►
From YouTube: Meshery CI Meeting (Feb 25th, 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
A
This
get
nighthawk
is
the
new
project
that
we
are
working
on,
and
this
issue
is
for
create
some
artifacts
for
specific
platforms
as
get
a
body
has
in
in
their
web.
So
I
will
work
on
on
this.
One
is
just
a
github
passion,
and
if
I
have
some
questions,
I
will
reach
out
avery
check
or
cuma.
B
Sounds
good,
do
you
have
a
plan
of
action
on
how
we're
going
to
create
it
by
any
chance
a
a
plant?
Do
you
mean
yeah
yeah,
a
plan
on
how
the
action
is
going
to
be
created.
B
A
Yeah
thanks
for
that,
I
will
be
working
on
that
this
weekend,
and
the
second
topic
in
the
meeting
knows
is
that
we
need
more
cyprus,
good
first
issues.
We
shall
already
request
this
because
he
wants
he
wants
to
to
start
creating
a
cypress
test,
so
it
is
what
this
we
can
also.
I
will
create
some
other
issues
to
to
create
in
cyprus
tests.
C
On
that
topic,
I
wonder
the
the
messory
console
adapter
it
has.
It
has
a
collection
of
scripts,
bats
scripts
that
you
know
perform
you
know,
but
would
perform
mostly
integration
tests.
I
guess
is
what
you'd
refer
to
them
as
and
ignoring
bats
or
cypress,
or
the
the
the
tech
that
gets
used
to
do
those
tests
or,
for
my
part,
I'll,
be
ignoring
them.
C
I'm
bringing
this
up
to
say
that
the
other
adapters
could
really
use
you
know
once
we
get
that
figured
out
that
can
be
proliferated
across
other
adapters,
and
so
I'm
tossing
it
up
as
a
food
for
thought.
Like
I
don't
know,
if
we
don't
have
to
make
it
a
discussion
necessarily
on
bats
versus
cyprus
or
we
can,
it
doesn't
really
matter
to
me
which
one
you
I
don't
or
I
don't
have
an
opinion
rather.
C
Integration
tests-
yep
good-
if
you
can
are
you,
can
you
navigate
to
the
measuring
console
repo.
C
And
if
we
look
through
this
read
me
a
little
bit,
I
thought
that
so
michael
gefeller
had
taken
on
this
initial
collection
of
things
to
do.
If
you
look
at.
C
A
C
C
Well,
if
you
yeah,
when
you
think
about
it,
you
know
we,
we
have
the
the
master
test
strategy,
doc
that
that's
mostly
empty
at
the
moment,
but
it
like
these
types
of
thoughts
would
probably
go
in
there
and
it
would
be
things
like:
okay,
hey,
if,
if
you
were
just
working
on
an
adapter
and
you're
making
some
changes,
you'd
like
to
confirm
that
that
things
are
working
in
accordance
with
how
they
should
be,
you
don't
have
any
regression
issues.
You
probably
you
know
you're
going
to
have
some
a
couple
of
unit
tests.
C
Probably
an
integration
test
like
this,
because
management
software
like
measuring
you
know,
combines
a
few
different,
two
things
together,
and
so
you
know
in
order
to
guarantee
that
it
works.
You
do
an
end.
You
know
functional
testing
like
that,
okay,
good,
so
so
you'll
want
to
like
confined
to
the
system
that
you've
changed.
In
this
case
the
adapter
you'd
like
to
be
able
to
run
some
of
those
integration
tests.
C
Great,
so
something
like
this
facilitates
that
as
you,
but
if
you
step
back
and
say
well,
hey
as
an
overarching
system,
what's
what's
the
manager
of
these
adapters
or
how
do
these
adapters
run?
Well,
they
get
invoked
by
meshri
that
measuring
gets
invoked
by
either
users
in
the
ui
or
users
in
the
cli.
C
And
I'm
pointing
out
is
that,
like
it's
valid
just
to
have
integration
tests
here
at
this
component
level,
it's
also
valid
to
have.
You
know,
function,
integration
or
functional
tests
at
that
high
level,
to
the
extent
that
collectively
we
can
write
an
integration
test
once
and
and
either
execute
that
same
thing
at
a
higher
level
or
at
an
individual
component
level.
C
If
we
do
that,
that
sounds
like
an
immediate
win,
because
there's
a
template
to
follow,
it
seems
to
be
working.
You
know
it
seems
to
be
relatively
stable.
It's
you
know
it's
bash,
which
you
know
the
majority
of
the
contributors.
C
You
have
some
comfort
level
with
or
there's
lots
of
help
out
there
you
know,
but
is.
Are
we
duplicating
our
efforts
if
we
have
end
to
end
ete
tests
performed
through
cyprus
at
that
higher
level?
At
measuring?
You
know,
measuring
system
level
measuring
server
level
and
then
different
tests
written
in
a
different
framework
over
here?
C
And
then
okay
good
good,
perfect,
and
then
I
guess
to
try
to
answer
that.
That
second
part
of
the
question
is
like
when
the
functional
tests,
some
of
the
functional
tests,
we
have
from
actuary
server
that
get
invoked
through.
You
know
as
and
when
someone
does
a
pull
request.
We
invoke
a
few
different
tests
in
cyprus,
some
of
what
and
it's
fine
that
github
as
our
workflow
engine
or
our
ci
engine
repeats
itself
or
does
duplicative
work
like
tests
the
same
things
a
couple
of
times
like
great
so
much.
C
Rather,
things
get
tested
19
times
than
zero
times
so
and
in
some
respects
to
the
extent
that
a
different
test
harness
a
different
framework
cyprus
over
here
and
bats
over
here,
are
used
to
attack
that
same
problem.
Great
the
you
can
get
likely
we'll
have
a
more
robust
and
higher
quality
feet
function
because
you
know
it,
but
the
trade-off
could
can
be
the
duplicity
and
the
sustaining
the
maintenance
of
the
sustaining
of
those
two
and
we
may
not.
Maybe
we
might
not.
I
I
don't
want.
C
I
hesitate
to
offer
up
a
suggestion
or
or
here,
but
I'm
I'm
raising
it
up
as
a
per
question,
for
you
know
all
of
you
or
or
maybe
rudolfo,
if,
if
this
is
the
week
to
potentially
give
an
update
in
the
community
call
about
you
know
goings-ons
in
this
ci
meeting,
that
might
be
a
question
that
either
you
can
answer
like
just
say
hey.
C
This
is
your
your
sense
of
it,
and
this
is
what
you
think
we
should
do
or
raise
it
up
as
a
question
for
everybody
else
to
pontificate
on
for
a
while
and
then
even
at
that
in
the
end,
I'm
you're
still
gonna
have
to
say
hey.
This
is
the
answer
like
you
know,
unless
somebody
else
has
got
a
super
strong
opinion.
A
Yeah
sounds
good
in
which
meeting
do
you
refer?
Tomorrow's
community
call
tomorrow's
okay.
If
you're.
C
C
C
You
know,
first
of
all,
what
a
sigh
of
relief
from
on
my
part
that
I
was
taking
it
in.
Second,
there
was,
you
know
the
concern
around
fragility
of
the
test
and
like
oh
no,
this
is
just
you
know.
Yet
another
thing
to
try
to
you
know
maintain
and
the
things
that
you
had
highlighted
about
how
to
avoid
fragility
of
how
you're
writing
your
test,
like
you,
know
great,
that
inspired
a
lot
of
confidence
and
and
and
so
but
yeah.
A
Yeah
but
yeah
we
can
start
working
on
that.
The
other
topic
or
the
next
topic
in
the
meeting
notes,
is
consideration
of
measure
production
deployments.
C
This
a
little
bit
more,
it's
what
hap,
when
sean
just
you
know,
joins
the
meeting.
That's
what
happens
it
so
shawn
to
try
to
poke
at
you
and
harass
you
a
bit
someone
it
was
yesterday
in
and
by
the
way
before
I
explain
this,
just
real,
quick
sean.
You
can
hear
me.
C
C
Okay,
good
deal
so
sean.
I
say
this
with
a
big
grin
on
my
face.
So
as
a
don't
clearly,
I'm
not
committing
you
to
anything
that
you
don't
want
to
do,
but
but
you
came
to
mind
when
this
issue
arose
yesterday,
so
to
recap
it
for
you
real
real.
C
Briefly,
we
were
on
the
weekly
mesh
recall
on
wednesdays
the
message
development
call
one
of
the
contributors
he's
creating
he's
refactoring
the
command
line,
client
for
meshri
meshri
ctl,
he's
refactoring
how
it
uses
the
oh,
a
a
jaw,
a
jwt
token
from
mescheri
server,
so
that
whenever
you
know
so
that,
whenever
a
user
of
the
command
line
client
wants
to
invoke
certain
operations,
certain
api
operations
inside
of
mesh
reservoir,
you
need
to
also
present,
you
know,
be
auth
authenticated,
and
yet
we
really
don't
have
an
authorization
system.
C
Yet
that's
just
roadmap
there,
but
so
he
was
just
asking
the
question
around
well
like
okay,
great
so
so
it
makes
intuitive
sense
to
him
that
it,
you
know,
as
as
a
user,
that
signs
into
mescheri
server
mastery
hasn't
that
you
have
a
session.
You'll
grab
your
token
you'll
use
that
token
you'll
reference.
It
on
the
command
line
and
great
you'll
be
able
to
invoke
you
know.
Do
administrative
things
do
do
powerful
things
from
there
measuring
server
meshery.
C
When
you
go
to
sign
in
you
can
today,
you
can
sign
in
you
can
authenticate
through
a
couple
of
different
or
four
different
social
profile
systems,
social
logins-
or
you
can
say
you
know
I
actually
don't
I'd
rather
not
I'd
rather
just
use
this
other
provider.
It's
called
none
and
basically
you
use
meschery
in
a
stateless
way.
So
it
doesn't
persist
it
don't
really.
You
don't
really
have
a
persistent
session.
C
It's
still
it's
it's
an
imp.
It
becomes
implicitly
it
becomes
a
single
user
tool,
even
though
user
is
sort
of
a
it
becomes
a
second
class
citizen,
the
user
as
a
construct,
becomes
a
second
class
citizen
when
you're
using
it
like
that,
and
so
his
question
was
well.
Okay,
hey
great
so
as
I'm
using
mastery
ctl
and
the
command
line.
Client-
and
I
want
to
present
a
token,
but
let's
say
that
I'm
using
meschery
in
anonymous
mode
in
none
mode.
C
What
then,
and
so
we
went
through
this
exercise,
just
on
the
call
of
showing
what
the
behavior
was
and
that
the
behavior
was?
Is
you
you
sign
into
mastery
server
on
authenticated
or
I'm
sorry,
not
unauthentic
in
a
non-user
mode
like
and
a
session
is
generated,
you
can
download
that
anonymous
user's
token
present,
that
to
the
command
line,
client,
and
it
will
use
it
to
invoke
operations.
C
E
C
And
the
first
action
item
coming
out
of
that
was
well,
you
know,
there's
some
this
discussion
around!
Okay!
Well,
hey
you
know
in
in
a
production
environment,
maybe
disable
the
none
provider.
You
know
maybe
mandate
that
everyone
has
to
authenticate
so
that
sort
of
removes.
E
The
one
of
the
problems
we
have
with
something
like
that
is
that
there's
nothing
other
than
those
social
media.
You
know
like
linkedin,
google
ways
to
authenticate.
It
would
be
nice
if
there
was
a
way
to
link
it
into
like
a
ldap
or
an
active
directory
infrastructure.
That
way
we
could
know
like
who
and
what
and
then
people
could
be
assigned
different
roles.
E
That
might
be
no
small
task,
but
it's
something
I
think
it'd
be
nice
to
think
about
in
the
long
run,
because
there's
some
environments
where,
based
on
the
need
for
that
log,
in
with
like
a
social
media
provider,
you
know,
like
our
my
organization,
can't
use
that
it
wouldn't
be
allowed.
C
That's
just
what
you
need.
It's
a
phone
sign
or
a
phone-sized
ahead
of
me,
so
you're
you're,
one.
E
Of
the
things
you
know
we're
doing
some
of
our
engineering
in
in
in
cloud
platforms
like
azure
and
aws,
and
you
know
we're
looking
at
like
you
know,
active
directory,
you
know,
infrastructures
that
teams
can
can
leverage
for
access
control
so
that
might
yeah.
I
don't
know
what
the
lift
is
there
like.
I
don't
know
what
the
level
of
effort
is
to
to
take
measuring
and
make
it.
You
know
active
directory
integrated.
If
you
will.
E
The
the
ability
to
to
do
the
social
media
piece
I
think
there
would
be
some
sort
of
plug-in
or
some
sort
of
module
or
something
that
would
make
it
not
too
hard.
But
it's
also,
I'm
not
sure
of
the
you
know
the
level
of
interest
from
the
or
I
mean
the
level
of
need
from
the
community
right,
because
it
seems
like.
E
D
E
E
You
know
and-
and
I
can't
pronounce
it
correctly
visit
titrate
the
guys
did
some
of
the
istio
work
at
google
and
then
they
formed
their
spin
off
and
they
they
had.
You
know
they're
they're,
pretty
tightly
integrated
now
with
national
institute
of
standards
and
technology.
C
E
C
C
Yep
so
I'll
say
it
like
this
in
the
beginning
of
the
mastery
project,
this
the
need
for
or
consideration
for
the
inevitability
around
ldap
or
oidc
blah
blah
blah
the
the
azure
active
directory
sort
of
the
list.
So
to
speak,
the
I
you
know,
aws.
I
am
etc
like
that.
C
Those
were
noted
down
from
the
start,
like
hey
as
soon
as
we
start
to
auth
or
authenticate
anything
like
in
the
and
yeah
you're
right
that,
like
often
historically,
and
it
continues
to
be
the
case
that
a
lot
of
times
you'll
find
the
there
in
line
the
line
of
a
demarcation
point
between
open
source
and
enterprise
or
like
the
enterprise
version
of
that,
like
an
enterprise
add-on.
If
you
will
right,
we
are
that's
totally
in
the
plans
or
like
yes,
and
I
don't
mean
necessarily
in
an
evil
scheme
to
go.
C
You
know
to
charge
people
for
the
for
that
component.
That's
not
what
I
meant
so
much
about
in
inputs
in
the
plans,
but
rather
there's
been
a
ton
of
work
put
in
to
make
sure
that
the
providers,
the
two
that
are
there
today
mesherie
and
none
that
that's
a
total,
completely
extensible
interface
and
that
right
to
support
just
just
authentication,
yeah,
pretty
darn
like
that
ain't
that
it
that's
not
such
a
pretty
yeah.
C
We,
I
think
I
think
the
community
can
pretty
quickly
go
accomplish
that
for
authorization
to
be
able
to
say
all
right.
Wait
wait
a
second!
So
we
know
you
know,
I
don't
know
why.
I
would
be
explaining
authentication
and
authorization
to
you
so
never
mind
anyway,
for
authorization,
maybe
a
little
bit
of
a
longer.
We
need
to
lay
out
some.
That's
totally
in
you
know
it's
it's
actually
listed
as
a
two.
Is
it
three
milestones
away
to
have
a
full
authorization
framework
where
yeah
it
will?
It's
considerations
I
think
and
actually
sean.
C
This
would
be
something
great
for
you
to
get
to
hear
part
of
your
perspective
on.
I
think
it's
considerations
for
a
couple
of
things
that
that
we'd
like
to
provide
a
lot
of
authorization
around
you
know
who
well
it's
not
just
who
can
do
what
on
the
service
mesh
who,
what
personas
or
what
user
groups
are
mapped
to
being
able
to
change
this?
This
can
you
know
this
category
of
configure
configuration
on
a
mesh
versus
this
category
of
configuration
for
that?
C
That's
more
like,
like
there's,
there's
some
of
the
tools
out
there
that
might
bucket
eyes
these,
and
I
don't
think
it's
inappropriate,
but
they'll
bucket
is
saying
like
hey.
If
you're
in
the
app
dev
group,
you
are,
you
know,
afforded
access
to
change
the
retry
count
for
between
your
services.
If
you're
in
the
platform
user
group,
then
you're
authorized
to
make
changes
to
the
sidecar
attachment
policy
or
what
you
know
like,
and
those
are
good,
our
perspective
is-
and
this
is
this-
is
landing
in
a
significant
way
in
the
next
measuring
release.
C
There's
early
there's
early
aspects
of
it
now,
but
ours
is
our.
The
mastery
authorization
framework
will
cover
in
high
detail
the
workloads
as
well
the
applications
themselves,
irrespective
of
how
that
reflects
on
the
service
mesh
config,
we
will
also
allow
people
to
control
the
behavior
of
third-party
components
like
a
flagger
or
an
argo
cd
or
like
whatever
canary
tool,
that
you're
using
such
that
their
control
authorization.
There's
controls
afforded
around
that
as
well
yeah.
C
Some
of
that
was
pie
in
the
sky.
Would
I
just
some
of
this
is
not
pie
in
the
sky,
because
the
it's
behind
the
sky
that
there
is
an
authorization
framework
to
come
forth.
It's
not
pie
in
the
sky
that
those
control
that
meshery
provides
those
or
has
control
over
those
things.
Today
it
does
you
just
need
to.
C
Not
that
I
want
to
talk
more
about
competitors
like
tetra8,
on
the
call
but
curious.
So
you
know
highly
familiar
with
have
been
have
been
there
during
their
founding
and
it
had
there's
a
reason
why
layer
five
is
a
thing
separate
from
from
tetrad
much
respect
for
those
guys
from
actually
do
have
a
real
great
relationship
with
them.
In
specific
we've
got
this
next-gen
next-gen
our
back
or
what
do
they
call
it,
but
it?
But
it's
a
it's
a.
E
C
F
E
C
E
You
know
and
looking
at
kubernetes
security,
because
what
we're
finding
is
that
people
will
deploy
pipelines,
but
it's
the
security
not
in
the
pipeline
but
of
the
pipeline.
We've
kind
of.
D
E
Looking
at
oasp
the
oauth
software
assurance
maturity,
model,
separate
conversation
but
interesting.
C
C
Along
so
you
had
had
some
notes
I
think
coming
out
of.
Was
it
the
messiah
development
call.
C
But
I
guess
the
the
point
here
is
like
we
do
have
a
little
bit.
We
do
have
some
time
on
this
call
and
also
I'm.
I
wonder
if
we
can
significantly
advance
your
effort
by
giving
this
a
review
here
of
the
a
review
of
the
selected
plug-in
and
then
a
review
of
how
that
plug-in
and
its
requirements
is.
It
will
affect
the
rest
of
the.
B
C
D
F
Yeah,
I'm
sorry.
I
was
trying
to
to
tell
you
that
the
plugin
doesn't
work
at
all.
Oh
yeah,
I
don't
know
which
are
the
conditions
for
it
to
to
make
it
work
with
the
with
the
repo,
but
I
didn't
be
I
I
couldn't.
I
didn't
make
it
work
with
anything
with
the
post.
You
know
the
jekyll
generates
like
a
post
with
a
date
in
the
in
its
name
and
such
it
works
very
good
with
posts
in
that
manner.
F
C
F
Alternative
and
that
I
saw
on
youtube
and
and
that
work
and
and
the
the
thing
with
it
is
a
plugin
that
in
the
default
main
and
the
default,
I
forgot
the
word
sorry.
But
in
the
default
template
it
kind
of
gets
the
language
variables
from
the
front
matter
of
the
site
and
on
from
the
pages.
F
C
It
sounds
like
it
sounds
like
we
were
sort
of
aligning
for
it
in
advance
already
that
on
each
individual,
docs
page
by
the
way,
I'll
just
familiarize
everybody
real
quick,
because
because
I
know
as
because,
because
if
sean
keeps
showing
up,
he'll
he'll
eventually
be
in
in
here
as
well,
along
with
all
the
rest
of
us
and
so
there's
this
is
on
in
the
mesherie
repo
there's
this
slash
docs-
and
this
is
this-
is
the
jekyll-based
site
for
each
of
the
pages
we
had
in
the
recent
past
gone
about
putting
a
small
variable
at
the
top
of
each
file,
and
hopefully
this
doesn't
make
me
a
liar
okay,
this
one.
C
Okay,
so
this
one
doesn't
so
this
particular
markdown
file.
It's
it's
a
document
about
extent,
mesher's
extensibility
as
a
matter.
This
is
funny
it's
a
document
about
measures
extensibility
on
providers,
which
is
talks
about
our
back
and
authenticating
and
authorization,
so
it's
kind
of.
But
when
we
look
at
the
the
markdown
itself,
the
this
isn't
the
best
easiest
way
to
read
this,
but
just
there's
this
this
section,
these
these
three
dashes
are
the
is
the
metadata,
is
the
front
to
use
the
jackal
nomenclature.
C
It's
the
front
matter
for
a
particular
page,
whether
that
page
is
a
post
or
a
piece
of
news
or
a
document
or
a
whatever's
inside
there.
It
doesn't
doesn't
really
matter.
That's
in
fact,
that's
just
up
to
us,
whatever
that
is,
and
in
in
also
the
variables
that
are
used
here.
C
These
are
also
up
to
us,
there's
some
built-ins
that
jekyll
comes
with,
but
but
we
can
make
up
stuff
just
like
this,
when
we
just
made
up
this
language
one
and
we're
beginning
to
tag
each
of
the
individual
markdown
files,
each
of
the
individual
pages
with
you
know,
language
moniker,
and
so
it
sounds
like
along.
So
this
last
plugin
that
you
were
looking
at
you
know
leans
into
that
type
of
thinking
that
type
of
a
system.
Yes,
it
does.
It's
not.
F
F
We
have
some
variables
that
you
may
see
redundant
that
that
there
are
a
lot
of
a
lot
of
variables
that
repeat
the
same
language
moniker,
but
it
has
to
be
like
that
for
us
to
to
make
the
the
the
filtering
and
showing
sometimes
the
the
tree
of
of
contents
on
the
left
of
the
of
the
sidebar
and
sometimes
and
in
the
front
of
the
indexes
of
each
section
and
sometimes
for
changing
the
language
of
the
whole
page.
F
So
there
are
a
couple
of
of
variables
in
the
firmware
that
are
filtering
some.
If
statements
in
several
files
also,
I
grabbed
the
the
data
folder
in
the
in
the
docs
and
yeah
folder,
yeah
and-
and
I
have
been
making
there
some
changes
to
make
this
work
with
a
lot
of
the
sections
that
were
only
accessible
in
english.
F
C
The
the
only
complaint-
the
only
thing
you
could
possibly
complain
about
is
you're,
saying
hey
in
the
you
know
in
the
front
matter,
depending
upon
the
type
of
content
that
you're
dealing
with
you
may
end
up,
it
may
take
a
couple
of
variables
in
the
front
matter
to
differentiate
between.
F
But
with
this
script
scripting
it
working,
it
is
working
right
now.
The
only
thing
that
I
need
right
now
is
the
fall
turkish
documentation,
translations
and
well
done
we're
done
with
this
wow.
C
F
Yeah
I
have
some
when
I
make
a
pr:
it's
the
the
netlify
new
preview
is
generated
every
time
and-
and
it
has
functioned
me
yeah-
that
one
so
I
I
have
seen
in
that
it
works
and
and
it
kind
of
relief,
because
yeah-
and
maybe
there
are
some
broken
links
that
I,
as
I
told
yesterday,
to
chandra
that
broken
list
in
the
actual
english
site.
C
Sure,
yeah
totally,
I'm
also
very
so
this.
This
is
fantastic
and
nice
work
as
a
matter
of
fact,
alonso
like
just
a
two
to
three
just
a
couple
minute
demo
of
this
on
tomorrow's
community
call
might
might
make
some
sense,
it'd
be
good
for
you
to
get
some
air
time
and
it's
something
you've
been
working
on
for
a
long
time,
so
good
good
for
you
to
get
some
feedback
on
it.
For
my
part,
I'll
take
a
look
look
later
today
to
try
to
help.
C
You
know,
I'm
support
and
I'll,
probably
message
mert,
the
turkish
guy,
who
was
doing
some
of
that
translation.
C
Sure
and
then
yeah,
if
don't
be
yeah,
please
don't
be
shy
about
if
you
do
find
broken
links
on
the
in
the
current
version.
But
you
know
please
don't
be
shy
about
raising
up
those
issues.
There's
one
other
thing
here
that
it's
a
bit
bothersome
and
I
it
sounds
like
from
the
way,
the
way
that
the
script
works.
It
don't.
C
I
think,
it'll
just
be
a
sequencing
of
potentially
two
different
scripts,
the
docs
right
now
they
don't
have
a
version
like
you're
not
able
to
switch
between
one
mesh
reversion
and
the
next
in
terms
of
like
the
documentation.
C
C
F
F
C
So
very
good
alonso
then
thank
you
very
much.
I'm
rudolfo
and
back.
A
A
So
to
the
next
topic
and
final
topic
on
the
mini
notes
and.
C
A
Yep
and
another
thing
that
I
just
remembered
is
a:
I
have
an
issue
about
ec2
github
actions
in
order
to
automate
some
deploys.
Could
you,
let
me
know
when
you
or
abhishek
or
kuma
will
release
a
machine
cloud
and
version,
because
I
I
want
to
to
see
that
in
order
to
automate
that
with
github
actions.
B
Currently,
we
don't
have
an
action
that
would
deploy
to
ec2,
but
rather
we
just
build
and
release
the
docker
image
and
we've
got
a
watcher
instance.
That's
running
in
the
to
instance,
which
would
update
the
image
as
soon
as
it
has
been
pushed.
B
C
B
So
the
ec2
instance
would
be
of
help,
because
if
we
need
to
revert
and
do
other
other
release
operations,
it
would
be
straightforward
with
the
ec2.
Rather,
we
don't
need
to
basically
ssh
into
the
easy
to
do
manual
stuff.
The
manual
stuff
would
be
would
be
automated.
C
So
rudolph
was
asking
for
some
specifics
on,
as
he
investigates
an
ec2
action.
What
should
he
have
the
action.
B
Do
is
that
a
issue
I
can
list
it
down
in
that
yeah.
A
B
Right,
okay
is
just
a
set
of
make
comments.
I
can
maybe
put
that
in
as
well.
D
D
B
B
I
couldn't
I
didn't
understand.
Yeah.
C
Yeah
dino
was
saying:
hey,
that's
interesting
when
you
guys
talk
about
that.
Can
you
loop
me
in?
Can
you
loop
in
adina
and
so.
B
C
Okay,
nice,
one
other
all
right
good
deal.
We
just
have
a
couple
minutes
left,
there's
one
other
thing
that
it's
related
to
this.
I
hope
that
this
is
interesting
to
everyone.
It's
related
to
what
rodolfo
was
just
saying
about.
Like
hey.
You
know
I
want
to.
I
want
to
help
out
here,
I'm
a
little
bit
some
of
the
details
about
how
it
is
that
a
release
is
made
are
are
unknown
to
him
because
he
hasn't.
C
And
I
thought
we
could
it's
it's
actually
a
very
short
topic,
so
I
thought
we
could
show
it
real,
quick
and
actually
we
can
confirm
whether
or
not
rodolfo
you
have
the
necessary
permissions
to
do
so
in
the
system.
So
rudolph.
If
you
don't
mind,
if
you're
able
to
share
the
message
repo,
when
you
know
every
time
that
someone
submits
a
pull
request,
various
workflows
kick
off,
various
tests
are
done,
humans
come
and
do
peer
reviews
and
comments,
mixing.
C
Eventually
the
thing
merges
when
a
pr
merges
in
any
of
the
repos
an
edge
release
is
made.
So
it's
a
new
edge
release
of
of
that
release.
Channel
becomes
available
the
meshry
stable
release
channel.
There's
the
a
new
release
only
goes
into
the
stable
channel
when
not
when
a
pr
merges
to
master
that
that's
when
an
edge
release
is
made,
a
stable
release
is
made
when
a
human
goes.
If
you
go
when
they
go
to
the
releases
section.
C
So,
if
rudolph,
if
you
were
to
navigate
back
to
the
main
folder,
okay
and
the
lower
right
hand,
side,
there's
a
releases,
there's
like
it
says,
82
so
releases
here
so
so
the
process
is
right
right
here,
a
very,
very
simple
process.
It's
so
it's!
What
we're
seeing
here
on
the
measure
measuring
releases
is
that
there's
a
release
in
draft
now.
C
It
says
for
it
says
that
lee
drafted
this
release,
although
technically
I
didn't
all
that
I
think
has
happened
is
like
is
that
I
was
the
last
one
to
touch
the
system,
so
it
sort
of
reflects
as
me
being
that
that
person,
but
anyway,
there's
a
small
github
action
that
every
time
you
merge
a
pr
anyone
merges
a
pr
a
small
github
action
runs.
It
invokes
a
small
bot
called
release
drafter
and
that
small
bot
just
captures
the
subject
line
of
the
pr
the
person
who
the
author
of
the
pr
and
the
pr
number.
C
So
basically
what
you're
seeing
right
there,
the
general
the
documentation
it
just
comes
over
and
in
advance
of
a
release
being
made
it
just
it
just
automates
the
compilation
of
release
notes.
So
it's
just
every
time
a
pr
is
merged,
it
grabs
the
title
and
link
to
it
and
it
puts
it
in
here
and
it
also
generates
what
it
thinks
will
be
the
title
of
the
next
release.
C
It
thinks
the
title
of
the
next
release
is
v051
and
it's
wrong,
but
it's
taking
a
guess
based
on
semantic
versioning,
because
we
follow
semantic
versioning
and
so
does
release
drafters.
So
do
a
lot
of
things
if
you
scroll
and
so
that's
under
draft,
it's
not
public.
It's
not
it's
not
part
of
the
stable
release.
Yet
if
you
the
way
that
a
release
is
actually
made,
then
is
if
you
go
over
and
click
on,
so
that
that
bot
is
just
helping
us
get
to
a
point
where
we
can
just
click.
Okay,
release
so
root
alpha.
D
C
Final
release
out,
hopefully
today
we
want
to
announce
that
we
want
to
tell
people
right
now
the
latest
release,
if
you
scroll
down
just
a
little
bit,
it's
v05
it's
release
candidate
for
we've,
gone
through
like
three
different
beta
releases
and
four
different
release
candidates,
and
I
think
we've
worked
out
most
of
the
bugs
and
I
think
you
know
I've
been
asking
people
publicly
for
a
while.
Are
you
ready?
Are
you
ready,
I
think
we're
ready.
C
Here's
what's
manual
for
making
a
release
and
it's
intentionally
manual,
because
it's
you
know
a
human
should
decide
we're.
Okay
with
it.
You
know
this
is
good.
So
in
this
case
you
would
change
the
suggested
tag
number
to.
D
C
And
then,
in
the
other,
in
the
subject
line,
it
would
be
measury
space
v0.5.0
and
that's
it
then,
when
you
click
publish
release
and
that's
it,
then
the
workflow
kicks
off.
It
builds
the
stuff
and-
and
so
everything
else
is
automated.
So
at
one
point
we
were,
we
were
saying:
hey
I
was
asking
in
the
community:
does
anyone
want
to
be
the
release
manager
for
this
release
or
for
the
next
one,
or
what
have
you
and,
and
naturally
people
say
well,
that
sounds
pretty
interesting.
C
Like
I,
you
know
I
want
to
raise
my
hand,
but
I'm
not
really
sure
what
that
entails
and
part
of
what
it
entails
like.
This
is
the
smallest
part.
It's
like
okay,
I
actually
released
the
thing.
The
harder
part
is
like
no,
it's
the
herding
of
cats
and
the
hey.
We
need
to
write
up
a
couple
of
blog
posts
that
explain
what
these
new
features
are.
We
need
to
make
sure
that
that
was
documented
there.