►
From YouTube: Contributing to Keptn Tutorials
Description
This knowledge transfer provides overview of the Keptn Tutorials site internals. We look into the repository structure, Codelabs, and tutorials content organization. Then we discuss how to build the site and run it locally. Last but not least, we discuss how we want to rework Keptn Documentation with Antora.
Tutorials website: https://tutorials.keptn.sh/
Meeting notes: https://docs.google.com/document/d/1M-8d7HnZgswGHrOzal_zl2Jaau1AWN07fyf2PgI2GcI/edit?usp=sharing
Learn more: https://keptn.sh
Contribute: https://keptn.sh/community/contributing/
A
Hello,
so
today
we
are
doing
ad
hoc
knowledge
transfer
session
about
captain
tutorials
and
how
to
develop
them.
I
will
share
my
screen
and
show
how
it
works
and
thanks
to
everyone
for
joining
so
try
to
do
more.
Such
knowledge
transfers
going
further
release
my
screen,
yes,.
B
A
Is
okay
yeah,
so
we
will
be
looking
at
tutorials.
Captain
sh,
first
of
all
feel
free
to
interrupt
me
at
any
moment.
So
it's
an
informal
session,
it's
mostly
for
coney.
I
also
started
a
google
doc,
so
if
you
want
to
make
some
notes
etc,
I'll
just
make
it
editable
for
everyone.
A
A
Also,
we
have
a
support
for
multiple
versions.
So,
for
example,
if
you
go
to
0.7,
you
can
see
that
there
is
a
lot
more
tutorials
which
have
been
written
at
some
time,
including
cargo,
cd,
aws,
gc
engine,
etc.
So
they
were
a
lot
of
tutorials,
but
for
recent
questions
that
are
much
less
and
one
of
the
reasons
for
that
is
that
there
are
super
hard
to
maintain
at
the
moment
due
to
many
reasons
and
yeah,
it's
one
of
the
problems
we
have
in
the
community.
A
Also,
if
you
go
to
tutorials
so
when
I
was
working
on
them
in
december,
I
spent
some
time
to
stream
outline
captain
and
parameters,
because,
even
if
they
were
present,
they
were
not
working
for
users
set
up
some
learning
paths
etc.
But
generally,
these
tutorials
are
still
a
big
problem
for
us
and
thanks
to
everyone,
looking
into
them
I'll
show
how
they
are
organized
internally.
A
At
the
moment,
full
disclaimer.
I
would
be
happy
to
completely
rework
the
engine.
We
have
a
pending
task
of
moving
everything
to
enter,
and
hence
those
key
doctors
so
that
we
could
remove
duplication,
introduce
common,
macros,
etc,
etc.
Right
now,
it's
not
there.
So
I
will
show
how
it's
implemented.
Now
we
have
a
repository
captain
tutorials.
A
These
tutorials
are
basically
based
on
google
code
lab
at
least
some
bits
of
these
tutorials,
so
google
code
labs
is
a
project
where
they
have
a
bunch
of
tutorials.
So
you
see
a
familiar
interface.
A
It's
open
source
project,
the
problem
with
that
that
we
actually
have
a
copy
of
that.
So
we
are
not
really
using
the
recent
version
older
one,
and
there
is
no
real
support
for
us
to
even
update
the
version,
because
it's
just
a
part
of
the
same
repository.
So
it's
basically
a
hard
fork
for
us
and
just
unwrapping
the
things
would
be
a
change.
This
is
why
you
can
see
various
finds
like
field
bezel
here
so
because
google
uses
bezel
to
build
everything.
A
I
have
no
idea
where
it
even
works
for
the
website,
so
I'm
going
to
show
how
it
works
actually
for
us
right
now.
So
what
we
have,
I
think
it
actually
works,
but
yeah
it's
a
separate
story.
So
what
you
have
now
you
can
see
that
there
is
contributing
candy
with
me
in
the
here.
A
Still
there
they
just
commented,
and
there
is
there
are
some
other
things,
but
basically
this
very
very
basic
introduction
guide.
Also
we
have
contributing
md
on
the
top
level
and
contributing
md
is
a
copy
of
codelabs.
It
hasn't
been
updated.
It
has
zero
relevance
to
this
tutorial
at
the
moment.
C
A
Yeah
yeah.
This
is
why
I
didn't
want
you
to
explore
it
on
your
own,
because
it's
a
bit
complicated.
So
what
we
have
else
here
we
have
site
so
site
is
basically
content
that
actually
gets
published.
A
You
can
see
that
now
there
is
go
file
et
cetera
and
they
get
some
hint
how
to
actually
get
it
running.
So
this
with
me
is
what
you
need
to
follow.
If
you
want
to
get
it
running,
so
npm
install
would
set
up
the
environment
for
you
well,
mostly
because,
due
to
reasons
unknown,
for
example,
it
doesn't
install
gulp,
so
you
will
have
to
install
it
manually,
but
basically
to
do
the
development
you
have
to
follow
this
guide,
so
tutorials
site,
good
labs
and
yeah.
This
is
something
you
can
follow.
A
A
So
our
tutorials
look
like
that.
So
you
can
see
a
bunch
of
stuff
and
yeah.
You
can
already
see
what
one
of
the
problems
is
notification,
so,
for
example,
captain
dyna
trace.
It
doesn't
exist
for
previous
version
for
a
thirteenth
version,
but
for
example,
if
you
open
parameters
it
exists
and
you
can
see
how
many
versions
and
basically
all
of
them,
are
copy
and
paste.
A
So
you,
if
you
open
photons,
you
can
actually
see
a
bunch
of
content,
including
images,
etc.
Images
are
not
copy
and
paste.
The
images
are
stored
in
a
separate
location,
but-
and
you
can
also
see
some
macros,
so
basically,
cod
labs
tries
to
work
around
some
limitations
of
markdown
by
adding
some
additional
functionality
on
the
top
of
it.
So
basically
it's
pattern
matching
and
also
all
these
snippets
so
way.
Snippets
selected.
A
So
for
every
version
you
have
to
copy
paste
all
the
snippets,
so
javane
did
it
seven
days
ago
for
version
13,
and
you
can
see
that
there
is
a
bunch
of
content
there
so
including
some
files
like
feedback
md,
which
obviously
have
no
reasoning
to
be
snippets.
It's
just
a
matter
of
the
current
site
architecture
to
get
it
working
pretty
much
what
else
we
have
in
this
needs.
I
also
have
some
files
like
setup,
quality,
etc.
A
It's
also
in
snippets
because
it's
referenced
a
few
times
from
the
tutorial,
so
they
decided
to
actually
extract
it
again.
It's
if
you
had
to
ask
the
doctor
if
you
had
under
a
macros,
it
could
have
been
done
with
native
tools
quite
but
this
is
how
it's
implemented
now
and
yeah.
It's
quite
a
challenge
to
go
through
that.
A
So
these
are
snippets
and
yeah.
There
is
a
bunch
of
them.
So
basically
there
is
no
particular
structure,
so
everything
you
want
to
duplicate
goes
here.
At
the
same
time,
there
is
no
macros,
for
example,
for
versions
etc.
So
when
you
copy
the
things
you
actually
have
to
copy
all
the
stuff-
and
sometimes
you
have
to
ensure
that
the
links
are
correct.
A
A
Well,
so
it
depends
because
you
have
to
manually
update
the
things
so
obviously
sometimes
it
gets
mixed
missed
it
just
becomes
a
mess,
so
this
part
would
have
been
actually
moved
to
templates
legitimately,
because
this,
what
are
templates
for
you
could
have
a
macros
here
that
updates
version
etc.
This
is
what
I
wanted
to
do,
but
more
look
at
the
site.
I
just
want
to
run
it
with
fire
and
the
new
one
from
scratch.
A
A
Okay,
so
what
else
important
about
the
tutorials
you
can
see
that
we
are
using
cloud
is
basically
a
part
of
codelab
which
basically
does
bits
of
the
build
and
here's
one
of
the
limitations
coming
that.
Well,
basically,
you
need
to
use
one
of
the
platforms
supported
by
card
and
even
if
you
have
windows
here,
it's
not
actually
going
to
work
right
now,
because
right
now,
in
order
to
build
all
this
happiness,
all
these
macros
et
cetera.
You
need
description,
and
this
script
supports
only
linux
and
macos.
A
A
So
basically,
you
can
also
see
that
there
are
some
interesting
things
like
putting
google
analytics
right
here
in
the
script,
etc.
I'm
not
sure
from
where
it
comes.
You
can
take
blame
but
yeah.
I
guess
jurgen
did
it
at
some
point
so
yeah
the
structure
for
this
thing
is,
but
actually
so,
if
you
really
want
to
get
it
running,
it's
not
enough
to
just
run
a
gulp
file
like
we
recommended
here,
because
golf
file
won't
serve
you
tutorials.
You
firstly
need
to
run
builder
to
generate
the
final
tutorials.
A
I
will
show
you
how
it
looks
like
in
the
code
in
a
second,
so
I'm
using
psl.
A
So
yeah
how
it
looks
like
yeah,
so
basically
we
have
this
side.
You
can
see
that
there
is
a
bunch
of
additional
change
files
here,
all
over
again.
C
B
A
So
tutorial
site
yeah,
and
here
we
have
tutorials.
So
first
of
all,
we
go
tutorials
and
yeah.
Here's
all
this
happiness,
yeah
so
mike,
if
you
use
so
you
use
the
avi
right.
Yes,
you
have
a
problem
yeah
for
this
particular
episode.
A
A
Okay
and
yeah,
you
can
see
that
already
you
have
some
interesting
events
happening
on
my
clean
build
because,
for
example,
stupid
is
missing.
I
have
no
idea
why,
let's
see.
D
A
Basically,
it
means
that
the
repository
is
already
bogus,
so
we
can
fix
it
while
we
okay,
so
okay.
A
A
Okay,
then,
I
will
definitely
fix
it
later,
but
you
know
so
in
theory,
you
run
this
builder
and
then
it
prepares
all
the
content,
but
it
actually
can
serve
yeah
so
and
to
serve
it.
We
actually
use
gold
so
for
gulp
we
have
by
the
way
it
doesn't
fully
clean
up
when
you
run
rebuild,
which
is
also
a
problem,
because
you,
if
something
is
wrong,
you
have
a
kind
of
messed
up
tutorial.
A
Okay,
so
for
real
me
yeah,
you
basically
see
all
the
commands
there.
So
we
after
you
build
the
tutorials.
You
first
need
npm
install,
so
you
can
get
all
these
tools
and
you
also
need
belt.
So
in
galpa
there
are
two
ways.
Actually,
so
you
can
use
a
gulp
serve.
A
A
A
Just
the
running.
A
You
can
browse
all
the
changes
etc.
Again
now
I
have
no
idea
what's
happening,
for
example,
version
12.,
I
guess
well,
we
can
see
feedback
so
something
gets
generated,
but
again,
I'm
not
sure
how
it
works
with
this.
If
it's,
what
is
broken?
What's
not
currently,
my
repository
should
be
a
complete
mess
and
actually,
if
you
run
the
script
without
minus
e,
I
guess
it
actually
gets
generated.
A
So
marquee
mark
is
basically
this
templating
engine,
but
why
the
heck
they
need
to
authorize
to
google
drive,
I
don't
know,
and
they
have
strong
suspicion
that
it's
jurgen's
google
drive
so
jorge
used
to
work
for
dana
trace.
Then
he
left.
So
am
I
better
that,
after
his
account
got
deleted,
this
thing
doesn't
work
I
yeah.
So
again
I
have
no
idea
what
exactly
that
I
will
take
a
look.
A
So
for
live
development.
If
you
use
linux
or
macos,
you
are
perfectly
fine.
You
can
just
develop
locally.
You
can
browse
the
things
for
windows.
It's
a
bit
more
complicated
because
for
windows
you
either
need
to
mount
a
directory
and
one
of
the
things
there
that
by
default,
if
you
use
vsl2,
so
you
can
just
go
to
linux,
ubuntu.
A
A
So
you
would
need
to
actually
configure
a
mapping
and
volume
so
that
mount,
I
don't
yeah,
it's
smart
context
on
windows,
so
you
would
need
to
map
this
repository
if
you
want
to
edit
visual
studio
code,
otherwise
it
won't
work.
We
can
do
mapping
now
if
you
want-
or
we
can
just
try
to
do
the
bypass.
I
think
it's
better
touching
up
this
directory,
because
otherwise
they
will
be
hitting
this
crop
every
time.
B
A
A
A
So
this
is
basically
the
same
repository,
but
it's
local,
so
I
usually
do
development
just
in
visual
studio
code
because
it
provides
a
decent
level
of
markdown,
browsing,
previews,
etc.
Obviously,
none
of
that
really
works
with
tutorials
with
cloud
with
all
this
templating,
so
you
don't
get
support
for
that
at
all.
A
You
would
need
to
rebuild
this
build
aside
sh
every
time
and
it's
also
broken
by
now.
So
I
guess
we
will
need
to
do
a
lot
of
hot
fixes
to
just
get
the
contributing
flow
running
right
now.
B
Can
you
go
back
to
any
the
the
main
like
where
you
go
into
the
prometheus,
the
full
tour
or
whatever?
Okay
there's
another
weird
thing
there
that
I
mean
just
from
a
usability
thing:
the
display
of
it,
the
the
rendered
okay.
A
A
Basically,
the
layout,
you
can
see
that
some
bits
are
already
tricky
and
not
really
synthesis
syntactically
right.
So,
for
example,
even
here
you
can
see
that
renderer
breaks,
because
that
is
how
markdown
works
again.
It
totally
depends
on
your
engine,
but
syntax
wise.
It's
not
correct,
and
here,
if
you
go,
you
can
see
all
this
needed.
B
Go
on,
I
want
there's
a
specific
thing
here
happening
in
this
tutorial.
That's
also
bizarre
just
construct,
but
it's
in
go
to
like
the
next
section
or
something
I
think
it's
not
in
the
welcome
section,
but
in
most
the
others
it's
there
yeah.
A
A
A
A
So
here
you
have
duration,
10
minutes
and
the
code
labs
parses
it
and
automatically
tracks
it.
Obviously,
whether
it's
10
minutes
or
not,
it's
a
separate
story,
but
yeah
this.
What
you
have
pretty
much
the
same
for
other
entries,
so
all
of
that
modulus
legit
markdown
files,
some
template
didn't
but
generally
it's
part
now.
A
A
Maybe
I'm
just
using
code
version,
so
actually,
if
you
open
13
right,
we
didn't
have.
We
have
full
run
parameters
for
13..
It's
actually
quite
interesting,
because
we
have
no
blob
file
for
that.
Okay,
we
will
figure
out
it
later,
but
you
can
see
that
there
are
also
some
codelab
files.
So
it's
basically
additional
metadata.
A
It
shows
how
to
actually
pass
through
the
things,
and
it
defines
an
entry
point
for
that.
So
you
can
see
url,
you
can
see
feedback,
metadata,
etc
and
also
tags.
So
tags
are
what
used.
So
you
can
see
advanced
and
advanced,
gets
mapped
to
this
level
advanced
again.
It's
a
kind
of
hardcoded
logic
that
allows
to
capture
way
is
what
they're
pretty
much
the
same
so
for
log
etc.
A
A
A
A
A
You
can
see
that
these
things
don't
make
much
sense
these
days,
so
they
update
updated.
We
also
open
11th
by
default.
So
when
you
open
the
default
page
right
now,
11.
So
when
I
was
updating
this
site,
but
then
it
was
johannes
and
giovanni
doing
that
and
looks
like
they
forgot
to
modify
so
here
by
default,
we
go
to
tutorials
for
elevens,
not
for
13,
and
this
one
of
the
problems.
I'd
say
we
also
need
to
fix.
D
I
think
there's
also
a
problem
with
if
you
go
back
to
the
tutorial
site
to
the
webpage,
if
you
click
on
captain
version
6
or
something
and
then
from
take
the
prometheus
tutorial
to
the
left,
left
left
first,
one:
oh
yeah,
okay,.
A
A
Well,
yeah,
it's
probably
some
of
the
versions
were
removed,
but
here
one
of
the
problems
is
that
it
has
no
russian
tech
at
all,
because
basically
it
was
the
first
version
of
captain
when
all
this
engine
was
introduced
and
basically
when
they
needed
a
new
version.
Instead
of
redesigning
this
site,
it
was
basically
extensive,
so
just
copy
whatever
files,
and
it
has
been
extensive
ever
since
so
every
month
there
is
a
bunch
of
copy
based
across
all
these
repositories
and
basically
one
could
get
completely
lost.
B
B
A
When
you
clean
the
top,
because
nobody
cleans
it
up,
okay,.
C
A
Yeah
also
tutorials
works
in
funny
ways,
so
sometimes
I
mean
cod
labs,
so
I'm
pretty
sure
that
these
versions
were
hidden
before,
but
now
they
appear.
For
example,
in
my
case
I
tried
to
hide.
A
Closed
it,
but
the
patch
which
removed
this
tutorial
after
you
disabled
to
the
public
demo,
it
stopped
working
again
with
this
amount
of
different
bugs
and
the
build
script
without
checking.
I
have
no
idea
what
exactly
breaks,
but
by
now
I
assume
that
our
engine
is
half
broken,
still
displays
the
things
but
yeah,
that's
it.
A
So
what
is
important
about
these
views
that
yeah?
You
can
see
that
there
are
some
such
visualizations
and,
for
example,
view
so
it's
also
view
for
version
13..
So
this
is
a
version's
vector.
Basically,
this
version
inspector
you
have
is
just
this
feature
of
use
from
codelab,
and
here
when
you
take,
for
example,
the
version
captain.13.
A
It
opens
you
this
view
and
this
view
you
can
see
that
it
matches
basically
things
by
tag,
so
it
displays
all
codelab
entries
which
have
this
tag.
It
also
excludes
some
items
because
otherwise
it
just
pushes
everything
back
on
the
page
as
well,
because
this
is
the
default
behavior
of
good
luck,
anything
that
will
copy
all
the
images.
A
A
separate
story
and
yeah
one
thing
or
with
all
these
filters
that
some
something
definitely
stopped
working.
I
didn't
know
what,
but
all
the
things
will
be
fetched
just
by
a
label
and
in
tutorials
in
principle
you
should
be
able
to
just
disable
them.
So
let
me
open
tutorials
again.
A
Just
yeah
a
couple
captain
public
demo,
so
this
is
basically
metadata
for
what
I
wanted
to
remove,
and
here
you
can
see
that
there
is
no
longer
time
for
that,
so
it
shouldn't
have
been
even
fetched,
but
it
was-
and
maybe
it's
just
because
we
do
not
clean
up
the
workspace
correctly
but
yeah.
I
don't
know
it's
a
public
json.
So
here
we
have
the
charts,
but
here
I
believe
it
was
explicitly
disabled,
maybe
not
get
status
published,
so
maybe
something
changed
but
yeah
overall.
This
is
how
it
should
work.
A
So
you
touch
this
metadata
and
hide
tutorial
through
that
you
can
put
additional
labels
so
that
it
gets
fetched
correctly.
I
also
try
to
use
this
labels
to
implement
learning
paths.
For
example,
if
you
click
quick
start,
you
get
level
introduction
same
for
football,
etc.
So
all
of
that
actually
gets
through
category
and
sometimes
level
matching.
B
B
A
It's
actually
a
good
question,
because
if
we
would
just
build
something
quickly
with
entera,
maybe
it's
something
we
should
consider
but
yeah
the
problem
that
migration
from
markdown
to
ask
talk
in
this
case
won't
be
trivial
because
all
of
all
these
macros
etc.
A
So
you
cannot
easily
automate
that
otherwise,
what
I
would
have
done,
I
would
have
just
used
one
dock
as
converter
and
basically
a
few
github
actions
and
you
would
have
got
everything
running,
but
right
now
it
doesn't
seem
to
be
possible
for
this
particular
repository.
So
yeah.
You
need
to
create
a
hard
fork
of
everything,
but
in
this
case
it's
also
a
question
whether
it's
just
tutorials
or
the
entire
documentation
side
and
whether
it's
a
project
of
a
size.
A
B
Yeah
another
question:
do
we
I
mean,
is
there
do
we
want,
apart
from
the
what
the
tools
currently
are?
Why
do
we
want
tutorials
in
a
repo
separate
from
the
rest
of
the
documentation?
I.
D
B
A
So,
for
example,
for
parameters,
so
primitives
mostly
describes
parameters,
integration
service
and
for
them
I
would
like
to
see
these
tutorials
actually
being
integrated
to
captain
country,
and
here
we
have
primitive
service
maintained
by
christian
and
the
team.
A
So
my
vision
would
be
that
all
the
communication
for
parameter
service,
whether
it's
tutorial,
whether
there
are
docks,
pages
etc
going
inside
this
repository,
so
that
whomever
does
patches
etc
as
a
part
of
their
pull
request.
They
update
the
documentation
as
well.
So
this
is
what
we
call
documentation
is
called
for:
real,
not
keeping
docs
in
down.
D
Yeah,
that's
a
very
good
idea,
because
also
the
ownership
might
change
a
little
bit.
This
is
what
we
have
with
the
same
problem
with
the
dyna
trace
service
in
captain
country
and
we're
referencing
lots
of
stuff,
and
if,
if
we
have
lots
of
snippets
in
every
place,
then
we
basically
end
up
changing
something
that
somebody
else
needs,
probably
in
a
different
way
and
also
the
captain
examples
come
into
play
then
as
well
right.
C
A
So
yeah,
this
is
basically
how
I
would
like
to
see
and
if
you
want
to
take
a
bite,
for
example,
just
putting
some
tutorials
in
the
repository
and
see
how
we
could
build
it
with
entera.
I
actually
wanted
to
do
that.
So
if
I
ever
get
a
few
days
of
non-interrupted
time,
this
is
actually
the
first
thing
I
would
like
to
do
just
rebuild
this
engine
and
see
how
it
works
well
and
do
not
hope
today
to
do
that.
So
I'm
not
sure
whether
you
saw
it
york.
A
We
actually
created
the
google
summer
of
code
project
id
for
that,
so
we
added
a
project
id
a
new
documentation
site
engine
which
is
basically
well
basically
describing
all
this
runs
in
text
more
or
less
and
yeah.
So
as
a
part
of
that,
you
would
actually
want
to
rework
that
so,
if
you're
potentially
interested
to
be
a
kind
of
stakeholder
or
a
mentor
for
this
project,.
B
A
That
there
was
a
lot
more,
but
it
wasn't
sustainable,
because
if
you
go
to
version
0.7
here
you
get
argo
cd,
which
is
essential,
which
we
promote
not
no
tutorial
for
the
recent
version,
even
though
even
a
captain
team
uses
rxcd
in
some
cases,
hd
devops
imports.
A
Custom
notification
service-
I
think
it's
obsolete,
because
now
we
have
a
web
hook,
service,
etc,
but
okay,
also
installation
guidelines
which
are
again
quite
important.
So
if
someone
asks
how
to
install
captain
on
gt,
then
this
is
where
you
would
like
to
point.
But
again
this
is
dated
it's
no
longer
relevant
because
we
mostly
moved
to
hand,
charts,
etc
and
then
mini
cube
again
important
for
those
who
run
on
the
macos
want
to
evaluate
it.
A
Openshift
again
important,
then,
and
we
want
to
have
more
tutorials
for
tools
for
integrations,
but
with
the
current
engine
is
just
not
sustainable.
I
mean
we
still
have
an
open
question
who
maintains
dana
trees
tutorials
I
mean
on
a
regular
basis
and
use
all
the
current
structures
just
impossible
to
actually
get
it
maintained.
So
it's
a
huge
endeavor.
B
C
A
This
is
correct,
although
even
with
that
there
are
a
few
optional
obstacles,
because
we
also
have
this
compatibility
matrix
and
it
means
that,
for
example,
remedial
service
and
hence
the
tutorial
would
be
compatible
only
this
particular
version
of
capital,
and
it
means
that
you
actually
getting
the
troubles
with
configuration
matrix
again.
I
don't
think
it's
a
really
big
problem,
because
I
would
rather
prefer
to
have
just
tutorial
for
the
recent
personal
primitive
service
and
for
the
recent
version
of
captain
and
that's
it.
D
A
B
I
mean
I'd
like
to
go
back
just
looking
at
it.
It
seems
to
me
that
we've
got
way
too
many
tutorials
that
are
covering
the
same
material
yep,
and
then
we've
got
all
these
things
about
how
you
do
this
and
then
we've
still
got
installation
stuff
stuck
into
the
what's
called
reference
pages
for
like
sli
and
slo,
and
I
mean
it's:
it's.
B
B
A
A
A
Remove
the
most
of
version,
specific
content
from
here
I
mean
any
of
these
things
again
should
go
to
repositories
so,
for
example,
installation
etc.
So
this
just
goes
straight
away
to
premiere
service
right.
A
A
Yes,
so
some
content,
like
gpr
et
cetera
well,
we
can
aggregate
that,
but
my
vision
for
that
would
be
that,
once
a
new
release
is
done,
we
just
freeze
the
recommendation
for
the
previous
version.
Take
it
from
whatever
branch
or
tag
basically
do
not
touch
it
unless
it's
really
needed.
So
we
should
stop
doing
this
much
massive
patches
by
developers
who
have
to
update
multiple
versions
so
that
they
don't
get
lost.
A
Yes,
so,
for
example,
yeah,
I'm
just
looking
so
here,
we
have,
for
example,
patches
that
update
multiple
versions,
how
it
looks
like
it's
like
that,
so
you
have
to
duplicate
all
the
content
for
12's,
then
update
13
with
the
same
content
and
obviously
something
will
get
good.
A
So
it
complicates
reviews,
because
you
need
to
update
multiple
places
and
basically
makes
no
much
sense
to
update
the
older
version
in
the
most
of
the
cases,
because
we
recommend
using
the
latest
version
the
well
basically
this
one
months
for
this
cycle.
I
think
it's
okay
for
the
captain
and
the
current
stage
you
might
introduce,
let's
say
one
year
cs
or
whatever
later,
but
maintaining
these
tutorials.
A
B
Okay,
yeah
we're
jumping
around
so
yeah.
A
Yeah,
I
think
we
go
beyond
tutorials,
but
I
guess
we're
on
the
same
page.
We
need
to
do
something
about
that,
and
also
there
is
this
page.
I
have
no
idea
from
so
it
basically
points
to
all
these
filters,
etc,
but
it
works
only
as
long
as
you
have
a
default
version.
I
pinned
to
developments.
A
Yeah,
so
my
suggestion
would
be
to
just
take
newcomer
friendly
tickets,
we
discussed
so
the
main
point
is
just
to
ensure
that
tooling
works
for
these
things,
but
yeah.
I
wouldn't
expect
you
to
spend
too
much
time
on.
B
C
A
Yeah,
so
for
me,
my
recommendation
would
be
to
just
understand
how
it
works
using
the
goal
of
reworking
that,
because
tutorials
as
they
exist
now
they
are
just
well.
Basically,
we
can
try
to
watch
them,
invest
much
of
effort
to
keep
them
running,
but
I
don't
think
it's
the
right
thing
to
spend
time
on
yeah.
B
I
mean,
like
we
had
a
bunch
of
engineers
who,
instead
of
doing
good
documentation,
set
just
did
all
these
tutorials
I
mean
most
of
these
are
not.
I
don't
see
the
point
of
most
of
these
tutorials.
You
know
when
you
look
at
them.
It's
like
what
is
the
goal
here.
A
Well,
they're
reasonable
the
program
that
was
too
many
installation
steps
and
not
clear
what
would
be
the
goal
well.
Actually,
there
are
quite
good
introductory
things
for
some
cases.
I
think
what
we
would
achieve,
etc,
but
actually
yeah
so.
B
I
wouldn't
put
that
in
a
tutorial.
I
would
put
that
in
the
intro
to
the
product
and
like
the
getting,
I
guess
whether
you
write
the
tutorial,
you
know
for
the
tutorial.
The
tutorial
is
going
to
be
basically
a
cookbook
sort
of
thing.
Do
this
and
look
at
this
happen?
Do
this
and
look
at
this?
It's
not
look
at
the
architecture.
These
are
all
the
pieces
that
go
together.
B
We're
only
going
to
mention
it.
You
know
it's
like
that's
that
should
be
in
regular
guide
material.
A
Well,
basically,
yes,
I
don't
see
much
difference
between
guides
and
tutorials
to
be
honest.
Well
maybe
for
technical
writers
there
is
no
difference,
but
yes,.
B
B
These
tutorials
are
full
of
information
that,
if
I'm
using
the
system,
I
want
to
go
back
and
reference
a
tutorial
is
something
I
go
through
once
maybe
twice
to
get
started
and
get
the
big
picture
and
then
and
the
guide
material
I
mean
we
there's
a
limited
amount
of
guide
material
that
people
should
be
looking
at.
Very
often,
we
need
really
good
reference
material
and
instead
of
a
reference,
material
is
written
like
tutorials
and
doesn't
have.
I
don't
know
somewhere
in
there,
there's
those
blogs
back.
B
You
know,
I
finally
saw
you
know
a
thing
for
shipyard,
but
you
know
what
I'm
saying
is
yeah
so,
and
I
mean
I
don't
know,
then
how
we
you
know.
Obviously
we
can't
fix
everything
at
once.
So
how
do
we
go
at
it?
But
so,
okay,
because
I
was
thinking
you
wanted
to
really
try
to
get
these
tutorials
straightened
up.
You
just
want
me
to
cruise
through
and
get
a
view
of
what
a
mess
they
are
right.
A
Yep
so
yeah,
there
are
two
ways:
firstly
yeah.
Let's
say
you
get
something
like
one
week
of
smallest,
uninterrupted
time.
We
just
work
together
and
get
the
under
engine
running,
or
maybe
we
do
it
as
a
part
of
gsoc,
but
so
my
concern
about
having
this
project
in
jsoc
that,
if
it's
done
by
the
end
of
summer,
so
it's
still
multiple
months
until
we
actually
have
it
running.
B
Yeah
all
I
know
I
mean
I'm
not
a
tools
person,
but
I
also
know
at
cloud
b's
the
just
before
we
merged
training
in
docs.
We
brought
in
this
guy
named
jeremy,
who
was,
as
my
grandfather
said,
I'd
like
to
buy
him
for
what
I
think
he's
worth
and
sell
him
for
what
he
thinks
he's
worth.
B
He
knew
it
all
they
are
there.
You
know
there
was
nothing.
He
did
not
know
more
about
than
anybody
else
in
the
world
and
the
antora
that
he
was
setting
up
was
an
absolute
disaster
and
then
we
got
merged,
and
that
was
something
that
tammy
and
ben
actually
knew
what
they
were
doing
and
we
just
converted
over
and
used
their
stuff.
So
that's
the
one
thing
that
concerns
me
about
doing
this
with
gsoc
is:
do
we
get
somebody
who
knows
what
they're
doing
enough
to
set
it
up
right
or
do
we
end
up
with
a.
A
Well,
yeah
I
actually
reached
out
japan
asked
for
some
expertise.
Maybe
they'll
cut
cut
this
part
from
the
recording
but
see
I
didn't
see
anything
secret.
So,
okay,
if
we
could
get
some
advice
on
how
how
cloud
is
on
terra,
maybe
if
they
would
be
willing
to
open
source
the
engine.
B
Probably
tammy
tammy
is
of
the.
Why
would
I
do
anything
for
open
source.
B
B
A
It
should
work
with
mark,
I
haven't
tried,
but.
C
A
A
Yeah,
so
I
have
no
idea
what
it
actually
gets
displayed,
but
let
me
know
if
I
forget:
the
tutorial
still
works.
B
B
Right,
no,
I'm
just
let's
not
fill
it
in
I'm
just
saying
what
do
we?
Let's
say
we
have
nothing.
We
have
something,
but
let's
say
we
have
nothing.
What
tutorials
do
we
need.
A
B
A
So
basically
the
idea
was
for
learning
past
that
you
start
from
quick
start
or
maybe
watching
it
live.
So
it's
currently
disabled
on
this
page.
Then
you
go
through
full
tour
and
then
you
explore
some
particular
use
cases
like
quality,
keys,
etc
for
separate
tutorials
regarding
the
installation.
So
quick
start
is
what
we
actually
need.
A
In
any
case,
the
current
quick
start
is
okay,
it's
focusing
basically
on
installing
captain
a
few
other
items.
So
you
see
that
steps
are
really
small
because
it
actually
points
here.
The
captain's
website.
Right
again,
it's
something
called
I
just
had
to
pick
patch
but
yeah.
This
is
basically
what
I
rewritten
in
december
and
it
used
to
work.
A
C
A
Etc,
but
still
it's
quite
heavy,
so
this
is
minimum
requirement
and
even
on
this
minimum
requirement
it
doesn't
work.
So
what
andre?
Sorry?
What
adam
created
he
created.
A
That
is
based
basically
on
a
kubernetes
under
kind
instead
of
kit
3d,
it's
a
much
more
lightweight.
Well,
it's
still
quite
heavy,
but
it
gets
you
quickly
to
a
hello
world
running.
A
A
A
A
There
is
another
one
which
is
hello
world,
so
this
one
basically
shows
how
to
set
up
captain
in
kind
out
of
the
box,
and
there
is
another
new
tutorial
which
shows
just
captain
this
kind.
B
B
D
B
B
A
So
yeah
for
the
tutorials
sorry
I
had
to
go
into
my
inbox
to
find
it
quickly.
A
A
A
Yeah,
so
it's
also
something
I
will
probably
need
to
report
to
this
repository,
but
I
think
that
in
principle
it
would
be
a
really
good
to
start
demo
if
you
could
get
it
working
properly.
B
B
Want
to
put
on
that
list
the
strong
possibility
that
we
should
just
abandon
the
the
tutorials
repo
and
whatever
we're
doing.
A
What
this
one.
C
A
D
A
Well,
firstly,
showing
everything
from
kind,
because
again
our
tutorials
are
super
heavy.
It
requires
users
to
set
up
kubernetes.
I
mean
it's
always
a
challenge,
so
he
took
kind
because
this
kind
you
can
actually
use
native
docker
image,
can't
find.
A
So
I
will
probably
play
around
with
this
tutorial
after
the
meeting,
because
I
actually
want
to
get
it
running
again
but
yeah.
I
think
it's
a
good
bet
to
actually
have
a
lightweight
starter,
and
now
it's
just
requests
cleaning
up
everything
manually,
because
this
tab
script
isn't
full
tolerant.
So
when
something
fails
in
the
middle,
we
are
basically
screwed.
A
A
I
understand
and
yeah
thanks
for
your
concern.
I
actually
agree
with
you,
but.
B
I
know
this,
but
I
mean
the
thing
is:
is
with
you
also
what
you
thrive
on
is
doing
stuff
like
this,
so
but
so,
let's
we'll
just
be
careful,
I'm
not
saying
don't
do
it,
but
okay,
so,
okay,
so
I'm
doing
analysis
only
for
now
and
I'm
not
even
worrying
about
trying
to
fix
this.
I'm
just
scoping
out
what.
A
B
Right
well,
like
you
said
what
I
did
well
what
I
started
with
the
contributing
I
just
going
through
and
reading
everything
and
noting.
What's
where
and
making
comments,
and
I
did
something
like
that.
With
the
reference
pages
and
stuff
I
haven't
yeah,
I
haven't
gone
through
everything,
but
I'll
go
through
because,
as
I
say,
some
of
these
tutorials
have
some
really
good
information
that
I
don't
think
belongs
in
a
tutorial,
but
it
belongs
someplace
else
where
people
can
find
it
so
yep.
A
But
yeah
right
now,
I
just
I
think
it's
a
wasted
effort
to
rebuild
the
documentation
account,
I
think
so.
Why
would
just
create
spin-off
for
dogs?
Captain
io
based
on
cuba,
sorry
based
on
entera
and
then
actually
think
what
do
we
do
with
the
rest
of
the
content,
the
remaining
huge,
but
maybe
moving
over
the
things.
B
B
B
And
what
I'm
thinking
is,
I
think,
a
fairly
simple
antora
implementation
would,
let
us
do
asciidoc
versions
of
the
markdown
files
that
we've
got
now
pretty
easily.
Then
we
it
gets
fussier
as
we
start
having.
You
know
all
the
different
stuff
that
we
pull
in,
but
but
maybe
but
maybe
that's.
The
thing
is
that
we
do
a
quickie
just
to
get
everything
into
ascii
doc
under
antora
and
then.
A
Actually,
what
we
could
try,
we
could
try
ascii
dock.
We
could
start
just
a
new
repository
dogs,
kaplan
captaindocs.github.com.
A
So
I
could
create
a
simple
github
action
that
just
shows
the
thing,
so
it
would
be
kind
of
our
sandbox
to
play
around.
For
example,
we
could
offer
adam
to
be
a
kind
of
pilot
project
to
be
moved
to
the
new
tutorial
to
the
new
recommendation.
B
A
Yeah,
I
think
we
set
up
a
new
repo
for
documentation,
deliberately
this
plan
to
not
put
too
much
content
there.
B
A
Well,
the
problem
is
not
this
conversion.
The
problem
is
putting
documentation
in
the
right
place
right
also,
we
will
have
a
question
about
readme
files,
because
readme
files
for
services
also
hosted
on
artifact
hub,
so
you
can
just
switch
towards
keyboard
if
it's
not
supported
in
our
default
hub.
A
Well,
artifact
hub
script
is
also
kind
of
yeah.
Well,
it's
not
dark
magic,
it's
quite
straightforward,
but
I
would
rather
be
careful
and
maybe
start
from
supporting
markdown
in
the
enter
engine
so
that
we
can
gradually
move
over
the
things.
B
But
what
back
to,
but
I'm
I'm
sure,
some
place,
we're
going
to
have
some
place
on
the
web
or
in
the
doc
set
where
it
says
tutorials
and
then,
if
we
have
some
of
some
tutorial
information
in
prometheus
and
in
the
dynatrace
are
those
going
to
get
referenced
from
there.
A
B
A
Well,
I
wanted
to
work
on
that
for
two
months,
because
every
time
I
look
at
these
tutorials
and
this
documentation,
I
just
want
to
nuke
it
from
the
orbit,
because
you
can
ever
anything,
that's
not
a
good
joke
these
days,
yeah
yeah.
Otherwise
I
want
to
get
rid
of
it.
B
Right,
it's
like
when
I
look
at
the
other
docks
I
mean
yeah
and
it's
like
there's
so
much,
there's
stuff.
You
know
I
can't
do
internal
cross
references
very
well
in
markdown,
and
I
I
can't
make
inline
comments,
and
you
know
this
sort
of
stuff,
but
there
is
stuff,
could
incrementally
work
on
but
yeah
when
you
looking
at
these
tutorials.
A
B
Fact
that
an
existing
tutorial
lacks
10
of
what
you
want,
you
do
not
write
a
new
tutorial.
You
fix
the
existing
one
that
we've
got
too
much
of
that
going
on.
So.
B
A
B
Yeah
yeah
once
there's
a
repo.
Well,
I
could
take
some
file
and
put
it
over
there
and
switch
it.
Tasky
talked
or
you're
saying
just
build
markdown
initially
right
and
do
ascii
doctor
conversion.
Next,
okay!
Well,
anything
you
want.
If
you
want,
you
should
not
be
spending
the
time
converting
files
from
markdown
to
ascii.
A
Oh
well,
I
will
probably
create
an
engine
that
converts
to
them
right
away
because
we
already
have
experience
with
pandora
and
with
creating
github
actions
for
jenkins.
So
this
is
what
I
will
do
here
as
well.
I
guess
okay.
B
Cool,
okay,
okay,
this
was
fun,
and
now
I
have
a
better
idea
of
what
I'm
supposed
to
be
doing.
I'm
just
reading
this
crap,
and
it's
like
what
am
I
supposed
to
be
doing
with
this
so
now
I
know
what
to
do
fabulous,
so
this
is
probably
it
for
you
I
hear
is
well.
They
said
in
vienna.
They
take
friday
afternoon
off.
Is
that
true
in
switzerland
also.
A
C
A
A
A
Well,
hey
it's
spring,
I
can
send
you
photos
later
yeah.
If
I
want
some
captain
chad
but
yeah,
that's
it.
B
And
we
will
talk
next
week
or
unless
you
have
something
I'll
be
watching
slack
over
the
weekend
too.
Don't
worry
if
you
need
something
before
that,
so
paying
me
if
you
need
me,
thank
you
so
much
for
all
of
this.