►
From YouTube: Agones 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
our
first
meeting
of
2021
super
exciting
small
group.
Today,
some
new
faces,
but
that's
great,
does
anybody
wanna
before
we
dive
into
mark's
agenda?
Does
anybody
wanna
tell
us
anything
about
cool
that
they're
working
on
or
anything
you
don't
have
to
I'm
just
giving
you
the
free
time.
B
I
mean
I
haven't
checked
in
with
y'all
in
a
while,
I
could
say
hi,
I
suppose
hi
there's
a
big
old
wave.
I
guess
my
mic
is
working
yeah,
so
daniel
I've
been
working
a
lot
with
daniel
who
you
guys
know
from
answering
every
single
question
in
the
chat
every
time
anybody
has
a
question.
B
Yeah,
the
dude's,
a
freaking
monster,
but
yeah,
so
we've
been
working
on
our
game.
We
we
actually
did
a
big
multiplayer
demo
on
new
year's
eve
and
we
got
a
bunch
of
people
that
caught
on
and
we
crashed
our
game
like
three
times
and
it
was
just
instantly
right
back
up.
Just
that's.
A
B
Yeah,
so
it
was
just
it
went.
It
went
really
well
for
like
as
far
as
demonstrating
agonists
and
and
the
infrastructure
that
we
have
built
demonstrating
the
actual
game
terrible
it
was.
It
was
a
nightmare.
It's
horrible,
but
demonstrating
the
infrastructure
went
really
really
well,
so
that.
C
B
C
B
Games,
no,
it
didn't
go
that
bad.
It
didn't
go
that
bad,
but
it
it
was
a
good
infrastructure
demo.
It
went
really
well.
A
Good
to
hear
mark,
do
you
want
to
talk
about
the
release
candidate
that
you
just
did.
D
Yes,
I
put
that
in
the
agenda,
so
I
should.
I
should
want
to
talk
about
that.
That
would
make
sense.
I
will
highlight
a
couple
of
things.
I
put
it
in
the
release
notes,
but
I
have
found
that
unless
people
like
scream
it
at
you,
maybe
people
won't
listen,
probably
there's
a
there's.
A
couple
of
things
that
moved
up
levels
so
like
getting
an
initial
copy
of
the
game.
Server
on
watch
on
connection
has
now
moved
to
beta.
D
So
I
stand
up
by
default
so
just
being
aware
of
that
that
small
behavioral
change
a
couple
of
other
new
feature
flags.
But
the
big
thing
is:
you
should
have
better
upfront
validation
of
your
pod
spec
inside
the
game
server,
because
now
the
game
server
crd,
includes
the
full
pod
template.
Spec
json
schema
ripped
straight
out
of
kubernetes,
because
I
wrote
a
bunch
of
bash
the.
D
If
you
need
to
turn
that
off,
you
can
actually
enable
preserve
unknown
fields
which
basically
just
turns
off
all
like
the
validation
in
the
field
pruning
and
if
you
run
into
any
issues
with
that,
the
other
thing
that
may
be
worth
tricking
out
early
and
it'll
be
lovely.
If
you
could
check
it
during
release
candidate
this
week,
even
though
tuesday
next
week
is
when
we
do
the
stable
release
to
see
if
there's
anything
wrong
or
anything,
you
run
into
any
issues.
D
It
is
a
big
crd
change,
so
I
have
found
instances
where
actually,
if
you're
using
helm,
do
a
full
uninstall
and
then
reinstall,
because
sometimes
it
complains
or
I've
actually
actually
in
our
entwine
system,
we
had
to
upgrade
our
helm
version
because
it
just
was,
like
I
don't
know,
what's
going
on
and
so
just
giving
a
heads
up
if
you're
doing
like
maintenance
on
an
individual
cluster,
you
may
need
to
do
a,
and
you
just
want
to
do
that.
D
You
might
just
need
to
rip
it
out
and
put
it
back
in
rather
than
do
an
upgrade,
but
you
can
try
it
first
worst
case
it'll
just
fail
and
then
you'll
have
the
old
things
still
running.
D
Trying
to
think
there's
anything
else,
the
it
is
probably
also
better
in
some
ways
to
do
a
clean,
install,
because
one
thing
I
have
seen
as
well,
which
is
probably
fine,
it's
not
really
like
a
problem,
is,
if
you
are
from
the
last
release
when
we
created
the
new
crd
stuff.
D
So
we
went
to
crd
version
one
and
we
actually
have
a
a
a
validating
spec
for
for
the
customer
resource
definition,
so
it
actually
validates
the
data
coming
in
creates
this
thing
called
preserve,
unknown
fields
which
basically
says
like
it
doesn't
prune
anything.
It
keeps
everything
rather
than
pruning
what
it
doesn't
know
about
when
it
kubernetes
is
neat
in
that
the.
When
the
this
custom
resource
definitions
were
in
beta,
it
would
actually
do
the
conversion
to
v1
for
you,
which
is
cool,
but
it
turns
on
preset
runner
field
to
everything.
D
So
if
you're,
installing
11
or
12,
you
may
want
to
do
a
clean
install
just
so
that
you
have
a
consistent
behavior
across
all
your
clusters,
rather
than
an
upgrade
just
something
to
be
aware
of,
but
other
than
that.
Just
lots
of
bug,
fixes
and
stuff
usual
shenanigans.
E
It's
true:
if
you
don't
turn
up
for
meetings,
can
you
get
basically
voluntold
what
you're
doing
next
next
month.
F
Okay,
so
here.
D
D
And
it's
in,
should
I
get
my
did
I
get
my
acronyms
right?
Eks.
Yes,
aks
is
your
okay.
There
was
a
way
I
think
of
doing
this
in
here
here.
Try
it
sign
in
with
my
azure
account.
I
don't
have
oh
yeah
there
we
go.
A
Like
we
have
a
lot
of
friends
that
are
in
the
open
source,
you
know
cloud
world
so
yeah.
It
is
kind
of
nice
like
when
something
on
github
happens.
I
actually
have
someone
I
can
just
spoke
and
be
like
hey
fix
this
so
yeah.
It
is
very
nice.
D
Do
you
have
a
thing
that
lets
me
run?
It
no
run
just
run
it
yeah.
Thank
you.
It's
literally
like
it's
good.
They
have
a
website
thing
that
you're
they're,
like
you,
can
run
it
in
the
website
and
I'm
like
fine
and
they're
like
let
you
sign
in
and
I'm
like,
I'm
hitting
the
sign
in
button
and
it
won't.
Let
me.
A
D
A
D
D
D
H
D
D
I
think
it's
probably
the
same,
I'm
just
copy
pasting
stuff
that
hasn't
literally
just
moved,
I
think.
Well,
I
think
both
of
these
could
get
bumped.
I
think
we're
at
the
same
place.
D
We
were
last
month,
which
is
the
player
tracking
stuff,
still
needs
the
update
to
the
rest
api,
which
has
been
talked
about
in
a
bug,
and
that
just
is
work
that
just
needs
to
happen
before
that
gets
bumped,
and
someone
just
needs
to
take
that
on
board
fix
rolling
update
skill
down
that
could
probably
also
get
bumped,
but
we
just
need
someone
to
do
that
work.
I
think
there's
that
I
think
there's
a
ticket
already
for
that
too.
D
No
okay,
there's
no
ticket
for
that.
Anyone
disagree
that
this
couldn't
get
moved.
It's
been
several
releases.
D
D
Do
so
we
moved
no
in
the
latest
release,
as
I
mentioned
before,
sdk
send
what
game
server
on
watch
execution
just
went
to
beta
m112
and
we
just
created
a
new
one,
which
is
utilize,
node
external
dns,
in
additional.
D
So
basically
we
use,
we
would
first
go
external
ip
internal
np.
When
we
look
at
nodes
now
we
look
at
external
dns
and
then
external
ip
and
then
so
on
and
so
forth.
That's
currently
in
alpha
2..
D
D
D
E
I
had
a
myriad
of
tabs
open,
which
wasn't
helping
that's
just
gonna,
be
the
theme
for
the
meeting.
Is
everyone's
gonna
have
microproblems?
That's
fine
where
where
was
I?
Yes?
So
I'm
still
working
on
a
ue4
backwards
compatibility
issue
to
get
four
point:
two
six
and
4.25
working
my
main
issue
around
this
is
I
just
don't
either
have
enough
time
or
I
don't
have
the
both
versions
of
unreal,
because
I
don't
want
to
mess
my
dev
machine
up.
E
E
So
still
waiting
out
on
that,
and
the
only
other
thing
I
want
to
mention
on
ue4
was:
I
will
probably
implement
some
kind
of
poor
man's
watch
because
I'll
need
it
soon
anyway.
So
I'll
do
a
first
pass
of
that
when
I
get
there,
which
will
probably
be
might
be
a
couple
of
weeks,
maybe
a
month
away,
well,
yeah
I'll,
definitely
need
it.
E
So
if
it's
not
there
by
the
time
I
come
to
it
I'll
implement
it
yeah
that
was
kind
of
just
an
update
and
where
that
was
this
needs
to
be
fixed
and
then
there's
a.
I
was
trying
to
find
a
link
for
the
actual
lining
code.
So
with
the
moving
on
to
the
next
point,
the
there's,
the
golang
kubernetes
client,
which
I
think
you
use
inside
the
allocator,
probably
like
deleting
and
adding
game
servers.
E
There's
this
concept
of
great
there's
a
load
of
options
on
the
delete
api
to
specifically
delete
a
running
game
server
that
has
been
allocated
right.
So
it's
fine
to
do
it
when
it's
ready,
but
it
seems
as
though
you
can't
delete
an
allocated
game
so
which
is
fine
in
production
in
dev.
I
want
my
engineers
to
be
able
to
to
click
and
delete
an
allocated
game
server.
D
E
E
Okay,
cool
ignore
that
point.
Then
that
was
no
point.
If
that
is
supported,
and
then
my
last
point
was:
is
there
any
alternatives
in
the
pipeline?
So
fleet
iq
seems
to
be
this
thing
from
amazon,
whereby
they
almost
use
a
demon
set
to
load
balance
spot
spot
price
nodes,
so
you
can
basically
have
short-lived.
They
basically
won't
delete
the
the
spot
instance
for
a
period
of
time,
apparently,
which
seems
very
nice
to
have.
E
At
google,
I
would
say:
that's
a
really
nice
product
now
I'd
hate
to
have
to
move
every
shade.
E
Well,
it's
cool!
If
anyone
wants
to
check
it
out,
it
would
be
nice
if
something
like
that
appeared
as
a
request
for
the
things
that
might
come
in
in
the
future.
Okay,.
D
Questions
have
one
other
item
that
I
wanted
to
bring
up
yeah
I'd
love
to
see
more
engagement
in
slack
on
hash
users.
It
feels
like
it's
just
me
and
and
daniel
answering
everyone's
questions,
and
I
know
there's
a
bunch
of
people
who
use
this
product
and
get
value
out
of
it.
A
B
Creating
swag
right,
I
do
have
a
question.
Curiosity
question
is
there?
I
know
that
dom
kind
of
has
been
shouldering
a
lot
of
the
ue4
compatibility.
B
D
B
D
B
So
our
game
is
the
vision
for
the
game
is
very
physics
based
and
ue4
is
not
real
friendly
on
physics,
and
one
of
our
contributors
is
on
the
rocket
league
team
and
has
just
straight
out
said:
don't
mess
with
physics
until
you
get
ue5,
because
the
ue5
that
the
internal
previews
of
ue5
have
been
like,
really
physics
friendly.
So
we're
just
like
sitting
on
the
edges
of
our
seats
waiting
to
start
development
in
ue5,
and
so
thus,
the
question
of
is
that
anywhere
on
the
roadmap
for
agonist?
B
Yet,
and
you
answered
that
just
wanted
to
give
you
more
context.
B
No,
no
so
they've
been
kind
of
going
back
and
forth
between
a
vague,
like
2021
and
then
when
covet
hit.
They're,
like
you
know
what
we're
gonna
bump
it
up
to
spring
2021
and
then
they're
like
at
gdc,
is
now
the
like.
Most
recent
one
is
like
it,
but
there
there's
no
there's
no
clear
point
of
when
it's
going
to
be
public
they're
internally,
developing
with
it
all
the
internal
epic
type
teams
and
like
psionics,
has
access
to
it
and
everything.
D
A
B
Yeah
thanks,
that
was
just
a
curiosity,
just
wondering
if
maybe
maybe
y'all
were
lucky
enough
to
have
something
in
the
works
already
just
curious.
E
A
Yeah
we
know
that
I
mean
I
know.
We've
talked
with
like
the
godot
folks
at
once
before
and
they're
awesome,
I
think
they're
just
limited
on
resources
and
it's
one
of
those
things
where
they'd
like
to
do,
but,
but
I
think
most
of
it
is
just
whatever
people
are
using
that's
what
they
are
motivated
for.
D
A
No,
it
is
I'm
just
saying
like
as
far
as
we'd
love
to,
I
think,
argonas
would
love
to
have
all
the
implementations.
It's
just
a
matter
of
like
haven't
run.
I
haven't
run
into
a
lot
of
people
that
are
interested
in
the
other
ones,
but
that's
always
the
big
challenge
is
just
people
and
then
the
expertise
I
mean.
That's
the
other
big.
D
Which
is
super
annoying?
I
was
going
to
ask
you
dom
with
the
with
the
unreal
stuff.
I
know
very
little
about
it
and
very
little
about
c
plus
in
general,
but
is
there
a
way
even
to
like
what
is
it
pre-progressing
like
blocks,
basically
like
if
you're
in
this
version,
you
get
this
chunk
of
code
and
if
you're,
in
this
version
you
get
this.
E
D
E
It's
kind
of
going
to
come
up
to
one
thing:
that's
going
to
say
about
the
old
engine
thing
is,
I
think
we
formed
this
very
weird
spot
where
most
of
us
in
here
are
probably
more
infrastructure,
focused
and
hardly
working
with
game
engines
which,
for
me,
is
like
the
perfect
storm
of.
I
don't
really
know
that
much
about
unreal
either.
E
So
there
are,
if
deaths
we
can
do,
but
I
don't
know
if
it's
per
version
but
that
the
change
I
made
recently
came
directly
from
the.
B
D
A
A
I
know
this
is
not
necessarily
something
that
is
on
their
radar,
because
it's
improv,
but
you
know,
like,
I
think,
there's
a
fair
amount
of
some
recruitment.
We
can
kind
of
do.
I've
tried
before
with
some
people,
but
like
I,
I
think
that,
like
that's,
ultimately,
what
we
need
is
just
to
build
the
base
of
people
that
are
contributing
and
bring
in
like
the
different
specialties.
A
E
D
B
We
found
a
memory
leak
an
hour
before
the
deployment
that
I
was
talking
to
you
about,
because
daniel
was
like
wait.
We
don't
have
any
visualization
for
our
monitoring.
Let
me
throw
it
up
real
quick.
He
throws
it
up
and
it's
we
just
watch
this,
the
memories.
It's
just
going
straight
up,
we're
like
yeah.
We
should
probably
fix
that.
We
should
probably.
D
Fix
that
that's
genuinely
useful,
I
did
remember
one
thing
I
should
also
bring
up.
I
mentioned
it
in
the
issue
once
once
we
do
stable
of
1.12,
I'm
gonna
move
master
domain
inside
the
github
repo.
So
just
be
aware
that
that's
coming
so,
if
you
have,
if
you
have
forks,
you
will
need
to
update
those.
A
D
Brad,
did
you
want
to
talk
about
your
actually?
I
haven't
looked
at
it
again,
the
pr
for
running
a
fedora
running
to
build
scripts
on
fedor
sure
I.
D
J
That's
what
I
figured
no,
I
I
was
just
booting
up
on
I'm
running
fedora
on
my
laptop,
which
I'm
doing
going
to
be
doing
the
builds
on
and
yeah.
There
were
just
a
couple
little
changes
because
of
the
I
assumed
because
of
the
sc
linux
and
podman
that
I
had
to
make
during
the
build
the
initial
build
kind
of
testing
go
and
following
that
documentation.
J
So
I
thought
it'd
be
good
to
make
those
changes
in
a
branch
and
push
them
back.
If
you
guys
want
to
check
it
out
and
pull
that
in.
D
D
Like
a
lot
of
them
are
just
grabbing
the
local
cages
that
you
have
and
shoving
them
back
up
into
the
container,
because
it's
useful
so
they
are
host
mounts.
I
get
worried
about
what
permission
changes
that
dash
that
z,
thing's
gonna
make
and
that
kind
of
freaks
me
out
a
little
so
that
that
should.
J
Be
personally,
I
have
never
seen
a
major
issue
with
other
tools
using
that
that's
really
sc
linux
contacts,
labeling
content,
so.
J
D
Yeah
it's
about
it's
passing.
I
guess
the
only
like
in
cloud
build
it's
not
such
a
big
deal,
because
if
it
makes
any
permission,
change
or
anything
like
that,
cloud
builds
like
ephemeral,
so
it
disappears
like
who
cares
right.
D
J
It
is
managed
through
the
container
engine,
so
if
it's
run
c
with
podman,
then
pondman
will
have
the
changes
that
it
will
interpret,
that
command
versus
docker
and
and
moby
will
do
we'll
interpret
from
the
operating
system.
What
changes
it
believes
it
needs.
D
D
J
Oh,
I
just
forgot
what
it
was
called
dyson
fear
program,
which
is
a
factorio
light
game
for
multi-planet,
dyson
spheres,
pretty
interesting
so
far,
18
or
so
bucks
on
steam.
C
A
Yeah,
I
was
gonna,
say
brad.
There's
a
lot
of
interest
in
this
type
of
you
know:
factorio
satisfactory
type
games.
Lately.
I
think
it's
like
a
mind,
not
mindless,
but
like
relatively
repetitive
thing
that
people
can
do.
I
know
for
me,
it's
great,
you
know,
doesn't
there's
no
anxiety,
there's
no
stress.
A
I
heard
that's
a
really
popular
game,
so
yeah,
I
have
not
started
it
yet
either
largely
I
there's
the
cyberpunk
subreddit,
where
people
are
just
like
constantly
sharing
the
weird
bugs
and
things
and
it's
interesting,
and
it's
entertaining
enough
for
now.
A
A
Unpopular
opinions
on
the
internet,
what
can
I
tell
you?
I
did
find
it
slow
at
the
start,
which
I
think
is
partly
why
I
put
it
aside
for
a
while.
C
A
D
E
D
I
struggle
with
the
the
game
mechanic
of
like
deteriorating
weapons
and
I
get
it,
but
it
also
just
frustrates
me.
A
D
D
E
Quick
question
from
mark:
what's
right
here,
the
controller
service
does
that
have
a
hdp
api
that
does
exactly
the
same
as
the
crds.
E
I'm
using
that
I
just
wondered
because
I'm
not
doing
it
from
inside
the
server
inside
the
cluster
I'm
calling
from
outside
the
cluster.
D
Oh
authentication
can
be
a
pain.