►
Description
Join us for Kubernetes Forums Seoul, Sydney, Bengaluru and Delhi - learn more at kubecon.io
Don't miss KubeCon + CloudNativeCon 2020 events in Amsterdam March 30 - April 2, Shanghai July 28-30 and Boston November 17-20! Learn more at kubecon.io. The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy, and all of the other CNCF-hosted projects
Merging Quickly in a Cloud Native World - Lucas Roesler, Contiamo
You are really excited by a new project, you have a great idea to contribute, you click submit ... and wait. The latest StackOverflow community report states that 80% of surveyed developers are coding in their free time but only 44% are contributing to open source projects. While this number seems good, can we do better? How can we ensure that both contributors and maintainers enjoy and benefit from participating?
In this presentation Lucas goes behind the scenes of the OpenFaaS project to share his knowledge from contributing to OpenFaaS and becoming a core contributor. We will walk through the standard git flow to understand what the maintainers of a large open-source project are looking for in new contributions, what we are doing to improve the experience, what the Kubernetes community at large is working on to help, and what you can do to get your pull request merged more quickly.
https://sched.co/MPZM
A
Hello,
I'm
Larry,
Donato
I'm,
one
of
the
maintainer
of
Falco,
and
please
join
me
in
welcoming
Lukas
rasler,
a
senior
software
engineer
at
contaminants,
our
core
contributor
of
open
fast.
He
we
will
share
with
us
valuable
insights
and
all
his
knowledge
in
out
to
contribute
to
open
source
project
so
enjoy.
B
Thanks
for
joining
me
right
before
lunch,
so
primarily
what
I
want
to
talk
about.
My
inspiration
here
today
is
that
I
think
we
all
hate,
waiting
and
I
find
that
this
is.
You
know
28
days
had
a
ploy
request.open,
no
one
likes
doing
this
and
a
lot
of
times
I
find
an
open
source,
since
the
communication
is
kind
of
hard
and
what
can
and
what
have
we
done
to
make
the
waiting
less?
B
How
do
we
do
this
more
quickly
and
actually
find
that
the
tools
that
we
have
in
the
cloud
native
world
have
have
actually
made
this
better
and
now
we're
getting
better.
So
specifically,
I
want
to
share
with
you
some
of
the
things
that
I've
learned
both
on
the
contribution
side.
So
when
I
came
to
open
fast,
I
was
looking
for
a
project
to
help
me
do
something
at
work
and
I
said.
Oh,
this
is
great.
B
It
solves
my
problem
and
then
but
I
wish
it
could
do
something
else
or
I
need
to
do
like
I
need
to
add
to
this
project
somehow
and
then
eventually,
I
started
adding
so
often,
then
I
became
a
maintainer
contribute.
You
know
a
core
contributor,
and
so
then
I
saw
the
other
side
of
what
this
is
like
when
you
are
receiving
contributions,
and
finally,
I
want
to
talk
to
you
about
what
some
of
the
technologies
and
changes
that
are
coming
in
the
broader
community.
That
I
think
are
helping
us
make
this
easier
and
faster.
B
So
first
off
start
with
the
contribution
side,
I've.
Basically,
two
overarching
things
is
communicate
with
the
team,
so
especially
in
these
this
day,
that
usually
means
get
on
slack,
say:
hi
reduce
yourself
and
the
more
you
communicate
with
the
team
and
I
think.
This
is
a
pattern
that
we
saw
yesterday
during
the
keynote
where
they're
talking
about
these
young
developers
contributing
to
communities
and
a
lot
of
it
starts
with
small
contributions.
B
Talking
with
just
interacting
with
the
team
and
the
more
you
interact,
the
more
they
know
you
the
more:
they
trust
you,
the
faster
things
are
going
to
be,
and
then,
of
course,
the
more
approachable
your
changes
are.
That's
always
going
to
make
it
easier
as
well,
because
if
it's
easy
for
me
to
understand
it's
easy
for
you
to
communicate,
what's
going
on,
then
of
course
that
also
helps.
So
some
of
these
lessons
are
actually
hard
learned.
B
I've
done,
you
know
my
own
mistakes,
so
primarily
what
I'm
gonna
try
to
focus
on
is
where
I've
managed
sake
and
kind
of
what
I've
learned
from
that
mistake.
To
avoid
it
and
making
small
changes
is
really
hard
as
a
developer.
Sometimes
you
get
in
the
flow
and
you're
just
like
AHA
low
change.
This
I
changed
that
I
need
that
and
all
of
a
sudden
you
have
this
really
big
diff
and
someone
looks
at
it
and
goes.
B
This
is
all
great
and
but
you
you
said
you
were
changing
like
this
one
line
and
one
line
changes
are
usually
the
fastest
thing
to
get
merged,
so
the
smaller
you
know,
your
dip
is
the
better,
and
you
know
this
particular
example:
I
think
what
it
was
is
I
had
prettier
installed
and
I
was
modifying
some
JavaScript
and
prettier
said
I'm
gonna
make
your
JavaScript
look
a
little.
As
the
name
says.
Prettier
everything
got
rearranged
for
a
one-line
change
and,
of
course,
I
mean
it
looks
nice
for
the
JavaScript
community.
B
It
kind
of
fits
these
standards,
but
when
someone's
going
in
to
look
at
this
and
like,
but
we
just
needed
this
one
bug
fixed
but
I
didn't
want.
You
know
a
thousand
line
change
that
obviously
slows
things
down.
So
what
I
found
I
finally
have
learned
to
love,
get
a
P
I'm,
generally
a
command
line
person,
but
my
recommendation
is
get
ad
P.
B
I
want
just
this
one
line,
change
added
and
honestly
I
love
the
command
line,
but
I
find
myself
actually
using
the
github
z'
GUI
tool
for
this,
because
you
can
just
go
in
and
say
that's
the
line
at
it
and
then
it's
very,
very
simple:
I'd
go
back
to
it
quite
frequently,
actually,
because,
if
you're
using
it
EDSP,
this
is
what
it
looks
like,
which
is
the
first
time
I've
looked
at
this.
Someone
showed
it
to
me.
It
was
like
hey.
You
should
use
this
I'm
like
why.
Why
would
I
want?
B
What
are
all
these
options
down
at
the
bottom?
Like?
Yes,
no
quit
ad
D
for
delete,
what
slash
supposed
to
mean
in
the
UE
and
the
gooey.
It's
just
add
this
thing:
double
click
to
add
the
line
and
and
you're
good
to
go
so
actually,
depending
on
what
it
is
that
I've
will
actually
come
to
learn
to
love
this
to
a
degree,
but
I
do
find
myself
going
back
to
a
GUI
every
now
and
then
the
other
thing
that
I've
learned
is
that
context
is
king.
So
once
you've
decided
here's
my
one-line
change.
B
Let's
add
it
fixed
secret
handling,
you
know,
as
I
actually
went
back
to
find
one.
This
is
not
a
great
commit
message.
If
I'm
writing
it
as
a
developer,
you
want
to
just
kind
of
like
blast
through
stuff
and
you
write
things
down
and
you
want
to.
You
want
to
share
because
you're
excited,
but
when
I
received
it
on
the
other
side,
so
this
kind
of
fits
on
both
sides
of
the
aisle
as
a
contributor
and
a
reviewer.
B
It's
a
lot
harder
for
me
to
pay
well
what
type
of
fix-
and
this
is
something
I've
been
doing
for
awhile,
but
you
can
actually
force
it
to
remind
you
to
add
more
context.
So
there's
this
dot
get
message.
If
you
add
this
to
your
home
folder
get
wool
when
you
say
get
commit
it
will
just
copy
whatever's
in
there
into
your
commit
message
right
away.
So
whatever,
whenever
I
open
up
my
commit,
it
just
is
pre-populated
with
this,
and
it
just
reminds
me:
I
need
to
add
more
than
just
fix
something.
B
It's
just
I
can
add.
Some
bullet
points
there
real
quickly
and,
of
course,
you
can
put
whatever
you
want.
This
is
just
what
I
have
started
doing
and
so
99%
of
my
commits
usually
looked
like
this
I.
Don't
necessarily
keep
everything!
That's
in
that
template
I,
don't
always
need
a
what
section
or
a
note
section,
but
I've
actually
been
trying.
I
went
as
I
was
reviewing
this
as
talking
with
some
people
over
the
last
couple
months.
B
This
is
probably
not
the
thing
if
you're
actually
going
back
to
the
history
you're
like,
but
that
you
were
going
to
look
for,
and
so
one
of
the
things
that
I've
been
really
kind
of.
I
decided
to
rewrite
this
game
message,
not
in
the
project
itself,
but
just
for
demonstration
purposes.
Support
image,
pool
secrets
for
functions
like
I
highlight
the.
Why
the
the
what
isn't
actually
quite
as
important
you,
usually
it's
the
why
I
wanted
to
support
polling
from
private
registries
for
functions?
B
Why
are
the
benefit
and
as
a
reviewer
when
I
see
that
poor
crust
is
much
easier
for
me
to
say
oh
you're,
trying
to
fix
image,
pool
secrets,
did
you
consider-
or
maybe
you
missed
this-
it's
easier
to
detect
edge
cases
or
think
about
the
impact
and
review
and
see?
Did
your
change
achieve
that
desired
impact
versus
tell
me
what
you
did.
Of
course,
your
changed.
If
you
just
tell
me
what
you
did
is
gonna
match.
So
this
is
the
other
thing
that
I've
been
trying
to
do.
B
It
takes
a
little
bit
more
time,
but
I
also
find
it's
much
easier
to
open
a
poll
request
after
the
words,
because
almost
all
of
the
same
content
is
asked
for
in
the
poor
crust
anyway
and
if
I've
managed
to
magically
keep
it
to
exactly
one
commit
github
will
help
pull
that
into
my
pull
request
body
immediately,
so
I
don't
have
to
spend
this
much
time.
You
know
writing
this
a
second
and
then.
Finally,
the
other
thing
is,
of
course,
writing
tests.
B
This
is
a
easy
thing
to
stay
to
say,
but
this
is
the
thing
that
I
found
the
hardest
when
I
started
contributing
in
a
project.
That's
you
know
deeply
connected
to
communities
and,
fortunately,
for
us
we
use
go.
I
went
looking
for
this
in
the
Python
package
for
the
kubernetes
client
it
doesn't
exist.
I
saw
a
ticket
that
it
would
be
nice
to
have
one
in
the
Python
package,
I'm,
not
sure
about
the
other
versions
of
the
client
but
I
think
so
far.
B
I,
don't
think
anyone
I've
seen
data
scientists
talking
about
Python
and
Java,
but
everyone
else
seems
to
talk
a
lot
about
go
so
if
you're
using
go,
there's
this
really
nice
kubernetes
interface,
and
you
should
pass
that
around
don't
pass
a
client's
head
around
because
now
you
can
actually
write
a
test.
So
one
of
the
things
I
did
when
I
wanted
to
fix.
This
was
I
wanted
to
really
fix
it.
B
I
wanted
to
see
how
this
would
behave
and
do
with
a
small
integration
test,
but
I
also
didn't
want
to
spin
up
a
kubernetes
cluster
in
order
to
do
this
until
recently,
this
wasn't
very
easy
to
do
and
I
did
this,
maybe
a
year
and
a
half
ago.
So
back
then
I
was
like
I,
don't
know
if
I
want
to
do
that
so
much,
but
instead
of
passing
in
the
client
set
for
cube,
I
passed
in
this
interface
and
then
what
this
means
is
on
a
unit
and
an
integration
test.
B
You
can
do
something
that
looks
like
this.
They
have
the
test
client
package.
Then
you
have
a
new,
simple
client
set
and
it
just
mocks
in
memory.
The
entire
kubernetes
api
for
you,
it's
not
gonna-
do
everything
that
a
kubernetes
api
does
is
not
gonna
do
maybe
all
quite
all
the
validations
are
like
the
life
cycle.
It's
not
replicating
the
whole
server
in
memory,
but
it's
gonna
get
you
at
least
some
of
that
integration.
B
Did
my
method
push
a
secret
or
get
a
secret
from
the
kubernetes
api,
and
then
I
can
check
that
so
I
create
my
secret
handler
with
my
new
MOC
cube,
I
call
my
secret
handler
and
then
ask
the
mock
kubernetes.
Does
that
secret
exist
and
I
actually
test
against
it
in
much
the
same
way
if
you've
written
or
deal
with
mocks?
It
just
behaves
the
way
you'd
want
in
the
market,
and
it's
one
import.
B
B
B
So
those
are
kind
of
both
as
a
contributor
things,
I've
done
contributing
that
I
noticed
made
it
easier
on
my
my
fellow
contributors
and
maintainer
x',
but
specific
things,
I
think
the
project
and
the
community
are
doing.
We
are
actually
making
maintaining
easier
and
making
contributing
easier
as
well
and,
most
importantly,
I
think
this
is
around
setting
up
a
dev
environment.
This
is
getting
so
much
simpler
than
it
used
to
be,
and,
of
course,
once
you
can
set
up
a
dead
environment
very
quickly.
B
You
can
also
set
up
a
testing
environment
very
quickly
and,
of
course,
once
you
can
do
these
things,
so
you
should
probably
tell
people
how
to
do
this
and
even
better
how
to
automate
it.
If
I'm
the
maintainer
side
is
our
responsibility
to
tell
you
you
know,
we
want
you
to
make
some
contributions.
We
want
you
to
test
your
contributions.
The
more
we
tell
you
the
more
easy
we
make
that
for
you.
B
Of
course,
it's
gonna
be
faster
for
everyone,
so
for
for
me,
I
work
on
a
Mac
and
for
a
long
time
it
was
docker
for
Mac
or
doctor
for
Windows.
They
just
had
communities
baked
in,
and
this
was
great.
It's
actually
really
really
great
experience.
Cuz,
it's
just
you
build
an
image,
deploy
image
things.
Just
kind
of
work
really
really
nicely,
but
it's
not,
as
se
portable
to
tell
you
know
like
my
friend,
our
other
contributors
that
are
on
Linux,
hey
by
the
way,
just
build
the
image
and
deploy
it.
B
It
doesn't
quite
work
the
same,
and
so
of
course
we
have
things
like
kind,
which
is
from
one
of
the
special
interest
groups.
You
have
mini
cube
and
micro,
Cates
and
I've
really
really
grown
to
love
kind
in
the
past
several
months,
I'm
small
mentioned
that
I
didn't
include
on
here
was
the
k3s
and
KQED
I.
B
Think
are
another
really
good
ideas
that
when
I
get
to
end
I'll
mention
them
again,
but
for
me
I
think
what
we
need
to
do
to
make
it
easier
and
faster
for
people
to
actually
contribute
when
you're
living
inside
kubernetes
is
telling
people
how
to
test
and
I've
done
this
one
which
is
I
just
tested
it
by
hand
in
a
cluster
doing
some
stuff.
It's
like
yeah,
I,
tested
it,
and
probably
because
I'm
a
core
contributor
someone
said
yeah.
He
probably
did
that
that
works.
B
Ok,
but
if
it's
your
first
or
second
commit,
then
we're
probably
like
well,
what
do
you
mean
tested?
What
type
of
cluster?
What
sort
of
environment
was
it
in?
What
sort
of
operators
did
you
have
installed?
There's
lots
of
questions
there's
so
many
things
to
plug
in
extend
kubernetes
these
days,
which
is
a
great
thing,
but
it
also
means
that
there's
more
moving
parts
to
test
you
know
to
to
account
for
it,
so
this
isn't
very
easy
to
validate
what
I've
been
doing
recently.
This
is
a
ploy
across
I
had
is
using
kind.
B
So
this
is
something
that
using
kind
has
made
so
much
easier
for
us
to
do
manual,
testing
and
just
even
just
local
development.
This
is
my
go-to
now
just
for
whenever
I
need
it.
Kubernetes
cluster
I
want
to
start
with
some
clean
local
development.
I
don't
want
to
muddy
up
some
existing
cluster
I.
Don't
want
to
deal
with
potential
interactions
to
stuff
when
I
use
kind,
I
get
a
35
line,
bash
script
that
just
starts
it,
so
the
slides
are
online.
B
If
you
want
to
go
see
what
this
thing
looks
like
it's
there
or
you
could
try
and
remember
this
shop,
but
I
actually
want
to
show
you
what
this
experience
looks
like
a
little
bit.
So
it's
Cape
here
flip
over
now.
I
tried
this
this
morning
and
so
I
decided
to
actually
record
the
install
using
ASCII
cinema
because
it
took
200
seconds
on
the
the
Wi-Fi
to
do
it.
So
I'm
gonna
speed
it
up
a
little
bit.
So,
let's
see
so
I
just
run
my
script
and
in
about
since
it's
sped
up
15
30
seconds.
B
It
just
runs
through
all
these
steps
and
we'll
have
a
kubernetes
cluster
with
open,
fast
already
running
now.
I'm
gonna
come
back
to
this
in
just
a
moment,
because
I
want
to
highlight
move
forward.
One
three
specific
commands
I
think
really
made
this
easy,
and
so,
if
you
are
also
trying
to
do
this,
if
you
want
to
make
your
kubernetes
based
project
easier
to
test
and
install
so
of
course
you
have
the
kind
create
command
I
like
to
give
my
environments
actual
names.
You
don't
have
to
do.
B
This
they'll
start
one
with
a
default
name,
but
it's
nice
to
the
name.
Things
next
cube,
detail,
I,
don't
know
when
they
added
this,
but
I
discovered
it
and
loved
it
is
this
rollout
status
command
it'll
actually
tell
it
tell
you
when
those
pods
hit
the
running
state.
So
if
you're
saying
hey
start,
my
kubernetes
cluster
install
open
fast
you're
sitting
there
you're
waiting
for
it
to
finish
like
it
takes
a
little
bit
of
time.
Maybe
it
has
to
download
a
docker
image
and
has
to
start
up.
B
You
don't
want
to
start
running
some
other
command
immediately
afterwards.
So
this
will
just
do
exactly
what
you're
waiting
for
and,
of
course-
and
this
is
the
thing
that
I
really
loved
well
kind-
is
this
load
command,
so
the
one
of
the
big
pains
with
at
least
in
the
past,
then
why
I
love
docker
for
Mac
is
I
would
build
my
docker
image
and
it
would
just
be
available
tonight.
Kubernetes
I
didn't
have
to
push
it
anywhere
and
kind
load
allows
me
to
completely
mimic
this
on
any
system.
B
B
That's
why
I
didn't
want
you
to
sit
through
that
whole
thing
and
then
I've
done
a
little
bit
of
helper
here,
because
it's
a
test
environment,
I
I
set
up
a
known
standard
user
password
because
we
enable
basic
off
since
we
have
basic
auth,
that's
better.
The
start
was
secure
by
default
and
figure
out
how
things
break.
B
If
I
forget
about
this
fact
and
then
I
give
myself
a
reminder
of
the
command,
I
need
to
run
and
then
finally,
the
last
step
I
do
kind
does
not
expose
itself
on
my
local
network,
so
I
have
to
do
a
port
forward,
and
this
means
that
now
that
pod,
in
our
case,
open
fast,
is
actually
available
to
be
used.
So
because
I
ran
this
to
the
ASCII
cinema,
it's
not
running
or
not.
Well,
it
is
running
in
the
background
because
I
left
it
in
the
background,
but
I
need
to
do
that
port
forward
again.
B
This
is
all
good
sign.
There's
my
open
pass
in
dark
mode
because
of
a
plug-in
I
have
installed
I,
don't
think
we
support
native
dark
mode
yet,
and
I
could
easily
just
deploy
any
store
function
if
I
wanted.
But
that's
not
the
thing
I
wanted
to
show
you,
because
what
I
wanted
to
actually
talk
about
is
not
so
much
a
I'm
a
function
developer
wanting
to
test
out
open
pass.
Of
course.
This
could
help
you
that
way,
but
more
of
I
want
to
work
on
open,
pass
itself.
How
do
I
run
and
test
my
stuff?
B
B
Just
so,
we
can
double
check
what
image
I'm
actually
running
right
now.
So
here
in
my
top
right
corner,
I
just
asked
the
kubernetes
to
tell
me
what
is
my
gateway
running
it's
running.
We
have
two
side
cars,
so
you
have
a
gateway,
and
then
we
have
fast
90s
and
that's
the
thing
that
I
want
to
kind
of
develop
locally
and
it's
running
zero,
seven
and
five.
So
this
is
just
the
plain
it
if
you
use
helm
to
install
open
pass
right
now,
that's
the
docker
image.
It's
gonna
pull!
B
B
Though
you
can
see
last
night
so
13
hours
ago,
I
built
a
new
open,
Fast,
Pass,
Nettie's,
and
also
so
we
can
just
check
what
the
logs
look
like
this
current.
The
current
version,
not
my
latest
when
it
starts
up,
looks
like
this.
So
for
demo
purposes,
I
went
in
last
night,
add
a
little
Hello
flag,
so
you
can
actually
trust
me
that
this
really
is
loading
my
image,
it's
not
just
changing
some
values
and
but
it'll
actually
print
out
something
new.
C
B
B
Yep
and
there
we
go.
My
new
pod
started
with
my
local
development
version
and
all
of
these
commands
the
thing
that
I
loved
about
it.
The
thing
I
think
speeds
everything
up
and
the
thing
I
think
any
project
that
is
doing
stuff
on
top
of
food
kubernetes
is
that
every
single
command
I
can
just
put
in
a
script.
I
can
just
put
in
github
pull
request
as
instructions.
What
did
I
actually
do
to
test
it?
B
Every
single
line
verbatim
can
be
put
there
copied
to
someone
can
look
at
it,
copy
paste
it
and
run
exactly
those
same
instructions
and
say
yeah.
This
worked
for
me,
so
this
is
something
I've
been
doing.
I
found,
it's
actually
been
helped
me
a
lot
as
a
maintainer.
I'll
get
instructions
from
people,
maybe
without
this
sort
of
scripting
and
what
I'll
do
is
I'll
just
I'll
look
at
and
be
like:
okay,
I
know
this
is
some
pretty
well
because
I
spent
a
lot
of
time
in
it.
B
So
I'll
just
kind
of
create
this
little
script
run
through
using
this
and
then
I
can
say:
hey
you
didn't
have
instructions,
but
here's
what
I
did
to
test
it.
So
I
can
show
them
what
I
thought
they
did,
because
they
didn't
tell
me
what
the
instructions
were.
But
now
there
is
some
instructions,
oh
yeah
that
looks
right
or
they
can
now
verify
the
same
thing
that
I
did
and
it's
much
easier
to
communicate.
Something
that's
scripted
like
that
than
it
is
to
display.
Well,
did
you
do
this
to
your
kubernetes
cluster?
B
Did
you
remember
to
install
that?
It's
like
no
run
this
command
and
we
now
have
the
same
environment.
So
it
brings
a
lot
of
the
I.
Remember
when
I
first
learned
about
containers,
it's
like,
oh
well,
you
run
the
same
thing
locally
that
you
do
in
prod.
It
gives
me
some
of
that
feeling
back
and
it's
really
easy.
So
I
love,
easy
and
ultimately
I
also
love
fast.
So
those
things
together.
Is
it
amazing
and
it's
not
just
for
your
local.
Of
course,
you
can
do
anything
that
you
can
scripts.
B
You
can
do
in
continuous
integration
and
we
gotta.
Can
a
community
member
contribute
this,
where
they
actually
test
our
helm,
install
using
kind
so
any
environment
with
docker?
We
can
just
do
this
and
it
uses
exactly
the
same
kind
of
sequence
of
commands
create
a
cluster
at
home.
Wait
for
everything
to
finish
add
commit
are
at
open
pass.
B
So,
at
the
end
of
the
day,
the
automate
everything
the
closer
you
get,
the
scripting
it
the
better,
and
this
is
always
going
to
make
things
faster.
So
the
last
thing
that
I
want
to
kind
of
highlight
is
that
living
inside
communities
also
makes
it
easier
to
automate
things
and,
as
we
saw
earlier
today,
where
they
talked
about
sock,
BOTS
bots
are
everywhere.
B
Okay,
well
is
this
person
you
know
somehow
good
enough
are
contributed
enough
to
become
somehow
get
special
permissions
and
github
we're
gonna
say
everyone
can
have
permissions
in
github
to
a
degree
and
using
the
bots
allows
us
to
have
that
fine-grained
control
and
in
in
some
cases,
do
automatic
enforcement's
of
things.
So,
since
we're
open
open-source,
we
want
to
make
sure
all
the
commits
are
signed
off
that
we
have
this.
Was
it
developer
certificate
of
origin?
B
B
So,
at
the
end,
what
I've
learned
between
a
number
of
kind
of
personal
changes,
making
sure
that
I
write
tests
make
sure
that
I
write
small
commits
and
reminding
myself
of
doing
this?
At
this
point,
I,
like
with
the
automation
that
we've
put
in
place
with
the
scripts
that
we
can
share
for
your
local
development,
we
really
can
get
to
the
point
where
contribution
really
is
just
kind
of
a
linear
flow,
create
a
branch
build,
it
run
it
locally,
and
these
are
things
we
can
actually
say:
yeah
just
build
and
run
it
locally.
B
So
now,
of
course,
this
is
the
moment-
I
say
there
are
plenty
of
things
to
do.
We
want
to
make
this
yet
even
faster
I
mentioned
briefly
earlier:
the
concept
that
k3s
in
k3d
and
one
of
the
things
I've,
maybe
there's
other
ways
of
doing
this-
the
idea
that
we
could
build
a
k3s
cluster
that
you
can
just
install
that
has
open
path,
kind
of
they're.
All
ready
to
go
out
of
the
box
is
really
enticing
to
me.
I
think!
B
That's
something
that,
if
you're
really
interested
in
like
automation
or
have
some
experience
with
that,
I
would
love
to
hear
because
I
think
that
200
seconds
we
could
probably
get
it
down
to
like
10
seconds.
You
know
we
could
say
here
is
the
docker
image
that
is
k3s
with
open
pass
on
it.
It
boots
everything
up.
You
use
a
command
to
load
your
death
image
in
and
you're
done.
You've
done
all
of
that
you've
done
the
cluster
setup.
B
You've
built
your
image,
and
since
it's
go
buildings
pretty
quickly,
pretty
quick
too,
you
done
that
whole
cycle
in
under
a
minute.
I
think
that's
possible
and
I
think
that's
possible
because
of
the
emphasis
that
our
community
has
put
on
building
some
of
this
tooling.
So
if
you
want
to
contribute,
let
me
know
talk
to
us
on
slack:
try
out
open
pass.
B
Especially
towards
the
end,
the
the
discussion
around
kind
is
something
I
think
any
project
that
uses
kubernetes
can
really
take
in
in
reuse,
whether
you're.
Actually
in
our
case,
we
interact
with
the
kubernetes
api
very
directly
or
if
you
just
want
to
have
an
environment.
Maybe
you
have
some
micro
services
and
you
need
like
it's
easier
on
kubernetes
or
that's
what
you
normally
deploy
on.
You
could
use
kind
to
mimic
that
for
your
dev
environment
too,
so
I
think
that
and
of
course,
finding
the
kubernetes
interface
inside
the
go
package.