►
From YouTube: Scalable Multiplayer Game Design with OpenShift (E1)
Description
This monthly series explores what it takes to design a scale-out multiplayer video game architecture using containers and OpenShift. There will be live coding, philosophical design discussion, and everything in between.
In this first episode, we'll choose a game genre and debate how the game design (rules and internal game systems) inform the architecture. We'll also discuss how, where possible, to break up the traditional monolithic game server into microservices components to support more massive scale.
GitHub repo: https://github.com/RHGamingCoP
Host:
Erik Jacobs
Guests:
* Derek Reese
* Michael Clayton
* Roderick Kieley
* Jared Sprague
A
Hey
everybody
and
welcome
to
our
first
live
stream
of
scalable
multiplayer
games
with
openshift.
I
am
your
host
eric
jacobs,
joined
by
several
illustrious
other
red
hatters
that
I
will
allow
to
introduce
themselves.
Firstly,
in
order
of
the
way
they
appear
in
my
zoom
conferencing
tool,
roddy
keeley.
B
Hey
everyone
radically
here
from
the
messaging
engineering
team,
at
red
hat,
I've
been
at
red
hat
for
about
five
years
working
on
open
shift
operators,
back-end
messaging
technologies.
I
also
did
a
stint
or
a
few
stints
really
in
the
gaming
industry
at
certain
canada.
If
anybody
remember
those
folks
other
ocean
interactive
and
a
couple
of
my
own
binary
dawn,
interactive
and
unified
softworks
doing
mobile
apps
back
around
2010
or
so.
A
Oh
man,
you've
done
a
lot
more
gaming
stuff
than
I
even
knew
see.
This
is
this
is
going
to
be.
Today's
interesting
thing
is
that
I
am
the
no
talent
clown
of
the
group
and
I'm
really
just
going
to
drag
the
entire
experience
down,
but
the
others
here
are
definitely
going
to
do
their
best
to
subvert
all
of
my
efforts
to
screw
everything
up.
So
thanks,
roddy
and
then
next
we
have
michael
clayton.
C
Hey
everybody,
I'm
michael
clayton,
I'm
a
front-end
developer
on
the
red
hat
customer
portal
and
very
amateur
game
developer
jared,
and
I
have
have
built
quite
a
few
games
both
in
our
spare
time
for
game
jams
and
some
for
red
hat,
like
interactive
experiences
at
conference
booths
and
virtual
arcades
during
the
covet
era.
So
it's
been
been
a
little
while,
since
I've
had
my
fingers
in
in
game
code
but
glad
to
be
here.
A
Awesome
jared's
spray
sprocket,
I
don't
I'm
gonna
butcher
everybody's
name
too.
Apparently
today.
D
Yeah
sprague,
hey
I'm
jared
sprague
and
I
work
at
devops
red
hat.
I
also
do
some
gaming
projects
with
four
red
hat
and
also
outside
with
with
michael,
have
developed
a
few
games
for
events
like
red
hat
summit
and
other
things
as
well
as
participating
in
some
game
jams
like
ludum,
dare
with
michael
and
yeah.
That's
it
for
me.
E
Hi,
I'm
derek
reese,
I'm
principal
engineer
at
a
red
hat
and
I
work
on
an
acceleration
team
internally
for
our
web
web
properties.
I
am
also
the
cto
of
landfest,
which
is
a
gaming
charity,
and
most
of
my
experience
in
games
has
been
from
releasing
a
game
called
first
offensive
back
in
2009,
as
well
as
releasing
numerous
tools
for
blender.
E
As
unreal
and
unity,
plugins.
E
A
Too
man,
I'm
learning
all
kinds
of
cool
stuff
today,
so
yeah.
We
we
had
this
idea
not
that
long
ago,
that
we
we
were
gonna,
do
sort
of
multiplayer
game
stuff.
We
really
wanted
to
explore
how
games
would
work
on
openshift
if
they
could
work
on
openshift
and
kind
of
go
into
some
of
the
gory
details
there
and
one
of
the
things
that
we
sort
of
figured
out
and
for
the
other
folks
on
the
line
here,
you
know,
feel
free
to
cut
in
at
any
point.
You
know.
A
We
just
gonna
have
like
monolithic
game
server
that
does
everything
and
kind
of
do
what
everybody
does
or
could
we
could
we
use
microservices
to
to
build
something
more
scalable
you
hear
about
all
these
horror
shows
people
crunching
and
crashing
and
thrashing
because
you
know
the
servers
don't
perform
well
or
whatever,
and
it's
like,
but
couldn't
something
like
openshift
just
be
magic
and
and
my
joke
commonly
that
I
don't
think
any
of
the
folks
on
the
line
I've
heard
before
is
a
you
know:
openshift
doesn't
solve
your
bad
application
architecture.
A
A
In
that
regard,
I
started
doing
some
research
into
sort
of,
because
I'm
again
the
no
talent
person
here
I
started
doing
some
research
into
kind
of
you
know
microservices
game
design
like
is
it
a
thing?
Has
anybody
blocked
about
it
or
wrote
about
it
or
whatever,
and
what
I
found
was
a
gentleman
in
the
nordics?
I
think
it's
a
gentleman.
I
don't
know
that
might
be
assigning
gender,
which
makes
me
a
horrible
person.
A
So
this
person
mano
wrote
a
thesis
paper
on
microservices
game
design
and
I
read
it
and
I
was
like
wow
this.
This
actually
doesn't
seem
too
bad.
Maybe
we
should
use
this
as
sort
of
the
background
of
you
know
what
we're
trying
to
think
about
and
we'll
probably
go
and
actually
pull
up
the
pdf
and
look
at
it
on
this
call.
A
But
one
of
the
things
that
we
wanted
to
figure
out,
I
think,
on
on
this
initial
show,
for
the
series
is
like
what
what
kind
of
game
do
we
want
to
make
seems
that
we've
already
settled
on
something
multiplayer
and
something
on
the
internet
right
so
like,
but
we're?
Where
do
we
go
from
there?
You
know
I'm
thinking,
probably
not
turn-based
strategy.
A
If
we're
going
to
really
talk
about
something
that
needs
to
be
super
scalable
and
have
you
know
a
very,
very
large
number
of
players
right
like
how
many
people
can
you
really
put
in
a
civilization
type
game
before
it's
just
a
disaster?
You
know
six,
like
that's.
That's
not
very
interesting
right
about
seven.
E
Yeah,
my
usual
recommendation:
go
to
for
building
things
in
the
cloud
that
show
off.
Scalability
is
an
rts
game,
because
that's
something
where
you
know
scalability
is
a
primary
gameplay
feature.
E
And
rts
stands
for
a
real-time
strategy
yeah.
So
that
would
be
something
where
you
would
have
a
map
of,
perhaps
even
infinite
size
right
using
your
openshift
scaling
tools
and
you
can
have
even
infinite
players
right
if
everybody's
allocated
their
own
space
to
start
and
then
move
towards
other
players
for
your
conflict.
A
Yeah,
yes,
that's
an
interesting
one,
so
that's
one!
We
got
one
option
there
kind
of
rts.
A
I
would
say
that,
like
an
rts
at
that
scale,
almost
feels
more
like,
like
r4x,
almost
like
a
like
a
real-time
strategy,
as
opposed
to
wait,
isn't
that
what
rts
stands
for,
I'm
like
losing
myself
with
my
own
words.
E
You're
talking
about
a
4x
game,
though,
where
there's
a
kind
of
an
overarching
like
large
scale,
map
right,
another
items
that
kind
of
purport
themselves
to
the
gameplay
design
things
such
as
politics
or
you
know,
diplomacy
that
sort
of
stuff.
That's
a
4x
game.
Is
that
what
you're
talking
about.
A
A
Argue
sorry
tune
two
yeah,
the
dune
one
was
totally
different
yeah,
so
so
you've
got
those
on
that
hand,
but
then,
on
the
r4x
side,
you
have
more
stuff.
That's
more
like
you
know,
I
don't
know.
What's
still.
E
I
think
solaris
is
a
real-time
forex
and
there.
D
E
The
largest
scale,
one
that's
popular
right
now,
I
would
say,
would
be
supreme
commander.
There
still
is
a
pretty
dedicated
following
planetary
annihilation,
which
is
a
spiritual
sequel,
is
about
battling
on
multiple
planets,
most
of
the
massive
scale
like
rts
type
games
out.
There,
though,
are
definitely
turn-based
where
everybody
has
a
turn,
and
you
look
through
all
10
000
players
and
everybody
does
their
thing.
E
You
know,
and
it's
kind
of
a
web
browser,
but
when
I'm
kind
of
experimenting
with
cloud
stuff,
the
reason
I
like
an
rts
game
is
because
it's
you
can
have
a
fun
game
with
very
limited
input,
so
your
input
might
be.
You
know
I'm
going
to
send
a
directional
for
my
units,
but
you
can
have
a
simulation
running
that
calculates.
You
know
all
the
vectors
for
the
attacks,
all
of
the
mechanics
between
the
units
right.
You
can
run
that
sim
in
the
cloud
and
you
no
longer
have
to
worry
about
doing
so
much
work.
E
C
E
Exactly
yeah
you
let
that
happen
in
the
server
right
and
I
think
that's
a
really
good
example
of
what
things
like
openshift
are
capable
of,
because
you
can
do
all
that
some
of
the
server
right
you
have
infinite
resources
in
the
cloud
right
and
that's
kind
of
the
selling
point
there,
like
sure,
it'd
be
great
to
build
a
you
know:
full
rpg
simulation
right
to
show
how
cool
all
this
stuff
is,
but
then
you're
doing
so
much
work
on
the
client.
You
know
the
client
has
to
do
so
so
much
work
prediction.
E
It
has
to
be
able
to
have
you
know,
interaction
and
ui
for
all
the
different
interactions
that
you're
gonna
have
in
the
game.
So
if
you
have
like
different
spells
or
abilities
right,
that's
very
expensive
to
create
all
of
those
things.
You
know
if
you
watch
there's
a
lot
of
good
tutorials
these
days
on,
like
using
an
out
of
the
box,
unity
or
unreal
setup
to
build
those
different
types
of
games
right.
E
E
I
really
like
that
as
something
that's
a
good
example,
and
you
can
do
neat
things
too
in
the
cloud
that
a
lot
of
people
don't
think
of,
like
you
can
pre-calculate
physics
in
the
cloud
and
send
that
blob
to
the
client
and
just
have
the
client
render
it
right.
The
client
doesn't
need
to
calculate
physics
same
thing
for
even
like
graphical
effects
right.
You
can
actually
pre-calculate
graphical
effects
in
the
cloud
and
then
send
that
to
a
mobile
device
to
render.
B
B
You
still
have
your
authentication
and
authorization
and
all
that
kind
of
stuff
too
right,
so
I
mean
the
game.
Mechanics,
I
think
certainly
extend
themselves
well
there,
but
I
think
there's
certain
other
segments
that
are
going
to
be
more
common.
That
might
give
us
an
opportunity
to
create
a
platform.
A
Yeah
I
mean
we
had.
We
had
had
a
lot
of
internal
discussion
amongst
ourselves
that
this
group
on
the
on
the
on
the
red
hat
side,
and
we
had
also
talked
about
sign
of
some
kind
of
like
mmo
type
thing
now,
not
necessarily
as
fancy
as
the
current
sort
of
world
of
warcrafty.
A
You
know
age
of
empires
like
what
not
age
of
empires
path
of
exile,
and
you
know
those
kinds
of
like
really
complex
fancy
ones.
I
think
you
know
there's
potential
for
something
simpler,
but
then
that
gets
into
derek
what
you
were
saying
about.
Like
you
know,
the
individual
character
is
the
unit,
but
that
one
unit
has
lots
of
abilities
that
then
have
to
interact
with
other
units
and
abilities,
and
so
perhaps
from
a
complexity
perspective.
E
Another
great
option
to
talk
about
is
a
platformer
right
where
you
can
see
all
the
other
people
playing
the
platformer.
At
the
same
time,
right
I
mean
it
still
runs
you
into
the
problem
of
you
have
to
build
a
platformer
which
is
already
pretty
complicated
right
and
that's
the
main
problem
you're
trying
to
solve
for
in
game
development.
Isn't
you
know
how
do
I
scale
this
right?
That's
a
secondary
concern,
or
how
do
I
build
all
this
infrastructure?
It's?
How
do
I
find
the
fun
and
everything
else
is
secondary
right.
It
feeds
into
that.
E
F
D
D
So
the
goal
of
this
I
mean
the
title
of
the
series-
is
scalable
multiplayer
architecture
using
openshift,
so
I
don't
maybe
not
get
too
hung
up
on
the
actual
genre
or
the
game,
mechanics
or
the
game
architecture
itself.
But
there's
some
like
multiplayer
like
how
can
we
use
openshift
to
make
a
game
scale
infinite?
Well,
almost
infinitely
as
far
as
you
have
resources,
but
so
let's
take
this.
Let's.
What
I
would
like
to
do
is
solve
this
question.
D
How
could
we
have
a
game
running
on
openshift
that
has
like
millions
of
people
playing
at
the
same
time,
not
on
the
same
server,
like
I'm
talking
more
of
like
small
small
instances
like
like
set
game
sessions
with
like
between
two
to
like
20
people,
so.
F
D
So
here's
the
here's,
how
I,
how
I
look
at
it,
there's
there's
the
old
style,
multiplayer
game
like
let's
say
a
centralized
single
cluster
where,
like
hundreds
of
thousands
of
people
are
maybe
are
connected,
not
maybe
100.
You
know,
tens
of
thousands
of
people
are
connected
to
it
right
on
one
and
they
can
and
everyone's
running
on
the
same
server
right.
I
don't
think
that
is
a
good
use
case
to
solve
with
openshift.
D
I
think
a
better
use
case
is
because
it's
monolithic
right,
I
think
a
better
use
case
is
more
like
the
like
a
lot
of
the
current
games
that
have
small
sessions
of
like
8
or
12
people
like
fortnite
or
apex
or
cfds
or
whatever,
but
there's
millions
and
millions
and
millions
of
people
playing
them
all
at
the
same
time.
But
it's
just
all
like
you
know
it's
scaled
out
right,
so
that
kind
of
use
case,
I
think,
would
be
something
openshift
would
excel
at
right.
Being
able
to.
E
And
you
say:
yeah
fortnight
right
chat,
says
hybrid
cloud.
Royale.
D
Okay,
anything
yeah
cloud.
Exactly
so
and
again,
the
genre
isn't
like
is
super
important,
but
it's
more
of
like
the
type
of
multiplayer,
so
a
multiplayer,
where
there's
a
ton
of
smaller
servers
with
like
two
to
twenty
people
or
two
to
forty
or
whatever,
but
versus
like
one
server
with
like
10
000
people
running
on
it,
because
that's
harder.
A
A
Why
is
it
so
hard
to
do
something
like
a
fortnight
with
more
than
100
players
in
the
same
server
instance,
and
somebody
I
forget
who
was
mentioned:
state
right
and
distribution
of
state
and
keeping
track
of
everything?
And
so
I
think
that
there
is
a
a
spectrum
if
you
will
of
distribution
of
stateful
information
over
it's
a
three-dimensional
spectrum,
so
you've
got
the
distribution
of
stateful
information
in
terms
of
the
quantity
of
state,
as
it
relates
to
the
number
of
players
and
the
amount
of
time
that's
elapsed.
A
The
number
of
state
changes
and
calculations
of
of
state
and
interactions
and
physics,
and
all
those
things
goes
through
the
roof.
As
you
increase
the
number
of
players,
and
then
you
have
to
get
into
stuff
where
it's
like,
okay.
Well,
we
can
break
players
into
multiple
servers
that
are
still
in
the
same
game,
but
each
server
is
managing
the
state
for
his
own
and
then
we
have
to
worry
about
like
zone
interactions
and
all
this
stuff.
This
was
a
gdc.
No.
A
This
is
in
the
that
white
paper,
this
white
paper,
the
rpg
white
paper,
and
so
when
you
start
talking
like
what
derek
was
talking
about
with
an
rts
where
the
number
of
state
changes
is
potentially
much
smaller,
so
slow
it
down
to
turn
based
and
look
at
something
like
civilization
right.
So
you
have
a
unit
that
attacks
a
unit
and
there's
basically
a
dice
roll
based
on
some
stats
and
like
either
one
unit
dies
or
it
doesn't.
A
You
could
probably
scale
that
out
to
bajillionity,
because
it
doesn't
have
to
happen
that
fast
and
the
distribution
of
that
stateful
information
probably
wouldn't
be
that
difficult
to
do
even
at
like
really
really
really
massive
scale
in
an
rts
setting.
If
you
sort
of
make
the
game
time
or
a
little
slower.
D
A
E
Don't
want
to
point
out,
like
you
said,
that
kind
of
like
3d
graph,
you
know
with
the
different
variables
in
it.
It's
a
sort
of
solved
problem
actually,
and
this
isn't
something
that's
very
commonly
discussed
in
game
design.
But
basically
the
way
I
like
to
call
this
concept
is
gravity.
State
needs
to
be
extremely
accurate
per
player
and
a
small
region
around
the
player
and
that's
very
strong
gravity
and,
as
you
move
distance
exponentially
from
the
player,
the
state
no
longer
has
to
be
as
accurate.
That's
what
we
call
decreasing
gravity.
E
So,
if
you're
looking
at
like,
let's
say
we're
playing
a
space
game
right
and
you're
playing
as
you're
in
a
fighter
right
and
you're,
along
with
all
these
other
fighters
and
you've
got
these
spaceships
that
are
right
here.
Everything
that's
right
in
front
of
you
right
that
needs
to
have
100
accurate
state,
which
means
the
server
needs
to
be
responsive.
It
needs
to
be
tracking
all
that
stuff
really
well.
But
if
you're
like
looking
and
there's,
you
know,
100
000
kilometers,
there's
a
battleship
coming
in
you
actually
don't
care.
E
If
that's
you
know
a
mile
or
60
miles
off
right,
because
you
won't
be
able
to
tell
visually
yeah,
and
so
this
works
for
a
lot
of
different
games,
even
like
mmos
or
fps's
right.
If
you
build
out
the
scale
enough
after
a
certain
point,
state
no
longer
has
to
be
consistent
or
accurate,
and
that
means
you
can
start
ballparking
stuff
right
and
reduce
server
load,
and
you
no
longer
have
to
maintain
main
state
for
everyone
in
that
area.
E
It
does
break
down
in
certain
cases
like
let's
say
you
have
over
a
million
players
and
they're
all
evenly
spaced
right
at
that
point.
You're,
like
you
know,
the
sim
can't
handle
it
right,
but
that's
where
you
start
looking
at
gameplay
design
that
facilitates
maintaining
a
manageable
state.
You.
F
A
C
C
Players
out
there,
but
if
your
actions
can't
affect
them,
unless
you,
I
suppose
you
take
some
time
to
travel
somewhere
else,
then
it's
almost
like,
like
it's
almost
like
you're,
not
playing
the
same
game
and
and
you're
not
able
to
have.
Let's
say
like
like
in
the
space
same
example
like
you
want
players
to
have
a
warp
ability
where
they
can
warp
instantly
like
halfway
across
the
universe.
C
Some,
suddenly
you
like
the
moment
they
do
that
ability
you
have
to
do
some
really
sudden
shifts
in
where
state
is
like.
You
have
to
move
them
to
say
a
different
pod.
You
have
to
locate
what
god
the
players
in
that
region
are
on
and
switch
them
to
that
and
bring
along
whatever
state
your
ship
is
carrying,
and
so.
D
D
I
think
I
think
that's
I'm
glad
you
brought
that
up,
michael,
because
that,
in
my
opinion,
that's
a
really
good
example
of
how
openshift
can
shine,
because
in
the
space
game
where
you
have
like.
Oh,
let's
see
you
have
a
whole
galaxy.
I've
been
playing
a
lot
of
elite
dangerously
if
anyone's
played
that
game.
D
Yeah
so
like,
when
you
do
a
jump
between
star
systems,
it's
like
that's
kind
of
what
I
was
talking
about
of
like
you're.
It's
it's
everyone's,
not
all
running
on
like
each.
I
think
each
system
has
its
own
like
kind
of
sub
state
and
sub
server,
that's
going
on
like
like,
and
you
can
distribute
that
across,
like
many
different
pods
running
in
the
cloud,
or
something
like
that.
D
So,
like
you're,
saying
when
you're
jumping
between
between
systems,
you're
gonna
go
you're
gonna
connect
to
a
new
node
like
in
the
infrastructure
itself
and
there's
like
a
there's
enough
time
like
it
doesn't
have
to
be
instant.
There
has,
like
a
you,
know,
there's
a
transition
period
between
jumping
between
star
systems.
That
is
plenty
of
enough
time.
I
think,
to
transition
your
character
to
a
new
node
of
the
full
cluster
which
is
distributed
across
everything.
E
And
I
do
want
to
point
out
here
this.
This
conversation
is
really
awesome
because
you
guys
are
thinking
from
an
engineering
perspective
right
from
a
player's
perspective,
yeah
yeah
from
a
game
designer
perspective
from
a
player's
perspective,
it's
about
finding
that
fun
right
and
when
you
say
like
oh
man,
it
sucks
to
have
these
restrictions
and,
like
have
to
think
through,
like
oh,
I
don't
want
to
limit
the
game
design,
but
some
of
the
best
game
designs
have
been
complete,
total
accidents
that
have
occurred
because
of
the
technological
restrictions
in
place.
Right.
E
A
Know
I
was
just
listening
to
a
podcast,
so
reid
sorensen
does
a
podcast
called
designer
notes
and
there
was
an
episode
with
amy.
I
have
to
show
I
have
to
mess
with
my
podcast
app
here,
amy
something
and
I'm
trying
to
remember
what
her
name
was
I'll
pull
it
up
later.
Maybe
we'll
put
it
in
some
notes
or
something
but
like
there
was
a
game
design
where
there
were
two
sort
of
realms
and,
like
the
character,
shifted
between
the
underworld
and
the
regular
one.
A
A
You
know
throughout
a
whole
environment
and
make
sure
that
you
get
that
eventual
consistency
so
again
we're
sort
of
thinking
from
the
engineering
side
out
before
we
have
a
fun
game
but
like
if
the
stateful
information
is
such
you
might
be
able
to
just
put
it
all
in
the
data
grid
and
then
whatever
thing
that
needs
to
interact
with
stateful
information
just
asks
the
data
grid
for
a
particular
piece
of
information
or
just
for
the
delta.
Like
here's.
A
What
I
last
knew
tell
me
what
I
need
to
know
now
and
maybe
there's
some
middle
layer
that
sits
in
front
of
the
data
grid
that
calculates
those
deltas
and
again,
when
we
talk
about
game
clock
versus
reality.
Like
is
there
enough
slowness
to
the
game
where
that
little
bit
of
latency,
of
asking
a
microservice
for
state
delta
and
hitting
the
data
grid
like
if
that's
only
90,
milliseconds
or
100
milliseconds
like
is
that?
Okay,
we
don't
know,
you
know
we
don't
even
have
a
game
yet
so
anyway,
yeah.
E
A
It
fun
I've
been
reading,
I've
been
trying
to
dive
into
gdc,
slide
decks
and
presentations
and
stuff,
and
there's
there
was
one
that
was
about
modern
warfare,
and
they
were
talking
about
exactly
how
they
do
lie
to
the
player,
which
is
like
they
actually
look
at
things
like
which
direction
the
player
was
going
and
how
fast
and
they
use
ai
techniques
to
basically
interpolate
like
well.
A
A
Of
that
nature,
so
yeah
it's
totally
cool
to
make
it
up
as
long
as
the
players
interpretation
of
that
is
okay,
whereas
yeah
like
in
a
financial
thing
like
you,
don't
want
to
make
up
somebody's
bank
balance
that
that
usually
results
do
not.
E
They
brought
up
a
bunch
of
examples
of
like
really
cool
things
that
have
happened
just
because
of
the
technology
restrictions,
so
shout
out
to
chat
for
that
awesome
input.
A
Oh
yeah
no,
and
I
mean
I
kind
of
wanted
to
look
at
some
of
the
ideas
that
somebody
had
had
in
the
chat
so
dweller
tunes
says
you
know
there
were
some
issues
with
like
if
the
game
ran
long
enough
and
the
total
unit
count,
he
was
running
out
of
memory.
Tech
jedi
was
talking
about.
You
know
we
probably
shouldn't
worry
about
like
sharding
players
into
groups
or
server
sides
or
whatever,
because
who
knows
what
the
eventual
game
is
going
to
be
so
yeah.
There's
there's
totally
a
bunch
of
interesting,
interesting.
D
Yeah
dungeon
instancing,
that's
kind
of
like
what
I
was
talking
about,
which
is
not
having
every
one
connected
to
the
exact
same
server
all
the
time
and
managing
state
that
way,
but
having
a
whole
bunch
of
smaller
instances,
with
less
less
connect
like
concurrently,
connected
players
to
the
same
node
or
instance
that
you're
on,
but
then
you
can
have,
but
that
you
can
have
you
can
use
openshift
then
to
scale
out
as
many
of
those
as
you
want.
Like
that's
what
I
I
think
we.
A
No,
I
think,
yeah.
So
what
we're?
What
we're
getting
at
is
kind
of
a
conversation
of
of
things
that
people
have
said,
which
is
we
can
have
truly
massive
scale
in
terms
of
the
game
universe,
but
potential
player
interaction
at
the
whatever
ends
up
being
the
server
level
may
be
lesser
number
of
players
depending
on
various
factors.
So
we
will
probably,
if
we
do
this
right,
we'll
be
able
to
end
up
with
both
millions
of
simultaneous
players
in
a
universe.
A
Yes,
but
area
of
influence
and
interaction
may
be
on
a
much
smaller
scale
and
and
going
back
to
the
space
theme,
it's
like
using
the
dangerous
or
star
wars
or
whatever
you
want.
You
know
we
can
have
a
game
design
element
where
I
can
move
across
great
distances
of
space
in
short,
relatively
speaking
periods
of
time,
but
we
can
use
things
like
fog
of
war
or
information
horizon
so
like.
A
A
Yeah
yeah,
like
how
good
are
your
scanners?
How
far
can
you
go
right?
So
I
guess,
if
we
step
back
from,
like
you
know,
diving
into
technology
problems
right,
so
is
anybody
really
horribly
opposed
to
some
kind
of
rtsr-4xish
thing
like
raise
your
hand
if
that
sucks,
you
know
I'm
fine.
D
A
Okay,
michael,
what
kind
well
we'll
get
we'll
get
into
that
right,
so
we've
got
like
I'm,
I'm
neutral
derek
was
the
one
who
suggested
rts.
So
that's
one
vote
for
for
rts.
E
D
But
maybe
I
don't
unders,
maybe
I
I
didn't
understand
the
expectation
of
like
what
we
were
going
to
produce
here
like
when
we
were
talking
about
internally.
We
were
talking
more
about
setting
up
the
architecture
and
not
even
like
building
a
fully
functional
game,
because
we
don't.
D
A
E
Yeah-
and
I
want
to
say
like
when
we're
talking
about
finding
the
fun
for
non-game
developers
right,
that's
kind
of
like
a
myriad
of
oh.
What
does
that
mean
for
game
devs?
You
can
find
the
fun
in
five
minutes
right
and
there's
a
lot
of
companies
that
do
things
that
way.
For
example,
a
really
good
instance
of
this
is
bungie
when
they
were
designing.
You
know
their
awesome
destiny,
raid
right
and
they're.
Like
you
know,
this
is
ridiculous.
E
We
can't
how
do
we
find
the
fun
you
know
with
all
this
stuff
that
we
got
to
deal
with.
They
just
went
into
the
gym
and
they
grabbed
some
dodge
balls
right,
they're
like
what's
fun.
What
is
fun
to
do
as
a
person
right
interacting
with
other
people?
You
know-
and
maybe
we'll
have
some
guns
or
something
with
this,
but
that's
how
they
prototype
out.
What
is
fun
in
the
game?
So
we
don't
need
to
build
a
graphical
user
interface.
We
don't
need
to
build.
You
know
crazy
graphics.
E
We
don't
need
to
do
all
this
extra
stuff
that
you
might
be
kind
of
thinking
of
as
fun
in
the
game
right.
Those
are
the
icing
on
the
cake.
The
core
level
of
fun
right
is
going
to
be
a
very
simple
gameplay
element.
Is
it
fun
to
watch
your
stuff
go
boom?
You
know
right,
that's
fun!
That's
a
known
fun
thing.
Is
it
fun
to
have
a
spaceship
and
like
see
it
fly
through
space,
shooting,
lasers
and
rockets?
Right?
That's
that's
fun!
E
A
A
Like
when
all
your
actions
are
used
and
with
you
know-
and
you
find
some
crazy
like
you
know,
weird
thing
that
you
can
do
with
an
object
that
interacts
with
another
object,
you
know
that's
what
made
it
fun
is.
You
know
it
was.
It
was
a
thought
experiment.
In
many
cases,
or
in
my
case
it
was
like
you,
you
used
a
bot
that
you
weren't
allowed
to
use,
and
then
you
got
banned
from
the
mud
because
you
had
a
tank
that.
F
A
B
I'm
kind
of
I'm
kind
of
neutral,
I
mean
I
I
do
like
the
idea
of
rpgs.
I
love
rpgs,
I
love
rts.
It's
played
the
heck
out
of
total
annihilation.
You
know
play
pool
radiance
back
in
the
day
either.
One
of
those
genres
is
good.
I
don't
think
I'd
try
to
do
a
first
person,
shooter
or
racing
sim.
I
don't
think
that's
that's
a
fit,
but
I'm
kind
of
a
little
bit
more
on
the
technical
angle.
I
really
love
the
programming
and
the
development
angle.
A
Yeah
well,
and
so
the
reason
I
say
that
picking
the
genre
is
important
because
again
using
that
sort
of
thesis
paper
as
the
example
like
he
chose,
an
rpg
and
rpgs
have
specific
thematic
elements
that
necessitate
certain
architecture:
components
where
something
like
an
rts
may
not
have
those,
and
when
I
was
thinking
about
like
we
had
originally
talked
a
little
bit
about
mmo
and
I'm
thinking
like.
Oh,
how
would
you
actually
have
like
a
microservice
that
handles
you
know?
I
don't
know
player
leveling
like?
Would
that
be
its
own
microservice
or
something
like
that?
E
A
Jared
and
I'm
not
totally
not
trying
to
pick
on
you
jerry
but
like
what
is
it
that
you
don't
like
about
the
idea
of
doing
an
rts,
or
is
it
more
that
like?
Maybe
it's
just,
you
would
be
fine
with
it
if
we
had
a
little
bit
more
thematic
or
or
story
sort
of
sussed
out,
because
me
I
hate
rts
as
they
move
too
fast,
but
if
there
was
like
a
really
slow
rts
that
was
space
themed.
That
was
almost
more
r4x
like
I
could
be
down
with
that.
D
So
this
I'll
tell
you
a
lot
like
my
my
main
reasons,
why
I'm
not
in
favor
of
an
rts,
and
I
okay,
so
for
first
thing-
is
that
a
lot
of
rts
is
like
the
a
lot
of
the
rts.
Most
experience
is
a
1v1
kind
of
thing
like,
and
you
don't
even
need
to
have
like
multiple
like
it
can
just
be
like
you,
you
peer
to
peer
with
someone
else
right,
so
that
doesn't
really.
That
part
is
really
like.
You
know
it
doesn't
really
need
to
be
scaled
out.
D
It
works
just
fine
like
without
any
kind
of
infrastructure
improvement
at
all,
like
the
other
thing
is
that
if
we
want
to
like
showcase
our
like
re,
how
red
hat
can
help
big
companies
with
major
scalability
issues
that
they
have?
I
can't
think
of
any
any
like
studios
fighting
with
this
right
now
in
the
rts
space
start
starcraft
has
been
around
for
10
years,
but
the
like
the
new
popular
games.
I'm
sorry
like
starcraft
is
a
great
game,
but
it's
it's.
It's
an
old
game.
D
A
So
a
couple
things
I
think
one
you're
looking
at
the
existing
genre
and
what
hasn't
been
done
and
basically
declaring
that's
like
the
way
it
works
kind
of
thing.
So
you
know
yes,
rts's
are
mostly
smaller
scale
but
like
there
was
a
time
before
battle,
royale
existed
and
then
when
people
decided
that
battle
royale
was
cool.
Suddenly
everybody
wanted
to
do
it
and
so,
like
maybe
there's
room
for
a
really
massive
thing
that
looks
rts
like
like.
Maybe
there's
a
way
to
redefine
that
yeah.
C
C
A
Yeah
I
mean
fortnite
league
of
legends
like
battle,
royale
style
fps's.
You
know-
and
I
think
those
to
a
certain
extent,
though,
are
all
technologically
to
a
certain
extent
outside
of
the
wheelhouse
here,
unless
we
just
completely
ignore
certain
aspects
of
the
game.
But
then,
if
we
ignore
certain
aspects
of
those
genres
like
are
we
actually
proving
the
point
that
you
can
do
scalable,
microservices
or
whatever
so.
A
A
But
if
you
talk
about
a
shooter
and
a
functional
skeleton
for
that
like
without
some
level
of
physics
in
3d
or
whatever
like
you,
you
really
can't
prove
anything
and
that
gets
probably
I
don't
know
what
your
technical
skills
are.
But
you
know
I've
got
none
so
yeah.
I
do
want
to
point
out
like.
E
Target
audience
wise
right,
our
target
audience,
isn't
the
person
who's
building
the
next
battle,
royale
at
epic,
right
or
valve,
or
whatever
right.
Our
target
audiences
are
yeah.
Why
not?
Because
they're
already
doing
it
right,
our
target
audiences
are
the
developers
that
are
thinking
hey.
I
need
to
scale
my
service
right
now.
I
need
to
scale
this
out
for
the
game
that
I'm
building
and
so
to
me
like
when
we
say
like:
let's
do
an
rts
or
something
like
to
me
in
my
head.
E
The
most
important
thing
is
that
we
keep
our
client
simple
and
we
keep
we're
implementing
simple,
because
I
don't
think
I
have
time-
and
I
know
you
all
don't
have
time
to
sit
down
and
unreal
right
or
you
know,
source
or
unity
and
go
build
an
fps
client
right,
but
something
2d
right.
We
can
whip
that
up
in
godot
or
we
could
even
do
multiple
clients.
If
we
want
right
to
show
like
hey,
we
can
scale
the
server
right
and
you
can
connect
as
many
clients
as
you
want.
E
You
could
upgrade
them
over
time
right
with
this
persistent
universe,
and
I
think
it's
important
instead
to
focus
on
one
keeping
the
basis
of
what
we're
doing
simple
and
then
two
we
can
add
things
to
it.
Right,
like
we
mentioned
earlier,
that
you
know,
tall
nation
has
a
leveling
system
right,
well,
who's
to
say
our
rts
units
or
our
rts
single
unit
right.
We
could
do
a
single
unit,
it
doesn't
have
an
inventory,
it
doesn't
have.
A
leveling
system
doesn't
have
quests.
E
It
needs
to
do
in
this
giant
universe
with
thousands
of
other
players
or
whatever
right,
there's
a
future
like
that.
There's
all
sorts
of
possibility
there,
and
when
we
look
at
like
instances
where
we're
saying
hey,
we
want
to
showcase
this
particular
feature
of
open
shift
like
like
data
grid.
Right,
we
want
to
say,
like
hey,
you
could
have
a
massive
inventory
or
you
could
have
like
a
virtual
economy
system
that
runs
on
openshift
and
or
data
grid,
and
any
player
can
access
it
at
any
time
right.
E
We
add
that
on
right,
we're
not
trying
to
make
a
game,
you
know
to
sell
or
anything
right,
we're
making
a
game
that
we're
you
know
showing
what
openshift
can
do
and
so
to
me,
keeping
like
the
base
mechanic
of
the
game.
Very
simple
is
the
most
important
thing
to
me.
So
that
way,
when
we
do
add
stuff
on
right,
we're
not
like
dealing
with
like
such
a
complex
overlapping
of
systems
that
we've
written
ourselves
into
a
corner
for
demo
purposes.
A
Yeah
and
that's
why
I
I'm
I
kind
of
for
selfish
reasons.
I
like
the
idea
of
kind
of
rts
r4x
space-ish.
You
know
simulation-y
thing,
because
the
game
can
move
slow-ish
and
still
have
you
know
a
truly
massive
scale.
Long-Term
persistent
state,
you
know,
but
you
can
get
created
like
you-
can
keep
adding
complexity
that
doesn't
require
making
the
game
client
more
complex,
but
creates
other
demands
on
the
subsystems
and
scalability
and
all
these
other
things.
So
we
were
talking
about
earlier.
Like
imagine
a
perfect
example
right,
so
really
dangerous.
A
A
And
you
bought
things
in
one
place
and
you
sold
them
in
another
place
and
there
was
combat,
but
it
was
basically
like
you
know,
attack
other
ship
period.
You
know
hit
enter
and
then
it
would
roll
some
dice.
And
so,
if
you
sort
of
extend
that
to
be
more
of,
like
quote,
unquote,
rts
well,
okay,
there's
not
really
turns
there's
just
a
game
clock.
That's
constantly
running,
you
can
say
I
want
to
move
these
units
to
these
locations
and
then,
by
the
time
they
get
there.
A
You
know
whatever
thing
it
can
just
be
like,
like
civilization,
you
know
five,
where
the
tiles
just
move
on
to
one
another
and
then
there's
like
a
little
explosion,
and
you
know
there's
a
resolution,
so
I
I
think
you
know
I
don't
want
to
get
too
hung
up
on
just
the
term
like
rts,
because
that
does
connote
something
like
a
starcraft
or
a
dune
or
a
command
and
conquer.
Where
it's
like
all
right.
A
I
need
to
pick
up
these
resources
over
here
and
then
you
know
use
this
in-game
object,
which
is
like
a
building
to
convert
it
into
something
else,
like
I
think
the
general
thematic
rts
is
more
like.
I
have
some
resources
that
I
convert
into
stuff
and
then
I
have
units
that
can
move
around
and
interact
with
the
universe,
and
I
think.
F
E
E
C
Guys
think
about
about
this
idea,
what
about
we
start
with
no
genre?
We
start
with.
D
D
D
And
even
further
than
that,
like
michael,
I
like
that
so
much
better,
like
going
back
to
the
original
thing
is
like
how
can
you
have
like
millions
of
people
playing
your
game
at
one
at
concurrently,
right,
the
and
then
the
genre
like?
Doesn't
really
I
mean
it's?
D
It's
I
like
get
rid
of
the
genre
and
just
say
like
how
could
we
have
a
million
people
playing
your
game
concurrently
at
one
time
and
and
they
have
a
cap
like
every
server
has
like
five
players
in
it,
and
then
that
could
be
a
forget
about
the
genre
that
could
be
a
rts
that
could
be
a
space
sim.
That
could
be
a
whatever
battle
royale.
B
B
D
E
I
think
there's
probably
a
little
bit
of
confusion
here
now.
I
understand
what
you're
saying
we
weren't
saying
rts
and
we
have
to
make
a
game
that
competes
with
starcraft
or
looks
like
starcraft
or
something
right.
We're
saying
rts
is
like
as
a
generic
set
of
features
that
we
want
to
build.
We
need
to
have
you
know
one
or
more
units
that
a
player
can
control
or
influence
and
some
sort
of
resource
management
and
conflict
with
other
players
right
and
it's
in
real
time
right.
There's
no
like.
E
And
that
doesn't
mean
like
these
three
things
could
like.
Those
three
features
could
be
to
any
genre
right,
but
we're
trying
to
we're
not
trying
to
say
we're,
trying
to
make
an
rts
we're
saying
as
a
starting
genre.
Here
are
some
features
we
should
start
building.
You
know
as
a
baseline
resource
management
being
able
to
control
one
or
more
units
and
being
able
to
interact
with
other
players
in
conflict,
so
it
doesn't
have
to
be
an
rts
right
and
in
fact,
if
you
look
at
like
halo,
for
example,
halo
started
as
an
rts
game.
E
It
was
built
in
the
myth
engine.
It
was
built
entirely
as
an
rts
setup,
and
then
they
switched
it
to
third
person
shooter
because
they
found
that
was
more
fun
and
then
they
kept
playing
this
third
person
shooter
they're
like
yeah,
but
I
want
to
like.
I
want
to
be
the
guy
in
the
green
suit,
because
this
unit
is
really
cool.
This
one
particular
unit
is
super
cool
and
now,
of
course,
you
know
what
happened
right.
E
D
D
I
think,
at
the
end
of
the
series,
not
not
only
like
not
having
a
fully
built
game,
just
like
everyone
is
saying
we're
not
building
a
full,
fully
fledged
game
here,
but
if
we
take
a
a
load,
a
load,
balancer
script
or
a
load
testing
script
and
like
set
it
up
in
the
cloud
and
have
a
million
clients
join
this
game
or
connect
to
games,
say
play
and
have
those
clients
end
up
in
an
instance
or
a
note,
or
something
with
like
five
other
clients
and
and
have
it
not
crash
and
have
it
work
like
yeah.
E
A
We've
gotta
start
somewhere
right,
and
so
it's
like
you
know
so
so
far
high
level
thematically
we've
got
some
some
concept
of
resource
management,
some
concept
of
units.
You
know
the
the
units,
the
sorry,
the
players
via
their
units,
interact
2d
with
a
potentially
you
know.
I
don't
know
how
we
would
accommodate
an
infinite
map
size.
That's
going
to
be
interesting.
You
know
we
might
have
to
spawn
more
map
as
more
players
join
to
give
them
their
own
sort
of
base.
A
If
you
will
and
then
then
it's
weird,
but
who
knows
right-
maybe
some
kind
of
space
theme-
and
I
think
the
space
theme
is
a
good
idea,
because
it's
much
more
from
a
game
design
perspective
and
like
going
back
to
what
derek
said
about
having
fun
it's
much
more
easy
to
mentally
deal
with
the
idea
of
space
being
infinite
than
something
ground-based,
because,
like
theoretically
the
universe
is
infinite.
A
A
You
know
in
terms
of
game
elements,
theme
theme
elements
if
you
will
so
with
respect
to
units
in
terms
of
like
designing
for
the
opportunity
for
complexity.
I
know
that's
a
weird
way
to
talk
about
it.
Do
you
do
we
want
to
have
like
fungible
stats
for
units,
meaning
like
as
as
derek
was
alluding
to
with
total
annihilation
and
some
of
these
other
games,
where
the
units
have
some
kind
of
level
or
skill
or
veterancy
or,
like
you
know
other
than
just?
A
E
I
think
we
want
to
account
for
the
possibility
that
we
can
add
arbitrary
components
to
our
units,
so
that
means
it
could
have
an
inventory
in
the
future.
Let's,
let's
design
this
in
a
way
where
we
could
enhance
our
units
any
which
way
we
want,
it
could
be
a
single
unit
right.
We
just
have
one
unit
the
whole
game.
Maybe
it's
a
massive
unit.
It
has,
you
know,
50
cannons
on
it
or
something
we're
just
playing
battleships
at
this
point
right
or
you
know,
maybe
it's
a
swarm
and
each
one
gets
one
enhancement.
E
You
have
to
decide
which
you
know
there's
plenty
of
opportunities,
but
I
think
componentizing,
the
entities
in
the
game
that
the
player
can
control
so
that
they
can
be
enhanced
in
arbitrary
ways,
is
going
to
be
important
because
that's
something
that
people
who
are
engineering
servers
for
games
have
to
think
about.
Is
you
know
if
my
game
designer
comes
in
next
week,
and
it's
like?
Oh
my
god,
I
found
this
amazing
thing.
It's
so
much
fun.
We
just
need
to
add
all
this
stuff
to
our
units.
E
E
A
May
not
even
want
to
spin
up
a
pod
for
every
player.
You
know,
but
again
we
might
have
like
a
scale
out
unit
service
where
the
data
grid,
or
one
of
many
data
grids,
is
keeping
track
of
unit.
A
Even
know
right,
like
I
think
I
think,
we're
it's
even
too
early
to
think
about
pods
or
or
services,
or
even
at
that
level,
right.
We're
still
sort
of
thinking
about
thematic
elements
and
the
interesting
thing
I'm
thinking
about
as
we
talk
about
these
things
is
like
unity
is
a
game
engine
you
know.
Good
o
is
a
is
a
game
engine
there's
all
these
like
game
engines
that
control
more
the
client
side,
to
a
certain
extent,
the
server
side.
If
you
look
at
openshift
like
what
is
it?
A
A
B
C
And
when
it
comes
to
the
state
state,
synchronization
architecture
in
open
shift.
C
D
Think
there's
like
their
short-term
state
and
there's
like
persistent
state
like
when
it
comes
to,
like
even
thinking
of
like
an
rts,
for
example
like
when
you
finish
when
you're
playing
the
game,
their
state
going
back
and
forth.
But
when
you
finish
then
there's
persistent
state
like
when
like
did
you
win
or
did
you
lose,
you
know
what
did
your
character
gain
experience
that
needs
to
be
persisted
after
that?
So
there's,
like
I
think,
persistence
like
generically
of
course
needs
to
be
definitely
needs
to
be
part
of
this,
like
like
long-term
persistence,
yeah
yeah.
E
E
A
A
In
the
clash
of
clanslight
model,
like
things
can
happen
to
you,
when
you
are
not
in
the
game
and
in
the
elite
dangerous
model,
that's
not
really
the
case
right.
The
universe
persists.
Players
come
in.
They
manipulate
the
universe,
the
state
of
the
universe,
changes
they
leave.
You
can't
kill
them
when
they're
not
there,
whereas
the
clash
of
clans.
D
A
E
There's
there's
a
lot
of
questions
that
you
kind
of
there's
a
lot
of
questions
you
tied
up
in
there
by
bringing
in
like
clash,
clans
or
something
like
that
right,
like
the
first
one
is
just
like.
Is
your
player
unit
or
units
plural,
persistent,
while
you're
on
or
offline
right,
which
is
that's
a
game
design
question
to
me:
that's
not
even
like
an
architecture
design,
it's
not
a
genre
design.
That's
just
game
design.
You
know
figure
that
out
what's
more
fun
for
people.
E
The
second
part
of
that,
though,
is
like
what
stylistically
is
the
implementation
for
how
people
are
going
to
interact
with
the
game?
Is
there
a
persistent
state
and
then
you
have
matches
off
of
that
or
and
to
me,
like?
The
answer
is:
keep
it
simple
stupid?
Let's
not
worry
about
matches,
not
let's
not
worry,
about
match
making
right.
There
are
services
out
there
that
do
that
really.
Well,
there
are
companies,
so
literally
that's
what
they
do.
E
Is
they
make
matchmaking
services
and
sell
that,
to
you
know,
game
devs
and
that
sort
of
thing,
but
to
me
like
just
focusing
on
keeping
it
really
really
simple
one
universe.
You
know
at
a
time
right
instead
of
worrying
about
matches
in
a
home
base,
or
whatever
is
probably
the
right
way
to
handle
it
for
now.
Well,.
A
A
If
it's
like,
post-universal
apocalypse,
where,
like
something
crazy
happens
to
space-time
right
and
so
a
player
connecting
to
the
game,
is
effectively
like
a
a
fragment
of
the
universe,
has
re-entered
space-time
and
so
now
they're
able
to
interact
with
you
know,
whatever
other
people
are
connected
to
the
system
at
that
time,
because
everybody's
just
like
in
all
these
weird,
fragmented
space
time,
continuums
that
keep
coming
together
and
then
breaking
apart.
So
like
a
player
leaving
the
game
is
really
just
their
space
time
like
disappearing.
A
A
2020
says
just
be
ready.
F
D
D
I,
like
I,
like
keeping
it
simple,
which
is
any
any
multiplayer
game
that
I
know
well.
Most
of
multiplayer
games
that
are
fun,
have
some
sort
of
long-term
partic
persistence
in
them.
So
our
example
could
have.
Our
demo
can
have
some
sort
of
simple
persistence,
whether
it's
like
a
win
loss
ratio
or
something
like
that.
Just
something
simple.
A
D
A
Of
of
of
events
right,
so
I'm
gonna,
I'm
gonna,
you
know
some
kind
of
event
and
what
I
mean
by
events
is
like,
and
those
are
interactions
between
two
or
more
players
right.
So
an
event
that
occurs
is
when
two
or
more
players,
and
perhaps
the
two
or
more
players
interacting,
is
a
proxy
via
their
units
right.
So
that's
really
maybe
just
units
interacting
that's
an
event
of
sorts,
and
so
maybe
the
persistence
is
you
know.
A
F
A
The
the
the
performance
stats
in
an
rts
allah
starcraft
are
usually
used
for
matchmaking
purposes,
like
my
performance
characteristics.
Define
me
as
a
player,
therefore
put
me
in
with
players
who
look
like
me
to
a
certain
extent,
so
that
I
don't
get
creamed
and
so
that
I
don't
totally
cream
the
other
person.
You
know
targeting
an
ideal
win-loss
ratio
of
maybe
let's
say
50
right,
so
you
you're
always
evenly
matched
elite.
A
Dangerous,
statistically,
like
the
persistent
game
universe
is
affected
or
or
I
am
affected
by
my
statistics-
within
the
persistent
game
universe,
for
example.
If
I
do
a
lot
of
missions
supporting
one
faction,
then
that
faction's
enemies
may
decide
to
ambush
me
whenever
they
encounter
me
just
because
I
look
too
friendly
to
their
to
you
know
it's
like
the
enemy
of
my
enemy
is
my
friend
kind
of
thing.
A
A
I
think
those
are
going
to
be
more
interesting
than
not
that
we
don't
want
to
keep
track
of
like
when
law
stats,
but
as
an
example
like
if
I,
if
I
go
and
interact
with
another
player's
units
and
destroy
all
those
units
like
does
that
have
some
kind
of
effect
on
the
universe.
Does
it
shrink
the
player's
area
of
effect?
Does
it
increase
my
area
of
effect?
Like
I
don't
know,
those
are
sort
of
thematical
things
that
that
we
need
to
figure
out
as
we
sort
of
choose
as
we.
C
Yeah,
I
think
one
way
to
keep
it
simple
would
be
to
be
kind
of
an
easy.
Come
easy
go
game
where,
like
maybe
you
log
in
and
you
have
some
sort
of
persistent
record
leaderboard
kind
of
entry,
but
each
time
you
begin
playing
the
game,
you
start
with
like
a
fixed
loadout
of
stuff,
and
then
you
have
to
go
out
and
and
harvest
and
stuff.
So
it's
kind
of
a
short-term
thing
more
like
the
the
io
games.
C
E
That's
exactly
what
you're
talking
about
it's
an
fps
game
kind
of
like
pub
g
or
something
like
that,
but
it
is
exactly
that
mechanic
where
you
go
in
and
in
order
to
level
up
your
character,
you
have
to
go,
collect,
stuff
and
then
escape,
and
if
you
die
right,
you
leave
your
stuff
right.
Everyone
else
can
pick
over
what
you've
left
and
if
you
make
it
out,
you
get
to
keep
those
things
persistently
and
then
decide
do
I
want
to
go
back
in
with
my
new
kid
or
do
I
want
to
save
it.
C
E
A
I
think
there's
only
so
far
we're
gonna
get
without
really
nailing
like
a
game
theme,
because
it's
really
hard
to
design
a
so
another
one
of
these
spectrum
kind
of
things
as
the
genericality,
which
is
not
a
word,
but
you'll
go
with
it.
As
the
generic
goldness
of
the
system
increases
the
ability
to
be
really
good
at
providing
a
game.
Experience
probably
decreases
right.
It's
like
a
swiss
army.
Knife
is
kind
of
okay,
but
it
usually
sucks
at
everything.
A
A
A
Yeah
or
at
least
as
a
location
for
the
theme,
so
you
know
let
me
I'm
gonna,
take
that
away
from
a
question
mark.
I'm
gonna
put
that
as
a
very
first
thing
as
definitive.
So
you
know
there
is
a
theme
of
space.
I
think
2d
is
also
I'm
gonna
actually
do
this.
A
E
Yeah
so
here's
the
pitch
you're
a
player
you're
on
your
cell
phone.
You
fire
up
the
game
and
you
launch
in
and
you're
in
your
spaceship
right
and
you
come
in
and
you
see
some
other
people
fighting
and
you're
like.
Oh
man,
that's
a
good
one.
I
can
go
for
that,
so
you
fly
your
spaceship
over
to
them
and
you
pick
off
the
remains
of
that
battle,
and
now
you
get
to
loot
so
because
you
didn't
totally
blow
up
the
spaceships
right.
E
You
just
you
know,
took
them
out
enough
for
those
players
to
have
to
respawn.
You
now
get
to
loot
those
ships
and
you
take
those
components.
You
upgrade
your
ship
with
those
resources
and
now
you
go
in
and
now
you
can
just
go
straight
into
a
battle
and
you
can
see
somebody
hey,
maybe
they're
at
full
health,
maybe
they're
not
in
the
middle
of
fighting
right,
so
you're.
Now
in
this
galactic
space
you
know
battle,
that's
just
ongoing
right
forever!
Imagine
warhammer
40k!
E
The
fight
always
continues,
and
so
that's
your
point
as
a
player
right
is
to
see
how
far
you
can
keep
going.
Upgrading
and
moving
your
ship
around
just
very
simple
right,
stealing
components
from
other
players
or
even
ai
ships
that
have
spawned,
and
maybe
one
component
lets
you
spawn
fighters
and
another
component
is
a
giant
cannon
and
you
got
to
place
those
on
your
ship
as
you
go
and
trying
to
upgrade
and
hey.
If
you
get
blown
up,
then
you
got
your
experience
right
and
maybe
got
some
cash
that
lets.
E
You
start
when
you
respawn
at
a
little
bit
better
position
and
we'll
spawn
the
you
know
the
most
powerful
players
right
in
the
middle
of
the
map
and
the
newer
players
right
they're
at
the
edge
where
it's
you
know
they
get
more
of
the
pickings
and
they
kind
of
work
their
way
towards
fighting
towards
the
middle.
So
that's
your!
You
know
simple
gameplay
loop
right.
I
got
a
spaceship.
I'm
going
to
blow
up
other
people
loot
their
stuff,
so
I
can
upgrade
it
and
keep
fighting
my
way
towards
the.
A
Something
that's
bad,
though
right
or
something
yeah
I
mean,
but
it
kind
of
has
a
lot
of
those
elements
right
like
instead
of
the
map
being
the
thing
that
shrinks
like
the
the
more
powerful
players
end
up
in
the
center,
and
so,
if
you
want
to
be
like
the
top
dog,
you
kind
of
got
to
go
towards
the
center.
You
know
the
loop.
A
Whatever
right
yeah,
I
mean
I,
I
I
think
it's
interesting,
I
think
it
it
has.
Some
of
the
things
like
the
unit
is
basically
n
equals
one
and
it's
my
ship
and
it
has
all
these
different
stats
or
attachments
or
whatever.
So
that's.
A
Yeah,
it's
just
a
question
of
like
from
from
a
player
interaction
perspective
like
how
do
the
fighters
get
used?
I
would
think
from
a
simplicity
perspective.
They
would
just
be
a
stat
that
contributes
to
the
battle.
E
A
D
A
E
Know-
and
this
solves
a
secondary
game
design
problem,
which
is
that
of
difficulty
right.
You
want
players
to
be
able
to
choose
the
level
of
difficulty.
Some
players
really
want
to
challenge
some
players.
You
know
want
things
easier
right
if
the
game
increases
in
difficulty
as
you
go
towards
the
middle
and
if
it's
too
hard
for
you,
you
can
just
go
back
out
right.
A
Well,
and
so
that
gets
us
back
to
the
like,
jumping
like
hyper
spacey
kind
of
trope,
where
you
know
if
I
want
to
jump
to
a
really
hard
level
so
have
any
of
you
played
spaz
space
pirates
and
zombies
nope.
So
it's
it's
like
it's
a
it's
like
a
shooter,
basically
top
down.
A
Remember
remember
what
combat
space
combat
was
like
in
the
old
star
control
games
in
star
control,
2
and
then
even
the
new
star
control,
where
it's
like
you
and
another
ship,
and
you
just
kind
of
fly
around
with
the
keyboard
and
try
and
shoot
at
each
other
or,
like
I
don't
know,
joust
from
a
million
years
ago.
So
but
one
of
the
things
in
spaz
was
like
you
would
be.
A
If
you
wanted
to
go
to
a
new
area,
it
would
tell
you
what
it
thought
your
chances
were
of
surviving
going
there
kind
of
thing,
and
so
maybe
that's
like
the
trophy,
like
long-range
scanner,
hyper
spacey
element.
It's
like!
Oh,
like
you
could
jump
to
ring
seven
but,
like
you
might
get
your
butt
handed
to
you.
D
Yeah
I
like
this
idea,
but
one
thing:
I
think
that
we
in
order
to
make
it
scale,
so
it
would,
we
would
have
to
have
some
sort
of
like
hyperspace
jump
between
sectors
like
you
were
talking
about
eric.
A
I'm
thinking
of
a
game,
design
element
right
so
like
if,
let's
assume
we
end
up
with
a
million
players
right
and
and
we
go
with
the
concentric
rings
idea,
because
we're
in
2d
space.
Well,
if
there's
a
million
players
equally
distributed
like
how
many
jumps
or
or
moves,
does
it
put
me
from
the
game
center?
A
Well,
I
mean
we
have
to
move
somehow
in
order
to
fit
derrick's
theme
of,
like
you
know,
moving
towards
the
center
so
to
speak.
I
think
that
the
question
is
just
more
like,
as
the
scale
of
the
universe
increases
with
the
number
of
players.
How
do
we
keep
the
game
fun
and
I
think
that
forcing
a
new
player
to
have
to
you
know
basically
take
an
extremely
long
time,
because,
just
because
they
joined
the
game
late
like
that,
doesn't
sound
like
fun.
B
A
E
A
B
E
Yeah
I
mean
the
persistence
of
the
universe,
doesn't
necessarily
have
to
be
in
the
physical
universe,
space
itself
right,
it
can
be
in
the
inventory
it
can
be
in
the
players
like
starting
cash
values.
That
sort
of
thing,
I
think,
some
of
the
most
popular
games
out
there
right
have
a
gameplay
time
of
20
to
40
minutes.
Yeah.
E
The
beginning,
if
you
can
design
the
game
where
it's
feasible,
to
get
to
the
middle
in
20
to
40
minutes,
you
know,
maybe
the
upgrades
are
that
quick
right.
Maybe
the
gameplay
itself
is
like
a
little
bit
slower
but
like
as
soon
as
you
blow
up
another
player
right,
that's
a
pretty
pretty
big
leap
for
you,
so
you
can
just
start
pushing
closer
to
the
center.
E
You
know
if
we,
if
we
design
the
game
where
it's
like
hey,
you
can
make
it
from
the
edge
to
the
center
in
about
40
minutes
or
so
right,
and
then
declare
yourself
the
winner
as
you
capture
to
the
middle
or
something
like
that
and
hold
it
for
enough
time
and
then
the
universe,
collapses
right
you
get
to
keep
all
the
loot
you've
earned.
You
get
to
turn
that
all
into
you
know
cash
and
upgrades
or
whatever
and
then
start
again
right.
E
If
you
want
to
play
another
session,
you
know
another
40
minutes
to
battle
your
way
in
the
middle,
but
hey.
Maybe
this
time.
Somebody
else
blows
you
up
right,
you're
on
to
the
next
game,
so
there's
there's
definitely
some
opportunities
to
kind
of
finesse
which
way
we
want
the
game
design
itself
to
go
to
make
the
experience
fun
for
players.
A
Yeah
I
mean
that
that
makes
the
so
that
that
you're,
basically
describing
persistent,
loot
and
persistent
skill,
not
really
a
persistent
universe.
I
think
from
a
massive
scale
like
again
going
back
to
elite
dangerous
in
games
of
that
ilk
like
the
you
may
only
play
for
20
to
40
minutes
or
four
hours
or
whatever,
but
the
things
that
you
do
in
that
universe
are
are
permanent,
so
to
speak.
You
know
if
you
help
a
faction
move
from
being
minority
to
majority
in
a
particular
system.
A
You
know
because
of
your
and
other
players,
actions
like
that
has
a
meaningful,
longer
term
impact
on
that
particular
system
and
then
the
next
player
that
comes
through,
who
yesterday
maybe
was
aligned
with
the
leading
faction
all
of
a
sudden
today
finds
themselves
not
aligned
with
the
leading
faction
and
now
they're
screwed,
because
now
their
trade
route
became
more
dangerous
or
whatever
so.
E
D
E
Right
it's
well,
it
doesn't
have
to
be
a
sandbox
either
right
but
think
of
world
of
warcraft.
How
do
you
win
world
of
warcraft?
You
don't
right.
The
whole
point
is
that
you
get
to
go
on
new
and
interesting,
varied
experiences
with
your
friends
so
for
this
particular
game
right,
if
we're
going
to
have
a
persistent
universe.
In
my
opinion,
that
means
there's
no
winner
right,
the
winner
is
hey.
I
got
my
ship
to
the
next
level
today
and
I
got
the
you
know.
I
got
this
new
cannon.
I've
been
after
for
a
week.
E
You
know
I've
been
trying
to
save
up
and
I'm
trying
to
find
somebody
with
this
particular
upgrade.
And
finally
I
did
it.
I
found
somebody,
that's
powerful
enough
that
I
could
take
down
and
I
got
the
cannon
I
can
strap
that
on
and
I
that
bad
boy
lets
me.
You
know
fight
my
way
closer
to
the
harder
battles.
D
A
D
A
Derek's
idea
for
the
game
for
sure,
like
that
sounds
like
a
fun
game,
but
it
doesn't
sound
like
it's
real,
even
at
very
massive
scale.
It
doesn't
sound
like
it's
going
to
be
super
stressful
to
the
system,
because,
basically,
every
end
period
of
time,
you're
like
throwing
most
of
it,
it
sounds
mean
when
I
say
it.
E
A
A
Right,
I'm
I'm
I'm
fine
with
not
having
a
match,
timer
or
or
reset,
or
anything
like
that,
because
I
think
that
when
in
the
case
where
you
have
like
really
powerful
players
at
the
center,
if
we
go
back
to
a
conversation
from
earlier
with
matchmaking
and
something
like
starcraft
like
at
some
point
in
theory,
if
the
balance
of
the
game
is
right,
you
will
have
two
players
that
are
evenly
matched
at
the
center
and
there
will
be
either
some
element
of
randomness
or
some
element
of
skill
where
they
both
cannot.
A
There
can't
be
a
stalemate
like
somebody's
gonna
win
and
if
you
balance
the
game
correctly,
that
person
who
who
lost
shouldn't
basically
be
able
to
come
right
back
to
the
center
the
next
time
out
like
maybe
they
get
to
start
a
little
bit
closer
to
the
center
in
theory.
But,
like
you
know,
it
wouldn't
be
fun
if
you
didn't
have
to
struggle
a
little
bit
right,
and
so
it's
like.
Oh
man,
like
I
got
beat
you
know
the
player
has
to
feel
like
they
legitimately
got
beat
like.
A
If
it
feels
like
it
was
totally
a
die
roll,
then
they
might
be
really
upset,
but
you
know
I
got
outmatched
this
altercation,
like
crap,
I'm
gonna
have
to
claw
my
way
back.
If
I
want
to
get
back
to
the
center,
you
know
that
loss
can't
be
too
devastating,
because
then
the
player
is
discouraged
and
they
don't
want
to
try
again.
A
So,
for
example,
like
there
was
a
a
fun
story
with
eve
where,
like
basically
some
guy
spent
years
sort
of
laying
subterfuge
and
then
took
over
an
entire
like
corporation
and
stole
all
the
assets
or
whatever
like
it,
was
like
a
major
like
multi-year
kind
of
planning
event,
but
for
the
people
that
got
screwed
on
the
other
side
of
that
transaction,
like
they
too
had
years
of
effort
that
basically
legitimately
just
disappeared
down
the
drain,
and
so
it's
like
that
can
be
a
really
disheartening
event
for
a
player.
A
E
Yeah
and
it's
we're
saying
that's
a
great
game
design,
question
right:
some
games
and
very
popular
games
are
completely
non-skill-based,
they're
very
random.
Right
I
mean
gambling.
Casinos
is
the
stupidest
obvious
answer
right,
yeah
people.
E
Right
and
the
other
thing
too
is
we
do
want
to
design
a
you
know,
basic
game
that
we
all
enjoy
right
and
part
of
that,
I
think,
is
staying
away
from
the
frenetic-ness
of
you
know,
like
you
said,
like
the
very
intense
rts
games
right.
A
A
I
wouldn't
necessarily
call
it
an
mmo,
but
I
guess
it
is
an
mmo
in
some
regards,
but
basically
like
what
you're
describing
is
a
character,
build
and
every
season
so
to
speak
in
their
terms
like
they
changed
things
a
little
bit
and
then
a
build
that
you
used
last
time
may
not
work
this
time
around
and
then
the
other
thing
is
that,
as
you
introduce
new
elements
or
take
away
other
elements
like
oh,
like
the
last,
you
know
time
shift
in
the
universe
like
made
this
material
unavailable
and
now,
like
all
these
things
disappeared.
A
You
know
you
can't
use
those
anymore
whatever.
So
then
people
have
to
like
try
new
builds
again,
and
so
that
brings
some
of
the
skill
element
to
it.
Even
within
the
randomness,
where
I
may
encounter
a
player
that
has
this
stuff
and
I
destroy
them.
Do
I
need
any
of
this?
Do
I
want
any
of
it?
Does
it
help
me?
Does
it
hurt
me?
Is
it
more
powerful,
whatever
so
yeah?
I,
like
I,
like
that
aspect
of
it,
and
I
think
that
gives
us
some
flexibility
in
designing
sort
of
stateful
stuff.
A
You
know
and
also
makes
it
it
gives
it
gives
a
player
some
skill
that
they
have
to
learn
and
it's
not
super
random.
There
may
be
randomness
within
the
thing,
but.
E
What's
interesting
is
go
ahead,
this
sort
of
design
also
solves
a
secondary,
gameplay
concern,
which
is,
if
you're
a
player,
and
you
do
something
wrong
like
you
build
a
really
bad
ship,
because
you
don't
know
what
you're
doing
right
your
punishment
once
you
get
blown
up
right,
if
you
start
at
the
beginning,
is
pretty
severe.
But
if
we
say
hey
when
you
get
blown
up
everything
you've
collected
so
far,
we
just
exchanged
it
for
cash,
and
now
you
get
to
rebuild
your
ship
and
do
it
better.
A
I
think
there's
there's
that's
a
double-edged
sword
in
some
regards,
because,
on
the
one
hand,
what
you
will
end
up
with
invariably
is
like
a
public
wiki
that
lists
the
best
current
builds
kind
of
thing
so
like
in
elite
dangerous.
You
have
like
these
ship
database
websites
where
people
you
know
config
a
ship,
and
it's
like.
Oh,
if
you
want
to
do
long
range
trading
like
this,
is
the
best
build
for
that
right.
A
A
Yeah
exactly
right,
so
it's
it's!
It's
not
necessarily
a
terrible
problem
to
have,
I,
I
think,
there's
probably
again
a
balance
or
a
spectrum
between
like
a
hundred
percent
conversion
to
cash
of
everything
like
maybe
you
shouldn't,
maybe
there's
some
starting
level
of
cash,
and
then
you
get
a
little
bit
more
cash
if
you
made
it
to
the
center
or
had
some
battles
or
whatever,
but
like
not
a
100
conversion
at
market
rate
of
all
the
assets
that
you
that's.
A
Yeah,
it's
a
big
week,
numbers
later,
yeah,
yeah,
okay,
so
so
circling
back
to
committed
elements.
So
space
theme,
2d,
infinite
map,
infinite
players,
so
is
it?
Is
everybody,
okay,
with
this
kind
of
like
highlander-esque
circle
of
death
space
match
thing
I
kind
of
like
it?
I
think.
B
C
A
I
think
we
can
actually
this
so
I'm
going
to
have
to
disagree
with
you
on
that
one.
I
think
this
actually
sounds
pretty
simple
to
build
out,
because
the
first
thing
you
do
is
you
start
with,
like
the
map
and
players
getting
instantiated
like
okay,
once
we,
you
know,
run
an
automated
test
of
10
000
players
like
does
the
universe
end
up
the
correct
size
like
do
we
place
all
the
players
correctly?
A
A
So
I
think
this
is
actually
a
perfect,
agile,
devops
kind
of
game
design
thing
where
we
can
just
keep
adding
features
and
scaling
it
out
like,
without
I
mean
literally,
without
impacting
the
game
to
a
certain
extent,
if
we
do
it
right
and
get
like
the
low
level
fundamentals
right,
like
all
of
a
sudden
there's
a
new
stat
that
you
can
have.
That
starts
at
some
known
level
or
whatever
and
blah
blah
blah.
I
don't
know
I'm
I.
A
I
think
this
actually
sounds
like
something
we
can
scale
out
in
terms
of
the
design
itself,
like
from
a
complexity,
perspective
and
a
features
kind
of.
A
A
So
we
got
a
circle
of
death
fight
to
the
center.
The
map
scales
out
with
the
number
of
players
there's
some
concept
of
cash.
The
player
has
a
ship
that
they
can
modify,
I'm
assuming
that
all
ships
are
going
to
have
like
the
same
number
of
I
guess
at
first.
We
probably
just
want
to
have
like
one
type
of
ship
with
a
certain
number
of
slots
and
like
we'll
start,
there
kind
of
thing.
B
A
And
so
this
gets
us
into
sort
of
some
rpg
fps
mmo
like
tropes,
where
you
know
basically
like
loot,
drops
right.
They
have
rarity
and
you
know
specs.
Basically,
so
you
know
you,
you
may
encounter
something
randomly
in
the
universe.
That's
like
a
pretty
cool.
You
know
debris
from
some
other
conflict.
You
know
just
happens
to
waft
into
your
sector.
Like
oh
cool,
that's
a
gold!
You
know
super
rare,
you
know
delicious
crystal.
A
Know
something
in
silly,
but
that
that
in
and
of
itself
gives
us
a
system
of
you
know
state
that
needs
to
be.
You
know,
managed
and
distributed
and
and
so
on
and
so
forth.
Do
we
want
to
have
I'm
assuming
you
know
just
to
be
trophy
and
and
current
like,
we
need
to
have
some
kind
of
chat
system.
A
E
Yeah,
absolutely
I
mean
it's
one
of
those
things
where
you
could
sit
on
the
game
design
you
know
for
weeks
or
whatever
right,
but
that's
not
productive.
That's
not
really
relevant
to
what
we're
trying
to
do
here.
We're
trying
to
build
something.
That's
you
know
interesting
enough
and
exemplary
enough
right
for
us
to
be
able
to
show
what
openshift
can
do.
A
Yeah,
oh
so
this
is
quote
unquote
real
time
right
so.
E
And
the
nice
thing
is,
like
I
mentioned
a
couple
times
like
doing
this:
on
a
cell
phone
I'm
pretty
sold
on
you
know
a
cell
phone
or
laptop
or
something
right
like
a
browser-based
client
of
some
sort
right,
because
it
makes
it
way
easier
to
demo
for
people.
E
We're
not
making
a
complicated
client,
I
don't
know
if
we
need
to
get
to
the
point
of
making
a
client
right,
but
you
know
it
should
be
an
option
right.
We
shouldn't
say
we
can't
make
a
client
right
like
if
stakeholders
are
like
hey.
Can
you
demo
this,
like
with
real
people.
A
I
think
this
is
an
opportunity
also
to
potentially
flex
the
power
of
microservices
and
and
openshift
and
stuff
in
general,
where
the
minimum
game
experience
is
a
browser-based.
You
know
quote-unquote
client.
A
As
an
experience,
the
data
just
comes
to
the
client,
it's
up
to
the
client
to
decide
what
to
do
with
it
when
it
when
it
shows
it
so
that
browser-based
experience
might
be
just
like
2d
tiles
kind
of
thing,
whereas
a
full
client
like
if
somebody
wants
to
go
ape
and
make
a
cool,
3d
client
like
with
unity,
hey
man
go
for
it
like
here's.
The
api
just
talk
to
it.
E
There
you
go
yeah
and
that
that
does
lead
us
into,
though,
if
we're
going
to
allow
any
clients
right,
we
gotta
think
through
security,
which
is
nice,
because
a
lot
of
game
does
have
the
same
concerns.
We
have
to
think
through
like
cheating
right
like
how
much
data
are
we
sending
to
the
the
client,
you
know,
make
sure
we're
not
sending
too
much
data,
so
these
are
all
concerns
that
game
just
has
with
their.
You
know,
games
they're
actually
building
today.
So
this
is
a
really
good
direction
for
us.
B
B
A
Source
works
differently
than
users
back
in
the
restream
chat
for
those
who
are
interested
in
this
pdf.
So
again,
this
is
a
thesis
paper
from
a
guy
or
sorry
from
a
person
at
what
was
the
university.
A
University
of
olu
I
I
should
probably
look
up
what
country
that
is
in
so
this
is
mauna's
master's
thesis
and
I'm
gonna
skip
ahead
to
the
actual
microservices
thing,
but
he
this
person
goes
through
and
does
a
really
good
job
of
like
analyzing,
different
game
architecture,
types
and
blah
blah
blah,
and
then
figures
out
like
okay.
If
we
were
going
to
design
an
mmo,
what
would
we
need
from
an
actual
services
microservices
perspective
and
he
goes
through
and
lists
out
a
bunch
of
requirements
and
all
this
other
stuff?
A
B
B
A
Yeah,
so
this
is
this
is
the
graphical
picture
of
all
the
different
services,
so
they
they
had
a
concept
of
zones
and
realms,
which
again
was
tied
more
into
the
multi.
Sorry,
the
mmo
theme
they
had
a
logging
service
where
everything
sent
log
information
discovery
service.
I
forget
what
that
was
for
and
then
an
account
service
that
maintained.
You
know
the
fact
that
a
user
had
an
account.
The
zone
in
the
realm
were
basically
most
of
the
stateful
information
about
what
was.
F
B
Simulation
and
like
physics
and
that
kind
of
stuff
and
then
as
well,
it
also
had
the
chat
stuff
built
in
because
anybody
who
was
chatting
was
going
to
need
to
know
whether,
like
myself
and
derek,
for
example,
are
we
close
enough
to
actually
chat
or
are
we
not?
You
know?
So,
if
you're
checking
to
be
sure
that
people
can
actually
talk
to
each
other,
that's
a
part
of
it,
and
then
the
realm
was
kind
of
like
the
player
state
and
the
rules
about
the
game
simulation
that
kind
of
thing.
B
What
else
I
had
some
notes,
the
logger
was
kind
of
things.
A
So
account,
oh,
it
doesn't
say
what
they
actually
do:
account
discovery
game,
realm,
api
zone,
etc,
and
so,
if
we
think
about
our
game
design
here
right
like
do,
we
need
an
account
service.
I
think
yes,
like
we
probably
you
know.
A
Yeah
and
so
then
they
had.
I
don't
remember
what
the
discovery
is.
B
B
A
Like
you
you,
why
can
I
not
close
this
stupid
window
go
away,
let's
zoom,
so
the
discovery
service
was
for
finding
where
services
were
and
and
what
services
exist
it's
kind
of
like
and
they
use
like
a
service
registry
and,
to
a
certain
extent,
kubernetes
provides
this
to
us
in
the
sense
that
when
I
write
the
account
service,
something
either
needs
to
talk
to
the
account
service
or
the
account
service
needs
to
talk
to
something,
and
so,
as
long
as
I
use
a
standard
nomenclature,
when
I
have
my
service,
other
components
will
automatically
find
all
the
endpoints
for
the
account
service.
A
A
Why
would
the
zone
service
need
to
figure
out
if
the
logger
service
exists,
like
no
the
zone
service
knows
the
people
that
write
the
zone
know
that
there's
a
logging
service
and
somebody's
job
is
just
to
tell
people
what
the
logging
service's
name
is
and
then
you
just
talk
to
it.
So
I'm
not.
I
don't
think
this
discovery,
one
like
that's
pretty
much
taken
care
of
by
kubernetes.
Does
anybody
struggle.
B
What
is
that?
Does
that
change?
In
the
case,
let's
say,
let's
say
the
platform's,
an
incredible
success
and
the
game
is,
is
really
popular
and
we're
opening
up
our
api
to
external
clients,
because
lots
of
programmer
dev
types
play
it
and
they're,
like
hey
I'd,
really
love
to
like
write
some
stuff
against
this.
B
A
So
from
a
keeping
track
of
where
they
are
a
kubernetes
service
already,
does
that,
like
problem
solved
from
a
registering
things
like
that
implies
that
you
have
enough
intelligence
in
your
endpoints
to
query
some
kind
of
registry
to
discover
what
exists
yeah?
Somebody
mentions
like
eureka
ribbon
and
hysterics
are
a
bit
dated
in
the
kubernetes
world.
I
agree.
I
think
we
may
want
to
incorporate
things
like
service
mesh,
right,
istio
and
and
k
native
sorry,
not
istio,
and
jaeger
and
kiali
for
things
like
tracing
or
observability
on
top
of
service
mesh.
F
A
A
Yeah,
I
would
see
more
of
a
registration
of
like
a
thing
like
I'm
gonna
talk
to
the
statistics
or
the
the
I'm
gonna
talk
to
the
loot
server
to
register
this
new
type
of
loot
sub
system
that
you
know
blah
blah
blah,
but
that
that's
not
really
like.
Hey
I'm
a
new
service.
That's,
like
hey,
add
this
stuff
to
your
list.
Whatever
I
don't
know,
so
I
don't.
This
is
this:
we
just
totally
don't
need
this
discovery
service
thing.
E
A
A
A
Api
gateway
sure
I
don't.
I
don't.
F
A
That
we
would
use
three
scale
or
need
this.
I
think,
to
a
certain
extent
the
kubernetes.
The
way
that
we
do
routing
in
kubernetes
may
be
sufficient
for
this.
I
don't
know
that
we'll
need
a
full
api
gateway,
a
la
3
scale,
to
manage
that
type
of
stuff.
A
If
you
look
in
the
in
the
white
paper,
one
of
the
things
that
they
talk
about
for
the
api
gateway
is
that
you
know
you
have
this
character's
endpoint
that
would
go
through
the
gateway
and
the
gateway
knows
via
the
discovery
service
like
where
the
endpoint
is
like.
This
is
a
realm
endpoint,
but
I
think,
with
respect
to
openshift
routes,
if
we're
coming
in
via
outside
the
outside
world,
like
via
the
browser
and
we're
asking
for
data
at
slash,
characters
like
we
already
know
where
to
send
that
internally
without
a
quote-unquote
api
gateway.
A
A
So
then,
we've
got
logger
service.
Of
course
you
know,
but
this
also
may
suffice
to
be.
You
know,
does
the
existing
efk
in
openshift?
You
know.
Does
that
work
for
this?
I
think
we
don't
know
the
answer
to
that
until
we
really
get
into
it
and
start
seeing
if
it's
efficient,
but
I
think
the
key
here
is
that
at
a
minimum
each
service
needs
to.
B
A
A
A
So
then
that
leaves
from
this
list
that
they
had
established
is
like
this
realm
and
this
zone.
Oh,
we
want
a
chat
service
because
we
already
talked
about
some
kind
of
in-game
chat
right,
whether
that's
you
know
text
to
start
you
know
maybe
voip
later,
who
knows
so
now
we're
left
with
like
realm
and
zone
now.
I
think
that
brings
us
back
to
some
of
the.
A
You
know
conceptual
ideas
that
we
had
where
this
is,
where
we're
gonna
diverge,
I
think
from
his
design
or
sorry
from
from
this
from
the
author's
design
I
keep
doing
the
gender
simon
thing,
that's
real
bad!
A
E
Yep,
so
to
me
for
this
game
design
doing
realms
and
zones,
I
really
liked
your
pitch
of
rings
and
the
reason
I
like
that
is
because
the
game
design
problem
that
was
sitting
in
the
back
of
my
head
is
you're.
A
overpowered
player
right
you've
been
playing
for
a
long
time.
You
have
lots
of
awesome
strategies
that
you
figured
out
that
work
really
well
and
you
spend
most
of
your
time
around
the
center.
You
know
maybe
building
things
or
trading
stuff
or
unlocking
cool
stuff
for
other
players
to
find.
E
E
C
E
E
Space
in
it
right,
the
rings
don't
need
to
be
a
predetermined
size.
It's
just
the
next
ring
up
a
level
if
you
were
it's
just
by
default
by
definition,
we'll
likely
have
you
know
less
players
in
it
and
be
a
smaller
size,
and
we
just
want
to
say
players
can't
move.
You
know
back
out
right.
You
can't
go
back
to
newbieville
unless
you
want
to,
you,
know,
start
a
new
character
or
something.
E
A
A
You
know
prevent
this
thing,
so
I
I
still
think
we
want.
I
don't
know
that
we
need
a
server
per
sector,
make
you
know,
because
you
think
about
it
in
the
idea
of
rings.
Like
again,
let's
say
we
have
a
million
players
like
how
many
rings
would
that
be,
and
how
many
how
many
players
per
ring
server
like
I
don't
I
don't
know
we
might
run
into
pod
per
per
cluster
limitations
depending
on
how
we
design
this
thing.
So
I
think
I
think
that's
a
question
mark
right.
A
Yeah,
you
know
so
this
is
going
to
be
like
ring
versus
sector
on
ring
kind
of
thing.
We
don't
know,
you
know
how
many
players
in
a
sector
slash
ring.
You
know
we
don't
know,
because
I
think
I
think
this
gets
back
to
something
we
talked
about
really
early
on,
which
is
you
know,
information
horizon
and
like
what
state
I
need
to
know
and
what
state
is
relevant
to
my
either
server
or
client
or
whatever
you
know.
A
A
Exactly
yeah
totally
area
of
interest
100,
and
I
think
the
other
thing
is
that
it's
also
like
what
state
are
we
storing
about
what
and
where?
If
everything
is
just
in
a
massive
data
grid,
then
I
don't
necessarily
need
to
it.
Doesn't
matter
that
the
cli,
if
we
consider
the
server
as
a
type
of
client,
the
server,
doesn't
necessarily
need
to
know
about
the
entire
state
of
the
universe.
It
only
needs
to
know
about
the
players
that
are
relevant
to
it,
but
it
may
still
send
stateful
information
back
to
the
grid.
The
grid.
A
Just
has
you
know
a
massive
massive
trove
of
information
about
what's
going
on,
you
know
and
that
that
access
to
the
data
grid
gets
load
balanced
in
such
a
way
that
you
know
hey.
This
player
attacked
this
other
player.
Here's
the
result
of
that
role.
You
know
you
said
the
damage
the
or
the
health
points
was
this.
You
know
now
it's
this
here
you
go
update
and
then
you
know
at
some
point:
it's
like
okay,
cool.
I
know
I
now
know
how
much
health
points
this
player.
E
Has
so
so
what
we've
found
here
is
an
unanswered
question,
which
is
how
many
players
can
we
get
in
the
same
area?
If
we
have
this
underlying
game
mechanic
that
maybe
says
point
blank
is
insta-death
right.
If
you
ram
a
battleship
into
another
battleship,
you
guys
just
die
right,
so
that
limits
and
it's
2d
right,
so
that
limits
the
number
of
players
that
occupy
the
same
space
right.
So
then,
how
many
players
yeah
is
how
many
players
can
our
system
essentially
support?
You
know
and
is
it
a
problem.
A
A
A
E
Yeah
exactly
so
last
system
like
this,
I
designed
it
was
in
3d
and
it
used
spheres,
but
we
could
just
use
you
know
hexes
or
a
grid
system.
You
know
whichever
we
want
to
do
because
it's
2d,
which
makes
things
easier,
but
it
used
a
concept
that
I
mentioned
earlier
of
gravity
right
of
things
that
are
close
to
the
player
need
to
have.
You
know
very
synchronized,
immediate
state
updates
versus
things
that
are
far
away
from
the
player.
You
know
it's
lower
detail
right,
so
that
means
you
may
not
even
need.
F
A
E
Rings,
don't
necessarily
have
to
influence
our
system's
architecture
either.
It's
just
an
option
to
consider
rings
could
also
be
a
just
a
directional
limitation
like
we
don't.
It
can
all
be
the
same
map
right.
It's
all
persistent
right,
but
the
rings
are
hey.
You
can't
go
out
where
it's
past
a
certain
point,
because
you'll
collapse,
the
gravity
of
the
system
or
something
or
you
know
your
ship.
A
A
Right
and
so
the
the
idea
of
concentric
rings
is
basically
like
every
time
two
players
at
least
enter
the
game.
A
new
sector
is
spawned
and
if
there
are
no
more
sectors
available
on
an
existing
ring,
we
then
create
the
next
ring
at
which
point
now
there's
a
new
sector.
One
ring
out
for
these
two
players
to
enter,
and
then
at
some
point
we
add
more
players
to
the
sector
until
that
sector
hits
a
limit,
and
then
we
start
filling.
A
A
Is
there's
a
form
of
matchmaking
there
where,
if
I'm
the
first
player
to
join
and
there's
no
second
player,
but
all
the
other
sectors
are
full
like
what
do
I
just
have
to
wait
until
another
player
joins
so
there's
there's
other.
E
Sort
of
things
to
figure
out
there
you
mentioned
ai
right.
We
have
to
be
able
to
spawn
players
anyway
for
testing,
but
the
other
thing
the
thing
I
want
us
to
be
careful
of
from
a
game
design
standpoint.
Is
you
don't
want
to
limit
a
player
to
having
to
restrain
themselves
from
ascending
difficulty?
What
I'm
saying
is
you
don't
want
to
force
a
player
to
grind
if
the
grinding
is
not
fun
so
and
the
place
where
grinding
would
not
be
fun
in
our
game
right
is.
I
have
an
awesome
ship.
E
I
have
a
really
good.
You
know
way
of
built
it
out
and
every
player
I
come
across
is
instantly
destroyed,
but
I
have
to
kill
so
many
more
players
to
get
to
the
more
difficult
players
that
I
want
to
fight
right
that
stops
getting
fun
after
like
the
tenth
person.
So
we
don't
want
to
limit
anyone
from
going
towards
the
center
right.
If
they
want
to.
You
know
just
fly
straight
in
there
because
they
think
they're.
You
know
amazing
sure.
A
Why
not
go
for
it?
This
is
great.
So
I
have
this
idea
for
like
a
gravity-esque
trope
in
the
game
where
the
more
crap
your
ship
is
outfitted
with
like
the
more
gravitational
effect
it
has,
so
you
can't
you
actually
get
pulled
towards
the
center
just
by
the
nature
of
outfitting,
your
ship
kind
of
thing,
so
that
that
design
element
kind
of
also
influences
like
your
inability
to
get
to
the
outer
rings
to
fight
noobs.
A
A
Multiple
continents
so
yeah,
maybe
I'm
going
to
add,
like
you
know,
as
a
question
mark,
maybe
some
like
gravitational
trope
here.
A
A
Yeah
yeah
right
sure,
but
basically
it
was
there's.
A
netties
is
a
company
in
china
that
makes
lots
of
games
and
they
make
a
bunch
of
mmos.
They
actually
wrote
ai
to
analyze,
like
the.
A
F
A
Things
like
pick
up
an
object,
move
to
a
place
like
attack,
an
enemy
or
whatever,
and
so
they
basically
wrote
an
ai
to
to
do
the
quest
in
order
to
test
the
quests
to
make
sure
that
the
quests
weren't
broken.
It
was
a
super
cool
talk,
and
so
the
interesting
thing
about
this
ai
service
for
npcs
is
that
if
we
do
it
right
basically
like
we
can
just
have
npcs
play
the
game
all
by
themselves
and
and
that
in
and
of
itself
is
like
the
scale
test.
You
know
what.
A
E
A
Yeah,
we
should
always
be
able
to
spawn
at
least
one
npc.
You
know
of
a
similar
equipment
level
for
the
player
to
interact
with.
A
I,
like
it,
cool
well
we're
just
about
out
of
time,
and
people
would
watch
that
highlander
npcs,
so
yeah
we've
got
a
month
until
the
next
episode
of
this
show,
and
so
I
think
earlier
I
pasted
the
github
repo,
but
I
will
paste
it
again
if
I
can
find
it
I'm.
I
can't
find
it
because
I'm
a
dum-dum,
rh
gaming
cop
here
we
go
I'll
paste
the
repo
here
in
the
twitch,
I'm
thinking
that
we
should
probably
have
some
kind
of
public
chatty
area.
D
E
A
F
F
A
A
I
don't
know
if
that'll
be
the
one
but
we'll
let
you
know
so
stay
tuned
that
might
be
the
one
and
yeah
so
we'll
the
four
of
us
or
the
five
of
us.
You
know
we'll
be
meeting
offline
now
and
again
to
talk
about
this.
Maybe
we'll
do
some
pseudo
architecture.
Yes,
we
have
this.
Maybe
we'll
do
some
all
right.
There's
a
wicked
delay.
A
Maybe
we'll
do
some
pseudo-architecture
kind
of
stuff,
maybe
some
pseudo
code-
I
don't
know,
but
you
know
check
back
with
us
next
month-
we'll
have
something
interesting
on
this
topic
to
continue
with.
We
very
much
appreciate
your
participation
today.
This
was
super
fun
for
me,
because
I'm
a
nerd
and
I
like
games.
If
any
of
you
who
are
watching
work
for
gaming
companies,
please
find
me
on
eric
on
open
on
twitter
eric
on
open
on
twitter.
A
Please
please,
please
find
me.
I
am
looking
to
interview
backend
architects
developers.
What
have
you
on
this
very
type
of
topic
right,
like
microservices
design
of
multiplayer
games?
You
know
dealing
with
performance
considerations.
My
goal
is
to
try
and
influence
the
red
hat
portfolio
and
road
map
to
better
support.
The
gaming
industry,
of
course,
supported
by
my
friends
here,
michael
jarrett,
derek
and
roddy.
Again,
I'm
eric
jacobs
thanks
so
much
for
being
with
us.
Today
we
are
gonna
call
it
call
it
a
show
and
we'll
see
you
next.