►
From YouTube: WG Platforms Project Meeting - 2022-11-09
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
B
Well,
it's
only
been
one
week,
so
you
know,
but
they
a
lot
a
lot
of
good
things
like
today,
the
entire
structure,
organizational
kind
of
structure,
kind
of
change,
because
the
thing
that
we
did
didn't
work
so
people
were
like
well.
Let's
change
that
then,
and
it
just
like
completely
redid
everything,
because
you
know
you
can
you
should
be
able
to
you
know.
So
it's
a
totally
different
mindset
than
like
big
corporates,
things
yeah.
A
B
A
B
C
Yeah
especially
I
work
at
Red
that
in
Consulting
I've
been
thinking
about
platform
platform
and
developer
experience
for
for
about
a
year
now,
and
then
thanks
to
Josh
I
discovered
that
we
have
this
working
group.
So
I
would
like
to
start
participating,
contributing
whatever
but
I'm
very
passionate
about
this
topic.
B
Excellent
welcome
and
it's
nice
to
meet
you
all
over
actually
for
now,.
B
A
B
A
B
C
Yeah,
so
so
let
me
let
me
make
me,
let
me
understand
what
we're
doing
in
this
group.
What
is
the
the
yeah.
A
A
I
I
do
have
a
slight
agenda,
but
right
I've
never
count
on
people
showing
up,
but
I
mean
I.
Think
that
the
purpose
of
the
group
Center
I,
know
Rafa
you're
in
tag
storage.
So
you
can
give
us,
you
know
perspective
from
there,
but
what
I?
You
know
what
I
see
is
our
mission
is
to
support
the
projects
in
ways
that
we
can,
which
I
think
is
mainly
by
introducing
them
to
each
other
and
rationalizing
them
in
a
space.
A
Yeah
support
end
users
in
ways
that
we
can,
which
I
think
is
by
gathering
feedback
from
them
and
and
channeling
it
back
to
the
projects
or
something
back
and
forth,
and
then
Third
Kind
of
to
some
degree
showing
people
or
guiding
people.
You
know
with
with
examples,
and
things
like
that,
like
like
potato
head,
is
an
example
or
other
blogs
that
we
write
or
that
we,
you
know
point
to
those.
C
Are
kind
of
okay,
yeah
I
think
in
tag
storage?
The
first
step
was
to
Define
storage,
Define,
Cloud
native
storage
right,
so
here
I
don't
know
we
want
to
translate
it,
but
essentially
yeah
defined
and
rationalize
this
space.
You
can,
you
can
be
monitoring
the
tooling,
but
I
think
it's
better
first
to
have
a
definition
right,
and
you
know
that
internally
I
was
trying
to
start
with,
create
some
content
around
the
definition.
C
Then
we
went
on
writing
the
performance
guideline
and
the
disaster
recovery
guideline.
Those
are
storage,
specific
things
right,
but
yeah
speaking
of
tax
stories,
they
meet
at
the
same
exact
same
time
as
this
meeting
I
had
to
make
a
choice,
but
it
would
tell
me
a
lot
if
we
could
move
this
meeting.
I.
C
B
B
Think
it
actually
is
a
fitting
name,
but
it
kind
of
doesn't
explain
what's
going
on,
but
but
you
know
it
is
actually
kind
of
fitting
there,
because
that's
kind
of
what
we're
doing
with
the
platform
right,
but
but
the
renamed
to
to
platforms
just
makes
sense
and
Josh
has
been
doing
like
a
document
about
about
actually
those
kind
of
things
like
specifically
like
you
know
what
is
what
is
a
platform
and
so
on
and
so
forth
and
and
I
think
that
is
a
it's
it's
it's
kind
of
like
we
did
with
with
open
githubs
right
like
one
of
the
first
things
and
like
most
important
things
were
like
just
defining
what
is
get
Ops
and
that
turned
out
to
be
principles.
B
So
you
know
they're
four
principles
to
what
get
offs
actually
is
not
necessarily
saying
that
we
need
to
do
that
for
platforms,
because
I
think
it
is
it's
a
much
rounder
term
right,
but
platform
can
mean
a
lot
which
is
usually
what
I.
When
I
speak
about
plan
for
engineering
and
stuff,
like
that,
I
kind
of
have
to
go
back
to
and
explain
that
like.
If
you
ask
me
what
you
know
how
it,
how
does
your
plugin
look
like?
It
might
be
totally
different
than
something
someone
else
has
been
doing.
B
B
D
B
Again,
like
right
now,
I'm
I'm
actually
working
on
I'm
working
on
internal
platform
stuff
at
my
new
company,
but
it's
it's
actually
to
facilitate,
creating
platforms
for
clients
and
customers,
so
yeah
I'm,
building
a
platform
that
creates
platforms.
You
know
just.
B
It
is
a
really
small
company
called
it's
well,
the
company
name
is
actually
42..
The
full
name
is
a
master.
42.
B
is
a
Norwegian
organization
with
a
lot
of
money
and
they
need
some
good
I.T
people.
So
you
know
it's
a
basically
startup,
more
or
less
okay.
C
D
C
Very
specific
problems-
maybe
some
guidelines
right,
yeah,
so
I'm
eager
to
join
our
efforts,
maybe
come
up
with
the
document
that
we
cannot
agree
on.
Yeah
I
have
my
experience.
I
built
my
platforms
right,
but
I'm
yeah,
yeah.
B
It's
one
of
those
things
where
you
could
potentially
said
you
know
in
a
in
a
corner
and
create
your
like
definition
of
a
platform
and
build
your
platform
and
and
then
you
talk
to
someone
else
and
they
have
a
totally
different
approach
to
the
exact
same
problem
and
they
both
solve
it
like
both
approaches
solve
is
the
platforms
is
very
much
a
thing.
B
That's
it
I
I,
usually
don't
I
I,
don't
like,
like
absolutes
I,
don't
like
when
people
say
this
is
the
right
option
or
you
know
you're
supposed
to
do
something
in
a
certain
fashion.
Some
cases
that
is
true,
obviously
but
like
platform
is
such
a
hard
thing
to
Define.
You
can
say
you
need
to
take
this
thing
and
connect
it
with
this,
and
then
you
have
a
platform.
You
know.
C
C
B
B
B
C
A
So
we
did
have
I've
been
trying
to
bring
a
lot
of
people
together
and
thank
you.
I
really
took
the
advantage
of
the
transition
of
the
name
to
and
kubecon
to
to
bring
a
lot
of
folks
together,
I
even
brought
the
humanotech
folks
in
centasso
a
few
others
I've
been
asking
everyone
to
to
look
at
this
document
and
we
yeah.
A
Please
ask
anybody
else:
the
internal
developer
platform,
people,
if
you
know
them
rough
up,
yeah
bring
them
in
I,
I
kind
of
been
thinking
with
this
initial
paper
like
let's
I,
want
to
gather
as
many
educated
opinions
as
we
can
get
us.
You
know
put
something
down
that
we
can.
You
know,
debate
and
plan
to
iterate.
A
You
know
I
wanted
I,
said
kind
of
in
the
in
the
issue,
I
kind
of
said:
let's
get
something
done
by
the
end
of
the
year,
so
that
we,
you
know,
have
something
to
talk
about,
and
then
let's
I
really
want
to
get
more
experience,
reports
from
end
users
and
in
fact,
I
met
the
the
folks
that
work
for
diagram
who
are
productizing
Dapper
were
very
eager
to
help
out
to
connect
us
with
with
the
implementers
of
platform
engineering
that
that
will
give
us
feedback
so
yeah,
yeah,
so
kind
of
that
was
kind
of
my
agenda
for
the
day
actually,
which
I
kind
of
put
in
there
was
to
discuss
that
paper,
which
I
think
has
evolved
to
me.
A
There's
two
things
that
I
have
been
kind
of
drafting
out
there.
One
is
the
key
attributes
of
a
platform,
both
the
positive
attributes
and
the
in
the
caveats,
and
then
the
next
section
is
here's
the
10
core
capabilities.
A
How
that's
you
know
that
like
resource
compositions
and
service
catalog
and
observability
and
pipelines,
and
which
I
really
want
to
debate
and
I
want,
you
know
a
lot
of
opinions
here.
So
we
can
have,
you
know,
a
rough
consensus.
This
is
the
thing
which
I
think
actually
I
mean.
My
goal
is
having
the
users
in
mind
like
these
platform,
Engineers
where's,
their
checklist
as
you're
building
your
platform.
What
are
the
things
you
should
be
thinking
about,
including
yes,.
C
D
C
You
want
to
go
for
it.
That's
fine
too
either
way.
But
yes,
that's
my
inventory
essentially,
but
as
you
saw
it's
an
inventory
of
capability
and
processes,
there
is
no
product
right,
yeah,
I
think
that's
that's
what
that's
where
I
would
start
a
list.
C
And
that's
fine,
we
that,
but
in
my
opinion
that
comes
later,
we
cannot
totally
do
it
right.
Oh,
you
want
to
implement
a
secret
Management,
Service.
Okay,
these
are
your
options
right,
but
first
you
have
to
understand
where
it
go
like
what
role
it
needs
to
play
in
the
platform,
then
the
rest
is
technicality
and
I
don't
even
like.
If
we
it's
going
to
be
very
hard
to
maintain.
If
we
start
doing,
you
know
those.
If
we
go
too
deep,
yeah.
A
A
You
know
if
we're
passionate
like
I
know,
Alex
Jones,
in
from
the
main
tag
had
wanted
to
work
on
multi-tenancy
and
I
started
getting
working
with
him
on
that,
and
it's
come
up
like
with
the
cluster
and
kcp,
but
even
other
things
like
Secrets
management,
like
I'm,
think
like
if
you
look
in
the
stock
which
I
just
shared,
where
about
halfway
down
on
page
two
or
three
is
where
right
now
there's
a
list
of
tons
of
capabilities
which
to
you
what
you
were
saying.
A
Rafa
is
more
abstract
capabilities
with
projects
on
the
far
right,
I'm
kind
of
thinking.
If
we
select
some
of
those,
then
the
next
step
might
be
to
double
click,
as
they
say,
you
know,
in
a
specific
domain
that
is
needs
clarification,
like
Secrets
management,
needs,
verification,
I,.
B
I
I
think
like,
for
instance
before
that,
because
you
have
the
the
diagram
of
the
the
components
I
think
I
I.
Think
like
that.
As
a
topic
like,
for
instance,
like
what
kind
of
capabilities
do
your
platform
need
that.
That
is
a
a
very
important
question
that
you
need
to
ask
yourself
before.
You
start
building
stuff
the
tools
that
you
use
for
that.
Obviously,
as
long
as
you're
as
long
as
it's.
C
B
Know
you
know
doing
it
for
cncf.
You
obviously
want
to
promote
cncf
projects
for
these
type
of
things,
but,
like
first,
you
need
to
know
what
you're
supposed
to
be
doing
with
your
platform
and
I
and
I
I
think
we
could
even
start
doing
for
like
a
better
termed
content
around
like
where,
like
how
do
you
do
that,
like
how
do
you?
How
do
you
structure
that
as
a
project
and
like
how
do
you.
C
A
A
D
No
I
I
attended
your
the
coupon
platforms,
working
group,
Meetup
and
I
figure.
This
is
interesting,
so
I
wanted
to
just
join
and
see
what
we
do
and
maybe
contribute
so
yeah.
That's.
A
B
B
D
Had
my
meeting
going
on
so
I
could
talk
but
yeah.
No,
it's
exciting
stuff
man,
so
I
hope
to
read
up
on
the
charter
and
be
more
useful
by
next
time.
A
Robert,
when
you
fell
out,
you
were
talking
about
like
an
Exemplar
platform
which
makes
me
think
of
I.
Don't
know
if
you
saw
Mauricio
salatino's
keynote
at
kubecon,
but
that
was
like
an
example
of
a
way
to
you
know:
build
parts
of
a
platform
I
felt
like
I'm.
B
Very
much
behind
on
basically
everything
I
actually
been
today
watching
stuff
from
cubicle
I'm.
D
B
Trying
to
but
no
I
didn't
see
that
one,
the
Keynote.
B
Yeah
I
think,
like
potato,
is,
isn't
a
project
you
use
to
try
to
Showcase
some
stuff,
because
you
know
we
need
an
application
in
some
fashion.
D
B
Know
you
know,
there's
no
point
of
building.
You
know
platforms
without
having
something
to
run
on
it.
So
you
know
it's
good
that
we
have
that.
So
we
can
use
that
in
examples,
but
yeah
I
think
we
should
slowly
move
forward.
You
know
if
we
start
by
defining
what
a
platform
is,
or
you
know
it.
You
know
it
doesn't
have
to
be
like
a
definitive
get.
Ops
is
more
specific.
B
You
know
get
UPS
is
a
very
specific
thing,
there's
a
reason
why
a
lot
of
people
are,
you
know,
think
they're
doing
get
ups
and
they're
not
so
they
you
know.
Of
course
you
need
definitive
kind
of
rules
to
Define
that
term
operational
model,
while
a
platform
can
be
so
many
things,
but
some
you
know
not
rules,
but
you
know
some
definition.
I
think
would
be
good
for
the
community
to
have
something
to
look,
look
at
and
can
understand.
A
Yeah
yeah,
even
if
they
disagree
with
a
point
or
two
that's
kind
of
like
there's
some
things
in
the
GitHub
stuff
I,
don't
a
thousand
percent
agree
with,
but
yeah.
If
we
just
have
a
stake
and
that's
you
know
for
people
to
debate
like
if
we
include
I,
don't
know,
store,
artifacts
and
Source,
or
actually
development
environments
was
one
that
Abby
was
saying.
Oh,
that
should
be
a
top
level
capable
and
I
was
like.
Is
that
something
that's
in
a
platform
or
not?
You
know.
C
D
A
B
And
and
like
I
said,
we
could
just
start
at
one
end
like
because,
like
my
my
girlfriend
actually
asked
me
well,
I
was
talking
about
the
fact
that
I
was
going
to
talk
about
platform
engineering
at
devops
days.
You
know
slow
and
she
kind
of
asked.
So
you
know
what
it.
What
is
the
platform
and
I'm
gonna?
Well,
it's
this
magical
thing.
B
You
know,
so
you
know
having
that
as
a
definition.
First,
like
I
said
it
doesn't
have
to
be
like
a
very
explicit
one,
but
you
know
yeah
something
to
work
from
and
then
we
can
start
like
breaking
it
into
like
how
you
doing
environments
like
how
are
they
supposed
to
you
know
interconnect
when
it
comes
from
like
from
platform
standpoint,
all
the
way
down
to
how
do
a
developer
or
user
of
the
platform
interact
with
it.
You.
C
A
What's
really
struck
me
is
how
some
people
think
of
Backstage
as
a
plat
as
the
platform,
and
but
what
that
says
to
me
is
that,
when
you're
looking
at
the
platform,
that's
what
you
see
that's
what's.
First,
is
the
apis
and
the
portals
and
all
that
kind
of
stuff,
and
what
comes
behind
it?
That's
like
it's
I
think
it's
gonna
vary
like
some
companies
might
just
write
a
portal
over
I,
don't
know
Azure
services,
but
some
might
have
an
operator
that
Provisions
something
some
might
go
deeper.
That's.
B
And
you
know,
go
friends
is
like
golden
golden
pass,
a
good
example
that
might
be
as
simple
as
just
templating
certain
things
and
just
making
it
available
and
then
put
it
on
whatever
you're
doing
your
stuff
and
I
have
so
I
feel
like
I
wouldn't
platform,
but
you
know
it's
at
least
like
one
part
of
it.
It's
it's
making.
It
super
people
to
just
to
get
their
stuff
to
to
turn
on
something.
A
B
A
Curious
what
you
all
think,
I,
put
I,
put
docs
and
templates
next
to
portal
and
API
as
like
an
interface
to
the
platform
like
the
platform
needs
to
have
a
pipeline
Runner,
an
observability
system
and
the
resource
provisioning
system,
but
you're
gonna
have
a
template
built
on
top
of
that,
like
you're,
not
going
to
say,
hey,
go
in
and
write
your
Jenkins
or
techtown
pipeline
from
scratch.
You.
D
B
Yeah,
it's
it's
one
of
those
things
where,
when
you
know
for
me,
like
a
platform,
isn't
when
people
have
pipeline
yaml
files
that
they
copy
back
and
forth
into
different
pipeline
scenarios.
No,
you
know,
that's
not
a
that's,
not
a
platform
that
might
be
part
of
a
platform
if
it's
done
right,
but
it's
not
really.
So
you
know
having
that
kind
of
like
you
know
what
does
it
mean
to
build
a
platform
which
is.
C
B
C
C
D
D
B
B
D
All
right
here,
this
one's
gonna
be
easier.
No,
this
is
from
Whimsical.
This
is
basically
just
a
mind
map
of
how
I
look
at
I
just
keep.
This
is
me
keeping
an
idea
of
all
the
things
I've
sold
on
a
kubernetes
platform,
so
building
it
for
a
company
that
I
was
working
for
so
I'm,
not
sure.
Oh
thanks
man,
that's
what
I
kind
of
need
to
say,
I'm,
not
sure.
D
If
this
all
of
this
is
a
platform,
but
you
know
the
the
idea
of
the
Journey
of
like
everybody
starts
off
with
like
a
yaml
that
they
apply,
you
know,
and
then
they
go
to
kind.
And
then
you
know
they
go
to
production,
I
guess
but
like
then,
everybody
goes
from
that
to
customize
the
helm
to
all
riding
operators,
because
they
think
they
need
operators
like.
C
D
One
of
these,
so
so
yeah
I
mean
I
I
would
say
if
we
could
call
it
something
like
this
from
all
of
our
experiences
and
then
we
can
just
pick
into
like
this
was
platform,
but
this
was
something
I
specifically
solved
for
the
company
I'm
at
or
for
the
situation
I'm
in
or
something
that
would
be
one.
That
would
be
one
way
to
do.
It.
A
One
thing
I
was
thinking
so
Rafa
also
shared
a
diagram
yesterday
with
some
really
good
insights.
Maybe
we
we
chat
in
this
GitHub
issue,
you
know
put
those
diagrams
in
there
and
just
you
know,
go
back
and
forth
and,
and
you
know
we
could.
B
A
B
Tag
up
delivery
repository
is
not
very
active
and
I
think
it's
not
active,
because
people
don't
see
how
you
can
use
the
first
stuff,
like
this
I,
think
it's
a
good
idea
to
put
it
there.
It's
it's
it's
supposed
to
kind
of
be
where
those
type
of
interaction
you
know
gets
recorded.
B
If
you
want
to
call
it
that
I
also
think
I
like
the
I,
like
the
I,
like
the
idea-
and
you
know
it
would
be
great
to
have,
for
instance,
like
the
the
article
that
you
sent
me,
there,
there's
a
diagram
there
with
with
apps
and
then
like
a
platform,
and
it
says
apis
for
for
me,
like
the
self-service
part
of
this
is
like
key
for
a
platform.
That's
just
my
opinion.
B
I
think
a
lot
of
people
agree
with
that
because,
or
else
you're
kind
of
just
like
all,
are
working
with
the
same
tools
anyway,
for
me,
self-service
kind
of
elements
to
a
thing,
Ergo
platform,
but
but
I
think
all
these
kind
of
things
kind
of
describe
like
why.
Why
should
a
platform
have
apis
or
some
sort
of
fashion?
B
You
know
way
to
interact
with
it
and
and
what
order
is
it
necessary
for
it
to
become
an
actual
platform
which
I
would
say
stuff
like
observability
and
and
and
and
and-
and
you
know,
even
testing
again,
it
kind
of
depends
on
what
kind
of
platform
You're
Building.
You
know
what
purpose
platform
that
I'm
building
to
build
platforms
thanks
that
should
probably
platforms
the
platform.
D
B
A
A
Kidding
but
that's
kind
of
I
I
actually
have
been
thinking
that
that's
kind
of
who
we
and
the
you
know
the
10
people
that
came
to
the
Meetup
and
the
maintainers
we're
providing
a
platform,
or
at
least
building
blocks
for
platform.
Engineers
like
we
are
trying
to
reduce
the
cognitive
look.
Even
Abby
actually
said
that
in
one
of
her
comments,
if
I
had
to
reduce
the
cognitive
load
for
the
platform
Engineers
by
creating
Frameworks,
you
know
creating.
B
Patterns,
which
is
great
because
no
one
thinks
about
us,
because
for
me,
one
of
the
reasons
why
you
build
a
platform
like
this
is
to
reduce
the
cognitive
load
of
the
people
who
are
supposed
to
use
it,
which
is
the
developers
who
are
supposedly
apparently
supposed
to
know
everything
about
infrastructure,
security,
automation,
et
cetera,
Etc,
including
them
also,
then
actually
developing
something
you
know
so
so
yeah,
it's
like
we
can
provide
some
cognitive.
You
know
release
whatever
you
want.
C
To
call
it
well,
at
least
at
least,
create
some
order
in
the
current
cows
right
then
you
know
you
still
have
to
study
right
if
you're,
if
you're
a
platform
engineer,
there
is
no
way
around
it,
but
at
least.
B
B
Yeah
and
that
doesn't
help
and
I
think
that
it's
important
that
the
people
who
are
because
I
I
truly
believe
the
platform
engineering
is
the
kind
of
like
next
step
for
us
that
are
like
on
the
operational
side
of
things
we
we
need
to
do
better
on
that
side,
you
know
to
elevate
the
game,
so
I
think
I.
Think
platform,
engineering
as
like
a
domain
knowledge
is
is
is
is
is
the
way
to
go.
D
C
Yeah
I,
like
your
observation,
I
I,
agree
that
what
I
see
often
in
my
customers
is
infrastructure
teams
are
tasked
to
build
the
platform
and
they
they
don't
see.
The
word
as
the
developers
see
the
world
yeah
and
and
so
they're.
The
possible
Vision,
sometimes
they're
good,
sometimes
they're
bad,
but
their
vision
is
always
limited
and
it
never
really
go
and
then
get
to
the
point
of
solving
real
developer
problems.
C
My
mind
background
is
more
in
publication.
Development
actually
and
I
feel
that
some
at
least
sometimes
I've
been
able
to
bridge
the
gap.
A
little
bit
and
I
would
like
to
propose
that
when
we
choose
a
language
to
write
our
definition,
which
we
choose
the
developer
point
of
view
language,
not
the
infrastructure
side
of
the
world,
because
we
are
building
the
platform
for
them
right.
So,
for
example,
a
silly
example
that
I
was
discussing
internally
with
with
some
colleagues
right.
C
There
is
this
debate:
should
we
do
namespace
as
a
service,
or
should
we
do
cluster
as
a
service?
Is
the
wrong
language?
For
me,
this
is
irrelevant,
because
what
the
developer
wants
is
environment
as
a
service
does
it
do
they
care
if
they
run
on
a
namespace
or
a
cluster
by
itself,
they
don't
care.
If
the
platform
is
well
designed,
they
need,
they
don't
even
need
to
know.
Okay,.
D
B
No
not
to
interrupt
you,
but
I
I.
Think
I.
Think
that
again,
like
I'm
saying
like
there's
a
reason
why
I
think
like
platform
engineering
is
kind
of
like
the
right
like
approach
going
forward,
because,
like
one
of
the
things
that
that
I
always
talk
about
when
it
comes
to
platform
engineering,
is
that
we're?
B
You
know
we're
doing
operational
things,
but
we
have
a
customer
focus
and
in
most
cases
the
customer
is
a
developer
and
I've
talked
to
developers
about,
for
instance
like
using
kubernetes,
and
if
that
developer
is
just
doing
most
of
the
people.
That
I
talk
to
don't
understand
the
reason
why
there
needs
to
be
a
kubernetes
at
all,
because
they
can
just
go
other
places
and
put
their
code
in
and
it
works.
That's
you
know,
that's
that's
what
it
boils
down
to
like
the
experience
for
them
developing
right
and
yes,
I.
B
You
know
there
might
be
a
number
of
technical
reasons
why
you
wanted
to
do
it
in
namespaces
or
having
B
clusters
or
separate
clusters
or
abstract.
You
know
so
on
and
so
forth,
but
you
know
for
the
end
user
does
that
matter?
In
most
cases,
no
they
just
want
like.
B
This
is
how
we
get
our
environment,
and
this
is
where
we
put
our
code
and
now
it
works
the
same
that
I
talk
about
like
when,
when
it
comes
to
monitoring
like
you
say
you
build
it,
you
run
it,
but
then
they
need
to
figure
out
how
to
connect
all
the
dots
and
all
the
monitoring
and
how
to
get
all
the
metrics.
And
you
know
they
don't
want
to
do
that.
But
if
you
say
you
build,
you
run
it,
oh
by
the
way,
there's
the
dashboard
that
shows
you
the
status.
B
B
C
B
Like
the
idea
about
about
kind
of
writing
it
from
I
try
to
not
use
the
word
developer
because
I
I
have
had
several.
You
know,
projects
where
I
I'm,
not
necessarily
working
with
Developers
I,
might
be
working
with
people
or
just
using
you
know.
You
know,
platform
in
a
sense
and.
B
Like
I,
don't
like
to
call
it
inter
local
platform,
I
like
internal
equipment
platform,
rather
because
it's
or
it's
case
for
me,
but
but
at
least
like
write
it
from
like
in
any
perspective,
yeah
or.
B
A
B
But
yeah
it
can
be
written
in
such
a
sense
that
you're
like
explaining
like
this
is
the
angle.
Your
user
is
going
to
look
at
your
platform
now.
This
is
the
reason
why
if,
for
instance
like
if
you
work
with
people
that
you
know
it
might
be,
the
best
approach
for
them
might
be
to
have
an
API
and
then
create
a
CLI
tool
for
that
API
and
there
you
go
that
works
best
for
for
those
people
and
some
might
want
the
portal
some
might
want.
You
know
they're.
B
You
know
Etc
et
cetera,
there's
no
one
solution
that
solves
it
for
everyone,
but
you
need
to
kind
of
like
Get
It
from
like
a
and.
B
I
think
it's
important
to
get
that
across,
because
that
is
for
me
one
of
the
key
reasons
why
we
even
are
creating
platforms,
because
if
not
then
I
could
just
be
the
Ops
guy
and
then
Azure
is
the
platform.
Like
a
customer
said
to
me
when
I
ask:
where
is
the
platform
You're
Building?
What
was
that
sure?
You
know
that's
my
platform
of
a
where's,
the
platform
that
we're
building
yeah.
A
B
And
I
don't
think
there
is
well
there
isn't
and
I
don't
think
there
ever
will
be
a
a
one
solution
that
you
can
buy,
and
then
you
have
a
platform
because
you
know
we
already
have
a
platform
like
me
that
work
with
Azure
I
have
a
platform.
I
have
a
lot
of
tools
in
my
platform.
You
know
that's
great
for
me,
but
then
someone
else
have
a
different
perspective,
so
a
different
scope,
they're
supposed
to
create
like
these
Snippets
of
code
that
are
supposed
to
run
on
something
they
don't
really
care.
B
C
As
a
person,
you
know,
you've
wrote
the
Enterprise
Integration
pattern,
it's
it's
kind
of
a
personality
in
it,
but
here
recently
he
started
doing
platform
related
work
at
AWS.
He
has.
He
has
an
entire
talk.
Just
on
this
thing,
that
Robert
was
saying
we
could
invite
him
and
you
know
get
his
point
of
view.
He
is
obviously
working
for
AWS.
C
So
it's
an
interesting
party,
but
you
know
he
has
this
concept
of
thinking
platform
like
yeah,
you
build
a
platform
on
top
of
the
cloud,
but
then
the
cloud
adds
more
functionalities
and
they
start
eating
away
the
functionality
that
you
did.
What
do
you
do
to
you
know
to
keep
your
platform
afloat
and
prevent
it
from
becoming
useless,
as
all
of
this
kind
of
reasoning
is
really
interesting.
B
Yeah
and
I
don't
think
there
shouldn't
be
like
people
work
where
they
work
right
like,
and
everyone
has
some
sort
of
angle,
usually
in
some
fashion,
but
that
might
not.
That
doesn't
mean
it's
like
like
ill
content
or
something
like
that.
I
think
I
think
yeah,
that
that
sounds
good.
It
sounds
like
a
a
you.
B
A
C
A
C
A
A
C
Yeah
yeah,
so
my
idea
was
to
try
to
represent
processes
and
possible
technology.
C
You
know
areas
and
in
a
sort
of
also
in
a
sort
of
capability
model,
or
at
least
logical
progression,
how
you
would
build
the
platform
so
that
that's
how
I
started
this
diagram,
it
may
not
be
complete,
or
maybe
you
even
disagree,
but
just
to
just
to
show
some
ideas
so
really
the
foundational
capabilities
that
I've
seen
are
needed
are
in
in
IDP,
as
in
the
security
identity
provider
infrastructure
as
a
service
which,
obviously
by
itself
is
the
platform
like
is
another
platform
right.
It
could
be
yeah,
storage,
Network
compute.
C
You
know
that
that's
a
large
area
but
I
want
to
I
need
to
be
able
to
provision
infrastructure
right
and
then
there
is
cigarette
management.
I
I
finally
decided
that
it's
a
foundational
capability,
you
basically
without
that
being
you
know
so
I
put
it
at
the
same
level
as
IIs
and
IDP.
Without
that,
being
you
know,
defined
it's
really
difficult
to
build
the
rest
in
a
self-service.
You
know
API
driven,
maybe
GitHub,
driven
Bank
and
then
an
obserability
stack.
That
should
also
be
part
of
the
foundation.
I
think
you
know.
C
This
this
is
becoming
more
and
more
important
in.
In
the
you
know,
a
soft
protective
of
the
secure
software
supply
chain
space,
but
anyway
it's
a
dotted
line
because
I'm
not.
C
Yeah
that
could
be
a
way
to
do
it.
Yes,
okay,
but
yeah,
but
you
know
if
you're,
if
you're
in
kubernetes,
maybe
you
have
a
way
to
attest
what
the
part
is
right.
C
B
C
D
C
Okay,
so,
and
then
in
red
we
have
the
processes
right
and
I
have
identified
the
song.
Maybe
there's
something
missing,
but
I
need
a
way
to
onboard
right.
I
need
I
need.
It
needs
to
be
easy,
easy
way.
Almost
you
know
self-service
way,
let's
say
zero
ticket
experience
where
I
can
come
to
to
my
to
this
platform.
I
get
some
space
where
I
can
do
my
stuff
right,
I
I
get
the
right
credentials,
the
right
quotas
and
whatever
it
is.
That
makes
the
platform
multi-tenant.
C
We
don't
know,
but
that
needs
to
be
taken
care
of
in
that
step
right
and
then
I
need
a
way
to
provision
an
environment
or
multiple
environments,
depending
what
I
need
that
should
be
again
as
a
service.
C
So
let's
say
I
want
to
annotation
I
will
need
the
environments
for
the
sdlc
once
I
am
on
boarded
I
have
a
way
to
request
that
right
and
then
credentials.
I
I
am
deeply
convinced
that
if,
if
let's
take
a
namespace
I,
if
I
provision
a
name
special,
then
and
and
if
I
tell
you
what
I'm
going
to
use,
the
namespace
for
the
platform
should
just
populate
the
secrets
that
are
needed
in
the
credentials
that
are
needed.
C
C
I'll
show
you
what
I'm
I
can
show
you
technologically
what
how
I
do
it?
But
the
point
is:
if
I,
if
I
provision,
an
environment
and
I'm
telling
you
that
I'm
gonna
build
applications
in
this
environment,
then
I
probably
need
my
git
credential
and
my
you
know
container
registered
credential
at
the
very
least.
If
you,
if
you
make
me,
create
that
credential
and
maybe
I
have
a
way
to
get
the
credentials
from
somewhere.
C
But
if
you
make
me
create
that
credential
as
part
of
my
deployment
in
that
names
in
that
environment,
then
you
you're
gonna
broadcast
security.
It
needs
to
just
appear
magically
because
I
told
you
what
I
was
going
to
do
with
that.
If
I
tell
you
I'm
going
to
run
a
database
in
this
namespace
which
or
this
cluster
or
this
environment,
which
comes
later,
but
if
I
tell
you
I'm
going
to
run
another
base,
the
database
credentials
appear
magically.
B
Right
and
I
I
agree
because,
like
environment
as
a
service
sure
you
could
populate
that
environment
with
a
lot
of
things.
But
at
some
point
you
have
some
sort
of
connection
to
something
potentially
external.
You
know
I'm
not
saying
that
happens
all
the
time
again,
because
again,
no
absolutes,
but
and
and
for.
B
You
know
again,
my
perspective
is
kind
of
like
the
Azure
stack
of
thing,
and
then
you
know,
part
of
density
is
kind
of
the
way
to
get
credentials
into
a
pod
to
use
with
the
platform
under
on
the
light
platform,
but
at
some
point,
I
might
need
to
reach
out
to
someone
else's
Azure
environment
I
need
to
maybe
reach
out
to
Azure,
devops
or
GitHub.
I
might
need
to.
You
know
Etc
at
that
point
right.
B
Those
credentials
should
exist
in
a
in
a
fashion
that
makes
sense
like
from
an
automation
perspective,
but
also
like
just
to
make
it
a
a
way
that
makes
it
easy
again
like
it's
something
that
the
end
users
are
not
are
not
supposed
to
have
to
think
about.
They
shouldn't
need
to
just
now.
You
need
to
go
in
and
do
these
commands
to
get
your
secret
in
or
something
like
that
that
shouldn't
be
necessary
right.
C
I
agree,
okay,
but
the
next
level
for
me
is
to
build
the
developer
capabilities
again,
not
trying
to
make
it
easy
at
this
point
for
the
developers
just
have
the
capabilities,
because
you
know
if,
if
we
stop
here
as
a
developer,
I
could
still
you
know,
provision
my
environment
and
then
I
have
to
provision
my
build
tool
and
my
repo
tool
and
my
all
of
these
things.
Instead,
we
could
recognize
that
these
are
share
capabilities.
They
just
exist
and
people
are
will
be
on
board
and
some
some
hours
they
will
be
available
to
them.
C
So
I
have
some
here,
but
you
know
the
list
could
be
very
long.
You
know
depending
what
you
need
to
do
with
your
code
right
when
you
build
it
and
when
you
run
it,
but
here
then
you
have
component
onboarding
application
and
component
of
boarding
I'm
using
the
backstage
concept-
words
here,
but
you
know
so
it's
the
ability
to
say
hey
there
is
this:
I
have
a
component.
There
is
needs
to
run
in
this
environment
and
you
can
find
it
at
this
repo
right
and
then
from
there.
C
The
platform
should
be
able
to
understand.
Then
the
ability
to
record
and
do
inner
loop
I
can
I
can
explain
some
of
the
problems
that
I've
seen
here.
But
you
know
we,
if
we're
here
to
enable
developers
in
a
loop
should
be
in
scope.
C
In
my
opinion,
a
pipeline
and
build
the
code,
so
the
Outer
Loop
ability
to
the
or,
if
you
will
do,
is
just
more
the
CI
and
then
you
have
the
CD
I
was
really
tempted
to
say
you
need
to
do
githubs
everywhere
and
especially
in
your
city,
but
we
can
leave
it
for
for
later,
but
you
need
to
have
an
ability
to
do
it
right
and
then
ability
to
observe
and
be
alerted.
C
So
I
can
use
this
capability
and
start
building
my
own
Dasher
and
my
own
others,
and
then
the
next
level
is
okay.
Now
we're
we're
starting
to
make
it
easy
for
our
consumers
for
our
users
right.
So
we
have
started
to
have
golden
path
and
middleware
as
a
service.
Those
are
the
two
main
six
goes
running.
C
State
football
results,
so
we
want
to
take
that
cognitive
load
away
from
the
developers.
I
really
I
believe,
and
we
do
middleware
as
a
service
in
some
shape
or
form,
and
the
other
is
golden
path
where
we
capture
we
standardize
and
capture
all
the
best
practices,
and
we
give
you
a
way
to
use
it
to
use
them
right.
C
They
shouldn't
be
mandatory,
but
they
should
exist
so
that
if
I
happen
to
build
something
that
is
in
the
golden
plus
and
and
I,
don't
want
to
learn.
All
of
these.
You
know
level
of
complexity.
I
can
just
on
board
on
the
golden
pass
and
I
have
a
you
know,
an
immediate
passage
to
production,
and
then
maybe
there
are
other
Advanced
things
like
feature
flagging
and
Progressive
deployments.
C
This
I'm
not
too
sure
that
this
may
be
too
little
too
too
much
detail
for
this
slide,
but
maybe
some
people
discuss
those
and
finally,
only
at
this
point.
For
me
it
makes
sense
to
have
a
developer
portal
right.
You
wrap
everything
up
with
some
nice,
UI
and
and
so
I
agree
with
Josh.
Today,
it's
like,
if
you,
if
you
install
backstage
you,
have
some
all
of
your
problems
and
it's
not
really
true.
In
fact,
backstage
the
main.
B
B
A
D
B
Well,
as
I
think,
that's
that
that
diagram
kind
of
yeah
it
puts
a
lot
of
things
into
perspective,
I
think
it's
a
really
good
diagram.
I,
don't
I,
didn't
notice
anything
that
I
felt
was
missing,
which.
D
B
That
just
means
that
I
I,
don't
I,
can't
think
of
anything
else
to
put
in.
So
that's
probably
more
on
me
on
rather
than
you,
but
but
I
think
it
looked
good,
and
these
are
the
things
that
I've
I
often
refer
to
when,
when
I
talk
about
like
building
platforms,
I
think
I
think
these
are
at
least
like
key
elements
and
and
obviously
there's
there's
different
stages
of
maturity
or
faces
or
whatever
you
want
to
call
it
right.
Just
you.
B
You
need
to
start
somewhere
right,
but
and-
and
it's
not
like,
even
though
I
like
to
run
my
stuff
in
kubernetes
and
and
so
on
and
so
forth.
That
doesn't
mean
that
I'm
solving
everything
through
there,
you
know,
for
instance,
I,
don't
wanna
I,
don't
want
to
host
my
own
gits,
because
why
would
I?
There
are
tools
that
do
that
which
are
which
works.
Fine,
you
know
so
again.
I
think
I
think
that
that's
a
good
overview
for
all
the
types
of
things
that
you
might
need
to
put
into
something
like
a
platform.
B
I
think
it
would
be
a
really
good
follow-up.
After
this
cut
and
again
I
say
defining,
but
at
least
like
scoping
or
scopic,
maybe
scoping
is
Right
scoping
down
what
a
platform
is
in
the
context
of
the
the
work
that
we're
supposed
to
be
doing
because
I
know
other
contexts
might
be
something
different.
Yeah.
A
C
A
B
I
I
I
like
the
idea
of
it
being
totally
open,
but
I
don't
like
the
end
result
of
that,
which
is
usually
a
lot
of
chaos.
I
think
a
controlled
flow
is.
D
A
B
But
you
need,
you
need
guard
rails,
or
else
is
going
to
become
a
yeah
words
I,
don't
supposed
to
use
in
public
yeah.
No,
it's
good
and
I
think
I
think
also
the
the
the
the
the
drawing
that
you
showed
Rafael,
also
kind
of
like
kind
of
also
is
reflected
in
the
drawing
that
you
had
Josh
over
the
different
like
capabilities
and
stuff
like
that,
although
probably
to
more
details.
C
We
have
very
similar
point
of
views
like
like
I
was
saying
to
Justice
yesterday
I'm
trying
to
also
add
processes
instead
of
just
technological
capabilities
right,
because
I
I
even
think
that
processes
should
come
first
almost
and
you
need
to
think
a
bit.
You
need
to
model
them
in
in
in
a
way
that
makes
it
makes
sense
for
your
users
right,
which
is
going
to
be
always
slightly
different
in
every
context.
Yeah.
B
I
I
agree
it's
one
of
the
reasons
why
I
don't
like
add
the
word
like
I,
say:
I,
think
everyone
should
do
terraformed
I,
think
everyone
should
do
infrastructures
code.
B
B
You
know
the
different
tools
that
solve
different
problems,
but,
first
of
all
you
need
to
figure
out
what
what
are
your?
What
are
you
supposed
to
be
able
to
do
with
this
platform?
Yeah
and
and
then
kind
of
tie
the
tie,
everything
together
and
so.
B
A
But
yeah,
let's
continue
conversation,
yeah
and
GitHub
and
and
Slack
are
you
in
our
slack.
You
should
join
there
and
maybe
more
people
than
me
can
speak.
A
C
A
Right
guys,
so
thanks
so
much
thanks.
B
A
The
conversation
and
looking
forward
to
making
this
getting
this
out
getting
this
getting
getting
some
opinions
and
publishing
something
all.