►
From YouTube: Open Match Community Meeting Jan 2021
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
All
right:
well
thanks
everybody
for
joining
us
for
the
first
open
match
community
meeting
of
2021.
there's
nothing
on
the
agenda
at
the
moment.
So
this
could
be
a
very
short
meeting.
It
seems,
like
john,
is
typing
something
about
release,
which,
oddly
enough,
is
what
I
was
going
to
ask
about
so
john.
Would
you
like
to
talk
about
release.
B
Yeah,
sorry,
I'm
in
two
meetings.
At
the
same
time,
so
the
release
is
underway,
yeah,
I'm
in
all
hands,
so
the
release
is
underway.
Currently
we're
dealing
with
some
issues
where,
with
some
testing
issues
we're
hoping
to
resolve
it
pretty
soon,
I'm
looking
into
it
now
that
I
have
a
little
bit
more
free
time
as
well.
So
definitely
we'll
be
messaging
back
and
forth.
Trying
to
see
if
we
can,
you
know,
find
some
type
of
resolution
to
these
issues,
but
that's
what
it
currently
is
at
the
moment.
B
Oh
no,
I
now
that
I'm
well
I'm
going
to
be
working
on
it.
A
little
bit
as
well,
but
I
believe,
alex
and
and
team
are
also
looking
into
it
as
well.
Okay,
cool.
A
D
We've
been
on
yeah.
C
We've
been
on
a
four
month
or
five
month
march,
to
kind
of
refresh
our
closed
beta,
which
is
complete
feature
complete
on
monday.
Yesterday
I
mean
monday
sorry,
the
days
blurred
together
in
this
phase
of
development
and
so
for
this
community
meeting.
We
are
deploying
more
open
match
instances
to
customers
next
tuesday
and
we're
actually
finalizing
our
total
complete
migration,
the
following
tuesday,
so
all
of
our
customers
will
be
running
on
on
open
match.
C
That
was
delayed
from
our
previous
update
because
we
had
to
do
a
kubernetes
cluster
migration
which
for
live,
live.
Matchmaking
is
a
challenge
we
should.
We
could
probably
write
a
paragraph
about
you
know
things
can
be
ephemeral,
which
is
nice
about
matchmaking,
but
that
doesn't
mean
that
doesn't
make
it
as
it
doesn't
make
it
seamless.
I
would
say,
however,
what
that
does
do
is
we're.
C
Gonna
have
a
push
for
for
multi-tenancy
as,
like
our
main
theme
going
forward,
which
means
we're
gonna
have
some
engineers
come
back
through
and
open
up
open
match
and
try
to
figure
out
how
to
add
partitioning,
possibly
some
hero
constructs
like
developer
pools
or
something
to
to
open
match
and
get
those
feature
specs
out
there.
So
I've
briefed
the
team
and
we've
got.
I
think
one
or
two
engineers
who'll
be
who'll,
be
stepping
in
as
part
of
like
our
core
product
goal,
so
it's
actually
aligned.
C
I
think,
for
the
first
time
in
a
while
that
that's
something
we're
doing
so.
That's
promise
of
future
engagement,
but
yeah.
That's
kind
of
the
next
big
thing
that's
coming
up
and
that
will
be
our
roadmap
for
the
next
six
months.
I'd
say
is
to
go
in
and
push
push
partitioning
in
open
match.
C
So
that's
that's
gonna,
be
the
next
focus
other
than
that
when
we
finish
the
migration
I'll
send
up
a
flag
that
says
all
the
case.
Studies
we've
published
are
now
running
open,
match.
C
I
think
we
had
a
bug
in
backfill
that
I
need
to
send
andy
around
alexander
to
talk
talk
through,
but
I
think
it
was
just
that
we
had
a
process
that
died,
that
there's
this
interesting
honey
pop
behavior,
but
we
can
take
that
to
a
technical
discussion,
since
this
is
a
community
meeting.
A
It's
awesome
that
you
said
that,
because
I
was
literally
gonna,
make
some
sort
of
snark
about,
like
we've
been
in
this
meeting
for
five
minutes,
and
no
one's
mentioned.
That,
though,
so.
C
On
getting
the
one
of
our
tens
of
matchmakers
sort
of
like
fixed
up
prepped
for
a
dry
run,
we're
freeing
up
our
on
our
managing
matchmakers
can
take
a
lot
of
time
and
operating
them
can
can
be
a
drain
so
for
everyone
in
the
community
running
matchmakers,
it's
hard
keeping
them
up
is
hard,
they're
very
custom.
C
If
you
have
customers
that
are
using
the
custom
functionality,
you've
doubled
your
pain
if
you're,
the
one
using
the
custom
functionality
great
shouldn't,
be
too
bad.
Freeing
yourself
up
to
go
work
on
more
features.
Just
in
terms
of
general
engineering.
Is
you
got
to
prioritize
that
sometimes
so
that's
what
we're
doing
so
we're
hoping
to
have
some
agility
to
come
back
through
and
actually
do
some
do.
Some
open
match
work
for
the
next
six
months.
A
Well
again,
alex
look
like
you
again
on
the
backfield
topic.
Alex
has
got
some
stuff.
Anybody
else
have
anything
they
want
to
add
to
the
agenda.
Otherwise
alex
go
ahead,
and
let's
talk
back,
though,.
F
Yeah,
so
I
just
want
to
tell
john
that
currently,
the
flakiness
of
the
tests
not
yet
resolved,
I
can
cannot
get
the
exact
expectation
date
when
we
can
do
this,
but
yeah.
We
are
trying
to
find
the
reason
why
probably
some
tests
are
running
in
parallel
and
causing
this.
F
So
if
a
bit
switched
to
requests
of
how
to
do
how
to
make
breakfail
id
accessible
from
the
game,
server
and
other
documentation
and
like
tutorials
on
how,
for
example,
we
can
use
agonize
with
backfill
and
what
director
changes
are
needed.
What
mmf
changes
are
needed
because
currently,
for
example,
today,
we
discussed
with
andrew
that
mf
actually
should
always
update
backfield,
not
not
only
game
server,
so
that
was
not
mentioned,
and
that's
why
yeah.
Thanks
to
him.
I
will
add
this
into
the
and
this.
F
F
I
think
I
can
use
another
open
source
project,
which
was
by
let
me
check
so
here
or
a
very
good
project.
I
will
add
from
daniel
elevator,
so
er
gonna
switch,
optimize
and
also
scott's
project
which
also
about
this
yeah.
This
is
probably
so
any
questions
on
this.
We
can
yeah
definitely
for
documentation.
It
would
be
more
updates
coming
because
it's
like
cart
right
now
to
see
where
what
a
mama
does.
What
director
should
do.
F
Some
like
corner
cases,
for
example,
when
bookfield
is
created,
how
we
will
allocate
games
error
and
then
what
am
I
should
do
if
games
are
already
allocated,
for
example,
so,
like
this.
F
Switches-
the
with
this
I
mean
my
things
like
manual
tests
and
some
example
games
here
would
be
good,
at
least
for
me,
to
understand
that
we
are
not
missing
something.
F
Right
now
we
have
only
scale
tests.
I
have
one
demo
of
two
big
fields:
it's
similar
to
getting
started
demo,
which
is
the
not
yet
in
prs,
but
in
my
branch
I
I
will
make
a
person,
so
we
have
scale
tests
and
pen
tests
and
this
demo
of
dropping
these
two
backfills.
That's
it
how
we
have
tested
this.
F
We
have
a
report
yeah,
we
have
our
parts,
it
was
done
by
dmitry,
so
I
definitely
push
this
link
forward
performance
test.
So
the
idea
here
we
in
this
request
was
to
gather
assignment
time
time
to
assignment
so
and
we
have
created,
for
example,
200
2
000
back
fields.
F
There
was
some
8
seconds
delay
before
assignments
come
to
the
stickers,
so
then
yeah
more
more
information
would
be
from
dmitry.
Actually
I
just
was
yeah
the
one
who
like
started,
tune
current
scale
tests
in
order
to
add
backfills
and
pass
over
this
work
due
to
meter
yeah-
and
this
is
the
reports.
F
So
actually
now
there
was
memory
dictators
by
the
way
which
sean
asked
us
to
do,
and
we
have
like
run
for
five
hours
for
eight
hours.
Sorry,
and
it
was
no
reasonable
changes.
I
mean
to
consumption,
to
memory,
consumption
cpu
consumption,
so
we
have
not
recognized
any
memories
so
far
and
was
the
only
memory
was
in
the
wesh
assignment
in
the
there
was
a
pr
just
today
about
how
we
can
treat
like
closing
context,
etc.
F
Yeah.
So
we
are
already.
C
There
with
the
test,
so
we
we
share
the
same
kind
of
the
the
heart
of
our
backfill
design,
because
we
had
an
early
circumvention
of
of
the
open
match.
Backfield
design
should
lend
well
to
us
trying
it
out
at
some
point,
so
maybe
in
in
a
few
weeks,
probably
the
earliest,
because
we're
gonna
have
to
support
this
release,
we're
doing
but
sometime
in
the
three
to
six
week
time
frame.
C
So
maybe
with
the
next
release
of
open
match,
we
might
be
able
to
go
ahead
and
try
flipping
a
bit
and
getting
some
of
our
stress
testing
that
we
do
internally
to
run
on
the
open
match.
Backfill,
and
maybe
we
can
get
a
sort
of
like
apples-to-apples
comparison
on
how
things
are
going
and
at
that
point
might
even
be
able
to
help
identify
or
squash
any
type
of
bugs
that
come
up.
F
Yeah
this
is
this
would
be
great
yeah.
We
had
like
constant,
not
peak
performance
like
number
of
ticket
requests
per
second
was
100
and
then
in
in
the
backfield.
It.
F
Yeah,
okay
and
we
have
plenty
of
rest
amount
of
one
hour,
which
is
not,
I
think,
the
real
per
value,
but
for
the
tests.
Actually,
sean
asked
us
to
do
like
happy
paths
when
there
is
no
delays,
for
example
between
acknowledgements
of
the
field
before
this
acknowledgement.
F
This
was
no
like
a
delays.
So
it's
like
a
good
scenario.
C
Do
we
have
any
community?
Is
anyone
either
on
boarding
to
the
new
backfill
or
have
plans
to
use
the
new
backfill
in
the
coming
coming
months
as
part
of
the
next
release?
C
I
know
we.
I
know
we
have
the
initial
justification
for
for
backfill,
which
was
we
needed.
Help
for
backfill
for
someone
from
the
community
wanted
had
a
need
for
backfill,
but
I
was
wondering
if
anyone
else
was
gonna
was
gonna
hop
on
using
it
and
could
help
provide
that
sort
of
like
multiplicity
approach,
because
otherwise
we're
going
to.
I
guess
we're
going
to
wait
a
few
weeks
until
we
can
sort
of
take
a
shot
at
it.
C
Okay
and
that's
probably
what
I
expected
so
we'll
we'll
try
to
get
that
second
look
in
to
try
and
have
that
kind
of
like
circle
back
and
then
log
anything
else.
That's
missing.
It
sounds
like
a
ton
of
awesome.
Stuff
has
been
done,
though,
in
the
back
full
space.
If
we're,
finally
to
the
to
the
sort
of
the
end-to-end
testing,
I
will
go,
read
those
reports
and
are
those
going
to
be
put
up
as
issues
or
are
those
going
to
be
checked
in
as
as
performance,
documentation.
F
Right
now
there
is
a
pr
with
steps
to
reproduce
once
it
would
be
mesh.
There
is
some
targets
like
making
styles
skill
chart
they
think
and
of
some
and
then
yeah.
F
We
can
add
more
to
this
to
original
issue.
C
C
For
folks
that
come
hunting,
they're
gonna
want
to
know
how
how
you
know
we
have
some
of
those
initial
performance
reports.
I
know
scott
and
john,
I'm
not
sure
where
those
are
at
the
moment,
but
expanding
them,
or
maybe
making
a
section
for
for
these.
For
folks
who
want
to
come
in
and
discover
what
open
match
does
from
a
size
perspective
would
be
pretty
cool,
especially
since
we've
done
all
this
work
to
produce
the
artifact.
F
B
Yeah,
I
agree
that
we
could
probably
add
a
section
for
some
for
some
probable
steps
on
setting
up
scale
tests
and
probably
some
findings
that
we
currently
have
at
the
moment
and
anything
that
we
sh
we
can
find
or
share
within
the
community,
probably
probably
possibly
be
added
as
well.
So
we
can
look
into
it.
B
E
C
A
B
I
guess
I
guess
is
one
last
thing
I
am
following
up
with
the
cloud
run
team.
They
had
some
ideas
about
open
match
and
cloud
run,
so
I'm
providing
some
feedback
with
them
today.
Hopefully,
by
next
week
we
can
have
more
conversation
around
what
cloud
run
needs
or
what
changes
need
to
be
made
to
open
match
that
wouldn't
break
anything
for
developers
who
want
to
use
it.
C
Yeah,
that's
really
cool,
you
know.
We've
talked
about
control,
plane,
style
features
for
actually
managing
the
mmfs
and
k
native
at
the
time
wasn't
quite
ready
and
we
kind
of
called
it
a
1.1
style
problem.
But
then
we
went
out
and
didn't
really
relaunch.
C
The
idea
feels
it
feels
like
that's
a
cool
conversation
to
go
chase
now.
B
Yeah,
that's
actually,
where
I'm
at
right
now,
it's
I
I
tried
to
do
like
the
full
shift
of
everything
over
to
cloud
run
as
much
as
possible.
So
gonna
push
that
feedback,
but
as
of
right
now
it
looks
like
the
mmf
for
cloud
run
seems
like
more
of
an
option.
So
that's
something
I'm
gonna
be
exploring
in
the
coming
weeks.
That'd
be
super
cool.
C
I
know
I
know
there
was
some
some
community
member
who
is
looking
at
k
native
nate,
like
they
were
already
using
k
native
for
their
mmfs,
but
it
was
an
internal.
My
understanding
was
kind
of
an
internal
thing,
so
they
were
using
for
their
own
management
versus
it
being
like
a
control
plane
feature
of
open
match.
E
Hoses
okay,
I've
currently
sort
of
butchered
k
native
out
of
it
and
okay
running
them
is
running
them
as
regular
kubernetes
services.
E
Just
balancing
a
few
different
things
in
in
the
cluster
k
native
vistio
their
eyes,
starting
to
get
a
bit
messy
so
but
yeah
like
theoretically,
there's,
there's
no
difference
other
than
the
management
stuff
that
you
get
from
k18.
Obviously,.
C
C
So
what
I
never
understood
about
k-native
was
whether
or
not
you
could
still
bind
scale
on
compute
usage
versus
just
having
a
sort
of
when
web
requests
activate
spin
up
to
a
certain
number
of
web
requests,
because
the
the
compute
can
just
vary
depending
on
what
the
function
is
processing
right,
and
that
was
something
I
never
quite
understood,
was
how
k-native
was
going
to
solve
that
or
if
it
was
like
built
in,
but
by
the
way.
That's
also
how
we
run
it.
We
run
them
as
kubernetes
services,
kind
of
just
generic,
very
vanilla.
D
C
I'm
kind
of
care
I
mean,
and
we
can
probably
stop
soon
after
this,
but
maybe
an
interesting
topic
for
folks
to
think
about
or
do
the
soul
searching
is
like.
Do
you
need
as
a
person
who's
using
overmatch?
Do
you
need
a
control
plane
for
managing
match
making
functions,
or
are
you
very
comfortable
going
it
alone?
C
And
if
folks
had
opinions
or
someone
wanted
it,
because
you're
not
interested
in
operating
sort
of
match,
function
management,
then
that's
something
that
that
openmatch
can
maybe
have
an
opinion
or
convention
for
doing
or
or
even
a
plane
or
api
for
doing
it.
C
So,
if
you're
seeing
this
on
the
internet
for
next
time,
otherwise,
like
it's
hard
to
say
whether
or
not
it
should
be,
the
next
thing
to
work
on
backfill
is
clearly
the
the
important
thing
and
we're
gonna
go.
Do
some
partitioning
stuff
because
we
would
like
it
so
pull
requests.
Welcome
for
any
other
things,
but
also
feature
requests
or
issues.
A
I
just
say
we
can
always
you
know,
do
it
in
slack
as
well
and
in
the
mailing
list
too.
Just
like
hey,
you
know
what
features
are
people
interested
in
things
like
that.
C
It's
been
a
pretty
active,
actual
slack
now
that
now
that
we're
kind
of
talking
about
it,
a
lot
of
people
client-side
so
in
the
game
domain,
trying
to
figure
out
how
to
use
open
match
to
achieve
a
sort
of
like
mature
services,
ecosystems.
The
way
I
would
interpret
it,
but
without
necessarily
having
like
open
matches
and
exactly
batteries
included,
because
it's
intended
to
sort
of
be
like
a
service
back
in
infrastructure
and
so
maybe
as
open
match
sort
of
reaches
a
sort
of
maturity
and
there's
fewer
things
we
can
extend.
C
We
start
talking
about
the
ecosystem
and
other
ecosystem
projects.
We
need
to
stand
up
because
otherwise,
I
just
don't
know
if
you
know,
I
think,
that's
going
to
continuously
be
a
conversation
that
folks
want
to
have.
Is
I'm
making
a
game,
and
I
want
to
use
open
match
and
our
answer.
You
know
the
answer
to
that
is
like
build
a
service,
build
like
a
game
back
in
service
and
then
put
open
match
behind
that.
It
still
works
really.
C
C
So,
if
you're
out
there
we'll
continue
to
do
our
best
and
and
there's
a
few
heroes
in
the
room,
I'd
say
worth
calling
out
who
have
done
a
great
job
answering
as
many
questions
as
they
can.
The
best,
I
could
say,
is
there's
we
talk
about
it,
especially
here.
So
if
you
want
to
come
here
and
talk
about
it,
let's
do
it.
E
You
were
doing
some
database
evaluation
work
at
unity.
Did
you
guys
get.
C
C
We
we
got
to
the
point
where
we
had
an
aerospike
db
stood
up
and
we
had
a
enterprise.
Redis
stood
up
and
we
were
trying
to
figure
out
if
redis
search
was
going
to
be
a
good
answer
for
what
we
currently
have
as
a
sort
of
problem,
which
is
the
the
redis
is
the
currently
the
bottleneck
throughput
we're
very
much
in
a
technical
discussion
a
bit.
C
The
once
in
summary,
is
it's
unclear
at
this
point
if
it's
just
about
replicating
and
scaling,
because
all
the
functions
need
access
to
all
the
data
all
the
time
and
we
constantly
need
to
re-snap
it
and
therefore
there's
like
a
network
throughput
consideration
versus
in
a
partitioned
world.
C
A
match
function
could
actually
hold
several
hundred
thousand
tickets
in
memory
and
operate
over
those
very
quickly,
and
so
one
of
the
discussions
we're
having
is
like
well,
we've
taken
this
opinion,
which
is
we
have
a
global
pool
or
partition
pools
and
everything
needs
to
be
able
to
be
stateless
and
query
state
in
and
then
lose
state,
but
that's
actually
kind
of
inefficient
and
has
a
ton
of
bottlenecks.
So
is
there
a
compromise
you
make
where
you
can
make
this
functions
more
stateful
kind
of
like
what
query
omquery
does
and
sort
of?
C
C
How
do
we
as
open
match
sort
of
enter
the
domain
of
understanding
more
about
match
functions
and
maybe
prescribe
a
way
that
the
database
enables
faster
matchmaking
as
a
practice
or
as
a
pattern?
So
basically
we're
going
to
check
full
split,
we're
going
to
check
the
databases,
see
if
there's
like
a
more
obvious
choice
to
like
kind
of
remove
the
need
for
like
an
om
query
type
of
thing,
which
is
a
custom
query
database
cache.
C
If
there's
something
we
can
do
with
om
query
to
make
it.
You
know
exponentially
more
make
it
horizontally,
scalable
in
a
partitioned
world
and
still
have
decent
throughput
or
maybe
even
consider
a
type
of
mode
that
we
could
add
to
openmatch,
where
we
actually
enable
functions
to
understand
partitioning
natively
in
memory
and
actually
like
have
a
way
for
them
to
subscribe.
C
To
ticket
cues
and
hydrate
tickets
into
memory,
and
maybe
even
achieve
another
layer
of
scale
on
top
of
that,
so
three
three
phased
approaches:
we're
going
to
investigate
and
hopefully
come
up
with
we're,
definitely
going
to
come
up
with
like
a
backwards
compatible
but
possibly
achieve
another
level
of
scale
for
open
match,
at
least
the
chief
partition,
which
is
another
which
is
a
type
of
scale,
and
maybe
push
for
a
little
past
that,
but
we
basically
paused
it
because
we
needed
to
the
dev
team
wasn't
getting
access
to
the
problem.
C
And
we're
gonna,
I'm
pushing
guillaume's
gonna
join
and
we're
gonna
have
him
kind
of
do
it
in
like
the
open
match
open
way
with,
like
the
issues
that
we're
all
used
to
used
to
doing
we're,
trying
to
bring
that
out
out
of
the
the
unity
corporate
wall?
Type
thing
that
I
know
has:
is
there
right?
It's
always
there,
but
we're
gonna
pull
it
out
of
that.
E
Yeah,
I
mean
we're
partitioning
at
the
moment
a
little
bit
on
the
cheap,
so
essentially
every
ticket
that
comes
in.
We
pull
a
value
out
of
our
jwts
and
stick
that
on
as
like
a
value
to
to
shard
on
and
then.
E
C
Right
yeah,
we
do
kind
of
the
same
thing.
I'm
we're
going
to
try
to
see
if
there's
like
a
a
type
of
feature
that
open
match
needs
to
expose
to
enable
partitioning
as
like
a
like,
we
could
just
say,
put
a
partition
key
in
and
you're
good.
We
could
also
define
a
type
of
configurable
thing
where
it
can
dynamically
create
partitions
right,
like
group
by
build
version,
type
language,
and
so
maybe
openmatch
understands
partitions
as
a
sort
of
like
it
knows
how
to
build
partitions
versus
just
having
predefined
partitions.
C
If
we
have
that,
then
like
problems
like
this
become
very,
very
horizontally,
there's
very
little
configuration
you'll
have
to
do.
You
just
have
to
define
the
rule
that
describes
how
partitions
are
built
and
you'll
be
you'll,
be
done
at
that
point,
but
that
very
much
depends
on
what
type
of
database
we're
able
to
put
up.
So
hopefully
that
gets
easier
with
this
change
and
on
that
it'll
still
be
sort
of
like
on
the
cheap.
Is
the
idea.
E
Yeah
like
for
us,
it's
a
bit
bit
of
a
weird
one
because
sometimes
like,
for
example,
something
like
cross
play.
Some
games
will
have
it
so
that
games
won't.
So
if
you
wanted
to
partition
you're
going
to
have
to
do
it
in
a
reasonably
flexible
way,
because
you
might
want
to
partition
for
a
title,
but
you
you
might
also
want
to
partition
for
a
title
on
a
platform
yeah.
So.
C
C
Some
of
them
are
grouped
together,
some
aren't
and
then
by
like
game
mode
or
maybe
you
have
a
global
pool
or
you
have
a
multi-select
style
where
you
can
actually
enter
multiple
queues
at
a
time
so
also
another
optional
one
and
then
the
last
one
is
sort
of
like
version
of
the
game,
grouped
by
concept
where,
like
you're
now
talking
about
compatibility
between
clients
or
like
having
the
right
dlc
can
you
you
know
effectively,
those
are
like
string
and
maybe
an
integer
match,
or
something
like
that.
C
Can
you
express
a
partition
by
saying,
hey
group
by
platform
group
by
mode
group
by
build
version
and
you'll
end
up
with
16
partitions,
that
kind
of
change
over
time
that
can
grow
to
18
or
go
down
to
four,
depending
on
your
current
development
cycle?
C
And
maybe
there's
a
world
where
openmatch
knows
about
that
kind
of
rule
concept
today
it
has
pools
and
filters
right,
but
it
assumes
a
global
pool,
and
so
maybe
what,
if
we
didn't
have
a
global
pool?
Well,
maybe
open
match
can
understand
a
way
to
break
apart
pools.
That
is
like
a
really
easy
to
work
with
convention
that
maybe
solves
your
your
current
partitioning
id.
C
Great,
I
have
nothing
else:
everything,
okay,
on
your
end,
ish.
I
Sorry
I
couldn't
find
my
mute
button.
Yeah
honestly,
we
haven't
I'd
like
to
eventually
look
into
the
backfill
stuff,
but
just
haven't
had
the
time
right
now,
so
hopefully
eventually,
but
it's
it's
probably
not
going
to
be
for
a
few
months,
every
everything's
working
great
on
our
end.
So
it's
kind
of
a
if
it
ain't
broke
kind
of
a
thing
right
now.
C
Yeah
I've
been
playing
and
and
and
and
and
bragging
a
little
bit
here
and
there,
but
you
know,
but
it
it's
it's
yeah.
It
seems
to
be
going
pretty
well
as
a
player,
cool.
D
F
I
just
was
thinking
anyone
tries
to
create
to
use
open
mic
in
two
different
namespaces
I
mean
install
it.
I
tried,
but
no
luck,
but
I
have
tried
on,
though.
C
C
We
haven't
quite
hit
like
a
hundred
yet,
but
we
we
have
several
tens
of
them
running
in
in
one
cluster
across
namespaces.
Is
there
a
particular
question
that
yeah.
F
Yeah
so
yeah
you
mean
multiple
optimized
cars
and
every
every
new
one
in
a
separate
new
space
right.
Okay,
that's
why
I
think
I
just
missing
some
config
okay.
F
C
I
don't
think
we're
using,
I
want
to
say,
we've
rendered
out
the
helm
chart
into
like
vanilla
into
like
kubernetes
yamls
that
we
might
have
touched
or
manually
manipulated
we're
trying
to
move
back
to
the
vanilla
helm
chart
though
so,
if
the
vanilla
helm
chart
isn't
working,
I
don't
think
we're
launching
any
work
to
move
back
to
that
unless
there's
a
new
feature
in
open
match
which
we're
going
to
be
developing.
C
So
we're
probably
eight
weeks
away
from
being
able
to
help
with
that
question
on
our
side
because
we're
we
are
going
to
go
back
to
the
vanilla
helm
chart.
But
it's
going
to
be
a
little
while.
C
Yeah,
but
it's
definitely
possible,
I
don't
think
we've
changed
many
settings.
I
can
ask
someone
how
many
settings
we've
changed
if
that
might
help,
but
I'm
pretty
sure
we
just
we
changed
the
namespace
and
that
was
it
and
we
changed
some.
F
Resourcing
I
think
I
was
just
thinking
about
like
a
separate
new
space
with
backfill
like
mmfs
and
some
namespace
with
openmatch,
without
it
just,
for
example,
to
to
integrate
it
like
one
by
one
or
step
by
step,
but
yeah.
I
will
try
this.
I
think.
C
If,
if
you
know
in
a
few
days,
it's
you're
still
not
making
progress,
getting
it
running
in
multiple
namespaces,
throw
up
a
slack
message
and
I'll
tag,
one
of
the
unity
engineers
who
has
made
it
past
this
point
and
we
can,
we
can
figure
it
out.
We
run
with
our
development
in
qa
and
cicd.
We
run
hundreds
of
open
match
instances
in
our
clusters
at
any
given
point
in
time
so
and
they're
all
in
different
name
spaces
in
in
several
environments,
yeah.
D
This
is
great:
okay,
just
want
to
to
check
this
okay,
sure,
okay!
Well,
do
we
do
four
weeks
or
two
weeks
or
six
weeks.
A
For
releases
or
meetings,
this
is
always
the
last
wednesday
of
the
month.
D
C
C
A
Needs
to
get
one
tab
and
then
your
life
will
be
easier.
C
E
A
Yeah,
no,
I
have
a
lot
of
chrome
open
so,
but
I
did
just
change
all
my
monitors
around
and
try
to
make
it
a
little
bit
easier
to
read
and
keep
track
of
things
so.
F
Yeah
and
I
I
gave
my
my
chromebook
to
my
father
and
he
just
used
a
guest
mod
and
he
liked
it
very
much.
It's
like
no
extensions,
it's
very
fast
and
he
always
starts
from
a
blank
page.
G
A
Yeah,
it's
kind
of
like
the
like
pro
tips
on
reddit
and
like
running
joke.
It's
like
the
real
life
pro
tip
is
always
in
the
comments.
So
it's
kind
of
like
you
know
for
us.
What's
everybody
playing
video
game?
Why
is
anybody
playing
anything
fun.
F
Yeah,
I
I
was
playing
keeper
punk
and
like
finished,
one
act
first
act
and
it
was
like
some
kind
of
virus
or
it's
very
common,
like
a
coronal
virus
yeah,
because
this
girl
suffering
from
something
yeah,
it
was
crazy
stuff,
any
any
game.
F
Yeah,
just
one
guy
wants
to
use
some
like
narrower
some
kind
of
chip,
and
he
just
have.
Instead
a
gun
like
he
entered
a
gun
instead
of
cheap.
Oh.
C
H
C
Some
pretty
crazy
maps
that
just
like
cars
flying
everywhere,
but
the
the
game
itself
is
kind
of
a
little
30
second
track.
Sometimes
multi-lap
arcade
game.
It's
really
fun.
It
scratches
that
itch
of
like
I
kind
of
want
to
perfect
something,
but
I
don't
want
to
spend
like
an
hour
on
speed
runs,
so
it
just
allows
you
to
do
like
little.
Speed
runs.
A
Yeah,
I'm
not
this
looks
interesting.
I'm
not.
I
will
totally
race
you
in
real
time
in
a
car,
I'm
not
so
good
on
the
virtual
one
in.
A
And
he
attended
it
with
at
mario
kart
yeah.
We
would
always
play
together
and
I
would
always
lose
because
it's
hard.
A
That
does
look
interesting,
though
I've
just
been
playing
a
lot
of
final
fantasy
14,
because
over
the
holidays,
friends
were
like
hey.
We
should
totally
do
that
thing.
We
keep
talking
about
doing
getting
back
into
final
fantasy
and
now
I'm
obsessed
again
because
that's
how
it
gets
you
so
yeah.
C
A
Nobody
wants
to
like
hang
out.
That's
weird!
No
in
our
discord,
we
do
have
one
person
who
lives
in
australia,
new
zealand
and
so
every
once
in
a
while,
he'll
be
like
no.
I
can't
I
can't
play
tonight
because
I'm
going
out
with
some
friends
we're
like
show
off.
J
I
was
gonna
say
I've
been
playing
one
round
of
splunky,
two
rage
quitting
and
then
restarting
it
ten
minutes
later,
I'm
just
in
the
loop.
A
C
A
Apparently,
there
are
cats
you
can
like
adopt
in
the
game
now
so
and
dogs
as
well.
A
That's
the
thing
we're
really
all
playing
for
yeah,
so
m
said
she
bought
a
cat
in
cfds
and
now
can
walk
around
everywhere.
Holding
a
cat
and
my
arms
like
they're,
a
baby
left
click,
pets,
them
and
right
click
holds
them
out
in
front
of
me,
like
I'm,
probably
showing
somebody
look,
it's
a
cat.
This
game
is
perfect,
simple
pleasures
these
days.
J
If
you're
gonna
play
at
me
and
I
might
join.
E
A
E
Yeah
we
have
somewhat
of
a
pokemon
household
because,
like
my
kids
are
in
at
like
the
right
age
for
pokemon
early,
so
yeah,
my
the
eldest
can
tell
you
like
what
region
the
pokemon's
from
how
much
it
weighs
what
moves
it
can
learn.
E
C
Well,
I
actually
need
to
hop.
I
mean
I
am
going
to
grab
a
coffee
and
take
a
vile
break
before
my
next
commitment,
but
I
just
want
to
say
I
always
enjoy
these
meetings
and
I'm.
C
The
next
one
and
april,
thank
you
so
much
for
for
organizing
these
and
for
scott
for
and
john
for
having
a
a
sense
of
organization
around
sometimes
putting
things
into
the
into
the
community
discussions
same
same
for
you
alexander,
so
I
just
want
to
say
I
appreciate
it.
Thank
you.
Well,.
A
Yeah
thanks
everybody
for
joining
and
thanks
for
all
the
hard
work,
and
we
will
chat
again
soon
thanks
everyone.