
of
a
parity
mismatch.
So
a
thing
that
the
kubernetes
community
is
currently
discussing.
A
Quite
a
bit
is
multi
cluster
kubernetes,
but
in
the
world
of
Cassandra,
we've
been
doing
this
since
what
day
one
Multi
datacenter,
that's
like
the
most
basic
feature,
Cassandra
and
kubernetes-
will
turn
it
off.
If
we're
not
careful,
so
yeah,
we've
got
some
I
think
we've
got
some
work
there,
but
it's
exciting
because
it's
helping
us
modernize
a
lot
of
the
current
operations
that
need
to
be
done.
I
think
that's
everybody
seems
to
be
on
board
with
that.
Then
I've
talked
to
you
know
inside
that.
What
we're
talking
about
kubernetes,
it's
like!
Oh!
B
I
when
I
cud
car,
I
don't
really
have
a
strong
point
of
view
to
be,
to
be
honest,
like
this
is
something
that
like
jake
had
his
point
of
view
and
he
because
he
was
originally
envisioning.
This
whole
like
this.
This
UL
based
sidecar
thing
like
two
years
ago
and
I
know
you
guys
with
Reaper
we're
trying
to
get
things
integrated
with
the
sidecar.
Now
it's
like.
A
What
I
would
suggest
we
do
is
let's,
let's
get
to
a
happy
place
with
the
CR
D.
There
might
be
a
good
beginning
place,
because
that'll
show
I,
think
that
would
give
us
some
like
common
language
framework
to
work
from,
because,
if
I,
if
I
went
back
so
I
went
back
and
looked
at
some
of
the
original
discussions
around
sidecar
and
the
they
look.
A
Yes,
there's
a
yes,
but,
and
they
were
looking
very
basic
I
think
which
is
good
I
mean
it
modernized
some
part
of
it,
but
I
think
that
if
we
bring
kubernetes
into
this,
which
our
community
is
really
having
an
strong
pull
towards
right
now,
okay,
what
does
it
change
that
conversation?
Because
that
conversation
happened
over
a
year
ago?
You
know
the
original
sidecar
discussions
so
I
think
it's
worth
a
refresh
I'm.
B
Just
wondering
whether
that's
orthogonal
to
reaper's
needs
out
of
a
sidecar
and
I.
Don't
know
whether
or
not
Medusa
realized
in
a
sidecar
as
well,
and
if
we
start
talking
about
like
post
governance,
vote,
Reaper
and
Medusa
contribution
to
the
project
and
how
they
integrate
with
the
sidecar,
like
I,
could
barely
rapidly
open
Pandora's
box
for
all
the
things
we
have
as
a
project
that
are
rattling
around
and
in
the
relatively
near
future.
But
your
appointment
I
do
think.
B
B
There's
a
lot
of
things
in
a
Cassandra
that
11
12
years
ago
were
novel
and
nowadays
are
not
only
they're
novel
in
the
sense
that
we're
the
only
ones
are
the
implementation
we
have,
because
the
entire
rest
of
the
distributed
systems
world
has
settled
on
a
different
approach,
and
that
doesn't
necessarily
mean
that
we're
wrong
and
may
give
us
characteristics
as
a
project
that
are
very
uniquely
beneficial.
For
what
we're
doing,
but
anyway,
there's
there's
a
lot
of
these
kind
of
meta
discussions.
B
A
A
We
are
in
I
think
we're
in
high
danger
of
cramming
a
lot
of
logic
into
a
container
if
we're
not
careful
and
that
that
may
or
may
not
be
a
good
idea,
but
it's
worth
discussing
it's.
It
goes
beyond
the
container
just
running
a
binary
of
cassandra
with
all
of
the
ports
and
storage
externalized.
This
starts
putting
like
operational
logic
inside
the
container.
Whenever
we
start
adding
Sekhar.
A
B
A
F
F
A
D
A
F
I
do
think
that
we
should
think
a
little
bit
more
about
how
to
increase
the
involvement
in
these
meetings
again
thinking
about
where
they
stemmed
from
kind
of
in
those
in
person
they
area
ones
that
involve
like
a
bunch
of
users
and
different
companies,
etc.
I
feel,
like
we've
kind
of
dwindled
down
to
one
company
plus
one
or
two
other
people.
This
is
just
to
make
sure
we
continue
to
get
that
like
community
interaction
from
everybody,
I,
don't
I,
don't
know
necessarily
that
I
have
the
answers
to
it.
You
know,
but
moreso,
just
something.
B
F
I
think
part
of
it
was
that
those
earlier
meetings
had
a
little
bit
more
pre-planning
put
into
them
and
a
little
bit
more
of
that
outreach
where
we
would
kind
of
reach
out
to
people
and
be
like
hey
like
we
need
to
actually
do
something
with
this
time.
You
were
also
more
motivated
to
because
you
were
like
going
to
do
something
in
person
that
took
a
commute
together.
I
see
you,
it
was
like,
but
yeah
so
III
I
will
think
about
that
more.
F
It
was
just
something
I
I
noticed,
and
it
may
just
be
like
you
know,
myself
or
Patrick
or
someone.
You
know
like
pushing
a
little
bit
earlier
to
try
to
get
people
to
like
present
things
or
talk
about
things
etc.
But
to
go
back
to
your
question
like
I
feel
like
I
was
talking
in
person
is
useful
like
or
you
know,
digitally
like
voice.
A
B
Sorry
I
told
Patrick
earlier
today
or
yesterday,
like
Eileen
I,
don't
want
to
bring
up
any
topics:
I
don't
want
to
rock
the
boat
I
want
that
gun
to
go
through
and
so
I'm.
Just
like
stay
quiet,
I
know,
I
personally,
have
a
bunch
of
stuff
I'd
like
to
bring
up
like
a
view
of
the
state
of
our
testing
and
what
we're
gonna
do
and
where
Harry
fits
in
with
that,
where
fallout
fits
in
with
that
who's.
B
B
We
need
to
figure
out
how
to
scope
them
was
what
goes
inside
there
like
there's
all
kinds
of
stuff
that
we
could
be
using
this
time
to
really
to
really
dig
into
and
chew
on,
but
I've
also
been
trying
to
remain
very
mindful
of
the
fact
that
a
lot
of
the
contributors
in
the
project
aren't
paid
full
time
to
just
work
on
the
project
and
so
that
there's
a
massive
Delta
ingredient
on
you
know
time
available
to
commit
to
these
kinds
of
discussions
and
this
kind
of
work.
So
yeah.
F
And
I
think
there's
there's
the
balance
of
like
how
do
we
do
that
productively
without
like
making
people
feel
like
who
aren't
there
that
we're
making
decisions
like
outside
of
the
mailing
list
etc?
But
at
the
same
time,
like
a
forum
for
discussion
about
those
things
that
is
higher
fidelity
than
an
email
thread
is,
is
useful.
So
how
to
balance
that
is
hard.
F
A
A
B
F
Agreed
completely
I
mean
I,
think
it's
it's
really
hard,
sometimes
to
get
through
the
difficult
conversations
on
email,
but
yet
I
know
some
people
feel
strongly
that
like
they
should
eventually
and
at
least
end
on
email,
so
yeah
I
think
having
an
agenda
like
a
week
before
having
topics
like
that
available
and
then
probably
like.
We
have
to
play
with
time
zones
a
little
bit
for
some
people
so
that
we
can
include
like
the
European,
folks,
etc
and
I
know.
That's
gonna
probably
be
punishing
myself
and
most
people
on
this
call.
A
Do
it
well,
that
was
actually
one
of
the
things
we
sorted
out
early
was
the
rotation,
so
just
a
quick
reminder:
the
rotation
that
I
am
running
right
now,
it
ping
pongs
between
AIPAC
and
Europe
from
you
know
like
from
pain,
basically,
because
that
that's
generally
the
problem
we're
dealing
with,
we
have
everyone
from
Melbourne.
Australia
to
you,
know,
see
est.
So
we
that
is
a
massive
timezone
shift
to
deal
with,
so
it's
either
early
early
as
or
late
late.
A
For
you
know,
wherever
you
are
in
the
world,
if
you
live
in
the
West
Coast
United
States,
it's
great,
because
you're
cutting
in
the
middle
of
those
two
really
far-flung
time
zones.
But
what
we've?
What
we
learned?
We
tried
this
on
the
kubernetes
sig
was
we
tried
to
do
two
meetings,
one
that
was
friendly
to
one
timezone
and
then
another
meeting
that
was
friendly
to
another
time
zone
and
what
you
wound
up
getting
was
two
distinctly
different
stations,
but
never
intersected
so
that
didn't
work
out
so
good
I.
G
Was
thinking
about
the
agenda,
which
is
it
probably
could
we
do
something
different
like
having
in
mind
sections
which
are
kind
of
high-level
areas
that
are
important
on
the
project?
And
people
just
know
that
these
are
the
areas
and
then
everyone
can
come
up
with
the
exemption
agenda
like
seeing
them
and
saying
oh
yeah
by
the
way
in
that
scope,
I
have
something
to
add.
A
G
B
Operator,
what's
going
on,
yeah,
just
kind
of
a
quick
status
would
beyond
that.
For
instance,
like
I
know
that
Silvan
was
chewing
on
some
stuff,
with
consensus.
Algorithms
and
I
connected
him
with
some
people
that
I
knew
on
the
project
that
were
chewing
on
some
other
stuff,
and
it's
like
those
are
the
kind
of
things
that
be
great
for
us
to
talk
about
on
a
call
like
this,
like
is
anybody
thinking,
fast
packs
or
so
anything
good
raft?
Are
they
thinking
whatever
lamb
for
its
latest
crazy
concoction,
is
and
like?
B
Where
do
we
stand
as
a
project
on
those
things
and
riff
on
those
high
bandwidth
like
I,
think
there's
a
lot
of
opportunity
for
even
beyond
sub-project
check-ins,
like
actual
like
State
of
the
State
of
the
Union,
with
subsections
of
the
project
and
things
that
we're
thinking
about
where
we
could
take
it,
because
that's
one
thing
that
we
really
need
to
figure
out
to
your
point:
Ekaterina
is
like
okay.
So
what
does
4.1
look
like?
What
does
five
thought
it
look
like?
How
do
we
decide
that?
When
do
we
decide
that
you
know?
B
Where
did
where
and
one
of
those
discussions
gonna
happen
again.
I've
been
erring
on
the
side
of
don't
want
to
steal
oxygen
from
the
room
with
four,
oh,
but
I'm.
Also
seeing
that
there's
a
number
of
contributors
that
are
kind
of
bottle
necking,
which
is
why
there
was
the
mass
on
assigning
and
shifting
of
photo
quality
testing
stuff
in
the
last
couple
of
days.
But
yeah
we'll
have
to
figure
it
out.
F
B
And
I
I
mentioned
I
haven't
said
any
not
about
this
yet,
but
I
think
I
screwed
up
in
the
whole.
Let's
overload
assignee
is
Shepard
because
all
the
forward
quality
testing
tickets
look
like
they
have
an
assignee
and
there's
no
way
of
knowing
which
ones
are
or
not
being
worked
on,
and
so
there's
a
lot
of
those
that
are
like
blocked,
and
you
left
a
bunch
of
comments
and
I
think
March
5th
and
we
haven't
heard
anything
back
from
a
lot
of
those
tickets.
So
part
of
what
is
I
just
pulled
the
assignees
off.
B
Add
the
Shepard
to
the
description
of
the
ticket,
and
now
it's
very
clear,
like
here
stuff
that
no
one's
taken
we're
gonna,
try
to
broaden
the
pool
a
little
bit
on
both
who
is
shepherding
certain
areas
of
the
codebase,
because
there's
a
lot
of
people
that
know
a
lot
about
the
codebase
and
also
start
to
get
some
people
working
on
that,
because
what
I've
been
hearing
a
lot
from
from
the
contributors
that
I
work
with
it
I'm
accountable
for
their
alignment?
B
Is
that
they're
like
well
we're
completely
slammed
in
terms
of
stuff
for
beta,
like
there's
no
room
for
us
to
fit
in
there?
Those
are
totally
saturated.
Sam
keeps
telling
us
to
leave
them
alone.
He
basically
tells
me
to
leave
Malone.
So
what
do
we
do?
Next,
where
we
pipelining
for
the
release
tickets
and
then
that's
where
that
exercise,
which
I
haven't
surfaced
yet
of
okay,
which
tickets
don't
have
good
scope?
Yet
it's
kind
of
stemmed
from
is
being
able
to
pipeline
work
for
the
people
that
are
basically
clamoring
and
saying
what
comes
next.
F
Yeah,
so
back
to
the
sorry
for
that
tangent,
going
back
to
the
agenda
bit
and
stuff
like
Josh
to
what
you
were
saying
like
that's
what
we
used
to
use
the
meetings
for
like
the
bay
area
meetings,
that
I
think
we're
kind
of
the
catalyst
for
this
meeting
was.
It
was
a
high
bandwidth
place
where
people
could
talk
about
like
a
compaction
idea
or
we
could
walk
through
the
bugs
that
we
found
etc.
F
I
would
have
expected
that
we'd
use
this
time
say,
for
example,
to
figure
out
what
we
were
gonna
do
with
all
those
tests
Gira's
and
things
like
that,
but
I
think
like
in
order
to
have
a
discussion
about
that
people
need
like
a
week
to
think
about
it
at
least
and
stuff.
So
maybe
good
goal
is
like
an
agenda
a
week
before,
or
something
and
I
can
take
that
this
time,
I
think
all
of
us
can
but
yeah
yeah,
yeah.
F
E
A
Then
you
know
that
in
yes,
I
think
II.
This
is
human
nature.
If
you're
saying
hey
just
bring
up
anything.
That's
on
your
mind,
right
now
my
mind
is
on
something
completely
different.
Yeah
it's
and
I've
been
trying
to
balance
like
yeah
I
could
I
could
go
out
and
I
could
say
here's
an
agenda.
Please
show
up
with
a
with
some
talking
points,
but
I
don't
know.
If
that's
the
right
spirit,
it
needs
to
be
somewhat
organic
so
that.
F
And
and
sorry
I'm
paging,
the
following,
it
is
exactly
that,
so
we
would
start
where
it
like.
Whoever
set
up
the
meeting
that
month,
because
it
would
usually
rotate
between
hosts,
usually
the
host
in
the
case
would
be
like
here
are
two
or
three
items.
I
know
I
want
to
talk
about,
and
then
there
would
be
like
a
Google
Doc
and
everyone
could
add
to
it
and
then
we'd,
Reece
or
tit,
etc.
So
I
didn't
mean
to
imply
like
we'll,
have
a
set
agenda.
F
H
D
F
I
think
like
at
the
end,
maybe
we
do
what
you
were
like
we
were
saying
is.
We
also
then
have
like
us
a
sub-project
status
thing,
but
sometimes
when
you
do
the
status
stuff
at
first,
it
does
sort
of
kill
the
like
everything
else.
Discussion
cuz
people
are
just
like
this
isn't
pertinent
to
me,
but
yeah.
So
let's
try
that
for
next
time
for
July
all.
A
A
F
B
A
Another
twist
on
this,
instead
of
having
it
scheduled
so
much
what
about
making
it
somewhat
asynchronous
with
a
weekly
doubt
so
I'll
give
you
an
example:
JIRA
some
number
there's
a
lively
discussion:
hey
can
we,
you
know
and
and
I
would
be
happy
to
pick
it
up
like.
If
someone
said
hey,
we
need
to
have
a
contributor
meeting
on
this
JIRA.
A
Can
you
can
you
post
it
and
I
would
be
happy
to
post
it
and
I'll
put
up
the
link,
I'll
put
it
in
into
the
confluent
and
but
then
everyone
knows
that
hey
we're
gonna
meet
at
this
date.
You
know
about
this
topic,
so
it
kind
of
captures
the
energy,
because
I
think
those
are
the
meetings
you're
talking
about
it's
like
when
we
had
some
energy
about
something
and
we're
like
we
should
meet
about
that.
B
The
fact
that
meetings
is
ritual
tend
to
suck
something
on
the
calendar
where
you
get
together.
Just
to
talk,
I
mean
it's
good
for
social
connection
like
an
age
of
co-ed,
but
it
isn't
particularly
good
to
to
delay
when
there's
energy
and
focus
and
interest
around
something
the
flipside,
something
we
need
to
contend
with
and
make
sure
we
have
a
really
good.
The
posture
on
is
that
the
the
legacy
of
the
culture
with
the
Apache
project
is
one
of
accessibility
where,
if
it
didn't
happen
on
the
mailing
list,
it
didn't
happen.
That's
by
design.
B
That's
intentional,
because
it's
not
everybody
can
make
a
call
like
this.
So
I
think
as
long
as
we're
really
really
careful
about
the
fact
that
we
are
looking
to
explore
ideas
on
discussions
about
things
like
that,
but
not
come
to
conclusions
and
then
bring
what
you
know
what
points
of
view
we
may
have
refined
on
those
calls
to
the
mailing
list
afterwards.
This
would
be
great
I.
Think
right
because
there's
a
lot
of
stuff
like
yeah
talking
about
the
thing
with
checksumming
or
talking
about
what
we're
doing
in
terms
of
our
testing
etc.
Like
it's.
B
It's
a
somewhat
functional
model
for
someone
to
stand
up
a
strawman
proposal
and
kick
it
out
to
the
list,
but
it's
not
the
most
efficient
model
and
I
also
find
I
am
at
least
as
guilty
as
everyone
else
with
this.
But
communicating
via
writing
is
very
different
than
communicating
when
you
can
hear
someone's
intonation
and
see
their
body
language.
Oh.
B
D
F
Distinctly
remember
this
discussion
from
the
scala
community
because
the
scala
community
had
very
similar
like
online
discussions,
and
one
of
them
gave
presentation
at
the
Scala
days.
I
was
at
in
like
2012,
and
it
was
largely
that
it
was
like
this
is
scientifically
the
fidelity
of
communication
anyways,
so
yeah
Patrick
to
what
you
said,
like
that's
kind
of
what
I'm
saying
is
like
the
camp
leader
like
rotating
thing,
is
just
keep
an
eye
on
the
community.
F
If
you
see
something
like
a
ticket
like
the
consensus
thing
or
the
checksumming
thing
say:
hey
guys
or
like
hey
folks
like.
Why?
Don't
you
bring
this
to
the
contributor
meeting
next
month
or
if
it's
urgent?
Why
don't
we
set
up
a
call
specifically
for
this
just
depending
on
on
context,
but
those
are
the
things
I
think
that,
like
are
substantive
for
this
time,
yeah.
A
And
well,
and
if
I,
if
I,
maintain
I,
think
the
important
thing
is
the
structure
around
it's
like
Tuesday's
our
contributor
meeting
days
and
if
something
and
here's
the
round,
here's
the
round
robin
values
of
my
numerated
times
that
I
that
I
use
and
I
just
round
robin
through
them.
So
it's
completely
fair
in
that
case,
or
at
least
fair
enough,
and
so
I
would
really
like
to
get
into
the
mode
of
having
more
asynchronous
me
as
needed,
because
I
think
that's
where
the
reeling
that
that's
where
the
interesting
stuff
happens,
I
think.
B
Patrick
you
and
I
can
take
away.
There's
got
to
be
some
kind
of
an
app
that
does
this,
where
you
can
have
a
bunch
of
people
bid
for
a
time
and
it'll
figure
out
the
right
time
to
have
a
call
on
something.
So
why
don't
we
dig
around
and
see
if
that
kind
of
thing
exists
because
that'd
be
useful
to
say:
okay
like
next
week
at
some
point,
let's
talk
about
this,
everybody
that
is
interested
hits
up
a
place
says
here's
the
good
times
that
I'm
available,
and
it
tells
you,
okay.
A
H
A
A
Yeah
I'll
keep
the
Kerwin
on
the
schedule,
but
let's,
let's
take
our,
take
a
shot
at
doing
some
aids
asynchronous
because
with
beta
near
shipping
ready
almost
there.
You
know
this
I'm
just
plenty
to
discuss,
but
this
will
probably
be
a
good
muscle
to
to
learn
muscle
response
to
learn
as
we
get
closer
like
into
five
zone
like
what's
gonna
happen.
Whenever
we
start
talking
about
new
features,
that's
gonna
be
the
time
where
we
really
can
use
this
I
think
yeah.
C
B
F
A
So
yeah
Jordan
did
you
hear
the
the
summation
I?
Think
since
you
raised
the
topic,
it's
your
good
bellwether
for
this
I
messed
the
summation
summation
is
we're
gonna,
give
a
shot
at
doing
an
asynchronous
meeting
along
the
way.
Josh
and
I
are
gonna.
Look
for
a
way
to
try
to
figure
out
how
to
do
a
schedule.
Like
hey
here's,
a
topic,
then
here's
some
something
that
can
help
us
figure
out
a
time
to
do
this
great
and
we're
gonna
give
that
a
shot.
Okay,
the
other.
Let's.
A
A
B
G
B
F
Basically,
so
on
the
checksumming
thing,
because
again
sorry,
I
was
out
last
week
when
all
the
interesting
discussion
happened,
we
would
basically
say
the
new
protocol
that
supports
all
the
new
fancy
stuff
we've
added
is
beta
or
whatever
we're
calling
it,
but
you
can
use
v4.
So
if
you
were
upgrading
like
from
3x
4
dot,
o
beta,
you
would
still
have
all
the
features
like
nothing
is
broken
from.
Before
is
what
you're
saying
you.
B
There's
an
experimental
feature
still
and
there's
a
lot
of
like
unknown
in
the
scope
for
testing
that
what
else
is
in
there,
though
I
guess
I
I
haven't
actually
dug
into
it,
because
every
time
I
asked
I
can't
talk
to
Sam
about
it.
It
was
basically
like
and
I
talked
at
Libby,
a
and
I
talked
to
you.
Alex,
like
the
the
v5
protocol,
doesn't
really
a
ton
of
stuff
in
it,
so
making
that
check
something
mandatory.
It's
not
like
we're
gatekeeping
and
blocking
a
ton
of
stuff
from
people.
B
F
Okay,
I
think
it's
like.
If
that's
what
we
had
to
do,
it's
not
like
my
ideal
situation,
I
think
it's
important
like
when
we
try
to
make
promises
about
interface
and
stuff,
but
in
this
case,
like
at
least
like
we're
not
telling
people
hey
like
you
have
to
use
v5
to
talk
to
this
thing
like
you
can
still
run
all
of
your
existing
tests.
I
don't
expect
someone
to
like
have
transient
replication.
B
B
Server
doesn't
mean
it's
gonna
be
done
in
the
driver's.
So
if
we
say
that's
a
blocker
to
release
debate
on
the
server
we'd
be
releasing
a
server
where
we
have
a
v5
protocol
and
no
clients
that
can
talk
to
it.
So
it's
like
the
tree
fell
in
the
forest
and
no
one
was
there
to
hear
it
anyway.
So
there's
there's
value
in
decoupling.
F
Yeah
and
I
think
to
as
long
as
we
say
like
this
is
an
exception
and,
like
part
of
it's
like
really
more
an
artifact
of
like
we
did
all
these
release
lifecycle
things
halfway
through
a
release,
life
cycle.
That
was
done
another
way
right
so
like
next
time,
we
will
do
better
by
the
nature
of
like
having
for
site
instead
of
hindsight
yep.
F
F
Think
continuing
to
show
progress
because
we
have
merged
alpha
stuff
is,
is
good.
I
also
think,
like
part
of
it,
is
that
we,
we
all
sort
of
presumed
that,
because
it's
not
in
beta
people,
won't
test
it.
But
that's
also
because
we're
essentially
messaging
that,
because
it's
not
in
beta
you
shouldn't
test
it
like
I,
can't
speak
for
for
y'all,
but,
like
the
downloads,
probably
reflect
that
right
like
if
we
push
more
like
hey
we're
on
alpha
6.
This
is
starting
to
get
closer
to
beta,
like
maybe
people
will
will
look
at
it
more,
but.
B
G
G
G
D
Agree
that
the
good
agenda
item
that
type
of
sprint
planning
it
doesn't
seem
to
step
on
too
many
toes
and
it
shouldn't
if
it's
for
right,
I
made
a
sink
as
well
before
the
final
decision,
but
those
a
sprint
planning,
type
of
kind
of
moving
tickets,
around
cleaning
them
up,
there's
also
a
lot
of
tickets.
That
could
just
get
closed,
but
no
one
just
wants
to
step
up
and
take
that
lead
and
go
no
one's
touched.
It
there's
a
reason
there
should
be
closed,
closes
I.
F
F
B
Yeah
ekaterina
to
your
point
like
sylvans
comment
on
there
indicates
we
essentially
look
in
application
downtime
for
people
that
need
to
do
this
kind
of
a
migration
of
the
specific
feature
and
that
kind
of
sucks,
unless
we
think
a
lot
longer
and
harder
about
giving
them
a
pathway.
We
should
you
said
you
have
something
on
the
mailing
list
about
this.