►
From YouTube: Keptn Developer Meeting - June 02, 2021
Description
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
A
Experienced
a
bug
in
the
shipyard
controller
which
was
binded
to
dispatching
the
new
sequences
directly,
even
if
there
were
other
sequences
in
the
trigger
state
waiting
for
execution.
So
the
sequences
were
not
executed
in
the
order
they
were
triggered.
A
But
if
some
kind
of
inconsistent
state
happens
and
this
inconsistent
state
was
a
small
period
of
time
where
no
sequence
was
started,
so
no
sequence
was
in
running
states
and
other
sequences
were
not
able
to
yet
start
to
wait
and
if,
in
this
time
slot
a
new
sequence
appeared
they,
it
was
kind
of
executed
immediately
because
nothing
was
blocking
it.
A
So
we
made
the
changes
that
every
sequence
is
not
blocked
only
by
running
sequences,
but
also
blocked
by
sequences
which
were
executed
before
this
sequence
and
are
in
the
triggered
state.
So
they
have
lower
timestamp
so
yeah.
I
won't
go
through
all
the
changes
but
yeah.
If
there
are
any
questions,
I
will
just
ask.
The
next
thing
was
introducing
a
webhook
config
version
check.
We
actually
forgot
to
introduce
webhook
configuration
file
version
check
today.
We
support
actually
two
versions,
the
alpha
version
and
the
beta
version.
A
So
we
were
not
checking
this
before
so
we
had
some
kind
of
problems,
nobody
spotted
this
fix,
but
it
was
rather
simple
and
did
not
make
any
damage
and
the
third
bug
fix
was
regarding
the
cli,
where
we
have
some
duplicated
sequence:
data
when
x,
when
triggering
a
sequence.
So
this
was
actually
we
just
removed
the
sequence
flag
and
there
is
no
need
to
duplicates
the
sequence
name
in
the
cli
yeah
and
the
second
part
of
the
sprint,
which
was
the
main
goal,
was
about
zero
downtime
testing.
A
This
was
mostly
about
executing
the
test
looking
at
logs
logs,
investigating,
where
the
problems
are
adapting
to
tests
also
some
time
outs,
for
example
the
bristol
hooks
or
some
grease
for
periods
and
yeah.
It's
still
a
little
bit
in
progress.
We
had
some
nitpicks
there,
but
we
need
to
finish
before
the
release,
but
with
the
release,
probably,
we
will
be
able
to
and
say
that
we
are
zero
downtime.
B
My
turn
yeah,
I
will
share
my
screen
just
a
second
cool,
so
my
sprint
as
well,
was
focused
on
zero
downtime
test.
I
also
had
the
future
bug
fixes.
So
I
start
briefly
from
those
first
nice
thing
was
we
finally
have
a
cp
common
folder
in
which,
for
now
we
have
added
a
few
things
among
those
we
do
have
now
the
interface
to
the
api
which
is
being
stored.
There
also
a
few
changes
like
adding
docker
files
for
enabling
testing
in
our
pipelines.
B
We're
done
here,
but
yeah,
not
much
few
things
will
be
in
the
future
migrated
in
this
folder
in
our
core
captain
core.
Then
there
was
a
bug
among
our
web
book.
The
beta
version
is
now
having
a
denied
list
which
is
built,
and
what
is
happening
right
now
is
that
basically,
some
of
the
internal
ips
were
not
filtered,
so
the
solution
for
the
bug
was
to
simply
adding
also
host
resolution
in
the
denial.
B
So
if
we
have
the
address
that
resolves
to
the
kubernetes
local
cluster,
any
address
that
resolved
to
that
cluster
then
will
be
denied.
B
These
other
issues
was
related
to
parallel
deliveries,
parallel
sequences.
If
we
had
two
sequences
on
the
same
project
same
stage
but
regarding
different
services,
the
wanted
behavior
would
be
that
we
can
execute
them
in
parallel,
since
there
is
no
conflict
among
them.
B
This
was
not
happening
because
we
had
a
filter
in
which
we
were
considering
the
staged
already
booked.
If
there
was
something
on
any
service,
so
it
was
as
simple
as
adding
a
filter
based
by
service
as
well,
so
that
now
we
could
do
all
of
them,
and
this
is
being
sold
here,
not
much
to
say
about
this
one.
B
B
To
select
things
such
as
whether
or
not
we
want
to
run
ui
test
together
in
parallel
with
the
zero
downtime
test
to
verify
that
also,
the
bridge
works
when
we
are
doing
upgrade
and
downgrade,
and
this
can
be
selected
by
this
tick
here,
then
what
else
has
been
added
the
possibility
to
select
in
the
pipeline
to
specify
a
full
url
for
a
certain
specific
commit
image
of
in
our
health
charts
and
also
the
ability
to
schedule
this
as
a
cron
job?
B
So,
basically,
for
now
we
have
that
every
monday
the
zero
downtime
pipeline
will
be
run
automatically.
B
For
now,
we
did
not
set
up
any
issues
auto
generated
because
we
are
still
perfecting
the
test,
so
we
don't
want
to
have
any
flukes
and
false
alarms.
B
A
few
refactoring
has
been
done
in
this
pr,
as
well
from
the
logic
of
how
we
run
at
the
same
time
the
go
test
and
the
cyprus
test,
but
not
much
really,
then.
Finally,
now
we
do
have
a
documentation
for
this
whole
pipeline
in
docs,
developer
and
zero
downtime
test,
so
it
is
explained
how
to
run
them
locally,
how
to
run
them
on
the
clusters
and
what
are
the
step
if
you
would
like
to
add
any
other
test
using
the
zero
downtime
test
suit?
B
B
B
B
C
So
yeah
not
actually
much
from
my
side
just
one
bug
fix.
For
example,
we
had
the
bug
in
the
evaluation
board
that
if
you
create
a
service
and
execute
an
evaluation
and
then
go
to
the
evaluation
board,
then
the
view
service
details
button
was
missing
here.
C
This
was
only
appearing
if
the
service
was
created
and
the
narration
was
triggered
on
the
same
page
without
refreshing
and
yeah.
This
is
fixed
by
now,
and
the
button
should
always
be
there,
and
you
also
took
the
opportunity
to
rewrite
some
of
the
the
whole
code.
It
was
here
in
the
aeration
board
and
yeah.
C
That's
basically
it
for
this
one
yeah
and
the
next
thing
comes
to
this
polling.
C
In
the
bridge
we
now
we
are
now
removing
some
polling
mechanisms
that
are
not
really
needed
and
now,
for
example,
for
this
one,
we
have
the
metadata
endpoint
in
the
bridge
back
end
that
updates
the
service
filter
information,
so
yeah
service
and
stage
information.
C
C
D
Thanks
yeah
from
my
site,
two
very
similar
requests
like
the
last
one
from
klaus:
it's
about
removing
polling
first
for
the
evolution
history
in
the
environment
screen,
so
in
the
environment
screen.
When
you
clicked
on
a
stage,
we
yeah
had
a
polling
mechanism
here
to
load
this
evolution,
results
which
is
not
needed
here
in
this
place
and
yeah
to
increase
performance
for
the
bridge
and
also
reduce
the
load
on
the
captain
core.
We
remove
the
pulling
here.
D
It
will
reload
when
you
click
on
a
stage
so
every
time
the
user
clicks.
Of
course
it
will
and
the
second
place
where
we
remove.
This
is
from
the
settings
screen.
So
in
the
settings
screen
services
list
yeah.
We
also
had
a
polling
mechanism
here,
since
the
user
visits
the
settings
screen
just
to
edit
or
create
a
service
yeah,
we
load
it
only
once
when
you
visit
this
page.
E
And
know
that
there
is
always
appalling
for
for
the
project
state
so
to
say,
but
there
is
also
one
hidden
behind
the
trigger
new
sequence
button.
Can
you
just
click
on
it
and
or
I'm
just
asking?
Is
there
a
polling
mechanism
that
checks
constantly
whether
there
is
a
new
custom
sequence
or
is
it
loaded
just
once.
D
I'm
not
sure
if
we
document
that,
but
if
there
was
a
pulling,
I
guess
it
should
be
in
our
list
where
we
discussed
the
polling
mechanisms,
but
I
think
here
it
will
reload
only
once
when
you
open
the
form.
I.
G
G
We
can't
access
your
screen,
come
on
and
restart
my
browser,
no
no,
oh
dear,
but
I
wanted
to
look
at
one
of
we
are
going
to
be
doing
some
major
restructuring
of
the
documentation.
One
part
is
to
create
reference
files
or
a
reference
page
for
all
of
the
configuration
files
that
we
have.
These
are
sort
of
for
those
of
you
who
are
used
to
old
man
pages,
that's
sort
of
what
I
have
in
the
back
of
my
mind.
G
They
are
not
for
the
novice
who's,
trying
to
figure
out
how
to
do
stuff
for
the
first
time
they
are.
You
know
they
are
a
summary
of.
They
should
be
a
summary
thing
that
one
can
do
and
oh
dear,
and
I
can't
show
you
what
I've
got
so
it's
hard
to
ask
how
we
do
this.
Anybody
have
any
bright
ideas.
H
G
G
Would
you
not
mind
if
you
I
was
going
to
show
you
my
local,
build
which
I
think
might
be
prettier
than
what's
on
here?
If.
D
G
To
that
pr,
there
are
three
files
here
and
the
third
one
is
the
one
we're
interested
in
the
other,
lateral
damage
that
I
won't
trouble
you
with
here.
G
There
we
go
and
then,
if
you
could
go
to
files
or
and
go
down
to
the
reference
file,
shipyard.
G
Okay,
so
brief,
the
shape
we've
got
here,
which
I
think
is
close
to
what
we
we
begin
with
a
very
brief
description,
kurt
about
what
this
file
is,
then
this
and
then,
if
you
scroll
down
section
where
we
describe
sort
of
every
bloody
time,
that's
in
here,
I
am
right
now,
but
that's
one
of
my
questions
below
that.
There's
a
usage
section
where
which
form
for
things
people
should
know
of
this.
G
It's
nice
to
have
examples
on
reference
pages,
but
shipyard
is
so
immense.
I'm
thinking
about
I'll
give
them
an
example,
but
we'll
give
them
links
to
other
files
where
we
will
discuss
different
parts
of
it
and
have
examples
below
that.
There
is
a
files
section
playing
a
little
bit
need
to
tell
them
where
the
file
is
located
where
they
created.
I
suppose
what
it's,
where
it's
going
to
be
in
the
upstream
github,
perhaps
a
link
to
the
github
page
github
section
that
does
that
differences
between
versions
is
an
old
favorite
of
mine.
G
It
is
a
place
to
summarize
right
now.
I
did
give
a
link
to
the
one
thing
I
have
to,
but
I'm
not
going
to
go.
Try
to
go
back
and
backfill,
although
if
anybody
wants
to
contribute
it,
but
but
as
we
go
forward
when
there
are
changes
in
the
file,
we
can
say
that
you
know
in
that
this
field
is
only
supported
in
17
and
later
or
this
field
stops
being
supported
in
16
or
whatever
it's
a
nice
summary
for
us
over
the
you
know.
G
We
when
we
first
make
the
release
it's
in
the
release,
notes,
but
a
couple
of
years
later
it
can
be
really
hard
to
figure
it
out
or
something
to
know
that
and
then
the
standard
c.
Also
so
that's
the
general
now.
My
first
question
is:
when
you
come
into
the
reference
section
for
files.
Do
you
want
to
see
this
called
shipyard
or
shipyard.yaml,
and
when
we
begin
and
say
the
shipyard.yaml
file
defines
your
project?
Do
you
want
it
to
say
shipyard.yaml
or
shipyard?
How
do
we
a
lot
of
times?
G
G
E
H
E
F
F
E
F
J
In
the
end,
just
from
myself
from
the
outside,
I
would
also
not
reference
the
shipyard
file.
I
would
say
the
shipyard
configuration
when
you
really
want
to
reference
in
the
documentation,
because,
where
the
configuration
is
stored
can
change
over
time
and
how
it
is
stored,
it
does
not
have
to
be
a
file.
Theoretically,
and
you
say,
look
at
the
shipyard
configuration
it's
more
referencing
to
any
kind
of
way
of
storing
the
configuration.
E
Yeah,
because
when
we
think
about
the
other
two
files
we
discussed
before
the
slo,
as
well
as
the
relation,
they
are
already
said
that
those
are
configs.
G
G
Right,
right
and
but-
and
we
can,
I
there's
gonna-
be
a
lot
of
iteration
getting
this
right.
We
can
all.
We
can
try
something
and
look
at
it
when
we
get
enough
and
we
can
always
change
what
we're
doing
too
so.
E
Are
yama
the
proposal
from
I
think
it
was
simon?
It
was
a
very
good
one
to
to
talk
about
the
shipyard
configuration,
not
the
file
as
an
artifact
more
like
the
configuration
of
of
the
security
configuration
of
an
slo,
the
configuration
of
a
remediation.
G
G
G
That
sounds
like
something
worth
trying
now:
the
header
that
I've
got
on
the
page.
That's
fields
conceal.
Can
you
scroll
back
up
to
that
one?
It's
just
under
synopsis.
G
And
that
they're
under
synopsis
there,
where
we're
gonna,
describe
all
of
these.
I
fields
parameters.
E
E
A
good
idea
to
take
a
look
at
how
kubernetes
is
calling
such
fields,
because
at
the
end,
they
somehow
look
like
a
custom
resource
definition
in
kubernetes,
and
we
should
learn
from
from
what
is
the
best
practice
or
from
what
from
how
kubernetes
is
doing
that.
Okay,
I
try
to
look
it
up.
G
Okay,
now,
within
the
description
where
you
really
can
see
it
go
down
to
the
tasks-
oh
I'm
so
sorry,
I'm
not
showing
it
the
the
old
documentation
that
I
stole
as
a
basis
for
this
had
a
little
synopsis
block
for
each
task,
and
I
started
to
propose
to
preserve
those
and
even
have
the
mark
synopsis.
G
And
then
I
was
thinking
having
that
synopsis
in
the
middle
of
the
description
is
annoying
and
can
we
scroll
down
like
just
a
little
bit
more?
The
first
couple.
Action
and
approve
do
have
a
synopsis
within
the
description
and
it
seems
like
a
good
idea
at
first,
but
as
I
look
at
it,
I
think
I
want
all
the
synopsis
to
be
up
in
the
synopsis
section,
and
here
we
just
discuss
it
and
if
they
want
they
can
roll
back
and
forth.
G
No,
no,
it
looks
good.
Okay
and,
let's
see,
I
think
those
are
actually
my
big
issues.
There
will
be
a
lot
of
anybody
who
wants
to
review
their.
You
know
there's
going
to
be
a
lot
of
reviews,
I'm
I'm
hoping
to
fairly
quickly.
G
I
was
hoping
for
16,
but
we're
not
going
to
make
16,
but
maybe
by
release.
17
have
these
good
enough
that
we
can
release
them
the
exist.
It
isn't
that
we're
replacing
perfect
documentation
and
then
they're
going
to
you
know
have
to
evolve
over
time
a
lot.
G
That's
how
it
always
goes
with
these
things
so,
but
does
the
shape
of
it
does
it
look
reason
I
mean
is:
is
the
writer
on
some
good
drugs
and
thinks
this
is
a
nutsy
thing
to
do?
Or
does
this
look
like
a
useful
thing
for
us
to
have.
C
E
F
G
Okay,
because
also
I've
noticed
a
lot
of
there's
some
that
are
not
even
in
but
they're
in
readme
files,
and
I
want
to
also
make
sure
that
the
developers
aren't
going
to
object.
You
know
it's
good
to
have
some
of
the
source
close
to
the
source
code,
but
it's
okay.
If
we
pull
all
that
stuff
that
I
can
into
the
documentation,
you
won't
feel
like
I'm
usurping
the
readme
files.
F
Do
you
also
describe
somewhere
on
this
flight
level
in
this
document
here?
What
are
the
possible
formats
you
have
to
put
into,
for
example,
for
the
triggered
after
property?
How
does
the
user
know
that
this
must
be
of
these
in
that
format,
for
example,
10
seconds,
then,
as.
G
We
have
a
separate
file
that
will
be
over
in
the
guide
material
about
triggers,
which
at
first
I
got
sucked
in
that
triggered
on
was
all
about
triggers,
and
somebody
pointed
out
to
me
that
no,
you
can
also
trigger
through
the
ecli
or
al
api,
so
we
will
get
that,
but
it
will
be
sort
of
an
iterative
thing
this
will
describe
triggered
on
and
then
we'll
link
to
the
article
about
triggers,
which
we
can
beef
up
and
put
in
everything.
We
want
with
lots
of
examples
and
discussions
and
recommendations.
G
F
So
this
document
just
describes
what
what's
the
purpose?
What's?
What's
the
behavior
of
of
all
the
things
you
can
do
in
the
shipyard
thingy,
and
it
will
contain
that
then
links
to
more
detailed
information
which
speaks
about
okay?
What's
now
exactly
the
format
of
the
time
frame,
I
need
to
use
when,
when
I
want
to
specify
a
trigger
after
thing.
G
Right
right,
this
is
about
my
fourth
attempt.
It's
you
know
some
of
the
other
rev
files
or
the
configuration
that
I
started
were
fairly
easy,
but
this
is
about
my
fourth
one
on
shipyard.
At
one
point,
I
think
I
was
you
know
at
a
30-page
document
and
still
just
getting
started
and.
E
F
G
That's
what
we
got
is
we'll
keep.
This
will
be
minimal,
but
it
will
be
my.
My
thing
is
always
if,
if
I've,
if
I'm
familiar
with
how
to
do
things-
and
I
know
what
I'm
doing
but
then
there's
always
that
little
thing,
what
exactly
is
this
called?
Where
does
it
go,
etc?
And
I
can
look
at
this
quickly
and
get
that
information
and
that's
sort
of
what
the
purpose
is
and
also
a
statement
of
it
and
over
time
we've
got
some
flexibility,
something
that
we
think
really
needs
to
be
here.
F
Yes,
absolutely
I,
like
the,
I
think,
that's
just
the
right
flight
level
so
to
say
to
describe
the
overview
of
that
yeah.
G
It'll
get
more
useful
when
we
have
the
other
docs
I
mean
most.
This
information
is
out
there,
but
I
know
I
still
I
stocks
and
ready
stocks
and
then
okay.
I
know
I
saw
that
someplace
and
I
can't
find
it
the
fact
that
we
don't
have
a
search
engine
on
our
documentation.
Engine
is
a
a
problem
that
will
be
solved
fairly
soon,
so,
but
but
meanwhile
yeah
so
we'll
get
a
little
better
and
then
I'll
be
able
to
sort
of
work
down
the
levels.
G
E
Just
did
a
quick
research
and
was
yeah,
taking
a
look
at
how
kubernetes
is
explaining
the
custom
resource
definition
of
a
deployment
and
there
the
term
field
is
heavily
used
and
also
a
very
common
term.
G
Okay,
excellent
excellent
coolness,
okay
and.
J
J
I
want
to
talk
about
two
certain
things:
okay,
because
for
the
documentation
I
opened
two
pull
requests.
One
is
the
improvement
of
the
documentation
build
because
I
saw
that
there
are
a
lot
of
that
some
dead
links
and
things
like
that
and
I
fixed
quite
a
lot
of
them,
but
I'm
still
stuck
with
with
with
30
issues,
currently
sorry,
which
I
picked
a
few
more
last
night
and
which
actually,
those
are
12
issues,
because
a
lot
of
them
are
in
there
for
different
kind
of
versions.
J
I
only
fixed
14,
15
and
16
issues
so
far
and
not
older
link
versions
and
those
are
sometimes
things
pointing
outside,
like
to
code
owner
files
on
the
captain
community,
which
does
not
exist
anymore
and
things
like
that.
But
I
cannot
I'm
not
sure
where
I
find
this
information
or
this
context,
I'm
not
that
familiar
with
the
project.
J
So
there
would
be
some
guidance
from
the
team
needed
somehow
to
fix
those
last
30
issues,
because
then
we
can
merge
this
pull
request
and
there
will
be
a
build
within
each
pull
request,
checking
if
there
are
some
kind
of
dead
links
in
there
pointing
to
some
resources
which
do
not
exist
anymore,
to
just
give
a
better
overview
of
the
documentation.
J
Currently,
I
deactivated
some
checks
because
they
are
just
they're
just
too
many
faults
too
many
issues
regarding.
J
J
What
I'm
looking
for
is
what
we
are
querying
and
what
we
are
so
what
we
are
testing
and
what
we
are
not
testing
and
what
what
rules,
for
example,
are
ignored
within
the
documentations
for
checks.
Some
of
them
are
ignored
with
com
edit
comments
for
those.
Why
I
ignored
them
just
to
give
you
an
overview
why
they
are
ignored
and
the
build
will
fail
afterwards,
if
those
so
if
the
check
is
is
failing.
J
Additionally,
to
that,
we
also
have
now
a
make
file
in
there
to
make
it
easier
for
people
to
start
the
server
with
docker
images,
to
build
the
documentation
with
docker
images
and
also
to
manually
execute
this
test
locally
with
with
a
make
goal
so
that
for
people
it
will
be
easier
in
the
future
to
set
up
the
documentation,
build
and
work
with
the
documentation.
G
That
should
be
a
big
help
to
get
us
through
now,
I
think
we'll
be
moving
to
docusaurus.
I
think
this
fall
and
dr
saras,
I
think,
has
its
own
html
test
that
we'll
have
to
look
at
that
and
see
if
it's
as
good
as
this
or
whatever
but
yeah
I've
I
fixed
it.
But
but
my
you
know,
the
test
I
can
do
is
to
read
through
the
docs
and
click
on
every
link
and
see
if
it
takes
me
where
it
should
so,
which
is
inefficient.
E
Or
not
yeah,
because
I
know
quite
a
lot
of
things
that
are
around
tomorrow.
G
J
No,
that's
not
that's
not
in
there
and
checking,
but
I
fixed
within
my
police
wait.
Why
does
it?
J
I
fixed
a
lot
of
wait
a
second.
I
fixed
a
lot
of
references,
for
example,
where
we
had
other
where
we
had
hard-coded
links
in
there
and
I
fixed
already
a
lot
of
links
which
were
not
appropriate
and
also
apps.
I
fixed
a
lot
of
absolute
links
in
there.
Also
there's
just
a
problem
with
the
cli
commands.
We
cannot
because
they
are
general,
as
I
saw
I
just
realized,
they
are
generated
automatically
there.
J
We
always
have
this
hard
coded
public
link
in
there,
which
we
might
should
remove
from
the
documentation
part
before
merging
it
in
there.
So
that's
the,
but
that's
outer
generated
by
a
bot
I
see
so
maybe
somebody
can
just
configure
the
bot
to
go
over
the
files
and
remove
this
hard
coded
url
for
the
current
version,
somewhat.
K
G
J
And
the
second
part
is
this:
is
this
is
just
the
first
part
regarding
those
links
and
everything
like
that,
so
this
should
fix
a
lot
of
links.
Also,
for
example,
there
were
links
to
github
with
release
in
front
of
the
text,
but
the
attacks
are
just
with
the
version,
so
those
configurations
are
properly
working
now
and
everything
and
the
second
part.
J
This
is
now
a
draft,
and
it
is
based
on
this
branch
because
there's
also
the
tool
markdown
link
cli,
which
helps
you
to
to
generate
a
more
concise,
markdown
link
and
a
more
consistent
markdown
for
all
your
files,
I'm
not
sure
if
doc
casaros
will
also
use
markdown
or
not,
but
in
the
end
we
also
do
have
there.
A
lot
of
there
are
a
lot
of
issues.
I
just
need
to
switch
to
my
to
my
branch
to
because
the
actions
are
not
running
here.
J
So
when
you
take
a
look
at
the
markdown
link
file
with
the
checks
we
have,
there
are
a
lot
of,
for
example,
headings
which
are
not
incremented
the
correct
way
and
all
those
kind
of
things
and
markdown
lint
is
actually
quite
powerful,
because
you
also
have
a
fixed
property
where
you
can
fix
most
of
these
issues
manually
already,
so
this
is,
for
example,
there's
a.
There
are
things
where,
for
example,
in
the
contributing
md,
the
the
for
example.
There
were
things
bro
with
contributing
md.
J
I
looked
if
here
I
am
there.
For
example,
the
the
ordering
of
the
numbers
was
not
correct,
for
example,
and
things
like
that
which
just
would
which
popped
up
further
down
here.
So
then
the
ordering
was
not
correct.
There
was
one
as
a
I
changed
that
already
and
things
like
that,
so
it
will
provide
a
more
concise,
markdown
file
experience.
Let's
say
that
way
and
you
can
automatically
fix
it
and
also
there
is
on
the
ones
hand,
there
is
again
this
markdown
link
file
where
you
can
exclude
rules
or
configure
rules.
J
We
have
an
ignore
file
to
not
fix
old
documentation,
more
likely
and
as
usual
I
put.
Where
is
the
first
make
file
just
a
second
I
normally
also
put
in
here,
and
I
also
put
in
a
make
file
target
so
that
people
can
actually
use
marked
only
locally
and
also
fix,
markdown
links
locally,
so
within
with
an
automatic
approach.
So
this
this
should
just.
J
G
G
J
But
those
are
the
two
things,
so
I'm
not
sure
I
I
just
started
with
them,
because
because
I
saw
them,
I'm
not
sure
if
I
should
pursue
with
those
issues
or
not,
as
I
will
just
put
a
one
for
the
html
test.
I
think
it's
a
good
idea
to
finish
those
ten
issues
and
I
will
get
in
contact
with
johannes
regarding
regarding
fixing
those
topics.
Markdown
lin
cli
is
another
topic
where
the
question
is.
G
You
said
something
by
the
way
I
I've
got
my
whole
long
list
of
improvements.
I
want
to
make
to
the
documentation
and
structure,
but
if
you
have
a
list
of
priorities,
I
would
love
to
know
that,
because
you
may
have
some
things
that
you
know.
I
I
look
at
stuff
as
to
what
I
like
and
don't
like,
or
something,
and
you
may
know
where,
where
the
real
pain
points
are.
G
E
G
Okay,
cool
yeah:
I
have
my
master
map
that
I'm
working
on.
It's
not
complete,
but
I'll,
send
you
a
link
to
it
and
you
can
see
where
I
am
and
feel
free
to
do
anything
you
want
to
to
it's
cool,
okay,.
J
Okay,
but
regarding
markdown
lindsail,
I
will
just
leave
this
for
now,
because
I
still
think
it's
a
great
benefit
to
edit,
but
we
will
take
a
look
at
all
the
rules
if
they're
really
an
issue
or
not
an
issue
and
take
a
look,
if
we
want
to
fix
them
or
not,
I
think
it
will
just
get
together
with
mac.
Regarding
that
to
take
a
closer
look
at
the
current
issues
reported
and
the
other
one,
I
will
just,
as
I
said,
get
in
contact
with
johannes.
I
I
Now
you
can
enable
that
via
helm
flag.
So
there
is
this
bridge
dot
heat
map,
dot,
enable.
If
you
set
this
true
when
you
install
captain,
then
the
operation
port
for
the
heat
map
view
only
would
be
using
d3
instead
of
high
charts.
And
if
you
want
to
change
your
instead
deployment
of
the
bridge,
you
just
need
to
change
the.
I
Okay
and
then
there
was
also
in
the
chat
suggestion
from
suraj
to
also
provide
some
documentation
directly
in
the
cli,
like
ctl
does,
with
the
explain
command.
So
we
might
want
to
have
captain
explain,
shipyard.
K
It's
it's
basically
like
you
can
do
cube,
ctl,
explain
and
the
resource
name.
So,
for
example,
deployment.template.spec.
K
I
G
What
is
the
feeling,
as
I
just
I
haven't
gotten
deeply
into
the
cli
docs,
but
as
I
glanced
at
them?
Well,
I'm
an
old
unix
person,
so
they
look
so
much
like
man
pages.
I
thought.
Oh,
this
looks
in
good
shape.
G
G
K
That's
that's
in
the
definition,
so
you
have
a
yaml
which
defines
the
resource
in
kubernetes
and
it's
a
part
of
the
code.
Basically,
so
you
have
to
define
it
in
your
yaml
custom
resource
definition
file
that
this
is
what
this
field
means
and
then
cube.
Ctrl
command,
auto
generates
the
documentation
for
you
when
you
need
to
cube
ctl,
explain
and
deployment
or
something
something.
K
Not
really
sure
how
feasible
this
is,
but
it
just
popped
up
in
my
head
might
be
nice
to
have
something
like
that.
G
We
can
certainly
look
at
you
know
we
could
define,
look
like
and
then
see
if
we
see
a
way
to
do
it.
What
about
the
api
documentation?
G
I
I'm
I
was
looking
into
this
two
weeks
ago
and
found
a
way
to
show
the
swagger
ui
as
part
of
our
website,
and
we
might
want
to
versioning
this
regular
ui
in
a
way
that
you
can
check
the
api
documentation
for
version.
16
version
15
14,
but.
G
K
Sorry
do
we
do
we,
I
have,
I
don't
know
if
we
have
this
feature,
but
do
we
have
a
feature
in
cli
where
you
could
put
forward
to
the
this
api
in
the
kaplan
cluster
and
then
it
would
show
you
the
api
documentation
on
localhost,
because
this
is
really
good.
One
problem
I
see
with
the
captain.slash
api
documentation
is
you.
K
This
is
the
documentation
for
latest
version,
so
there
can
be
confusion
as
in
if
my
captain
cluster
is
on,
let's
say
1.12
and
the
latest
documentation
is
1.16,
then
I'll
be
looking
at
one
on
16's
documentation
when
I
should
be
looking
at
1.12
documentation.
G
Or
to
be
able
to
have
a
differences
between
versions,
which
I
kind
of
I
favor
something
like
that
just
because,
if
well
I
don't
know
if
we
have
it
with
captain,
I've
lived
in
worlds
where
one
site
might
have
different
release
different
releases
running,
and
they
don't
want
to
have
to
look
at
three
different
sets
of
almost
identical
documentation,
trying
to
do
a
grip
with
their
eyes.
To
find.
Aha,
that's
different.
G
That's
the
nice
thing
about
differences
between
versions.
We
can
at
least
highlight
where
things
change,
but
but
even
just
being
able
to
see
it
would
be
nice
giovanni
anything
you've
got
there.
We
might
want
to
throw
that
at
our
gsat
kid
as
a
requirement
he's
pretty
sharp,
too
or
as
a
wishlist
thing,
or
something
to
see
what
he
you
know
he's
taken
off
pretty
strongly
so
there,
and
it
seems
like
docusaurus,
has
all
sorts
of
powers,
so
it
might
be
able
if
it
had
input.
It's
you
know.