►
From YouTube: Meshery CI Meeting (April 22nd, 2021)
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
And
but
but
rodolfo
yeah,
let
me
jump
into
the
meeting
minutes
as
well.
There's
there's
a
couple
of
topics
to
touch
on
okay,
one
topic
we
didn't
touch
on
in
the
past
and-
and
maybe
today
is
a
good
day.
Maybe
not
is
air
gapped
deployments
kind
of
relates
it's
sort
of
tooling
around
mesh
resource
sort
of
relates
to
ci.
A
A
Well,
that's
fine,
like
we've
made
considerations
for
that
in
the
architecture
we
haven't
made
it
we
haven't
built
a
lot
of
tooling
around
it
to
make
it
easier
to
really
to
facilitate
that
is
so
right
now,
there's
manual
steps
for
what
that
is,
and
in
the
future,
some
changes
to
both
ci
and
to
mesherie
itself
can
facilitate.
They
can
just
make
it
easier
to
do
that.
One
example
is
when
you
when
we
were
to
build,
so
let
me
jump
into
the
medium,
so
I
can
help
with
these
notes.
A
Okay,
one
example
is
when
you
go
to
build
a
measuring
adapter
one
of
the
things
the
adapter
is
capable
of
is
lifecycle,
management
of
service
meshes
and
to
install
service
mastery
or
the
adapter
specifically
needs
a
copy
of
that
service.
Meshes
installation
files
like
you
know,
and
right
now
the
way
that
the
adapter
the
adapters
have
that
ability
to
take
a
cached
copy.
As
a
matter
of
fact,
what
they'll
do
is
they'll
download
the
bundle
for
istio
or
for
linker
d,
or
for
whatever
surface
mesh
and.
B
A
The
first
time
that
they
deploy
it
then
they'll
cache
it,
and
I
forget
how
long
it
will
how
long
the
cache
is
before
it
expires,
but
it
we
could
equally
as
well
create
and
image
create
images,
alternative
images
of
each
of
the
adapters
that
already
have
a
ver.
You
know
the
latest.
You
know
a
version
of
each
service
meshes
installation
files
bundled
into
the
image
so
that
it
that
just
makes
it
convenient
for
them
to
be
able
to
deploy
on-prem.
A
That's
one
approach.
Another
approach
is
when
you
so
mesherie
cloud
as
one
of
the
remote
providers
that
you
can
sign
into
take.
That's
just
a
container,
it's
a
container
and
a
postgres
database.
So
and
so
it's
a
small
service
that
has
a
postgres
database.
You
can
take
that
postgres
database.
You
can
take
it
and
deploy
it
in
a
container.
It's
it's
running,
containerized
as
a
service,
and
it's
not
much
of
a
service
you're
just
running
and
that
same
docker
compose
file.
A
And
so
my
point
is,
you
can
either
bundle
the
service
meshes
installation
files
inside
of
an
adapter
and
pre-bake,
an
image
that
has
them
or
which
is
fine
or
or
I
should
say,
andor
in
meshri
cloud.
If
you're
running
that
locally
well,
part
of
the
configuration
of
the
adapter
or
of
meshri
can
just
be.
A
A
A
You
might
only
want
to
bundle
one
copy
of
that
one
version
of
that
service
mesh,
and
so
if
they
were
working
with
a
version
that
you
didn't
bundle
well
then,
then
they
go
back
to
the
manual.
You
know
manual
process
of
like
loading
into
the
file
system
themselves
and
okay.
Well
also,
if
they're
running
mesher
for
a
long
time.
Eventually,
the
service
mesh
has
a
new
version,
and
then
they
want
to
grab
that
new
version.
It's
like
well
so
like
at
some
point
having
something
like
mesh
recloud
that
facil.
A
That
is
a
single
point
of
reaching
out
to
the
internet,
that
like
facilitates
that
type
of
a
thing
or
that
will
connect
to
a
local
registry,
because
a
lot
of
the
meshes
are
released
as
containers
themselves,
and
so,
if
someone
caches
to
a
local
or
if
they
store
those
images
in
their
local
registry,
then
measure
you
could
use
that.
So
there's
some
considerations
and
asuko.
A
I
had
put
down
some
considerations
because
we
have
a
user,
a
bank
here
in
austin,
that's
wanting
to
use
messry
and
the
user,
his
name's
pulberg
he's
in
the
slack
he
was
asking
for
instructions,
so
we
started
to
write
down
some
things
and
a
suco
took
and
put
those
into
a
document.
So,
okay
grab
that
link.
A
So
so
I
see
jubrilla's
on
this
jubril.
I
know
we
don't
ask
you
to
do
enough.
You've
been
getting
bored
recently,
so
so
I'll
definitely
highlight
this
air
gap
deployment,
design
spec,
because
it
touches
on
some
things
that
you're
familiar
with.
A
Okay,
yeah,
the
brief
recap
is:
there
are
certain
users
that
want
to
deploy
mesherie
in
a
non-internet
connected
environment
and
they
can
do
that.
A
You
know
automating
some
of
those
steps
or
changing
so
so
that
air
gap
deployment
dock
starts
to
outline
those.
Could
you
describe
what
that
looks
like
yeah?
Oh.
C
A
It's
the
first
comment
in
there,
so
that's
the
user,
mr
polberg
he's
in
our
slack.
So
if
you
guys
want
to
talk
to
me
just
so
happens
to
live
in
the
same
city
as
me.
I
I
didn't
know
him
before
he
before
he
became
interested
in
mastery
or
whatever.
A
I
don't
even
know
how
you
heard
of
mystery,
but
so
the
things
that
were
top
of
mind
for
me
was
the
first
thing
I
was
talking
about
about
the
adapters
needing
to
have
a
copy
of
the
installation
files
for
a
mesh
so
that
that
describes
it
more.
The
second
consideration
is
like
well.
This
is
more
like
their
issue.
We
can
facilitate
it,
but
the
container
images
for
measuring.
If
you
run
measuring
ctl
well,
it
goes
out
to
docker
hub
and
pulls
them
down.
A
A
The
third
cons
and
measury
cloud
can
help
with
both
of
these
things:
caching,
installation
files,
caching,
other
container
images
from
sri-
I
shouldn't
say
caching,
but
I
mean
being
a
repository
of
that.
The
third
thing
is
measuring
cloud
itself.
If
you
want
to
use,
you
can
use
the
local
provider
with
no
need
for
internet,
but
if
you
would
like
to
take
advantage
of
some
of
the
things
that
mastery
cloud
has,
then
it's
free
to
do
so.
We
just
run
it
as
a
simple
service
for
everyone
today,
but
we
can
make.
A
But
those
container
images
are
public
and
we
can.
We
just
need
to
write
down
instructions
for
like
how
to
download
those
and
run
them
locally,
and
then
that
way
you
can.
The
one
thing
about
measuring
cloud
right
now
is
when
it
lets
you
authenticate.
A
We
had
to
do
it
kind
of
cheaply
because
there's
a
lot
of
stuff
that
we
only
get
to
you
know,
because
it's
only
because
of
all
of
you
that
show
up
that,
like
any
of
this
stuff
gets
done
so
so
we
can
only
get
so
much
done
and
in
measuring
cloud
you,
you
have
all
seen
that
it.
It
allows
you
to
socially
authenticate,
and
that
was
done
because
it's
just
faster
than
having
to
deal
with
email,
addresses
and
password
resets
and
you
know,
and
then
ldap,
integration
and
stuff.
A
But
as
we
go
forth,
ldap
integration
active
directory
integration,
those
that's
those
will
be
highly
called
upon
features
in
rightfully
so,
and
so
that's
something
that
we
don't
intend
to
build
into
the
local
provider,
be
for
a
bunch
of
reasons,
one
of
the
reasons
being
because
if
you're
using
measuring
in
a
local
provider
way
with
a
nun
provider,
you're
really
just
using
it
as
an
ad
hoc
tool,
something
you
reach
into
the
toolbox,
you
pull
it
out,
use
it
for
a
point
in
time,
and
then
you
throw
it
up.
A
A
A
Those
are
some
of
the
those
are
some
of
the
considerations
about
what
it
would
be
to
run
mescheri
air
gapped,
and
he
this
guy
polberg
forget
his
first
name.
He
is
entirely
fine
with
just
whatever
the
manual
instructions
are.
He
just
needs
those.
A
Nice
guy
he's
kind
of
suffering
in
his
environment
because
he's
sort
of
like
there's
a
small
collection
of
like
the
devops
people
or
what
have
you
and
everybody
else
you
know
is
holding
him
back
so.
A
A
Them
cool
that
was
just
a
you
know,
kind
of
top
of
mind,
because
suco,
like
asuko,
has
asked
yeah
osugo's
asking
for
feedback
in
that
aircap
deployment
talk,
and
one
of
us
needs
to
copy
my
comments
from
the
issue
into
that
doc
and
sort
of
engage
asuko
in
there,
and
he
just
put
this
together
a
couple
of
days
ago
or
so,
and
he's
got
a
couple
of
suggestions
that
I
think
are
might
be
bigger
than
we
can
bite
off.
A
The
guy,
the
guy
who
was
asking
for
this
pullberg
like
so
feel
free
to
engage
him
in
our
slack.
Like
you
know,
he's
a
seems
like
a
nice,
a
nice
user,
we
definitely
need
more
users
and
more
users.
Giving
feedback
and
he'd
also
highlighted,
like
he'd,
given
an
example
of
what
he
was
comfortable
with
and
I'll
try
to
find
the
doc.
He
had
pointed
out
that
there's
some
of
you
are
familiar
with
longhorn
longhorn.io
is
just
another
cloud
native
project:
it's
for
storage
and
to
install
longhorn
on-prem
there
is
they
have
a
dock?
A
A
A
A
A
A
So
yeah
to
all
of
you
or
that
on
this
call
it.
This
is
a
a
medium-sized
semi-juicy
like
area
for
you
to
go,
make
an
impact,
and
I
don't
think
you
can
really
get
it
wrong
to
start
with.
Maybe
you
provide
a
bad,
maybe
all
you
do
is
provide
instructions.
Maybe-
and
it's
all
manual-
maybe
there's
a
simple
bash
script-
maybe
you're
helping
identify
how
some
of
the
other
things
I
was
saying
in
the
way
in
which
we
could
change
mesh
for
you
to.
Maybe
you
want
to
work
to
deploy
measuring
cloud
locally?
A
A
There
are
two
other
potential
topics
for
us,
as
everyone
absorbs
this,
the
other
one
of
the
other
topics
is
well,
it's
helm,
charts
and-
and
I
think
this
one
is
easier
than
this
third
one.
The
third
one
is
more
like
documentation,
versioning
and
that's
kind
of,
but
but
helm
charts.
So
so
the
project
itself,
mastery
has
a
helm
chart
and
the
helm
chart
tends
in
the
past
has
been
treated
as
a
second
class
citizen,
and
we
all
know
that
no
one
likes
to
be
treated
like
that.
A
So
we're
trying
to
be
more
kind
to
the
measuring
helm
chart
and
have
it
released.
You
know
pretty
much
every
time
that
measure
is
released,
the
helm
charts
are
stored
within
the
meschery
repo,
the
source
code
for
the
helm,
charts
like
you
know,
they're
there
and
there's
occasionally
changes
to
them.
There
aren't
nearly
as
many
changes
to
the
helm
chart
as
there
are
to
mastery
itself,
but
occasionally
we
add
an
adapter
or
we
remove
an
adapter
or
we
have
a
an
operator
or
we
have
or
I
don't
know
we
do.
A
We
do
other
changes
and,
and
that
you
know,
and
there's
a
github
action
that
will
that
you
can
configure
to
point
to
your
the
your
file
system,
location
in
your
repo,
for
where
helm
charts
are,
and
it
will
every
time
you
make
invoke
the
action
it
will
build.
The
chart
create
the
artifacts
and
publish
them
to
another
repo
and
the
other
repo
that
we
have
is
is
the
repo
it's
the
website.
Measurey.Io.
A
It
has
a
public
facing
http
server
where
we
can
put
publish
the
helm
chart
so
that
people
can
just
grab
it
and
if
you
go
to
artifact
hub
dot
io,
I
think
you'll
find
yeah
artifact
hub.io.
If
you
search
for
measuring
you'll,
find
meshrey's
helm,
chart
and.
A
So
I
think
he
ended
up
kind
of
making
it
another
one
and
he's
got
it
committed
to
our
repo
and
it
will
be
invoked
whenever
there's
a.
I
think.
The
way
it's
configured
right
now
is
that
a
new
chart
release
will
be
made
any
time
that
someone
does
some
someone
changes
the
code,
the
files
under
slash,
install,
slash,
charts,
slash,
kubernetes
or
whatever.
A
I
think
he
has
that
that
github
workflow
only
to
be
triggered
on.
You
know,
detection
of
changes
under
those
files,
and
so
I
I
went
ahead
and
tried
it
out
a
couple
of
weeks
ago
when
he
I
was
just
trying
to
help
him
advance
it
along
and
yeah.
I
think
that's
the
one
I
think
when
we
get
it
working
he'll
end
up
merging
it
into
layer.
Five,
if
you
look
in
mesherie,
there's
there's
a
open
pull
request.
A
In
which
it's
the
fifth
one
down
at
ci
testing
home,
so
I
had
just
opened
up
a
quick
small
change
under
that
directory.
I
just
like
bumped
the
version
of
something
to
test
it
out
and
it
ended
up.
The
action
associated
ended
up
failing,
and
I
believe
that
was
because
of
maybe
oh,
I
think
asuka
might
have
been
referencing
well,
we
can
rerun
if
you
want
to
click
rerun,
maybe.
A
A
A
Yeah
to
fix
the
issue
that
to
to
fix
the
file
reference
with
the
github
action.
D
Now
but
I
was
finding
excuses
that
I
don't
know
where
you
publish
and
blah
blah,
but
you
said
earlier
it's
on
the
party
fact.
I
also
yeah.
A
A
Eventually,
another
thing
to
do
is
not
that
it's
on
line
49,
the
github
actor,
is
asuka
and
like
I,
if
you
guys
can't
tell
I'm
all
about
recognizing
the
work
that
everyone
here
does
and
highlighting.
Folks.
A
In
this
case
it
is
a
liability
to
or
like
it's
we
we
would
do
well
to
change
it
to
like
a
service
account,
a
robot
account,
just
as
a
point
of
order
of
like
it's
a
robotic
thing:
that's
kicking
off
and
stuff,
but
but
asuco
is
listed
as
the
one
of
the
maintainers
of
of
this
and
stuff
and
something
like
you
know
it
isn't
about
not
recognizing
it's
about
making
sure
that
it
works.
A
C
A
A
Well,
we're
all
fall,
I'm
not
sure.
If
we
have
speaking
of
energy,
I'm
not
sure.
If
alonso
do
we
have
the
energy
for
a
mesh
redox
versioning
this
guy,
like
it
pains
me
that
your
pr
has
been
out
there
for
about.
As
long
as
my
beard
is,
it's
been
a
while.
E
A
D
A
Three,
that
and
and
by
the
way
I
don't
think
that
we
want
to
actually
keep
so.
Okay.
Let
me
not
regret
digress
here,
but
but
let's
say
that
we
made
a
real.
Let's
say
they
made
the
v060
release.
Like
you
know,
it's
it's
a
stepwise
release
like
it's
a
significant
thing.
Changes
have
happened,
we've
added
stuff,
etc.
A
So
so
so
part
of
the
release
workflow
in
an
ideal
world
like-
and
I
don't
know
why-
I'm
using
my
hands
to
describe
this
but
would
snapshot
the
wood
build,
would
do
a
jekyll
build
grab
the
output
of
the
underscore
site
folder.
It
would
just
basically
grab
the
output
of
the
statically
com.
A
History
use
the
word
compiled,
but
the
statically
generated
website
the
docs
website,
rename
the
folder
to
v060
and
shove
it
into
a
folder
underneath
the
docs
folder
somewhere
like
and
then
yep
you'd
have
a
point
in
time.
Snapshot
of
the
entire
site,
inclusive
of
whatever
was
in
the
menu
and
whatever
was
in
the
footer
and
the
header
and
like
for
that
point
in
time,
which
you
know
over
time
like
could
be
a
dramatically
different,
looking
site
their
problems
with
that
potential
problems
with
that.
A
But
my
sense
is
that
the
problems
with
that
you
know
like
the
problem
would
be.
What
would
be
the
problem?
Well,
problem
would
be:
a
user
would
be
on
v08.0
the
latest
release
they
want
to
go
back
in
time
to
six
zero
and
in
v080,
like
the
latest
release
they're.
Looking
at
this
dock,
it's
this
nice
air-gapped
deployment,
instructions,
dock
and
they're
like
oh
yeah.
This
is
great.
Okay,
good,
oh,
except
we're
not
ready
for
8.0.
A
A
Well
like
yeah,
because
we
didn't
support
it
then
or
like
it
wasn't
written
up
then,
and
then
like
that's
just
how
it
is.
You
know,
or
in
the
footer
it
didn't,
have
the
link
to
the
youtube
channel,
because
we
didn't
have
one
back
then,
or
something
like
that
like
like.
Well,
that's
just
how
it
like
yeah.
That
would
be
the
worst
of
the
experiences.
A
I
think
that
they
would
get
on
the
flip
side
if
what
we
versioned,
if
we
didn't,
take
a
complete
snapshot
of
the
entire
site,
if
what
we
ended
up,
versioning
was
just
the
the
internal,
the
guts,
the
content
of
not
the
menu,
not
the
header,
not
the
footer.
If
we
left
those
as
perpetually
the
same,
we
would
be
in
the
gut.
We
would
be
versioning
the
guts
of
the
actual
contents
right.
The
actual
now
the
issue
becomes.
A
A
Well,
hopefully,
you
guys
see
where
I'm
going
with
this.
That
you're
going
to
do
is
something
much
something
worse
like
if,
if
someone
goes
back
in
time
to
six
so
they're
going
to
be
seeing
a
bunch
of
broken
links
like
at
least
at
least
when
they
have
a
different
experience,
when
they
go
back
in
time,
they
won't
have
broken
links
and
weird
stuff
going
on
they'll.
Just
things
will
be
missing
because
they
weren't
present
at
the
time,
which
is
exactly
why
you
version
things
which
is
exactly
the
point.
A
Yeah
yep
could
be
yeah
yeah.
There
could
be
some.
I
think
if
you
thought
about
it
long
enough
and
got
clever
enough,
you
might
be
able
to
only
modify
that
yeah
you're
right
the
alonso.
It
depends
on
highly
likely
most
likely
not
worth
discussing,
there's
a
chance
that
you
wouldn't
have
to,
but
that
would
only
be
if
we
had
done
really
good
design
in
relative
menus
and
relative
hyperlinks
and
anyway,
to
your
point,
yeah,
there's
a
that's
a.
E
But
with
this
a
different
approach,
the
only
thing
that
will
have
to
be
being
taken
care
of
will
be
that
that
that
prevent
generating
a
link
of
a
non-existent
document
of
a
previous
version
that
was
snapchatted
in
that
document
that
you
were
saying.
Yes,
let
me
think
on
it
and
see
how
we
started
with
that
yeah.
A
There's
if
it
begins
to
like
there
might
be
some
things
that
I'm
gone,
there
might
be
some
un
unknown
issues
that
actually
occur
with
what
I
had
said
about
snapchatting
the
folders.
So
one
of
the
issues
is
that
the
vj
brought
this
up
before,
like
that,
hey,
you
might
run
out
of
disk
space.
That
github
is
giving
us
and
be
like
okay.
What
we
would
do
is
well
we
would.
We
would
not
persist.
A
The
v
0
5,
one
v,
zero,
five,
two
b,
zero,
five,
three,
those
would
be
deleted
and
we
would
just
keep
v
zero
five
zero
v,
zero,
six
zero.
Only
the
the
minor
dot
zeros,
not
the
incremental,
not
the
30
incremental
in
between
is
one
thing,
that's
a
kind
of
an
issue
anyway.
If
it
starts
to
like
there's
a
guy
in
there's
a
contributor
that
joined
not
too
long
ago,
his
name's,
aditya,
krishna
and
I'll
put
his.
You
know
slack
his
name,
he'd
been
hot.
Well,
my
wife
hates
it.
A
When
I
use
this
term,
he'd
been
hot
to
trot
on
potentially
using
docusaurus
instead
of
jekyll,
and
I
agree
that
docusaurus
is
more
feature-rich
by
example.
It
has
the
ability
to
version
your
documentation,
it
takes
care
of
this
like
we
don't
have
to
come
up
with
something
janky
it
does
this,
and
it
has,
I
think,
some
helpful
stuff
around
api
docs
and
and
like
it's
better.
A
The
there's
two
reservations
on
my
part,
just
solely
on
my
me
specifically,
my
my
individual
part,
is
anything
that
we
do
here
is
nothing's
ever
even
the
easiest
of
things
to
is
take
some
effort
to
get
right
and,
like
I
can't
tell
you
how
many
times
my
plate
spills
over
and
I
can't
assist
folks
and
so
people
will
start
something
and
then
I'll
feel
you
know
we
won't
get
it
done,
because
I'm
not
there
to
help
not
that
I'm
the
only
one
that
helps
and
get
stuff
done
and
stuff.
A
That's
not
it's
very
much
so
not
the
case.
I
very
much
desire
for
all
of
you
to
be
making
decisions
and
doing
doing
stuff,
but
my
point
was:
is
that
the
the
the
other
issue
around
it?
Is
that
so
it's
not
easy,
he
was
saying
he'd
do
the
whole
thing
he's
like
well,
that
would
be
fantastic.
There's
no
way
in
hell
that
that's
going
to
ever
happen
like
it's
evident
from
him
contributing
elsewhere.
A
I
wish
he
was
here
because
I'm
not
trying
to
be
rude
to
him
in
any
way,
and
he
already
acknowledged
that
and
stuff.
But
the
point
is
the
second
point
is
that
that's
a
react-based
framework
is
fine.
Technology
is
just
different
than
jekyll,
which
means
that
as
alonso
and
as
elizabeth
and
as
all
the
rest
of
us
there's
this
economies
of
scope
of
like
oh,
what's
the
mastery
io
sight
in
jekyll
the
s
p
site
jackal,
they
get
nighthawk.
Second
jackhaul
chuckle
jacket,
oh
great,
you
get
competent
in
one
and
you
can
go.
A
You
know,
go
round
the
projects
and
jekyll
is
not
in
any
way
shape
or
form
better
than
you
know
like
the
best.
It's
the
most.
It's
like
pearl
as
a
language,
if
any
of
you
know
pearl,
which
I'd
be
curious,
if
you
do
but
the
reason
that
it
continues
to
survive.
Even
the
authors
of
the
pearl
books
from
o'reilly
like
like,
like
these
individuals,
are
hunchbacked
long.
D
A
People
who
are
quite
wrinkly
and
old
in
age,
because
pearl
is
also
what
I'm
trying
to
say.
Why
is
pearl
even
still
around,
because
it's
ubiquitous
like
every
linux
system,
any
any
inx
system
you
go
to
has
got
pearl.
So
if
you
want
to
have
an
install
script
or
you
need
to
do
some
regex,
for
example,
like
man
pro,
is
there
guaranteed
like
you're,
even
though
it's
technically
not
part
of
the
you
know,
it's
not
part
of
the
kernel,
but
so
point
is
that's
kind
of
the
same
thing
around
jekyll.
A
Is
it's
not
the
best?
It's
been
there
forever.
It's
simpler,
it's
easier
to
use.
Unfortunately,
we
chose
a
theme
that
didn't
have
versioning
in
it.
Ironically,
and
actually
the
other
choice
here
alonso
would
be
to
use
hugo
because
we're
actually
literally
using
the
same
theme
that
tensorflow
and
kubernetes
that
I
o
and
others
they're
using
that
same.
This
is
a
ported
theme
that
we're
using
from
a
hugo
theme
and
the
hugo.
The
original
hugo
theme
actually
has
versioning
supported
and
we
could
move
to
hugo
and
it's
very
simple.
A
It's
very
similar
to
jekyll,
like
you,
don't
need
to
know,
go
at
all
for
jekyll.
You
end
up
touching
kind
of
touching
some
ruby.
You
get
around
some
ruby
and
mark
and
liquid
templating
and
stuff,
and
then
hugo.
I
don't
know
what
the
templating
is
hugo's
written
and
go.
But
the
point
is
like
the
docs
that
you're
writing
is
just
it's
just
markdown
again
with
front
matter
variables,
and
so
now
I've
really
expanded
up
the
potential
considerations,
but
I'll
list
aditya's
name
and
there's.
There's
a
new
there's,
a
dot.
There's
a
mesherie.
A
Sorry,
there's
a
repo
in
the
layer,
five
org,
it's
called
dox
ng
for
next
generation,
and
it
was
it
has
the
start
of
a
docusaurus
test
that
that
this
gentleman
aditya
was
doing
so.
A
A
So
alonzo
that
doesn't
mean
you
need
to
go
and
do
something
different.
I
just
was
calling
it
out
since
it's
relevant.
E
Yeah
yeah:
I
think
that
this
it
is
a
very
good
moment
to
see
another
alternative
even
to
jaco.
Yes,
I
I
agree,
because
we
have
not
now
seen
some
difficulties
of
how
this
is
going
to
be
a
minor.
I
don't
know
struggle
in
the
future
with
with
all
these,
these
other
actions
trying
to
automate
things
and
jekyll
being
in
the
in
the
way.
So
that's
not
good.
A
A
A
D
D
F
Yes,
so
I
was
thinking
you
know
that
a
documentation
is
part
of
the
application,
it's
part
of
the
source
code,
and
so
when
you
make
a
change
to
something
you
make
a
change
to
the
source
code.
You
because
change
the
documentation.
You
make
a
change
to
the
test.
You
know
the
whole
thing
goes
together,
it's
one
full
package.
F
So
as
you,
when
you
have
an
issue
or
vr,
you
fix
the
pr
you
then
then,
then
you
merge
it
and
then
it
goes
to
one
release.
And
then
you
know
the
documentation
always
stays
in
sync
with
the
development.
A
F
Okay,
just
just
just
as
an
example,
the
previous
one,
the
air
gap
right.
So
you
go
you
you,
you
go
to
six
point.
Air
gap
is
not
there
so
now
somebody
has
made
air
gap
possible.
Okay,
so
they
put
in
they
put
in
the
code
changes
and
assuming
that
there
are
code
changes
around
the
scripture,
just
they
put
in
all
the
code
and
all
and
and
they
make
the
dark
recommendation
right
and
then
they
they
check
in
the
whole
thing
at
once.
F
So
you
know
when,
whenever
you
make
a
chain
so
and
if
there
are
tests
to
go
with
it,
you
also
check
in
the
test
at
the
same
time,
so
that
there's
no
part
of
the
project
that
is
you
know
I
like,
like
me,
was
saying
earlier.
That
is
then
like
a
like
a
lost
last
friend,
or
something
like
that.
It
is,
you
know
they
all
go
together.
F
So
when
you
make
this
change
automatically,
when
they
go
to
that
particular
version,
they
know
that
this
aircap
is
there
and
when
they
go
to
the
previous
version,
it's
not
there
and
it
all
these
things
to
happen.
D
A
Nice,
I
actually
just
leave
them
for
a
long,
so
this
visual
is
a
kind
of
kind
of
a
good
example
of
like
well.
I
had
said
that
kubernetes
is
using
basically
the
same
thing
theme
that
we
are
using,
that
measure
io
is
using
or
I'm
sorry
docs.measuringio
is
using
and
you
can
kind
of
tell
by
the
same
font
and
and-
and
you
can
tell
you
can
really
start
to
tell
as
you
look
through
here.
It's
like
wow
yeah,
that
sort
of
looks
like
exactly
like
measures.
A
I
guess
that
is
they're
using
hugo
we're
using
jekyll
and
but
now
their
version
of
the
theme
actually
has
so
so
I
I
think
their
theme,
inclusive
to
the
theme
I
think,
has
support
inherently
for
versions,
but
even
at
what
we're
looking
at
it's
like
you
are.
This
is
a
little
bit
weird.
This
could
be
something
to
look
at.
It's
like
you
are
viewing
the
documentation
for,
and
so,
if
we
go
to,
if,
as
a
user,
we
go
over
and
we're
like.
How
do
we
run?
A
How
do
we
spin
up
a
pod
it'll?
Have
this
up
there
the
whole
time
and
if
that
command
ever
changed,
then
that's
why
you
would
want
to
know
like
hey,
there's
a
different
version.
Interestingly
in
the
url,
oh
here
it
is
you
see
this
url
up
here,
it's
so
they're
using
this
trick,
which
is
something
we
could
do,
they're
doing
a
subdomain
version
such
and
such
and
then,
when
you
want
to
go
back
to
the
latest,
it's
just
kubernetes.io
docs.
A
But
if
you
wanted
to
go
to
the
old
stuff,
if
we
went
back
to
v17,
then
so
that's
that's
another
way
of
like
okay,
hey,
we
published
a
totally
separate
site,
a
snapshot
of
it,
so
it
the
thing
I
was
sort
of
hand
waving
over
or
that
I
didn't
include
the
thing
that
I
said
there
might
be
some
unknowns
there,
like
one
for
like
look
at
this.
First
of
all,
they've
got
a
broken
link
because
it
you
know,
because
some
of
this
stuff
is
kind
of
hard
to
do.
F
This
is
it,
I
mean
it
would
be
really
nice
if
you
have
some
tool
that
builds
a
current
version
from
from
github,
I
mean
for
a
particular
version
for
github
and
generate
the
documentation
real
time.
A
A
But
but
that's
the
problem
is
because
it's
auto-generated
from
the
code
I
mean
the
great
thing
is
it
has
to
be
there's
no
way
you
can
do
that
for
a
language
like
it's
and
if
it's
an
api
like
yeah,
you
want
it
auto
generated
and
stuff
because
who's
who's
there
looking
at
it's
a
developer
who
wants
to
know
like,
but
for
the
user
facing
docs,
it's
like
well
even
here
this
is
the
user
facing
godox
and
these
are
not
auto-generated.
These
are
handwritten
they're,
still
not
pretty.
A
A
A
Like
for
idina
and
vijay,
and
everybody
that
there's
it's,
these
are
open
challenges
for
anybody
to
go.
You
know
latch
onto
and
figure
out.
A
C
So
if
not,
we
are
sorry.
D
Oh,
I
I
I
wanna
only
the
camera,
but
I
put
the
mic
also
well,
I
do
have
an
idea.
I
think
we
have
quite
some
bots
right
with
the
pipelines,
so
we
can
have
also
like
a
doc
or
or
whatever.
D
A
Oh
yeah,
yeah,
okay,
just
declare
yeah
you're,
not
talking
about
like
slack
bots
but
yeah.
No
there's
there's
no.
A
Yeah
there's
a
number
of
workflows
and
then
those
workflows
will
call
it's
confusing
because
the
language
is
like,
but
we
have
a
number
of
workflows
that
get
triggered
based
on
different
events
and
within
the
workflow.
They
may
refer
to
any
number
of
github
actions,
but-
and
yes
and
there's
lots
of
things
to
do
within
them,
like
one
of
the
things
that
I
think
is
going
to
upset
rodolfo
in
the
best
of
ways
is
I
bet
if
we
make
a
release
right
now
we
don't
get
auto-generated
and
auto-published
release.
Notes
like
I
bet.
A
D
Just
that,
the
last
time
or
I
don't
know
what
our
meeting
was
was
that
you
want
something
more
often
like
the
deployment.
I
think
it's
actually
releasing
more
more
often.
A
Yeah,
I
think,
to
help
with
that
is
something
that
rudolfo
had
done,
which
is
initiated
the
framework.
The
use
of
cyprus,
as
is
one
of
the
more
helpful
things
toward
raising
confidence
level
of
just
making
releases
more
often,
and
so
there's
a
lot
of
tests
or
there's
more
tests
to
be
written
for
cyprus.
A
That's
kind
of
a
nice
thing
because
you
don't
use
because
you
don't
have
to
go
pick
up
golang
or
go
pick
up.
I
don't
know
any
language
per
se.
You
do
have
to
look
at
the
code
and
you
know
pick
out
a
few
things,
but
even
I
could
do
it
is
what
I'm
saying
so.
A
Well
shoot
I'm
stressed
out:
I
need
to
go,
approve,
link,
rd's,
graduation,
I'm
getting
pounded
by
the
toc.
So,
okay,
that's
good
thanks!
Thank
you.