►
From YouTube: Keptn - GSoC Preliminary Reports - July 20, 2021
Description
Preliminary discussion on GSoC projects ahead of the presentations on the week of July 25th. Meeting notes: https://docs.google.com/document/d/1y7a6uaN8fwFJ7IRnvtxSfgz-OGFq6u7bKN6F7NDxKPg/edit
Learn more: https://keptn.sh
Get started with tutorials: https://tutorials.keptn.sh
Join us in Slack: https://slack.keptn.sh
Star us on Github: https://github.com/keptn/keptn
Follow us on Twitter: https://twitter.com/keptnProject
Sign up to our newsletter: https://bit.ly/KeptnNews
B
The
captain
community
meeting
on
july
20th
2022
a
reminder
that
we
abide
by
the
captain
code
of
conduct
play
nice.
This
is
a
special
session.
It's
dedicated
to
an
early
view
of
our
gsoc
projects.
For
this
year.
We
have
four
of
them
and
the
phase
one
ends
next
monday
and
they
will
all
be
making
formal
midterm,
appreciations
and
being
evaluated.
B
So
this
is
a
preliminary
a
chance
for
us
to
give
them
some
feedback
and
then
to
get
any
help.
They
have,
and
mostly
just
be
wowed
and
stunned
by
how
marvelous
they
are
so
format
we're
gonna,
let's
well,
so
we've
got
15
minutes
for
everybody.
Let's
try
to
do
it
12
and
then,
if
we've
got
more
discussions
we
want
afterwards
we
can
anybody
who
wants
to
can
hang
around
for
them,
but
let's
try
to
make
sure
that
everybody
gets
to
get
in
in
the
first
hour.
B
So
turning
it
over
to
first
project
is
maha
with
mentor,
brad
and
thomas
so
brad.
Take
it
away.
A
A
You
know,
such
as
argo
and
flux
at
the
moment,
is
in
square,
so
essentially
we're
extending
get
ups
with
ci
event
driven
crcd
and
it
doesn't
have
to
be
crcd,
but
we
we
get
a
notification
from
the
the
tool
that
we're
out
of
sync,
and
then
we
essentially
invoke.
A
We
trigger
a
captain
sequence,
so
I'll
share
my
screen
and
then
let's
go
over
sorry.
I
just
have
one
screen
today
as
well,
so
everyone
can
see
the
notification
controller,
yep,
yep,
okay,
so
in
flux
we
it's
sort
of
broken
down
into
components
as
well
similar
to
captain.
A
A
Well,
I
wouldn't
really
say
their
events,
but
they're
called
notifiers,
so
you
can
a
notifier
can
be
like
slack
teams
discord
generic
web
hook,
there's
there's
a
lot
of
different
notif
notifiers
and
it
auto
sorter
does
the
other
way
as
well.
So
you
can
have
web
hooks
on
github,
etc
and
then
each
of
those
down
below
you'll
see
the
get
ups
reconciled.
So
the
source
controller,
customize
controller,
helm.
A
What
we're
saying
here
is
that
give
me
an
alert
when
that
you
can
specify
your
events
severely
and
so
give
me
an
alert
when
the
event
source
is
a
filter.
So
you
could
say
when
a
get
repository
has
changed
when
a
helm
release
has
changed
when
customized
has
changed
and
you
can
be
very.
You
can
filter
this
right
down,
so
you
can
have
values
for
name,
namespace
labels
etc.
Sorry,
so
when
an
event
happens
on
those
specifications,
it
will
then
trigger
the
alert.
A
A
In
this
case,
it's
a
slash
so
it
will
send.
You
can
see
that
sending
a
web
book
similar.
A
D
A
C
A
C
A
And
then
so
at
the
moment,
the
challenge
that
we
have
with
flux
today
is
that
it's
an
opinionated
payload,
so
you
can't
customize
what
it
is.
So
obviously
that's
not
ideal
for
sending
a
cloud
event,
because
then,
although
it's
templated
so
an
example
of
what
that
payload
looks
like
is
here
so
you
can
see.
This
is
me
just
mimicking
what
flux
sends
the
notification
controller
so
you'll
see
here
we're
using
the
pod
tattoo
head
as
well.
E
A
Too
much
but
but
this
is
enough
to
so
so
that
that
was
a
challenge.
You
can't
output
a
cloud
event.
So
therefore,
until
we
I
I
ideally
we
want
to
actually
make
the
changes
in
flux.
So
it's
submitted
where
you
don't
have
to
have
another
service
that
we
have
to.
A
So
you
know,
we've
probably
already
got
too
many
already,
so
they
get
hard
to
update
for
breaking
changes
and
documents
and
yeah.
We
ideally.
A
That
back
to
flux
and
then
fox
will
just
send
the
cloud
again.
So
therefore,
that's
the
problem
that
we're
trying
to
solve
so
we've.
Actually,
we
have,
if
I
deciding
what
the
let
me
just
get
my
code
up
and
we'll
walk
through
the
code.
A
So
what
we've
done
is
we've
just
made
a
helm
chart
for
this.
That
will
help
folks
to
and
be
able
to
play
around
with
this
and
change
different
providers,
different
alerts
as
well,
so
I've
been
working.
We
have
been
working
with
adam
to.
C
A
So
we
would
do
changes
in
flux
and
then
we
would
trigger
like
a
k6
test,
which
is
quite
cool,
so
I
I
we've
done
a
bit
of
a
readme
here
as
well.
So
the
first
thing
you
need
is
flux,
which
is
very
easy
to
do.
You
get
your
github
token
and
you
just
have
another
repository
similar
to
the.
When
you
do
your
captain
project
with
the
shipyard,
you
need
a
repository,
and-
and
this
is
the
secret
that
I
was
talking
about,
so
what
what
we've
done
in
this
integration
is.
A
A
A
So
we
would
put
that
to
captain
once
we
have
the
ability
to
send
the
cloud
event-
and
you
can
see
here
that
we've
added
body-
and
this
is
using
go
templating
to
the
the
minimum
fields
that
you
need
is,
as
you
probably
know,
when
you're
triggering
the
sequence,
is
the
captain
service,
the
project
and
the
stage
which
is
defined
in
the
the
shipyard
file,
except
for
the
service
and
then
yeah
the
type
as
well,
which
you
know
technically,
would
be
hardcoded?
A
No,
actually
it
can't,
because
that's
the
that's
the
event
that
we
need
to
push
to
so
one
example
would
be
dev
trigger
div
dot,
something
dot
trigger,
and
then,
if
you
want
to
get
started
as
well,
you
can
we
we're
just
using
the
helm
releaser
to
build
the
charts
every
time
we
change
it,
and
then
that
uses
leverages
github
pages
as
well.
A
Yeah-
and
you
can
see
that
you
can
specify
your
captain
url
there
as
well,
so
pretty
much
the
only
thing
that
you
need
to
do,
and
then
this
is
examples
in
the
documentation
of
alerts
that
you
can
do
so
we're
testing
with
potato
head,
and
you
can
see
here.
So
this
is
the
beauty
of
what
we've
done
so
for
easiest
acceptance.
A
This
is
the
type
I
was
talking
about
before
too
so
for
easy
as
acceptance,
we've
we'll
put
these
key
value
pairs
that
we
need
for
this
template
in
the
provided
labels
because
they
in
flux
they
use
api
machinery
to
get
the
the
communities
object,
which
would
be
a
provider,
and-
and
this
is
already
what
they
have.
A
We
can
just
call
you
know
and
go
provide
our
dot
labels,
and
it
will
give
us
all
of
these,
and
then
we
can
use
a
simple
yeah.
D
B
A
Okay,
is
she
I,
I
thought
it
was
just
the
mentor
talking
today.
Sorry
yeah!
Oh
sorry,
okay,
mayhew,
do
you
want
to
talk
more.
C
F
Yes,
like,
as
brad
explained
like
we
are
yet
to
create
the
cloud
event
to
trigger
the
captain
sequences.
So,
yes,
I
am
do
that
thing
in
my
computer
right
now.
B
A
F
No
like
what
I
have
learned
so
far,
and
this
midterm,
like
worked
with
job
executor
service
and
yes,
I
have
implemented
that
slack
web
hook.
Integration.
F
A
C
A
And
then
yeah,
it's
been
cool,
seeing
maya
answer,
questions
in
the
help
channel
as
well
from
things
that
she's
learned
and
it's
helped
other
people
that
are
picking
it
up.
So
it's
been
really.
C
B
A
Maybe
sorry
I
was
confused
like
I
thought
the
mentor,
so
maybe
we
can
do
like
seven
and
a
half
minutes
for
the
mentor
then
hand
it
to
the
mentee.
I
I
talk
too
much
sorry.
B
We'll
figure
it
out:
okay,
next
up,
k-6
shainam
and
adam
and
pepe
pepe's,
not
here
so.
C
Oh
and
I'll
I'll
I'll,
even
you
out
brad
I'm
gonna
hand
it
straight
over
to
journal
so
yeah.
The
the
k6
project
is
obviously
just
to
create
an
integration
or
the
ability
to
run
k6
from
from
captain.
C
B
C
G
Okay,
cool
yeah,
so
I'll
start
start
off
with
a
brief
introduction
regarding
what's
k6.
So
k6
is
a
open
source
tool
which
is
used
for
load
testing,
just
like
the
jmeter
integration
that
we
have
in
the
already
in
the
organization,
but
it's
more
cloud
native
we
can
say
but
yeah
so
the
flow
it
goes
like
this.
G
So
for
k6
k6
test,
we
have
a
k6
script,
which
will
be
a
javascript
code
which
will
this
is
just
doing
the
http
get
on
this
test.k6.io
and
it
will
get
all
the
matrix
like
how
much
was
the
latency
and
how
much
was
the?
How
many
requests
got
failed
and
all
that?
So
this
is
generally
yeah.
So
this
is
mainly.
These
are
the
kind
of
matrix
that
we
needed
and
we
are
trying
to
get
this
integration
done
for
the
of
k6
for
captain.
G
So
for
that,
for
the
step
for
the
first
initial
step,
it
was
to
see
and
look
how
k6
will
perform
and
does
job
executor
service
run
for
it.
So
we
did
the
exercise
of
running
it
via
a
job
executor
service,
directly
importing
a
k6,
docker
image
and
then
running
it
to
a
job
executive
service
passing
the
test
step.
So
this
was
the
result,
so
we
were
able
to
do
that,
and
job
executor
service
was
also
able
to
fail
the
k6
script,
so
k6
has
a
feature
called
k6
thresholds.
So
what
is
it
does?
G
Is
that,
based
on
the
load
testing,
we
can
add
thresholds
like
hey.
I
want
the
p95
percentile
of
the
request.
Duration.
The
latency
of
the
http
request
should
be
less
than
let's
say,
200
milliseconds.
So
if
the
p95
95
percentile
of
the
request
is
greater
than
200,
so
here
it's
284,
so
it
will
fail,
and
so
that
see
so,
the
sequence
in
captain
will
also
fail
using
job
executive
service
so
yeah.
This
was
done
and
we
also
documented
the
tutorial
for
it
and
after
this
the
next
step
was
mainly
how
to
use
k6
extension.
G
So
what
is
k6
extension
k6
extension
is
basically
way
of
yeah,
so
k6
extension
is
basically
a
way
of
using
k6
as
a
load
testing
tool
and
writing
the
data
that
k6
provides
the
matrix
that
k6
provides
to
a
remote
host
like
prometheus
or
influx
gb
or
anything
so
k6
has
this
extensions,
what
we
so
this
extensions
has
have
the
ability
to
run
the
k6
and
also
output,
its
matrix
to
the
prometheus
or
whatever
output
source.
We
want
so
data
source
you
want
so
over
here
for
this
instance.
G
This
is
an
this
is
the
repository
for
xk6
output
prometheus
remote.
So
what
we
have
to
do
is
that
we
have
to
use
this
xk6
command
to
build
its
binary
and
then
use
that
binary,
providing
the
providing
the
sufficient
environment
variables
for
this.
So
for
the
for
this,
we
also
tried
to
see
if
we
can,
you
know,
make
a
docker
image
for
k6,
make
a
docker
image
of
this,
the
binary
k6
and
after
that
run
it
using
job
executor
service.
So
currently
I
was
working
on
that.
G
So,
for
that
it
was,
I
was
able
to
run
the
k6,
but
I
was
not
exactly
able
to
write
the
matrix
to
the
prometheus
output,
the
matrix
matrix
to
the
prometheus,
and
I
was
facing
some
errors
over
here.
It's
mainly
related
to
that
endpoint
that
I
need
to
give
over
here
to
write
down
the
prometheus.
So
for
that,
I'm
fixing
that
so
for
that
the
prometheus
needs
to
be
promises
so
k6,
our
captain
is
running
in
cuban.
G
It
is
cluster,
so
it
was
not
able
to
interact
with
localhost
prometheus,
so
I
had
deployed
it
in
ubuntu's
cluster
and
also
enabled
the
remote
right
and
all
that,
but
still
I'm
facing
some
issues
over
there.
So
yeah,
we'll
work
upon
that
and
another
way
of
implementation
that
we
came
across
is
that
we
can
also
write
a
new
service
for
k6
or
for
k6
integration.
So,
for
that,
we
we
are
still
discussing
whether
we
want
to
do
this
or
the
we
want
to
go
with
the
job
executor
service
way.
G
So
for
this
I
just
had
written
a
service,
so
a
new
captain
service.
So
if
I
do
a
trigger
over
here,
so
what
will
happen
is
that
my
yeah
okay,
sorry
yeah,
so
my
server
service
will
receive
the
triggered
event
and
it
will
sleep
for
five
seconds.
Send
the
started
event.
The
test
dot
started
event
and
after
that
it
will
sleep
for
a
five
second,
just
a
random
sleep,
and
after
that
it
will
send
the
finished
event
to
the
captain.
G
So
after
this
I
mean
in
five
or
some
seconds,
it
will
get
into
a
successful
state
over
here
and
so
right
now.
The
code
for
this
has
been
that
it's
written
so
the
whole
flow
of
getting
the
test,
the
triggered
event
from
the
captain
and
sending
started
event
and
finished
event.
So
the
main
we
can
say
the
main
fruit
of
running
k6
test
is
remaining
so
that
we
can
do,
but
first
we
also
have
to
evaluate
whether
we
need
this
or
not.
So
I
can
also
go
through
the
code
over
here.
G
Do
we
have
time
or
no,
I
I'll
go
in
two
three
minutes
quickly,
yeah.
So
basically
this
is
the
so
the
using
helm
we,
I
have
created
a
hand
chat.
So
we
can
directly
install
this
service
in
the
captain
name,
space
and
directly
basically
find
it
over
here
in
the
integrations
of
captain
and
we
can
add,
active
subscriptions
corresponding
to
it
and
yeah
base.
G
G
Yeah,
okay:
there
we
go
yeah
yeah
and
now
you
should
be
able
to
see
yeah.
So
I
just
showed
this
flow
of
triggering
the
event
and
in
my
k6
service
and
my
k6
service
will
subscribe
to
this
test.triggered
event
and
what
it
does
it's
that
it
simply
listens
the
event
and
sends
a
test.
Dot
started
event
back
to
the
captain
and
after
that
just
adds
a
normal
sleep
of
five
seconds
and
sends
the
finished
event
back
to
the
captain.
So
this
is
the
flow.
So
right
now
the
main
thing.
G
So
we
just
need
to
add
the
k6
test
execution
after
in
this
step,
so
this
part
is
remaining
and
over
here
this
is
the
main
function.
So
what
main
function
does
it?
It
simply
starts
the
so
we
have
the
helm
charts
for
this
and
what
helm
charts
does
is
that
it
deploys
the
k6
service
through
the
kubernetes
cluster.
So
captain
is
able
to
see
that
integration
and
it
can
subscribe
and
all
that.
So
this
is
the
main
helm
chat.
G
So
this
flow
can
be
used
for
the
dive
work
and
over
here
it
will
just
receive
the
events
and
after
that,
if
it
it
sees
the
so
what
it
does,
it's
that
it
checks
the
type
of
the
event
and
if
it's
a
test
dot
triggered
event,
then
it
will
go
to
this
event
handler.
So
we
have
this
handle
test,
dot
triggered
event.
It
will
take
the
captain
instance
and
incoming
event
and
data
corresponding
to
it
and
it
will
see
if
it's
functional
or
not.
The
test
strategy
is
functional
or
not
so
test
date.
G
There
is
a
format
of
data
which
we
need
to
send
to
the
captain,
so
it
checks
that
and
see
if
it's
functional
and
send
the
started
event.
Sleeps
for
five
seconds
sends
finished
event
for
now
over
here
there
will
be
the
real
implementation
and
we
also
have
a
test
event
handler
test.
So
what
it
does
is
that
it
simply
mocks
the
test.triggered
event
and
sends
this
payload
to
the
sends
this
payload
to
the
yeah
sends
this
payload
and
checks
whether
this
handle
test
trigger
event
is
working,
fine
or
not
so
yeah.
E
D
We
have
a
dock
engine
which-
and
we
have
a
lot
of
problem
like
versioning
multi-dock
support.
We
don't
have
a
search
seo
so
that
type
of
problem
we
are
solving
and
we
had
a
discussion
on
which,
like
like
frameworks
or
something
tool
we
should
use.
So
we
had
a
comparison
between
antona
and
darkosaurus
and
we
decided
to
move
with
dockasaurus
so
like
it
was
like
a
react-based
like
just
an
optimized
way
to
make
a
dock
site.
D
It
is
open
source
tool
by
facebook
and
we
have
a
lot
of
features
which
are
like,
like
things
supported
by
us,
so
yeah.
So
my
topic.
D
So
I'm
going
to
showcase
the
how
the
dark
side
look
like.
I
have
dark
side,
and
so
so
and
let's
see,
let's
search
something
so
now
search
we
had
in
the
docuseries.
So
let's
search
something
and
the
search
or
whatever
okay.
So
it
won't
work
in.
D
Okay,
so
so
whatever
we
are
going
to
search
it,
it
is
coming
from
the
latest
version.
So
we
have
a
latest
version
of
0.17.1x.
D
So,
let's
something
cli
so
it's
this
gives
us
cli
things.
D
So
like
like
earlier
in
hugo,
we
are
like
manually
adding
it
to
add,
so
we
have.
C
Like
versions
like
zero
and
fifteen,
sixteen
and.
D
D
C
D
So
it
is
going
to
clone
the
latest
version
to
a
new
enable.
Then
we
can
inside.
We
can
go
it
and
change
the
dock
like
something
we
want
to
change
like
what
many
versions
should
contain,
and
all
this
is
all
about.
Versioning,
we
okay,
so
we
have,
we
also
improved
the
seo.
So
when
we
go
to
this
road
map,
this
website.
D
So
somewhere
like
what
was
somewhere
like-
let's
say
this
check
this
in
meta
text:
the
meta
meta
image,
so
we
it
should
generate
our
image.
There's
no
description
of
the
thing.
Yes,
so
we
have.
D
Go
meta
tags
it's
going
to
generate
the
description,
change
it
and
how
we
are
doing
it
right
like
we
just
have
to
just
description
and
all
in
front
matter
of
front
matter
of
any
document.
Even
we
have
our
tax
and
keywords.
So
I
just
given
one
keyword.
So
when
you
go
to
the
end
of
document,
this
is
like
older
docs.
We
don't
have
any
keyword
or
something
tags
and
all
when
you
go
to
here
text,
so
roadmap
kptn.
So
we
can
tag
it
like.
D
If
your
roadmap
is
there,
so
we
can
tag,
we
can
get
all
documents
related
to
tag
a
roadmap,
so
we
can
define
here
all
the
tags
here
in
the
front
matter.
Even
we
can
define
image.
So
if
you
want
something
like
every
every
page
should
have
some
different
image,
so
we
can
define
that
also.
So
this
is
all
about
improving
the
seo
of
page.
So
the
main
feature,
the
multi-doc
support,
so.
D
Let
me
see
so
okay,
so
here
we
have
so
this
is
like
docs
and
we
here
we
have
two
mode
links
kpt
and
go
utils,
and
so
this
is
a
depot
in
the
capital.
I
just
give
an
example
here,
I'm
not
sure
about
what
a
repo
we
are
going
to
put
it
here
and
we
have
one
more
repo
here
like
this
one.
So
we
are
fetching
so
docuseries
provide
a
multi
docs
support.
D
So
we
have
we
have
a
docs
folder
here
and
we
can
have
a
different
need
to
add
content
here.
So
I
just
know
we
are
yeah,
so
we
wanted.
We
wanted
some
script.
So
so
that's
that's
what
one.
D
D
Any
content
here
we're
just
fetching
the
redmi
file
of
that
repository
see.
Okay,
I
think
that
repository
read
me
file,
so
we're
using
zero,
but
we
want
some
action
for
it.
E
Okay,
meanwhile,
I
can
share
a
few
things
around
the
project
as
well.
So,
like
rajiv
said
it's,
it's
mostly
like
the
version
2
of
our
documentation
site
that
we
have
right
now
and
it
attempts
that
fixing
a
lot
of
problems
in
the
current
documentation
site.
Like
rajiv
explained,
we
haven't
transferred
all
the
content
from
the
whole
documentation
website
to
the
new
one.
E
We
just
have
enough
to
demo
the
capabilities
of
the
new
documentation
engine
and
there
are
a
lot
of
other
features
which
are
going
to
come
in
phase
two
as
well.
The
demo
that
rajiv
is
showing
right
now
is
the
first
draft
we
have
and
we
would
really
appreciate
any
kind
of
feedback
of
for
the
actual
demo
next
week
and
if
you
have
any
questions
around
the
engine,
if
you
have
any
comments,
any
feedback,
I
would
really
appreciate
it.
B
C
H
Awesome
yeah,
so
let
me
see
here.
Let
me
show
my
screen
as
well.
H
Sorry
excuse
my
voice.
It's
a
little
bit
early,
where
I
am
so.
I
might
sound
a
little
bit
raspy
I
apologize
but
yeah.
I
guess
I
guess
for
me,
like
I'm.
H
I'm
really
excited
to
just
see
everyone's
progress
like
the
other
g-stock
contributors
and
how
far
they've
come
so
really
cool
I'll,
say
the
the
biggest
update
that
I
had
is
that
in
fact
my
timeline
is
a
little
bit
different
than
the
others,
just
because
I
was
a
little
bit
out
of
commission
for
for
the
past
few
weeks,
and
so
I
had
to
extend
my
timeline
into
october.
H
My
progress
as
well
as
explain
my
project
a
bit
and
then
I
think
next
week,
I'll
have
something
more
concrete
to
demo,
but
yeah
with
that
being
said,
yeah,
so
so
my
project
is
really
building
on
top
of
of
the
work
that
thomas
has
done,
specifically
on
top
of
the
the
get
ops
operator
prototype
and
that
is
to
enable
sort
of
a
similar
workflow
for
a
lot
of
the
other
entities
that
haven't
been
implemented
yet
so,
for
example,
sli
configs,
slos,
remediation,
configs
and
and
weapon
configs
for
captain.
H
So
currently
I'm
working
on
on
the
sli
configs
and
that's
something
that
I
hope
to
demo
next
week
but
yeah.
So
so
I
guess
the
main
benefit
here
is
to
to
achieve
like
or
to
to
bring
kept
in
closer
to
a
git
ops
workflow
as
well.
H
I
guess,
in
addition
to
that,
being
able
to
manipulate
the
entities
via
like
keep
cuddle
and
and
the
kubernetes
api
versus,
what's
currently
happening,
which
is
using
the
internal
git
repository,
which
we
want
to
try
to
limit
so
yeah
for
our
project,
I'm
heavily
heavily
leveraging
operator
sdk,
which
is
obviously
also
built
on
top
of
a
keyboard
there.
H
Yeah,
so
if
you,
I
guess
like
like,
I
said,
I'm
I'm
working
on
top
of
this
repo,
which
is
the
get
offs
operation.
H
So
if
you
take
a
look,
it's
it's
really
nicely
laid
out
thanks
to
the
operator
sdk
framework,
where
a
lot
of
so
like
the
the
entities,
you'll
you'll,
be
able
to
recognize
right
away
through
through
the
api
which
you
can
generate
via
the
cli.
So,
for
example,
like
kept
an
instance,
this
is
the
resources.
This
will
generate
the
resource
itself,
as
well
as
the
corresponding
controller.
H
H
For
for
these
resources,
and
that's
that
that's
sort
of
yeah
that
that's
that'll
be
that'll,
take
me
to
I
guess,
like
october
and
yeah.
So
this
is
obviously
my
project
page
and
then,
if
I
go
back
to
you.
H
So
again,
in
terms
of
implementation,
I'm
just
not
as
far
along
as
as
the
others,
just
because
I'm
a
again
like
a
few
weeks
like
shifted
in
terms
of
the
timeline,
but
let
me
see
yeah
this
right
here
is.
H
H
I
guess
currently
implemented
and
and
monitored
in
captain
currently
so
yeah.
I
think
this
is
already
like
most
of
the
pseudo
code
for
the
resource
itself,
and
then
the
next
piece,
obviously
is
the
corresponding
controller,
which,
hopefully,
by
next
week,
I'll
be
able
to
demo
yeah.
That's
about
everything
from
me
again,
not
as
much
not
as
much
progress
but
but
yeah.
If
anyone
has
any
questions
or
feedback,
I
think
we're
mostly
on
the
get
ops
operator
working
group,
page
whoops.
E
Yeah
and
I'm
not
sure
if
this
was
mentioned,
but
eric
just
switched
his
job
and
switched
to
a
different
due
to
a
different
country.
So
it's
been
really
hectic
and
we
had
him
timeline
by
a
couple
of
weeks,
but
otherwise,
I
think
he's
been
really
communicative
about
it
and
really
on
the
really
owning
what
he
is
doing
for
those
of
you
are
not
aware
of
the
cap
and
get
off
software.
It
essentially
allows
you
to
run
captain
without
the
cli
command.
E
You
just
get
custom
resources
to
set
up
your
captain
set
up
your
entire
captain,
and
what
eric
is
doing
here
is
to
add
support
for
some
extra
custom
resources
like
sli
as
a
lot
of
type
of
conflicts,
and
all
of
that
which
would
really
enrich
the
already
existing
ecosystem
that
we
have
for
captain
setting
up
captain
using
custom
resources
and
controllers.
E
Yep
really
interesting
stuff
to
come
in
the
future.
That's
all.
B
C
D
So
has
a
multi-talk
support.
D
Putting
it
in
document
like
you
can
go
if
you
have
like
all
the
docs
will
be
the
different
different
report.
Docs
will
be
here
like
I
have
a
go:
captain
go
utils.
So
when
we
go
over
here
captain
going
to
repository,
we
have
a
readme
file,
so
I'm
just
rendering
it.
D
D
So
we
wanted
some
like
like
github
action,
which
can
like
say
save
a
lot
of
time
like
we
don't
have
like
we
have
so
many
nested
contents
and
all-
and
it's
very
probably
it's
working
like
we
have
github
gs,
the
devops
collection
documents
and
change
log
for
that
it
is
also
rendering
0mb
so
yeah,
so
that
I'm
working
on
also
like
onto
myself
like
we
have
to
move
a
lot
of
documents
like
we
have
current
hugo
dockside.
D
We
have
a
lot
of
documents
from
older
to
new
and
hugo.
Have
some
templates.
We
can't
run
into
servers
and
we
have
to
see
that
whole
documents
where,
where
the
templates
are
there
and
convert
to
js,
so
that
what
that
also
have
to
tackle
that.
So
I
think
so.
Okay,
so
I
think
we
need
to.
I
don't
think
if
I
need
to
rescue
the
work,
because
I
think
only
the
github
actions
left.
C
E
A
B
H
H
So
I
moved
from
vancouver
canada
to
currently
I'm
in
boston
in
the
u.s,
so
yeah
it
was
yeah.
There
was
a
bit
of
like
visa
issues
and
some
unexpected
things
that
that
kind
of
put
me
out
of
commission
for
a
few
weeks,
but
I
definitely
appreciate
the
flexibility
of
the
mentors
but
yeah.
H
I
think
like
in
terms
of
I
I
would
say,
like
one
thing
that
I
learned
or
that
I
know
I
need
to
be
better
at
is
like
communication
and
like
just
and
being
quick
with
that.
Just
because,
like
the
time
zone,
difference
might
add
like
half
a
day
or
a
day
to
when
people
can
respond,
and
so
that
just
adds
a
little
bit
of
lag.
That
that
I
need
to
be
more
aware
of.
I
think.
A
Wow
yeah,
it's
a
it's
a
common
problem.
I
I
personally
always
never
leave
the
ball
in
my
court.
So
if
I
ever
need
something
from
someone,
I
just
slack
them
straight
away
and
then
it
sort
of
sits
with
them
until
they
can
do
it.
But
time
zones
are
a
common
problem,
especially
I
guess,
if
you're,
not
in
certain
parts
of
us
or
europe,
you
know
you
pick
like.
No
one
really
cares
about
so
yeah.
I
feel.
C
H
Oh
sorry,
I
was
gonna
ask
if
there
was
like,
because
I
think,
like
what
I'm
trying
to
do
is
like,
like
I've
set
up
like
like
an
asynchronous
kind
of
stand-up
type
thing
as
as
a
way
to
at
least
keep
others
updated.
But
I
was
wondering
if
you
guys
have
any
like
tips
for
for
dealing
with
for
stuff,
like
that,
I
guess.
B
No,
I
think
that's
we're
all
having
to
learn
to
get
better
at
asynchronous
because
boy,
it
is
a
24-hour
day.
A
Also
eric
I'm
in
columbia
at
the
moment,
so
I'm
sort
of
closer
to
your
time
zone.
If
you
ever
wanna
move.
A
Yeah,
not
british
columbia,
the
spanish
columbia.
E
I
I
think
that
the
problem
is
in
case
of
eric's
project
is
thomas,
is
in
utopian
time
zone.
I
am
an
indian
time
zone
and
axel
is
also
in
I'm,
not
sure
where
it's
from,
but
it's
I
might
be
wrong
is.
I
think
it
is
also
in
a
time
zone
closer
to
indian
time
zone,
so
it
gets
really
difficult
for
eric,
and
if
he
wants
to
ask
something
one
of
the
things
that
we
could
try
is
obviously,
I
think,
with
someone
who's
willing
to
help
from
years
time
zone.
E
That
would
be
great
other
would
be
getting
good
at
asynchronous
communication.
But
what
do
we
really
mean
by
that?
One
thing
could
be
working
out
loud.
That
is,
if
you're
working
on
something
just
start,
a
thread
in
the
in
the
public
channel
and
type
as
you
do
things
so
that
others
could
catch
up,
but
that
also.
E
E
That
could
be
one
and
other
something
that
I'm
trying
to
do
now
is
just
see
if
we
can
find
a
time
slot
for
you,
so
that
you
could
just
reach
out
and
slack,
and
we
could
get
on
a
column
here,
program
and
stuff.
So.
E
B
I
just
we're
all
learning
how
to
do
stuff.
It's
asynchronous.
You
know.
Another
thing
that
sometimes
works
for
is
like
you
file
an
issue,
and
you
can
have
the
discussion
in
the
issue
and
then
it
can
be
all
asynchronous,
but
you've
got
it
all
together
in
one
place
where
you
can
keep.
It
depends
on
what
you're
trying
to
address
you
know,
but
but
that
can
be
useful
and
we're
all
if
anybody
figures
out
how
to
handle
these
time
zones.
Please
let
us
know.
A
Maybe
we
can
get
dramatrix
to
pay
for
our
flights
and
we
can
meet
in
vegas
or
somewhere.